PB82-244013
Diesel Emissions Symposium Proceedings
(U.S.) Health Effects Research Lab.
Research Triangle Park,,NC
Jul 82
                   U.S. DEPARTMENT OF COMMERCE
                 National Technical Information Service
                                NTIS

-------
                                              PB82-2U4013
     DIESEL EMISSIONS SYMPOSIUM

            PROCEEDINGS
         Project Officer

          James R. Smith
   Research Coordinations Office
Health Effects Research Laboratory
Research Triangle Park, NC  27711
OFFICE OF RESEARCH AND DEVELOPMENT
HEALTH EFFECTS RESEARCH LABORATORY
US ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NC  27711

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
  600/9-82-014
           3. RECIPIENT'S ACCESSION
                 mi
                                            SION NO-
                                            4 A o i
4. TITLE AND SUBTITLE
    Diesel Emissions  Symposium Proceedings
                             5. REPORT DATE
                                 July 1982
                                                           6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS

    Health  Effects Research Laboratory
    U.S. Environmental Protection Agency
    Research  Triangle Park, N.C. 27711
                             10. PROGRAM ELEMENT NO.
                                    9XA1C
                             11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
    U.S.  Environmental  Protection Agency
    Office  of Research  and Development
    Health  Effects Research Laboratory
    Research  Triangle Park. N.C. 27711
                RTP,  NC
           13. TYPE OF REPORT AND PERIOD COVERED
               Proceedings
                             14. SPONSORING AGENCY CODS

                                   EPA/600-11
15. SUPPLEMENTARY NOTES

    P.O. James  R.  Smith
16. ABSTRACT
          The high fuel efficiency of diesel  engines is expected to  result in a
    significant increase in the production  of diesel-powered passenger  cars.
    Major research programs were initiated  in the late 1970s by governments,
    industry,  and the academic community  in order to understand the physical and
    chemical  characteristics of emissions from the diesel engine, and the
    biological  effects of these emissions.   In October of 1981, the U.S.
    Environmental Protection Agency sponsored a Diesel Emissions Symposium to
    report and  review the major scientific  and technical  information developed
    from  these  research programs.

          This  proceedings volume contains 21 review papers and 79 short papers
    covering all  the oral and poster presentations of the 1981 Diesel Emissions
    Symposium.   The meeting spanned the following subject areas:  diesel
    emissions  characterization and control  technology; chemical and bioassay
    characterization; pulmonary function, toxicology, and biochemistry;
    mutagenesis;  carcinogenesis; exposure and risk assessment.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                b.lDENTIFIERS/OPEN ENDED TERMS  C. COSATI Field/Group
    diesel  emissions
    characterization
    carcinogenesis
    mobile  source emissions
    pulmonary function
    mutagenesis
bioassay
toxicology
vehicles
diesel emissions
18. DISTRIBUTION STATEMENT
                                              19. SECURITY CLASS (This Report/
                                                                         21 . NO. OF PAGES
                                              20. SECURITY CLASS (This page)
                                                                         22. PRICS
EPA Form 2220-1 (R«v. 4-77)   PREVIOUS EDITION is OBSOLETE

-------
                 NOTICE






THIS DOCUMENT HAS  BEEN REPRODUCED



FROM  THE BEST COPY  FURNISHED  US BY



THE SPONSORING AGENCY.  ALTHOUGH IT



IS RECOGNIZED  THAT CERTAIN PORTIONS



ARE ILLEGIBLE, IT  IS  BEING  RELEASED



IN THE INTEREST  OF MAKING  AVAILABLE



AS  MUCH INFORMATION AS POSSIBLE.
                       I

-------
                               DISCLAIMER


Papers included in this document authored by U.S.  Environmental  Protection
Agency researchers have been peer and administratively reviewed and approved
for publication.  Work described in papers authored by invited speakers out-
side the agency and not funded by the U.S. Environmental  Protection Agency
do not necessarily reflect the views of the agency and no official  endorse-
ment should be inferred.
                                      11

-------
                                   FOREWORD
     The Health Effects Research Laboratory conducts a coordinated
environmental health research program in inhalation toxicology, genetic
toxicology, neurotoxicity, developmental and experimental biology, and
clinical studies using human volunteer subjects.  These studies address
problems in air pollution, non-ionizing radiation, environmental
carcinogenesis, and the toxicology of pesticides and other chemical
pollutants.

     The high fuel efficiency of diesel engines is expected to result in a
significant  increase in the production of diesel-powered passenger cars.
Major research programs were initiated in the late 1970s by governments,
industry, and the academic community in order to understand the physical and
chemical characteristics of emissions from the diesel engine, and the
potential biological effects of these emissions.

     In December 1979, the U.S. Environmental Protection Agency Health Effects
Laboratory at Cincinnati, Ohio, sponsored the first symposium on the Health
Effects of Diesel Engine Emissions.  The 1981 Diesel Emissions Symposium,
sponsored by the U.S. Environmental Protection Agency Office of Research and
Development during October 1981 In Raleigh, North Carolina, fostered the
exchange of more recent scientific and technical information derived from the
various research programs.

     This proceedings volume contains 21 review papers and 79 short papers
covering all the oral and poster presentations of the 1981 Diesel Emissions
Symposium.  The meeting spanned the following subject areas:  diesel emissions
characterization and control technology; chemical and bioassay
characterization; pulmonary function, toxicology, and biochemistry;
mutagenesis; carcinogenesis; exposure and risk assessment.


                                        F. Gordon Hueter
                                        D i rector
                                        Health Effects Research Laboratory
                                        U.S. Environmental Protection Agency
                                        Research Triangle Park, North Carolina

-------
                                   ABSTRACT
     The high fuel efficiency of diesel  engines is expected to result  in a
significant increase In the production of diesel-powered passenger cars.
Major research programs were initiated in the late 1970s by governments,
industry, and the academic community in  order to understand the physical and
chemical characteristics of emissions from the diesel  engine,  and the
biological effects of these emissions.  In October of  1981, the U.S.
Environmental Protection Agency sponsored a Diesel Emissions Symposium to
report and review the major scientific and technical  information developed
from these research programs.

     This proceedings volume contains 21  review papers and 79  short papers
covering all the oral and poster presentations of  the  1981 Diesel  Emissions
Symposium.  The meeting spanned the following subject  areas:   diesel
emissions characterization and control technology; chemical and bioassay
charcterization; pulmonary function, toxicology, and biochemistry;
mutagenesis; carcinogenesis; exposure and risk assessment.
                                      iv

-------
                                 ACKNOWLEDGMENTS
     The assistance of the many individuals who contributed to the planning
and execution of the symposium and to the compilation of the proceedings  is
gratefully acknowledged.  Special  appreciation is due to the members  of the
symposium organizing committee:  James Smith, General Chairman; Joel!en
Lewtas, Organizing Chairman; Stephen Nesnow, and Larry Claxton, Health
Effects Research Laboratory; and Ronald Bradow, Environmental  Sciences
Research Laboratory.  Special appreciation is also due to Ms.  Olga Wierbicki
and Ms. Barbara El kins of Northrop Services, Inc., the symposium coordinators.

-------
                                   CONTENTS
Foreword  ........................... .....   i i i
Abstract  ............. . ..................    'v
Acknowledgments   ............  ................    v

1.  Diesel Emissions Characterization and  Control Technology  ......    1
    Diesel Emissions, a Worldwide Concern ................
        Karl J. Springer
        Southwest Research Institute

    Diesel Participate Emissions:  Composition,  Concentration,
    and Control ..................  ...........   14
        Ronald L. W! I Mams
        General Motors Research Laboratories

    Diesel Particle and Organic Emissions:   Engine Simulation,
    Sampling, and Artifacts .... ...................   32
        Ronald L. Bradow
        Environmental Protection Agency

    Particulate Emissions from Spark-ignition Engines .  .  .  .  ..... .   47
        Ted M. Naman and D.E. Seizinger
        U.S. Department of Energy
        Charles R. Clark
        Inhalation and Toxicology Research  Institute

    Particulate Emission Characterization Studies of In-Use
    Diesel Automobi les  .........................   52
        Richard Gibbs, James Hyde, and Robert Whit ley
        New York State Department of Environmental Conservation

    Diesel Exhaust Treatment Devices:   Effects on Gaseous  and
    Particulate Emissions and on Mutagenic  Activity ...........   55
        R.A. Gorse, Jr., J.J. Florek,  W. Young,  J.A. Brown,  Jr.,
        and  I. Salmeen
        Ford Motor Company

    Characterization and Oxidation of Diesel Particulate .........  53
        David A. Trayser and Louis J.  Hillenbrand
        Battel le-Columbus Laboratories
                                      vi

-------
    Heavy-Duty Diesel  Engine Emissions—Some Effects of
    Control  Technology  	    62
        J.M. Perez and R.V. Bower
        CaterpiIlar Tractor Company

2.  Chemical and  Bioassay Characterization   	    63

    Methodology of Fractionation and Partition  of Diesel
    Exhaust Particulate Samples 	    64
        Bruce A.  Petersen and Cheng Chen Chuang
        Battelle-Columbus Laboratories

    The Utility of Bacterial Mutagenesis Testing in  the
    Characterization of Mobile Source Emissions:   A  Review   	    81
        Larry D.  Claxton
        U.S. Environmental Protection Agency

    Emission Factors from Diesel and Gasoline Powered
    Vehicles:  Correlation with the Ames Test	,  95
        Roy B. Zweidinger
        U.S. Environmental Protection Agency

    Analysis of Volatile Polycyclic Aromatic Hydrocarbons
    in Heavy-Duty Diesel Exhaust Emissions   	   109
        Walter C. Eisenberg and Sydney M. Gordon
        I IT Research Institute
        Joseph M. Perez
        CaterpiIlar Tractor Company

    The Chemical  Characterization of Diesel  Particulate Matter   	   Ill
        James Alan Yergey and Terence H. Risby
        Johns Hopkins University
        Samuel S. Lestz
        Pennsylvania State University

    The Analysis of Nitrated Polynuclear Aromatic Hydrocarbons
    in Diesel Exhaust Particulates by Mass  Spectrometry/Mass
    Spectrometry Techniques 	   115
        T. Riley, T. Prater, and D. Schuetzle
        Ford Motor Company
        T.M. Harvey and D. Hunt
        University of Virginia

    Contribution of 1-Nitropyrene to Direct Acting Ames Assay
    Mutagenicities of Diesel Particulate Extracts 	   119
        Irving Salmeen, Anna Marie Durisin,  Thomas J. Prater,
        Timothy Riley, and Dennis Schuetzle
        Ford Motor Company
                                      VI I

-------
    Dinitropyrenes:   Their Probable Presence in Diesel  Particle
    Extracts and Consequent Effect on  Mutagen I c Activations  by
    NADPH-Dependent S9 Enzymes  ...........  .......... 1Z1
        T.C. Pederson and J-S. Siak
        General  Motors Research Laboratories
3.  Pulmonary Function
    Inhalation Toxicology of Diesel  Exhaust Particles .......... i 124
        Roger 0.  McClellan,  Antone L.  Brooks,  Richard G.  Cuddihy,
        Robert K. Jones,  Joe L.  Mauderly,  and  Ronald  K.  Wolff
        Lovelace  Biomedical  and  Environmental  Research Institute

    EPA Studies on the Toxicologica I  Effects of  Inhaled  Diesel
    Engines Emissions ..........  ..............  .  .
        Wi I I iam E. Pepelko
        U.S.  Environmental Protection  Agency

    Deposition and Clearance of  Diesel  Particles  from the Lung   ..... 168
        Jaroslav  J. Vosta I ,  Richard  M.  Schreck, Peter S.  Lee,
        Ta i  L. Chan, and  Sidney  C. Soderholm
        General Motors Research  Laboratories

    A Subchronic  Study of the Effects  of Exposure of  Three
    Species of Rodents to Diesel Exhaust   ................ ,185
        Harold L. Kaplan
        Southwest Research Institute
        Wil I iam F- MacKenzie
        University of Texas Medical  School
        Karl  J. Springer
        Southwest Research Institute
        Richard M. Schreck and Jaroslav J.  Vosta I
        General Motors Research  Laboratories

    Pulmonary Function Testing of Rats Chronically Exposed to
    Diluted Diesel Exhaust for 612 Days ................. 207
        K.B.  Gross
        General Motors Research  Laboratories

    Pulmonary Functional  Response in Cats  Following Two  Years
    of Diesel Exhaust Exposure  ..................... 209
        William J. Moorman and John  C.  Clark
        National   Institute for Occupational  Safety and Health
        William E. Pepelko and Joan  Mattox
        U.S.  Environmental Protection  Agency

    Deposition and Retention of  Surrogate  and  Actual  Diesel
    Particles  .............................. , 215.
        R.K.  Wolff, L.C.  Griff is, G.M. Kanapilly  and
        R.O.  McClel Ian
        Lovelace   Inhalation Toxicology Research  Institute

-------
    Lung Clearance of Radioactlvely Labelled  Inhaled Diesel
    Exhaust Particles 	   220
        P.S. Lee, T.L. Chan,  and W.E.  Hering
        General  Motors Research Laboratories

    CompartmentaI Analysis of Diesel Particle Kinetics  in the
    Respiratory  System of Exposed Animals  	   222
        S.C. Soderholm
        General  Motors Research Laboratories

    Response of  Pulmonary Cellular Defenses to the  Inhalation
    of High Concentrations of Diesel Exhaust   	   225
        Kenneth  A. Strom
        General  Motors Research Laboratories

    The Effect of Diesel Exhaust on Cells  of  the  Immune System   	 i  227
        D. Dziedzic
        General  Motors Research Laboratories

    The Participation of the Pulmonary Type  II  Cell Response
    to Inhalation of Diesel  Exhaust Emission:   Late Sequelae   	   229
        H.J. White and B.D.  Garg
        General  Motors Research Laboratories

4.  Pulmonary Toxicology and Biochemistry  	   231

    Response of  the Pulmonary Defense  System  to Diesel Particulate
    Exposure	  232
        Jaroslav J. Vostal,  Harold J.  White,  Kenneth A. Strom,
        June-Sang Siak, Ke-Chang Chen, and Daniel Dziedzic
        General  Motors Research Laboratories

    Investigation of Toxic and Carcinogenic Effects of Diesel
    Exhaust in Long-Term Inhalation Exposure  of Rodents  	   253
        U. Heinrlch, L. Peters, W. Funcke
        Fraunhofer-lnstitut fur Toxicologie und Aerosolforschung
        F- Pott
        Medizinisches Institut fur Umwelthygiene
        V. Mohr
        Medizinische Hochschule
        W. Stober
        Fraunhofer-lnstitut fur Toxicologie und Aerosolforschung

    Morphometric UItrastructura I  Analysis  of  Alveolar Lungs of
    Guinea Pigs  Chronically Exposed by Inhalation to Diesel
    Exhaust (DE)  	   271
        Marion I. Barnhart,  Steven 0.  Sal ley,  Shan-Te Chen, and
        Henry Puro
        Wayne State University
                                      i x

-------
    Biochemical  Alterations  in  Bronchopulmonary Lavage Fluid
    after IntratracheaI  Administration  of  Diesel  Particulates
    to Rats	  Z89
        C.D.  Eskelson, M.  Chvapil,  E. Barker,  J.A.  Owen
        University of  Arizona Health Sciences  Center
        J.J.  Vostal
        General  Motors Research Laboratories

    Lipid Changes in Lung  of Rats  after IntratracheaI                     I
    Administration of  Diesel Particulates	:
        C.D.  Eskelson, E.  Barker,  M. Chvapil,  J.A.  Owen
        University of  Arizona Health Sciences  Center
        J.J.  Vostal
        General  Motors Research Laboratories

    BioavailabiIity of Diesel Particle  Bound  [G-3H-1 Benzo(a)pyrene
    (3H-BP)  after IntratracheaI  Instillation   	j  295
        P.K.  Medda,  Sukla  Dutta, and Saradindu Dutta
        Wayne State University  School of Medicine

    The Potential for  Aromatic  Hydroxylase Induction in the  Lung          ;
    by Inhaled Diesel  Particles 	i  298
        K.C.  Chen and  J.J. Vostal
        General  Motors Research Laboratories

    Xenobiotic Metabolizing Enzyme Levels  in Mice Exposed to
    Diesel Exhaust or  Diesel Exhaust Extract   	  300
        WiI Iiam Bruce  Peirano
        U.S.  Environmental Protection Agency

5.  Mutagenesis		305
                                                                         i
                                                                         i
    Mutagenic Activity of  Diesel Emissions  	 i  306
        Joel I en  Lewtas
        U.S.  Environmental Protection Agency

    Genotoxicity of Diesel Exhaust Emissions  in Laboratory Animals   ...  328
        Michael  A. Pere-lra
        U.S.  Environmental Protection Agency

    Human Cell Mutagenicity of  Polycyclic  Aromatic  Hydrocarbon
    Components of Diesel Emissions  	  340
        Thomas R. Barfknecht
        Massachusetts  Institute of Technology
        Ronald A. Hites
         Indiana University
        Ercole L. Cavaliers
        University of  Nebraska  Medical  Center
        William G. Thilly
        Massachusetts  Institute of Technology

-------
    Cytotoxicfty, Mutagenictty,  and Comutagenicity in Diesel
    Exhaust Particle Extracts on Chinese Hamster Ovary Cells
    In Vitro	Jb°
        A.P. LI, R.E. Roger,  A.L. Brooks, and R.O. McClellan
        Lovelace Inhalation Toxicology Research Institute

    Mutagenic Activity of Diesel Particles in Alveolar
    Macrophages from Rats Exposed to Diesel  Engine Exhaust  	   363

6.  Carclnogenesis	365

    Skin Carcinogenesis Studies  of Emission  Extracts   	 ;  366
        S. Nesnow, C. Evans,  A.  Stead, and J. Creason                    '
        U.S. Environmental Protection Agency
        T.J. Slaga and L.L. Triplett                                     \
        Oak Ridge National Laboratory

    Dermal Carcinogenesis Bioassays of Diesel Particulates and
    Dichloromethane Extract of Diesel Particulates in C3H Mice  	   392
        Linval R. Depass, K.C. Chenn, and Lynn G.  Peterson
        General Motors Research  Laboratories

    Respiratory Carcinogenicity  of Diesel Fuel Emissions
    Interim Results 	   399
        Alan M. Shefner, Bobby R. Collins, Lawrence Dooley,
        Arsen Fiks, Jean L. Graf, and Mauriine M.  Preache
        I  IT Research Institute

    Carcinogenicity of Extracts  of Diesel and Related
    Environmental Emissions upon Lung Tumor  Induction in  Strain
    "A" Mice	421
        R.D. Laurie, W.E. Peirano, W. Crocker, F.  Truman,
        J.K. Mattox, and W.G. Pepelko
        U.S. Environmental Protection Agency

    The Influence of Inhaled  Diesel Engine Emissions  upon Lung
    Tumor Induction in Strain "A" Mice	425
        William E. Pepelko, John G. Orthoefer, W.  Bruce Peirano,
        Wai den Crocker, and Freda Truman
        U.S. Environmental Protection Agency

    Objectives and Experimental  Conditions of a VW/AudI Diesel
    Exhaust Inhalation Study   	   429
        U. Heinrich, F. Pott, and W. Stober
        Fraunhofer-lnstitut fur  Toxikologie  und Aerosolforschung
        H. K I  ingenberg
        Volkswagenwerk AG
                                      XI

-------
7.  Exposure and Risk Assessment  	 .....•••   435

    Potential Health Risks from Increased Use of Diesel Light
    Duty Vehicles	   436
        Richard G. Cuddlhy, Roger 0. McClellan, William C.
        Griffith, Fritz A. Seller, and Bobby R. Scott
        Inhalation Toxicology Research Institute
        Lovelace Blomedical and Environmental Research Institute

    Health Effects of Exposure to Diesel  Fumes and Dust in
    Two Trona Mines	   451
        M.O. Attf-leld and Aremlta Watson
        National  Institute of Occupational Safety and Health
        G.W. Weems
        Mine Safety and Health Administration

    Mutagenicity and Chemical Characteristics of Carbonaceous
    Partlculate Matter from Vehicles on the Road  	   453
        William R. Pierson, Robert A. Gorse, Jr., Ann Cuneo
        Szkarlat, Wanda W. Brachaczek, Steven M. Japar, and
        Frank S.-C. Lee
        Ford Motor Company
        Roy B. Zweidlnger and Larry D. Claxton
        U.S. Environmental Protection Agency

    Emissions of Gases and Partlculates from Diesel  Trucks
    on the Road	• •   457
        Raisaku Klyoura
        Research  Institute of Environmental Science

    Diesel Bus Terminal Study Effects of  Diesel Emissions on
    Air Pollutant Levels  •	   459
        Robert M. Burton, Robert Jungers, and Jack Suggs
        U.S. Environmental Protection Agency

    Diesel Bus Terminal Study:  Characterization of Volatile
    and Particle Bound Organlcs 	   466
        Robert H. Jungers and Joseph E. Baumgardner
        U.S. Environmental Protection Agency
        Charles M. Sparacino and Edo D. Pellizzarl
        Research Triangle Institute

    Diesel Bus Terminal Study:  Mutagenicity of the Particle-
    Bound Organlcs and Organic Fractions   	   469
        JoeNen Lewtas, Ann Austin, and Larry Claxton
        U.S. Environmental Protection Agency

    Nttro Derivatives of Polynuclear Aromatic Hydrocarbons  in
    Airborne and Source Particulate 	   472
        Thomas L. Gibson
        General Motors Research Laboratories
                                     xi

-------
    Risk Assessment of Diesel Emissions 	   476
        R. Albert
        New York University Medical Center
        T. Thorslund
        U.S. Environmental Protection Agency

8.  Poster Presentations  	  477

    Mutagenicity of Particle-Bound Organic Chemical  Fractions
    from Diesel and Comparative Emissions 	  .  478
        Ann Austin, Larry Claxton, and Joellen Lewtas
        U.S. Environmental Protection Agency

    Scanning Electron Microscopy of Terminal Airways of Guinea
    Pigs Chronically Inhaling Diesel Exhaust (DE)  	  482
        Marion I. Sarnhart> Fatma Mohamed, and Ahmet Kucukcelebi
        Wayne State University School of Medicine

    Emission of Diesel Particles and Parttculate Mutagens at
    Low Ambient Temperature	.1  484
        James N. Braddock
        U.S. Environmental Protection Agency

    The Design of the CCMC's Long-Term Inhalation  Program to
    Investigate the Possible ToxicologicaI Effects of Diesel
    and Gasoline Engine Exhaust Emissions 	  .  487
        J. Brightwell, R.D. Cowling, X. Fouillet,  R.K. Haroz,
        H. Pfeifer, and J.C. Shorrock
        Battelle

    Chronic Inhalation Oncogenicity Study of Diesel  Exhaust in
    Sencar Mice	   490
        K.I. Campbell, E.L. George,  I.S. Washington, Jr.,
        P.K. Roberson, and R.D.  Laurie
        U.S. Environmental Protection Agency

    Species Differences in Deposition and Clearance of Inhaled
    Diesel Exhaust Particles 	   492
        T.L. Chan and P.S. Lee
        General Motors Research Laboratories

    Species Comparisons of BronchoaIveolar Lavages from Guinea
    Pigj and Rats Exposed In Vivo to Diesel Exhaust (DE)	495
        Shan-te Chen, Mary Ann Weiler, and Marion  I. Barnhart
        Wayne State University School-of Medicine
                                      XI I I

-------
Chemical Characterization of Mutagenic Fractions of Diesel
Part I cut ate Extracts	
    Di I  ip R. Choudhury
    New York State Department of Health

Preliminary Report of Systemic Carcinogenic Studies on
Diesel and Gasoline Particulate Emission Extracts Applied
to Mouse Skin	!
    N.K. Ciapp, M.A. Henke, T.L. Shock, T.  Triplett, and
    T.J. Slaga
    Oak Ridge National Laboratory
    S. Nesnow
    U.S. Environmental Protection Agency

Influence of Driving Cycle and Car Type on  the Mutagenlcity
of Diesel Exhaust Particle Extracts  	 	  501
    C.R. Clark, A.L. Brooks, and R.O. McClellan
    Lovelace Inhalation Toxicology Research Institute
    T.M. Naman and D.E. Seizinger
    U.S. Department of Energy

CCMC's Health Effects Research Program 	  505
    Members of the Emissions Research Committee of the CCMC
Fractionation and Identification of Organic Components
in Diesel Exhaust Particulate  ....... ..... .  ......    509
    Mitchell D. Erickson, David L. Newton,  Michael  C.  Saylor,
    Kenneth B. Tomer, and E.D. Pellizzari
    Research Triangle Institute
    Roy B. Zweidinger and Sylvestre Tejada
    U.S. Environmental Protection Agency

Effect of Chronic Diesel Exposure of Pulmonary Protein
Synthesis in Rats . . ......... .............  .  .   513
    R.G. Farrer, Sukla Dutta, and S. Dutta
    Wayne State University School of Medicine

The Effect of Exposure to Diesel Exhaust on Pulmonary
Protein Synthesis ..........................   516
    C. Filipowitz, C. Navarro, and R. McCauley
    Wayne State University School of Medicine

The Rapid Analysis of Diesel Emissions Using the TAGA  6000
Triple Quadrupole Mass Spectrometer .................   517
    J.E. Fulford, T. Sakuma, and D.A. Lane
    SCIEX, Inc.
Preparation of Diesel Exhaust Particles and Extracts  as
Suspensions for Bioassay
    Jean L. Graf
    I IT Research Institute
                                  xi v

-------
Compounds in City Air Compete with 3H-2,3,7,8-
Tetrachlorodibenzo-p-Dioxin for Binding to the Receptor 	   520
    J.-A. Gustafsson, R. Toftgard, J. CarIstedt-Duke,  and
    G. lofroth
    Karolinska Institute and University of Stockholm

GC/MS and MS/MS Studies of Direct-Acting Mutagens in
Diesel Emissions  	   523
    T.R. Henderson, J.D. Sun, R.E. Royer, and C.R.  Clark
    Lovelace Inhalation Toxicology Research Institute
    T.M. Harvey and D.F. Hunt
    University of Virginia
    J.E. Fulford, A.M. Lovett, and W.R. Davidson
    Sciex, Inc.

Research Plans for Diesel Health Effects Study  	  ,528
    Hironari Kachi and Tadao Suzuki
    Japan Automobile Research Institute, Inc.                         :

Neurodepressant Effects of Uncombusted Diesel Fuel   	   531
    Robert J. Kainz
    Environmental Industrial Safety Consultants
    LuAnn E. White
    Tulane University School of Public Health and
    Tropical Medicine

Evaluation of the Release of Mutagens and 1-Nitropyrene from
Diesel Particles  in the Presence of Lung Macrophage Cells in
Culture	   535
    Leon C. King, Silvestre B. Tejada, and Joel I en  Lewtas
    U.S. Environmental Protection Agency

Bacterial Mutagenicity of a Diesel Exhaust Extract  and Two
Associated Nitroarene Compounds after Metabolism and Protein
Binding	   538
    Mike Kohan and Larry Claxton
    U.S. Environmental Protection Agency

Characterization of Particulate Emissions from In-Use
Gasoline Fueled Motor Vehicles  	   541
    John M. Lang, Roy A. Carlson, and Linda Snow
    Northrop Services, Inc.
    Frank M. Black, Roy Zweidinger, and Silvestre Tejada
    U.S. Environmental Protection Agency

Surface Reactivity of Diesel Particle Aerosols  	   546
    Magnus Lenner, Oliver Lindqvist, and Evert Ljungstrom
    University of Gothenburg and Chalmers University of
    Technology
    Inger Lundgren and Ake Rosen
    Volvo Car Corporation
                                  xv

-------
Effects of Ozone and Nitrogen Dioxide Present During
Sampling of Genuine Part IcuI ate Matter as Detected by
Two Biological Test Systems and Analysis of Polycyclic
Aromatic Hydrocarbons	« •  •  550
    G. Lofroth
    University of Stockholm
    R. Toftgard, J. CarIstedt-Duke, and J.-A. Gustafsson
    Karolinska Institute
    E. Brorstrom, P. Grennfelt, and A. Lindskog
    Swedish Water and Air Pollution Research Laboratory

Alumina Coated Metal Wool  as  a Part IcuIate Filter for
Diesel Powered Vehicles	  553
    M.A. McMahon, W.T. Tierney, K.S. Virk, and C.H. Faist

Isolation and Identification  of Mutagentc Nitroarenes in
Diesel-Exhaust Particulates ... 	 .....  556
    X.B. Xu, Joseph P. Nachtman, Z.L. Jin, E.T. Wei,
    Stephen Rappaport, and A.L. Burlingame
    University of California

Comparison of Nitro-PNA Content and Mutagenicity of Diesel
Emissions	559
    Marcia G. Nishioka and Bruce A. Petersen                          .
    Battelle Columbus Laboratories
    Joel Ien Lewtas
    U.S. Environmental Protection Agency

1-Nitropyrene Emissions from  Five Production Model  Diesel
Vehicles and the Effect of Damping Valve on the Emission	 563
    Nissan Motor Company,  Ltd.

Analysis of the Factors Affecting Unusually High BaP  Emission
from a Nissan SD-22 Diesel Engine Vehicle Observed  at EPA  ......  568
    Nissan Motor Company,  Ltd.

Capillary Column GC/MS Characterization  of Diesel Exhaust
Particulate Extracts	  .  534
    T.J. Prater, T. Riley, and D. Schuetzle
    Scientific Research Laboratory

Respiratory Health Effects of Exposure to Diesel  Exhaust
Emissions	5gg
    R.B. Reger
    National Institute for Occupational  Safety and  Health
                                  xv i

-------
Physico-Chemical Properties of Diesel  Part IcuI ate Matter	   589
    Mark M. Ross and Terence H. Risby
    Johns Hopkins University School  of Hygiene and
    Pub Iic Health
    Samuel S. Lestz and Ronald E.  Yasbtn
    Pennsylvania State University

Some Factors Affecting the Quantitation of Ames Assays   	   591
    Irving Salmeen and Anna Marie  Durlsin
    Ford Motor Company
                                                                      [
Chemical  and Mutagenic Characteristics of Diesel  Exhaust
Particles from Different Diesel Fuels  	   593
    D.S.  Sklarew, R.A. Pelroy, and S.P. Downey
    Battelle Pacific Northwest Laboratories
    R.H.  Jungers and J. Lewtas
    U.S.  Environmental Protection  Agency

Fractionation and Characterization of  the Organics from
Diesel and Comparative Emissions  	   598
    C. Sparacino, R. Williams, and K.  Brady
    Research Triangle Institute
    R. Jungers
    U.S.  Environmental Protection  Agency

SWRI-SFRE Diesel Health Effects Exposure Facility 	   603
    Karl  J. Springer
    Southwest Research Institute

Post-Exposure Diesel Particle Residence in the Lungs  of Rats
Following Inhalation of Dilute Diesel  Exhaust for 6 Months  	   605
    K.A.  Strom and B.D. Garg
    General Motors Research Laboratories

Trapping Gaseous Hydrocarbons 	   608
    Fred Stump
    U.S.  Environmental Protection  Agency

Analytical Methods for Nitroaromatic Compounds  	   611
    Sylvestre B. Tejada
    U.S.  Environmental Protection  Agency

Total Luminescence Spectroscopy of Diesel Exhaust Particulate  ....   614
    Gregory Wotzak
    Cleveland State University
    Robert Whitby
    New York State Department of Environmental  Conservation
                                  XVI I

-------
    Evaluation of the Metabolic Requirements of Diesel  and
    Comparative Source Samples in the Salmonella typnimurium Plate
    Incorporation Assay .... 	  616
        Katherine Williams and Joellen Lewtas
        U.S. Environmental Protection Agency                             ;

    MS/MS Characterization of Diesel  Particulates .  .	.619
        Karl V. Wood,  James D. Ciupek, R.  Graham Cooks,  and
        Col in F. Ferguson
        Purdue University

9.  Perspectives	622

    Perspectives on  Diesel Emissions  Health  Research   	  623
        Norton NeI son
        New  York University Medical Center
                                     xv i i i

-------
SECTION 1



DIESEL EMISSIONS CHARACTERIZATION AND CONTROL TECHNOLOGY

-------
DIESEL EMISSIONS, A WORLDWIDE CONCERN

KARL J. SPRINGER
Department of Emissions Research, Southwest Research Institute, 6220 Culebra
Road, San Antonio, Texas, USA
   Recent visits to Japan and Europe plus scores of visitors from other coun-
tries have convinced me that there is a worldwide concern over the possible
health effects of diesel exhaust.  Not all of these visitors come to San
Antonio to visit the Alamo or stroll by the river.
   Laboratory tests with bacteria, animal cells and tissues have shown some
components of diesel exhaust to be toxic, mutagenic or carcinogenic.  In addi-
tion to gas phase compounds that have both direct and secondary effects in the
atmosphere, diesel exhaust contains particulate matter of both solid (soot) and
liquid (aerosol) type.  The soot particles are less than one millionth of a
meter in size and provide a surface for the aerosols to condense or absorb.
For example, benzo(a)pyrene, a well known carcinogen, is but one of the mate-
rials that are in diesel particulate.  Nitropyrenes are currently a popular
group of compounds for study.
   Studies in 1977 by Southwest Research Institute's Emissions Research Depart-
ment proved that diesel passenger cars produce particulate on the order of 50
                                         1               2
times their gasoline-fueled counterparts.   A 1981  report  gave 0.31 g/km (0.5
g/mile) as an emission rate from cars.  On the average, about 15 percent of
the particulate weight is soluble organics (i.e., extractable with dichloro-
methane solvent).
   • It is the soluble fraction of the particulate  that has sounded the
     alarm si-nce this contains the materials that have been found to be
     direct acting mutagens by the Ames bioassay test.
   • It is this fraction, first collected by SwRI3  and then evaluated in
     the Ames test in 1977, that resulted in the precautionary notice
     published by EPA that same year.
   • It is this fraction which has caused this symposium to be held and
     the previous CRC Dearborn meeting in March 1981, the EPA Cincinnati
     Symposium in December 1979, and the EPA Ann Arbor Symposium in
     May 1978.
   • It is this fraction which has caused the legislators, Federal
     officials in the DOE, DOT, EPA, Bureau of Mines and others so much
     frustration and confusion.

-------
   • It is this fraction which has given executives of diesel engine
     manufacturers and car makers chronic nightmares.
   • It is this fraction which has given the voices for environmental
     protection another item for argument.
   • It is also this fraction which has given many of us a scientific
     challenge.
   I loosely gather the governmental policymakers, the corporate executives,
and all those for whom you work as "THEY Who Must Be Obeyed."  THEY have sup-
ported us with money, facilities and the opportunity to investigate diesel par-
ticulate.  Until recently, THEY have been, on the surface at least, patient,
realizing that such research is tedious and long term in nature.  Underneath
that thin veneer of patience I sense the growing imperative to decide, to rule,
and to move on.  It is that basic impatience of the U.S. or us.
   Can we decide at this conference if there is or is not a problem with that
organic fraction?  If there is, do we have enough data to prove it?  If we
don't, what is needed in time and money?  Can we convince ourselves and others
that the studies should continue?  In the meantime, what about dieselization
of the passenger cars and trucks in the D.S.?  Is it business as usual?  Do
we suggest diesels be limited in urban and congested areas?  Or do we give
diesels the green light?
   THEY, who must be obeyed, are faced with these and related questions.  What
to do?  Do not think that because your research is incomplete or that you need
another five years .to complete a health survey of a specific population that
THEY will necessarily wait.  Diesels in cars and trucks are a quick way to
reduce fuel consumption, energy costs and foreign dependence.  It is clear
from recent changes in environmental thinking that a new policy is emerging.
   I am reminded of the analogy with the city traffic engineer who, as traffic
at an intersection increases, decides to install a stop sign on the side street
to assure orderly traffic flow and prevent accidents.  Contrast that with the
                                                       •
same traffic department, for whatever reasons, who must wait for three acci-
dents within a year at that same intersection before making a traffic survey.
Has environmental policy toward the diesel changed?  To read the July 30, 1981
Wall Street Journal article.4 it may be inferred that it has.  The headline on
the article by Andy Pasztor reads
       "Studies That Find Diesel Fumes Benign Encourage the Easing
       of Engine Controls."
Incidentally, the medical definition of benign is "of a mild character."  News
of this article reached me by way of some visitors who began by saying, "Well,

-------
 I guess research on diesel exhaust particulate is over since the National Re-
 search Council says diesel exhaust is OK." Not having any response other than,
 "I don't think anyone as yet really knows that answer," I obtained a copy of the
 article thinking that if it were true, there would be little need for this
 speech, much less this symposium.
   To illustrate my point, I want to read the first three paragraphs.
     "WASHINGTON—Increasing numbers of prominent scientists appear
     ready to grant the diesel engine a clean bill of health.  And
     that's bound to add fuel to the Reagan Administration's effort
     to roll back diesel pollution rules.
       After years of controversy, many public health experts say
     they are becoming convinced the particles in diesel exhaust
     don't cause cancer or chronic respiratory difficulties as they
     once feared.  Several scientists, in fact, are urging the admin-
     istration to loosen a variety of pollution rules for diesel-
     powered cars and trucks, which are considerably more fuel-
     efficient than conventional models.
       The latest sign of this trend is a report on the effect and
     future of diesel technology, just completed for the government
     by a blue-ribbon study group formed by the National Research
     Council.  In the report, expected to be released in the next
     few weeks, more than a dozen prominent engineers, medical
     scholars and qther experts conclude that air pollution from
     anticipated wide-spread use of diesel engines won't pose a
     major health hazard or environmental problem."
As of September 30, 1981, the National Research Council report had not yet
been released.
   From these recent pronouncements, it appears we have to prove to THEY, who
must be obeyed, that there is a clear and present danger from diesel particu-
late.  I wonder if we are able to do this in the next few years, much less in
the next few days.  In any event, this is the major challenge facing those of
us in the health effects business.  If we can not prove it, then, like the
traffic department, we may have to wait for sufficient statistics before pre-
scribing a cure.  The case may be argued in many ways, and without facts, it is
hard to say who is right.
   Regardless of the arguments, THEY, who must be obeyed, have an approach which
goes essentially like this.  A broad program has been in progress for, sav,
three to four years with no proof that diesel exhaust is hazardous beyond the

-------
positive Ames test  (which we knew  of in  1977).  What  are  the  chances  of proving
a hazard exists,  given another  three to  four  years?
    It is the same question we ask  of ourselves  on many  topics every day.   I
submit that the delegates to this  conference  have a job of  convincing THEY,
who must be obeyed, of the need to continue.  Otherwise,  THEY, who must be
obeyed, quite likely will become impatient and  make those decisions for us
using whatever facts are available at the time.  You  may  not  agree, but the
decision point is very near.
    Let us now consider the importance of the  diesel and its particulate contri-
bution.  Diesels  are the workhorse of modern  society.   Practically all goods
are transported by  diesel locomotives, and diesel trucks.   Practically all
construction, farming and tunneling is by diesel tractors of  one type or other.
Practically all ships of the world are diesel-powered.  Practically all emer-
gency or standby  power units are diesel.  Yet their particulate emissions  are
a tiny fraction,  less than 5 percent, of the  particulate  emissions from all
sources.   The on-highway diesel constitutes  about one-forth  of this  5 percent.
Of  that, trucks and buses account  for most of the particulates.  To illustrate,
Table 1 shows the annual particulate emission rate per  vehicle for four dif-
ferent type vehicles.  In one year, a. diesel  bus or truck generates about  17
times the particulate of a diesel  car.
TABLE 1
ANNUAL PARTICULATE EMISSIONS FOR TYPICAL VEHICLES
Based on Three Year Old Vehicles5
Vehicle Type
Gasoline Car
(Unleaded Fuel)
Diesel Car
Diesel Bus
(2-Stroke Cycle)'
Diesel Truck
Emission Rate
g/mile
0.014
0.50
1.77
1.61
Miles/
Year
14,000
14,000
69,000
69,000
Pounds/
Year
0.43
15.4
273
248
   Figure 1, from data in a recent APCA paper7 by Ingalls and Bradow, projects
several rates of dieselization of passenger cars-.  The 25 percent best estimate
agrees with GM!s prediction that by the late 1980's, nearly 25 percent of its
new car fleet will be powered by diesels.
   Figure 2 shows the expected sales of medium- and heavy-duty trucks.  Note
that 50 percent of medium-duty vehicles of 8500 to 10,000 Ibs are expected to
be diesel and by 1996, essentially all new trucks and buses over 10,000 Ibs

-------
CO
z
o
co
HI
CO

Q
 16

 14


 12

 10

  8


  6

  4


  2
                             TOTAL GAS & DIESEL
                  50% EST.
                                       10% EST.
      1975
           1980
1985
1990
1995
        FIGURE 1. PROJECTED SALES PENETRATION
            OF  DIESEL POWERED LD VEHICLES
2000
CO
O
z
O
X
CO
ID
CO

X
o
D
DC
   400-
 .  300 -
200 -
   100-
      1975
            1980
 1985
 1990
                                       1995
         2000
FIGURE 2. PROJECTED SALES OF MEDIUM AND HEAVY DUTY
              DIESEL TRUCKS AND BUSES

-------
will be diesel-powered.
   Figure 3 places the current interest in diesel passenger cars in perspective.
Given a 25 percent sales penetration  (by 1985  and thereafter) of cars powered
by diesel engines, it will be 1998 before diesel cars equal diesel truck and
bus particulates.  With moderate particulate control of both classes of ve-
hicles, particulate parity is not reached until after the year 2000.  Diesel
car particulates equal gasoline car particulates in 1986.  With regulation,
diesel cars reach the gasoline car total particulate level in 1989.  Control
assumes the 0.6 g/mile and 0.2 g/mile light-duty particulate standards take
           /
effect as proposed in 1982 and 1986.  Control also assumes that new HD diesels
must meet about a 30 percent reduction in particulate starting in 1985.
   Figures 4 and 5, like Figure 3, project what may happen given a 10 percent
and 50 percent penetration of diesels without and with some control of parti-
culates.  In the case of ten percent sales of diesels, which is achieved in
1982, the diesel car is only about one-third that of the truck and bus by the
year 2000.  With some regulation, the light-duty diesel car contribution is
less than the gasoline car and is only one-sixth that of heavy-duty by the
year 2000.
   Assuming a very rapid rate of dieselization of 50 percent,  Figure 5 shows
that by 1991, light-duty diesel equals heavy-duty diesel and with regulation,
1996 is when diesel cars emit the same tons per year as heavy-duty vehicles.
This rate of dieselization is greater than the most optimistic projections,
but indicates what may occur if diesel cars become 50 percent of new car
sales by 1995.
   What do these graphs portray?  First, diesel trucks and buses are the major
producer of diesel particulate and will continue for many years, even with the
best estimate of 25 percent car penetration by diesel engines.   Second,  com-
parisons to gasoline particulate are interesting from the total tonnage  stand-
point but tell us nothing of the relative hazards of each.   Third,  we still
have a very viable alternative to diesel cars albeit a bit less energy effi-
cient whereas we are hooked on the diesel in heavy-duty applications.   There
simply is no alternative.  Fourth, more and more diesels are going  to replace
more and more gasoline engines in all types of highway vehicles,  a  trend that
is gaining more momentum.  Fifth, ought we not consider other  circumstances
than the total tonnage figures or the big picture?
                                 Q
   In a recently published report  to EPA, the emphasis was on how  to model
small or microscale situations in which self-contamination might occur.   Self-
contamination can occur in parking garages, street canyons, tunnels,  express-

-------
    700

co   600
i
o
 *  500
w
o  40°

y  300
 g
 H
 cc
 <
 0.
    200
100
    YEAR 1980
                 1990
2000  1980
                                               1990
                                                     2000
         FIGURE 3. PROJECTED PARTICIPATE POLLUTION
                   (25% LDD PENETRATION}

-------
   700
   600
CO
I
o
x 500
 CO
    400
 O
 t-
 ul  300

 ^  200
 g
 cc  100
 0_
      0
     YEAR  1980
1990
2000  1980
1990
2000
         FIGURE 4. PROJECTED PARTICULATE POLLUTION
                     (10% LDD PENETRATION)

-------
o
                  n
                  >
                  o

                   x
                  Z
                  O
                  tu
                  l-
                  <
                  _i
                  D
                  g

                  H
                  CC
                  <
                  Q.
800



700



600



500



400



300



200



100



  0
HD DIESEL


LD DIESEL


LD GAS


HD GAS
                       YEAR 1980
                                       1990
                           2000  1980
                                             2000
                                FIGURE 5. PROJECTED  PARTICIPATE POLLUTION
                                          (50% LDD PENETRATION)

-------
ways or almost anywhere crowded conditions occur with limited dilution.  The
current maximum 24 hour limit for total suspended particulate (TSP) is 260
yg/m .  For example, in a typical above ground multilevel parking garage, the
TSP can be exceeded by having as few as 12 percent of the cars diesel-powered.
These localized concentrations may be important, if not from health, from
public welfare, such as odor and eye irritation and reduced visibility, in-
creased soiling, etc.
    Most tunnels, underground parking garages and all underground mines have
forced ventilation to dilute and remove exhaust fumes.  This is an installation
and operating cost of some significance and depends on the type and amount of
exhaust generated.  I know of a tunnel in Switzerland which must close occa-
sionally due to reduced visibility from diesel traffic.  I would assume that
during peak conditions, one would be advised to take a deep breath on entering.
   Earlier, I challenged this conference to decide whether a health problem
exists with diesel exhaust or not.  Now, I want to direct a few, more pointed
remarks to the audience and especially THEY, who must be obeyed.  I know you
are here or will learn of this from the publication of this book.  Your task
is not to merely endure, although the veterans of the last such symposium in
Cincinnati surely deserve a commendation for valor and courage above and
beyond the call of duty.  Your challenge is to assimilate or in plain English,
to take up, absorb, incorporate,  digest and compare.  Then, you are to rumi-
nate.  In other words, turn it over in your mind, reflect on and think about
it.  Remember to assimilate and ruminate.   When you are being shown a rat lung
all black and sooty for the 92nd time, or the one-hundred and forty-sixth
macrophage slide, don't forget to assimilate and ruminate.
   My next challenge is to the speakers.  Please note that the official lan-
guage of the session is English.   I have asked the symposium organizers to
raise a white flag (to indicate surrender)  when a speaker loses  control with
the very complex medical terms.  Plain English is needed to summari-e the
importance and if we can not do this ourselves, our years of hard work may be
of reduced value to "THEY, who must be obeyed."
   My next message is to the organizers.  You have done a superb job of bring-
ing us a program that has 56 prepared papers in a three-day period.   It is
quite ambitious and by all standards, the topics to be covered are certainly
adequate.  By way of introduction, the topics are:
     Diesel Emission Characterization and Control Technology
     Chemical and Bioassay Characterization
     Pulmonary Function
     Pulmonary Toxicology and Biochemistry
     Mutagenesis and Carcxnogenesis

-------
      Exposure  and Risk Assessment
    If this were not enough, we have poster sessions this evening and tomorrow
 evening.  I asked someone what a poster session is, and was told that "It  is
 an  adult version of a high school science fair."  as an ex judge of  several
 senior division science fairs, I only hope our displays are as good.
    Buried within this mountain of macrophages and mutagens are two most  inter-
 esting developments.  They are both relegated to Poster Session  No.  2 on
 Tuesday evening.  Two posters describe the CCMC (Committee of Common  Market
 Automobile Constructors) Long-term Inhalation Health Effects program  being
 conducted in Geneva, Switzerland.  The other development is the planned  Diesel
 Health Effects Study by tha JARI (Japan Automotive Research Institute) for
 the Japanese automakers and government.
    These are two major long-term programs that may well carry on much of the
 research that has been performed in the U.S.  As the diesel health effects
 program in the U.S. is de-emphasized, and priorities revised, as appears to
 be  happening, we may have to look to Europe and Japan for answers to  the
 questions I posed earlier.  Let us hope that these programs have sufficient
 pesos,  patience and perseverance to finish the task.
    There is little doubt that the worldwide worry over diesels has given
 health  effects researchers a golden opportunity to study the lung in  ways
 heretofore never tried.   The pulmonary system has never been better understood
 and that in itself is a positive result of this work,  although few of us.
 emphasize its significance.
    My final challenge is quite simply to
                                 Assimilate
                                 Ruminate
                       and then
                                 Communicate.
We must absorb, digest,  compare and then think about it and get the meaning
into simple  terms  so that THEY,  who must be obeyed, will understand.   Only
then,  will we have done  our job well*
                              12

-------
REFERENCES

1.  Springer, K. J. and Baines, T. II., "Emissions from Diesel Versions of
    Production Passenger Cars."  SAE Paper 770818 presented at the Passenger
    Car Meeting Detroit Plaza, Detroit, September 26-30, 1977.
2.  Hare, C. T. and Black, F. M., "Motor Vehicle Particulate Emission Factors."
    APCA Paper 81-56.5 presented at the 74th Annual Meeting of the Air Pollu-
    tion Control Association, Philadelphia, Pennsylvania, June 21-26, 1981.
3.  Hare, C. T., Springer, K. J. and Bradow, R. L., "Fuel and Additive Effects
    on Diesel Particulate Development and Demonstration of Methodology."
    SAE Paper 760130 presented at Automotive Engineering Congress and Exposi-
    tion, Detroit, Michigan, February 23-27, 1976.
4.  Wall Street Journal, "Studies that Find Diesel Fumes Benign Encourage
    the Easing.of Engine Controls," July 30, 1981.
5.  National Air Quality and Emissions Trends Report, 1976.  EPA Report
    EPA-450/1-77-Q02, December 1977.
6.  Mobile Source Emission Factors.  Final Document, EPA-400/9-78-005,
    Environmental Protection Agency, March 1978.
7.  Ingalls, M. N. and Bradow, R. L., "Particulate Trends with Increasing
    Dieselization 1977 to 2000."  APCA Paper 81-56.2 presented at the 74th
    Annual Meeting of Air Pollution Control Association, Philadelphia,
    Pennsylvania, June 21-26, 1981.
8.  Ingalls, M. N. and Garbe, R. J., "Estimating Mobile Source Pollutants in
    Microscale Exposure Situations."  Final Report prepared for Environmental
    Protection Agency, EPA-460/3-81-021, July 1981.
                               13

-------
 DIESEL PARTICULATE EMISSIONS:   COMPOSITION,  CONCENTRATION, AND CONTROL
 RONALD L.  WILLIAMS
 Environmental Science Department,  General Motors Research Laboratories
 Warren, Michigan,  U.S.A 48090-9055
 INTRODUCTION
    The history of diesel particulate emission studies is still being written.
 The time and attention of many scientists  and engineers are currently being
 directed at trying to understand the formation,  atmospheric impact,  and
 health significance of diesel particulate  emissions.   In the past five years,
 •considerable information has been assembled on the chemical composition, the
 atmospheric concentrations,  and the prospects for controlling the fuel and
 lubricant by-products of diesel combustion.   The investigators of the last
 five years have surely benefitted from older studies  which defined the problem
 and developed the vocabulary and the concepts for most of the current work.
 However, new techniques and  approaches to  study diesel emissions coupled with
 changes in diesel-engine technology guarantee new findings which must be re-
 ported, digested, and occasionally reviewed.
    This paper will attempt to review recent work on the composition of diesel
 particulate and compare it with particulate from other combustion sources.
 Likewise, the current and projected concentration of  diesel particulate in
 urban areas and in other situations will be considered relative to the concen-
 tration and composition of other airborne  particles.   Finally, the limited in-
 formation available on the effects of experimental control systems on particu-
 late composition will be discussed.

 COMPOSITION
 Particle Size
    The size distribution of  diesel particles has been studied by a variety of
 techniques.   The most straight-forward approach has used electron microscopy
•to  look at diesel particles  deposited on various collection surfaces. ~   The
 chain-like configuration of  diesel particles is often cited as a characteristic
 feature.  Variations in both size and substructure have been studied using
 different engines, fuels,  and operating conditions to obtain clues concerning
                                              4
 the details of particle formation mechanisms.   The electrical aerosol analyzer
 has been used for rapid accumulation of particle size information. '   Inertial
 impactors and dichotomous samplers have been used to  fractionate diesel
                                 14

-------
particles by aerodynamic size. ' '   However, the proliferation of particle  size
studies was truncated by the common observation that the mass median diameter of
diesel particles is a few tenths of a micrometer.  For inhalation studies, it is
important to note that the number median diameter is considerably smaller,
i.e., fewer than 1% of the particles are larger than 0.05 micrometer.  Conse-
quently, diesel particles are easily transported by airstreams and are readily
removed from air only by high-efficiency dust filters.  Likewise, the submicro-
meter diesel particles have low settling rates and their deposition velocities
                          Q
are difficult to quantify.   The impact of small carbonaceous particles on
visibility and atmospheric chemistry and the mechanisms which remove such mate-
                                                      9
rial from the atmosphere have been discussed recently.

Black carbon
   The analysis of diesel exhaust particulate has stimulated the development
of carbon analysis methods which are useful for studying other carbonaceous
particles as well.  The most common approach is solvent (Soxhlet) extraction
which separates the particulate into a soluble and an insoluble fraction.
Similar mass fractionations have been made by thermogravimetry in several laho-
                                                                           n
                                                                            13
ratories.       Cadle and others in our laboratories have developed a carbon
analyzer which thermally distinguishes organic carbon from elemental carbon."
Generally, the nonextractable carbon is likewise not volatile, which is consis-
tent with the low hydrogen content of the nonextractable material.    The degree
of crystallinity of this black carbon may be important in understanding the
                                               4
formation of these particles in diesel engines.   This material is chemically
unreactive and presumably not toxic.  Its accumulation in and clearance from
animal lungs will be discussed in this symposium.

Organic materials
   The molecular weight distribution and the carbon number distribution of the
organic (extractable) fraction have been reported.  '    This material resem-
bles slightly oxidized engine oil.  The carbon number distribution determined
by gas chromatography begins at about 15 carbon atoms, peaks near C., , and
tails to  C  , which is roughly the practical limit for gas chromatographic
         14              12
analysis.    Cadle et al.   have found that carbonaceous material continues  to
volatilize to 700 C.  This observation is consistent with gel permeation re-
sults which show molecular weights as high as 5000, about 400 carbon atoms.
However, the vast majority of the extractablas are lighter paraffinic hydro-
carbons which derive from the diesel fuel and engine oil.''   These compounds
                              15

-------
 are generally not expected to be toxic.  Because of their  low vapor pressures,
 the extractables readily condense or adsorb on small carbon particles.   This
 means  they are not likely to be involved in gas-phase photochemical reactions
 so  they do not contribute to smog chemistry.  Similar organic materials  are
 found  as  10 to 30% of the mass of ambient particulate collected  in  urban areas.
    Some subtractions of the extractables have been subjected  to  intense  study.
 Fractionation on silica and alumina columns was conducted  on  the first diesel
 extract for which Ames mutagenicity was reported.    Operational names for  the
 fractions were assigned to aid inter laboratory comparisons.   One of the  most
 widely used methods of fractionation today is the high-performance  liquid chro-
 matography (HPLC) separation using a Biosil A column.  The time  trace of the
 fluorescence of the eluent has been subdivided into regions which roughly
 correspond to classes of compounds with increasing polarity,  from aromatics to
 highly oxygenated compounds.  This method was examined by  several laboratories
 in  a recent interlaboratory comparison conducted by the Chemical Characteriza-
 tion Panel of the Coordinating Research Council's air pollution  research
 program.
     The  polynuclear aromatic hydrocarbon (PNA) fraction of the  extract  has
 probably  received the most attention."  Methods for the measurement  of PNA in
 diesel particulate have greatly improved since the early work of Falk and co-
        18 19                       i-
 workers.  '    In particular, benzo(aTpyrene (BaP) can now be measured in diesel
 extract at the picogram-per-milligram' level using high-performance  liquid chro-
 matography   or thin-layer chromatography with fluorescence detection.    BaP
 has been  determined in the particulate from all types of vehicles   and  from
 other  combustion sources.    BaP continues to be used as an indicator of com-
 bustion particulate and of potential carcinogenicity.  Efforts to understand
 the biological activity of BaP are unequalled and the data base  of  BaP measure-
 ments  in  ambient air is one of the most comprehensive among air  pollutants.
     Another subtraction of diesel extract which contains  the strongly muta-
 genic  nitro-PNA derivatives has received considerable attention  in  the past two
       24-26
 years.       Analysis methods for the nitro-PNA in carbon  black,  in combustion
 particulate, and in ambient air will be discussed by several  investigators  at
 this symposium.  Modifications of the Ames bioassay will be described which
 raveal the pathways by which nitro-PNA display mutagenicity even when the
                                                                     "^7
 nitro-PNA are present at extremely low concentrations in the  extract.''
     Attempts at complete analysis using powerful mass spectrometric methods
 have revealed the chemical complexity of the extract fractions.   The variety of
individual compounds detectable in diesel extract was most recently  shown in a
                              16

-------
                                 28
study by Schuetzle and coworkers.    This variety indicates that diesel combus-
tion processes can arrange atoms of C, H, 0, N, and S in almost every chemi-
cally allowed configuration, albeit at extremely low concentrations.
     The detailed chemical analysis of diesel emissions holds challenges for
several more generations of researchers.  However, the methods available today
provide the tools to evaluate sampling methods and sampling conditions, as well
as the potential formation of artifactual materials during the sampling process.
The work in this latter area has been described by Bradow in this symposium and
by others.

Distinguishing features
     Despite the application of the best analytical methods, no unique feature
of diesel particulate has been identified which clearly distinguishes it from
the particulate from gasoline-burning engines or, for that matter, from other
combustion sources.  For example. Figure 1 shows that the relative amounts of
organic and elemental carbon in the exhaust particulate of gasoline-powered
vehicles depends on the vehicle type and the operating conditions.  The ratio
of elemental carbon to total carbon in the particulate emitted from wood and
natural gas shows a similar range of values.  The per-mile emissions of elemen-
tal carbon from vehicles appears to depend more on the air-to-fuel ratio than
they do on the fuel or engine type.  Diesel particulate is also not distinguish-
able from particulate emitted from gasoline engines of the stratified-charge
    29                                                                 23
type   or from homogeneous-charge engines operated in a fuel-rich mode.
     The emission rate of BaP from a variety of passenger cars likewise shows
                            22
that diesels are not unique.    On a per-mile basis, noncatalyst, gasoline-
powered cars commonly emit more BaP than diesels.  Even the relative amounts of
the individual PNA from diesel engines are not distinguishable from the rela-
                                                     30
tive amounts of individual PNA from gasoline engines.
     Likewise, the presence of direct-acting mutagens in diesel extracts and
the association of this activity with nitro-PNA are not unique to diesel en-
gines.    A study will be reported later in this symposium which has found
measurable concentrations of nitropyrene in the exhaust particulate from non-
catalyst and catalyst vehicles.  And in the same study, the presence of nitro-
pyrene in ambient air samples suggests that nitro-PNA derivatives are formed in
many combustion processes and perhaps in atmospheric processes and, therefore,
have always been present in ambient air.  It is thus clear that diesel exhaust
particulate is very similar to particulate from other combustion sources.

-------
                             Furnace
            Normal
              Rich
                         Fireplace
          Hardwood
          Softwood
          Synthetic
CO
            Automobiles
Pre-Catalyst  Detroit
Pre-Catalyst  Denver
   Catalyst  Detroit
   Catalyst  Denver
     Diesel  Detroit
     Diesel  Denver
                                                           0.2
                                              0.4           0.6
                                                  Ce/Cr
0.8
                             Figure 1.  The ratio of elemental carbon  to  total carbon from selected sources.

-------
CONCENTRATION
     Particulate carbon is an atmospheric pollutant which has not been  studied
extensively, but it has been suggested that carbon can cause visibility reduc-
tion, perhaps promote chemical reactions, and possibly disturb the global heat
        9
balance.   In the preceding section, I tried to demonstrate that diesel partic-
ulate has no unique properties which distinguish it from combustion particulate
from other sources.  However, since diesel vehicles emit larger amounts of par-
ticulate carbon on a par-mile basis than gasoline vehicles, it is important to
project the increase in the ambient concentration of particulate carbon from
expanded use of diesel-powered cars and trucks.

Modeling approaches
     The most commonly used approach for predicting the concentration of diesel
particulate for a given time and place has been simple modeling based on carbon
monoxide and/or lead as surrogate exhaust components.    The data base  for CO
and lead is very broad because they have been measured at many air monitoring
stations and in several special studies around the country.  The simplicity of
this approach and its direct tie to measured concentrations of vehicle  emis-
sions makes it the most reliable predictor because it automatically takes dis-
persion and source distribution into account.
     The input information for predicting the concentration of diesel particu-
late at any location can simply be scaled to the surrogate emissions.   One must
define the per-mile emission races and the vehicle miles traveled by all vehi-
cle classes.  For the location of interest, the predicted diesel particulate
concentration can be calculated from the lead emission rates, the vehicle miles
traveled, and the measured lead concentration using currant or historical data.
The critical assumption for the surrogate model is that lead particles  and
diesel particles disperse and transport identically in the atmosphere.

Current and projected diesel particulate concentrations
     Estimates of the current input of diesel particulate to the nation's at-
mosphere range from 80t000 to 120,000 metric tons per year (MT/y)." '   '
More than 90% of the diesel particulate currently is from heavy-duty diesel
engines in trucks and buses.  Based on the urban versus rural miles traveled by
heavy-duty diesel vehicles, about 25% of the total national diesel particulate
                          34                      3
is emitted in urban areas.    Of the 20 to 30 x 10  MT/y of diesel particulate
emitted in urban areas, we estimate that about 75% of it is elemental carbon,
which occurs in submicrometer particles.
                              19

-------
     On a national basis, residential wood burning releases a quantity  of ele-
mental carbon roughly equal to that from dieaels.    Since residential  wood
burning is concentrated in heavily populated areas, elemental.carbon  from
diesel engines accounts for less than half the total elemental carbon currently
emitted in urban areas.  This estimate has been confirmed using a broad set of
sources of elemental carbon.  Wolff and coworkers   found that diesel emissions
accounted for 24% of the elemental carbon found in Denver's air during  Nov-Dec,
1978.  The average concentration of elemental carbon in Denver at that  time was
5.4  microarams  P6* cubic meter, which means that the diesel particulate  con-
centration in Denver was 1.7 micrograms  per cubic meter (24% times   5.4  micro-
grams  per cubic meter divided by 75% elemental carbon).  Similar concentra-
tions (in micrograms per cubic meter) of elemental carbon were found  in other
cities, such as 13.3 in New York, New York, 4.1 in Downey, California,  3.2  in
Pleasanton, California, 3.6 in Pomona, California, 1.7 in Abbeville,  Louisiana,
and 1.1 in Pierre, South Dakota.    Assuming diesels contribute 25% of  the  ele-
mental carbon in urban areas and that the elemental carbon content of diesel
particulate is 75%, the concentrations of diesel particulate in these cities
were New York, 4.4, Downey, 1.4, Pleasanton, 1.0, Pomona, 1.2, Abbeville, 0.6,
and Pierre, 0.4.  Strictly speaking, these estimates apply to specific  sampling
periods in several different years   during which diesel emissions didn't
change appreciably.
     Using the current data as baseline information, we can estimate  the  numer-
ical relationship between the total urban emissions of diesel particulate and
measured values in several cities.  Total urban emissions of 20 to 30 x 10
MT/y result in annual-average concentrations of 4 to 6 micrograms per cubic
meter in cities with the highest vehicle populations, such as New York  and  Los
Angeles.  Correspondingly, ambient concentrations are 1 to 2 micrograms per
cubic meter in most other urban areas.
     Estimates of diesel particulate concentrations for the future depend on
the total amount of diesel particulate emitted and the distribution of  diesel
vehicles geographically.  Particulate emissions from light-duty diesels result
in proportionately larger increases in urban particulate concentrations com-
pared with particulate from heavy-duty diesels which travel a smaller percenr-
                                        34
age of their total miles in urban areas.    On a national basis, passenger
cars would emit 80 to 120 x 10  MT/y  (which equals the total mass of  heavy-
duty particulate emissions in 1981) if the passenger car fleet contained  12 to
18 million diesel cars  (10 to 15%) with an average emission rate of 0.6 g/'mile.
About 60% of the total would be emitted in urban  areas and passenger  car
                               20

-------
diesels would contribute 10 to 15 micrograms per cubic meter of particulate  in
New York and Los Angeles, and 2 to 5 micrograms per cubic metar in many othar
urban areas.  Because only 10 to 15% of the cars would emit diesel particulate,
the fleet-average emissions would be 0.06 to 0.09 g/mile.
     These estimates are consistent with the values reported by Bradow and co-
workers   which were calculated for St. Louis, Missouri, using a more complex
dispersion model.  For a fleet-average emission rate of 0.195 g/mile, that
estimate was 8 to 10 micrograms per cubic meter in the center city with a local
maximum value of 13 micrograms per cubic meter.  Our estimates are likewise  not
inconsistent with projected particulate concentrations we have published previ-
ously, which assumed a stablized fleet of 25% light-duty diesels with an emis-
sion rate of 0.2 g/mile.    The equivalent fleet-average emission rate was
therefore 0.05 g/mile.  In that work, the direct application of the lead surro-
gate model gave estimates of 6 to 10 micrograms per cubic meter for worst-case
cities and 2 to 4 micrograms per cubic meter in other major urban areas.
     It is clear, then, that urban concentrations of diesel particulate in the
future will depend on the growth rate in miles traveled by light- and heavy-duty
vehicles in urban areas and on their emission rates.  Based on current measure-
ments of elemental carbon in urban locations, each 25 x 10  MT/y of urban emis-
sions nationally would produce annual-average values of 1 to 2 micrograms per
cubic meter of diasel particulate in major U.S. cities.

CONTROL
     The prospects for increased use of diesel engines has stimulated efforts
to develop new technologies for reducing their particulate emissions.  In this
section, recent results from two different control approaches will be examined
to determine the effect they have on the composition of the particulate as well
as on the total quantity.  We emphasize that the experimental particulate
control systems described here are selected from the broad range of possible
control systems simply because they are available for particulate characteriza-
tion studies at this time.  Because the health-affects studies being conducted
predate the availability of particulate control systems, we think it is
important to make a preliminary assessment of how potential control approaches
affect the composition of the particulate.

Experimental catalyzed particulate trap
     Catalytic devices in several configurations have been applied to diesel
exhaust systems to lower the emission rate of gaseous hydrocarbons from diesel
                              21

-------
engines.37'38  As early as 1979, reports appeared in the literature which showed
that catalyst temperatures sufficient to oxidize hydrocarbons were generally
sufficient to convert significant quantities of sulfur dioxide to sulfate.
This has been given as the reason for an increase in the particulate  emission
                                                    39
rate with the installation of a catalytic converter.   At the same time,  one of
the most common approaches to reducing the particulate emission rate  of diesel
engines is to use a trap to remove the particulate and to periodically burn  the
                                       41—43
material which accumulates in the trap.       In order to minimize the ignition
temperature of the trapped material, trap surfaces are sometimes coated with
noble-metal catalysts.
     The composition of the particulate from one system of this type  has  been
determined.  The particulate trap was a single underfloor trap of metal mesh
with a coating of alumina impregnated with precious metals.  The car  which was
equipped with a 5.7-L diesel engine had been driven about 1000 miles  and  the
trap had been regenerated about 10 times before we tested it.  Ignition of the
trapped particulate was initiated by throttling the engine manually.
     A brief program was run to determine the chemical composition of the par-
ticulate emitted during several driving modes and during trap regeneration.  The
mass emission rate and composition of the particulate emitted in the  EPA-
specified (FTP) driving cycle were determined.  Higher speed driving  cycles
were also, used to determine the sulfate emission rate at higher exhaust temper-
atures.  Steady-state tests of 64 km/hour were run to characterize the emis-
sions during storage and during regeneration of the trap.  This speed, i.e.,
64 km/hour, was a convenient driving condition for monitoring the regeneration.
     Particulate samples were collected from our large dilution tunnel on Dexi-
glas filters.    The filters used for sulfate determination were pretreated
with hydrochloric acid to pacify the basic sites normally found on fiberglass
filters.  The determination of sulfate, extractables, benzo(a)pyrene, and
organic and elemental carbon in diesel particulate have all been described
previously.  '  '  '    In each of the trap regeneration tests, a white cloud
was observed through the polyacrylate walls of the dilution tunnel.   At the
same time,  the white cloud, was visible at the outlet of the tunnel system on
the roof of the laboratory.  The white cloud was also observable during regen-
eration on the road.
     The results of the chemical characterization for various conditions, in-
cluding regeneration, are summarized in Table 1.  Because no tests were run  on
this car without the catalyzed trap, direct measurements of removal effi-
ciencies for the individual components of the particulate are not available.
                              22

-------
 TABLE 1

 COMPOSITION OF THE EMITTED PARTICULATE  (in mg/mile)
                     Total Par-  Extract-          Organic   Elemental
 Test Mode           ticulate    ablas      BaP    Carbon    Carbon     Sulfate
Federal Test
Procedure cycle
Highway Fuel
155
181
8.0
71
0.0003
<0. 00001
7.6
3.4
109
53.9
4.6
50
 Economy Test cycle
 Sulfate Emission
 Test cycle
 64-km/h-storage
 64-km/h regeneration
 229     76       <0.000008  5.1
  55      6.2      0.00017   4.2
5360   4000        0.0002   30
35.2
40
  63
   3.1
2530
    Calculated from cold-start and hot-start 18-cycle results; weighted 43%
    cold-start and 57% hot-start.

However, I will assume the composition of the particulate emissions from this
                                                 23 44
engine was typical of other 5.7-L diesel engines.  '    Normally, organic car-
bon is about 15% of the total carbon in the FTP cycle while, with the catalyzed
trap, organic carbon was only 6%.  In the higher speed cycles, organic carbon
is typically 25%, while here it was 6%.  The organic carbon emission rate was
consequently lowered from normal values of 50 to 70 rag/mile down to 3 to 8 mg/
mile.  In addition, the BaP reduction was about 90% in the FTP cycle and was
greater than 99% in the higher speed cycles.
     Correspondingly, the amount of conversion of sulfur dioxide to sulfata in-
creased from the FTP cycle to the higher speed cycles.  The sulfate emission
rates of 50 and 63 mg/mile in the latter cycles represent conversions of 5.7
and 7.8% of the fuel sulfur, whereas typical diesels emit 1 to 2% of the fuel
                  45
sulfur as sulfate.    These sulfate conversions are considerably lower than
those observed for similar precious-metal catalysts on gasoline-powered cars,
                                                          46,47
presumably because of the differences in the temperatures.
     Throttle-initiated regeneration of this experimental catalyzed trap gave a
oarticulate emission rate of 5360 + 1020 mg/mile.  In each case, a whits cloud
of emissions was observed.  The variability of the regeneration ^mission rate
presumably reflects the difficulty in igniting the stored particulate and repro-
ducing the combustion conditions during the regeneration which lasts from 5 to 3
minutes.  The total carbon was only about 1% of the total mass of particulata
emitted during regeneration.
                              23

-------
     The total particulate emissions, as shown in Table 1, are  not  simply the
sum of the components listed.  A better estimate of the composition of the
particulate can be obtained by adjusting the chemical components of the partic-
ulate to the mass they represent in the gravimetric determination of total par-
ticulate.  For example, the carbon analysis ignores the hydrogen, oxygen,  and
other noncarbon elements which actually contribute to the mass  of the two  car-
bonaceous fractions.    Mass adjustment factors have been determined using
several diesel engines which were not equipped with exhaust-trapping equipment.
These results showed that the organic carbon is only 70% of the mass of the or-
ganic material volatilized from the sample in the first step of the  carbon
analysis, and that elemental carbon is only 90% of the residual black carbon
which is oxidized in the second step.    Therefore, the emission rates  of  the
carbonaceous fractions must be increased appropriately to account for the  total
mass.  Likewise, the sulfate determined analytically is only 45% of  the mass of
the particulate sulfate actually weighed under the balance room conditions.
The remainder of the particulate sulfate mass is water, which associates with
the sulfate in the balance room as it also does in ambient air.  Using  these
mass-adjusted components. Figure 2 shows the composition of the exhaust par-
ticulate in five different test modes.  These three components  account  for 96
+ 4% of the total mass, which supports the use of the mass adjustment factors.
     The most dramatic change in the particulate composition occurs  during re-
generation.  At 64 km/hour, the emission rate of black carbon during regenera-
tion is identical with its emission rate during storage.  However, the  total
mass emission rate during regeneration is 100 times the emission rate during
storage.  During regeneration, the sulfate material accounts for more than 98%
of the total particulate emitted.  During storage, it appears that a small
amount of diesel particulate passes through the trap and carries with it the
hydrocarbons normally associated with diesel particles.  However, during re-
generation the temperatures are apparently sufficient to volatilize  some or-
ganic carbon and to selectively burn high-molecular-weight organic compounds
such as BaP.
     These results suggest that the biological and health effects studies on
typical diesel particulate may not be applicable to this diesel car.   In addi-
tion, the catalyzed conversion of sulfur dioxide to sulfate must be  avoided
because of the high sulfur content of diesel fuel.  Particularly, traps which
store sulfur and release it as sulfate during regeneration may  cause unaccept-
able sulfate concentrations in localized situations.
                              24

-------
                                                                  5360 mg/mi
         300 -
         200 -
Paniculate
Emissions
 (mg/mi)
         100 -
                             Residual

                       tS:S3 Paniculate Sulfate

                            | Organic Material

                             Black Carbon
                                      :•:•:•":•':
                   FTP
HFET         SET
        Test Mode
Storage    Regeneration
             Figure 2.  Composition  of  Exhaust Particulates from Diesel
                        Car Equipped with a Catalyzed Trap.
                             25

-------
Coppered-fuel with experimental fiber trap
     The second particulate control system had the following key  features:
 (a) the addition of copper to the diesel fuel, (b) the use of a tube-type trap
coated with small ceramic fibers which filter the particles out of the  exhaust,
and (c) the spontaneous auto-regeneration of this trap under normal driving
conditions.  Based on these features, the objective of our testing was  to de-
termine the carbon, sulfate, extractable, and BaP content of the  resulting par-
ticulate for several test modes as well as to measure the emission rate of par-
ticulate copper.
     The coppered fuel and experimental particulate trap had been used  on the
vehicle for about 1000 miles before we tested it.  Cold- and hot-start  urban
test cycles were used since these cycles produce a wide variety of exhaust
temperatures and exhaust flow rates and since these cycles are the most common
basis of comparison between diasel cars.  The highway fuel economy cycle was
used for higher but variable exhaust flow rates.  A series of steady-state tests
was run at 88 km/hour to allow for accurate monitoring of the inlet pressure of
the trap.  In the cyclic test modes, the inlet pressure differences which re-
sulted from changes in the exhaust flow rate obscured the inlet pressure differ-
ences caused by accumulation of particulate in the trap. 'At 88 km/hour, we were
able to observe a progressive increase in inlet pressure and its  rapid  return to
baseline values during the regeneration periods.  In this way, we were  able to
collect separate filter samples during storage and regeneration and to  determine
any composition differences.  The extent to which continuous combustion occurs
in the trap cannot be determined by our tests.
     Elemental carbon.  Elemental carbon normally dominates the mass of total
particulate emitted from a diesel engine.  The fuel-additive/trap system re-
duced the FTP emission rate of elemental carbon 99%, i.e., from 398 mg/mile to
2.0 mg/mile, as shown in Table 2, which demonstrates the high efficiency of
this trap system for removing solid particles from the exhaust.   The emission
rate of elemental carbon was less than 3 mg/mile in all the tests with  the
experimental trap.
     Particulate organic carbon.  Particulate organic carbon is normally emit-
ted at a relatively constant emission rate which appears to be independent of
driving mode and the emission rate of elemental carbon.    In the absence of
the particulate trap, this vehicle displayed normal emission behavior.   The
emission rate of organic carbon ranged only from 50 to 75 mg/mile while ele-
mental carbon varied 6-fold from FTP to 88 km/hour steady-state driving.  The
trap reduced the FTP emission rate of organic carbon 77%, i.e., from 67 mg/mi
                               26

-------
TABLE 2.

SUMMARY OF EMISSION RATES OF PARTICOLATE MATERIALS
Driving Mode
Material
                  Emissions  (ing/mile)
                                             Reduction
without trap
FTP






Highway Fuel
Economy





88 tan/h
a
store





88 tan/h
regenerate





Total Particulate
Elemental Carbon
Organic Carbon
Extractables
Benzo(a)pyrene
Sulfate
Copper
Total Particulate
Elemental Carbon
Organic Carbon
Extractables
Benzo (a)pyrene
Sulfate
Copper
Total Particulate
Elemental Carbon

Organic Carbon
Extractables
Benzo (a) pyrene
Sulfate
Copper
Total Particulate
Elemental Carbon
Organic Carbon
Extractables
Benzo (a) pyrene
Sulfate
Copper
649
398
67.4
103
0.00360
30
8.4
351
172
56.8
86
0.00200
17
5.6
217
75.1

47.6
68
0.0012
15
4.8
217
75.1
47.6
68
0.0012
15
4.8
with trap
36
2.0
15.2
28
0.00002
3.1
0.14
60
2,7
35.9
50
0.00006
2.1
0.16
27
0.7

13.7
-
-
1.5
0.18
42
0.5
19.2
-
-
4.2
0.25

94
99
77
73
99
90
98
83
98
37
42
97
88
97
88
99

71
-
-
90
96
81
99
60
-
-
72
95
  In the without-trap configuration the engine emissions for comparison with
  storage and regeneration are assumed identical.
                             27

-------
to 15 rag/mile.  At the same time, more than  99% of  the BaP was apparently
burned,  effectively eliminating BaP from the emissions of this diesel car.
Similar  trapping efficiencies were observed  for organic  carbon in the other test
modes.   The low emission rate of organic carbon during the trap regeneration
suggests that most of the stored organic material burns  up with the elemental
carbon.
     Sulfate.  In the no-trap configuration, sulfate emissions from this car
were similar to those of other diesels, ranging from 30  mg/mile in the FTP  to
15 mg/mile in the 88 km/hour cruise mode.  The sulfate was 4.6% of the total
particulate in the FTP, 4.8% in the highway  fuel-economy cycle,  and 6.9% in the
cruise mode.  However, with the trap in place, the  sulfate emissions were about
90% lower in all the test modes.  As a result, sulfate emissions were only  4
mg/mile  during regeneration.
     Particulate copper.  The rate of injection of  copper into the engine under
any test condition can be calculated from the copper content of the fuel and
the fuel consumption.  In the no-trap configuration, the emission rate of par-
ticulate copper ranged from 4.8 mg/mile at 88 km/hour to 9.5 mg/mile in the
cold-start portion of the FTP.  The copper emitted  accounts for 62 to 73% of
the copper in the fuel consumed.  Thus, it appears  that  this metal-additive
system is reasonably self-scavenged and that only a small part of the copper
consumed remains in the engine as deposits or accumulates in the engine oil.
The emission rate of total particulate from  this car in  the no-trap configura-
tion (0.65 g/mile) is typical of 5.7-L diesel cars  which indicates that copper
in the fuel does not appreciably affect the  engine-out emission rate of car-
bonaceous particulate.
     With the experimental trap in place, the copper emission  rate was reduced
by more  than 95%, as shown in Table 2.  Under trap-storage conditions,  the
copper emission rate was 0.14 mg/mile in the FTP, 0.16 in the  highway fuel
economy  cycle, and 0.18 at 88 km/hour cruise.  Under all these operating condi-
tions,  the trap was presumably accumulating  copper  at a  rate of 5 to 10 mg/mile.
During trap regeneration, the copper emission rate  was only 0.25 mg/mile.   Even
during oxidation of the carbonaceous particulate on the  trap  (regeneration),
most of the copper in the particulate was retained  by the trap.   The net result
was that only 0.8 to 3.2% of the copper consumed in the  fuel was emitted.   No
attempt was made to determine the chemical form of  the emitted copper under any
of the test conditions.
     Composition of carbonaceous emissions.  These  emission results show that
the efficiency of this control system for removing  elemental carbon from the
                              28

-------
exhaust is greater than 99%.  At the same time, the efficiency for removing

organic carbon is about 70%.  The emission rata of participate organic carbon

ranged from 15 to 35 rag/mile, but organic carbon accounts for 88 to 97% of the

total carbon as particulata.  It is important to note that the emission rate

of BaP from this fuel-additiva/trap system was also very low.  This particulate

control system markedly changed the composition of the particulate emissions

from the composition of typical diesel particulate.  However, the use of copper

in diesel fuel raises questions about compatibility with current engines and

fuel systems as well as questions about the potential environmental and bio-

logical effects of particulate emissions from such vehicles.  Its application

to real-world diesel vehicles is uncertain.


     Clearly, the history of diesel particulate emission studies is still

being written.


REFERENCES

 1.  Lipkea, W. H. , Johnson, J. H., and Vuk, C. T. (1979) in: The Measurement
     and Control of Diesel Particulata Emissions, Progress in Technology Series,
     PT-17, Society of Automotive Engineers.
 2.  Dolan, D. F., Kittleson, D. B. , and Pui, D. Y. H. (1980)  Diesel
     Exhaust Particle Size Distribution Measurement Techniques, Society
     of Automotive Engineers, Paper 800187.
 3.  Roessler, D. M., Faxvog, F. R., Stevenson, R., and Smith, G. W. (1981)
     in:  Particulate Carbon:  Formation During Combustion, Siegla, D.  C.
     and Smith, G. W., eds., Plenum Press, New York.
 4.  Siegla, D. C. and Smith, G. W., eds  (1981)  Particulate Carbon:  Formation
     During Combustion, Plenum Press, New York.
 5.  Groblicki, P. J. and Begeman, C. R.  (1979)  in:  The Measurement
     and Control of Diesel Partioulate Emissions, Progress in Technology
     Series, PT-17, Society of Automotive Engineers.
 6.  Kittleson, D. B. , Dolan, D. F., Diver, R. B., and Eberhard, A. (1978)
     Diesel Exhaust Particle Size Distribution, Society of Automotive
     Engineers, Paper 780787.
 7.  Chan, T. L. and Lawson, D. R. (1981)  Atmospheric Environment, 15, 1273.
 8.  Chang, T. Y. , Modzelewski, S. W., Norbeck, J. M., and Pierson, W.  R.  (1981)
     Atmospheric Environment, 15, 1011.
 9.  Wolff, G. T. and Klimisch, R. L., eds.  (1981)  Particulate Carbon:
     Atmospheric Life Cycle, Plenum Press, New York.
10.  Williams, R. L. and Begeman, C. R.  (1979) Characterization of Exhaust
     Particulate Matter from Diesel Automobiles, General Motors Research
     Laboratories, Warren, MI, Publication GMR-2970, May.
11.  Cuthbertson, R. D., Stinton, H. C., and Wheeler, R. W.  (1979)  The Use of
     a Thermogravimetric Analyzer for the Investigation of Particulates and
     Hydrocarbons in Diesel Engine Exhaust, Society of Automotive Engineers,
     Paper 790814.
12.  Cadle, S. H. and Groblicki, P. J.   (1981)  in: Particulate Carbon:
     Atmospheric Life Cycle, Wolff, G. T. and Klimisch, R. L., eds. Plenum
     Press, New York.
                              29

-------
13.  Cadle, S. H., Groblicki, P. J., and Stroup, D. P.  (1980)
     Anal. Chem., 52, 2201.
14.  Black, F. M. and High, L. E.  (1979)  in:  The Measurement  and  Control of
     Diesel Particulate Emissions, Progress in Technology  Series, PT-17,
     Society of Automotive Engineers.
15.  Mayer, W. J., Lechman, D. C., and Hilden, D. L.  (1980)  The Contribution
     of Engine Oil to Diesel Exhaust Particulate Emissions, Society of
     Automotive Engineers, Paper 800256.
16.  Huisingh, J. , Bradow, R., Jungers, R., Claxton, L., Zweidinger,  R.,
     Tejada, S., Bumgarner, J., Duffield, F., Water, M., Simmon, V.  F.,
     Hare, C., Rodriguez, C., and Snow, L.  (1978)  in:  Application of Short-
     Term Bioassays in the Fractionation and Analysis of Complex Environmental
     Mixtures, U.S. Environmental Protection Agency, 600/9-78-027,  September.
17.  Schuetzle, D. and Perez, J. (1981)  A CRC Cooperative Comparison of Extrac-
     tion and HPLC Techniques for Diesel Particulate Emissions, Air  Pollution
     Control Association, Paper 81-56.4, June.
18.  Falk, H. L. , Steiner, P. E., Breslow, A., and Hykes, R. (1951)   Cancer
     Res., 11, 318.
19.  Falk, H., Kotin, P., Thomas, M. (1955)   Archives Ind. Health,  10, 113.
20.  Swarin, S. J. and Williams, R. L.  (1980)   in:  Polynuclear Aromatic
     Hydrocarbons:  Chemistry and Biological Effects, Bjorseth, A. and
     Dennis, A. J., eds., Battalia Press, Columbus, OH.
21.  Swanson, D., Morris, C., Hedgecoke, R., Jungers, R., Thompson,  R., and
     Bumgarner, J. (1978)   Trends in Fluorescence, 1  (2), 22.
22.  Williams, R. L.  and Swarin, S. J.  (1979).  in:  The Measurement  and Control
     of Diesel Particulate Emissions, Progress in Technology Series,  PT-17,
     Society of Automotive Engineers.
23.  Muhlbaier, J. L. and Williams, R.  L. (1981)   in:  Particulate Carbon:
     Atmospheric Life Cycle, Wolff, G.  T. and Klimisch, R. L..  eds,  Plenum
     Press, New York.
24.  Gibson, T. L., Ricci, A. I., and Williams, R. L. (1981) in:  Chemical
     Analysis and Biological Fate:  Polynuclear Aromatic Hydrocarbons,
     Cooke, M. and Dennis, A. J., eds,  Battelle Press, Columbus, OH.
25.  Rosenkranz, H. 3., McCoy, E. C., Sanders, D. R., Butler, M., Kiriazides,
     D. K., Memelstein, R. (1980)  Science,  209,  1039.
26.  Schuetzle, D., Riley, T., Prater,  T. J.,  Harvey, T. M., and Hunt, D. F.
     (1982)  Anal. Chem., in press.
27.  Pederson, T. C.  and Siak, J. S. (1981)   J. Appl. Toxicol., 1, 54.
28.  Schuetzle, D., Lee, F.S.C., Prater, T.  J., and Tejada, S. B.  (1981)
     Intern. J. Environ. Anal. Cham., 9, 93.
29.  McKee, D. E., Ferris, F. C., and Goeboro, R. E.  (1978) Unregulated Emis-
     sions from a PROCO Engine-Powered Vehicle, Society of Automotive
     Engineers, Paper 780592.
30.  Kraft, J. and Lies, K. H. (1981) Polycyclic Aromatic Hydrocarbons in the
     Exhaust of Gasoline and Diesel Vehicles,  Society of Automotive  Engineers,
     Paper 810082.
31.  Williams, R. L.  and Chock, D. P.  (1980)  in:  Health Effects of  Diesel
     Engine Emissions, Pepelko, W. E.,  Danner, R. M., and Clarke, N.  A., eds.,
     EPA-600/9-80-057, November.
32.  Baines, T. M., Somers, J. H., and Harvey, C. A.  (1979)  J. Air  Pollut.
     Control Assoc.,  29, 616.
33.  Ingalls, M. N. and Bradow, R. L.  (1981)  Particulate Trends with Increasing
     Dieselization, 1977 to 2000, Air Pollution Control Association,  Paper
     81-56.2, June.
34.  MVMA Motor Vehicle Facts and Figures (1981)   Motor Vehicle Manufacturers
     Association, Detroit, Michigan.
                              30

-------
35.   Wolff, G.  T. ,  P.  J. Groblicki, Cadle, S. H., and Countess, R. J.  (1981)
     in:   Particulate Carbon:  Atmospheric Life Cycle, Wolff, G. T. and
     Klimisch,  R.  L.,  eds.,  Plenum Press, New York.
36.   Bradow, R. L.   (1980)  Bull. N. Y. Acad. Med., 56, 797.
37.   Amano, M., Sami,  H., Nakagawa, S.,  and Yoshizaki, H.  (1976)  Approaches to
     Low Emission Levels from Light-Duty Diesel Vehicles, Society of Automotive
     Engineers, Paper 760211.
38.   Seizinger, D.  E. , Eccleston, B. H. , and Hum, R. W.  (1979)  Particulates
     and Associated Emissions from Two Medium-Duty Diesel Engines, Society of
     Automotive Engineers, Paper 790420.
39.   Bassoli, C.,  Cornetti,  G. M.,  Biaggini, B.,  and DiLorezo, A.  (1979)
     Exhaust Emissions from a Europaan Light-Duty Turbocharged Diesel,
     Society of Automotive Engineers, Paper 790316.
40.   Hunter, G., Scholl, J., Flibber, F., Bagley, S., Leddy, D., Abata, D.,
     and Johnson,  J. (1981)   The Effect of an Oxidation Catalyst on the
     Physical,  Chemical, and Biological Character of Diesel Particulate Emis-
     sions, Society of Engineers, Paper 810263.
41.   Wade, W. R.,  White, J.  E. , and Florek, J. J.  (1981)  Diesel Particulate
     Trap Regeneration Techniques,  Society of Automotive Engineers, Paper
     810118.
42.   Murphy, M. J., Hillenbrand, L. J.,  Trayser,  D. A,, and Wasser, J. H.
     (1981) Assessment of Diesel Particulate Control — Direct and Catalytic
     Oxidation, Society of Automotive Engineers,  Paper 810112.
43.   Howitt, J. S.  and Montierth, M. R.   (1981)  Cellular Ceramic Diesel Particu-
     late Filter,  Society of Automotive Engineers, Paper 810114.
44.   Gabele, P. A., Black, F. M., King,  F. G., Zweidinger, R. B., and
     Brittain,  R.  A. (1981)   Exhaust Emission Patterns from Two Light-Duty
     Diesel Automobiles, Society of Automotive Engineers, Paper 810081.
45.   Cadle, S.  H.,  Nebel, G. J., and Williams, R. L.  (1979)  Measurements of
     Unregulated Emissions from General Motors' Light-Duty Vehicles, Society
     of Automotive Engineers, Paper 790694, June.
46.   Begeman, C. R., Jackson, M. W., and Nebel, G. J.  (1974)  Sulfate Emissions
     from Catalyst-Equipped Automobiles, Society of Automotive Engineers,
     Paper 741060,  October.
47.   Trayser, D. A., Creswick, F. A., Blosser, E. R., Pierson, W. R., and
     Bauer, R.  F.  (1976)  Effect of Catalyst Operating History on Sulfate Emis-
     sions, Society of Automotive Engineers, Paper 760036.
48.   Letkauf, G. ,  Yeates, D., Wales, K. , Spektor, D., 7-Oiert, R. , and
     Lippmann,  M.  (1980)  Amer. Ind. Hyg. Assoc.  J., 42, 273.
                              31

-------
DIESEL PARTICLE AND ORGANIC EMISSIONS:
ENGINE SIMULATION, SAMPLING, AND ARTIFACTS

RONALD L. BRADOW
Mobile Source Emissions Research Branch, Environmental Sciences Research
Laboratory, Environmental Protection Agency, Research Triangle Park, NC  27711
INTRODUCTION
     Research into the emission of diesel particles is being conducted by
government, industry, and academia.  While each group is examining specific
aspects, the underlying concern is the potential human health hazard of the
emitted particles.  This paper will discuss many of the mechanical details of
testing vehicles and measuring emissions.  In particular it will summarize the
current techniques for simulating road operation, discuss the various diesel
sampling systems for exhaust particles and organics, and consider artifactual
generation of mutating agents.  These details, while appearing at present to be
relatively unimportant in the determination of effects, are necessary to
develop methods of emissions regulation.

SIMULATION OF ROAD OPERATION
     Due to the scale of the test equipment involved, it is impractical to
measure pollutant emissions from a vehicle operated in traffic.  Therefore,
considerable effort has been expended in creating an engineering test system
capable of physically simulating the emissions of a vehicle operated in traffic
while physically stationary in a laboratory.
     The road-simulation system used for this purpose is called a chassis
dynamometer, a roll-test machine capable of producing resistive forces at the
drive wheels of a car or truck.  These forces can reasonably match those
experienced on the road.  Roll-test dynamometers with computer controls are
now available to match most of the important road force conditions experienced.
     Figure 1 depicts such a dynamometer installed in the U.S. Environmental
Protection Agency (EPA) laboratory located in the Research Triangle Park
(RTP), North Carolina.  Dynamometers with similar capabilities are available
at a few other institutions, notably General Motors  (GM), Ford, Volkswagen
(VW), and the New York State Department of Environmental Quality.  Basically,
these systems are capable of simulating the aerodynamic drag, inertial forces,
and tire rolling resistance of a vehicle operated on a level road.  The
simulation equations are as follows:
                             32

-------
                 BELT  DRIVE
                 ROLL
                 COUPLER
RTP    -    CELL     1
                                                                                                   DYNAMOMETER
                                                                                                   ROAD   SZMULATOR
                                                                                          SPEED   TRANSDUCER


                                                                                                   FLYWHEEL
CO
CO
                                           TORQUE   TRANSDUCER
                 FORCE  -
                                •«-  bV  -*-
                 POWER  RANGE•   - I SB  TO  160  hp
                 INERTIA  SIMULATION!
                       1600 TO  10,000  Ib.
                       IN  I   Ib.   INCREMENTS
                                                                                                 CONTROLLER
                                                 Fig.  1.  Layout of EI'A-RTP  dynamometer.

-------
                                                      2
     1.  Aerodynamic drag:                Forcea = CiV

                                                         dv
     2.  Inter!tal forces:                Forcej = (Mass)j^-

     3.  Tire rolling resistance:         Forcet = C2 - C.jV; where C2 » CjV

     4.  Total force:  Forcea + Force-j. + Force t = C1V  + C2 (Mass)^r

They are graphically represented in Figure 2.  The variations in resisting
force applied to the drive wheels of a passenger car can be seen as a func-
tion of steady-state forward vehicle speed caused by aerodynamic drag and
tire resistance.  In addition to these forces, inertial force equivalent to
the mass times acceleration rate is applied either by the use of flywheels,
i.e., physical mass, or electrically by sensing acceleration rate instantane-
ously and controlling applied torque from the electric motor.  Next, the
dynamometer is capable of controlling resistive forces to the drive train by
any simulation equation of the form:

                       Force = a + bV + CVd + Mass (||)

Thus, most conditions experienced by real vehicles operated on roadways can be
simulated.  There are some real limitations in the availability of road data,
particularly dealing with transient driving, to insure that the road simulation
is nearly perfect.  In the next few years, adequate road data will probably be
available to meet this need.
     In the meantime, we have several road driving cycles available, generated
principally by using Monte Carlo methods and screening procedures on large
speed-time data bases.  The available passenger car cycles are shown in Table 1.
Each of these is a speed-time route representing some facet of normal operation
in which emissions could be a problem.  Table I shows a few of the pertinent
features of these cycles, including average speed, distance, and number of
complete stops.  The routes currently used span a wide range of driving
conditions from cross-town driving in lower Manhattan to driving on a two-lane
country highway.

     In dealing with the forces applied, an automobile engine experiences a
variety of speed-load conditions.  Figure 3 shows what happens to a VW Rabbit
passenger car as it operates on the highway fuel economy test cycle (HWFET)
                              34

-------
20  HP
      III
      3
      Q
      1
      hi
      0)
      £
      0
      I
                      AERO  POWER  -  FORCE X  VELOCITY  -  C, Vs
                     DRAQ  POWER -  ex 1 0-*  V»
                                       SPEED
                                                                    30  MPH
           Fig.  2.  Graph of VW Rabbit aerodynamic drag and tire factors.




TABLE 1


STANDABD DRIVING CYCLES DSHD IN VEHICLE TESTING
Driving Cycle
New York City Cycle (NYCC)
Morning Commuter: Federal Test
Procedure (FTP) , Los Angeles
Cycle No. 4 (LA-4)
Crowded Urban Expressway (CUE)
Highway Fuel Economy Test Cycle
Average
(km/h)
11.4
31.5
56
77.5
Speed
(mph)
7.1
19.5
34.8
48.2
Distance
(km)
1.9
12
21.7
16.5
Number of
Stops
10
18
3
1
  (HWFET)
                              35

-------
test just described.   These traces were obtained with the real-time system
described by Gabele1  and constitute actual dynamometer test data.  The middle
speed trace (in mph)  shows that this cycle is basically a steady-state route
with an initial acceleration,  a final stop, and small speed variations
associated with vehicle control.   The upper trace shows engine intake air flow
(in SCFM) on the same scale; after this initial acceleration, there is
relatively little variation in this parameter.  This air rate is a rough
indication of how hard the engine is working.  The bottom trace shows one of
the important sampling parameters, dilution ratio on the same scale, for a
dilution air flow rate of 350 CFM.  For this condition, the idle dilution
ratio is about 15 to 1, while the steady-state is about 5 to 1.  Increasing
the blower rate to about 1000 CFM would increase the dilution ratio to about
45 at idle and 15 at steady-state.  To achieve a dilution ratio of near
atmospheric conditions with this car, say 500:1 at steady-state, one would
need a blower of about 10,000 CFM flow rate.  Work is in progress to install
such a system in cell #2 in the RTF facility.
I 06
                             TIME  CSECOMDSr-
Fig. 3. Speed traces showing VW Rabbit operation under the highway  fuel  economy
test cycle (HWFET).  The upper trace is engiae intake inflow  in.  standard cubic
feet per minute (CFM); the middle trace is speed in miles per hour  (mph);  and
the bottom trace is the dilution ratio.
                              36

-------
     Figure 4 shows a similar set of traces for the New York City Cycle  (NYCC).
Here the speed trace shows the typical large number of stops incurred driving
across lower Manhattan from about 10th Avenue to 1st Avenue.  Random slow-down
portions are made as if this car were caught in traffic and stopping for
traffic lights, spending about 4% of its time in idle.  This car spends a fair
amount of time accelerating to mid-block speeds, and rather high intake air
flows periodically occur.  The dilution ratios are again between about 15 to 1
and 5 to 1, but a greater portion of the time is spent in the high dilution
ratio condition.
180
                                                                           60S
                              TIME CSETCONDS3
Fig. 4. Speed traces of a VW Rabbit on the New York City Cycle  (NYCC).  The
upper trace is engine intake inflow in standard cubic feet per minute (CFM);
the middle trace is speed in miles per hour  (mph); and the bottom trace is the
dilution ratio.
                            37

-------
     The final figure in this series, Figure 5, shows what happens on the
Federal Test Procedure (FTP), a Los Angeles commuter route simulation.  Here,
all the curves are very spikey and a bit hard to follow.  Still the intake air
flow rates and dilution factors are within the ranges of the other cycles.
tea
                                                                     I90a
                           TIME CSECONDSD
Fig. 5. Speed traces of a VW Rabbit on the Federal Testing Procedure (FTP)
commuter route.  The upper trace is engine intake inflow in standard cubic
feet per minutes (CFM); the middle trace is speed in miles per hour (mph); and
the bottom trace is the dilution ratio.
     So, despite prominent differences among these cycles, the engine
experiences at least some high load and some idle operation in every one of
them, and only the percentage of the time spent in-each mode changes.
Therefore, cycle-to-cycle differences are expected only for those engine-
produced pollutants that experience a substantial difference in emission rate
as a function of driving mode or work required.  Ample evidence exists that
for carbon monoxide, e.g., from gasoline-fueled spark-ignition engines,
some dramatic effects of this type occur.  However, for current generation
diesel passenger car engines, there appear to be only minor variations in
particle emissions with change in driving cycle.
                            38

-------
     Figure 6 is a chart of emissions values for four driving cycles using the
same VW Rabbit.  Total hydrocarbons (THC), particle mass, and the Ames activity
in revertants/mile for TA98 -S9 show relatively little variation with driving
                                                            o
pattern, especially for particle mass and mutagen emissions.   Therefore, at
least with this car, driving patterns do not greatly influence the important
emission rates.  Gibbs et al.,  Hare and Baines,  and Naman et al.  have all
come to the same conclusion.  From a data base encompassing about 30 in-use
diesel passenger cars, it appears that driving cycles and load simulation are
relatively unimportant matters within reasonable limits.    Lang et al.  have
recently conducted a study of the particle emissions of 20 in-use gasoline
cars.  These authors have also concluded that mutagen emission rates are only
weakly dependent on driving cycle.
                                                           FTP
                                                        MMTT
           1.82
  ^
<-v
  a.aa
Fig. 6. Emissions data  for a VW Rabbit tested under various driving conditions.
                               39

-------
     Diesel and gasoline trucks can also be tested on a  chassis  dynamometer
using driving cycles similar to those used for passenger cars.   Dietzmann
et al.7'8 have reported results from four diesel and two gasoline  trucks
                                                                 Q
operated over the EPA transient driving cycle.  Dietzmann et al.   also split
the  samples by mode for this driving pattern to develop  particle emissions
factors for trucks.  Figure 7 shows particle emission rates for  the  six trucks
on a gin/km and gm/kg of fuel basis.  For comparison, data from 20  in-use
diesel passenger cars  and 20 in-use gasoline passenger  cars  are  included.
                          <>? Z L^

                            CO
                                                         a
                                                              §
                                                            3
                                                            CO
      Fig. 7. Bar graph showing particle emission rates from six trucks.

     Clearly, significant differences among vehicle types exist in gin/km and
gm/kg of fuel emission factors.  These differences will probably be reflected
in the relative influence of vehicle types on atmospheric concentrations of
particles.  The Dietzmann papers also indicate a substantial bias among driving
cycles for heavy duty trucks, both gasoline and diesel.  More than half the
particle mass is emitted in the Los Angeles freeway mode for both types of
trucks.  These differences in emissions factors between classes of trucks and
passenger cars are important in that they influence the need for and cost of
control.
                              40

-------
SAMPLING OF DIESEL EXHAUST PARTICLES AND ORGANICS
     A variety of sampling systems have been used for diesel exhaust particles.
Figure 8 shows a rather elaborate system currently in use at the Mobile Source
Emissions Research Branch at EPA-RTP.  Actually, two dynamometer cells are
similarly equipped in this facility with cell #1 devoted to gasoline spark-
ignition engine work and cell #2 devoted to diesel.
     It is difficult to say that any one of these is "typical," since a variety
of shapes and sizes of equipment are used.  The equipment shown in Figure 8 is
especially interesting, however, since this cell was used to generate the
large samples for biological testing from several cars.  In this case, room
air is drawn through a filter and charcoal bed, then mixed with exhaust to
produce a diluted stream in which the exhaust stream is cooled to near room
temperature.  Using typical gaseous emissions test conditions, e.g., a blower
speed of 350 CFM, dilution ratios of about 5 parts of air to 1 part of exhaust
are typical for a small car like the VW Rabbit, as shown previously.  For
larger cars or trucks greater quantities of dilution air are needed, and it is
rather common to use bigger constant volume samplers (CVS) blowers for sampling.
In this case, flow rates of up to about 1200 CFM are available with two parallel
positive displacement pumps; a still larger heavy-duty CVS with a capacity of
10,000 CFM is currently being installed in this facility.  Dietzmann et al.
have used multiple pumps to achieve flow rates as high as 12,000 CFM in testing
diesel tractors.
     It is interesting to note that very similar results are obtained from a
                                             o
number of different systems.  Tai Chan et al.  have recently reported a study
of raw exhaust electrostatic precipitation compound with air dilution.  These
authors find very similar results, both in terms of particle mass and in Ames
mutagenicity values for both methods.  The cooled raw exhaust collector samples
produced somewhat larger amounts of organics, mainly acidic-salt components,
but even chemical fractions had similar mutagenic activity with the exception
of acidic fractions.
     In the past, condensation and filtering of raw exhaust have been used to
acquire samples for analysis of polynuclear aromatics (PNA).   This technique
is still very common in Europe.    Recently, some studies have been reported
                                                                    11 12
involving Ames mutagenicity results from the University of Stockholm  '   and
comparative studies of polynuclear aromatic hydrocarbon (PAH) emissions from
           13
Volkswagen.    These works have shown similar levels of mutagenicity and PH
                                                           14
for gasoline and diesel vehicles.  A report by Stump et al.   strongly suggests
that condensation procedures produce results very similar to those obtained by
                              41

-------
                                                             SAMI'lt flOW
                                                             AMI (1C f 10*
                                                        	UIGIIAIFIOW
                                        HtCOIIUEH
Fig.  8.  Diagram of EPA-RTP diesel  sampling system.

-------
air-dilution techniques.  Therefore, it appears that different sampling
procedures result in samples with similar chemical characteristics and
mutagenic activity.

GENERATION OF ARTIFACTS
     Several phenomena have been proposed as mechanisms by which artifactually
high levels of organics or mutagens could exist in air-dilutioa samples of
diesel particles.  Dolan and Kittleson   suggested that air-dilution procedures
could produce particle samples containing greater organics levels than exist
in the ambient and postulated that a condensation mechanism might be possible
in the typical range of dilution ratios used in sample collection. However,
Plee and MacDonald   conducted a series of experiments and developed a model
of air dilution processes that argues against this postulate.  These authors
calculate a particle-organic association energy of about 7K cal/mole, which is
somewhat in excess of latent heat values.  Ross et al.   have studied the
physical chemistry of this association process and found it to be somewhat
more complicated.  Nevertheless, the global energy of dissociation appears to
                                                         1 f\                IS
be near the 7K cal/mole calculated by Plee and MacDonald.    Pierson et al.
suggest that air samples are similar to dilution tube samples in activity and
composition.
     In either case, the principal argument deals with the importance of
particles.  Gaseous organics, however, appear to be readily absorbed in the
           19
human lung.    Studies have now been done with toluene, benzene, methylene
chloride, nitrobenzene, and other gaseous hydrocarbons at low level and all
                                                                           19
seem to be absorbed very readily; usually, uptake is 30 to 70% of the dose.
So, whether the mutagenic material is in the particle or gas phase may be
immaterial.  The only significant issue is the relative amounts of active
                                           13
material in the two phases.  Kraft and Lies   report that about 70% of the PAH
in condenser experiments is trapped on a filter and only 10% is in the conden-
                   14
sate.  Stump et al.   find virtually no mutagenic activity in gasoline car
porous polymer gas trap samples. From diesel cars, the picture is more
complicated, but even there at least two-thirds of the activity is in particles.
Therefore, it appears little if any activity is lost in the gas phase.
     The final artifact mechanism proposed deals with nitrogen dioxide (NO.).
             20
Gibson et al.   have shown that re-exposure of filter samples of diesel
particles to the gas phase of diesel exhaust can elevate mutagenic activity.
      21
Bradow   described some diesel experiments in which dilution-air NO, levels of
about 100 ppm roughly tripled mutagenic activity.  However, more recent work
                              43

-------
has revealed that NO. levels above about 5 ppm are needed to produce this
effect.  At blower speeds of 350 CFM with a VW Rabbit-size vehicle, the N02
levels in any cycle never exceed 5 ppm.  With a larger Oldsmobile-size car, a
blower speed of 600 to 800 CFM is needed to maintain N02 concentrations below
5 ppm.  For gasoline cars, NO- is virtually undetectable.   Such N02 effects
apprently have not seriously complicated any of the exhaust samples collected
so far.  Very high dilution ratio experiments could be conducted to confirm
these findings.
     The filter media efficiency studies recently reported by Black and Dober-
     ryry
stein   (1981) have shown that several media give essentially identical mass
emission rates and extractable mass in handling diluted exhaust; therefore,
within reasonable limits, choice of media does not bias samples in these
                       23
respects.  Clark et al.   demonstrated that commonly used filter media all
produce the same sample mutagenic activity; therefore, this property is also
not biased by filter media.

SUMMARY
     Several working hypotheses can be drawn relative to measuring diesel
exhaust particles and organics.  They are:
1.   Passenger car dynamometer simulation and driving cycles are adequate
     for determining mass emission rates.  The choice of a simulation
     condition appears to make little difference.  Evidence for trucks is
     fragmentary, but high load freeway driving produces especially high
     particle emissions.
2.   Sampling of particles and organics can be done in a variety of ways.
     Since most sampling systems yield similar results, the choice is
     not critical.  If measuring mutagenic agents is the issue, then the
     distribution between gas and particle phase is also not critical.
3.   Artifactual generation of mutating agents in sampling seems to be
     relatively unimportant.  Filter media studies are negative, with
     respect to artifacts.  Although there is evidence of artifact
     formation from NO. concentrations above 5 ppm, this condition is not
     generally present under normal sampling conditions.
     Consequently, in the context of laboratory studies of diesel emissions,
it appears difficult to choose conditions that will produce totally invalid
samples.   Differences in particle emission factors from various vehicle types
are substantial, however, and these influences on air quality are important.
                             44

-------
REFERENCES

 1.  Gabele, P.A. and Colotta, J. (Oct. 1981) A Computer-Controlled, Real-time
     Auto Emissions Monitoring System, SAE Paper, Society of Automotive
     Engineers, Tulsa, OK.
 2.  Gabele, P.A., Black, F.M., King, F.G.,  Zweidinger, R.B. and Brittain,
     R.A. (Feb. 1981) Exhaust Emissions Patterns from Two Light-Duty Diesel
     Automobiles, SAE Paper No. 810081, Society of Automotive Engineers,
     Detroit, MI.
 3.  Gibbs, R.E., Hyde, J.D. and Whitby, R.  (Oct. 1981) Particulate Emission
     Characterization Studies of In-Use Diesel Automobiles, Paper presented at
     1981 EPA Diesel Emissions Symposium, Raleigh, NC.
 4.  Hare, C.T. and Baines, T.M. (Feb. 1979) Characterization of Particulate
     and Gaseous Emissions from Two Diesel Automobiles as Functions of Fuel
     and Driving Cycle, SAE Paper No. 790424, Society of Automotive Engineers,
     Detroit, MI.
 5.  Naman, T.M., Seizinger, D.E. and Clark, C.R. (Oct. 1981) Particulate
     Emissions from Spark-Ignition Engines,  Paper presented at 1981 EPA Diesel
     Emissions Symposium, Raleigh, NC.
 6.  Lang, J.M., Snow, L., Carlson, R., Black, F.M., Zweidinger, R. and
     Tejada, S.  (Oct. 1981) Characterization of Particulate Emissions from
     In-Use Gasoline-Fueled Motor Vehicles,  SAE Paper No. 811186, Society of
     Automotive Engineers, Tulsa, OK.
 7.  Dietzmann, H.E., Parness, M.A. and Bradow, R.L. (Oct. 1980) Emissions
     from Trucks by Chassis Version of 1983 Transient Procedure, SAE Paper No.
     801371, Society of Automotive Engineers, Baltimore, MD.
 8.  Dietzmann, H.E., Parness, M.A. and Bradow, R.L. (Jan. 1981) Emissions
     from Gasoline and Diesel Delivery Trucks by Chassis Transient Cycle, ASME
     Paper No. 81-DGP-6, American Society of Mechanical Engineers, Houston, IX.
 9.  Chan, T.L., Lee, P.S. and Siak, J.-S. (1981) Environ. Sci. Technol., 14,
     89-93.
 10.  Grimmer, G., Hildebrandt, A. and Bohnke. (1973) Zentralblatt Bakyt. Hyg. ,
     I Abt., 158, 22.
 11.  Lofroth, G. (1980) in Health Effects of Diesel Emissions, Vol I, Pepelko,
     W.E., Danner, R.M. and Clarke, N.A. ed., EPA 600/9-80-057a, U.S. Environ-
     mental Protection Agency, Cincinnati, OH, pp. 327-344.
 12.  Egeback, K.E., Tejle, G., Stenberg, U., Westerholm, R., Alsbert, T.,
     Rannug, U. and Sundvall, A. (Nov. 1981) A Comparative Study of Diesel and
     Undiluted Automobile Exhuasts Utilizing Polynuclear Aromatic Hydrocarbons
     Analysis and Mutagenicity Tests, Paper presented at the International
     Symposium on Polynuclear Organic Compounds, Columbus, OH.
 13.  Kraft, J. and Lies, K.-H.  (Feb. 1981) Polycyclic Aromatic Hydrocarbons in
     the Exhaust of Gasoline and Diesel Vehicles, SAE Paper No. 810082, Society
     of Automotive Engineers, Detroit, MI.
 14.  Stump, F., Bradow, R.L., Ray, W., Dropkin, D., Zweidinger, R.B., Sigsby,
     J.E. and Snow, R. (Oct. 1981) Trapping Gaseous Hydrocarbons For Mutagenesis
     Testing, Poster presented at 1981 EPA Diesel Emissions Symposium, Raleigh,
     NC.
 15.  Dolan, D.F. and Kittelson, D.B. (Feb. 1979) Roadway Measurements of
     Diesel Exhaust Aerosols, SAE Paper 790492, Society of Automotive Engineers,
     Detroit, MI.
 16.  Plee, S.L. and MacDonald, J.S. (Feb. 1980) Some Rudiments of Diesel
     Particulate Emissions, SAE Paper No. 800251, Society of Automotive
     Engineers, Detroit, MI.
 17.  Ross, M.M., Risby, T.H., Lestz, S.S. and Yasbin, R.E. (Oct. 1981) Physico-
     Chemical Properties of Diesel Particulate Matter, Poster presented  at
     1981 EPA Diesel Emissions Symposium, Raleigh, NC.
                               45

-------
18.  Pierson, W.R.,  Gorse,  R.A.,  Szkarlet,  A.C., Brachaczek, W.W., Japar, S.M.
     and Lee, F.S.-C.  (Oct. 1981) Mutagenicity and Chemical Characteristics of
     Carbonaceous Particulate Matter from Vehicles on the Road, Paper presented
     at 1981 EPA Diesel Emission  Symposium, Raleigh,  NC.
19.  Astrand, I. (1975) Scand. J. Work Environ.  Hlth.,  1, 199-218.
20.  Gibson, T.L., Ricci, A.I. and Williams, R.L.  (Nov.  1980) Measurement of
     Polynuclear Aromatic Hydrocarbons, Their Derivatives and Their Reactivity
     in Diesel Automobile Exhaust, GMR No.  3478, General  Motors, Dearborn, MI.
21.  Bradow, R.L. (Nov. 1980) Bull.  NY Acad. of  Med.  56,  797-811.
22.  Black,  F.M. and Doberstein,  L.  (June 1981)  Filter Media for Collecting
     Diesel  Particulate Matter, EPA Report  No. 600/52-81-071, U.S. Environmental
     Protection Agency, Research  Triangle Park,  NC.
23.  Clark,  C.R., Truex, T.J., Lee,  F.S.C.  and Salmeen, I.T.  (1981) Atmos.
     Environ., 15, 397-402.
                             46

-------
              PARTICULATE EMISSIONS FROM SPARK-IGNITION ENGINES

                                     by

                                Ted M.  Naman
                               0. E. Seizinger
                          U.S. Department of Energy
                    Bartlesville Energy Technology Center
                           Bartlesville, Oklahoma

                              Charles R. Clark
                Inhalation and Toxicology Research Institute
                           Albuquerque, New Mexico
Experiments  were  conducted at the U.S.  Department  of Energy's Bartlesville
(Okla.)  Energy  Technology Center to quantify  particulate  and gaseous emis-
sions from current-production vehicles equipped with spark-ignition engines,
to  determine the  influence  of fuel and  ambient  temperature on particulate
emissions,   and  to  characterize  particulates  in  terms   of  their  carbon
content, soluble organic  fractions, and biological activity.

Four  1980-81  model-year  vehicles  equipped  with  oxidation  and  three-way
catalysts  and  spark-ignition engines ranging  from  1.6  liter, 4-cylinder to
4.3  liter,  V8  (see table  1) were  tested  on a  climate-controlled  chassis
dynamometer  using  the driving  cycles  of the  1975  Federal  Test Procedure.
The vehicles were  operated at 20°, 50°, 75°, and 100° F (-7°,  10°, 24°, and
38° C,  respectively) ambients on gasoline and at 75° F ambient on four fuel
blends:   90  percent gasoline/10 percent ethanol, 90 percent gasoline/10 per-
cent  methanol,  93 percent  gasoline/7  percent methyl  tertiary butyl  ether,
and a commercial gasohol.

Particulate  matter was  collected on 40-  by  40-inch  filters using the total
volume  of  the exhaust,  and  on conventional  47 mm  filters  using a sampling
probe and  a  portion  of the  exhaust.   Bioassays  of dichloromethane extracts
of  the  samples were  carried out  using  the Salmonella  mutagenicity  (Ames)
test at the  Lovelace Inhalation Toxicology Research Institute.

The results  from  the Federal Test Procedure (figure 1) showed  a significant
reduction  in particulate emissions with the alcohol/gasoline fuel  blends
when  compared  to  gasoline alone.  The methyl  tertiary butyl ether/gasoline
fuel  blend showed a  slight   reduction in particulate  emissions.  The  carbon
monoxide  emissions  were  slightly  reduced  with  the  alcohol/gasoline  fuel
blends.    Hydrocarbon  emissions  remained relatively  unchanged,  oxides  of
nitrogen  emissions were  slightly increased,  and fuel  economy was  2  to 3
percent lower (data not shown).  Ambient temperature seemed to  have a slight
effect  on  particulate  emissions.    Overall,  particulates   emitted  from the
vehicles with spark-ignition engines were 90 to 100 times lower than partic-
ulates emitted from current-production diesel vehicles.
                                      47

-------
Samples  of  participate  extracts  from  vehicles  fueled  with  gasoline  and
gasohol were separated using a silica column.  The results obtained  from the
separation showed a higher percentage of polar compounds in the gasohol  than
in the gasoline samples (figure 2).

Benzo(a)pyrene  and  nitropyrene levels  in the  gasohol  participate  extracts
were  at  least  50  percent  lower  than the  measured levels  in particulate
extracts from gasoline-fueled vehicles.

Dichloromethane extracts  of  the particulate exhaust from vehicles operating
on  gasoline  were evaluated  in Salmonella  strain  TA-100,  with  and without
the  addition  of a  liver  enzyme preparation (S-9) as a  source of metabolic
enzymes.   Extracts  of particulate  exhaust from the  four  vehicles  operated
on  gasoline  and/or on  alcohol/gasoline fuel blends  produced direct, dose-
related  increases  in mutagenicity  (see  table 2).    The   addition  of  S-9
either decreased or did not alter the observed direct mutagenicity (data  not
shown).

The addition  of 10  percent ethanol to  gasoline  either  decreased or did  not
significantly change  the mutagenicity of  the resultant  exhaust particulate
extracts (table 2).   Operation of the Mercury Monarch on the methanol blend
increased the mutagenicity of the exhaust particulate extracts but decreased
mutagenicity  in  the  Chevrolet  Citation.   Commercially available  gasohol
resulted in particulate extracts that were less  mutagenic in the Ford Escort
but not significantly different in the other cars.

The mutagenic  potencies  of  the extracts  do  not reflect differences among
cars in the mass of mutagenic material  associated with the particulate emis-
sion rates.  Therefore,  the  mutagenicity data were normalized  so  that com-
parisons of the amount  of mutagenicity could be made  among cars.   This was
done by dividing  the  mass of dichloromethane extractable material  from each
filter by the number of miles of vehicle operation (26),  to yield mg/mile of
particulate  associated  organic  material.    This  number was  multiplied by
revertants per microgram (ug) to estimate the amount of mutagenicity emitted
from each car (revertants per mile).

When compared to gasoline, the addition of 10 percent ethanol  or methanol to
gasoline,  or operating the  vehicles  on the commercial gasohol,  reduced the
mass emission rate (mg/mi) of organic materials  associated  with the particu-
lates.   This resulted in  significant reductions in revertants  per mile for
all  of the alcohol  fuel  blends tested (table 2).
                                       48

-------
                           Table 1.   Test vehicles
Ford
Escort
Oldsmobile
Cutlass
Chevrolet
Citation
Mercury
Monarch
Engine displacement,         98 (1.6)    263 (4.3)    151 (2.5)    250 (4.1)
  CID (liters)
Carburetion                  2 bbl       2 bbl         2 bbl        1  bbl
Compression ratio            8.8         7.5          8.2          8,6
Transmission                 Manual      Auto         Auto         Auto
                             4-spd
Emission Control System:
  EGR                        Yes         Yes          Yes          Yes
  Air pump                   Yes         Yes          No           Yes
  Air injection              No          No           Yes          No
  Oxidation catalyst         No          No           Yes          Yes
  Three-way catalyst         Yes         Yes          No           No
  Charcoal canister          Yes         Yes          Yes          Yes

Axle ratio                   3.59        2.29         2.84         2.79
Inertia weight, Ib           2375        3750         2875         3625
Actual dyno load, hp         6.4         11.5         6.6          11.1
                                      49

-------
        Table 2.  Influence of alcohol fuel blends on mutagenicity of
                  spark-ignition engine exhaust particulate extracts
 Vehicle
   and
  Fuel
Revertants/ug
   Extract
    TA-1003
      Emission of
Particulate Associated
   Organic Material       Revertants
        (mg/mi)            per Mile
Ford Escort
  Gasoline
  Ethanol blend
  Commercial gasohol

Oldsmobile Cutlass
  Gasoline
  Ethanol blend
  Commercial gasohol

Chevrolet Citation
  Gasoline
  Ethanol blend
  Methane1 blend
  Commercial gasohol

Mercury Monarch
  Gasoline
  Ethanol blend
  Methanol blend
  Commercial gasohol
      10
       9
       4
      10
       5
      13
      17
      14
      11
      10
      16
      12
      26
      20
          1.5
          1.1
          1.2
          1.7
          0.6
          0.6
          1.9
          0.9
          0.8
          1.0
          7.1
          2.8
          3.3
          2.2
 15,000
  9,900
  4,800
 17,000
  3,000
  7,800
 32,300
 12,600
  8,800
 10,000
114,000
 34,000
 86,000
 44,000
 Slope of linear portion of dose-response curve,  without S-9.

Research performed in part under U.S.  Department  of Energy Contract Number
OE-AC04-76EV01013.
                                      50

-------
   40
    30
 2  20
    10
          Gasolina  Gasoline  Commercial  Gasoline   Gasoline
                  + 10%   gowhol   +10%    +7%
                  EIOH          MeOH    MTBE
    Figure  1.   Influence of fuel  extenders
                on particulate  emissions.
7O
60
U)
UJ
s 50
t—
O
< 40
X
H-
z
UJ
u
£ 2°
Q_
10
o







~
KM

1
1
5v!
1




T*7

•';'-';

;.';!
'•v!"







&
S
^
1
,

•'''.'
'* "•



^"•trT-'h Gasoline
(- » '-i GasohoL





            MECL
                   TEMPERATURE, 75'F
Figure  2.   Particulate  extracts  from vehicles
            operating on  gasoline  and gasohol.
                        51

-------
                PARTICULATE EMISSION CHARACTERIZATION STUDIES
                                     OF
                          IN-USE DIESEL AUTOMOBILES
                                     by


                Richard Gibbs, James Hyde, and Robert Whitby
                                Division of Air
           New York State Department of Environmental Conservation
                              Albany, New York


     A sample of 20 in-use diesel automobiles has been repeatedly tested over
a two-year period to accumulate emissions characterization data with major
emphasis on particulate.   Each vehicle test included replicate sample col-
lection for "as received" and "control fuel, control oil" vehicle conditions.
Driving cycles tested at each vehicle condition included:  FTP, CFDS (CUE),
HFET, 50 mph cruise, NYCC, and Idle.  Measurements in each test cycle included
gaseous emissions, fuel economy, and particulate emissions; also, individual
particulate samples were collected for each individual test cycle in suffi-
cient quantity for subsequent chemical and bioassay analyses.  These particu-
late samples were quantified for soluble organic fraction (SOF) and Ames test
direct-acting mutagenlc response by TA98 (-).  These parameters are examined
for 60 vehicle tests to indicate particulate character effects for vehicle
types and test cycles.  Results will be presented for three vehicles in the
sample group which have been tested over sufficient mileage accumulation
intervals to provide limited insight into vehicle aging effects.

     Table I gives average values of the mutagenic activity of the SOF
directly.  In general, FTP samples exhibit a higher specific activity than
samples from the same vehicle for other driving cycles.  Expressed as a per-
cent of total particulate, however, the SOF (7.) is generally less for the
FTP than other cycles as seen in Table II.

     When SOF mutagenic response and SOF (7.) data are combined to express
mutagenic response on a particulate mass basis, these differences are mostly
indiscernable as shown in Table III.

     From Table III, with some exceptions, it can be seen that the bio-activi-
ty on a total particulate basis is quite uniform within a vehicle type.  This
is in spite of rather large variations in activity among the various vehicles
tested to generate the averages.  Thus, we conclude that, as a first
                                       52

-------
approximation, for a given vehicle teat a "gram of particulata is a gram of
particulate" regardless of driving cycle.  When variations in vehicle partic-
ulate emission rate per distance travelled are incorporated to the data of
Table III, the per-mile emission of bio-active material can be calculated as
given in Table IV.

     Since all test cycles except the FTP begin with a warm vehicle, the
foregoing conclusions may not include parameters related to vehicle start-up.
A battery of tests were performed on a VW diesel in the Winter of 1981 to
preliminarily investigate cold-start effects on particulate, SOF, and muta-
genic activity.  Overnight vehicle soak at laboratory and outdoor ambient
conditions were followed by:  FTP, Bag I, 10 min. pause, FTP, Bag III to
give separate particulate samples for each bag at each condition.  The cold-
ambient tests were repeated, and average values from two runs are reported.
Continuous temperature recordings of ambient temperature, crankcase lubri-
cating oil, and fuel temperature (between pump and injectors) were obtained.
When results for the normal FTP Bag III were used as a basis for comparison,
the other three test conditions yield ratios as given in Table V.

     Comparison of the 0°C Bag I results to the 20°C base condition showed a
74% increase in particulate emission, 11% increase in SOF emission corres-
ponding to a 36% decrease in SOF expressed as a percentage of particulate.
SOF mutagenic response was increased by a factor of 3.6 and 4.0 when expres-
sed as revertants/mile.  These results are in general agreement with those
presented above in comparison of FTP and other driving cycles for the in-use
sample group.
        GM
        VW
        MB
               Table I.  Mutagenic Activity Per Microgram SOF
                             (Revertants/^g SOF)
             FTP
HFET   SOmph Cruise  CFDS   NYCC   IDLE   Sample a
3.8
11.3
4.8
2.4
10.9
4.4
2.1
8.1
3.5
2.7
11.0
5.9
1.4
16.8
1.8
2.4
3.4
2.7
57
36
18
                             Table II.  SOF (%)

             FTP    HFET   SOmph Cruise  CFDS   NYCC   IDLE   Sample n
GM
VW
MB
25.1
20.0
14.0
34.1
21.5
13.6
39.3
20.7
15.2
31.2
22.7
14.6
31.7
33.4
14.8
24.6
55.0
14.6
57
36
18
                                       53

-------
  Table III.  Mutagenic Activity Per Microgram Particulate
                        (Revertants/^g  Particulate)


                                           NYCC   IDLE   Sample n_
GM
VW
MB
0.77
1.95
0.53
0.75
2.06
0.45
0.72
1.53
0.40
0.74
2.26
0.68
0.36
5.82
0.27
0.43
1.94
0.39
57
36
18
       Table  IV.   Mutagenic Activity Per Vehicle Mile
                          (105 Revertants/Mile)
      FTP
CFDS   NYCC    IDLE
GM
VW
MB
6.5
6.9
2.8
2.7
6.1
1.7
2.2
4.4
1.3
3.8
7.4
3.0
8.6
27.0
6.2
0.74
0.47
0.17
57
36
18
*  c
 10-* Revertants/Minute
                 Table V. VW Cold Sure Partlculaca Coa?«rlaone


Vehicle Teat Condition
Baae Condition .
(FT? Bat III)
Mean TeaBera^uref °C
Overnight 1 Injector
Soak 1 Fuel Line

20 1 25
Nota: Reanlta below ihadad areaa are ratloa
to bale condition
Bag III after ambient
Cold Soak Bag I
normal FT?
Bag I
Cold Ambient Soak
FTP Bag I

0

20

0

18

21

5
Crankeaee
Lube

90



98

48

38
Pirtlculata

(•/ml

0,34
1%
'' - \ !*"

1.0

1.18

1.7ft
SOF

(T)

24.8



0.87

0.79

0.64
SOF

(«Vm)

0,084



0.87

0.93

1.11
TA98 (-) Eevartanta

R/d* SOF

5.8
-


1.5

1.8

3.6

a/,,« Part.

Ir44

', '• "• V

1.3

1.4

2.3
.
^ ^/oil

4.9



1.3

1.7

4.0
                               54

-------
            DIESEL EXHAUST TREATMENT DEVICES:  EFFECTS ON GASEOUS
             AND PARTICULATE  EMISSIONS  AND ON  MUTAGENIC ACTIVITY
                                     by
                   R. A.  Gorse, Jr., J. J. Florek, W. Young,
                      J.  A.  Brown,  Jr. and I.  Salmeen
                                Research  Staff
                              Ford Motor  Company
                              Dearborn, Michigan
      In  conjunction with  the Ford research effort on the control of diesel
 exhaust  particulate emissions we have  investigated and characterized the
 emissions  from  four diesel particulate  emission  control devices.  These
 include  two ceramic honeycomb monolithic filter  traps, one of which is coated
 with  a     catalyst, a  compactsd wire-mesh particulate trap with a precious metal  (PM)
 catalyst coating, and  a free-flow monolithic catalyst with a PM coating.
 The                       four devices  and the test vehicle are described in
 Table I.

     Gaseous and particulate emissions  have been compared to those from the
 uncontrolled baseline  vehicle.  The particulate  material has   been fractiona-
 ted into the sulfate,  soluble organic and inorganic fractions by Soxhlet
 extraction using dichloromethane.  The  extract was analyzed by high perform-
 ance liquid chromatography (HPLC) analysis with  fluorescence and ultra-
 violet absorption detectors and subsequently characterized by the Salmonella
 typhimurium plate incorporation assay  (Ames) using strains TA-98, TA-100
 and TA-1538, without metabolic activation.  The  results are summarized in
 Table II.

     The ceramic honeycomb traps are extremely efficient filters for remov-
 al of the  inorganic fraction (mostly elemental carbon) and are less effic-
 ient for removal of the soluble organic fraction.  The catalyst coated
honeycomb  trap shows some activity for  CO and for gaseous hydrocarbon but
 is also capable of producing sulfate at 507. above the baseline emission
value.

     The catalyst coated wire-mesh trap shows high activity for CO, HC and
also for the soluble organic fraction but in addition produces largt yields
of sulfate.  It is not an extremely efficient filter for the inorganic fract-
ion of the particulate material.
                                     55

-------
      The catalyst coated ceramic honeycomb trap and the catalyst coated wire-
 mesh trap significantly reduce the number of Salmonella revertants per mile
_travelled.  Provided  that  the wire-mesh trap could be modified to minimize sulf-
Tte formation,  both of  the above traps have the potential for reducing the total
 particulate emission  rates to values near or below the 0.2 g/mi level.


      The HPLC- results show that the extracts from the trap systems in general
 are qualitatively similar  to those from the baseline vehicle.  The traps seem
 to show the highest  collection efficiency for the aliphatic type compounds in
 the soluble organic  fraction.

      Regeneration and  durability  of the  trap systems studied remain as areas
 for further research  and will  not  be addressed in this report.

                                                                    a,b
           Table I.  Description of Diesel Exhaust Treatment Devices
Designation     Type          Catalyst       Volume   •    Cell  Density   Filter
                                                                        Area
                                              In3          cells/in2       in2
A

B'

C

D
Ceramic Honeycomb
Filter Trap
Ceramic Honeycomb
Filter Trap
'Compacted Kni-tted
Wire-Mesh Trap
Free-Flow Monolithic
None

PM

PM

PM
119

119

186

80
100

100

NAC

100
1970

1970

8370

NAC
»
 The control devices used in this study represent early  experimental  versions
 of potential particulate control devices and hence  the  results  do  not reflect
 future design modifications.

Vehicle = 2.3 1 Opel (European) Diesel, 3000 Ibs IW, 10.3  hp  PAU,  3.89 Axle
 Ratio, A Speed Manual Transmission, CVS Driving Cycle.  During the  course  of
 this study the original engine (A) was replaced with a  new engine  (B) due to
 excessive HC emissions.

CNA = Not Applicable
                                       t>o

-------
          Table II.  Gaseous and Particulate Emialona from Diesel Partlculate Control Devices
Device Engine
o
Baseline A
Vehicle
B
A, Glean A
B
A, Loaded A
B
B, Loaded A
G A
D A
HC
g/mi
0.99

0.61
0.91
0.98
0.64
0.77
0.52
0.05
0.89
CO
g/mi
1.65

1.50
1.03
0.93
1.05
0.93
0.56
0.08
0.72
Total
Particulate
mg/mi
436

303
Ratio to Baseline
0.44
0.21
0.21
0.25
0.24
0.68
0.89
Sulfate
mg/mi
13

23
Emission
0.31
0.61
0.31
0.52
1.54
12.5
1.08
Organic
mg/mi
309

138
Rate
0.56
0.41
0.27
0.46
0.25
0.06
0.84
Inorganic
mg/mi
127

142
0.02
0.00
0.05
0.05
0.07
0.87
0.99
Ames
rev/ug
1.9

2.0
1.9
2.6
1.9
2.2
0.42
1.2
1.2
rev/mi
xlO'5
5.9

2.9
1.1
1.1
0.51
1.0
0.10
0.07
1.2
 Engine A replaced by Engine B due to excessively high HC emission rate.
 Trap !iad just been regenerated.
cl'rap loaded with soot by mileage accumalatlon.

-------
                       CHARACTERIZATION  AND OXIDATION
                           OF DIESEL PARTICULATE

                                     by

                 David A.  Trayser and Louis J.  Hillenbrand
                       Battelle-Columbus Laboratories
                              505 King Avenue
                           Columbus, Ohio  43201
INTRODUCTION
          A study was recently completed by Battelle-Columbus Laboratories
for the U.S. Environmental Protection Agency to evaluate emissions control
on light-duty diesel vehicles by postcylinder oxidation.  The primary objec-
tive of this program was to determine the feasibility of thermal  or
catalytic oxidation as a means of diesel particulate emissions control.

          Two major efforts in this program involved characterization of the
particulate for chemical and physical parameters relating to ignition and
oxidation, and development of a catalytic ignition concept for reducing the
ignition temperature of the particulate in the exhaust system.

CHARACTERIZATION OF THE PARTICULATE

          Particulate characteristics were measured using samples collected
directly from the surface of the exhaust pipe and samples collected by filter
from a dilution system.

Physical Properties

          The particulate physical properties included surface area, size
distribution, and mass concentration.  These were measured using dilution
system samples, hence, representing the particulate after entering the atmos-
phere.  The surface area was approximately 100 m2/g, the mass median particle
diameter was in the range of 0.1 to 0.3 ym and was observed to increase in
size as engine speed and load increased, and the mass concentration in the
engine tail pipe was observed also to vary with engine speed and load. Mass
concentrations of particulate in the exhaust gas varied from 20 mg/m3 at the
lowest speed and lightest load to 500 mg/m3 at maximum speed and load. These
values translate to 0.1 to 3.8 grams per mile, respectively, a fairly "dirty"
exhaust for an automotive diesel (and not representative of current diesel
engines).
                                      58

-------
Chemical  Properties

          The chemical properties were measured using the exhaust-pipe
particulate samples and included soluble organics (using toluene), carbon,
hydrogen, oxygen, ash, and trace minerals.  The soluble organics ranged from
2 to 10 percent, with no evident correlation with engine speed/load condi-
tions.  By comparison, the soluble organic content measured in particulate
collected in the dilution system, for the same range of engine speed and load
conditions, varied from 3 to 25 percent, increasing as speed and load decreased.

          Carbon content ranged between 73 and 93 percent, hydrogen content
varied from 0.5 to 1.8 percent, and ash content ranged from 0.1  to 2.2 per-
cent.  Again, no correlation with engine conditions was noted for any of
these parameters.  About 7 percent oxygen was found in two particulate
samples, and 23 percent in a third sample.  Part of the oxygen is present in
the particulate as $04 and part may represent partial oxidation  (to C02,
perhaps) but with the products still bound in the particulate.

          The trace mineral analyses revealed considerable calcium, iron,
phosphorus, and magnesium in the particulate.  Similar analyses  of samples
of the fuel and lube oil indicated that the lube oil was most likely the
source of these elements as well as of chromium, copper, manganese, and lead,
found in lesser quantities in the particulate.  These results imply that a
substantial portion of the particulate may derive from the lube  oil.

Ignition Properties

          A Differential Thermal Analyzer technique was used to  measure
ignition temperatures and maximum oxidation temperatures for the particulate
samples obtained in the engine exhaust pipe.  The mean ignition  temperature
for the samples evaluated was 594 C, with a variation range of 583 to 604 C.
The only significant correlations with engine operating conditions and with
physical and chemical properties appeared to be with the presence of hydro-
carbons in the exhaust and with aluminum and lead in the particulate. Lower
ignition temperatures occurred with higher exhaust gas hydrocarbons and
higher particulate aluminum and lead.

          More significant changes in ignition temperature were  observed when
the bulk density and the amount of the particulate sample in the DTA were
altered.  Increasing the sample size and the bulk density by factors of 10
resulted in a decrease in ignition temperature of about 150 C.  This result
suggests that if a method could be devised for compacting the particulate
in a trap, it would be more easily ignited under normal exhaust-gas conditions.

CATALYTIC IGNITION

          In this study the catalytic ignition of diesel particulate was
initially developed in bench-scale experiments.  Final experiments were
conducted in the exhaust system of an engine to verify the applicability of
the concept to the actual engine environment.
                                      59

-------
Bench-Scale Experiments

         The bench-scale experiments were carried out using a hot-tube
reactor in which small samples of the diesel particulate were subjected  to
a gradually increasing temperature until ignition and oxidation occurred.
Experiments were conducted to identify potentially catalytic materials,  to
explore methods of catalyst application, and to determine the magnitude  of
the catalytic effect in relation to the amount of catalytic material used.

         Metal salts such as copper chloride, manganese chloride, and cobalt
chloride were found to be capable of reducing the ignition temperature of the
particulate by as much as 200 C.  The addition of sodium or ammonium salts
to the metal salt reduced the ignition temperature another 50 C, for a total
reduction of 250 C (resulting in an ignition temperature of about 350 C  for
the catalyzed particulate).  For most of this work, copper chloride equiva-
lent  to 7.5 mg Cu/g soot was used to achieve the catalyzed ignition.

Exhaust System Experiments

         Final experiments were conducted in the exhaust pipe of a production
diesel engine, where the results of the bench-scale experiments were confirmed.
Various particulate trap materials and configurations were tried, including
quartz wool, stainless steel wool, foamed alumina, and porous ceramic honey-
comb.  The porous ceramic honeycomb traps proved to be the most suitable,
providing high surface area in a small volume, good collection efficiency at
low pressure loss, and tolerance to the high temperatures reached during
oxidation of the particulate.

         Successful regeneration of the trap in the engine exhaust was
achieved by intermittent injection of a water solution of the copper chloride/
sodium chloride catalyst mixture into the exhaust stream immediately upstream
of the trap, after a period of particulate collection on the trap.  The
actual ignition temperature in these tests ranged from 350 to 400 C.  The
amount of soot burned in each of these final trials is unknown but the copper
concentration is believed to have been somewhat greater than used in most
of the bench-scale tests.

CONCLUSIONS

         The concept of using a catalyst material  introduced into the exhaust
of a  diesel engine in a manner which allows the catalyst to associate with
particulate collected on a trap has been successfully demonstrated.   The
catalyst acts to reduce the ignition temperature of the trapped particulate
by about 250 C, which results in particulate burnoff and trap regeneration
at significantly lower exhaust gas temperatures than would be required
otherwise.   This concept could provide more highly controlled burnoff of the
trapped particulate,  leading to lower peak oxidation temperatures in the
trap (promoting longer trap life and improved trap reliability)  and to
minimization of the danger of unscheduled trap burnoff as a fire hazard.

         Practical  application of this concept requires development of suit-
able hardware and identification of the optimum operation cycle for the trap-
oxidizer/catalyst injection system.   The enclosed figure illustrates, in
simplified  form, a particulate control  system that might be developed based
on the metal-salt-catalyst concept briefly described in this paper.
                                     ou

-------
             EXHAUST
             MANIFOLD^,
ENGINE
  PARTICULATE
  TRAP
                               PUMP
              CATALYSTJ TRANSFER  CATALYST
              INJECTOR   LINE      RESERVOIR
        PARTICULATE
        TRAP
                                 TRANSFER
                                rLlNE
                                 SOLENOID
DIESEL ENGINE PARTICULATE CONTROL SYSTEM
BASED ON BATTELLE METAL-SALT-CATALYST
                 CONCEPT
                     61

-------
   HEAVY-DUTY DIESEL ENGINE EMISSIONS -- SOME EFFECTS OF CONTROL  TECHNOLOGY

                                      by

                           J.M. Perez and R.V. Bower
                              Research Department
                            Caterpillar Tractor Co.
                               Peoria, Illinois


     Use of various control technology methods to reduce specific emissions
such as NOX or particulates usually result in changes to other constituents in
the exhaust of diesel engines.  The effects of control technology on  EPA
Advisory Circular No. 76 emissions are reported.  Unregulated emission
trade-offs as a result of timing, EGR, catalysts, and engine modifications  are
discussed.  Fuel consumption increased with most changes.

     Although the emission levels are changed as a result of the  control
technology, the emissions pose no obvious health risk based on estimated
exposure levels and available health effects data.
                        Table 1.   Engine Change Tradeoffs
J • Increase | « Decrease • • » No Change
Change
PC-OI
EGR
Timing
Advance
Retard
AftercooHng
Injector SAC
Volume Increase
Catalyst*
Fuel
(BaP Increase)
Partle
Total
t
t
ulates
SEF
H
I
1
n
ti
ti
t
t!


_t
t
ti
t
t*


HC
NO,
nit
i
ALO
11
U 1

—
BAP
t
1

t 11
H I
UIU
t
1
t
L



-*
1
t
1
I
t
\
t
\
1

T
Fuel
Cons
1
t

Power
—
n

*i 	
t
H
—
~*
I
i
—
—



                          NH3  f
                                      62

-------
SECTION 2



CHEMICAL AND BIOASSAY CHARACTERIZATION
                        63

-------
METHODOLOGY OF FRACTIONATION AND PARTITION OF  DIESEL
EXHAUST PARTICIPATE SAMPLES
BRUCE A. PETERSEN AND CHENG CHEN  CHUANG,  Battelle-Columbus  Laboratories,
505 King Avenue, Columbus; Ohio,  USA
INTRODUCTION
   The decision by  the U.S.  Environmental  Protection Agency  to  examine  the
health effects of diesel particulate  emissions has  resulted  in  a  great  increase
in the number of investigations  to analyze their physical, chemical  and bio-
logical characteristics.   Diesel particulate emissions  are  a very complex
mixture of carbonaceous matter containing  adsorbed  and/or condensed  organic
components from the combustion of fuel and lubricants.   The  organic  solvent
extractable fraction of these particulates is extremely  complex and  has been
                                                     2
reported to contain hundreds of  individual compounds.    A number  of  studies
have reported that  these solvent extracts  are mutagenic  as determined by the
                                 3—8
Ames Salmonella microbial assay.      The greatest activity has  been  found to be
present in certain  compound  class fractions within  these extracts which do not
require metabolic activation.      Chemical characterization studies have been
used in an attempt  to identify the compound classes or specific compounds
present in these fractions.  '       In addition, polynuclear aromatic hydro-
carbons (PAH), particularly  pyrene, have been reported to readily react with
nitrogen oxides to  form nitrated derivatives which  are powerful direct-acting
         12 13
mutagens.  '    Both the PAHs and nitrogen oxides are present in  the exhaust
of diesel engines and thus the formation of nitroaromatics may  indeed be
possible.  These results have prompted increased attention to the characteri-
zation of diesel particulate extracts to identify the compounds or compound
                                                          4 14-17
classes which are responsible for the biological activity. '       Identifica-
tion of specific mutagens or classes  of mutagens is important to  determine
whether these components are present  in the exhaust which is emitted to the
atmosphere or are formed as  a result  of the dilution particulate  collection
and analysis procedures.  This information must be  established  before potential
effects of these compounds on the environment can be addressed.
   The objective of this paper is first to briefly  review the general tech-
niques used for chemical characterization  of diesel particulates.  This review
is followed by a description of  a procedure which can be used to  quantitatively
characterize the chemical and direct-acting mutagenic properties  of  the soluble
                             64

-------
organic fraction of these participates.  This procedure  is  based  on  a mass
fractionation and Salmonella bioassay to isolate classes of compounds which  are
responsible for the mutagenic activity.  Finally, applications  of this  pro-
cedure are presented for particulate extracts from  two light  duty diesel
engines and one heavy duty diesel engine.

MATERIALS AND METHODS
    Review of General Techniques.  Figure 1  illustrates a typical  experimental
approach used in the characterization of the soluble organic  fraction of  diesel
particulates.  This general approach involves extraction of the organic mater-
ial from the collected  particulate  using an organic solvent,  concentration of
the extract for subsequent analysis, fractionation  into  less  complex non-polar
and polar  sub-groups and a final  concentration  to facilitate  chemical and
biological  analysis.  Ames Salmonella bioassay  on the fractions requires  that
the solvent be exchanged with one which is  compatible with  the  assay, such as
dimethyl sulfoxide  (DMSO).
                                                                CHEMICAL
                                                                ANALYSIS
   EXTRACTION
CONCENTRATION


FRACTIONATION
CONCENTRATION
                                                                 SOLVENT
                                                                EXCHANGE
                                                                 BIOASSAY
Fig. 1.   General Experimental Approach
                             65

-------
Chemical analysis of the fractions for identification of individual  compounds
has been carried out using a variety of chromatographic and mass spectrometric
techniques.  The non-polar fractions have been well characterized and consist
of PAH and aliphatic hydrocarbons.18"22  Identification of specific  compounds
in the polar fractions is difficult since these compounds are thermally labile,
highly polar, low in volatility and very low in concentration. '     Three
specific techniques have been used to identify constituents in these polar
fractions:   (1) high resolution gas chromatography/high resolution mass
spectrometry (HRGC/HRMS) , (2) combined liquid chromatography/chemical ioniza-
tion mass spectrometry (HPLC/MS)  , and (3) high resolution gas chromatography/
negative ion chemical ionization mass spectrometry (HRGC/NCI-MS) using on-
column injection.    The latter two methods provide injection of the fraction
at ambient temperature to avoid thermal degradation of the polar organics.  In
addition, since the energetics of chemical ionization are less than  that of
conventional electron impact ionization, the ionized molecules fragment less
and a greater amount of the total ion current remains with the molecular ion.
This feature aids the identification of the individual compounds.  Negative ion
chemical ionization is especially sensitive and selective for detection of
several types of polar organics such as nitro or oxygenated PAH, because of
the electronegativity of the polar substituent.
   Within this general experimental approach, the procedures for extraction,
concentration and fractionation can directly influence the integrity of the
extract.  To relate analysis data to the initial particulate sample, these
procedures should not change the chemical and biological characteristics of
the sample.  An overview of these three procedures is presented in the follow-
ing section.
   Extraction.   Two widely employed methods for removal of organics  associated
with diesel particulates are soxhlet extraction and ultrasonic extraction using
a suitable organic solvent.  The total quantity of material extracted by these
techniques is influenced by the type of particulate being extracted, contact
time with the solvent, temperature of the solvent during the extraction, and
polarity of the solvent.  Recent studies conducted by the Coordinating Research
Council indicate that either technique can be used satisfactorily to measure
the total soluble organic fraction.    Extraction by sonication is usually more
rapid (0.5 to 1.5 hours) than using the Soxhlet technique (3 to 24 hours) but
sonication requires several additional analytical steps to remove suspended
soot and filter particles.  Soxhlet extraction can be conducted unattended and
many particulate samples can be extracted simultaneously in separate apparatuses.
                               66

-------
    The type of organic solvent used for extraction has received much attention
recently.  In general, methanol and aromatic-alcohol mixtures have been found  to
extract the largest quantities of material from the particulate presumably due
to increased extraction of inorganic material. '  '    Aromatic solvents have
                                                  28
been shown to extract the largest quantity of PAH.    Methylene chloride,
however, has been reported to extract the largest quantity of biologically
active material from particulate, and is usually the solvent chosen for bio-
                  26
logical screening.
    Techniques such as vacuum sublimation and thermal desorption have been used
as rapid methods for determining the organic content of particulate filter
samples.  Although these techniques may provide reasonable results for the
volatile fraction of the organics associated with diesel particulates, the
nonvolatile components are not completely removed.  Collection of the organics
using these techniques is difficult and chemical characterization of individual
compounds and compound class is not possible.
    Concentration.  Concentration is necessary within various steps of the
overall procedure.  The solvent must be removed from the extract, or an aliquot
of the extract, to determine the residue mass for calculating the total organic
mass.  To facilitate separation into fractions, the extract must be concentrated
to a small volume, typically 50 ml or less.  Finally, once fractionation is
complete, the resulting fractions must be concentrated for chemical analysis by
instrumental techniques.
    Typical methods of concentration include removal of the solvent by rotary-
film evaporation  , vortex evaporation  , nitrogen stream blow-down" '" , and
Kuderna-Danish concentration.  The method of concentration will in many cases
depend on the volume of solvent to be removed.  Rotary film evaporation is a
very rapid means of removing large quantities of solvent (1000 ml to 10 ml),
and is probably the most widely used technique.  Vortex evaporation is a very
convenient method for reducing solvent volumes from 10 ml to 1 ml or lass.
Using this method, the solvent is transferred to a centrifuge tube, which is
placed in the vortex unit.  During the vortex mixing, the solvent is removed
while heating (20-60 C) under vacuum (30 inches of H_0).  Nitrogen stream
evaporation is used to concentrate the solvent to several hundred microliters
or to dryness.  Rates of evaporation are controlled by the nitrogen flow rate
and the solvent temperature.  Kuderna-Danish concentration has not received
much attention for the concentration of particulate extracts primarily due to
the overall time of concentration.
                              67

-------
     Concentration steps are generally a major source of compound loss through
 adsorption, evaporation, incomplete transfer between containers, and/or reaction
 with other constituents.  Extreme care must be taken to insure minimal sample
 loss and maintaining sample integrity.
     The total organic mass can be determined by total removal of the solvent
 under vacuum using a rotary-film evaporator, transfer of the residue to a tared
 vessel, and drying under nitrogen to constant weight.  An alternative method is
 to reduce the volume of extract to a known value and remove a small aliquot for
 a residue mass measurement.  The total extracted mass can be calculated by
 multiplying the mass concentration in the aliquot by the total extract volume.
 Using this alternative procedure, it is not necessary to bring the entire
 extract to dryness and may help preserve the chemical and biological integrity
 of the extract.
     Fractionation.  Fractionation of the particulate extract is usually required
 before further chemical and biological characterization can be achieved.  A
 variety of partitioning methods and chromatographic techniques have been used
 to effect this fractionation on the basis of chemical functionality. '
     Acids and bases can be removed from the neutrals by liquid-liquid partition-
 ing sequenctally with aqueous solutions at low and high pH.  The neutral organic
 compounds can be further separated using solvent partitioning, and/or column
 chromatography.  Pellizzari has developed a solvent partitioning scheme to
                                                                32
 separate the neutrals into non-polar, polar, and PAH fractions.    Column
 techniques using Sephadex LH-20,  silica gel, and alumina have been developed by
 numerous workers to separate the neutrals based on compound classes.  '
 Jewell has used a combination of ion-exchange coordination and adsorption
 chromatography (known as SARA technique) for the separation of acids, bases,
 neutral nitrogen-containing compounds, saturated hydrocarbons and aromatic
 hydrocarbons.
     The application of high performance liquid chromatography (HPLC) as a means
 of rapid fractionation of particulate extracts has been reported by Huisingh,
       4
1 et al.   This method has recently been modified by Schuetzle and co-workers to
                                2
 produce six specific fractions.   A complete fractionation is possible in
 approximately 40 minutes.  In an interlaboratory validation study of HPLC
 fractionation, recovery of total mass is reported to be near 100 percent.
 However, biological activity of the fractions have been shown to be highly
 dependent on the influence of fractions containing very small mass.  Although
 HPLC can be used to rapidly fractionate particulate extracts, it is limited in
                                68

-------
the quantity of mass O15 mg) which can be fractionated during a single analysis.
Thus many separations may have to be carried out to obtain the necessary
quantity of material for chemical and biological characterization.  There are
several problem areas that are associated with the procedures for fractionation.
The various steps involved with fractionation can certainly increase the
preparation time, which is generally not desirable when large numbers of samples
are to be analyzed.  Contamination can readily occur during fractionation and
can lead to discrete artifacts and interferences.  Incomplete extraction from
the fraction procedure can be a major source of sample loss.  Liquid-liquid
partitioning may cause the formation of tars, insoluble material, or stable
emulsions, which result in inefficient separation.
    To remove sources of contamination, all materials must be routinely demon-
strated to be free of interferences under conditions of the analysis by running
parallel laboratory blanks.  The use of high purity solvents helps minimize
interferences.  Solvents should be always checked for purity.  Glassware must
be scruptuously cleaned usually by detergent washing, distilled water rinse,
solvent rinse and heating at 450 C for several hours.
    Procedure for Fractionation of a Particulate Extract into Specific
Compound Classes.  A fractionation method to separate chemical classes from
extracts of particulate samples was developed by Battelle in 1978.  Several
refinements in the original procedure have been made, and considerable experi-
ence in its use has been obtained on a wide variety of particulate sample types.
There are two advantages associated with this procedure which are pertinent to
the analysis of the organic material extracted from diesel particulate samples:
          •  The method is sensitive enough to fractionate small
             quantities of total extracted mass.  Particulate
             extracts containing 1 to 4 mg of total extracted
             mass have been sufficient for compound class
             separation and chemical analysis.
          •  The method can be conveniently scaled up for
             fractionation of much larger samples when biological
             as well as chemical characterization is required.
             Particulate extracts containing several grams of
             extracted mass have been successfully fractionated.
    A total quantity of extracted mass of at least 50 mg is required for both
chemical characterization and biological screening.  The following description
of the procedure is based on this minimum quantity.  For larger or smaller
samples, appropriate adjustments to the procedure can be made.
                               69

-------
    A schematic of the overall procedure  is  shown  in Figure 2.   The particulate
sample is first extracted and two aliquots are  removed;  one for a residue mass
measurement and one for a bioassay measurement.  These measurements are used
to calculate the total quantity of mass and  biological activity in the extract.
The remaining extract is separated into six  compound class fractions.  Two
aliquots are removed from each fraction for  determination of their residue mass
and biological activity.  A material and  bioassay  balance is calculated as
percent recovery by summation of the fraction values,  dividing  by the original
value and multiplying by 100 percent.
                                   Compound*
                               FraethHMtlofl 
-------
    The Fractionation Procedure. Acids and bases are first separated by liquid-
liquid partitioning.  The neutrals are further partitioned by  silica gel  column
chromatography.  A schematic representation of the entire fractionation proce-
dure is shown in Figure 3, and the generated fractions are listed in Table  1.
For purposes of discussion, the fractions are referred to numerically as  Ifi,
#2, etc.
                TABLE 1
                COMPOUND CLASSES GENERATED" BY FR&eT-!6WAT-ING SCHEME

                        Fraction                 Fraction #

                 Bases                              1
                 Acids and Phenols                  2
                 Aliphatic HyUiuc             3-
                 Aromatic Hydrocarbons              4
                 Moderately Polar Neutrals          5
                 Highly Polar Neutrals              6

    Separation of Bases.  In the fractionation scheme, the bases such as  amines
are first separated from the extract by liquid-liquid partition with 100  ml of
5 percent sulfuric acid solution three times.  The aqueous phases are combined
and the bases then back extracted into methylene chloride after adjustment of
the pH of the solution to 12-13.  The volume of the methylene  chloride solution
containing the bases is then reduced to exactly 5 ml by rotary and vortex
evaporation.  An aliquot of 10-100 yl (depending on the quantity of mass present)
is removed to measure the organic mass in this fraction.  The  bases constitute
Fraction #1.  After an aliquot of 50-100 ul is removed for chemical analysis,
the solvent is exchanged with DMSO and the fraction submitted  for Ames bioassay.
    Separation of Acids.  The acidic (carboxylic) and phenolic compounds  are
extracted from the organic solution with 100 ml of 5 percent sodium hydroxide
solution four times.  The aqueous phases are combined and the  acids then  back
extracted with methylene chloride after addition of Na SO  (for "salting  out")
and adjustment of the solution's pH to about 1.  An aliquot of 100 ul is
removed to measure the organic mass in this fraction.  The acids and phenols
constitute Fraction //2.  Ames testing and chemical analysis is carried out on
the acids by the same procedure as used with the bases.
                              71

-------
    •n
    H-
   00
a  ID
H1 0-
    c
H-  O
O  H
(0
rt  rt
ro  n
    H-
    §

    s
    ex
    o
    1-1
   (W
    01
Add 4O\ KoH i« pH 13

trgtwl b^md pun lion WM*I

CH4TI (100ml  lun,..)

Or» with •nhtdioui N* SO.
AddOMSOto* ml

tv*poiMr 10 1 m»

Add DMSO 10 Am)
                                                                                               Ncutrcls to Organic Solution
                                                                                                              AM h«.«M M 10 Ml

                                                                                                              E«*p Solution
                                                                                                                                       »IUI* Otgcntc M«*» kn

                                                                                                                                       tBClion 2
                                                                                          Measure Organic Mast in Ftacttoni 3 6
                                                                                                               Ev*po(«l«Mriiff»c|Mmli> I nri
                                                                                                                                                                                        flwaporaii u I ml

                                                                                                                                                                                        AM DMSO la • ml

                                                                                                                                                                                        E>*p«iM«l<> 1 ml

                                                                                                                                                                                        Add DMSO la • ml
                                                                                                                                                                                      loM*ev on FrKikm 2
                                                                                       Amet MuligantU Bk>ai«ay on Fraction* 3-B

-------
    Separation of Neutrals.  The organic solution containing the neutral com-
pounds is further partitioned into four fractions by open column chromatography
on 5 percent H_0-deactivated silica gel.  The silica gel columns (2.5 cm i.d. x
15 cm long) are packed with 20 g of the silica in a hexane slurry and the gel
retained with a glass frit.  Columns are prepared for each organic solution  to
be partitioned.  An additional column is also prepared to check the accuracy
of the silica gel deactivation before partitioning the neutral organic solution.
This is done by measuring the volume of hexane required to elute 500 yg of
anthracene.  The migration of the anthracene is monitored by a 366 run UV lamp,
and the volume of hexane is measured during the migration.  When the silica  gel
is deactivated 5 percent, anthracene starts to elute from the column after the
addition of 140 + 10 ml of hexane.  For 3 percent and 7 percent deactivation,
the volume of hexane required is 270 + 12 ml and 115 + 8 ml, respectively.
    Upon assurance that the silica gel is 5 percent H.O deactivated, the neutral
organic compounds are further fractionated.  Separation of the neutrals into
compound classes using silica gel is carried out by increasing the polarity of
the eluting solvent.  Therefore, the separation scheme must begin with the most
nonpolar eluent, and the solvent must be exchanged with hexane before trans-
ferring the solution to the silica gel column.  This is accomplished by
reducing the volume to 5 ml, adding 45 ml of hexane and reducing again to 5 ml.
Repeating this process will result in a complete solvent exchange.   The solu-
tion is transferred to the silica column using a Pasteur pipette.  Polar
neutral compounds will be present in the original vessel as insoluble material
after the solvent exchange into hexane.  As the column elution solvent is
changed it should first be introduced into the vessel to dissolve the polar
material and then transferred to the silica column.  Four elution solvents are
used in the following sequence:  60 ml hexane, 100 ml hexane/benzene (1:1),
100 ml methylene chloride, and finally, 200 ml methanol.  The eluent is
collected individually for each solvent and corresponds to fraction numbers  3,
4, 5 and 6.  These fractions are evaporated to exactly 50 ml and 100 yl removed
to measure the organic mass.  After measurement, 1-4 ml is removed from each
fraction for chemical analysis and the remaining solution submitted for Ames
bioassay after solvent exchange with DMSO.  Saturated and unsaturated hydro-
carbons are present in the aliphatic compound class.  The aromatic fraction
will contain the polynuclear aromatic hydrocarbons, alkyl substituted benzenes,
nitrogen and sulphur heterocyclic hydrocarbons and mono nitro derivatives of
2, 3 and 4 ring PAH.  Chlorinated hydrocarbons as well as silicones will also
be present in this fraction.  The moderately polar neutral fraction will
                               73

-------
contain mono nltro substituted PAH greater than 4 rings as well  as  all dinitro
and trinitro PAH.  Mono oxygenated PAH of 2, 3 and 4  rings such  as  fluorenone
will be present in this fraction.  The highly polar neutral  fraction will
contain substituted PAH with more than one type of functionality (i.e.,  nitro
and keto groups such as in nitrofluorenones) and poly oxygenated PAH.   Any
plasticizers, such as phthalates and sebacates in the analytical system will
also be present in this fraction.
    Concentration of Fraction Aliquots for Chemical Analysis.  The  fraction
aliquot is transferred to a 5 ml pyrex  Chromaflex tube, and capped with a
piece of aluminum foil.  The tube is placed in a water bath  so that the  water
line is even with the solvent.  The temperature of the bath  is maintained at
10 C above the boiling point of the solvent.  A nitrogen stream  (200-500 yl/miij
is introduced to the Chromaflex tube using a stainless steel capillary,
positioned 0.5 cm above the reflux line of the solvent.  This allows a portion
of the solvent to be removed slowly while the remaining solvent  condenses and
continuously rinses the walls.  As the solvent is removed, the position  of the
Chromaflex tube is adjusted to maintain the solvent line even with  the level
of water bath.  The solution is concentrated to 100 ul, which takes about 2-4
hours, depending on the volatility of the solvent.  Concentrated solutions
are stored at -70 C until analysis to minimize sample degradation.
    Ames Salmonella Mutagenesis Bioassay.  The assay  is conducted by adding a
0.1 ml aliquot of an overnight broth culture of tester strain TA-98 (without
metabolic activation, or S-9) to 2 ml of molten top agar supplemented  with
biotin and a trace of histidine.  Subsequently, dose  levels of the  test
extracts and compound class fractions are added to the appropriate  tubes.
Dose levels used are 25, 50 and 100 )jg.  The contents of these tubes are mixed
thoroughly and poured over the surface of selective agar plates.  Following
solidification, the plates are incubated for 72 hours at 37 C and scored for
the number of colonies growing on each plate.
    The number of revertant colonies on the duplicate mutagenesis plates are
averaged for each concentration of test material and  control.  The  control
value or number of spontaneous revertants are subtracted from the counts of
the test dosages.  The resulting number of revertants are then divided by the
dosage to ascertain the activity of the extracts or fractions per unit weight.

RESULTS AND DISCUSSION
    The application of the chemical and biological characterization of organics
associated with particulates has been demonstrated for particulate  filter
                            74

-------
samples from two light-duty passenger car diesel engines  (LDDI and LDDII) and
one heavy-duty truck type diesel engine (HDD).  All particulate filter  samples
were collected on Pallflex T60A20 PTFE impregnated glass  fiber filters  using an
exhaust dilution tunnel.  Particulate filter samples were collected from the
test engines as they were operated under the following conditions:
                     LDDI:   steady-state, medium speed/medium load
                     LDDII:  highway fuel economy test cycle
                     HDD:    steady-state, rated speed/rated load
Table 2 summarizes the extraction and fractionation results from these  three
engines.  A discussion of these results for each engine is presented in the
following section.


TABLE 2

                           EXTRACTED ORGANIC MASS DATA


                                   FRACTION MASS, mg

ENGINE
TYPE
LDDI
LDDII
HOD
TOTAL ORGANIC
EXTH ACTABLE
MASS, mg
61.90
142.03
63.24


BASE
14.86
0.20
1.48


ACID
1.S6
10.11
9.48


ALIPHATIC
17.43
69.49
8.23


AROMATIC
3.46
IB. 12
5.61
MODERATELY
POLAR
NEUTRAL
4.11
11.38
12.07
HIGHLY
POLAR
NEUTRAL
14.60
27.21
23.40
TOTAL
FRACTION
MASS. nt«
(PERCENT
RECOVERED!
56.11 1911
133.E1 (94)
60.28 136)
    LDDI.   For the LDDI engine,  a total  of 61.90  mg of  total  organic  mass  was
 extracted from the particulate  filter samples.   The base,  aliphatic,  and  high-
 ly polar  neutral  fractions  contained significant quantities  of  material,  the
 sum of  which accounted for  approximately  75  percent of the total  extracted mass.
 In general,  the aliphatic and highly polar neutral  fractions contained the
 majority  of  mass  in extract,  however, the quantity of  mass in the base  fraction
 was still unusually large.   A possible  reason for this large quantity may be
 due to  the presence of amines as a result of the high  oil  consumption of  this
 specific  engine.   The aromatic  and moderately polar neutral  fractions contained
 about 6 and  7 percent of  the extracted  mass  which are  considered  average.
 Acids accounted for 3 percent of the total mass  and is considered low.   Typical
 quantities in the acid fraction are 8 to  12  percent of the total  mass.   The
 recovery  of  total mass through  the fractionation procedure was  determined to
                             75

-------
 be  91 percent  and  is considered  satisfactory.
    1DDII.   The  total organic mass  extracted from the particulate filter samples
 was 142.03 mg.   Again,  the  aliphatic  and  highly polar neutral fractions con-
 tained  significant quantities of mass and accounted for 48 and 20 percent of
 the total extract.  The base fraction contained less than one percent of the
 total mass,  and  the aromatic and moderately polar neutral fractions represent
 10  and  7  percent of the total mass.   Acids accounted for approximately 7 percent
 of  the  total mass.  Recovery of  the extracted mass through the fractionation
 procedure was  94 percent.   In general,  the distribution of mass within the
 fractions is typical for particulate  extracts of light-duty diesel engines.
    HDD.   The total organic  extract  was  63.24 mg, which represented 3.1 percent
 of  the  total collected  particulate  mass.   This low percentage of total organics
 as  well as the distribution of mass within the six compound class fractions  is
 typical for  heavy-duty  diesel engines operated at rated speed and rated load.
 The aliphatic  fraction  contained only 13  percent of the total extract while
 the polar and  acid fractions account  for  the majority of the mass.  Both the
 base and  aromatic  fractions contained average quantities of mass.   Recovery  of
 the extract  through the fractionation procedure was 95 percent.
    Ames Mutagenesis Bioassay Results.  Table 3 summarizes the bioassay data
 using tester strain TA-98 without metabolic activation of the extracts and
 their individual fractions. The data is  presented as the total number of
 revertants within  each  extract and  fraction.  These values were determined by
 multiplying  the  average mutagenic response at the 50 u g dose level by the
 total mass in  the  extract or fraction.
    Both light-duty extracts contained greater than 300,000 total revertants
 whereas the  HDD  extract only contained  about 50,000.  A bioassay balance was
 calculated by  comparing the sum  of  the  revertants within the fractions to the
 total revertants found  in the extract.  The recovery of biological activity
 was as  follows:  LDDI,  72 percent;  LDDII, 95 percent; and HDD, 97 percent.
    Table 4  presents the bioassay data in terms  of  specific  activity  as
revertants per milligram (rev/mg) of  mass.   The  LDDI and LDDII extracts  had
specific activities of  5,049 and 2,290 rev/mg.   Specific activity  for the HDD
was much lower at 819 rev/mg.   The moderately polar neutral  fraction  of  all
three extracts is very active.   For the two  light  duty  diesel  extracts,  the
specific activity of this fraction dominates  the activity of  the other
fractions as well as that of the total extract.   Activity of  the moderately
polar neutral fraction  in the HDD extract  is much  less  than  that of the  LDD
                              76

-------
fractions, however, it is still significant  in relation to the other IIDD
fractions.
TABLE 3
        AMES MUTAQENESIS BIOASSAY (TOTAL REVERTANTS) OF PARTICULATE EXTRACTS AND
                   COMPOUND CLASS FRACTIONS USING TESTER STRAIN TA-S8
                                   (NONACTIVATED)'



ENGINE
TYPE
LOOI
LDDII
HDD



TOTAL
EXTRACT
312.592
329,1 SO
81 .824


MODERATELY
POLAR
BASE ACID ALIPHATIC AROMATIC NEUTRAL
2.867 3.603 1.830 20.886 101.43*
132 39,866 - 22.227 220.020
1.038 14.424 — 777 16.647


HIGHLY
POLAR
NEUTRAL
99.360
2s.es*
18.663
TOTAL
FRACTION
ACTIVITY
(PERCENT
OF TOTAL)
226.7*0 (721
308. 12k (961
50.340 (97)
  •AVERAGE MUTAQENIC RESPONSE AT 60 ug DOSE LEVEL X MASS OF EXTRACT OR FRACTION
TABLE  4
    AMES MUTAQENESIS BIOASSAY (REVERTANTS/mg) OF PARTICULATE EXTRACTS
          AND COMPOUND CLASS FRACTIONS USING TESTER STRAIN TA-98
                                (NONACTIVATED)
ENGINE
TYPE
LOOI
LOO II
HOO
TOTAL
EXTRACT
5.049
2.290
819
BASE
1.549
660
ess
ACID
5.019
3.946
1.521
ALIPHATIC AROMATIC
1O4 5.969
— 1 .472
— 139
MODERATELY
POLAR
NEUTRAL
24.679
19,300
1.288
HIGHLY
POLAR
NEUTRAL
6.600
962
793
                               77

-------
CONCLUSIONS
   Fractionation into compound classes using  the  described procedure resulted in
a satisfactory material balance for a variety of  particulate extracts.  Recovery

of the total extracted material was 91 percent for  the  LDDI extract; 94 percent
for the LDDII extract; and 95 percent for the HDD extract.   A satisfactory

biological activity balance was also demonstrated for these samples.  Recovery

of the total revertants was 72 percent for the LDDI extract;  95 percent for the

LDDII extract; and 97 percent for the HDD extract.
   Two important conclusions can be drawn from these data:   (1) the mass and

mutagenicity, using TA-98 (-S9), of the extract is  approximately equivalent to
the sum of the mass and the mutagenicity of the fractions,  and (2)  the fraction-

ation procedure does not significantly influence  the mutagenicity of the

extract.  The data presented in this paper demonstrates  the utility of this
fractionation procedure for the general chemical  characterization of organics

associated with diesel particulates.  Furthermore,  the  ability to conduct a

general biological screening for direct acting mutagenicity within the parti-
culate extracts is also demonstrated.  This procedure separates the particulate

extracts into less complex fractions which can be analyzed by a variety of

analytical techniques for identification of specific compounds.


REFERENCES

1. U.S. EPA.  Precautionary Notice  on the Handling  of  Exhaust Products From
   Diesel Engines.  Issued under the signature of S. Gage, Office of Research
   and Development, Washington, D.C.  November 4, 1977
2. Schuetzle, D., Lee, F. S.-C. , Prater, T. J., Tejada,  S. (1981),  International
   Journal of Environmental Analytical Chemistry, 9, 93-144.
3. Claxton, L.  (1979).  Mutagenic and Carcinogenic  Potency of Diesel and
   Related Environmental Emissions:  Salmonella Bioassay.   Health Effects of
   Diesel Engine Emissions, Proceedings of an International Symposium, Pepelko,
   W. E., Danner, R. M. and Clarke, N. A., ed., U.S. EPA-600/9-80-057,
   Cincinnati, Ohio.
4. Huisingh, J., Bradow, R., Jungers, R., Claxton,  L., Zweidinger, R., Tejada,
   S., Bumgarner, J. , Duffield, F., Waters, M., Simmon,  V., Hare, C.,
   Rodriguez, C., and Snow, L.  (1978).  Application of  Short-Term Bioassays
   in the Fractionation and Analysis of Complex Environmental Mixtures.
   Waters, M., Nesnow, S., Huisingh, J., Sandhu,  S., and Claxton, L., eds.
   U.S. EPA-600/9-78-027, Research  Triangle Park, North Carolina.
5. Liber, J., Andon, B., Kites, R., and Thilly, W.  (1979).  Diesel Soot:
   Mutation Measurements in Bacterial and Human Cells.   Health Effects of Diesel
   Engine Emissions, Proceedings of an International Symposium, Pepelko, W. E.,
   Danner, R. M. and Clarke, N. A., eds., U.S. EPA-600/9-80-057, Cincinnati,
   Ohio.
                              78

-------
 6.  Mitchell,  A.,  Evans, E.,  Jotz, M. , Riccio, K. , Mortelmans,  K. ,  and  Simmon,  V.
    (1979),  Mutagenic and Carcinogenic Potency of Extractsof Diesel and Related
    Environmental Emissions:   In vitro Mutagenesis and DNA Damage.   Health
    Effects  of Diesel Engine Emissions, Proceedings-of an International  Symposium,
    Pepelko, W. E.,  Danner, R. M. and Clarke, N. A., eds. , U.S.  EPA-600/ 9-80-051,
    Cincinnati, Ohio.
 7.  Siak,  J.,  Chan,  T.,  and Lee, P.  (1979).  Diesel Particulate  Extracts  in
    Bacterial Test Systems.  Health  Effects of Diesel Engine Emissions,  Proceed-
    ings of  an International Symposium, Pepelko, W. E., Danner,  R.  M. and
    Clarke,  N. A., eds., U.S. EPA-600/9-80-057, Cincinnati, Ohio.
 8.  Wei, E., Wang, Y., and Rappaport, S. (1980).  Journal of the Air Pollution
    Control Association, 30,  267-271.
 9.  Huisingh,  J., Bradow, R., Jungers, R., Harris, B., Zweidinger,  R.,  Gushing,
    K. ,  Gill, B., and Albert, R.   (1979).  Mutagenic and  Carcinogenic Potency
    of Extracts of Diesel and Related Environmental Emissions:   Study Design,
    Sample Generation, Collection, and Preparation.  Health Effects of  Diesel
    Engine Emissions, Proceedings of an International Symposium, Pepelko, W.  E.,
    Danner,  R. M. and Clarke, N. A., eds., U.S. EPA-600/9-80-057, Cincinnati,
    Ohio.
10.  Petersen, B. A., Chuang,  C. C.,  Margard, W. L. , and Trayser, D.  A.  (1981).
    J. Air Pollution Control Assoc., 81, 1-15.
11.  Lofroth, G. (1979).   Salmonella/Microsome Mutagenicity Assays of Exhaust
    from Diesel and Gasoline Powered Motor Vehicles.  Health Effects of  Diesel
    Engine Emissions, Proceedings of an International Symposium, Pepelko, W.  E. ,
    Danner,  R. M. and Clarke, N. A., eds., U.S. EPA-600/9-80-057, Cincinnati,
    Ohio.
12.  Jager, J.  (1978) J.  Chromatography, 92, 575.
13.  Pitts, J. N. ,  Van Cauwenberghe,  K. A., Grosjean, D.,  Schmid, J.  P.,  Fitz,
    D. R., Belser, W. L., Knudson, G. B., and Hynds, P. M.  (1978).  Science,
    202, 515.
14.  Zweidinger, R. B., Tejada, S. B., Dropkin, D., Huisingh, J., and Claxton, L.
    (1979).   Characterization of Extractable Organics in  Diesel  Exhaust
    Participates, U.S. EPA Laboratory Report.
15.  Wang,  Y-Y., Talcott, R. E., Sawyer, R. F., Rappaport, S. M., and Wei, E.
    (1978)  Mutagens in Automobile Exhaust.  Symposium on Application of  Short-
    Term Bioassays in the Fraction and Analysis of Complex Environmental
    Mixtures, Williamsburg, Virginia.
16.  Santodonato, J., Basu, D., and Howard, P. (1978) Health Fffects  with  Diese]
    Exhaust  Emissions, Literature Review and Evaluation,  EPA-60011-78-063
17.  McGarth, J. J.,  Schreck,  R. M. and Siak, J. S. (1978) Mutagenic Screening
    of Diesel Particulate Material.  Proceedings of the 71st Annual Meeting
    of APCA, Houston, Texas.
18.  Grimmer, G. (1977) Analysis of Automobile Exhaust Condensates.   Air  Pollution
    and Cancer in Man, Mohr,  V., Schmahl, D., and Tomatis, I.,  eds.
19.  Kaden, D. A., Thilly, W.  G.  (1979) Genetic Toxicology of Kerosene Soot.
    Cancer Research, 39, 492.
20.  National Academy of Science  (1972) Particulate Polycyclic Organic Matter.
    Committee on Biological Effects  of Atmosphere Pollutants, Washington, D.C.
21.  Lee, F.S.C., Prater, T. J. and Ferris, F. (1979) PAH  Emissions  from a
    Stratified-Charge Vehicle With and Without Oxidation:  Sampling and Analysis
    Evaluation in Polynuclear Aromatic Hydrocarbons.  Jones, P.  W.  and  Seber,P.,
    eds.,  Ann Arbor Science Publishers, Inc., Ann Arbor,  MI, 83-110.
22.  Grimmer, G., Bahnke, H.,  and Glaser, A.  (1977) Zbl. Bakt. Hyg.  I. Abt.
    Orig.  B., 164, 218.
                              79

-------
23. Erickson, M. D., Newton, D. L., Pellizzari, E. D.,  Tomer,  K.  B.,  and
    Dropkin, D. (1979) Identification of Alkyl-9-Fluorenones  in Diesel Exhaust
    Particulate, J. Chromatographlc Science, 17, 450.
24. Nishioka, H. G., Petersen, B. A., Lewtas, J.  (1981)  EPA Diesel Emissions
    Symposium, Research Triangle Park, NC.
25. Petersen, B. A., Chuang, C. C., Kinzer, G. W., Hayes,  T.  L.,  Meehan, P.  W.,
    and Trayser, D. A. (1980) Diesel Engine Emissions of Particulates and
    Associated Organic Matter, NTIS No. PB-80-221963.
26. Brooks, A., Wolff, R., Royer, R., Clark, C., Sanchez,  A.,  and McClellan,
    R. (1979)  Biological Availability of Mutagenic Chemicals  Associated with
    Diesel Exhaust Particles.  Health Effects of Diesel  Engine Emissions,
    Proceedings of an International Symposium, Pepelko,  W.  E. ,  Danner,  R. M.,
    and Clarke, N. A., eds., U.S. EPA-600/9-80-057, Cincinnati,  Ohio.
27. Swarin, S. J., and Williams, R. L. (1979)  LC Determination of BaP in
    Diesel Exhaust Particulate:  Verification of the Collection and Analytical
    Methods, presented at the Fourth International Symposium on PAH,  Columbus,
    Ohio, GM Research Paper GMR-3127, ENV #69.
28. Choudhury, D., and Doudney. C. (1979)  Isolation of  Mutagenic Fractions
    of Diesel Exhaust Particulates as an Approach to Identification of the
    Major Constituents.  Health Effects of Diesel Engine Emissions, Proceed-
    ings of an International Symposium, Pepelko, W. E.,  Danner,  R.  M.,  and
    Clarke, N. A., eds.,  U.S. EPA-600/9-80-057, Cincinnati, Ohio.
29. Hare,  C., and Baines, T. (1979).  SAE Technical Paper  Series,  Publication
    No. 790424.
30. Lyons, M. (1962) Comparison of Aromatic Polycyclic Hydrocarbons from Gaso-
    line Engine and Diesel Engine Exhaust, General Atmospheric  Dust,  and
    Cigarette Smoke Condensate.  Paper presented at the  Symposium on  the
    Analysis of Carcinogenic Air Pollutants.  National Cancer  Institute
    Monograph No.  9, 193-199.
31. Risby, T., and Sigsby, J. (1980)  Exhaust Emissions  From a  Diesel  Engine.
    Annual Report  on U.S. EPA Grant No. R-806558.
32. Rodriguez, T., Fisher, J., and Johnson, D. (1979) Characterization of
    Organic Constituents  in Diesel Exhaust Particulates.  Health  Effects of
    Diesel Engine  Emissions, Proceedings of an International Symposium,
    Pepelko, W. E., Danner, R. M. and Clarke, N. A., eds.,  U.S.  EPA-600/9-80-057,
    Cincinnati, Ohio.
33. Jewel, D. M.,  Weber,  J. H., Bunger, J. W., Plancher, H., Latham,  D.  R.  (1972)
    Anal. Chem., 44, 1391-1395.
                              80

-------
THE UTILITY OF BACTERIAL MUTAGENESIS TESTING IN THE CHARACTERIZATION
OF MOBILE SOURCE EMISSIONS:  A REVIEW
Larry D. Claxton
Genetic Toxicology Division, Health Effects Research Laboratory,
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina
INTRODUCTION
   In 1978 Huisingh et al.  reported that organic fractions chemically extracted
from the exhaust particles of diesel vehicles were mutagenic in the
Salmonella typhimuriinn plate incorporation  (Ames) test.  This report and an EPA
cautionary notice  for laboratory workers exposed to diesel exhaust sparked
expanded efforts in industry and government to understand whether or not these
bacterial mutagens also presented any potential health risks.  This question was
especially important since the automotive industry was planning expanded light-
duty dieselization for fuel economy.  Although this original report was quite
extensive — examining multiple vehicles, multiple fuels, chemical fractions of
exhaust organics, etc. — some obvious questions had not been answered or even
approached.  Researchers in government and industry, striving to serve public
interest, wanted to answer several critical questions as rapidly as possible:
   •  Could these bacterial mutagens be artifacts of the testing process?
   •  Could these mutagens extracted from carbonaceous particles by the use
      of chemical solvents also be extracted by physiological fluids, enzyme
      systems, or cells within the human body?
   •  Would interaction with an ambient air situation alter the activity of
      these genotoxic materials?
   •  Could these bacterial mutagens cause hereditary or carcinogenic effects?
   •  What portion of the population would be exposed to these mutagens and to
      what degree would they be exposed?
   •  And ultimately, what hazard and risk would further dieselization present
      to man and to future generations?
   Obviously, bacterial mutagenicity testing cannot answer some of these
questions.  Bacterial tests cannot define a substance as a human carcinogen
and cannot be directly used for quantitative risk estimation.  These microbial
tests can be used, however, to aid in overcoming difficulties encountered while
trying to understand the potential genotoxic effects of mobile source emissions.
Table 1 lists some of the stages at which difficult research questions are
                              81

-------
TABLE 1
DIFFICULTIES ENCOUNTERED IN THE INTERPRETATION OF MOBILE
SOURCE GENOTOXICITY STUDIES
Stage                                       Types of Difficulty

Generation                    Characterization of the source;
                              Repeatability of testing procedure;
                              Vehicle-to-vehicle variation.
Collection                    Artifact formation (NOX, organic interaction,
                              ozone);
                              Control  versus ambient conditions.
Extraction                    Physiological versus organic;
                              Selective loss and recovery of compounds;
                              Artifact formation.
Fractionation                 Recovery efficiencies;
                              Complexity.
Bioassay                      Relevance;
                              Reproducibility;
                              Endogenous activation;
                              Bioassay versus human metabolism.
Statistical                   Collection/recovery efficiencies;
                              Data transformations;
                              Data summary methods;
                              Statistical/correlation methods.
generated and the problems to which these questions are related.  Bacterial
testing has provided and will continue to provide many of the needed mechanisms
for recognizing, defining, and answering these difficult problems.  Bacterial
tests have proven valuable because they are rapid, relatively inexpensive indi-
cators of genotoxic activity that can be used in a unified effort with other
research methods.  For the cost of one life-time animal chronic-toxicity study
of one exhaust sample, more than 100 different vehicles can be compared in a
simple bacterial screen.  When exhaust organics are fractionated by physical/
chemical means, bacterial testing can follow the distribution of genotoxic
activity before compound identification is complete.  When road-way and smog
chamber samples are collected, microbial tests allow for comparison of genotoxic
activity with that of dilution tunnel samples.  The purpose of this paper,
therefore, is to recognize and document the role that bacterial mutation tests
                               82

-------
have played in characterizing mobile source emissions for genotoxic activity.
This presentation will be divided into four areas:  1) generalized observations,
2) assessing factors that modify the mutagenicity of mobile source emissions,
3) the comparison of various mobile source emissions, and 4) the effect of
physiological fluids and enzyme systems.  This review will assume that the
reader has a general knowledge of mutagenicity testing and mobile source
research.

GENERALIZED OBSERVATIONS
   Most of the bacterial mutagenesis research has been conducted in the same
manner as that originally reported by Huisingh et al.   They, and most other
investigators, have used the Salmonella typhimurium plate incorporation assay
                           4
as described by Ames et al.  as the primary test protocol.  Some investigators
have used all five strains as recommended by Ames.  However, many investigators
have used only one or two strains (generally TA98 and/or TA100),  primarily for
two reasons.  First, the sample amounts available have been relatively limited,
and secondly, TA98 and TA100 have been the strains most responsive to organics
extracted from mobile source emission particles.  TA1535, which responds by
base-pair substitution, has been negative or only marginally positive with total
extracts; however, since fractionation studies have not generally used this
strain, mutagens that cause base-pair substitution may be overlooked.  Perhaps
even more interesting are the observations concerning strain TA1538.  In
contrast to TA98, which exhibits the same or a decreased response with S9 addi-
tion, TA1538 exhibits an increased response upon the addition of an exogenous
activating system.  This response means that TA1538 (although not providing as
many revertants per plate) is detecting indirect-acting mutagens that are not
readily recognized with strain TA98.  Again, researchers may be failing to
identify another class of mutagens.  Nitroaromatic activity in diesel exhaust
was first recognized by the use of another group of bacterial tester strains —
                                     5—7                                      8
the nitroreductase-deficient strains.     Additional work by Rosenkranz et al.
                      g
and Mermelstein et al.  has provided and characterized additional nitroreductase-
deficient strains since this initial work.  Work with other bacterial strains and
systems has generally been neglected.  While some investigators '  '   have
demonstrated the usefulness of the 8-azaguanine forward mutation assay, most
others have not used this simple, more quantitative system.
   Many investigators have also recognized that exogenous activation systems
reduce the mutagenic response of diesel exhaust organics, with the exception
                              83

-------
of the TA1S38 results.  Few have emphasized, however, that  these  same exogenous
activation systems generally enhance the mutagenic response of  gasoline exhaust
organics.    This simple observation clearly demonstrates the presence of
different mutagenic compounds in exhaust organics from diesel and gasoline
engines.

ASSESSING FACTORS THAT MODIFY THE MUTAGENICITY OF MOBILE SOURCE EMISSIONS
Fuels and oils
   One of the first questions posed for diesel emissions work was:   "Is the
mutagenicity due to mutagens in the fuel or due to the combustion process?"
The diesel fuel used by Huisingh et al.  was negative when  tested directly in the
Salmonella bioassay.  Lebowitz   also reported diesel fuel  as being  negative.
    1 A 1C
Wang      reported that a diesel fuel, JP-4, and two types  of gasoline were
negative when tested with TA98.  Other investigators, however,  have  reported
various crude oils and some of their distillates to be positive in the Ames
     16—18         18 19
test.       Calkins  '   reported that some natural, syncrude,  and shale  oil
crudes, and some of their distillates are positive in the Ames  test.   In  each
                                                                  20
case, however, the naphtha distillate was negative.  Epler  et al.    and Guerin
et al.   demonstrated that coal-derived petroleum substitutes could  exhibit ten
times the bacterial mutagenic activity of a similar natural product  and that the
mutagenic response could occur over several orders of magnitude.   They also
reported that the activity of petroleum crudes was found mainly in the neutral
fraction, while significant activity was found in both the  neutral and basic
fractions of derived fuels.  Within the neutral fraction, Guerin  et  al.    found
that aromatic amines were the predominant mutagenic constituent.   When
Henderson et al.   separated diesel fuel into an aromatic and an  aliphatic
fraction, they found both fractions mutagenic using strain  TA100.  They also
noted that exposure to NO. dramatically increased the response  of both fractions.
This work was similar to the work of Pitts,   who exposed benzo(a)pyrene  to N02
and generated a direct-acting derivative.  The effect of fuel type upon the
mutagenicity of the emission organics was shown by Huisingh et  al.    They tested
the effect of seven different fuels in two different vehicles and found a wide
activity range in the mutagenicity for the emission organics.   The results from
         24
McClellan   are similar, and both studies tend to demonstrate that fuels  high in
aromatic content produce a more notable mutagenic response.  Crankcase oils have
                                                        14                "*5
also been examined for mutagenic activity.  Wang et al.,    Herman et al.,' and
Lofroth et al.  each stated that unused crankcase oils are  not  mutagenic,  but
that used crankcase oils from gasoline engines give a positive  response.   In
                              84

-------
addition, L<5froth states that 1) metabolic activation increased any mutagenic
response seen, 2) the response increased with vehicle mileage, and 3) this posi-
tive response was not seen with used oil recovered from a. diesel engine.  In no
case has it been demonstrated that the mutagenic emissions of diesel or gasoline
engines are due primarily to mutagens within the fuel.

Test cvcles
   Although most papers describe the test cycle(s) used when generating exhaust
samples, only a few researchers have published any direct comparison of test
cycles.  Vihen reporting data as revertants per microgram of organic material,
             26
Gabele et al.   found no great differences between six different test cycles.
Gibbs et al.   was able to examine five different cycles with six different
cars.  They made the following observations:
   •  expressing the data as revertants per gram of particulate gave "widely
      divergent" results,
   •  expressing the data as revertants per mile, "cycls-to-cycle" trends were
      more pronounced and reproducible,
   •  when ranking cycles by revertants per mile, activity decreased in the order
      FTP > CFDS * HFET > 50C, and
   •  a general reduction in revertants per mile was found as the mileage of the
      vehicle increased  (upon close examination of Gibbs et al.   data, it was
      noted that very low mileage cars {<4000 miles} demonstrated a very
      enhanced mutagenic response for all cycles except idle).
                     24
McClellan (July 1980)   examined four test cycles using a single automobile
and noted that in his study "cycles with lower speeds and more stops and starts
(NYCC and FTP) had higher mutagenic activity."

Collection methods and artifact generation
   The greatest potential problem in collection methodology is the generation of
artifacts, i.e., the generation of substances that do not exist in the natural
situation or the elimination of substances that would normally exist.  A number
of investigators   '      have demonstrated by various means that mutagenic
nitroaromatic compounds are contained in organic extracts of filter-collected
particles.  However, since diesel and gasoline also emit varying levels of NO
gases that pass across the filters and collected particles, these nitroaromatic
compounds may be artifacts.  There are basically three places where these nitro-
aromatics may be produced:  1) in the combustion process, 2) in the exhaust
process as organics interact and condense upon the particles, or 3) in the
                              85

-------
collection process as an artifact.  Pitts et  al.    were the first to show that
the passage of nitrogen oxides across a polynuclear aromatic hydrocarbon (PAH)
compound upon a filter could generate a nitroaromatic  (NO.-PAH)  that is direct-
                                                     22
acting in the Ames bacterial assay.  Henderson  et  al.    exposed 1-g samples of
fuel aromatics and fuel aliphatics to excess  NO2 at 25°C and generated direct-
acting mutagens for strain TA100.  The aromatic NO. fraction was the most active
and nitro-PAH compounds were identified in this fraction.   In some preliminary
experiments Bradow34 and Claxton3  passed artificial gas streams containing high
levels of N0_ across filters with diesel particles and  observed increased muta-
                                                         2930
genie activity of the extracted organics.  Gibson  et al.   '    re-exposed filter-
collected diesel particles to the gas-phase portion of  similar diesel emissions
and found increased levels of 1-nitropyrene,  nitrobenzo(a)pyrene,  and mutagenic
activity.  Although the extent and the relevance of this artifact problem has
not been fully resolved, bacterial testing has  paved the way in identifying and
providing the methods for examination of the  problem.

Extraction and chemical methods
   The mutagenic response of different chemical fractions  from the organic
emissions of a diesel engine was initially done with an organic  extract from
emission particles of two heavy-duty engines.   In this study,  the two most
active fractions were eluted from a silica gel  column and  were designated as the
transitional and oxygenate fractions.  Choudhury and Doudney   fractionated
organic emissions into three fractions and noted that 1)  the acid  fraction  had
the highest specific activity and showed direct-acting  mutagenicity,  2)  the basic
fraction was enhanced by the addition of S9,  and 3)  the neutral  fraction
accounted for 94% of the mass and was predominantly direct-acting.   Subsequently,
Choudhury subfractionated the neutral fraction  with adsorption chromatography
methods into seven subfractions.  Subtractions  3 to 7 were positive;  however,
the paraffinic subtraction was not active.  Upon examining emissions from both
a diesel and a gasoline vehicle, LSfroth  noted that the aromatic  and an oxyge-
nate fraction were the most mutagenic.  McClellan's work,37  using  a Fiat with
varying conditions, showed that upon Sephadex fractionation  three  of five frac-
tions were mutagenic to bacteria.  The classes  of  compounds  said to contribute
to the mutagenicity of these fractions were alkyl-substituted PAH  compounds and
                               38
oxygenated PAH.  Ohnishi et al.   examined the  fractionated  emissions of two
heavy-duty vehicles and one small diesel and  found each fraction tested as
                                 39
being positive.  Rappaport et al.   examined  16 liquid  chromatography fractions
of organic emissions from a Cummins turbodiesel engine  and postulated that
                              86

-------
pyrene-3,4-dicarboxylic acid anhydride and similar compounds accounted for a
                                                        40
sizeable portion of the mutagenic activity.  Siak et al.   summarized his
fractionation study with emission organics from a GM 5.7 L diesel engine by
stating that "more than 90% of the biological activity was accounted for in the
neutral-nonpolar II, neutral polar, weak and strong acid fractions."  Using
normal phase and reverse phase thin layer chromatography, Pederson and Siak
showed that some major mutagenic constituents were the mono-substituted nitro-
PAH compounds and other more polar compounds.  The examination of diesel exhaust
organics for active nitroarenss seems to have been spurred on by the earlier
reports of Claxton  and Lofroth,  who demonstrated the presence of nitroaroma-
tics by performing the bacterial bioassay with nitroreductase-deficient strains
and anaerobic conditions.  It is also interesting to note that very few of these
investigators used the indicator strains TA1535 and TA1538; therefore, some
mutagens that cause base-pair substitution and that need activation to be a
frameshift mutagen will be missed in these bioassay-directed fractionations.
Although the activity of these types of mutagens may have been missed in the
total organic extract, bioassay-directed fractionation has been established as
the primary means of identifying biologically active compounds in complex
organic mixtures from combustion sources.

Ambient conditions
   Ambient conditions can affect the condensation of organic compounds onto
particles, influence the interaction between organics, alter the organic species
emitted by a source, and provide for various other interactions.  Most diesel
and gasoline emission studies, however, have been done under non-ambient condi-
tions, i.e., by the use of dilution tunnels, tail pipe filters, etc.  The few
studies that have been done have taken different approaches.  Claxton and
      41
Barnes   studied ambient-like conditions in a controlled manner through the use
of the Calspan smog chamber.  In their studies, they found that the presence of
ozone in the chamber tended to reduce the mutagenic response of the organic
material collected; however, irradiation without other mitigating factors such
as ozone did not alter the mutagenic response.  A mutagenic sample also was
collected when (without diesel exhaust) propylene, SO , NO, and NO^ in an ambient
                                                                  ^ ^8
atmosphere were irradiated, so as to produce ozone.  Ohnishi et al."  examined
road-side particles collected within a highway tunnel.  When examined with TA100
and an activating system, they found 60 to 88 revertants/m  for particles
collected during daytime hours.  However, particles collected at night with a
high density of diesel traffic exhibited 121 to 238 revertants/m .  Alfheim and
                              37

-------
Miller    collected ambient air particles at a road-side site, on a  roof, and
within a  park.  They  found that the contribution of traffic to mutagenicity of
air samples was significant and that the mutagenicity at street level varied
                                                        43
with traffic  frequency.   In the Allegheny tunnel study,   it was shown that
roadway-produced diesel aerosol is similar in activity to organics  recovered in
dilution  tunnel studies.   Secondly,  it demonstrated that in revertants/km
traveled, the mutagenicity of  diesel vehicle exhaust is several times that of
                                                                         44
gasoline  engine exhaust.   The  New York Port Authority bus terminal  study   pro-
vided some real contrasts  to other ambient air studies.  Although the air
particle  concentration inside  the bus terminal was three times the  outside con-
centration, the mutagenicity outside based on revertants/m  was greater than
                                     45
that of inside air.   (Jungers  et al.    describes the technology available for
ambient air studies.)  These studies demonstrate that the production, chemical
alteration, distribution,  and  concentration of mutagenic mobile source particles
is dependent  upon traffic  patterns,  amounts of reactive gases and vapors, level
of ozone  present, meteorological conditions,  and the presence or absence of
other ambient air particles.
COMPARISON OF VARIOUS MOBILE SOURCES  USING  BACTERIAL BIOASSAYS
   A number of different diesel and gasoline engines and vehicles have been
used for the generation of emission samples for mutation testing.  Although a
few authors have not given descriptions, most have described the engine and/or
vehicles used in their studies; however, these descriptions are basically very
limited.  Generally, even less information  is provided about the test cycle
or run procedures, the fuel and lubricants,  the description of the dilution and
collection devices, and the precise methods of sample preparation.  In addition,
although most investigators described their bioassay as following the procedure
               4                                  AC
of Ames et al.,  recent work by Toney and Claxton   shows that most "Ames testing
laboratories" have made specific modifications to this somewhat standard proto-
col.  The mutagenicity results are also analyzed and summarized in a variety of
ways.  Once these facts are understood, one is aware that the comparison of
engines and vehicles between different studies (and even within the same study)
must be done with caution and only in a qualitative manner.
   In the first paper by Huisingh et  al.   (Parts I and II), two heavy-duty
engines and three light-duty vehicles were  used.   Although a direct comparison
of these sources was not the primary  purpose of this report, it provided the
first mobile source comparison based  on bacterial mutagenicity.  In this
study, particle exhaust organics from the heavy-duty engines were tested in fi?e
                              88

-------
tester strains (TA98, TA1535, TA100, TA1537, and TA1S38) both with and without
exogenous activation.  Qualitatively, the two anginas showed very similar  re-
sults with the four positive strains having decreasing activity in the order
TA100 > TA98 > TA1538 > TA1537.  Without activation, TA1535 was negative with
samples from both engines; however, with activation one engine  (a Caterpillar
3208, 4-stroke V-8 engine) gave a marginally positive response.  When examined
in a more quantitative manner, the samples from these two engines demonstrated
a difference greater than ten times in response for each of the tester strains;
however, results were examined only as revertants/mg of particles.  Due "to
sample amounts available, the exhaust organics from the three light-duty engines
were tested using only strain TA1538.  In this study, in which a comparison was
made based upon the fuel used, results for even a single vehicle  (using different
                                                         5-7 47-49
fuels), could vary greater than 100 times.  Other studies          report  the
results of multiple vehicle and/or engines in their reports.  Qualitatively,
the results are in agreement with the report of Huisingh et al.
   In order to demonstrate the effect of some sampling parameters, Claxton and
     6
Kohan  reported three specific types of comparison:  1) between different runs
with the same diesel engine, 2) between gasoline vehicles of the same make,
model, and configuration, and 3) between different makes of diesel vehicles.
The results are summarized in Table 2.  The coefficient of variation for the
revertants/mi for the above three cases was respectively 0.11, 0.49, and 0.59.
Assuming the coefficient of variation is a good estimation of the standard
deviation in relation to the mean, and assuming a normal distribution for  the
test values, one can estimate the 99% confidence limits and percent of variation
from the mean expected in all three cases.  For the above three cases, a value
could be within 99% confidence limit values and vary by 33, 147, and 177%,
respectively.  If multiple testing facilities, fuels, and bioassay laboratories
are used, the variation between results would be expected to increase.
Recognizing that the Ames assay is a semi-quantitative test for screening
substances over a dynamic range of ~10  in dose/response slope, and recognizing
that other parameters (such as percent extractable of the particles) show
broad variation, then this seemingly large variance for a complex testing
situation should not be considered excessive.  Together, these studies indicate
that semi-quantitative comparisons can be done within a single study and that
cautious qualitative comparisons can be made using results from multiple studies.
                              89

-------
TABLE 2
COMPARISON OF SUMMARY DATA DEMONSTRATING THE EFFECT OF DIFFERING SAMPLING
PARAMETERS6

Different Runs Within Some
Slope
Rev/ug
%
Ext.b
Rev x 105/
g Part0
P.E.R.d
g/Mile
Rev x 105/
Mile6
Automobile: (Diesel)
Mean 3.68 11.8 4.35 0.524
SD 0.42 1.0 0.64 0.037
Coefficient Var. 0.11 0.09 0.15 0.07
Vehicles of Same Make, Model, and Configuration: (Gasoline)
Mean
SD
Coefficient Var.
Different Diesel Vehicles i
Mean
SD
Coefficient Var.
7.03
3.51
0.50
1.98
0.80
0.41
7.52
7.83
1.04
36.6
18.0
0.49
3.16
0.87
0.28
6.96
4.06
0.58
0.0102
0.0048
0.47
0.687
0.256
0.37
2.27
0.26
0.11
0.032
0.016
0.49
4.38
2.59
0.59
 aslope of linear regression  line  (revertants per plate per microgram organic
 material)
 ^Percent extractable
 cRevertants x  10s per gram of particulate natter
 ^Particle emission rate
 eRevertants x  10^ per mile
EFFECT OF PHYSIOLOGICAL FLUIDS AND ENZYME SYSTEMS  ON THE MUTAGENICITY OF
DIESEL EXHAUST
   Since initial testing involved organic chemicals  extracted from particles
with strong organic solvents, researchers questioned whether or not chemicals
bound to carbonaceous particles would be released  into physiological fluids or
in the in_ vivo situation.  McGrath et al. ,    tested  whole particles suspended in
dimethylsulfoxide  (DMSO) and obtained results ranging from negative to moderately
positive in those tested using the Ames bioassay.  However, DMSO is a moderately
effective solvent.  In 1980, Siak et al.,    reported extracting particles with
four simulated biological fluids:   fetal  calf  serum,  0.5% bovine serum
lung surfactant, and saline.  The assay of  each biological fluid in the Ames
test was negative except for  a positive response with the fetal calf serum.
                               90

-------
fetal calf serum extract provided only about 6% of the response found with
                                                   52
extraction by dichloromathane (DCM).  Brooks et al.   found similar results with
dog serum, lung lavage fluid, saline, dipalmitoyl lecithin, and albumin.  How-
ever, they state that "the minimal inutagenic activity... may be due to a lack of
removal of mutagens from the particles or an inactivation of removed mutagens by
binding or some other process."  When Clark and Vigil   tested a DCM diesel
extract in the presence of Aroclor 1254 induced rat liver 39, an uninduced 39,
an 39 without NADP, bovine serum albumin, and fetal calf serum, they showed a
decreased mutagenic response in each case, suggesting that protein binding of
mutagenic components was at least partially responsible for the lack of activity
seen with incubated particles,  By following the mutagenic activity of the DCM
extracts in serum, lung cytosol, protease-treated serum and lung cytosol, and
                                54
extracted particles, King et al.   demonstrated the release of mutagens from
diesel particles and postulated that the lack of mutagenic response is due to
                                                     55
either protein binding or metabolism.  Siak and Strom   exposed rats to diesel
particles, recovered the lung macrophages, and extracted the macrophages with
DCM.  They showed that although the particles continued to contain mutagens,
"seven days after exposure, DCM extracts of alveolar macrophages had no detect-
able mutagenic activity, even though more diesel particles were recovered."
These effects may be due to either protein binding or metabolism. Wang and Wei
and Wang et al.   gave evidence that the antimutagenic effect of S9 is not
enzymatic by examining S9, heat-deactivated S9, S9 minus cofactors, and albumin
effects.  Somewhat in contrast, Pederson and Siak   used a nitroreductase-
deficient bacterial strain to show that some mutagens in diesel particle extracts
are activated by S9 and that 1-nitropyrene was also activated by NADPH-dependent
S9 enzymes.  Other studies presented within this volume will have an impact upon
our understanding of this issue.  In essence we now know the following: 1) muta
genie substances are released from diesel exhaust particles into certain physio-
logical fluids and cells, 2) physiological fluids and 39 decrease the apparent
mutagenic activity of diesel extracts and particles primarily because of protein
binding, and 3) some mutagenic components (e.g., 1-nitropyrene) are activated
by the microsomal fraction of S9, while other components are activated by the
cytosol fraction.
                               91

-------
SUMMARY
   In summary, the work presented demonstrates  that rapid, in vitro indicators

of genotoxicity have been and will continue  to  play a valuable role in under-

standing the toxicity of mobile source emissions.   Bacterial assays have had

tremendous value in the characterization of  mobile source emissions.  Specifi-

cally they have had four major uses:  1) comparative screening, 2}  analyzing

factors that alter the genotoxicants found in emission products, 3) directing

the chemical fractionation of emission organics for the identification of specific

genotoxicants, and 4) analyzing the interaction of complex emission products

with various mammalian systems.


REFERENCES
 1.  Huisingh, J., Bradow, R., Jungers, R.,  Claxton,  L.,  Zweidinger, R.,
     Tejada, S., Bumgamer, J., Duffield, F., Waters,  M., Simmon, V.F., Hare, C.,
     Rodriguez, C. and Snow, L. (1978) in Application of  Short-Term Bioassays
     in the Fractionation and Analysis of Complex  Environmental Mixtures,
     Waters, M.D., Nesnow, S., Huisingh, J.L.,  Sandhu,  S.S.  and Claxton,  L. ed.,
     U.S. Environmental Protection Agency, EPA-600/9-78-027,  pp. 1-32.
 2.  Gage, S.J. (1977) Precautionary Notice  on  Laboratory Handling of Exhaust
     Products from Diesel Engines, Office of Research and Development,  U.S.
     Environmental Protection Agency, Nov. 4, 1977.
 3.  Geomet, Inc. (1979) Proceedings of the  Scientific  Review Meeting on  the
     U.S. Environmental Protection Agency Diesel Emission Health Effects  Research
     Program held at the Pentagon City Quality  Inn,  Arlington, VA,  on
     Dec. 12-13, 1978, U.S. Environmental Protection Agency,  EPA-600/1-79/010.
 4.  Ames, B.N., McCann, J. and Yamasaki, E.  (1975)  Mutation Res. 31,  347-364.
 5.  Claxton, L.D. (1980) in Health Effects  of  Diesel Engine Emissions,
     Proceedings of an International Symposium  Dec.  3-5,  1979, Pepelko, W.E.,
     Danner, R.M. and Clark, N.A. ed., U.S.  Environmental Protection Agency
     EPA-600/9-80-057b, pp. 801-809.
 6.  Claxton L. and Kohan, M.  (1981) in Short-Term Bioassays in the Analysis of
     Complex Environmental Mixtures 1980, Waters,  M.D.,  Sandhu, S.S.,
     Huisingh, J.L., Claxton, L. and Nesnow, S.  ed.,  Plenum Press,  New York,
     pp. 299-317.
 7.  LBfroth, G.  (1980) in Health Effects of Diesel Engine Emissions,  Proceedings
     of an International Symposium Dec. 3-5, 1979,  Pepelko,  W.E., Danner, R.M.
     and Clark, N.A. ed., U.S. Environmental Protection Agency, EPA-600/9-
     80-057a, pp. 327-344.
 8.  Rosenkranz, H.S., McCoy, E.G., Sanders, D.R.,  Butler, M., Kinazides, D.K.
     and Mermelstein, R.  (1980) Science 209, 1039-1043.
 9.  Mermelstein, R., Rosenkranz, H.S. and McCoy,  E.G.  (In Press) in Proceedings
     of the Symposium on the Genotoxic Effects  of  Airborne Agents Feb.  3, 1981.
10.  Liber, H.L., Andon, B.M., Kites, R.A. and  Thilly,  W.G.  (1980)  in Health
     Effects of Diesel Engine Emissions, Proceedings of an International
     Symposium Dec. 3-5, 1979, Pepelko, W.E., Danner,  R.M. and Clarke,  N.A. ed.,
     U.S. Environmental Protection Agency, EPA-600/9-80-057a, pp. 404-412.
11.  Barfknecht, T.R., Andon, B.M., Thilly,  W.G.,  and Hites,  R.A.  (1981)  in
     Chemical Analysis and Biological Fate:  Polynuclear  Aromatic Hydrocarbons,
     Cooke, M. and Dennis, A.J. ed., Battelle Press,  Columbus, OH,  pp.  231-242.
                               92

-------
12.  Huisingh, J., Bradow, R., Jungers, R., Claxton, L., Zweidinger, R. ,
     Tejada, S., Bumgarner, J., Duffield, F., Waters, M., Simmon, V., Hare, C.,
     Rodriguez, C. and Snow, L. (1978) in Application of Short-Term Bioassays
     in the Fractionation and Analysis of Complex Environmental Mixtures,
     Waters, M.D., Nesnow, S., Huisingh, J., Sandhu, S.S. and Claxton,  L. ed.,
     Plenum Press, New York, pp. 381-418.
13.  Lebowitz, H., Brusick, D., Matheson, D., Jagannath, D.R., Reed, M.,
     Goode, S. and Roy, G.  (1979)  Environ. Mutagenesis 1, 172-173.
14.  Wang, Y.Y., Rappaport, S.M.,  Sawyer, R.F., Talcott, R.D. and Wei,  E.J.
     (1978) Cancer Letters 5, 39-47.
15.  Wang, Y.Y., Sawyer, R.F. and Wei, E.J. (1978) U.S. Environmental Protection
     Agency, EPA-600/9-78-027, pp. 587.
16.  Brusick, D. and Matheson, D.W. (1978)  Mutagen and Oncogen Study on
     JP-8, Aerospace Medical Research Laboratory, AMRL-TR-78-20.
17.  Brusick, D. and Matheson, D.W. (1978) Mutagen and Oncogen Study on JP-4,
     Aerospace Medical Research Laboratory, AMRL-TR-78-24.
18.  Calkins, W.H., Deye, J.F., King,  C.F., Hartgrove, R.W. and Krahn, D.F. (1980)
     Synthetic Crude Oils Carcinogenicity Screening Tests:  Progress Report,
     Department of Energy, DOE/EV/10127-3.
19.  Calkins, W.H. and Krahn, D.F.  (1979) Synthetic Crude Oils Carcinogenicity
     Screening Tests:  Progress Report, Department of Energy, DOE/COO-4758-2.
20.  Epler, J.L., Young, J.A., Hardigree, A.A., Rao, J.K., Guerin, M.R.,
     Rubin, J.B., Ho, C.H. and Clark,  B.R. (1978) Mutation Res. 57, 265-276.
21.  Guerin, M.R., Ho, C.S., Rao,  J.K., Clark, B.R. and Epler, J.L. (1980)
     Environ. Res. 23, 42-53.
22.  Henderson, J.R., Li, A.P., Royer, R.E. and Clark, C.R.  (1981) Environmental
     Mut. 3, 211-220.
23.  Pitts, J.N., Jr.  (1979) Philosp.  Trans. R. Soc. London, Ser. A 290, 551-576.
24.  McClellan, R.O. ed.  (1980) Diesel Exhaust Emissions Toxicology Program
     Status Report - July 1980, Department of Energy, DOE/DE-AC04-76EVO 1013.
25.  Hermann, M., Chaude, 0., Weill, u., Bedouelle, H. and Hofnung, M.  (1980)
     Mutation Res. 77, 327-339.
26.  Gabele, P.A., Black, F.M., King,  F.G., Zweidinger, R.B. and Brittain, R.A.
     (1981) Society of Automotive Engineers, SAE Technical Report 810081.
27.  Gibbs, R.E., Hyde, J.D. and Byer, S.M. (1980) Society of Automotive
     Engineers,: SAE Technical Report 801372."                  " "
28.  Schuetzle, D., Riley, T., Prater, T.J., Harvey, T.M. and Hunt, D.F.  (In
     Press) Anal. Chem.
29.  Gibson, T., Ricci, A. and Williams, R. (1980) General Motors, Research
     Publication No. GMR3478.
30.  Gibson, T.L., Ricci, A.I. and Williams, R.L.  (1981) in Chemical and
     Biological Fate:  Polynuclear Aromatic Hydrocarbons, Cooke, M. and
     Dennis, A.J. ed., Battelle Press, Columbus, OH, pp. 707-717.
31.  Pederson, J.C. and Siak, J.S.  (1980) General Motors, Research Publication
     No. GMR-3265.
32.  Pederson, J.C. and Siak, J.S.  (1981) J. Appl.Toxicol., 1:2, 54-60.
33.  Pederson, J.C. and Siak, J.S.  (1981) J. Appl. Toxicol. 1:2, 61-66.
34.  Bradow, R.L. (1980) Bull. NY Acad. Med. 56, 797-811.
35.  Claxton, L.D.  (1980) Detection and Comparison of Mutagens Associated with
     Complex Environmental Substances, pp. 95-132.
36.  Choudhury, D.R. and Doudney,  C.O.  (1980) in Health Effects of Diesel
     Engine Emissions Proceedings of an International Symposium Dec. 3-5,
     1979, Pepelko, W.E., Danner,  R.M. and Clarke, N.A. ed., U.S. Environmental
     Protection Agency, EPA-600/9-80-057a, pp. 263-275.
37.  McClellan, R.O. ed.  (1980) Diesel Exhaust_Emissions Toxicology Program
     Status Report - January 1980, Department of Energy, DOE/EY-76-C-04-1013.
                              93

-------
38.  Ohnishi, Y., Kachi, K., Sato, K.,  Tahara,  S.,  Takeyoshi, H. and Tokiwa, H.
     (1980)  Mutation Res.  77, 229-240.
39.  Rappaport, S.M., Wang, Y.Y., Wei,  E.J.,  Sawyer,  R.,  Watkins, B.E. and
     Rappaport, H.  (1980) Amer. Cham. Soc.  14:12,  1505-1509.
40.  Siak, J.S., Chan, T.L. and Lee, P.S.  (1979) General  Motors, Research
     Publication No. GMR-3171.
41.  Claxton, L.D. and Barnes, H.M.  (1981)  Mutation Res.  88, 255-272.
42.  Alfheim, S. and Miller, N. (1981)  in Short-Term  Bioassays in the Analysis
     of Complex Environmental Mixtures  1980,  Waters,  M.D.,  Sandhu,  S.S.,
     Huisingh, J.L., Claxton, L. and Nesnow,  S. ed.,  Plenum Press,  New York,
     pp. 85-99.
43.  Pierson, W.R., Gorse,  R.A., Szkarlat,  A.C., Brachaczek, W.W.,  Japar,  S.M.
     Lee, S.C., Zweidinger, R.B. and Claxton, L.D.  (1981) Diesel Emissions
     Symposium, Oct. 5-7, Raleigh, NC.
44.  Lewtas, J., Austin, A. and Claxton, L.  (1981)  Diesel Emissions Symposium,
     Oct 5-7, Raleigh, NC.
45.  Jungers, R., Burton, R., Claxton,  L. and Huisingh, J.L. (1981)  in Short-
     Term Bioassays in the  Analysis of  Complex Environmental Mixtures 1980,
     Waters, M.D., Sandhu,  S.S., Huisingh,  J.L., Claxton, L. and Nesnow, S.
     ed., Plenum Press, New York,  pp. 45-65.
46.  Toney, S. and Claxton, L. (In Press) Comparison  of Ames Salmonella
     typhimurium Plate Incorporation Test Protocols,  U.S. Environmental
     Protection Agency Report.
47.  Loprieno, N., DeLorenzo, P.,  Cornetti, G.M. and  Biaggini,  G.  (1980) in
     Health Effects of Diesel Engines Emissionss Proceedings of an
     International Symposium Dec.  3-5,  1979,  Pepelko, W.E.,  Danner,  R.M. and
     Clarke, N.A. ed., U.S. Environmental Protection  Agency, EPA-600/
     9-80-057a, pp. 276-308.
48.  Dukovich, M., Yasbin,  R.E., Lestz, S.S., Risby,  J.H. and Zweidinger,  R.B.
     (1981) Environ. Mutagenesis 3, 253-264.
49.  Dietzman, H.E., Parness, M.A., Bradow, R.L. (1981) American Society of
     Mechanical Engineers,  Publication No.  81-DGP-6.
50.  McGrath, J.J., Schreck, R.M.  and Siak, J.S. (1978)   Presented  at the  71st
     Annual Meeting of the Air Pollution Control Association June 25-30, 1978,
     78-33.6.
51.  Siak, J.S., Chan, J.L. and Lee, P.S. (1980) in Health  Effects  of Diesel
     Engine Emissions, Proceedings of an International Symposium Dec.  3-5, 1979,
     Pepelko, W.E., Danner, R.M. and Clarke,  N.A. ed., U.S.  Environmental
     Protection Agency, EPA-600/9-80-057a,  pp. 245-262.
52.  Brooks, A.L., Wolff, R.K., Royer, R.E.,  Clark, C.R., Sanakey,  A.  and
     McClellan, R.O. (1980) in Health Effects of Diesel Engine  Emissions,
     Proceedings of an International Symposium Dec. 3-5,  1979,  Pepelko, W.E.,
     Danner, R.M. and Clarke, N.A. ed., U.S.  Environmental  Protection Agency.
     EPA-600/9-80-057a, pp. 345-358.
53.  Clark, C.R. and Vigil, C.L. (1980) Toxicol. Appl. Pharmacol. 56,  110-115.
54.  King, L.C., Kohan, M.J., Austin, A.C., Claxton,  L.D. and Huisingh, J.L.
     (1981) Environmental Mut. 3,  109-121.
55.  Siak, J.S. and Strom,  K.A. (1981) Society of Toxicology Presentation.
56.  Wang, Y.Y. and Wei, E.J. (1981) in Short-Term Bioassays in the  Analysis
     of Complex Environmental Mixtures 1980,  Waters,  M.D.,  Sandhu,  S.S.,
     Huisingh, J.L., Claxton L. and Nesnow, S. ed., Plenum  Press, New York,
     pp. 359-368.
57.  Wang, Y.Y., Talcott, R.E., Seid, D.A.  and Wei, E.J.  (1981)  Cancer Letters
     11, 265-275.
                              94

-------
EMISSION FACTORS FROM  DIESEL AND GASOLINE POWERED VEHICLES:  CORRELATION WITH
THE AMES TEST
Roy B. Zweidinger
Mobile  Source  Emissions   Research  Branch,  U.S.  Environmental  Protection
Agency, Research Triangle Park, N.C. 27711

INTRODUCTION
      In  1978,  initial  findings  on the  mutagenic nature  of  diesel extracts
were  reported.     Since  that time,  the Ames 'Salmonella  typhimurium bioassay
has  been used  extensively  in the  investigation of mobile source emissions.
Both  government and  industry have  carried  out  numerous  studies  on the many
variables  effecting  the  mutagenicity of mobile  source samples.  These studies
for   the  most  part   fall   into  six general  categories:  1)  Sampling-exhaust
dilution  ratios,  gas phase vs particulate phase collection and examination of
filter  types; 2)  Bioassay  sample preparation-extraction solvents, extraction
procedures,  solvent  exchange studies   and  sample  storage; 3)  Vehicle types-
light duty diesels,  heavy  duty  trucks  and gasoline powered  cars; 4) Opera-
tional  characteristics-driving cycles,  fuel types, temperature,  mileage and
engine  malfunction  conditions;  5) Artifacts;  6)  Characterization studies-
qualitative  and quantitative studies on the nature  of  mobile source rautagens
and  their  precursors.
      The  areas of  sampling  and  bioassay  sample  preparation  impact  on all
resultant  data.   Some of the concerns  in  these  areas  will be briefly review-
ed.   The question of artifacts,  which  is  closely tied  to sampling consider-
ations,  is an  area  of  on-going study.   The status  of  these  efforts is  being
                                         2
reported  else-where  in  this  symposium .    Diesel  particulate  extracts have
been  characterized by many  investigators.   The  majority have employed liquid
chromatography  to  resolve  the extracts  into non-polar (aliphatic and aromatic
hydrocarbons),  moderately  polar  (transition)   and  highly polar  (oxgenates)
fractions;   the  latter  two  fractions contain most  of   the  Ames activity.
Various  nitroaromatic compounds,  e.g.  1-nitropyrene (1-NP)  are  potent  muta-
gens  which may be very important in the overall activity  observed  for diesel
         3 4
extracts.  '  '   The mutagenic response  of gasoline powered  vehicle particulate
extracts  relative  to diesels is  different  suggesting  the  importance of  other
compound  classes  (see text).  Little characterization  work has been reported
for gasoline particulate extracts,  however.
      The major  emphasis  of this  report  will be  various operational  parameters
and  their effect  on mutagenicity.  Finally, the  results  from several  recent
                              95

-------
studies  on  emission factors  and  Ames  test mutagenicities  of  gasoline and
diesel powered vehicles will be compared.
     Sampling.   The majority  of mobile  source  testing consists of  operating
a vehicle  or engine on a  dynamometer and passing the exhaust  into a dilution
tunnel  connected  to a  constant  volume sampler  (CVS).   The  total  air flov
remains  constant even  though  the engine exhaust flow may  vary due  to trans-
ient  operation  (acceleration,  deceleration).    Sampling  the  diluted exhaust
at  a  constant rate will give  an integrated  sample representative of a parti-
cular  driving cycle.   The average  dilution of  the exhaust is generally from
10:1  to  20:1  depending  on   the engine size   and  average  operating speed,
Actual  dilution  on  the  highway  reaches   1000:1 shortly  after  the exhaust
exits  the  tailpipe.  Concerns were  expressed that collecting diesel exhaust
particulate  samples  under the  relatively  low  dilution  used  in  normal  test
procedures  would  not represent  the  roadway situation,  especially  with re-
spect  to  the  particulate  bound  organics.      Several  investigations  have
indicated  that  the composition  of the  particulate  is  fixed  when  it leaves
the tailpipe and is not significantly influenced by  further  dilution. '   No
differences  were  observed in either  the  particulate  emission rates  or the
molecular  weight  distribution of  the soluble organic  fraction  (SOF).   More
recently,  studies  conducted at the  Allegheny Mountain tunnel  found  that the
diesel  produced aerosol at  Allegheny  to be very  similar  to  that encountered
in  dilution  tube experiments  with  respect  both  mutagenicity  in the Ames test
and molecular-weight  distribution.
     None  of these above  studies,  however, address the  question   of  what
happens to  the  particulate organics  on prelonged  exposure  to sunlight and/or
to  other  chemical  species  present  in  the atmosphere such  as  ozone.   The
destruction  of polynuclear  aromatic  hydrocarbons  (PAH's) under  atmospheric
conditions  is well known.  For example, Falk,   et al. found  that benzo(a)-
pyrene  (BaP) absorbed  on  combustion  soot  was decomposed  10%  on  exposure to
light and  air for 48 hours and 18% on  exposure  to light and  smog for only 1
     Q
hour.
     Preliminary  results  on   the  mutagenicity of diesel  exhaust  particulate
exposed to  light and ozone  in a  smog  chamber found  similar  TA 98 activities
for exhaust  samples which had been irradiated  (6 hours) or  aged  in  the dark
(4  hours).   However, exposure  to  ozone at  parts per  billion levels  (up to
650ppb) resulted  in  a  significant  decrease in  TA 98  activity,  the decrease
being most pronounced without  metabolic activation.
     Another  major area of  study  concerning sampling  has dealt  with filter

-------
types.   Glass  fiber filters  had  long  been  a standard  medium  for general
participate  sampling.   However, several  studies indicated  they were act the
filter  of  choice for  diesel application as they  (Gelman  GF/AE) were subject
to  adsorbing gas  phase hydrocarbons.    This  resulted  in  the  recommendation
of  a  teflon-coated  glass  fiber type  filter,  (e.g. Pallflex  T60A20J for the
determination of diesel particulate emission rates.
     The effects  the different filter media may have  on the Ames activity  of
the SOF of diesels have also  been  studied.   Diesel particulate was  simultan-
eously  collected  on  Teflon  membrane  (Zeflour),  Teflon  coated  glass-fiber
(Pallflex  T60A20)  and  quartz  fiber  (Pallflex 25000AO) filters.   No differ-
ences were observed in sample  mutagenicities  in strains TA 1538, TA 98 or  TA
100.     Particulate  loadings  (0.3 to  0.7  mgxcm  )   were  typical  of  those
 encountered in  diesel  studies. At these loadings the filter  matrix is quickly
covered with particles, and  particles are  then collected  on  particles;  any
interactions  of the  filter matrix with the particles is  minimized.
      Bioassay Sample Preparation.   Soxhlet extraction  has been the  method  of
choice  for  obtaining   the SOF from mobile  source  particulate.   In  the case
where gram quantities  of  the  SOF  are required for bioassay procedures such
as  skin painting,   it  is  the only practical procedure.   Investigations   of
solvents representing  a wide  range of polarities  have generally found dichl-
oro methane  (DCM)   to  be  the most efficient single  solvent  for extracting
mutagens from diesel particulate.
      The more polar solvents  also extract  variable amounts of inorganics  as
shown in  Table 1.      The % sulfate values are relative  to the total amount
of  sulfate present  in the  particulate phase.   On  the other hand, a benzene-
ethanol azeotrope  was  found  superior  for extraction of  benzo(a)pyrene,   a
known  carcinogen  and  am t a gen.   DCM  extracted only about  70%  of the  BaP
obtained by benzene-ethanol.
Table 1
SOLVENT STUDY
Solvent
Cyclohexane
Toluene
Methylene Chloride
Acetone
Acetonitrile
Benzene-Ethano 1


% Extractables
12.65
14.15
14.09
19.08
13-43
21.32


% Sulfate
0
0
0
25 - 50
11 - 50
16 - 25*
*Mass Balance Not Obtained.
                              97

-------
     Operational Characteristics.   Malfunctions:   The majority of  emissions
and Ames test data has been from vehicles operating according to manufacturers
specifications.  However,  many vehicles  in actual  use on  the  highway  may be
operating  under various malfunction conditions due  to  a lack  of  or improper
maintenance, component  failures,  and engine wear.   Some  of these malfunctions
                                                                             14
may  drastically effect  emissions  yet not  noticably  effect performance.
     Little  data is  currently  available as to how these malfunctions  effect
mutagenicity, but  indications are they may be  considerable.   Two studies have
indicated  that  injector  problems  with diesels  can  cause  significant increases
in mutagenicity.    A 1980  Volkswagen Rabbit diesel experienced  a 42% decrease
in Ames  mutagenic  activity (TA 98)  by  replacing a faulty  injector.     Nissan
Motor  Company  also  demonstrated  that problems  with secondary  fuel  injection
due  to a  lack  of pressure pulse  dampening in  the fuel injection  system  can
cause  increased mutagenicity and BaP emissions.
     A recent   investigation  of  in-use  gasoline vehicles    included several
with  various malfunctions  evidenced by their  emission factors as shown  in
Tables 2 and 3.   The specific malfunctions were  not identified,  except that
the  1976 Fury  obviously was an oil  burner  (82% SOF) and 1977 Dodge Aspen was
suspected  of having EGR problems.  The presence of oil in  the SOF  of the Fury
had  a  diluting  effect  on  mutagenic activity  expressed as revertants/pg,  but
mutagenic  activity  on a revertants per mile basis, particularly with activa-
tion was substantially higher than average for the  unleaded gasoline vehicles.
BaP  emissions  for  the  Fury were  also  the highest monitored for  any  of  the
catalyst vehicles  tested  (19.7  (Jg/mi vs fleet average of 2.1  Mg/mi,  excluding
the  Fury).   The other  catalyst  vehicles which failed  to  meet  certification
values for  regulated emissions  generally had  somewhat  lower revertant  per  Mg
activities, but higher revertant per mile activites, with the  exception  of the
1981 Citation.   While these vehicles with some apparent  malfunction generally
had  increased  revertant per  mile  activities,  it should  also be noted  that a
1979 Chevette which met  certification values  for  THC,  CO,  and  NO  ,  exhibited
                                                          o        x         •}
the highest revertant per mile activity (-S9 = 181.4 x 10 ; +S9  = 424.4  x 10 )
of any of the  catalyst  cars, including the Fury.   BaP,  Pyrene, and  1-nitro-
pyrene emission rates as  well  as  revertant  per  Mg SOF activities were also
much higher than the average.
     Fuel effects.   The  composition of  diesel fuel  might  be expected  to  in-
fluence  the  mutagenicity of  the  SOF  in two  ways: 1)  direct contribution of
amtagens,  a.g.  BaP  in  Fuel,  or  2)  by percursor  supply,  e.g.  Pyrene  in  the
                              98

-------
TABLE 2
EMISSION FACTORS FOR MALFUNCTION VEHICLES
Vehicle
1976 Fury
1977 Aspen
1978 Dodge Truck
1981 Dodge Van
1981 Citation
1977-79 Cert.
1981 Cert.

HC
24.5
2.30
2.06
0.65
0.79
1.5
0.41
g/mi
CO
32.2
28.0
22.6
7.1
10.1
15.0
3.4
mg/mi
NO
X
3.0
6.1
3.2
6.4
3.3
2.0
1.0
Part.
192
30.0
36.1
24.5
15.3
(3.1

SOF%
81.7
45.0
23.0
43.3
11.8
in 1976)

TABLE 3
PARTICULATE EMISSIONS FOR MALFUNCTION VEHICLES
TA 98 Activitv

Vehicle
1976 Fury
1977 Aspen
1978 Dodge Truck
1981 Dodge Van
1981 Citation
Fleet Average
SOF
mg/mi
156.9
13.5
8.3
10.6
1.8
4.9
BAP
pg/mi
19.7
3.8
7.9
1.3
0.1
2.1
Rev/Mg
-S9
0.4
3.6
4.3
12.6
3.8
8.0
+S9
1.3
7.5
12.4
5.6
6.8
14.2
Rev/mi *
-S9
65.9
48.1
35.4
133
6.9
40.6
+S9
204
101
102
59.2
12.3
71.0
*x 10 3

fuel  yielding  increased  levels  of nitropyrene  or more  indirectly,  certain
components  may  aid  or be  more prone  to  combustion synthesis  of particular
mutagens.  Table 4 lists  several fuel  parameters  and their  correlation with
Ames activity in revertants per microgram of extract.
     Highest  correlations  were  seen  with  nitrogen  content  of  the  fuel,  al-
though  the range  is  rather  limited.   In view of  this  correlation,  experi-
ments  were conducted  wherein  fuel  nitrogen levels were  varied  by  doping a
base fuel  with  isoquinoline.   However, results thus far  indicate no correla-
tion  between  mutagenicity and  fuel  nitrogen  levels  when  the  nitrogen  is
                           14
introduced as isoquinoline.
     The aromatic  content  of  the fuel  tended to  show  slight correlation with
activity,  particularly without  activation.   Henderson,  et  al.  have recently
reported that  treating diesel  fuel with NO.  greatly increased its mutagen-
icity.  Furthermore,  the activity  of the N0_  treated aromatic  fraction from
the  fuel  was  40  times  greater  than  the  N02   treated  aliphatic  fraction
(strain TA 100).  No 4 or 5 ring aromatics were detected in the  work by
                             99

-------
Henderson.  However, the  fuels  used with  the  Caterpillar  study in Table  4
were  found  to  have pyrene levels  from  185  to 14,000  pg/liter.   (The correla-
tion  of  fuel  aromaticity and fuel pyrene was 0.53).  Given  the high mutagen-
                                      4
icity reported  for the nitropyrenes,    the  pyrene and other aromatics  in the
fuel  may be important nitroaromatic precursors.   However, no correlation was
observed  between specific aromatics in the  fuel  and  the mutagenicity  of the
extracts  in the previous fuel studies.

TABLE 4
R-SQUARE  CORRELATION COEFFICIENTS FOR FUEL PARAMETERS  VS AMES ACTIVITY
Fuel Parameters S9
Aromaticity (14-39%)
Aromaticity (14-39%) +
Fluoranthene, pg/L
Pyrene, pg/L
BaP, pg/L +
Nitrogen, wt.%
Nitrogen, wt.% +
Cat. 3304
TA 1538
0.67
0.82
0.03
0.09
0.00
0.88
0.98

TA 98
0.66
0.37
—
—
—
0.98
0.83
DD-8V71
TA 100
0.95
0.59
—
—
—
0.88
0.20
Cat. - 13 mode composite, 6 fuels, %N, = 0.005 - 0.024
DD-1983 Transient Cycle, 5 fuels, %N, = 0.006 - 0.61

     The lack  of correlation between specific PAH's  in the fuel and mutagen-
icity  is  not  totally  unexpected,  however,  as  a significant  amount of muta-
genicity  may  apparently  be  derived  from  combustion  synthesis   of  hydro-
carbons.   Studies conducted  at Pennsylvania State University  with a single
cylinder  diesel   engine  run  on pure isooctane  and  tetradecane have yielded
                                18
SOF's  with high  mutagenicity.      Furthermore,  the levels  of 1-nitropyrene
found  in the  SOF were similar  to those  found in the SOF of a 1980  Oldsmobile
diesel operated on #2 Diesel fuel containing 21,000 yg/1 pyrene. 19
     Temperature  effects.   The  effects  of  temperature on the  FTP emissions
of  a  1980 Volkswagen  diesel Rabbit and 1980 Oldsmobile  diesel were recently
         20
studied.      The vehicles  were conditioned (soaked) overnight  at  a range of
ambient  temperatures  such  that crankcase  oil  temperatures at  the beginning
of  each  test ranged from 23°F  to  82°F.   The FTP's were  conducted  at ambient
temperatures  (an average of  12°F  above crankcase  temperature) without heat-
ing of dilution air.
     Decreasing  FTP test temperatures resulted in  slightly increased HC, CO,
and NO^  emissions (7-17%),  increased particulates  (17-30%),  and SOF (40-63%)
and decreased  fuel  economy  (-11  to  -17%).  The  majority of  the increased
                             100

-------
                   VOLKSWAGEN AnCS TEST ACTIVITIES
                       IREVlflTAWTS/olerogrw)


                 REV/ug I-S0I  HH          REV
                        Temperature
 Fig. 1.   Low Temperature Study  - Volkswagen  Ames test
 acti vi ti es ( Revertants/mi crogram) .
                    VOLHSWACEN »"£S TEST RESULTS
                         tREVERTANTS/«I!•>
           REV/Kl  >  IBM 1-091
                                     REV/.I  »  IBBO I«S8>
 sm
 see  •

 see  -
 zsa -
                         Temperature °F
Fig. 2.   Low Temperature Study   Volkswagen Ames  test
activities (Revertants/mi xlO j).
                  101

-------
participate and  SOF  was found to be  uncombusted diesel fuel.   On a revertant
per  Mg SOF  basis,  TA 98  activity  without  activation  appeared to  decrease
with  decreasing  temperature  (see  Figure  1).   A  mild correlation  of  Ames
                                                 o
activity  with soak  temperature was  observed,  r   =0.73  for both vehicles.
No correlation was found to exist between  soak temperature and  activity on a
revertant  per mile  basis.  (See  Figure  2)   This  would  be  expected if  the
increased  SOF emissions  at  low  temperature  were  uncombusted  fuel which  is
not Ames active and acts as a diluent.
     Vehicle Types.  As previously mentioned,  the  majority  of mobile source
emission  characterization  work has  been carried out on  test vehicles  which
are  operating according to manufacturers specifications.   However, the  emis-
                                                                    21
sions  from consumer  operated vehicles may  be  appreciably different   and  in-
deed,  it  is  these emissions which impact air quality and public health.  With
this  in mind the  following data  for comparison of various vehicle types  was
selected from representative in-use vehicle studies.
     Tables  5 and 6 lists emission  factors  and Ames  test data  from recent
studies of light  duty in-use vehicles  operated over the  cold start Federal
                15 22
Test  Procedure.   '    The  diesel  data  is  from 6  cars  (4 General  Motors,  1
Volkswagen, and  1  Mercedes Benz)  while the gasoline data is from  a  20 vehicle
study  (4  leaded gasoline  (GM,Ford,  Datsun, Honda) and  16 unleaded  gasoline
vehicles employing various catalytic control systems (8 GM, 4 Chrysler,  and  4
Ford).  Table 7 and 8  list some  limited data  for  heavy duty trucks  operated
over the proposed  1983 transient cycle, a  cold start,  soak, hot  start proce-
                                 	           03
dure  similar to  the light duty FTP sequence.     Several of  the light duty
gasoline  vehicles  had  some  apparent emission  control  malfunctions  as pre-
viously discused (see tables 2,3).
     The average as  well as the minimum and maximum values observed  for each
emission factor  are  given in tables  5  and  6.   In  the case of the  light-duty
gasoline  cars,   the  range was  over  two  orders   of  magnitude.   Vehicle  to
vehicle variation  far  exceeded those observed  in  repeditive  testing of  the
same  vehicle.   (This   implies  it  is  important to  test  a  large  number  of
vehicles in  establishing  emission  inventories which might be used in model-
ing  studies).   Excluding  the oil  burning Fury which  was   discussed  under
malfunctions,  all  the  light-duty  gasoline   vehicle particulate  associated
organic emissions  examined  with  the exception of  1-nitropyrene  were higher
for  the leaded  vehicles.   Ames  activity  in   revertants  per  |jg  of SOF were
similar but  the  revertant per mile levels were  considerably  higher in  the
case of the  leaded  vehicles, mainly as a  result of increased SOF  emissions.

-------
TABLE 5
FTP EMISSIONS FOR IN-USE LIGHT DUTY GASOLINE AND DIESEL VEHICLES
Emission Factors
THC, g/mi

CO, g/mi

NO g/mi
A
Particulates, mg/mi

SOF, mg/mi

Leaded,
Gas (4)A
2.74B
(1.66-3.48r
28.5
(15.0-61.0)
3.52
(2.5-5.2)
102
(49.3-128)
21.1
(6.8-33.5)
Unleaded
Gas (15)
1.05
(0.22-2.94)
12.2
(1.6-28.0)
2.35
(0.2-6.1)
21.0
(5.9-36.1)
4.9
(0.7-13.5)
Diesel (6)
0.38
(0.21-0.60)
1.27
(1.03-1.86)
1.27
(0.8-1.9)
607
(370-1070)
124
(45-290)
   Number of Vehicles
 ,  Mean Value
   Minimum and Maximum Values Observed
 TABLE 6
 FTP PARTICULATE EMISSIONS FOR IN-USE LIGHT DUTY GASOLINE AND DIESEL VEHICLES
Emissions Factors
Bap, (Jg/mi

1-Nitropyrene, Mg/mi

TA 98 -S9, Rev/Mg

TA 98 +S9, Rev/pg
c
TA 98 -S9, Rev/mi
C
TA 98 +S9, Rev/mi

Leaded.
Gas (4)A
14.6
(1.1-35.5)
0.20
(0.08-0.36)
7.3
(6.2-8.0)
12.5
(9.1-15.9)
152
(51.1-256)
258
(107-489)
Unleaded
Gas (15)
2.1
(0.1-12.4)
0.19
(0.004-1.21)
8.0
(0.4-19.6)
14.2
(1.3-42.2)
40.6
(2.3-181)
71.0
(3.3-424)
Diesel (6)
4,5B
(0.9-7-7)
7.8B
(3.4-10.5)
4.1

	

509
(260-670)
	

   Number of Vehicles
°  Typical Values, Not Reported In Ref. 20
C  xlO 3
Both  vehicle  groups exhibited  higher cold  start Ames activity  with activa-
tion,  suggesting the importance  of PAH or  other compounds requiring activa-
tion.   BaP emissions  of  the  catalyst  vehicles  (excluding  the Fury)  had a
                                                                    •7
fair  correlation with Ames  activity  in TA 98  with activation  (r"  = 0.81).
     The  Ames  test  data within  the gasoline  vehicle catagory  may  be quan-
titatively comparable  as most  all  the vehicles were tested  together in  two
experiments,  one with  activation (+S9) and one  without  (-S9).   Comparisons
of  the  data  between  the diesel  and  gasoline  groups  is more  qualitative,
however.  Ames  test results obtained by  different  laboratories or  even  the
                             103

-------
TABLE 7
HDV TRUCK TRANSIENT EMISSION CHARACTERISTICS
Gasoline


GVW Lbs. (xlO-3)
HC, g/mi
CO, g/mi
NO , g/mi
Particulate g/mi
SOF, mg/mi
Ford
370
19.7
20.9
129
13.0
0.58
40.6
HI
345
24.0
6.2
103
13.7
0.89
14.5
Cat.
320B
27.5
1.90
5.1
19.3
0.90
475
Diesel
Mack
676
80.0
1.43
9.5
29.6
1.95
193
Cunm.
290
80.0
2.32
8.0
27.4
1.61
336
DD
8V71
36.9
2.13
75.1
35.5
3.33
537
TABLE 8
HDV TRUCK TRANSIENT EMISSION CHARACTERISTICS
Gasoline


BaP, M8/i
TA98 -S9
TA98 +S9
TA98 -S9
TA98 +S9


ni
Rev/Mg
Rev/Mg,
Rev/mi*
Rev/miA
Ford
370
61.0
2.57
14.9
104
604
HI
345
17.1
8.05
17.4
117
253
Cat.
320B
1.61
1.10
1.08
523
515
Diesel
Mack
676
0.92
1.36
1.22
262
232
Cumm.
290
5.30
1.02
0.95
343
319
DD
8V71
1.33
0.04
0.09
21.5
48.3
"  xlO J

same  laboratory over a  period of  time may vary for various reasons  related
                   24-27
to assay  protocol.        An additional source  of variation may be  related to
the particular  form of  data reduction employed  as  several models  and  linear
regressions procedures  are available  to  calculate activities.28'29   Table 9
compares  the  results obtained   from  the  present study  of four leaded  gaso-
line  vehicles  (Group A)  with those  of four different leaded gasoline  vehi-
cles  (Group B)  obtained  two years  previously.   Although  the  vehicle  mixes
are  similar,  major  differences are  observed  in the Ames  test results.  If
these  earlier  four vehicles were  used  for  comparison,  with  the  recently
tested  unleaded  vehicles,  somewhat  different  conclusions  would  be  drawn.
Since  the sample is  so  small,  it  is  presently  not known  whether  these di-
verse results simply indicate the range of valves  possible  or whether they
alao  reflect variations  in the Ames  test  itself.
                            104

-------
TABLE 9
COMPARISON OF TA 98 AMES TEST RESULTS FROM TWO GROUPS OF LEADED  GASOLINE
VEHICLE - HWFET DRIVING CYCLE
Test
Group
A
B
Test
Group
A
B
HC
1.35
2.21
mg/mi
SOF
23.5
16.1
g/mi
CO NOx
8.15 6.0
32.1 3.5
Rev/Mg
-S9 +S9
8.6 10.7
0.6 1.47
mg/nu
Part.
276
222
Rev/mi
-39
163
9.85
%
SOF
10.9
7.27
xlO 3
+S9
232
23.7
   1970 Ford, 1973 Chevy, 1979 Datsun, 1979 Honda
   1963 Chevy, 1971 Chevy, 1976 Honda, 1978 Datsun

     In  comparing the  light  duty diesel  and  gasoline  vehicle  groups,   the
average  HC, CO,  and  NO  emissions  were lower  for the  diesels.   These  data
suggest  that the  diesel's  durability toward  regulated emissions is  superior
to gasoline vehicles.   Diesel total particulate FTP  emission rates were   5.9
z  the  leaded  gasoline  vehicle  rates  and   19.1 x  the  unleaded gasoline.
Diesel  highway fuel economy  test (HWFET)  values   (not shown)  were 1.3 x  the
leaded  and  13.5  x the  unleaded gasoline.   Diesel  FTP SOF emission rates  were
5.9 x the leaded and 8.6  x the unleaded  gasoline .
     Ames test  results  generally showed the gasoline  vehicles  to have higher
activity  with  activation while  the  reverse was the  case with diesels.   This
may  also  relate to  the  BaP  and  1-nitropyrene  (1-NP)  emissions observed  for
each  vehicle class.    (Note:   BaP &  1-NP levels  for  the   diesels  were  not
reported  in reference  20 but should  be representative  values).   The levels
of  cold  start  FTP BaP,  an  indirect acting  mutagen, were higher  than  the
diesels  in   the  case of  the  leaded  cars  and  similar for  the  unleaded.   As
previously  mentioned,   cold  start FTP  BaP  emissions  and  activity in TA 98
                                            i
with  activation correlated  fairly well (r~  =  0.81)  for the  gasoline vehi-
cles.   On the other  hand,  the  levels of 1-NP  a  direct  acting mutagen,  were
20 to  30 times greater for  the diesels than  the  gasoline cars.  Correlation
of  1-NP  emissions  from  the  gasoline  cars with   activity  in TA  98  without
activation  was  very  poor (r  =  0.25).   The  correlation  of  1-NP emissions
from diesels with  activity  in TA  98  without  activation while not possible in
this study  was found to  be   significant in some unpublished work.     Driving
                                                                              2
cycle studies conducted on  a  1980 Volkswagen  Rabbit  diesel  resulted  in an r
                             105

-------
                                                                            2
=  0.96  and artifact studies with  a Cummins 290  heavy duty diesel had an r  =
                                                                   2
0.97.   On  the other hand, the  above Rabbit diesel  had only an  r  =0.79 for
the  series  of  FTP  runs made  in  the  low temperature study  previously dis-
cussed.   The Ames  activity  therefore,  is  indicative of  the  types  of coo-
pounds  responsible but  individual  mutagenic species  may or  may not  act  as
mutagenic  markers.   The highest  levels of 1-NP  observed in the studies with
the  Rabbit  would  have  accounted  for  5% of  the activity,  using  1.7  rever-
tants/ng  for the activity  of  1-NP (4) and  assuming  their are  no  synergisms
involved.
     The  transient heavy  duty truck data  in Tables  7  and  8 shows  emission
levels  are much higher than the  light  duty vehicles,  with  the notable excep-
tion  of BaP which  is lower  for  the  heavy  duty diesels.   Within the  heavy
duty  (HDV)  class,  many  features   are  similar to  the  light  duty,  i.e., HDV
gasoline  HC  and  CO  emissions are  higher and  particulates and SOF are low
compared  to  the HDV  diesel.  Because  of  the  differences  in  gross  vehicle
weights  (GW),  only the  Caterpillar 3208  can be  fairly compared to  the two
HDV  gasoline vehicles.   Even  in  this  case,  however,  the  gasoline  vehicles
were  operated  at  wide  open  throttle  during much   of the  transient  testing
which was  not  necessary  with the more powerful  caterpillar engine.  The two
stroke  engine  (Detroit  Diesel, DD  8V-71)  has noticably higher particulates
and  SOF,  generally related to oil consumption in this type of  engine.  Ames
activities  in TA 98 in revertants  per  pg in general were much  lower for the
heavy duty diesels than the  light duty diesels, but are similar on  a  rever-
tant per mile basis due to the heavy duty's increased SOF emissions.

ACKNOWLEDGEMENTS
     The  author wishes  to acknowledge and  express gratitude to Susan Bass
for typing this manuscript.
                             106

-------
1.    Huisingh, J.,  Bradow, R.,  and Jungers, R. ,  et al. "Application of
     Bioassay to the  Characterization  of Diesel Particulata Emissions:  Parts
     I & II"  in Application of Short-Term Bioassays  in  the Fractionation and
     Analysis  of  Complex Environmental  Mixtures, EPA 600/9-78-027,  November
     1978.
2.    Bradow, R.L., "Diesel  Particle  and Organic Emissions, Sampling and Arti-
     facts,"  EPA  Diesel  Emissions  Symposium,  Raleigh,  N.C.  October,  1981.
3.    Schuetzle, D., Lee, F.S.C., Prater, T.J., and Tejada, S.B. "The Identifi-
     cation of Polynuclear Aromatic Hydrocarbon Derivatives in Mutagenic Frac-
     tions  of  Diesel Particulate Extracts,"  International  Journal of Envir-
     onmental Analytical Chemistry 9, 93, 1981.
4.    McCoy,  B.C.,  Rosenkrauz,  H.S.  and  Mermelstein, R.,  "Evidence  for  the
     Existance of  a  Family  of  Bacterial Nitroreductases Capable of Activating
     Nitrated Polycyclics to Mutagens," Environmental Mutagenesis 3, 421-427,
     (1981).
5.    Black, F.M. and High,  L.E.,  "Methodology for Determining Particulate and
     Gaseous Diesel Hydrocarbon Emissions,"  SAE Paper 790422 (1979).
6.    Williams, R.L.,  and  Chock, D.P.,  "Characterization of Diesel Particulate
     Exposure," Health  Effects  of Diesel Engine Emissions:  Proceedings of an
     International Symposium, EPA-600/9-80-057a, November 1980.
7.    Pierson, W.R.,  Gorse,  R.A. Jr., Szkarlat,  A.C., Bracheczek, W.W., Japar,
     S.M.,  Lee,  F.S.C.,  Zweidinger, R.B., and   Claxton,  L.D.,  "Mutagenicity
     and Chemical Characteristics of Carbonaceous Particulate Matter from Veh-
     icles  on  the  Road,"  EPA Diesel  Emissions  Symposium,  Raleigh,  N.C.,
     October 1981.
8.   Falk,  H.L.,  Markul, I, and  Kotin,  P.K.,   "Aromatic Hydrocarbons, Their
     Fate  Following  Emission into the  Atmosphere," AMA Arch.  Ind.  Hlth.,  13,
     13-17  (1956).
9.   Claxton,  L.D.  and  Barnes,  H.M.,  "The Mutagenicity  of Diesel  Exhaust
     Particle  Extracts  Collected  Under  Smog-Chamber Conditions  Using  the
     Salmonella  typhimurium  Test  System,"  Mutation  Research  88,  255-272
     (1981).
10.  Clark,  C.R.,  Truex,  T.J.,  Lee,  F.S.C.,  and  Salmeen, I,  "Influence  of
     Sampling  Filter Type  on  the Mutagenicity  of  Diesel Exhaust Particulate
     Extracts," Atmospheric Environment, 397-402  (1981).
11.  Black, F., and Doberstein, L., "Filter Media for Collecting Diesel Parti-
     culate Matter," EPA-600/52-81-071, June 1981.
12.  Siak,  J.S.,  Chan,  T.L.,  and Lee, P.S., "Diesel Particulate Extracts  in
     Bacterial  Test  Systems,"  Health  Effects  of  Diesel Engine  Emissions.
     Proceedings  of   an International   Symposium,  EPA-600/9-80-057a,  November
     1980.
13.  Zweidinger, R.B.,  and Winfield, T.W., unpublished data.
14.  Baines,  T.M.  "Summary of  Current  Status of EPA  Office  of  Mobile Source
     Air Pollution Control Characterization  Projects," EPA-OMSAPC  Report No.
     EPA/AA/CTAB/PA/81-18.
15.  Gabele, P.A., Black, F.M., King.F.G. Jr., Zweidinger, R.B.,  and Brittain,
     R.A., "Exhaust Emission Patterns from Two Light-Duty Diesel Automobiles,"
     SAE Paper 810081, February, 1981.
16.  "Analysis of  the  Factors  Affecting Unusually High  BaP  Emissions from a
     Nissan  SD-22  Diesel  Engine  Car  Observed  at EPA  Test."   Nissan Motor
     Company, LTD, private communication, September 1980.
17.  Lang, J., Snow,  L.,  Carlson, R.,   Black, F.,  Zweidinger,  R., and Tejada,
     S. , "Characterization of Particulate Emissions from In-use Gasoine Fueled
     Motor Vehicles," SAE Paper 81186,  Tulsa, October, 1981.
                             107

-------
18.  Risby, T. and   Lestz,  S., "Exhaust Emissions  from a Diesel  Engine," EPA
     Grant R806558020, private communication.
19.  Tejada, S.B., Private communication.
20.  Braddock, J.N.,  "Emissions of Diesel  Particles  and Particulate Mutagens
     at Low  Ambient Temperatures," EPA Diesel  Emissions Symposium, Raleigh,
     NC, October, 1981.
21.  "Mobile  Source Emissions  Factors," EPA-400/9-78-005,  USEPA,  Office of
     Air and Waste Management, Washington, DC, March 1978.
22.  Gibbs, R.E.,  Hyde,  J.D.,  and Byer, S.M.,  "Characterization of Particu-
     late Emissions  from In-Use Diesel Vehicles."  Paper 801372 presented at
     SAE Fuels and Lub Meeting, Baltimore, October 1980.
23.  France,  C.J.,   Clemens,   W.,  and  Wysor,  T.,  "Recommended  Practice  for
     Determining Exhaust  Emissions from  Heavy Duty Vehicles  under Transient
     Conditions," EPA Technical Report SDSB-79-08, February 1979.
24.  R.J. deSerres  and Shelby, M.D.,  "Recommendations  on Data Production and
     Analysis  Using the  Salmonella/Microsome Mutagenicity  Assay."  Mutation
     Research 64, 139-165, 1979.
25.  Cheli,  C.,  DeFrancesco,  D.,  Petrullo,  L.A.,  McCoy,  E.C.,  and  Rosen-
     kranz,  H.S.,  "The  Ames  Salmonella   Mutagenicity  Assay:   Reproducibi-
     lity."  Mutation Research 74, 145-150,  1980.
26.  Salmeen, 1., and  Durisin,  A.M.,  "Some  Effects of  Bacteria Population on
     Quantitation of  Ames Salmonella-Histidine  Reversion Mutagenesis Assays.
     " Mutation Research 85, 109-118,  1981,
27.  Chu,  K.C.,  Patel,  K.M., Lin,  A.H.,  Tarone,  R.E.,  Linhart,  M.S.,  and
     Dunkel,  V.C.,  "Evaluating  Statistical Analysis  and Reproducibility  of
     Microbial Mutagenicity Assays."   Mutation Research 85,  119-132,  1981.
28.  Myers,  L.,  Sexton,  N.,  Souther land,   L.,  and  Wolff,   T.,  "Regression
     Analysis  of  Ames  Test  Data,"   Environmental  Mutagenesis,   (1981)   in
     press.
29.  Stead, A.,  Hasselblad, V.,  Creason, J., and Claxton,  L.,  "Modeling  the
     Ames Test," Mutation Research, 85 (1981).
                             108

-------
             ANALYSIS OF VOLATILE POLYCYCLIC AROMATIC HYDROCARBONS
                    IN HEAVY-DUTY DIESEL EXHAUST EMISSIONS

                                      by

                   Walter C. Eisenberg and Sydney M.  Gordon
                          Analytical  Research Section
                            IIT Research Institute
                           Chicago, Illinois  60616

                                Joseph M. Perez
                              Research Department
                          Caterpillar Tractor Company
                            Peoria, Illinois  61629


     The breakthrough and/or loss of polycyclic aromatic hydrocarbons (PAH)
from particulate filters was investigated during the  collection of heavy-duty
diesel exhaust emissions.  The concentration of PAH associated with heavy-duty
diesel particulate has been monitored for several years.  During these studies
it was observed that appreciable quantities of organics including PAH pass
through the particulate filter during the sample collection (1).  The emissions
from a heavy-duty diesel engine were sampled using 70-mm Pallflex TX40HI20WW
filters.  A portion of the gas passing through the particulate filter was
sampled using 4.0-mm x 4.0-cm sorbent cartridges.  Chromosorb 102 and Tenax GC
were used to collect gas phase organics.  The particulate filter was extracted
in a micro-Soxhlet apparatus for 6 hours with methylene chloride.  The sorbent
cartridges were extracted by passing hexane/benzene (90/10, v/v) through the
cartridge at a flow rate of 0.5 ml/min.  Following concentration the PAH
fraction was isolated using open column silica gel chromatography.  The PAH in
particulate and gas phase samples were analyzed using gas chromatography/mass
spectrometry (GC/MS) and high performance liquid chromatography (HPLC).

     In the GC/MS analysis the samples were eluted on a 15.0-m x 0.31-mm i.d.
SE-54 fused silica capillary column directly into the ion source of a Varian
MAT 311A mass spectrometer operating in the repetitive scanning mode.  The data
were enhanced using a computer program by Dromey et al. (2) to locate peaks in
the raw data and provide a set of clean mass spectra  of the sample components
free of contributions from background and overlapping peaks.  Compounds were
identified with a library matching search algorithm,  and in the case of complex
spectra the data was manually interpreted.
                                      109

-------
     Twelve parent PAH were measured in the gas phase and participate samples
using reverse phase HPLC.  The analysis was performed using two coupled
4.6-mm x 25.0-cm Zorbax ODS columns and an acetonitrile water gradient.  The
eluant were monitored using an ultraviolet absorbance detector at X = 254 and
280 nm and a fluorescence detector, Xgx * 280 nm and Xgm = 389 nm.

     Over 40 compounds were tentatively identified in the gas phase after a
particulate filter during the collection of heavy-duty diesel exhaust
emissions.  Thirty-five of these compounds were volatile PAH and included
parent- and aldyl-substituted compounds ranging in size from two to five fused
rings.  A distribution quotient was defined as the ratio of the concentration
of the PAH in the gas phase to their concentration in the particulate phase.
It ranged from ~ 56 for fluorene to ~ 1 for benz[a]anthracene.  Experiments to
date show that at some operating conditions the concentrations of volatile PAH
were significantly lower when measured on the particulate only.   Since the role
of these compounds 1n the formation of artifacts and mutagens is unresolved,
the methods for measuring PAH 1n exhaust streams need to be modified to
account for gas phase compounds.


REFERENCES

1.  W.E. Pepelko, R.M. Danner,  and  N.A. Clarke,  eds.   1980.   Health  Effects  of
       Diesel  Engine Emissions.  EPA-600/9-8-057a.   Health  Effects Research
       Laboratory, U.S.  Environmental  Protection Agency:  Cincinnati,  OH
       45268.   pp. 138-174.

^.  R.G. Dromey,  M.J.  Stefik, T.C.  Rindfleisch,  and  A.M.  Duffield.   1976.
       Anal.  Chem. 48:1368.
                                     110

-------
          THE CHEMICAL CHARACTERIZATION OF DIESEL PARTICIPATE MATTER
                                      by
                    James Alan Yergey and Terence H. Risby
                      School of Hygiene  and  Public Health
                           Johns Hopkins University
                              Baltimore,  Maryland

                               Samuel S.  Lestz
                     Department of Mechanical Engineering
                         Pennsylvania State  University
                         University  Park,  Pennsylvania
INTRODUCTION
    A great deal of research has been directed toward elucidating the
potential health hazards of Diesel particulate matter.  The potential health
risks of the emmitted particles are due to a number of important factors.
The mass median diameter of the agglomerated particles found in Diesel
exhaust is less than 1 um (1-4), and their surface areas greater than 50
m^/g (3,4).  The large surface areas facilitate the adsorption of gas
phase combustion products, while the small diameters allow appreciable
residence times in the atmosphere.  In addition, it is generally accepted
that particles less than 1 um in diameter can be respired by humans, with a
significant portion depositing in the pulmonary regions of the lung (5-7).
An assessment of the potential health effects of Diesel particles must also
consider the fact that many of the surface adsorbed species which have been
isolated from the soluble organic fraction (SOF) of Diesel particulate
matter are potentially carcinogenic (8-10).

    A number of references have appeared in the literature in recent years
regarding the identification of the individual constituents of the SOF
(11-15).  The commercial Diesel fuels and lubricating oils which were used
in each of these studies contain such a wide variety of compounds that a
study of the combustion mechanisms which lead to the particle-bound products
becomes extremely difficult.  An understanding of the mechanisms governing
the formation of the adsorbed species is a requisite for complete
comprehension of the potential health effects of Diesel particulate matter.
The primary objective of this research was to simplify the combustion
chemistry in order to better understand the overall mechanisms governing the
formation of the particle-adsorbed species.
                                      Ill

-------
EXPERIMENTAL

    Diesel particulate matter was generated from a single-cylinder engine,
operated on a 1:1 by volume blend of n-tetradecane and 2,2,4-trimethylpentane.
This prototype fuel was used in conjuntion with a synthetic lubricating  oil,
in order to simplify the exhaust chemistry.  Air and argon/oxygen oxidant
systems were employed.  The Ar/02 was used in order to investigate the
products of nitrogen-free combustion.  Particle samples were collected on
142 mm Pallflex filters in an isokinetically drawn sample line, and gas
phase emissions were monitored.  Filters were Soxhlet extracted with
dichloromethane, and the resulting extract blown to dryness under nitrogen
and weighed.  Samples were analyzed by the Ames Salmonella and Comptest  (16)
bacterial assays in order to assess their mutagenic and potential
carcinogenic capaity.

    The simplifications introduced in this study allow the separation and
analysis of the entire SOF in a single pass.  Particle extracts were
analyzed by capillary gas chromatography, using flame ionization and
thermionic specific detectors, and high performance liquid chromatography,
using UV detection.  Both positive and negative chemical ionization mass
spectrometry were employed for direct analysis of the gas chromatographic
effluents, and for analysis of collected HPLC fractions, using a
pt-filament, direct-insertion probe.  Capillary gas chromatography/electron
impact mass spectrometry was utilized for substantiating identifications.
In addition, Diesel particles were heated under vacuum, and the evolved
gases analyzed by chemical ionization mass spectrometry.

RESULTS

    Figure 1 portrays a typical GC/PCIMS total ion profile for an air
oxidant sample.  Table 1 illustrates the compounds which have been
identified in the air oxidant samples.  Identifications were based upon  gas
chromatographic retention indices (17), molecular weights derived from CMS,
and EIMS library searchs.  In general, the Ar/02 samples exhibited the
same major components, with lower concentrations.  GC/TSD results indicated
a lower number of nitrogen containing compounds in the Ar/02 samples,
however, specific identifications of the nitrogenous species were not
possible.  HPLC separations coupled with the mass spectra generated using
the Pt-filament probe were especially useful for identifying higher
molecular weight PAH, while GC/NCIMS data were particularly sensitive to the
oxygenated species found in the SOF

    Results indicate that unsubstituted, non-linear polynuclear aromatic
compounds are the primary particle-bound combustion products from aliphatic
hydrocarbon fuel components.  The identified compounds are demonstrated  to
be fuel-independent products of the diffusion controlled combustion process
which exists in the Diesel engine.  This fact should be considered in any
future health effects studies of Diesel particulate matter.  The potent
carcinogen, nitropyrene, was tentatively identified in the soluble organic
fraction of the particles, and could possibly account for a large portion of
the observed mutagenic properties of the extractable organics.
                                      112

-------
            TICP FOR  GC/PCIKS OF SAMPLE I
            TOTAL ION CURRENT:   3683972       BASE PEAK: 48657
            MASS RANGE(S):  129-320
         I
         N J
         T I
         E i
         N-i
         s I
         I \
i
                                               il
             i 1 1 I .  I i I I 1  1 j I I I | I  I i | I i l ,  l i l ! I I  i I . i i  i i i  j . I I I I t  ! I I I < l  i i l
                   208       408       688      808     1088      1200      14!
                                           SCAN
    FIGURE 1. Total  Ion  Current  Profile  for GC/PCIMS of Air  Oxidant Sample
Naphthalene
Benzofuran,7-methy1-
Inden-1-one,2,3-dihydro-
Methylnaphthalenes
Phthalate-anhydride
Biphenyl
n-Tetradecane
l-Benzopyran-2-one
Biphenylene, or Acenaphthylene
Acenaphthene
Dibenzofuran
Fluorene
9-Fluorene
Anthracene
               TABLE  1. Compounds
                         Phenanthrene
                         Methyl-9-Fluorenones
                         Benzo[c jcinnoline
                         Fluorene Quinone
                         Phenanthrene  Quinone
                         Cyclopenta-phenanthrene-5-one
                         Naphtho[1,8-cd]pyran-l,3-dione-
                         Fluoranthrene
                         Pyrene
                         Methy Ipyrenes
                         Benzo(ghi]fluoranthene
                         CyclopentafcdIpyrene
                         Chryaene or Triphenylene
                         Benzofluoranthrathenes  or Benzopyrenes
                    Identified in Air  Oxidant  SOF
                                      113

-------
REFERENCES

1.  W. H. Lipkea, J. H. Johnson and  C.  T.  Vuk.  1978.  The Physical and
       Chemical Character of Diesel Particulate  Emissions- Measurement
       Tecnniques and Fundamental  Considerations.  S.A.E.  Paper No. SP-430.

2.  J. A. Verrant and D. A. Kittelson.  1977.  Sampling and Physical
       Characterization of Diesel  Exhaust Aerosols.  S.A.E. Paper No. 770720.

3.  J. W. Frey and M. Corn. 1967. Nature.  216:615-617.

4.  M. M. Ross. 1981. Physicochemical Characterization of Diesel
       Particulate Matter. Ph. D.  Thesis. The  Pennsylvania State University.

5.  T. F. Hatch and P. Gross. 1964.  Pulmonary Deposition and Retention of
       Inhaled Aerosols. Academic  Press: New York.

6.  Task Group on Lung Dynamics.  1966.  Health Physics. 12:173-208.

7.  P. Kotin and  H. L. Falk. 1959.  Cancer. 12:147-163.

8.  D. J. Barth and S. M. Blacker. 1978. Journal  of the  Air  Pollution Control
       Association. 28:769-771.

9.  J. McCann, B. N. Ames, E. Choi and  E.  Yanasaki.  1975. Proceedings of the
       National Academy of Sciences U.S.A.  72:5135-5139.

10. M. Dukovich, R. E. Yasbin, S. S. Lestz, T.  H. Risby  and  R.  E. Zweidinger.
       1981. Environmental Mutagenisis.  In  press.

11. E. F. Funkenbusch, D. G. Leddy and  J.  H.  Johnson. 1979.  The
       Characterization of the Soluble Organic Fraction of Diesel Particulate
      Matter. S.A.E. Paper No. 79418.

12. D. Schuetzle, F. S. Lee, T. J. Prater  and S.  B.  Tejeda.  1981.
       International Journal of Environmental  Analytical  Chemistry.  9:93-144.

13. F. W. Karasek, R. J. Smythe and  R.  J.  Laub. 1974. Journal of
      Chromatography. 101:125-136.

14. M. D. Erikson, D. L. Newton, E.  D.  Pellizzari,  K. B.  Tomer  and  D.
      Dropkin. 1979. Journal of Chromatographic Science.  17:449-454.

15. F. Black and L. High, Methodology for  Determining Particulate and
      Gaseous Diesel Hydrocarbon Emissions. 1979. S.A.E.  Paper  No.  79042.

16. M. Dukovich.  The Mutagenic and "SOS" Inducing Activity of Diesel
      Particulate. M. S. Thesis.  (1981). The  Pennsyvania  State  University.

17. M. L. Lee, D.  L. Vassilaros, C. M. White  and M.  Novotny.  1979.
      Analytical  Chemistry. 51:768-773.
                                       114

-------
  THE ANALYSIS OF NITRATED POLYNUCLEAR AROMATIC HYDROCARBONS
                    IN DIESEL EXHAUST PARTICIPATES BY          .
          MASS SPECTROMETRY/MASS SPECTROMETRY TECHNIQUES1

                                       by

                       T. Riley,  T. Prater and D. Schuetzle,
           Ford Motor Co, Scientific Research-Lab., Dearborn, MI 48121;

                            T. M. Harvey and D. Hunt
            Dept. of Chem., Univ. of Virginia, Charlottesville, VA 22901

INTRODUCTION

Recent investigations have indicated that organic solvent  extracts  of light-duty
diesel exhaust particulates exhibit direct-acting mutagenicity when tested using the
Ames assay.  Preliminary estimates indicate that a significant portion of this direct-
acting mutagenicity may be due to the presence of nitrated poiynuclear aromatic
hydrocarbons (nitro-PAH).  Mass analyzed ion kinetic energy spectrometry  (MIKES)
and  triple stage quadrupole (TSQ) analytical techniques  used to characterize these
compounds in diesel exhaust are described.

EXPERIMENTAL

Light duty diesel exhaust particulate  samples  were collected on  T60A20 Pallaflex
filters using a dilution tube and a chassis dynamometer test facility.  Filter  samples
were extracted with dichloromethane.

MIKES analyses were  performed  on a Vacuum Generators ZAB-2F mass spectrometer
using both electron  impact (El) and  negative ion methane chemical ionization (NICI)
procedures.   All experiments were  conducted using  a magnetic sector resolution of
approximately 2000 and  helium as a collision gas ( 1x10"  torr).

A Finnigan  TSQ mass  spectrometer  was used  to  perform  collisionally activated
dissociation  (TSQ-CAD)  and constant  neutral loss studies.   All  experiments were
conducted using positive ion methane  chemical ionization (PICI) procedures. In the
TSQ-CAD studies the first quadrupole was set to transmit a parent ion of  interest
into  the second quadrupole  which functioned as a collision cell  (N_, 5  x 10"  torr).
The third  quadrupole was scanned repetitively to collect daughter ion spectra. In the
constant neutral loss studies, the  first and third quadrupoles were scanned in parallel
with a 17  amu mass deficit.  Under  these conditions, only ions which experience the
loss  of a 17  amu  neutral  fragment  when coliisionally dissociated  in  the second
quadruple  are detected.  This  ion reaction was found to be characteristic of nitro-
PAH compounds.
                                       115

-------
RESULTS AND DISCUSSION

The  TSQ  constant  neutral loss analysis was found to be  a very useful screening
procedure for nitro-PAH compounds. Table I lists 20 different nitro-PAH  derivatives
which were tentatively identified  in diesel particulate extract using this  procedure.
It must be emphasized that this technique only monitors a reaction characteristic of
nitro-PAH compounds  and does not confirm  their presence.   The specificity of  the
constant neutral  loss analysis was assessed  by examining  the  response  of several
compounds  representative  of the  classes of  PAH derivatives  observed in diesel
exhaust.   The results of this study, as shown  in Table n, indicate that  the technique
exhibits good selectivity for the nitro-PAH derivatives.

Table in illustrates the concentration of 1-NP in the exhaust particulate extract from
four different diesel engines  as determined by MIKES and TSQ-CAD analysis. These
quantitation  studies  indicated  that  both  MS/MS  techniques  lacked   sufficient
resolution on the first mass filter to eliminate positive interferences in the daughter
ion spectra completely. The M-16 daughter ion fragment (M+l - OH) was found to be
specific for 1-NP and was used for quantitation by TSQ-CAD, but the electrostatic
sector of  the MIKES instrument did not resolve this daughter ion adequately.  It was
necessary to prefractionate  the OP-1  and PG-1  samples by  preparative scale high
performance liquid chromatography and to  use  NICI techniques  to accomplish an
interference-free MIKES analysis of 1-NP.

Neither MS/MS technique distinguished between nitro-PAH isomers. This information
was obtained by capillary GC-MS.


   Riley,  T., T. Prater, D. Schuetzle, T. Harvey and D. Hunt. 1981. The  analysis of
   nitrated polynuclear aromatic hydrocarbons in diesel exhaust particulates by mass
   spectrometry/mass  spectrometry  techniques.   Presented  at  the 29th  Annual
   Conference on Mass Spectrometry and Allied Topics, Minneapolis, MN.


   Schuetzle,  D., T. Prater,  T. Riley,  A. Durisin and I. Salmeen.  1980.  Analysis of
   nitrated derivatives of PAH and determination  of their  contribution to Ames
   assay  mutagenicity for   diesel particulate  extracts.  Presented  at  the  Fifth
   International Symposium on Polynuclear Aromatic Hydrocarbons, Columbus, OH.
                                       116

-------
Table I. Nitro-PAH Derivatives Tentatively Identified in Diesel Participate
           Extracts by TSQ Constant Neutral Loss Analysis

Nitroacenaphthy lenes
Nitro(acenaphthlenes, biphenyls)
Nitronaphthaquinones
Nitrodihydroxynaphthalenes
Nitrofluorenes
Nitro(methylacenaphthalenes, methylbiphenyls)
Nitro(trimethylnaphthalenes)
Nitro(naphthalic acid)
Nitro(anthracenes and phenanthrenes)
Nitro(fluorenones and methylfluorenes)
Nitro(methylanthracenes and methylphenanthrenes)
Nitro(anthrones and phenanthrones)
Nitro(pyrenes and fluoranthenes)
Nitro(dimethylanthracenes and dimethylphenanthrenes)
Nitro(methylpyrenes and methylf luoranthenes)
Nitro(pyrones and fluoranthones)
Nitro(pyrene and fluoranthene)quinones
Nitro(dimethylphenanthrene and dimethylanthracene) carboxaldehydes
Nitro(methylbenzo(a)anthracenes, methylchrysenes and methyltriphenylenes)
Nitro(benzo(a)pyrenes, benzo(e)pyrenes and perylenes)
 Table H.  Selectivity of the TSQ Constant Neutral Loss Analysis for Nitro-PAH

       Compound Class                          Selectivity Ratio3
                                            (Nitro-PAH/Compound Class)

           Amines                                     50/1

           Aldehydes

           Quinones
                                                       200/1
           Carboxylic Acids

           Acids
a interference level <5%
                                        117

-------
Table HI. Quantitation of 1-NP in Diesel Exhaust Particulate Extract
                   using MS/MS Techniques
Engine Sample


    Nl-1



    OL-1

    OP-1
    PG-1
Instrument


   TSQ

   MKES

   TSQ

   TSQ

   MKES

   MKES

   MKES
lonization


   PICI

   El

   PICI

   PICI

   El

   NICI

   NICI
Concentration
    (ppm)

  2285+230

  2080±220

   204+30

   77+15
   55+11

   150+30
                                     118

-------
          CONTRIBUTION OF 1-NITROPYRENE TO DIRECT ACTING AMES ASSAY
                MUTAGENICITIES OF DIESEL PARTICIPATE EXTRACTS


                                     by


                     Irving Salmean, Anna Marie Durisin,
            Thomas J. Prater, Timothy Riley, and Dennis Schuetzle
                       Engineering and Research Staff
                                  Research
                              Ford Motor Company
                              Dearborn, Michigan


     We have determined the percentage contribution, P, of 1-nitropyrene
(1-NP) to the direct acting Ames assay mutagenicities of dichloromethane
extracts of exhaust particles collected from three different diesel powered
passenger vehicles.  For strains TA98, 100, and 1538, respectively, P was
(16, 1, and 7%) for vehicle 1; (24, 9, and 7%) for vehicle 2; and (13, 4, and
13%) for vehicle 3.  We assumed:

                          P = C x (M1/M2) x 100%

     Ml and M2 are the slopes of the linear portion of the Ames assay dose-
response functions for 1-NP and for diesel particulate extract respectively.
C, the mass of 1-NP in the particulate extracts/mass of extract, was deter-
mined by collisionally activated dissociation mass analyzed ion kinetic
energy spectrometry (CAD-MIKES).  [CAD-MIKES is a double mass spectrometer
technique with which compounds often can be analyzed in complex mixtures
directly without chromatographic separation.)  C for vehicles 1, 2, and 3,
respectively, was 55 + 11, 2030 +_ 220, and 150 +_ 30 ppm.  We used fused-silica
capillary column GC-MS~ to show that the mass 247 0-NP) CAD-MIKES spectrum
was due to 1-NP and not to some other mass 247 isomer.  Several isomers of
nitrofluoranthene and nitropyrene were synthesized to assure adequate GC-MS
separation of these isomers from 1-NP.

     The slopes, Ml and M2, are approximately proportional to uN, where y is
the mutation rate per concentration of mutagen and N is the total number of
histidine auxotrophs in the background lawn Q).  N is proportional to the
initial inoculum and the average number of auxotrophs per individual back-
ground colony.  N cannot be measured in the Ames assay and it is not neces-
sarily the same for Ml and M2.  Consequently, whenever we carried out an
Ames assay of dtesel particulate extract we concurrently carried out an assay
of Ir.NP using equal amounts of the same broth culture, thereby ensuring that
the initial  inocula were equal.  We then obtained photomicrographs of the
                                      119

-------
background lawn (100X) on each plate at the time of counting revertants and
from these photomicrographs counted the background colonies and estimated
their average size.  If the background colony counts and their average sizes
were the same for both 1-NP and diesel particulate extracts, we took this as
evidence that N was the same and then used the corresponding slopes to
calculate P.

     Finally, we did experiments in which various known amounts of 1-NP were
added to a constant amount of extract and obtained the Ames assay dose-response
function of the 1-NP in the presence of the diesel extract.  The added amounts
of 1-NP were chosen to be within the same order of magnitude as that of the
1-NP already in the particulate extracts as measured by mass spectrometry.
We found that the slope of the 1-NP dose-response function in the presence of
particulate extract was the same as that for 1-NP alone in solution.   This
result supports an assumption, tmplfctt in the above equation, that the
mutagenicity of 1-NP is not altered by the other components in the mixtures.

     These data show that 1-NP is an important contributor to the mutagenicity
of these diesel particulate extracts.  Even a 1% contribution is important,
considering the thousands of compounds present in the particulate extracts.
This observation is very encouraging to experimenters seeking to identify
mutagens in complex mixtures because it suggests the possibility that a small
number of compounds may account for the majority of the mutagenicity  of these
complex mixtures.


                                REFERENCES


1.  Salmeen, I. and Durisin, A., Mutat. Res.. 85., 1QT-118, 1981.
                                     120

-------
      DINITROPYRENES:  THEIR PROBABLE PRESENCE IN DIESEL PARTICLE
      EXTRACTS AND CONSEQUENT EFFECT ON MUTAGENIC ACTIVATIONS
                       BY NADPH-DEPENDENT S9 ENZYMES
                            T. C. Pederson and J-S. Siak
                          Biomedical Science Department
                       General Motors Research Laboratories
                                 Warren, MI 48090


The direct-acting  mutagenic  activity  of diesel  exhaust  particle extracts  in  the
Salmonella mutation assay may be mostly due to nitroaromatic compounds activated by
bacterial  nitroreductase enzymes.   Chromatographic separations  have demonstrated
that much of the  extract's mutagenic activity is associated  with components that are
more polar than 1-nitropyrene or other monosubstituted nitro-PAH  (Pederson and Siak,
1981,  J.  Appl.   Tox., 1(2):54-60).   The  present studies,  employing TLC  and HPLC
separation techniques and the recently developed dinitropyrene-resistant  Salmonella
strains (Rosenkranz et al, 1981, Mutation Res.,  91:103-105),  investigate the  probable
presence of dinitro-,  trinitro-,  or tetranitropyrenes in the polar mutagenic  fractions of
diesel particle extracts.

The direct-acting mutagenic  activity of a diesel particle extract, reduced by 40% in the
niridazole-resistant  strain TA98NR, was decreased by 70%  in  the  dinitropyrene-
resistant strain TA98/l,8-DNPg. After fractionation  of the extract by  silica gel TLC,
all  mutagenic fractions exhibited reduced activity in TA98/l,8-DNPfi, as shown in the
figure.  The most  marked reduction  occurred with the fractions  wnich co-chromato-
graphed with reference samples of the multisubstituted nitropyrenes.  The mutagenicity
of these  fractions  and multinitropyrenes was markedly reduced in TA98/l,8-DNPfi, but
not in TA98NR.  HPLC separations on a cyano phase-bonded silica column indicate that
1,8-dinitro-and  1,6-dinitropyrene are  the predominant  mutagenic  components  in  the
material  recovered from TLC fractions 11 through 15. Some mutagenic activity was also
attributed to 1,3-dinitro- and 1,3,6-trinitropyrene.  The dinitropyrenes may account for
15-20% of the mutagenic activity in the particle  extract, but they are very  potent
bacterial mutagens and would  be present at concentrations  of less than 1 ppm in the
exhaust particulate.

The nitroreductase-deficient  bacteria have also been used in the salmoneJla/S9 muta-
tion assay to examine the effect of mammalian enzyme activities  on the mutagenicity
of diesel  particle  extracts and the nitropyrenes.  Under appropriate  conditions, the
activation of mutagens in diesel particle extract by rat liver S9  enzymes was evident as
a difference between assays with and without NADPH (Pederson and Siak, 1981, J. Appl.
Tox.,  l(2):61-66).  1-Nitropyrene was similarly activated  by S9  enzyme activity.  The
activation of 1-nitropyrene is located  in the microsomal fraction of the  S9 preparation,
but activation of diesel particle extract was more evident with the cytosol  fraction.
                                      121

-------
The mutagenicity of each dinitropyrene was greatly reduced by the NADPH-dependent
activity  of  S9  enzymes.   As  shown  in  the table  below,  the  NADPH-dependent
inaetivation is  catalyzed by  microsomal enzymes, but with the  cytosol fraction, the
mutagenic activity of the dinitropyrenes is increased.  The NADPH-dependent  increase
in mutagenicity of the dinitropyrenes includes both  a cytosol-independent reaction,
which is probably a direct reduction of the compounds by reduced pyridine nucleotide,
and an enzyme-catalyzed  reaction. NADH is just as  effective as NADPH in both the
cytosol-dependent and independent activation reactions.

The  NADPH-dependent increase in  mutagenicity of  diesel particle  extract in the
SaZflioneZL3/S9  assay  involves  both  multiple extract components  and  multiple  S9
enzymes.  The  dinitropyrenes presumably contribute to the cytosol-catalyzed activa-
tion.  The much smaller  effect  of microsomal enzymes on the mutagenicity of the
particle  extract  must  reflect competing activation  and  inactivation  reactions  as
evidenced by the difference between 1-nitropyrene  and  the dinitropyrenes.
                                               Net TA98NR Revertatits/Plate


1,6-Dinitropyrene
10 ng/plate

1 ,8-Dinitropyretie
2 ng/plate

Diesel Particle
30 ug/plate



-NADPH
+NADPH
Change
-NADPH
+NADPH
Change
-NADPH
+NADPH
Change
plus
S9 microsonies
282 +17
19 ± 5
-95%
196 ±13
7 ± 6
-95%
326 + 6
415 ± 8
+25%
plus
S9 cytosol
347 + 16
878 ± 98
+150%
392 ± 21
1129 + 51
+190%
344 + 17
605 + 20
+75%
                                                      ASSAY
                                                      CONDITIONS

                                                    CD TA98
                                                    BB TA98NR
                                                    •I TA98/1,8-DNPS
                                          15         20
                                          TLC Fraction
                                      122

-------
SECTION 3
PULMONARY FUNCTION
                       123

-------
 INHALATION TOXICOLOGY  OF  DIESEL  EXHAUST PARTICLES

 ROGER 0. MCCLELLAN, ANTONE  L.  BROOKS,  RICHARD G. CUDDIHY, ROBERT K. JONES,
 JOE L. MAUDERLY AND RONALD  K.  WOLFF
 Inhalation Toxicology  Research Institute,  Lovelace Biomedical and Environmental
 Research Institute, P.  0. Box  5890, Albuquerque, New Mexico, USA
 INTRODUCTION
   Diesel engines  have  found wide  application in heavy duty vehicles and equip-
 ment for many years  and in  recent  years  have been used to an increasing extent
 in light-duty vehicles.  This  latter usage  is projected to increase substan-
 tially  in the future, both  in  the  United States and worldwide.  It has been
 estimated that 345,000  tons of diesel  exhaust particles were emitted by diesel-
 powered heavy duty trucks,  off-road  equipment and railroad engines in the United
 States  in 1977.    Further,  it  has  been estimated that light-duty diesel vehicles,
 if they comprise 20% of the U.S. automotive fleet in 1995, will release to the
 atmosphere an additional 60,000 tons  of  diesel  exhaust particles per year.  This
 estimate has been  made  assuming light duty  vehicle emissions can be held to a
 level of 0.12 gm/km, a  level below that  currently being attained by most vehicles
 with existing exhaust emission control devices.
   In recent years,  increased  concern has developed for the potential, health
 effects of diesel  exhaust particles.   This concern stems from recognition that
 (a) diesel exhaust particles are small in size,  readily inhaled and deposited
 throughout the respiratory  tract with a  substantial fraction in the deep lung,
 (b) the relatively insoluble carbonaceous core  of the particles results in their
 tenacious retention  in  the  lung, and  (c) the cytotoxic and mutagenic properties
 of organic solvent extracts of the diesel exhaust particles may result in func-
 tional diseases or cancer,  especially of the lung.  Recent concern for the muta-
 genic, and potentially  carcinogenic,  properties  of the particles has no doubt
 been fostered by the development,  during the last decade, of relatively simple
 techniques for assaying  mutagenicity  in  bacterial and mammalian cells.  However,
 to some extent, the  concern generated by recognition of the mutagenic properties
of diesel exhaust  particle extracts  has  been countered by the failure to demon-
strate increased incidence of  cancer  in  human populations occupationally exposed
to diesel engine exhaust.
   The potential health  effects of diesel exhaust coupled with the projected in-
creased use of diesel-powered  vehicles has  stimulated research to resolve health
                              124

-------
risk uncertainties.  The central issue being addressed by this international
research effort is; does occupational or environmental exposure of people to
diesel exhaust result in increased health risks?  Because non-particulate emis-
sions of diesel engines are qualitatively similar and quantitatively not markedly
different than those of gasoline spark ignition engines, the research effort has
focused on the effects of diesel exhaust particles.  This paper will briefly
review this research, summarize some of the most significant findings obtained to
date and identify areas requiring further research.
EXPERIMENTAL APPROACH
   The research program to develop an improved understanding of the health
effects of diesel exhaust has proceeded along two inter-related avenues.  The
first is to develop as much information as possible from epidemiological studies
of populations that have been occupationally exposed to diesel exhaust.  Unfor-
tunately, the  information base obtained to date from the epidemiological studies
has been limited because there are relatively few suitable populations available
for study.  The major research in this area was summarized at a 1979 Symposium.
The most extensive epidemiology study that has yielded significant, albeit
negative to date, results is a study of London Transport Authority workers em-
ployed during  the years 1950-1974.  Although the findings of this study as re-
                                                4
gards lung cancer incidence are negative, Harris  has recently analyzed them to
provide an upper boundary estimate of the lung cancer risk.  (A 0.05 percent
proportional increase in lung cancer incidence for an exposure of 1 microgram of
particulate per cubic meter for 1 year.)  Other epidemiological studies are
currently being conducted.  However, useful data from them is not likely to be
available for at least several years.  In the absence of adequate information
obtained in man, it has been necessary to pursue a second avenue; the conduct of
studies with laboratory animals and other biological systems.  This approach is
based on recognition of the many similarities between different species such that
extrapolations can be made between species, including laboratory animals to man.
Moreover, it is assumed studies of laboratory animals and simpler systems, i.e.
organ, tissue, cell and subcellular entities can provide insight into the mecha-
nisms by which the human body responds to foreign materials such as diesel ex-
haust.
   The studies being conducted are directed to answering three questions: 1) Does
exposure to high levels of diesel  exhaust particles result in increased health
risks?  2)  What are the mechanisms by which the health effects are produced? and
3) Is there a basis for extrapolating these health effects to low level exposures
                             125

-------
by considering the  probable levels of human exposure,  the kinds of health effects
observed and the mechanisms by which the health  effects are produced?  A premise
inherent in this approach is that studies can be conducted at exposure levels
higher than those typically encountered in occupational or environmental settings
and extrapolated to exposure levels likely to be of concern.
   The overall approach  being taken is shown schematically in Figure 1.  As may
be noted, the research program consists of several  inter-related components
directed to obtaining  information that is required  to  provide the desired end
product;^an~assessment of "the~potential health risks to man from exposure to
diesel exhaust.  Let us  consider the results to  date for each component.
                                   Transport and
                                   Transformation
          I  Physical, Chemical and
            Biological Characterization
               of Emissions
                                                           ni
                                                            Effective Dose
                                                             • Deposition
                                                             • Retention
                                                             '• Fate in Body
                                              Health Effects
                                              • Cancer
                                              • Functional Disease
                          3t Integrated Health Risk Assessment
Fig.  1.  Schematic representation of  the research approach being taken  to eval-
uate  the health effects of diesel exhaust.
PHYSICAL, CHEMICAL  AND BIOLOGICAL CHARACTERISTICS
   Other  papers  in  this symposium have provided  an  excellent review of the cur-
rent  status of  our  knowledge of the physical,  chemical  and biological charac-
teristics of diesel  exhaust emissions.  Perhaps  the most striking features of the
data  are  the qualitative and, to a considerable  extent, quantitative similarities
between  the exhaust emissions resulting from a wide range of vehicle, fuel and
operating variables.  This is important since it  provides a basis for having
confidence that the diesel  exhaust exposure environments being utilized in the
whole animal studies  are representative of the varied environments likely to be
encountered by  man.   Thus,  the results of the  animal studies are not likely to be
unique to the specific exposure atmosphere being studied and its source.
                              126

-------
   An additional point that is worthy of note is the extent to which studies with
cellular and subcellular systems are starting to focus on consideration of the
effects of individual chemical compounds and classes of compounds.  They are pro-
viding more detailed dose-response orientation than earlier studies which were,
of necessity, more of a screening nature.  A more detailed level of knowledge
should provide a better basis for understanding the mechanisms that may be in-
volved in producing effects in the whole animal exposed to exhaust.  Past re-
search with extracts of particles has focused on their mutagenic properties.
Looking to the future it would be appropriate for additional attention to be
focused on the non-mutagenic properties of diesel exhaust particles and inter-
actions between mutagenic and non-mutagenic effects in cellular and subcellular
systems.  This is especially important recognizing that to date no convincing
evidence of carcinogenicity has been found in studies of animals exposed to whole
diesel exhaust.  Conversely, non-mutagenic changes, reflecting both tissue injury
and  repair, have predominated in the studies of animals exposed to whole diesel
exhaust.
TRANSPORT AND TRANSFORMATION
   In developing an assessment of the health risks of diesel exhaust, a key ques-
tion is; what reaches the breathing zone of man?  It is important that this
question be kept in mind while considering the laboratory research findings since
the majority of the research being conducted involves either animal exposures or
sampling of diesel exhaust particles at relatively low dilutions and within
seconds or minutes of their emission.  It must be kept in mind that these condi-
tions are not typical of all occupational or environmental exposure situations
for man.  In situations where people may be exposed to high levels of exhaust
components (parking garage workers, for example), the exposure atmosphere will
represent material that has had a residence time of at least several minutes in
the atmosphere.  Material having even longer residence times may predominate in
other atmospheres such as the street canyons of large metropolitan areas.  Fur-
ther, the atmosphere is likely to contain not only freshly emitted exhaust, but
also resuspended material that may have been exposed to sunlight or subjected to
other environmental variables for extended periods of time.
   The number of studies conducted on the transformation of diesel exhaust in the
atmosphere has been limited; however, they do give rise to concern for signifi-
cant changes in diesel exhaust particle constituents between the time they leave
the tailpipe and reach the breathing zone of man. '   The results of these
studies errrohasize the need for additional research in this area.
                             127

-------
    It  is  also  important  to  consider the fate of the particles, and especially
 their  organic  constituents,  after being released into the environment and whether
 they may  affect man  through  other routes of entry.  The reTease of several hun-
 dred thousand  tons of  particles  from diesel  vehicles to the atmosphere each year
 emphasizes  the importance of this question.   Obviously, we need to know the
 extent to which the  organic  constituents may be deposited on foodstuffs and be
 available for  ingestion.  A  related need is  to determine the fractional absorp-
 tion of these  materials from the  gastrointestinal  tract of man.
 DEPOSITION  AND RETENTION
    Adequate assessment of the health risks  of inhaled diesel  exhaust particles
 requires  knowledge of  the deposition,  retention and fate of these particles and
 their  organic  constituents in man.   There is  a substantial  body of information
 available on the deposition  of inhaled particles  in man for a wide range of
 particle  sizes.  The kinds of information available may be found in recent
                   78                      9
 papers  by Lippmann;  Chan and Lippmann;   and  Stahlhofen et_ aj_.    Unfortunately,
 most of the data have been obtained for  particles  larger in size than diesel
 exhaust paj^tKiles. _The most relevant  human  data were reported  by Chan and
 Lippmann  who  studied  the deposition of 0.2  um diameter particles in healthy non-
 smokers that breathed the test aerosol  through a mouthpiece,  the mean values
 obtained  in their studies for tracheobronchial  and  pulmonary  deposition are shown
 in  Figure 2.   For comparison  purposes, data  are also shown  for  dogs,  rats  and
 mice that had  inhaled radiolabeled  gallium oxide aerosols with  a volume median
                30r
                                                    Dog
              c
              o
              1 20
              o
              a
              CO
              Q
              c
              a>
              o
              5 10
              Man
                        Mouse
        1
                      Rat
                    Naso-
                   pharyngeal
Tracheo-
bronchial
Pulmonary
Fig.  2.  Deposition of 0.1 or 0.2 vm particles in mouse,  rat,  dons  or man.
                              128

-------
diameter of 0.1 pm (similar in size and shape to diesel particles).  The labora-
tory animal data are useful in two ways; first, they provide a means for extend-
ing the more extensive observations on man to smaller particle sizes, and second,
they provide essential information for use in interpreting studies of inhaled
diesel exhaust particles in laboratory animals.  With regard to the first con-
sideration, it is noteworthy that Cuddihy e_t aj_.   found good agreement between
data obtained in man and Beagle dogs for the fractional pulmonary deposition of
particles ranging from 0.4 to 5 ym in mass median aerodynamic diameter.  Further,
these results (Fig. 3) are in general agreement with the predicted values based
          11 1 ?
on theory.  '    This lends confidence to using the fractional pulmonary deposi-
tion values obtained with 0.1 ym volume median diameter particles in the Beagle
dog   for estimating the fractional pulmonary deposition of diesel exhaust par-
ticles  in people until such time as direct information can be obtained in man.
Looking  to the future, plans are being developed at our Institute to conduct
studies  of the deposition of surrogate diesel exhaust particles (radiolabeled
gallium oxide) in human subjects.
         .4
       o
       a
       o.
       ^  ->
       o .2
                                Theory ( 1CRP)
                                       Human and Dog
                                         I Da;a
            \
Jheory
Cfehand Schum)
               0.05
            O.I
           Diesel
                -Volume Median Diameter
10
                                 Mass Median Aerodynamic-*)
                                          Diameter
Fig. 3. Pulmonary deposition of inhaled particles in humans and dogs,   ihe data
points in the diesel particle size range are from studies with Beagle dogs.13
    In considering the data on deposition, it should be emphasized that  the range
of  uncertainty of the fractional deposition of diesel exhaust particles in the
respiratory tract of people is relatively low (approximately a factor of 2) and
                              129

-------
 as  noted,  this  uncertainty  can  be  further  reduced  by  studies  of people exposed to
 surrogate  diesel  exhaust  particles.   Unfortunately, information on retention of
 diesel  exhaust  particles  in the respiratory  tract  of  man  is much more limited.
 Short-term observations of  the  retention of  radiolabeled  materials will  provide
 insight into  the  rapid clearance phases in man  for particles  similar in  size to
 diesel  exhaust  particles.   It is unlikely  that  approaches can be developed  that
 will  give  reliable  information  on  the  long-term retention of  these small  size
 particles  in  the  respiratory tract of  people.   A possible exception is to study
 the retention of  iron oxide particles  in people using a magnetic detector system
 such  as  has been  used for larger sized particles.     Until such  time as  informa-
 tion  can be obtained directly in man,  the best  information available will be that
 obtained from laboratory animals such  as shown  in  Figure  4.   Chan et^alk    have
 reported on the retention of carbon-14 labeled  diesel exhaust particles  in  rats
 exposed  for short periods of time to diesel exhaust.  They have  interpreted the
 retention  curve as  having two components with approximate half-times  of  1 day and
 62  days.   These results are quite similar to those obtained in our institute with
 rats  exposed  to several different types of relatively insoluble  aerosols when the
 observations  have been restricted to a 100 days  or so post-inhalation  exposure.
 Typically, when the animals  have been  observed  for longer periods  of  time, an
 additional component is observed with a longer  half-time.  Also  shown  in Fig-
 ure 4, are data obtained by Griffis e£a]_.16 and Strom and Garg17  who  analyzed
 the lungs  of  rats for the carbonaceous particles at various times  following
 extended exposure to diesel  exhaust.  These data suggest  the  presence  of a
Zl00i
o
i-
2
LU 50
t—
LU
OC
O
z
3 20
LU
o 10
-
\ 18 Wk. Exposure
\ ^^ / (Griffis)
-\ ^^*»
s
^
sx Single Acute Exposure
*s/ 
-------
longer-term component  in  the  pulmonary  retention of diesel  exhaust particles
which is not apparent  during  short  observation  periods following acute exposures.
There is also a suggestion  that  chronic exposure to high concentrations of par-
ticles may prolong  the retention of particles.   It should be recalled that pul-
monary clearance of particles  is generally  more rapid, and  thus retention is
                                                      1 Q
lower in rodents than  in  larger  species such  as dogs.     This suggests the need
to evaluate diesel  exhaust  particle retention in dogs  and perhaps subhuman
primates as an aid  in  predicting particle retention in man.   Retention data are
essential for interpreting  the health effects studies  which  will be discussed
later.  They permit comparisons  to  be made  on the basis of  dose-response rela-
tionships where dose is expressed as retained particle mass  rather than the less
supportable exposure-response relationships which do not consider species dif-
ferences in deposition and  retention.
   The ultimate objective of  the studies of the deposition  and retention of
diesel exhaust particles  and  their  organic  constituents is  to develop an inte-
grated model of the fate  of diesel  exhaust  particles and their constituents as
shown in Figure 5.   The ideal  model  will provide a description of the temporal
pattern of distribution of  diesel exhaust particles and their constituents
throughout the body with  emphasis on the respiratory tract.   It should extend to
the  cellular level  providing  a description  of the time course of contact of
                                         Naso-
                                         pharynx
          Tract
    DISSOCIATED
     ORGANIC
     COMPOUND
  retained adjacent to
  particles or tiansported
  to other sites

'- interact in unaltered
  form, detoxified or
  activated
                        OTHER SITES
Fig. 5.  Schematic model of the fate of  inhaled  diesel  exhaust particles and
associated organic compounds.
                             131

-------
 particles with the various cell types of  the  respiratory tract.   Information of
 this type has been obtained by serial sacrifice  of  exposed animals with subse-
 quent morphological evaluation at both the light and  electron microscopic level.
 From such studies, it is clear that epithelial cells  lining the  conducting
 airway receive only brief exposure due to the effective  nature of mucociliary
 clearance mechanism.  Conversely the cells residing in or lining the alveoli  are
 in contact with the particles for a more protracted time due to  the longer
 residence in such structures.
   Due to their enormous phagocytic capability,  alveolar macrophages (AM)
 receive  the greatest potential exposure to inhaled  diesel  particles.  This is
 illustrated in a series of photomicrographs of lungs  from rats exposed by inha-
 lation for 19 weeks. 5 days a week. 7 hrs per day to  a chamber concentration  of
Approximately 4300 ygm/M  diesel exhaust particles  (DEP).   Figure 6 shows an
 increase both in number and in size of All's with  most distended  by phagocytosed
 DEP.  Another characteristic feature is the aggregations  of particle laden AM
 in alveoli adjacent to terminal bronchioles (Fig. 7).  Also shown in this figure
 is the increase in number of Type 2 pneumocytes  lining such occluded alveoli  and
 the presence of particles within alveolar and peribronchial  interstitial  tissues.
 In most  instances the particles reside within macrophages  and the macrophages
Fig. 6.  Section of rat lung exposed to 4300  ygm/M   for  19  weeks.   Number of
Type 2 cells  (arrow) increased in alveolus containing  diesel  particle laden
alveolar macrophages (arrowheads).  H and E.  X  320.
                             132

-------
 Fig.  7.  Section of rat lung exposed to 4300 ugm/M  for 19 weeks.  Clumped pig-
 ment  laden alveolar macrophages  (arrow) in occluded alveolus.  Increase in
 Type  2 cells  (arrowhead) lining  alveolar septa.  Interstitial pigment also
 present  in wall of terminal bronchiole.  H and E.  X 320.

 are in small  lymphatics.  Other  studies have shown that Type 1 pneumocytes are
 also  capable  of phagocytosing particles and thus they represent another cell at
                      10
 potentially high risk. "
   Although the precise mechanisms for subsequent transport to lung-associated
 lymph nodes is not known, it is  clear that DEPs are concentrated in both bron-
 chial associated lymphoid aggregates and in hilar lymph nodes (Fig. 8).  Ini-
 tially,  they  appear in histiocytes in the peripheral sinusoids but with time may
 be seen  in both the medullary and cortical sinousoids.  From these observations
 one may  conclude that biological responses will most likely occur within cells
 comprising the terminal bronchiole, the respiratory bronchiole, alveolar struc-
 tures, and in lung-associated lymph nodes.
   The information already available emphasizes the need for the model to con-
 sider particles that are (a) either free or within relatively mobile macro-
 phages,  and (b) particles sequestered in other tissues of the respiratory tract.
           	       20
At this  symposium, Soderholm   presented such a model  and the associated analysis
                             133

-------
Fig. 8.  Section of hilar lymph node from rat exposed  to  4300  ygm/M  for  19
weeks.  Particle laden macrophages (arrow) pack peripheral  sinusoids.   H  and
E.  X 130.
of the particle kinetics.  An additional key factor that must  be  considered in
the model is the dissociation of organic compounds from the  particles.  These
dissociated organic compounds may be retained adjacent to  the  particles or
transported to other sites within the respiratory tract, the regional  lymph
nodes or to other tissues.  Consideration must be given to the fate  of these
organic compounds, i.e. do they interact with cells and subcellular  organelles
in unaltered form, are they detoxified or activated?  At the present time there
are no direct observations of the dissociated organic compounds in vivo.  This
is not surprising recognizing that the compounds are probably  released very
slowly and at low levels.  Thus, it is difficult to envision experimental
approaches that will allow one to measure the disassociated  compounds  in vivo.
An insight into the processes that may be involved can be  obtained by studying
simpler systems.
FATE OF ORGANICS ASSOCIATED WITH PARTICLES
   Organic solvents such as dichloromethane readily remove the organic compounds
from the diesel exhaust particles and the extracted material is mutagenic to
bacterial and mammalian cells.  However, dichloromethane is  not typical of the
                            134

-------
solvents present within biological systems such as the respiratory tract.  This
raises a question as to the extent to which the organic compounds present in the
diesel exhaust particles may be extracted by materials that are more typical of
those found in the respiratory tract.  Apparently, organic compounds can be
removed by biological fluids, however, the removal process is not nearly as
efficient as with dichloromethane.  An example of the type of data that have
been developed is shown in Figure 9.  In this study a single cylinder diesel
engine was operated with 14C-labeled fuel and labeled exhaust particles col-
lected.  The   C-labeled material could be extracted from diesel exhaust par-
ticles by dichloromethane with a  very short extraction time.  Using serum,
however, only about half of the   C-labeled material was extracted during a
72-hour period and a much smaller amount was extracted with saline.  These data
suggest that in  the body, the organic compounds associated with particles are
likely to be removed from the particles very slowly.
                 I0°r     T •                     CH2CI2
                       6  12
   24           48
EXTRACTION TIME (nr)
                                                           72
                        14       14                                       ?1
 Fig.  9.   Extraction of   C from   C-labeled diesel  exhaust particles (Sun  )
    Having observed that biologically relevant materials such as serum can remove
 material  fron the diesel  exhaust particles, it is important to determine the
 mutagenicity of material  extracted by these more biologically relevant extracts.
              °2                23
 Brooks et.al..'"  and King  et_ al_.    have developed information to address this
                                                                 7?
 question.   Shown in Figure 10 are data obtained by  Brooks  ejt aj_.     In this
 particular study, Ames Salmonella strain TA-100 were exposed to extracts from
 1  mg  of diesel  exhaust particles per bacterial  culture plate.   Substantial
 mutagenicity was observed with the dichloromethane  extracted material  while  only
 a  low level  of mutagenicity was  observed for serum  and the levels  of mutageni-
 city  observed for material  extracted with lavage fluid or  saline were similar to
 the background  levels.  From the data presented in  Figures 9 and  10, it is
                              135

-------
                1500
              ill
                1000-
              (D
              z
             or
             UJ 500
             UJ
             cc
                                     CH2CI2
Lavage Fluid
                  '0   6        24            48
                             EXTRACTION TIME (hrs)
                   72
 Fig. 10.  Mutagenicity of diesel exhaust particle extracts  prepared  with dif-
 ferent solvents.

 reasonable to hypothesize that organic compounds were extracted  by the  serum and
 perhaps by the lavage fluid, however, these were rendered inactive to some
 extent in the test system.  Support for this is apparent from  the results of
           23
 King et^al_.   as shown in Figure 11.  In their studies, a high level of muta-
 genicity was apparent when the extract alone was evaluated  in  the Ames  test
 system, but the level of mutagem'city was markedly reduced  when  extract plus
 serum were evaluated.  When a protease was added, the activity level was inter-
 mediate between that of the extract alone and the extract plus serum.   One
 explanation for these observations is that the compounds responsible for the
 mutagem'c activity were inactivated by serum, but could be  released  by  the pro-
 tease and become available to the test system.  Very similar results were ob-
 tained when lung cytosol was added instead of serum.
     24
   Li   studied the cytotoxicity of diesel particle extracts using Chinese ham-
 ster ovary cells.  In his initial studies in which serum was included in thecui-
 ture media at levels typical of those used in cell culture  studies,  he  found
 little evidence for cytotoxicity.  More detailed studies, however, with varying
 levels of serum present in the culture media demonstrated that the serum had a
 protective effect, i.e. the relative survival of Chinese hamster ovary  cells was
 lowest when the lowest quantities of serum were present and survival was similar
 for extract treated and control cultures when more than 5 milligrams of serum
 protein were present per ml of culture media.  Using this same system,  he also
 studied the effect of addition of lung and liver S-9 fractions to the culture
media and found that these afforded a protective effect (Fig.  12).   This pro-
 tective effect was further enhanced when co-factors NADP (nicotinamide-adenine
                              136

-------
                 2000r      P"1
               LU
                cc
                LU
                co 1000
                Z
                HI
                CC
                                Extract
       Extract
       Serum
      Protease
Extract
Serum
       Extract
         Lung
       Cytosol
      Protease
 Extract
   +
  Lung
Cytosol
Fig. 11.   Influence of serum and lung cytosol on mutagenicity of diesel  exhaust
particle extracts.
                I00r
              LU
              cr
                 50-
                           No Treatment
                                                        Liver S9
                                                      •»• Cofactors
        Lung S9
      * Cofactors
                                    Lung S9
                             Particle
                             Extract
                               n
    n
                                                 Liver S9
Fig. 12.  Influence of liver or lung S-9 with and without co-factors on cyto-
toxicity of diesel exhaust particle extracts.

dinucleotide phosphate glucose-6-phosphate and magnesium) were added to the
culture media.
   Potential CO-mutagenicity, which may lead to an altered carcinogenic  response,
is another factor that must be considered in evaluating  the  release of  particle
                                              25
associated hydrocarbons and their effects.  Li   has addressed this question  by
evaluating the mutagenic response of Chinese hamster ovary cells  in in  vitro
cultures treated with benzo(a)pyrene alone, diesel exhaust particle extracts
                             137

-------
alone or the two materials in combination.  The  response  was  more than additive
suggesting an apparent synergistic response (Fig.  13).  Additional  studies of
this type are needed recognizing that our ultimate  objective  is to  assess the
health risks for people that are exposed to a wide  variety  of materials in ad-
dition to diesel exhaust.
                                                    Actual
300
CO
DC
O
S- 200
CO

"o
CO
Z 100
H
5

n
• » W » W*MI
DEPE
BaP

_

Projected
DEPE
BaP BaP
only | — i
1 I
DEPE '1
only |
r-i 1 i




















Fig. 13.  Co-mutagenicity of diesel exhaust particle extracts  and  benzo(a)pyrene.
HEALTH EFFECTS OF INHALED DIESEL EXHAUST
   In vitro studies with diesel exhaust particle extracts and  in vivo  assays
such as skin painting have identified potential health risks of exposure  to
diesel exhaust.  They have also provided insight into the mechanisms by which
effects might be produced or be minimized, i.e. detoxification.  However, stud-
ies with diesel particle extracts have typically involved delivery of  organic
compounds to cells over short times and at doses that are many orders  of  magni-
tude larger than those that could conceivably be encountered by cells  in  the
human respiratory tract.  The methods used frequently bypass normal protective
mechanisms of the body and the endpoints measured provide only indirect infor-
mation on the health effects of ultimate concern.  Thus, to provide a  more rele-
vant basis for predicting the potential health risks in man, it is important
that further studies be conducted in laboratory animals using  the  inhalation
route of exposure to observe the risks for causing cancer and  respiratory func-
tional disease as they are the major concerns for man.
   A number of studies in which animals have been exposed by inhalation to
diesel exhaust have been completed, are in progress or planned.  On first
                             138

-------
review, one is struck by the number of species and the range of exposure levels
studied.  However, closer examination reveals that only a few of the studies in-
volve observations for the total  life span, or the majority of the life span,
of the species being studied.  These studies are listed in Table 1.  In addition
to those listed, it is understood that studies of the health effects of diesel
exhaust in laboratory animals will  be initiated during the next year in Japan by
the Japan Automobile Research Institute, in West Germany by the Fraunhofer
Institute and in Geneva by the Battelle Memorial Institute.  Life span studies
with larger numbers of animals are especially important since the effects being
observed are subtle and are likely to occur in low incidence and late in life.
TABLE  1
MAJOR  LONG-TERM STUDIES OF THE HEALTH EFFECTS OF DIESEL EXHAUST COMPLETED OR IN
PROGRESS
Laboratory
Environmental
Protection Agency
Fraunhofer
Institute
General Motors
Lovelace Inhalation
Toxicology Research
Institute
Battell e-Northwest

Southwest Research
Institute
Reference
26
27
28

29
30
31
32
Species
Chinese Hamster,
Mice, Rats, Cats
Syrian Hamsters
Rats, Guinea Pigs

Mice, Rats
Rats
Syrian Hamsters
Syrian Hamsters,
Rats, Mice
Particle ,
Concentration (yg/m
6000-12000
4200a
250, 750, 1500

350, 3500, 7000
8300
7300
1:60, 1:120,
1:360 dilution5
 Also, exposures to gaseous emissions only without particles
 Particle concentrations not given
   In studies completed to date, all of the observed health effects have been
non-neoplastic in nature.  Although there are some species-related differences,
in general, the responses have been similar in all laboratory animals.  After
inhalation the biological sequence of events starts with the phagocytosis of
particles by alveolar macrophages (AM).  With time, there is an increase in both
the number and size of AMs and an increasing concentration of DEP within their
cytoplasm (Fig. 6).  Type 2 pneumocytes also increase in number and size within
alveoli containing pigment-laden macrophages.  There is no evidence to suggest
                              139

-------
 that Type 2 cells participate  in  the  clearance  of DEP,  but both neutrophils and
 eosinophils do  appear  to  be  recruited and  to phagocytize particles under condi-
 tions of high pulmonary loading. 19'33 There is  ultrastructural evidence that
 Type 1  pneumocytes also phagocytize DEP, particularly under conditions of high
 levels  of particle exposure. 19 With  time,  particle  laden  AM form dense aggre-
 gates within alveoli, most notably adjacent to  terminal  bronchioles (Fig.  7).
 The surrounding tissue response to the macrophage clusters is highly variable.
 In some instances, there  was a proliferation of  interstitial  cells and an  in-
 crease  in interstitial reticulin but  in other cases,  there was no elicited re-
 sponse.  Particles are also  translocated from alveoli to the interstitium  where
 they are usually contained in  interstitial  nacrophages.  Finally, it has been
 shown that particles are  transported  to local and regional  lung-associated
 lymphoid tissues (Fig. 8).  Although  at later times  these  tissues concentrate a
 significant mass of DEP within histiocytes,  there is  no evidence  that other
 surrounding cells are affected by their presence.  However,  there is an indica-
 tion that these nodes may have altered immunological  competence.
   The  responses in lung  and lymph nodes observed to  date  represent the usual
 response of lung to inhaled relatively insoluble  particles.   Longer-term obser-
 vations will be required  to ascertain whether the lesions  remain  the same,  or
 whether with time they become more functionally significant.   Substantial  effort
 has been directed to evaluating non-morphological  responses,  for  example,  bio-
 chemical and physiological alterations.  The  biochemical changes  observed  in
 tissues and airway fluids have in general been transient in  nature suggesting
 injury  followed by adaptation or repair.  The physiological  changes  have been
 minimal to non-existent even at the highest  exposure  levels.
   The  lack of outstanding effects, and especially the lack  of carcinogenicity,
 should  be interpreted (and extrapolated to  man) with  caution  for  several rea-
 sons.   First, as noted earlier, only  a few  of the studies  have involved life
 span observations.  Study of exposed  aninals  for  their life  span  provides  the
 best opportunity for detecting late-occurring effects.  Second, all  of the
 longer-term observations  have been made on  rodent species.   Ideally,  one would
 like to study not only rodent species for their  life  span,  but also  longer-lived
 species.  It is generally felt that rodents  have  more rapid  clearance,  and  con-
sequently, lower retention of particles than  do other species  including man.
Thus, per unit of exposure, the actual dose  to tissue may  be  less in the rodents
than would be the case for man.  The  best basis for extrapolation between
species will  probably be  particle or  chemical dose-response  relationships
rather than exposure atmosphere - response  relationships.   This emphasizes the
                              140

-------
need for periodic measurements of particle burdens in animals from the long-term
exposure studies.  Third, all of the laboratory animals have been- in excellent
health when placed on study and have been maintained under optimum conditions.
It would be desirable to study some animals whose health status has been altered
(for example, emphysematous animals) so they might be more representative of
potentially sensitive individuals in human populations.  Fourth, with the excep-
                             27
tion of the Fraunhofer study,   the treated populations have only been exposed
to whole diesel exhaust.  The Fraunhofer study is noteworthy in that it involved
exposure of animals to diesel exhaust with and without particles and pre-treat-
ment of animals with known carcinogens.  The former should provide insight into
the relative role of gaseous and particulate emissions in producing health
effects.  The pre-treatment of animals with known carcinogens may provide in-
sight into those situations in which people are concurrently exposed to poten-
tially toxic materials other than diesel exhaust.  Finally, the majority of
studies have been conducted at very high exposure levels.  The effects at these
levels may be dominated by alterations in protective mechanisms that may not be
as significant at lower exposure levels.
SUMMARY
  Our present knowledge of the health effects of diesel exhaust particles can
be summarized as follows:
     1.   Diesel exhaust particles are very small in size and consist of a
carbonaceous core with a myriad of adsorbed hydrocarbon compounds that are
readily extracted with organic solvents.
     2.   The particle extracts are cytotoxic and mutagenic in in^ vitro
bacterial and mammalian cell cultures.
     3.   The particle extracts are carcinogenic when painted on mouse skin
along with a suitable promoter.
     4.   Inhaled particles readily deposit in the respiratory tract, a portion
is rapidly cleared and a substantial portion is retained for long periods of
time (over 100 days) in the lung.
     5.   Adsorbed hydrocarbon compounds slowly dissociate from the particles
in biological  media and presumably in the lung.
     6.   Detoxification mechanisms act on the hydrocarbon compounds released
from the particles to minimize effects in in vitro systems and presumably jm
vivo.
                              141

-------
     7.   There is morphological and biochemical evidence  of lung tissue injury
and adaptation or repair after inhalation of very  high  levels  of diesel  exhaust.
     8.   To date, the longest term studies have not  been  demonstrated carcino-
genicity or major physiological changes.
RESEARCH HEEDS
                                                                            35,36
  From consideration of potential levels of diesel exhaust exposure for  people
and the foregoing summary of our current knowledge of diesel exhaust-induced
health effects, it appears highly unlikely that such  exposures  will  produce
substantial health effects.  Indeed, there may be  no  health effects  that will be
attributable to diesel exhaust emissions.  However, this should be viewed as a
provisional assessment pending development of additional information.  Some of
this information will be obtained from research now in  progress,  while some will
require the initiation of new studies.  The following information needs  are
viewed as being of highest priority for obtaining  an  improved  assessment of the
health effects of diesel exhaust emissions in people.
     1.   Additional information on the transport  and transformation of  particle
associated hydrocarbons from the point of release  to  inhalation or ingestion by
man.  This should include studies of the environmental  fate of  the several
hundred thousands of tons of diesel soot emitted per  year  into  the atmosphere.
     2.   Improved knowledge of the long-term retention of particles and their
associated hydrocarbons in the respiratory tract of man or species with  particle
retention characteristics similar to those of man.  Emphasis should  be placed on
evaluating the "microdosimetry" fpr^ tissue structures such as  the bronchi  and
bronchioles that are known to be especially sensitive to cancer induction, and
alveolar tissue, known to be sensitive to destruction,  often resulting in major
functional alterations.
     3.   Information on the fate of major constituents of diesel  exhaust par-
ticles in the body after inhalation.  The results  of  such  studies  using  exhaust
particles with added concentrations of specific1 compounds  should aid in  defining
exposure atmosphere - dose relationships.  These may  serve as  a prelude  to the
conduct of exposure-dose-response studies with such particles.   Such information
should be obtained at several exposure levels to define the influence  of expo-
sure levels, and possible associated tissue alterations, thereby  providing a
better basis for extrapolating to lower exposure levels.
     4.   Information on the response of model cell and tissue  systems derived
from respiratory tract epithelium.  Studies with such model  systems  should aid
                              142

-------
in linking In vitro results to those obtained in the whole animal.  They should
include the study of diesel exhaust particles and known mutagens and carcinogens.
     5.   Information on late-occurring tissue responses at various exposure
levels to complement the information currently available for high levels of
exposure and exposure times that have generally been one year or less.
     6.   Information on the response of individuals with altered health status
that may render them more sensitive to diesel exhaust.  This should include
appropriate models of important respiratory diseases of man.
     7.   Information on the response of individuals following concurrent expo-
sure to other materials known or suspected to be pulmonary toxicants.  Studies
have previously been conducted with animals exposed to diesel exhaust and coal
or uranium ore dust.  Similar studies should be conducted with oil shale dust
in view of likely occupational exposures to shale dust and diesel exhaust.
Recognizing the central role of cigarette smoking in human lung disease, the
effects of smoking in combination with diesel exhaust exposure should be inves-
tigated.
     8.   Additional information is needed on tissue burdens of diesel soot
particles and their organic constituents as a function of exposure level and
duration of exposure in animals being studied in long-term experiments of the
health effects of diesel exhaust.  Such information will provide a basis for
interpreting and extrapolating results between species on the basis of exposure
atmosphere-tissue-dose-response relationships.
     9.   Additional epidemiological information on any human populations with
sufficiently high exposure levels and population size to warrant study.
AN ANCILLARY MESSAGE
  It is our opinion that the substantial amount of information obtained on the
health effects (or lack of effects) of diesel exhaust particles and particle
extracts conveys a message that extends well  beyond the question originally
asked - does exposure of people to diesel exhaust result in increased health
risks?  The message comes from consideration on the one hand of the cytotoxic,
mutagenic and carcinogenic properties of diesel exhaust particle extracts.  And
on the other hand, recognition that inhalation and deposition of rug quantities
of diesel  exhaust particles in the respiratory tract of rodents has not elicited
a carcinogenic response.  These apparently contradictory results indicate the
need for further research to evaluate the utility of using short-term in^ vitro
studies for predicting late-occurring effects _j_n_ vivo and thus to recognize the
                              143

-------
limitations of such extrapolations.  The research conducted  on  diesel  exhaust

has indicated both the usefulness and the short-comings of short-term  tests as

predictors.  It has also provided insights that help explain  the  apparent con-
tradiction.  Additional mechanism oriented research will not  only aid  in better

understanding the likely human effects of exposure to diesel  exhaust,  but im-

prove our ability to predict the health effects of other complex  materials to

which man may be exposed.

ACKNOWLEDGEMENTS

   This work was performed under United States Department of  Energy  Contract

No. DE-AC04-76EV01013.  The information summarized in this paper  represents the
efforts of many scientists and technicians conducting research  both  in the
United States and abroad.  Grateful acknowledgement is made of  the contribution
of these colleagues whose research has profoundly influenced  our  knowledge of

the health effects of diesel exhaust and has done so in a remarkably short
period of time.


REFERENCES

 1.  U.S. Environmental Protection Agency (1980) 1977 National  Emissions Report:
     National Emissions Data System of the Aerometric and Emissions  Reporting
     System.
 2.  National Research Council (1981) Health Effects of Exposure  to  Diesel Ex-
     haust Report: Impacts of Diesel-Powered Light-Duty Vehicles.
 3.  Pepelko, W. E., Danner, R. M. and Clarke, N. A. eds. (1980)  Health Effects
     of Diesel Engine Emissions.
 4.  Harris, J. E. (1981) Report to the Diesel Impacts Study  Committee, National
     Research Council, National Academy Press, Washington, DC.
 5.  Pitts, J. N., Jr., Cauwenberghe, K. A. van, Grosjean, D.,  Schmid, J. P.,
     Fitz, D. R., Belser, W. L., Jr., Knudson, G. B. and Hynds, P. M.  (1978)
     Science 202, 515-519.
 6.  Claxton, L. and Barnes, H. M. (1980) in Health Effects of  Diesel  Engine
     Emissions, Vol. 1, Pepelko, W. E., Danner, R. M. and Clarke, N. A. 309-326.
 7.  Lippmann, M. (1977) in Handbook of Physiology, Section 9,  213-232.
 8.  Chan, T. L. and Lippmann, M. (1980) Am. Ind. Hyg. Assoc. J.  41, 399-409.
 9.  Stahlhofen, W., Gebhart, J. and Heyder, J. (1980) Am. Ind. Hyg. Assoc. J.
     41, 385-398.
10.  Cuddihy, R. G., Brownstein, D. Q., Raabe, 0. G. and Kanapilly,  G. H. (1973)
     J. Aerosol Sci. 4, 35-45.
11.  Task Group on Lung Dynamics (1966) Health Phys. 12, 173-207.
12.  Yen, H. C. and Schum, G. M. (1980) Bull, flath. Biol. 42, 461-480.
13.  Uolff, R. K., Kanapilly, G. M., DeNee, P. B. and FlcClellan,  R.  0.  (1981)
     J. Aerosol Sci. 12, 119-129.
14.  Cohen, D., Arai, S. F. and Brain, J. D. (1979) Science 204,  514-516.
15.  Chan, T. L., Lee, P. S. and Hering, W. E. (1981) J. Appl.  Tox.  1, 77-82.
16.  GHffis, L. C., Wolff, R. K. and Mokler, B. V. (1981) Personal  Communication.
17.  Strom, K. A. and Garg, B. D. (1981) Personal Communication.
18.  Thomas, R. G. (1972) in Assessment of Airborne Particles,  Mercer, T. T.,
     Morrow, P. E. and Stober. W., eds., Charles C. Thomas (oublisher),
     Sprinnfield, IL, 405-420!
                             144

-------
L9.  Barnhart, M. I., Chen, S. T., Salley, S. 0. and Puro, H. (1981) J. Appl.
     Toxicol. 1, 88-103.
20.  Soderholm, S. (1981) Personal Communication.
21.  Sun, J. (1981) Personal Communication.
22.  Brooks, A. L., Wolff, R. K., Royer, R. E-, Clark, C. R., Sanchez, A. and
     McClellan, P.. 0. (1980) in Health Effects of Diesel Engine Emissions,
     Pepelko, W. E., Danner, R. M. and Clarke, N. A., eds., Vol. 1, pp 345-358.
23.  King, L. C., Kohan, M. J., Austin, A. C., Claxton, L. D. and Huisingh, J.
     L. (1981) Environ. Mutagenesis 3, 109-123.
24.  Li, A. P. (1981) Toxicol. Appl. Pharmacol. 57, 55-62.
25.  Li, A. P. and Royer R. E. (1981) Mutation Research (in press).
26.  Pepelko, W. E. (1980) in Health Effects of Diesel Engine Emissions,
     Pepelko, W. E., Danner, R. M. and Clarke, H. A. eds., Vol.  2, pp. 673-680.
27.  Heinrich, U., Stober, W. and Pott, F. (1980) in Health Effects of Diesel
     Engine Emissions, Pepelko, 'J. E., Danner, R. M. and Clarke, U. A. eds.,
     Vol. 2, pp 1026-1047.
28.  Schreck, R. M., Soderholra, S. C., Chan, T. L., Hering, !-J. E., D'Arcy,
     J. B. and Smiler, K. L. (1980) in Health Effects of Diesel  Engine Emissions,
     Pepelko, W. E., Danner, R. ',}. and Clarke, N. A. eds., Vol.  2, pp 573-591.
29.  McClellan, R. 0. (980) Diesel Exhaust Emissions Toxicology Program,
     Lovelace Medical Foundation Report-31.
30.  Cross, F. T., Palmer, R. F., Filipy, R. E., Busch, R. H. and Stuart, B. 0.
     (1978) Study of the Combined Effects of Smoking and Inhalation of Uranium
     Ore Dust, Radon Daughters and Diesel Oil Exhaust Fumes in Hamsters and
     Dogs.  Pacific Northwest Laboratory Report-2744.
31.  Karagianes, M. T., Palmer, R. F. and Busch, R. H. (1981) Am. Ind. Hyg. Assn.
     J. 42, 382-391.
32.  Springer, K. (1981) Personal Communication.
33.  White, H. J. and Garg, B. D. (1981) J. Appl. Toxicol. 1, 104-110.
34.  -Bice, D. E. (1981) Personal Communication.
35.  Williams, R. L. and Chock, D. P. (1980) in Health Effects of Diesel Engine
     Emissions, Pepelko, W. E., Danner, R. [1. and Clarke, N. A., eds., Vol. 1,
     PP 3-33.
36.  Cuddihy, R. G., Seller, F. A., Griffith, W. C., Scott, B. R. and McClellan,
     R. 0. (1980) Potential Health and Environmental Effects of Light Duty
     Vehicles.  Lovelace Medical Foundation Report-82.
                              145

-------
 EPA STUDIES  ON THE TOXICOLOGICAL EFFECTS OF INHALED DIESEL ENGINE EMISSIONS
 WILLIAM  E. PEPELKO
 Health Effects Research  Laboratory,  U.S. Environmental Protection Agency,
 26 West  St.  Clair  Street,  Cincinnati,  Ohio 45268, USA
 INTRODUCTION
   Rapidly  rising  fuel costs in. recent years  have resulted  in  attempts to in-
 crease fuel  efficiency in passenger cars.  Because the  diesel engine uses less
 fuel  and has proven to be practical for use  in heavy trucks,  buses,  and farm
 machinery,  the interest in utilizing  the diesel engine  as a power  source for
 light duty  vehicles has  increased.  The Environmental  Protection  Agency  has
 estimated that up  to 25%  of new U.S. passenger cars could  be diesel powered by
 1985.    The  Department  of  Transportation  made  a  more  modest  estimate of
 10%.    Even at  the  lower  estimate,  a  larger  number  of  light  duty  diesel
 equipped vehicles  will be manufactured, and  combined  with  heavy duty  trucks
 and buses will make a   significant  contribution to environmental pollution.
   The exhaust from diesel engines contains most of  the pollutants common to
 the  gasoline  engine.   These include  carbon dioxide, carbon monoxide,  nitric
 oxide,  nitrogen dioxide,  ozone, sulfur  dioxide,  alkanes, alkenes,  aldehydes
 and many organic oxidants.  The toxicological  effects of most  of  these pollu-
 tants have  been investigated  in considerable  detail.   The  particles  emitted
 from diesel  engines,  however,  differ  greatly  in both quantity  and composition
 from those produced by gasoline engines.   Even a well tuned  engine produces 20
 to  100   times  more particulate matter than a  catalyst  equipped  gasoline  en-
 gine. '    While  the  gasoline  engine  particles  are  primarily  composed of
 sulfur compounds,  diesel  particles consist of  a  carbonaceous material with a
 large variety  of  high molecular  weight organic compounds  adsorbed' onto  the
particle  surface.    Over  80 such  compounds have  been  identified  in our  own
 studies.
   It is difficult,  if not  impossible,  to  predict accurately  the  carcinogenic
and other  toxicological effects of the diesel  exhaust  particulates.   Quanti-
tative measurements of  all the  components are not available.  Some of  the com-
pounds present have  not even been  identified.   Even if accurate,  quantitative
component data were available,  effects would be  influenced by  bioavailability,
efficiency  of  activation  or detoxification mechanisms,  transport and  elimi-
nation as  well as possible mutually  synergistic  and/or inhibitory  actions of
the many compounds present  on  the particle.   Thus,  in order  to  produce ade-
quate data  for risk assessment,  a large scale animal  exposure study  of over
                              146

-------
two years duration utilizing a  variety  of  animal species and testing a variety
of endpoints was planned by the Environmental Protection Agency at Cincinnati.
   The inhalation route was used as  the primary  means of exposure.  During the
second year  of the  study,  however,  intraperitoneal  injections  and intratra-
cheal  instillations  of diesel  exhaust  particulate were used  increasingly to
produce  greater  exposures,  and  in some  cases,  to   produce  dose  response
curves.  A single exposure  level was set for the inhalation studies because of
the limited  chamber  space  available for  the  large  number of  planned  experi-
ments  and because a  single  data point was considered adequate  for cancer risk
assessment using the linear no-threshold model.
   Since cancer risk assessment was  to  be  emphasized,  a near maximum tolerated
dose was  selected  to achieve the greatest chance of  producing a  positive re-
sponse.  In a preliminary 60 day study,  exposure to a  1:14 dilution of  exhaust
for 20 hours/day  resulted in decreased weight gains  and food  intake.   it ap-
peared that this exposure regime over a long time-period might  result  in mor-
tality and shortened life span. As  a result, the dilution  ratio  (DR)  was ad-
justed to produce  a  particulate concentration of  6 mg/m  (DR =  1:18)  and ex-
posure time was shortened to 8 hours/day.   After the first year  of exposure,
it became  apparent that  the  animals had  adapted to  the exposure  conditions
with little gross evidence  of  stress.  In  order  to approach  the  maximum tole-
rated  dose more closely the dilution ratio  was  decreased to produce a  parti-
culate  concentration  of  12  mg/m"
until  the completion of the study.
culate  concentration of  12  mg/m .    This  level  of  exposure  was  maintained
EXPOSURE CONDITIONS
Exposure facility
   Exhaust was  produced by  one  of two  Nissan CN  6-33,  6 cylinder  198 cubic
inch  (3.24  liter)  displacement  diesel engines coupled  to a  Chrysler  torque-
flite model A727 automatic transmission.   The  engines  were mounted on an Eaton
Model  758-DG dynamometer  and  run  in  the  federal  Short  Cycle Mode.    After
mixing with  filtered  and conditioned air  in  a dilution tube,  the exhaust  was
passed  into  a large  mixing  chamber,  and  from there  into the  exposure  cham-
bers.   The   24  exposure  chambers  were  constructed  of  stainless steel  with
plateglass windows and  each  had an interior  volume of 100 cubic  feet  (2.8 cu
m).   Animals  remained  in  the  chambers  continuously  during   the  exposure
period.  They were housed  in wire  cages  with the exception of  cats which were
allowed to roam free.
                            147

-------
Aerometry
   An  on  line data  acquisition  system  was  utilized for  measurement  of 6
gases:   carbon dioxide  (CO.),  carbon  monoxide  (CO), sulfur  dioxide  (SO),
nitric  oxide  (NO),  nitrogen  dioxide   (N02),  and  total  hydrocarbons  (THC).
The atmospheres  of each of  12 exhaust  and  4 clean  air chambers were  sampled
hourly to  give 8  data  points/day/chamber for each gas.  Acrolein,  formaldehyde
and total  aliphatic aldehydes  were monitored  daily from  several  (usually 4)
exhaust chambers.  Particulate mass concentrations were determined  daily using
samples collected  on glass fiber filters  from each  of the exhaust containing
chambers.
   Concentrations of the automatically  monitored pollutants, aldehydes, parti-
culate mass, as well as dilution ratios are  shown in Table 1.   Actual particu-
late concentrations were  slightly  greater  than  the  planned level  of  6 mg/m
during the first  61 weeks and  slightly less  than 12  mg/m   during the second
half of the study.  The ratio of the two concentrations was thus nearer to 1.8
than 2.0.   The ratios  of  the  various pollutant  gas concentrations for  the 2
periods generally ranged from 1.6 to 1.8 if  the background concentrations were
first  subtracted.   The  only major exception  was  SO   which  was  undoubtedly
influenced by varying  sulfur content in the fuel.   For further details of the
exposure system and aerometry see Hinners, et al.

TABLE 1
CONCENTRATIONS OP HOURLY  MONITORED GASES,   ALDEHYDES,  PARTICULATE MASS  AND
DILUTION RATIOS IN THE EXPOSURE CHAMBERS
                             Clean Air Chambers        Exhaust  Chambers
Compound              Units     Weeks 1-124      Weeks 1-61      Weeks 62-124
C02
CO
THC
NO
N02
S02
Acrolein
Formaldehyde
Total aliphatic
aldehydes
Particulate Mass
Dilution Ratio
%
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm

mg/m3
— ™"
0.04 + 0.002*
2.20 + 0.50
2.82 ••• 0.30
0.05 + 0.04
0.03 + 0.03
0.03 + 0.02
0.00 ~
0.00
0.00

0.00
—"••"•
0.30 + 0.04
20.17 + 3.01
7.93 + 1.42
11.64 + 2.34
2.68 + 0.80
2.12 + 0.58
0.025 + 0.003
0.106 + 0.029
0.177 i 0.043

6.34 + 0.81
18.16 + 1.72
0.52 + 0.04
33.30 + 2.94
11.02 + 1.04
19.39 + 3.80
4.37 + 1.19
5.03 + 1.03
0.034 + 0.009
0.251 + 0.059
0.338 + 0.057

11.70 + 0.99
9.37 ± 1.13
"Standard deviation of weekly means.
                              148

-------
Particle characterization
   Particle size was  estimated  by collection of  the  particulates on a nucleo-
pore membrane  filter  and by  subsequent examination using  a  scanning electron
microscope.  Generally,  the  basic unit was found  to  be 0.1 micron  or  less in
size.  These units were  intermixed  with larger  particles,  consisting of agglo-
merates of the basic  units,  reaching almost a  full micron  across.   Our obser-
vations were in general  agreement with previous studies indicating  that 90% of
the particles  (by  mass)  are less than 1  micron diameter and that  50%  are 0.3
micron or less.  Thus, they are almost all in the respirable range.
   Chemical  analysis  of  the diesel  exhaust particulate  was  carried out  by
              9
Pitts,  et al.    Glass  fiber  filter  samples  were  extracted  with  a  solvent
system containing  benzene,  methylene  chloride  and methanol  1:1:1  by  volume.
The  extract  was further  separated   into  acids  (27.7  rag),  bases  (2.0 mg)  and
neutrals  (171.3 mg).   The fractions were further divided  into aliquots  which
were used for qualitative and quantitative GC-MS analysis.
   Seventy compounds  were  detected qualitatively.   These  included  aliphatic
hydrocarbons,  polynuclear aromatic  hydrocarbons,  alkylated  polynuclear  aro-
matic  hydrocarbons,  aliphatic  acids,  aromatic acids  and   a  variety of  other
compounds not  falling in the above  groups.   Quantitative analysis  was  carried
out  for   12  compounds deemed  most  important  by  concentration  and  activity.
These are listed in Table 2.

TABLE 2
CONCENTRATION OF 12 SELECTED COMPONENTS OF DIESEL EXHAOST EXTRACT
            Compound                                   Concentration
                                                          ugm/gm
      Phenanthrene                                        145.2
      Fluoranthene                                        155.8
      pyrene                                              198.0
      Benz(a(anthracene                                    53.8
      Chrysene                                             71.6
      Benzo(k+b)fluoranthenes                              77.8
      Benzo(e) pyrene                                       28.6
      Benzo(a)pyrene                                       15.9
      Perylene                                              3.5
      Indeno(l,2,3-Cd)fluoranthene                         10.9
      indenod, 2,3-Cd) pyrene                               14.8
      Benzo(ghi)perylene                                   21.1
                             149

-------
 EXPOSURE ASSESSMENT
    All  exposures have  been terminated.  Most  of the  experiments are complete
 and have either  been reported previously or will  be  at the present Diesel Sym-
 posium.    Some  important  exceptions   include  a   detailed   biochemical  and
 morphometric  analysis of the lungs of  cats exposed  to diesel exhaust (DE)  for
 27  months.  A multigeneration reproduction study in rats  has been completed,
 but a compilation of results is still  in progress.   Also,  positive control  ex-
 periments for sperm quality assessment in  cats exposed to  DE is still in pro-
 gress.
    The  types  of  experiments that were  carried  out generally  fall into 3 cate-
 gories  emphasizing  carcinogenic,  mutagenic  or  toxicological endpoints.   The
 experiments  designed to  assess cancer  risk included  lung tumor  induction  in
 Strain  "A" and  SENCAR  mice and liver  island  induction  in rats.  Mutagenesis
 studies  were  designed  to  detect possible  increases  in heritable mutations  in
 mice and fruit flies;  sister  chromatid exchange  in  Syrian Hamster  lung cells
 and peripheral lymphocytes in mice and Chinese hamsters; micronuclei and cyto-
 genetic  changes  in peripheral  lymphocytes of  mice  and  Chinese  hamsters  and
 finally;  sperm  abnormalities  in mice  and  cats.   Toxicological  endpoints  in-
 cluded  behavioral and neurophysiological changes  in  rats,  resistance to  infec-
 tion in mice,  pulmonary function changes  in Chinese hamsters and cats,  repro-
 ductive  effects  in mice,  teratological  effects  in  rats  and  rabbits,  and a
 variety  of biochemical measurements in the lung.  A  discussion of genotoxicity
 and sperm quality will  be  covered  in  a  separate chapter.
    In  the following  studies  exposure levels will be referred  to in  terms  of
 either  6 or  12 mg/kg particulate matter.   A few studies will also be reviewed
 that took place during the preliminary 60-day exposure period.   Although  the
 engines were  run at a set  dilution ratio of 1:14  in  the preliminary study,  the
 particulate  mass  averaged  6-7 mg/m .   Thus,  these  exposures  will  also   be
 listed  as 6  mg/m  particulate,  but can be  differentiated  from  the  main study
 by  the fact that the engines were  run 20 hr/day instead of  8.

 Cancer risk assessment
    Lung  tumor  induction in Strain "A" mice exposed via  inhalation.  The  Strain
 •A' mouse was selected as a cancer risk assessment  model because  of  it's sen-
 sitivity  to induction of lung tumors, the  relatively short  exposure period  re-
quired,  and  the  large  volume of  background data available.10  The  mice were
 exposed  from  about  6 weeks to 9 months of  age unless stated otherwise. They
were    then    sacrificed    and    lung   tumors   counted    using    Standard
                             150

-------
Mechc^s.  '  '     The  first  experiments of  this series  were carried  out by
Orthoefar,  at  al.   at  a  particulate concentration  of  6  mg/m .   No  signi-
ficant changes in  lung tumor  incidence could be  detected  in 400 males exposed
to DE,  although  in 100 of  these  animals that were sacrificed at  11  months of
age instead of 9  there was a  tendency for a  lower  tumor incidence as compared
to controls.   In  the second experiment  170  females per group were  exposed to
either DE or  clean air.   One  half the mice  in each group  received  an initi-
ating dose  of  1 mg urethane prior to the start of  exposure.  The  incidence of
lung tumors in both initiated and non-initiated  DE exposed  mice  was slightly
but significantly  greater  than those  of  clean  air controls.  The  tumor inci-
dence in controls,  however, was  very  low compared  with historic controls ren-
dering the  results  inconclusive.
   Since the previous  report  a series of  3  experiments were carried  out  at a
particulate concentration  of  12  mg/m  .   In  the first  experiment  half the ex-
posed and half the  clean air controls  received  an initiating dose  of  5 mg ure-
thane.  In  the second  experiment  none of  the mice received  urethane but were
exposed until  12 months of age  instead of  9.   In  the final  experiment  the
light cycle was altered  so the chambers would be dark  during engine  operation
with the animals presumably awake, active and respiring at a higher level.
   The  results are  shown  in  Table  3.  In  every case  the  average  number  of
tumors per  mouse  was less  in DE  exposed mice  compared with  their  respective
clean air controls, and  differences were significant with only  one exception.
The lower incidence values in exposed mice was particularly noticeable  in the
ones pretreated with urethane.
   The reason  for  the  decreased  tumor incidence  in DE  exposed  mice  is uncer-
tain,  possibly,  diesel exhaust inhalation results  in an inhibition of the in-
duction of  enzymes responsible for  converting  procarcinogens to  their  active
forms.   It  is  also possible  that  the   immunocompetence  of  the  animals  was
altered by the inflammatory reaction to deposited materials.
   Lung tumor induction by  IP  injection  in Strain °fi" mice.   This  study is one
of several  being conducted by the EPA to compare the relative carcinogenicity
of diesel exhaust with that of other environmental  pollutants known to be car-
cinogenic in humans.   Comparisons were  made between exhaust  samples from the
Nissan diesel, an  Oldsmobile Diesel  run at   a  steady 40  mph, cigarette  smoke
condensate,  coke  oven mains and roofing  tar  condensate.
                             151

-------
TABLE 3
EFFECTS OF DIESEL EXHAUST INHALATION ON LONG TUMOR INDUCTION IN STRAIN  "A" MICE

            Age   Ilium!-      Survivors/  Mice with
Treatment    (Mo)  nation    Sex    Initial    Tumors     P     Tumors/Mouse   P
Clean air 9


Exhaust 9


Clean air 9
+ 5 mg
urethane
Exhaust + 9
5 mg
Drethane
Clean air 12
Exhaust 12
Clean air 9


Exhaust 9


Light


Light


Light


Light


Light
Light
Dark


Dark


H
F
H6F
H
F
MSF
M
F
MiP
M
F
M&F
H
H
H
F
MSF
H
F
H&F
44/45
43/45
87/90
37/45
43/45
80/90
38/45
37/45
75/90
39/45
36/45
75/90
38/
44/
97/108
140A42
237/250
111A15
139/143
250/258
10
11
21
5
4
10
32
34
66
26
16
42
22
11
28
31
59
13
9
22



NS
.05
.05



.10
.0001
.0001

.01



.01
.001
.0001
.227 + .071
.349 + .080
.287 + .054
.189 + .077
.093 ± .080
.138 + .056
2.368 + .263
3.243 + .314
2.800 7 .260
1.025 + .260
0.861 + .316
0.947 + .206
0.684 + .090
0.250 + .083
3.24 ± .047
.234 + .034.
.271 + .028
.135 + .047
.065 + .034
.096 + .027



.130
.05
.055



.001
.0001
.0001

.001



.01
.001
.0001
   The  particulate from  our own  study was  collected  from  the large  mixing
chamber on  Pallflex T60 A20  (teflon  coated)  filters  during the course  of  the
inhalation  study.   The  Oldsmobile sample  was  provided  by the  Environmental
Monitoring and  Support  Laboratory  of EPA.   Both  samples  were soxhlet extracted
with  dichloromethane.   Cigarette  smoke   condensate  was  produced  from  the
Kentucky reference  2  RI cigarettes.  Coke   oven  mains  and  roofing  tar were
collected  using procedures  described   by  Huisingh,  et  al.     Strain  A/Jax
mice approximately  8 weeks of age were  injected 3x weekly for  8 weeks with  the
test substances.   They were  sacrificed at 9 months  of age and examined  for
presence of pulmonary adenomas.
   Two separate experiments were carried out.   Results  are shown  in  Table 4.
The increase in tumor rates in the mice injected with urethane was comparable
                              152

-------
to  that  reported  in  earlier  studies    and  showed  that  the  mice  were  re-
sponding  normally.   The lack  of  a consistent  increase  in  tumors  in mice  in-
jected with  the  environmental pollutants indicated that either the carcinogens
present were vary  weak or  that the concentration  of  carcinogens reaching  the
lungs was below detectable  limits.
TABLE 4

EFFECTS  OF INJECTED POLLUTANTS ON THE  INDUCTION OF LUNG  TUMORS  IN STRAIN  "A"
MICE
Experiment I
Group Sex Tumors/mouse
Dninjected M
controls F
Vehicle M
controls F
Ore thane M
F
Nissan par- M
ticulate F
Nissan M
extract F
Olds M
extract F
Cigarette M
smoke F
Coke oven M
F
Roofing tar M
F
*Signif icantly
controls (P < .
"Significantly
0.6 + 0.2
0.6 + 0.2
0.9 + 0.5
0.7 + 0.2
22.5 + 1.9*
21.8 + 1.5*
0.4 + 0.1
0.5 + 0.1
1.4 + 0.3**
1.0 + 0.3
0.4 + 0.1
0.8 + 0.3
1.1 + 0.2
1.2 + 0.3
0.5 i 0.2
0.7 + 0.2
0.7 4; 0.3
different from
05) .
different from
Dose
_
0.05 ml/inj
20 mg/mouse
4 mg/inj
1 mg/inj
1 mg/inj
.20 mg/inj
.02 mg/inj
.02 mg/inj
Sex
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
Experiment
Tumor s /mouse
0.2 _+ 0.1
0.4 + 0.2
0.5 + 0.2
0.3 + 0.1
7.3 + 0.7*
11.3 ± 0.9*
0.3 + 0.1
0.3 + 0.1
0.4 + 0.1
0.7 + 0.1
0.4 ± 0.1
0.4 + 0.1
0.2 + 0.1
0.4 + 0.1
0.9 + 0.2**
0.7 4- 0.2
0.3 + 0.1
II
Dose
_
0.05 ml/inj
10 mg/mouse
2 mg/inj
1 mg/inj
1 mg/inj
. 20 mg/inj
.02 mg/inj
.02 rag/in]
both vehicle controls and unin3ected
uninjected controls (P < .05).
Tumor induction in Sencar Mice. This study
was
designed with several pur-
poses in mind;   (a)  to  evaluate the effects of chronic DE  exposure  for a near
lifetime upon  both pulmonary and  nonpulmonary tumorigenesis,  (b)  to separate
the tumor  promoting  from the tumor  initiating  effects  of DE, and  (c)  to pro-
                            153

-------
vide  further information  on the  nononcogenic pathological effects  of diesel
engine emissions, especially in the  lungs.
   Groups of approximatly 780  mice,  half males  and half females  were exposed
from  conception until  sacrifice to  either DE or  clean  air.   Each  group  was
separated  into 3 subgroups, initially  numbering 260  animals  each,  receiving
either a single injection of the tumor  initiator urethane,  multiple injections
of the  tumor  promoter  butylated  hydroxytoluene,  or  no injections.   The  mice
were  sacrificed between 15  and 16 months  of  age.   The particulate concentra-
tions in  the chambers  were maintained  at 6  mg/m   during  the first  3 months
of exposure  and 12 mg/m  thereafter.
   Preliminary  inspection of  histopathological  results  indicate  no  dramatic
differences  in  lung  tumor incidence between  the  exhaust and clean air groups.
Results are  very preliminary,  however,  and no final conclusions can be reached
at this time.  Further details  will  be  presented by  K.  I.  Campbell et  al. at
the 1981 Diesel Emissions  Symposium.
   Liver Island Assay.  The  liver island  test  was developed  as  a relatively
short term  in_ vivo carcinogenesis bioassay.   It is  a 2  stage initiation/pro-
motion test  using either  a  choline deficient diet or  phenobarbital  for  pro-
motion and  partial hepatectomy  tp enhance initiation.     The endpoint  is  the
focal appearance of  hepatocytes staining positive for gamma glutamyl transpep-
tidase (GGT).   These  studies were carried out by Pereira  et al. and  have  been
partially published.15
   In the first study young  adult male  Sprague Dawley rats were exposed to DE
at a  particulate concentration of 6 mg/m  for  3  or 6  months following  par-
tial  hepatectomy  and/or  after placement  on  a  choline  deficient diet.   No
islands could  be  detected  in  any  of  the  groups  after  3  months  exposure.
Results of 6 months  exposure are shown  in  Table  5.    Fewer  foci were detected
in the diesel exposed rats,  a  difference that was significant  in the most  sen-
sitive group,  those partially hepatectomized  and  fed  a  choline  deficient
diet.  The results generally agree with  those found in Strain  "A"  mice exposed
via  inhalation.  It  is premature, however,  to  conclude  that  exposure to DE
inhibits island development  since  results were quite variable  and  there was no
indication of a toxic  response in  the   liver,  suggesting  that  little of  the
test chemicals may actually  reach  the liver.
                            154

-------
TABLE 5
EFFECTS OF DIESEL EXHAUST ON THE DEVELOPMENT OF LIVER FOCI

                                                Number of GGT  (+) Foci/cm2
               Treatment                     Clean Air          Diesel Exhaust
Partial hepatectomy, choline deficient
Partial hepatectomy , choline sufficient
Sham operated, choline deficient
5.68* + 3.07
0.29 + 0.29
2.82 + 1.91
0.20 + 0.20
0.00
2.28 -(• 2.28
   In order  to increase  the  concentration of exhaust  components  reaching the
liver, an  additional group of rats  was  injected with 667 mg/Ttg body  wt of DE
extract  following  partial hepatectomy and using phenobarbital as  a promoter.
Dnder these  conditions DE had  no effect  with treated  animals  averaging 0.37
islands/cm  compared with 0.33 for controls.

Metagenesis
   Heritable  effect  on drosophila.   The drosophila  sex-linked  recessive bio-
assay was  chosen as  one  of  our  tests because  it is  considered  an excellent
                                                                            and
                                                                            17
screen  for  genetic  hazards    and  also because  of its  relative economy  and
small exposure  space requirements.   For  details see  Schuler  and Niemeier.
Male flies  were exposed 8 days  to whole exhaust filtered  to  remove particles
larger  than  0.3  microns  diameter,   resulting  in  a  mean  particulate  concen-
tration  of  2.2 rag/fa .   The males  were mated  one  and  8 days  post exposure,
thus utilizing  sperm exposed while mature and while in the spermatocyte stage,
respectively.   The  F.   generation females  were mated  with  their brothers.
The F  's were scored for  a sex  linked lethal  event.   A portion of  the F  's
were mated to produce an F  generation  which was also  scored.
   Among  F   generation  flies  sex linked lethal events  were  detected  in  4 of
1350 vials  (0.30%)  for  exposed  versus 5 of  1354  (0.37%)  for  control flies.
Among  F 's  no  sex  linked  lethal  events were  detected in  the  diesel  group
versus one  of 680 among controls.  With  the numbers used  the  test  is capable
of detecting  mutagens exhibiting  3  to 5 times  the background rate of 0.1 to
0.6%.   It appears  that  either  no mutagen was present  or  the  test was not suf-
ficiently sensitive  to detect  the weak  activity  present.
   Heritable  effect  in  mice.   The effectiveness of  inhaled whole  diesel ex-
haust  on the  induction of heritable  effects  in   mammals  was  studied   by  a
battery of tests utilizing  mice.  The  assays chosen were designed to detect a
                            155

-------
number  of  endpoints,  namely  chromosome breakage,  chromosome  interchange and
point  mutations.   Both  sexes  were examined  for  effects.   These experiments
were designed and  carried  out by  L.  B.  Russell and  co-workers  at  Oak Ridge
National  Laboratories.
   The  mice were exposed 8 hr/day,  7  days/week to  a  particulate concentration
of  6  mg/m3.   In the first experiment  the incidence  of heritable  point muta-
tions were  assessed by exposing recessive T stock males,  then mating them with
females homozygous for  7 easily  detected phenotypic traits.  No definite muta-
tions  were detected  among  42,512  offspring  ruling out with  97.5% confidence
that,  at the  exposure  level  encountered  by  man,  the  induced mutation  rate
could exceed  0.01  times the spontaneous  rate.
   To test  for induction of dominant lethals,  male  T-Stock  mice were bred  to 4
different stocks of females following  exposure to DE,  and  the pregnant animals
examined for dead  implants.  Again, no effects of diesel exposure  could be de-
tected  following examination of about  280 animals/group.
   To test  for heritable  translocations, 160 T stock  mice were  bred  to (SEC X
C57 BL) F.  females after 4.5  weeks of DE exposure.   The  male  progeny  were
weaned  and  subsequently  tested for  sterility.   One partially  sterile translo-
cation  was found  among  1466 control  male progeny  compared with  none  in  350
male progeny of  diesel exposed  mice.
   Effects  of  DE on oocyte  killing  in females  was studied  by  measuring repro-
ductive performance of  60  (SEC X C57 BL)  F^ females/group after   8  weeks  ex-
posure  to DE.   The  average  litter size  of 11 for  both exposed and  unexposed
groups  indicated a lack  of detectable chromosomal or cytotoxic  effects in the
oocytes.
   The  possible  induction of  dominant  lethals  in  females was  evaluated  fol-
lowing  exposure  of  54  (101 x  C3H)  F.  females for  7  weeks  to DE  followed by
mating  to  the same strain of males.   While there was  no evidence  for  the  in-
duction of dominant lethal effects,  fewer  corpora  lutea  were  found and a
longer  interval  between caging  and  copulation occurred in exposed females.
   In the final  experiment of  this series, the effects  of  DE on spermatogonial
survival was tested  in JH and H strain mice after 5 or  10  weeks of exposure to
DE.  No effects of  exposure   could  be  detected among the  8  spermatogonial
classes tested.
   In  summary,   results of  all the heritable tests using  flies or  mice  were
negative.   The  only  significant change was a  decrease  in  corpora lutea in
mice,   slightly  depressing  reproductive  performance.    The  absence of  genetic
effects indicated  that either  no active  metabolites reached the germ cells or
                             156

-------
the germ  cells  were refractory to the  induction  of  mutational events.   in  any
case  it appears that diesel exhaust does not pose a major  hazard.

lexicological effects
   Behavioral effects.   Certain behavioral endpoints,  such as activity levels,
have  been shown to be  quite sensitive to  the  inhalation  of engine exhaust  and
                                 18
other  environmental pollutants.     An  experiment  was  therefore  designed  to
assess the effects of  DE exposure in both adult  and  neonatal rats using spon-
taneous  locomotar activity, forced activity  and an operant  conditioning task
                                                    19
as endpoints.  For  further details see  Laurie et al.
   Spontaneous locomotor  activity (SLA)  was  measured using Sprague Dawley rats
housed in Wahman  LC-34  running wheels.   All  exposures  were conducted at a par-
ticulate  concentration of 6 rag/m .   SLA in adults  exposed 20 hr/day  was  de-
creased  to less  than  10% of  controls  after  6 weeks in diesel  exhaust.   Rats
exposed 8 hours/day showed  similar but  less  dramatic SLA declines in exposures
lasting 16 weeks.   A separate  series  of experiments  was also carried out using
rats  exposed to DE, 20 hr/day  on days 1-17 of age, and 8  hr/day  on days 1-21,
1-28, or  1—42 post parturition.   Testing, however, was conducted in clean air
starting  after  the rats reached 6-7 weeks of  age.   The  post-exposed rats were
less  active than  controls with the largest decrements in  SLA,  again,  noted in
the 20 hr/day exposure group.
   Forced activity  was  measured during  the final week of  a 42 day exposure in
young adult males.   Maximum tolerance was  determined  as  the time to refusal to
run on a  treadmill at a  speed of 19  meters/second.   The  animals were removed
from  the  chamber  and breathed  clean air during the  test.   The criterion aver-
aged  40.9  minutes  for  the   exposed  rats  compared with 107.5 minutes  for
controls,  a difference that was highly significant (P  .01).
   Another test involved  learning a bar pressing task  to  obtain food pellets.
The rats  were  exposed  20 hrs/day and  remained  in  clean  air  thereafter.   The
training  period  started  at 15 months  of  age and was  continued  for  42 days.
The controls showed a  short rise in  the  learning curve  starting at day  5  of
shaping and  all learned  to press the bar shortly  thereafter.   After  25 days
only one  rat previously exposed to exhaust learned to press the bar.
   It was  concluded  that  exposure to  diesel  exhaust  resulted  in alterations in
voluntary  activity,  forced  activity and learning ability.  Moreover,  at least
some of the changes were permanent since  exposure as  early as  the  first week
of life could affect adult learning.
                                157

-------
    Heurophysioloqical effects.   Investigations  in  this  area were  designed to
 provide  information explaining the behavioral changes  observed during exposure
 to diesel  exhaust.   The  endpoints  assessed were somatosensory   and  visual
 evoked  potentials  (SEPs  and  VEPs,  respectively)  collected  from rat pups at
 varying  ages.   For  further  details see Laurie and Boyes.
    Sprague  Dawley pups were exposed 8 hr/day, 7 days/week  from birth to 7, 14,
 21, or  28 days of age.  The exhaust was  diluted  to produce  a partioulate con-
 centration  of  6  rag/m .  Measurements  were made  immediately  after  completion
 of exposure.  SEP  was elicited by a  1 m  amp electrical  pulse to  the  tibial
 nerve of the left  hind limb.   VEP was elicited  by  a  flash  of  light.   Evoked
 potentials  were recorded from silver ball electrodes  located over  the  appro-
 priate projection area on  the  skull.   The responses consisted of a  series of
 positive and negative peaks.
    A discussion of  the details of the  responses is beyond  the scope of  this
 review.   However,  the primary  change detected was  an increase  in pulse  laten-
 cies in  the diesel  exposed groups.  This was  thought to be  due  to  differences
 in the  rate of nervous system development, which is most  rapid in  2 week  old
 rats.  While the  particular process is uncertain, it was hypothesized that the
 increased latencies were related  to  a lesser  degree of  myelinization  in  ex-
 haust exposed rats.
    Susceptibility to infection in niice.   Enhanced  susceptibility to infection
 by inhaled  Streptococcus  and Klebeiella pathogens has been shown following  ex-
 posure   to   both  gasoline     and  diesel  engine   exhaust.22   The  present
 studies  were designed to further  evaluate diesel exhaust  with  respect  to  en-
 hancing  susceptibility to  infection  under conditions  of longer-term  exposure
 and the  use  of pathogens  in addition to Streptococcus.
   Mice  were exposed  for acute,  subacute  or chronic  (up to several months)
 periods,  8  hr/day,  7  days/week,  to  DE diluted to  a particulate concentration
 of  6  mg/m .   Immediately after  test  exposure the  mice  were removed and  ex-
 posed very   briefly  to an  infectious  challenge  aerosol of   Salmonella  typhi-
 murium.  Streptococcus  pyogenes  or  A/PR8-3  influenza  virus.
   Test  exposures  to  DE  significantly and consistently,  as  was the case  for
 gasoline  engine exhaust, enhanced susceptibility to lethal  infection by  the
 streptococcal pathogen.   Susceptibility to the viral or  Salmonella  pathogens,
on the other hand,  was not  significantly affected by previous DE exposure.   At
least part  of  the  DE effect  was considered  possibly caused by the NO.  and
 short-chain  aldehydes present  in exhaust.  Further details  are available  in
 Campbell, et al.23
                             158

-------
Teratoloqical effects
   Although many  of the organic  compounds present  in  diesel engine emissions
may be potentially  teratogenic,  no previous  studies have been  carried out to
specifically  address this  issue.   An  experiment  was  therefore  designed to
evaluate  the  potential for  diesel emissions  to produce malformations in rat
and rabbit fetuses.
   Twenty Sprague Dawley rats  and  20  New Zealand white rabbits were exposed to
DE  at  a particulate  concentration of  6 mg/m   during  the  critical  period of
gestation for  development  of abnormalities  (days  5-16 for rats  and  days  6-18
for  rabbits) .   The  animals  were  sacrificed  one  day  prior  to  the  predicted
birth date.   A portion of the fetuses were fixed  in  Bouins  for determination
of soft tissue  abnormalities.  The remainder  were  fixed  by  the  Alizarin Red S
procedure and examined for skeletal anomalies.
   No effects of  diesel exposure  were noted for either  fetal visceral  or  ske-
letal abnormalities in either  rats or  rabbits.
   Lung function  and pathology in Chinese hamsters.  This  study was designed
to elucidate  the  effects  of  intermediate term exposure  to  diesel engine emis-
sions in a  small  laboratory  animal species.   For  further details see vinegar,
 *  i 24,25
et al.
   Groups of young  adult male  Chinese hamsters were exposed  8  hr/day,  7 days/
week for  6 months to DE diluted  to produce particulate  concentrations  of  6 or
12  mg/m  .   After  completion of  exposure, lung  volumes, diffusing  capacity,
and pressure volume curves were measured.   The lungs were then fixed for path-
ological evaluation.
   Macroscopic  examination  of  the lungs of exposed animals  revealed  the  pre-
sence  of  numerous  black   alveolar macrophages,  almost  filling  the   alveolar
spaces.   The  lining  of the  alveoli  was thickened  due  to  hyperplasia  of the
type II alveolar  cells.   Adding to this thickening was  the  presence  of edema
along with  a possible  increase  in  lung collagen.  Functional  parameters are
listed in Table 6.   Dose  dependent increases  in lung weights,  along  with de-
creases  in  vital capacity  and  diffusing  capacity were  detected.  Plots of
vital capacity  versus  transpulraonary pressure  (not  shown)  also  indicated  a
loss of  recoil pressure.   Despite the  large  decrement in lung  function,  the
animals apparently  adapted  to  the exposure conditions  since  body weights  were
normal.
                             159

-------
TABLE 6
EXPOSURE TO DIESEL EXHAUST UPON LUNG VARIABLES IN CHINESE HAMSTERS
Clean Air
Body wt.
Lung wt.
VC (ml)
(gms)
(gms)

DC (ul/ml/mmHg)
36.0 +
0.23 +
0.96 +•
2.9 +
4.1
0.02
0.10
1.1
DE (6 mg/m3)
35.
0.
0.
1.
2 +
37* +
75* +
5** +
3.4
0.00
0.12
0.9
Clean Air
31.
0.
0.
8.
5 +
20 +
98 +
0 +
3.4
0.02
0.10
1.6
DE (12 mg/ta3)
34.3*
0.48*
0.61
2.2*
+ 1.7
+ 0.06
+ 0.10
± 1.6
 •Significantly different  from control (P < .01),
"Significantly different  from control (p<.05).
VC - Vital Capacity
DC - Diffusing Capacity

   Pulmonary  function in  cats.   This study  was  designed  to evaluate  the  ef-
fects  of chronic exposure to DE upon lung  function  and  pathology in  a  large
animal species having a  lung  complexity more similar  to that of humans than is
found  in rodents.   Functional measurements  have  been  completed.   Pathological
evaluation along with quantitative morphometric measurements of the  lung, how-
ever,  are still in  progress  and results are not  yet available.  For  further
details  of methodology  and  results  of  testing  after one  year exposure  see
Pepelko  et al.26
   Twenty-five  adult male disease-free  cats of uniform  age  and  genetic  back-
ground were exposed  to  DE for approximately  2  years and 3 months.   During  the
first  year  of exposure  the  exhaust dilution  ratio was adjusted to produce a
                                      3
particulate  concentration of  6  mg/m .   During  the  second  year the  concen-
tration  was  increased to  12  mg/m   and remained there until  completion of  ex-
posures.  The cats were removed from  the chambers and anesthetized  with  Keta-
set Plus (Ketamine 100  mg.,  Promazine 7.5 rag/ml) at a dose  of  42 mg/kg during
testing.
   Results  (Table  7)  were generally  negative after  one  year of  exposure.   By
contrast  a clearly defined response was  noted  after the second year.   The  de-
crease in vital capacity  and total lung capacity  compared  with  normal values
for most  functional  measurements indicated a lesion which  restricted breathing
but did  not  cause airway  obstruction or loss  of  elasticity.   The  restrictive
disease  found was compatible  with a  diagnosis  of  pulmonary  fibrosis,  along
with chronic  inflammation,  interstitial  edema,  or  vascular engorgement.   In
agreement  with' this  diagnosis  was the  decrease   in  diffusing capacity.   The
comparable  decreased diffusing capacity, decreased  lung  volumes,  and  patho-
                             160

-------
logical  alterations  in  the lungs  of DE  exposed hamsters  provide additional
support for this diagnosis.
TABLE 7

PULMONARY FUNCTION AND LUNG VOLUMES  IN CATS EXPOSED TO DIESEL EXHAUST
One Year Exposure

RV
VC
TLC
MEF
MEF 50%/VC
MEF 25%/VC
MEF 10%/VC
Compliance
Resistance

Diffusion

CV
N2 washout
Units
ml
ml
ml
ml/sec
ml/sec
ml/sec
ml/sec
ml/cmH20
cmH20/
L/sec
ml/min/
nnnHg
ml
%N2
Clean
104 +
369 +
450 +
1041 +
761 +
491 +
222 +
23.7 +
10.3 +

1.21 +

36.0 +
29.0 +
Air
37.7
42
15
174
160
199
157
9.3
4.4

0.40

16.0
30.2
Exhaust
86.1
348
415
1016
728
490
197
23.5
10.7

1.18

25.6
32.0
+
+
+
+
+
+
+
+
+

+

*+
+
37.0
43
56
185
196
187
107
7.2
4.6

0.43

13.4
20.6
Two Years Exoosure
Clean
80.3
410
484
952
864
574
234
26.2
5.7

1.01

25.2
21.0
+
+
+
+
+
+
+
+
+

+

+
+
Air
28.2
58
68
111
122
153
102
7.1
2.3

0.14

19.3
18.2
Exhaust
66.9
369*
428*
887*
801
518
223
27.4
5.6

0.90*

27.2
39.0*
+ 14.3
+ 42
+ 56
+ 98
+ 125
+ 154
+ 110
+ 4.9
+ 3.2

+ 0.27

+ 17.6
+ 26.0
•Significantly different from controls  (P <.05).
 RV - Residual Volume                 MEF - Maximum Expiratory Flow
 VC - Vital Capacity                   CV - Closing volume
TLC - Total Lung Capacity
   Deposition  and  clearance.  The  large concentration of  respirable particu-
late in DE can enhance  the inherent toxicity of  adsorbed  organics by carrying
them deep into the  respiratory  tract where they  can  be slowly released.  Slow
clearance of  the particles  will result  in an increased  exposure to  the  re-
leased toxic substances.
   In order to investigate the deposition and clearance of inhaled DE,  Charles
River suckling rats were exposed  20  hours/day for 54  days? 20 hours/day for 16
days; 8 hours/day for 5 days,  and 8 hours  for one day  to  exhaust diluted to a
particulate concentration  of  6  mg/m .   Following completion of  exposures  a
small number  of animals from each  group  were   sacrificed  and   histologically
evaluated  for  particle deposition.    The  remaining  rats  were   sacrificed  at
varying times post exposure.  For further details see Moore, et al.
   The rats exposed for 8  hours  and then  sacrificed  showed only an  occasional
black particle in  a  few alveolar macrophages  (AM).    Twenty-eight days after
exposure, no particles  could be seen in the  AM.   The  rats  exposed for 5 days
all contained  a  moderate number of  granules  in   the cytoplasm.   After  28 days
                             161

-------
post exposure a few of  the AH contained no particles  with varying amounts pre-
sent in the  remainder.   The  AM in the rats exposed 20 hr/day  for 16 or 54 days
were  found to be  loaded  with phagocytized particles at  the end  of exposure
with  no  detectable  decrease  in  particle  numbers even  after   90 days  post
exposure.
   Moore  et  al.    concluded  that animals  exposed for  a single  8-hour  period
received  a low enough  dose  to effectively clear the  lungs within  a 28  day
period, whereas the longer exposures  overwhelmed the clearance mechanisms.

Biochemical alterations
   Enzyme  induction.   The primary function  of xenobiotic  metabolizing enzymes
is to  detoxify and/or  to transform  harmful environmental chemicals  into more
readily  excretable  forms.   unfortunately,  some  of   the  metabolites  formed
during those  reactions  are active mutagens and/or carcinogens.   Therefore,  en-
zyme induction could lead to increased risk.  This study was  designed to  eval-
uate the  potential for  induction of  these  enzymes  in mice  exposed  to either
inhaled DE or to  intraperitoneal injections  of DE extract.   For  further  de-
tails see  Peirano.28
   Male strain A/J mice were exposed  for either  6 or  8  months to DE  at a par-
ticulate  concentration  of  6  mg/m  .   Separate  groups  of  male  and female
strain  A  mice were  injected with  OE extract prepared  by soxhlet  extraction
with methylene chloride and dissolved in DMSO.   Each  mouse received  250  mgA9
body  wt/day on  2 consecutive days.   The  liver cytochrome  p        enzymes
present in microsomal preparations were measured using  a modified version of
                                29
the method of Cmura  and  Sato,    while the  enzyme aryl hydrocarbon  hydroxy-
lase (AHH) in liver and  lung was  determined by modified methods  of  Van Canfort
et al.     The  sensitivity  of the  cytochrome  P448_450  assay  was  tested by
injection  of  20 mg/kg body wt of  3-methylcholanthrene (3-MC)  or  80 mg/kg body
wt phenobarbital (PB) into separate groups of  mice.
   PB  induced  75%  and  131%  increases  in liver cytochrome  p    type enzymes
in males  and  females,  respectively,  while  3-MC  induced  39%  and  37%  increases
in liver   cytochrome  P,..  type  enzymes.   Liver  cytochrome  ?..„ ,rn levels
                        448                          ••            448—450
and liver  AHH activities were not influenced by  the inhalation exposure.   Lung
AHH activities calculated on a per mg microsomal protein basis  were  decreased
31% and 21%  in mice  exposed for 6 or 8  months to DE,  respectively.   When cal-
culated on a per  lung  weight basis,  however,  6  and  8  month  exposed  males
showed 23% and  47% increases,  respectively.   Injection of DE extract resulted
in a 23%  increase  in liver  cytochrome  P,,0 „.„ levels in  males   (P<.02),  but
                                         448-450
                            162

-------
only a  nonsignificant 7% increase in females.  AHH activity was  not determined
in DE extract  exposed mice.
   The  changes in  lung AHH were considered artifactual.  This was  based  on the
likelihood  that the non-microsomal  protein  of the DE  exposed  animals was  in-
creased due in part  to inflammation,  increased  collagen synthesis, etc.   The
limited inducibility  by  IP injection,  despite  the very  high  dose,  suggested
either  that DE contains only a limited amount of  inducing chemicals,  induction
inhibitors  are present,  or  the  bioavaliability  of  chemicals present  on  the
diesel  particle is  very low.
   Benzo(a)pyrene  metabolism  in  mice.   One  means of  estimating  the carcino-
genic potential of inhaled DE is to measure  the  effects  of  DE exposure on  the
absorption,  distribution,  metabolism and excretion  of  a  polycyclic aromatic
hydrocarbon (PAH)  known to be present in the exhaust.  Benzo(a)pyrene  (BP)  was
selected  as a characteristic  PAH  that  is  found adsorbed  to  DE  carbon par-
ticles.   The following study was performed  by Tyrer  et  al.    and  Cantreli et
al.32
   Male strain A/J mice were  exposed for  9  months to  DE diluted  to produce a
particulate concentration of  6 mg/m .  following completion  of  exposure  the
                                             14
mice  were  intratracheally instilled  with    C-BP and  then  sacrificed  2,  24
and  168  hours later.   Immediately  after sacrifice  the mice  were  frozen in
liquid  nitrogen,  sectioned and  autoradiographs  made.   The metabolism  and  ex-
cretion of  intratracheally instilled   H-BP  was determined in  separate groups
of mice.   The  mice were  instilled  and sacrificed at  2,  24,  and  168 hours as
before.   The liver,  lungs and testes were removed, frozen  and  later extracted
and analyzed via HPLC.  Drine  and feces were  also  collected and analyzed.
   The  results indicated  there were  no differences  in clearance  of  soluble
metabolites  between  clean  air  or   DE  exposed  mice.    The  DE  exposed  mice
appeared  to have  less free unmetabolized BP in  their tissues,  which  at  the
time was  speculated  as possibly due  to  enzyme  induction.   The  only meaningful
differences  found  between DE and clean  air  exposed  mice was the  inability of
the DE  exposed mice  to clear  small  amounts of BP one  week  after instillation.
It was  suggested that this was due  to adsorption  of BP onto deposited DE par-
ticulates.
   Biochemical  assessment  of  exposure induced lung damage.  Since  the lung is
the prime  target  organ for  inhaled  pollutants,  a series of experiments were
conducted to evaluate lung damage using a  variety of  biochemical  parameters.
Some of  the studies  have  been published by  Lee,  et  al.    The  remainder  are
unpublished.
                            163

-------
   All the experiments were  conducted using rats exposed  either  8 or 20 hours/
day, 7 days/week for periods ranging from one to 63 days.  Participate concen-
trations were maintained at  6 mg/ta  .
   Assays  were carried  out on  lung  lavage  fluid, pulmonary macrophages  and
lung tissue  homogenates.   Lung  lavage was analyzed  for      I-albumin  leakage
into the alveolar spaces as  an  estimate of lung integrity.   Lysozyme and total
protein concentrations  in the lavage were monitored  to  assess  the degree of
cell lysing  and  subsequent protein release.   The  number and  viability  of  pul-
monary macrophages were  determined  as an estimate  of  the lung's  defensive  re-
sponse  to  an  irritant  pollutant insult.   Analyses were  carried  out on  lung
tissue homogenates for  the following:  total  lung  proteins to assess  for pos-
sible lung fibrosis  and general  tissue  injury; lung  lipid peroxidation as an
indicator of the presence  of free-radicals; superoxide dismutase  assessing  the
ability of the lung  to destroy  harmful superoxide freeradicals that  may result
from DE exposure; glucose-6-phosphate dehydrogenase and 6-phosphogluconate  de-
hydrogenase  to assess  any  diesel  induced effects on the  pentose  phosphate
metabolic pathway.
   The only  significant  changes were  substantial  increases in lung  homogenate
protein concentration.   It was  concluded that  there were no  significant early
detectable  biochemical alterations that  would indicate  diesel  induced lung
tissue injury.
   Influence of  DE  on lung  proteins.  Diesel  engine  emissions  contain many
potentially  harmful  components  capable of  injuring lung  tissue.   Such  injury
may  lead  to  increased synthesis of  lung  proteins,   especially  collagen,  and
subsequently to  an  accumulation  of  lung  connective   tissue  matrix.   This  can
lead to morphological  alterations,  such as fibrosis,  or  to  lung scarring  and
loss of structural integrity as in  emphysema.
   To assess the potential  of  DE  to induce  these lung  alterations,  several
studies were conducted  using rats  exposed  20 hours/day,  7 days/week, or mice
exposed 8  hours/day,  7 days/week to  exhaust diluted  to produce a particulate
concentration  of  6 mg/m .   The  animals were  assessed  for  total  lung  protein
and the ability  of  the  lungs to synthesize  and accumulate  collagen and non-
ollagen proteins using  radiolabelled proline  and  leucine incorporation  assays
and, in some cases, lung prolyl hydroxylase activity.   '
   The total lung proteins in rats  exposed for 56 days were  increased  47% over
that of controls.   The in vivo leucine incorporation,  however, was  decreased
38*  suggesting a  decrease  in  overall protein  synthesis  in exposed  animals.
The increase in protein concentration, despite an apparent decrease  in overall
                                164

-------
protein synthesis, was  considered  as possibly due to either an  increased accu-
mulation  of  circulating  proteins  and migrating  cells in  the  alveolar  inter-
stitium or  other compartments,  or to connective  tissue  proliferation and  in-
creased deposition of connective tissue matrix.
   In  vivo  proline incorporation,  an estimate of collagen  synthesis, was  not
affected  by  59 days exposure to DE.   However, prolyl-hydroxylase activity, an
in vitro  assay,  was increased  in  rats exposed 33 days  to DE and  in rats  ex-
posed  to  DE in  utero.   Hie results  of  both of  these assays  suggest relative
increases in the synthesis  of  collagen as compared  with  the  synthesis of non-
collagen proteins in DE exposed  rats.
   In  order  to extend  these studies to compare  effects  upon  another species,
an experiment  was designed  utilizing mice exposed  for  3.5,  6 or  9  months to
clean  air or DE.  large increases  in  lung protein content were  found.  Colla-
gen synthesis  also increased in the  exposed  mice, reaching a value  1.5 times
that  of  controls  after  9  months  exposure,  while  overall protein   synthesis
decreased.
   The relative  increase  in collagen synthesis in both mice and rats  suggested
the occurrence of lung  injury,  leading to proliferation  of  connective tissue
and possible fibrosis.   The data from the mice suggests further that continued
exposure to diesel emissions may exacerbate lung  injury.

SOMMARY
   There  was little  evidence that  inhalation of  DE  resulted  in the  induction
of  tumors.   In  fact,  some of  the  results   suggested  a possible   inhibitory
effect of DE on  tumorigenesis.   Injection of DE particulate or particulate  ex-
tract  into Strain  "A" mice  or  rats  likewise  failed  to produce significant  in-
creases in the incidence of either  lung  tumors or liver islands.  Essentially
negative effects were also  noted in  attempts  to  induce  increases in  heritable
mutations in mice or fruit  flies;  teratological effects  in rats or rabbits; or
enzyme induction in mice  (with  the  exception  of a small  increase in males,  but
not females,   injected  with a  very  large  dose).   Results  of  most  genotoxic
studies, which have  been discussed by Pereira in another chapter,  were also
negative  using inhalation  or  injection of  particulates as the means  of   ex-
posure.  Only  after  injection  of  large  doses of   particulate  extract,  or mea-
suring effects in  the   lungs were  positive results  obtained.   The  most likely
reason for  the relative  lack of effects  is  a low  degree  of  bioavailability,
which could stem from slow  leaching of the particulate  coupled  with rapid  in-
activation and excretion of potential  carcinogens  and mutagens.
                             165

-------
   By  contrast,  a  wide  range  of  non-oncogenic  toxicological  effects  were

found.   Exposure to DE  resulted in a  decreased level  of both  voluntary  acti-

vity and exercise tolerance.   Exposure early in  life had a  detrimental effect

on learning  in adult rats.   In agreement with  behavioral changes  was evidence

for delayed  neuronal maturation.  A  variety of  changes were  detected in the

lungs.   Functional  decrements  suggested  the development of restrictive  lung

disease.   Biochemical  changes  indicative  of  increased collagen  deposition,

along with alterations in histology supported  this  conclusion.  Resistance to

infection decreased  markedly  after  DE inhalation,  and clearance  mechanisms

were overwhelmed.  Hie marked effects  found in the  non-oncogenic  studies are

likely due to the presence  of vapor  phase  components  such  as nitrogen oxides,

aliphatic aldehydes, etc.,  which  would be  expected  to show  a  much  greater

degree of bioavailibility than the potential carcinogens.

   In the areas  where  dramatic effects were  noted,  such as behavior,  lung mor-

phology,  infectivity,  etc., further  studies are necessary  to delineate thres-

hold limit values and  to isolate  and  identify components of exhaust  respon-

sible for producing  these changes.


REFERENCES

 1. U.S.  EPA  (1978)  Health  Effects Associated  with  Diesel  Exhaust  Emissions.
    OSEPA-600/1-78-063.
 2. U.S.  EPA  (1979)  The Diesel Emissions Research Program.  EPA-625/9-79-004.
 3. Springer,  V.J.  and  Baines. ,T.M.  (1977)   Society of  Automotive  Engineers,
    SAE  Report #770818.   Warrandale, PA.
 4. Springer,  V.J. and  Stahman, R.C.  (1977)  Diesel  car Emissions -  Emphasis on
    Particulates and  Sulfate.   Society of  Automotive  Engineers,  SAE Report
    1770254.   Warrandale,  PA.
 5. Menster,  M., and  Sharkey,  A.C.,  Jr.   (1977)  Chemical  Characterization of
    Diesel Exhaust Particulates.  NTIS,  PERC/RI-77/5.
 6. Hinners,  R.G., Burkart,  J.K., Malanchuk,  M., and  Wagner,  W.D.  (1980)  in
    Generation of Aerosols, Willeke,  K. ed.   Ann Arbor  Sci.,  Ann Arbor, HI,
    pp.  525-548.
 7. Hinners,  R.G., Burkart,  J.K., Malanchuk,  M., and  Wagner,  W.D.  (1980)  in
    Health Effects  of  Diesel  Engine  Emissions:    Proceedings  of  An inter-
    national  Symposium.   V. 2,  Pepelko,  W.E., Danner,  R.M.,  and Clarke, N.A.,
    eds.   USEPA,  Cincinnati, OH, pp. 681-697.
 8. Laresgoiti,  A.,  Loos, A.C., and Springer,  G.S.   (1977) Env. Sci.  Technol.
    11,  973-978.
 9. Pitts, J.A., van  Cauwenberghe, K.,  Winer,  A.M.,  and Belser,  W.L.  (1980)
    Final  Report,  EPA Contract No.  R806042.   USEPA,  Cincinnati,  OH.
10. Shimkin, M.B. and Stoner,  G.D.  (1975)  Adv. Cancer Res. 21, 1-55.
11. Humason,  E.L.  (1972)  Animal  Tissue  Techniques.    3rd  ed.  W.H. Freeman Co.,
    San Francisco CA, pp.  21-22.
12. Orthoefer, J.G., Moore,  W., Kraemer, D.,  Truman,  F.,  Crocker, W.  and  Yang,
    Y.Y.  (1980)  in Health Effects  of  Diesel  Engine  Emissions:  Proceedings of
    an International Symposium.  V. 2, Pepelko,  W.E.,  Danner,  R.M.,  and  Clarke,
    N.A.,  eds.   USEPA, Cincinnati,  OH, pp.  1048-1072.
                             166

-------
13. Huisingh,  J.L.,  Bradow,   R.L.,   Jungers,  R.H.,  Harris,  B.D.,  Zwidinger,
    R.B., Gushing, K.M., Gill, B.E.,  and  Albert,  H.E.   ibid.  pp. 788-800.
14. Sells, M.A. , Katyal, S.L. ,  Sell,  S.,  Shinzuka, H. ,  and  Lombard!,  B.  (1979)
    Brit. J. Cancer.  40, 274-283.
15. Pereira, M.A., Shinozuka,  H.,  and Lombardi, B.  (1980)  in  Health Effects of
    Diesel  Engine  Emissions:  Proceedings of  an International  Symposium.   V.2,
    Pepelko, W.E., Danner, R.M.,  and  Clarke,  N.A.,  eds.   USEPA,  Cincinnati,
    OH, pp. 970-976.
16. Wurgler, F.E. ,  Sobles,  F.H., and Vogel,  E.  (1977)  in  Handbook of  Mutage-
    nicity  Test  Procedures, Kilbey,  B.J., Legator, M.,  Nichols,  W.,  and Ramer,
    C., eds. Elsevier,  Amsterdam, Holland, pp.  335-373.
17. Schuler, R.J., and Niemeier  (1980)  in  Health  Effects  of  Diesel  Engine
    Emissions:   Procedings  of  an  international Symposium,  v.2,  Pepelko,  W.E. ,
    Danner, R.M.,  and Clarke, N.A., eds.  USEPA,  Cincinnati, OH,  pp.  914-923.
18. Lewkowski,   J.P.,  Malanchuk,  M.,  Hastings, L., Vinegar,  A.  and  cooper, G.P.
    in  Assessing  Toxic Effects  of  Environmental  pollutants,   Lse,  S.D.  and
    Mudd, J.B.  eds.   Ann Arbor Sci.,  Ann  Arbor, MI,  pp.  187-217.
19. Laurie, R.D.,  Boyes,  W.K.  and  Wessendarp,  T.  (1980)  in Health Effects  of
    Diesel  Engine  Emissions:  Proceedings of  an  International Symposium,  v.  2,
    Pepelko, W.E., Danner. R.M.,  and  Clarke,  N.A.,  eds.   OSEPA,  Cincinnati,
    OH, pp. 698-712.
20. Laurie, R.D. and  Boyes, W.K.  ibid.,  pp.  713-727.
21. Coffin, D.L.,  and Bloramer, E.J.  (1967) Arch.  Environ.  Health, 15,  36-38.
22. Lee,  S.D.,  Campbell,  K.I.,  Laurie, D.,   Hinners,  E.G.,  Malanchuk,  M. ,  and
    Moore, W.  (1978)  71st Ann. Meeting, Air Pol.  Control Assn.,  Houston, TX.
23. Campbell,  K.I.,  George,   E.L.,  and  Washington,  I.S.   (1980)  in  Health
    Effects of  Diesel Engine  Emissions:  Proceedings  of an  International Sym-
    posium.  V.2,  Pepelko,  W.E., Danner,  R.M.,  and  Clarke,  N.A., eds.  OSEPA
    Cincinnati,  OH, pp. 772-785.
24. Vinegar, A., Carson, A.I., and Pepelko, W.E.   ibid., pp. 749-756.
25. Vinegar, A., Carson,  A.,  Pepelko, W.E.,  and Orthoefer,  J.G.  (1981)  Fed.
    Proc. 40, 593.
26. Pepelko, W.E. , Mattox,  J.,  Moorman, W.J., and Clark, J.C.  (1980)  in Health
    Effects of  Diesel  Engine  Emissions:  Proceedings  of an  International Sym-
    posium.  V.2,  Pepelko,  W.E., Danner,  R.M.,  and  Clarke,  N.A.  eds.  DSEPA,
    Cincinnati,  OH, pp. 757-765.
27. Moore, W.,  Orthoefer,  J.,  Burkart, J., and  Malanchuk,  M.  (1978)  71st Ann.
    Meeting, Air Pol. Control Assn.,  Houston, TX.
28. Peirano, W.B.  (1981)  Proceedings EPA 1981 Diesel  Emissions Symposium   (in
    Press).
29. Omara, T.,  and Sato, R. (1964) J. Biol. Chem.  239,  2370-2378.
30. Van  Cantfort,  J., DeGraeve,  J.,  and  Gielen,  J.E.  (1977)  Biochem.  Biophys.
    Res. Coram.   79, 505-572.
31. Tyrer,  H.W.,  Cantrell,  F.T. ,  Horres,  R. ,  Lee,   I.P.,  Peirano,  W.B. ,  and
    Danner, R.M.   (1980) in Health  Effects of  Diesel  Engine  Emissions: Proce-
    edings  of  an  International  Symposium.  V. 1,  Pepelko,  W.E., Danner,  R.M.,
    and Clarke,  N.A.  eds. OSEPA, Cincinnati,  OH,  pp.  508-519.
32. Cantrell,  E.T.,   Tyrer, H.W.,  Peirano,  W.B.  and  Danner,  R.M. ibid.,  pp.
    520-531.
33. Bhatnagar,   R.S.,  Hussain,  M.Z.,  Sorensen,  K. ,  von  Dohlen,  F.M.,  Danner,
    R.M., McMillan, L., and Lee, S.D. ibid.,  pp.  557-570.
                             167

-------
DEPOSITION AND CLEARANCE OF DIESEL PARTICLES FROM THE LUNG

JAROSLAV J.  VOSTAL, RICHARD M.  SCHRECK, PETER S. LEE,  TAI L. CHAN, AND
SIDNEY C. SODERHOLM
Biomedical Science Department, General Motors Research Laboratories, Warren, Michigan
48090, U.S.A.
INTRODUCTION
   A projected increased use of light-duty  diesel engines on  U.S. roads has generated
considerable interest in determining the potential health impact of inhaled diesel emis-
sions.   Of primary concern  is  the potential effect  of  the  submicron carbonaceous
particulate  fraction of the exhaust, which is present in concentrations of 30-100 times
those of a spark-ignited,  catalyst-equipped, gasoline-fueled  engine.   Since these particles
                                                                  2
have a mass median aerodynamic diameter of approximately 0.2 ym,  they have a settling
velocity which is  almost  nil, and will persist in the  atmosphere for a  considerable time
after emission, allowing time for airborne transport and possible chemical transformation
in the atmosphere.  Studies  have shown the particle  to consist of smaller, elemental,
                                          o
carbonaceous  particles approximately 300 A in diameter, which  are  fused together  into
                                                             Q
agglomerates containing up to several hundred  elemental units.    Closely  associated with
                                                                   2
this  core  is a  variable fraction of  benzene-soluble organic material,   known to contain
literally thousands of compounds, including  polycyclic  aromatic  hydrocarbons  and other
                                                                M (? C •?
compounds  known to be biologically active in various assays. '  '  '    Since airborne
particulate material in this size range is not effectively removed in the upper respiratory
system, it is anticipated  that  a certain fraction of the particulate material inhaled with
each breath will be deposited  in various regions of the respiratory system, including the
deep lung.  The exact amount  of diesel  particulate material, and where in the respiratory
system it  will be deposited, are determined by the physics of the particle interaction with
the inspired current of air and the adjacent walls of the airways. The total mass deposited
determines  the dose to  a given  organ, and  together  with  other factors such as the
bioavariability of materials on  the particles for interaction with the surrounding tissues and
the rate of clearance, causes  the response which  may  occur  in  these  tissues.  It is this
issue, the determination of dose  to the lung, which  this chapter  will address through
experimental measurements of deposition and clearance of inhaled diesel particles.
   Deposition  efficiency  in this context will  refer  to  the mass fraction of particulate
which deposits  on  the surfaces of any of the  respiratory airways, divided by the  total mass
of particulate material entering during respiration.  This percentage of deposition has been
studied  extensively in man, and has been expressed as a function of the diameter of the
                             168

-------
inhaled particle  by the  well  known International  Commission Radiological Protection
              q
(ICRP) model.   On a smaller scale,  similar work has been  done for some species  of
laboratory animals; however,  the experimental  data  available to date  is  much more
       9
limited.    The form of  the deposition relationship  is derived  from consideration  of  the
deposition mechanisms of inertial impaction, sedimentation, and Brownian diffusion applied
to a particle during its transit of the nasopharyngeal, tracheobronchial, and deep pulmonary
regions of the lung. The model has been well verified for particles >  0.5 ym mass median
aerodynamic diameter in human exposure tests to  innocuous monodisperse aerosols through
measurements  of  the  inspired  and  expired mass  concentrations and  the air volumes
respired.   Recent data obtained  for submicron aerosol inhalation  studies  in  humans has
shown  that the amount deposited may be over-estimated by the ICRP model,   and that
experimental measurements may still be necessary to accurately predict the deposition  of
airborne materials in this size range.
    The second process  closely  associated with  particle deposition  is  the clearance  of
deposited  material from  the respiratory system via the mechanisms of physical  transport  or
dissolution.  Since  this process may account for  the removal  of a significant amount  of
deposited  material  from the airways, it is an important consideration in determining the
amount of material retained in the lung, and thereby the long-term  dose to this organ.  To
this end,  the following series of  experiments was performed  using  diesel particles, and a
mathematical model of  particulate transport in the respiratory system was developed  to
evaluate and interpret the experimental findings.

METHODS
    Test animals  were exposed to diluted diesel exhaust at controlled  particle eoncentra-
                                      11                                             12
tions in a  large-volume exposure facility  by methods described in an earlier publication.
Tracing the deposition,  clearance,  and subsequent retention of  the  diesel particles was
accomplished  using a  second  exposure apparatus in which  radioac lively-labelled diesel
particles  were inhaled by the animals, and their subsequent fate determined by tracing the
radioactivity.  The  radioactive  tagging was achieved by  introducing the compound (1- C)-
n-hexadecane  (ICN Pharmaceuticals) into the diesel fuel (AMOCO  Type 2D), based on its
                                                                         14
representative boiling  point, molecular weight, and  chemical structure.  A   C tag of an
aliphatic  hydrocarbon  in the fuel has several advantages  over  other markers  which were
considered for the study.   These include  the  fact  that  the  tagged carbon  atom  is
incorporated  into the diesel particle during  the combustion process in a manner indistin-
guishable  from atoms of other  fuel compounds, and the  fact  that the  tagged atom  cannot
be  leached from the  particle  by body fluids.  The latter characteristic  is  particularly
                                                                            13
important for  clearance and retention studies as discussed in detail by Chan et al.
    In order to utilize the radioactive compound most effectively, a single-cylinder direct-
                               169

-------
injection diesel engine  (Farymann K-54)  with a 242  cm3 displacement  was used.  The
engine was operated at  a constant speed of 3600 rev/min, and particles were sampled at
various engine loads on  microquartz  glass fiber filters to determine the distribution of  C
radioactivity between the insoluble particle core and the extractable fraction.  Generally,
the  extractable fraction  decreased  with  increasing  engine  load  both  for  the  mass
                                                  14
distribution determined  gravimetrically  and  for   C  activity distribution  determined
radiometrically.   In view of the  fact  that the  diesel fuel contained a large number of
aliphatic and aromatic hydrocarbons and the   C tag was only present in a single compound
(  C-n-hexadecane), the activity distribution did not exactly follow  the  mass distribution.
Under conditions of high engine load, 99% of the   C was preferentially incorporated in the
carbonaceous  particle core.  The particle size  distribution  was determined at different
engine loads using a multi-jet Mercer cascade  impactor, and  it was determined that when
the engine  was operated at loads  higher than 30%  of full load,  the  size distribution was
essentially identical to  that observed in the GM  5.7 L diesel engine operated at 20 mph
under road load conditions.
   Groups of  24 male Fischer 344 rats (Rattusnonregicas) weighing approximately 325 gor
24 Hartley guinea  pigs  (Cavia porcellus) weighing approximately 450 g  were exposed in a
nose-only inhalation chamber to  C-tagged diesel particles generated from the Farymann
engine operated at full load.   The  exposure chamber was constructed of  6  mm thick
stainless steel with exposure ports on  each of three tiers.  Radioactive diesel particles
entered through the bottom section of the  chamber, and rapid mixing with dilution air was
achieved in a 15 cm high mixing  section  inside  the exposure chamber  below an annular
perforated plate.  Precalibrated orifice meters in the airflow measurement section allowed
precise adjustment of the flow of dilution air and  diesel exhaust to the chamber. The total
flowrate through the chamber was 200  L/min, and a negative pressure of 1.5  cm of water
maintained within  the chamber prevented  any possible radioactive contamination  leaks.
Immediately after  the exposure, i.e., generally within 30 minutes from  the conclusion of
the exposure, groups of animals  were  sacrificed by an  intraperitoneal injection of Na-
pentobarbital.  A blood  sample was  obtained by cardiac puncture, and the lungs, heart,
spleen, liver,  trachea with hilar lymph nodes, and thymus with  mediastinal lymph nodes
were removed. The tissues were processed in a biological material oxidizer (Harvey). The
         14
resultant  CO2 was absorbed in a trapping solution (three  parts of ethanolamine and seven
parts  of  2-methoxyethanol) for  radiometric assay of  C activity using a liquid scintillation
counter (Searle Mark ffl). Expired air,  feces, and urine were also collected and analyzed
   14
for  C  activity during the first ninety-six hours post-exposure.  The measurement of the
  C activity in blood, urine, and expired air indicated that the elimination of  C activity
from  the blood through  the urine and expired air was surprisingly fast, and more  than 95%
of the initially absorbed  activity was eliminated from the circulating blood within the first
                                170

-------
six hours.
   Although  a  diffusion scrubber greatly reduced   CO.  concentration in the diluted
exhaust, a relatively large amount of remaining radioactive CO0 could still be absorbed in
                                                            ti
the circulating blood and significantly influence the measurements of particle retention by
14
  C activity assay.  To correct  for the potential artifact,  the amount of blood remaining in
the  excised  organs was  determined by  the analysis  of iron using atomic  absorption
sp<
14,
                           14
spectrophotometry, and the   C activities in the tissues  were corrected for  the  absorbed
  C bicarbonate contamination in the circulating blood.  Based upon these observations, a
numerical correction was necessary only for the initial deposition determination.  After 24
hours, no correction was necessary, since 99% of the activity in the blood had already been
eliminated.  Typically, the numerical correction  for the blood contamination in the initial
deposition was 10% of the total activity.
                  14
   In each of the  C  studies, the deposition efficiency was calculated by dividing  the
corrected lung particle activity determined immediately after the termination of  exposure
by the inhaled dose.  The inhaled dose was calculated as the product  of the specific activity
of the  exhaust  particles  in  the  inhaled  air,  estimated  lung  ventilation,  and  exposure
duration.   The  specific activity  of particles in  the exhaust  was determined from  the
radioactivity assay of filter samples, the exposure duration was experimentally known,  and
literature data were used to estimate  the lung  ventilation of  test animals.  '  '   Since the
lung ventilation correlates well  with the 3/4 power of  the body weight,  an  empirical
relationship was also used to estimate the lung ventilation based on variations in body
                         14
weight of the  test animals.

EXPERIMENTAL RESULTS
   Three separate projects investigated the kinetics of deposition and clearance  using the
14
  C-tagged diesel particles:
Long-term Clearance
              JO               1»7
   Chan et  al  and Lee et al   measured the long-term lung clearance of inhaled diesel
particles in animals acutely exposed to two different exhaust concentrations.  At the  lower
                                         o
diesel particulate  concentration (2000 yg/m ) and extended exposure duration (140  min), the
deposition efficiency of diesel particulates in the lung was estimated to be 20 ± 5% of the
inhaled  dose  for  the exposed Fischer 344 rats.  This  agreed  well  with  the calculated
deposition efficiency of 17 t 2% of the inhaled dose for the same strain exposed to a higher
concentration of diesel  particulates  (7000 pg/m ) for a shorter exposure duration  (45 min),
and it  indicated  that the  deposition  efficiency  was  not significantly influenced by  the
inhaled  concentrations of particles within the range of our observations.  It was encour-
aging to note  that the value of 15-20% deposition efficiency for inhaled diesel particles in
the Fischer  344 strain  was in good agreement  with  experimental  data  reporting 15-20%
                               171

-------
                                                                         ,1048
                                                                                It also
deposition  efficiency for particles  of comparable size  (0.2 ym) in  humans.
agreed  well  with the  Schum and  Yen theoretical model,   which  suggests that the
deposition  efficiency of inhaled submicron particles in  rats was more similar to  humans
than any other mammalian species used in laboratory testing.
   The activity of inhaled 14C-tagged diesel particles retained in the lungs at various time
periods after the exposure was determined by liquid scintillation counting of the retained
14C activity in the lung.  Since the initial lung burden could not be determined for every
test  animal, a normalized value to account for the variability in the initial  lung dose has
been estimated.  No significant difference in particle clearance has been observed between
the two experimental groups exposed to two different particulate concentrations with
comparable inhaled dose (7000 ug/m3 for 45 minutes and 2000 ug/m   for 140 minutes) for
an observation period up to 28 days after exposure as shown in Figure 1. This indicates that
at the two particulate concentrations studied, there were no major alterations in the lung
clearance process.
   At the  inhaled particulate concentration of 7000 ug/m , the retention of particles was
determined up to 6 months after the exposure.  The  experimental data, shown in Figure 1,
were analyzed by a  curve-stripping procedure  into three  components  with approximate
                                                 17
half-times of 1 day, 8 days, and 80 days, respectively  . The  biological meaning of these
           100
         c
         9
         9
        
-------
components can  be understood in  terms of clearance mechanisms.  The first mechanism
deals preferentially with particles deposited in the Uacheobronchial tree,  and represents
their transport by the mucocillary  escalator. The particles are finally cleared through the
gastrointestinal tract, and their elimination is clearly documented  by the presence of   C
activity  in  the feces.  The second  mechanism  can be  interpreted as the  transport  of
material deposited in  the proximal respiratory bronchioles, where only  a short distance is
required  for  transferring  the particulates to  the  mucociliary  escalator.   The  third
mechanism involves the removal of participate matter from the alveolar region where  the
clearance  may involve endocytosis,  passive and  active  absorption,  and  dissolution  or
metabolism of the deposited particles.
Species Differences
    Chan et al  analyzed the species differences  in the deposition and clearance of inhaled
diesel particles using rats  and guinea  pigs.  Lung clearance of  inhaled  diesel exhaust
                                                                  13
particles in the rat was divided into two distinct phases of clearance  up to 105 days post-
exposure.  An exponential  clearance  half-time of  1 day  for particles  removed  from the
tracheobronchial region by  ciliary  action represents the 'rapid' clearance phase.  A slower
clearance phase,  mediated by the alveolar  macrophages through phagocytosis and transport
of the particles out of the  respiratory airways, has a half-time of  62 days.  In the guinea
pig, the  clearance rates for diesel particles removed  in the first  few  days by  the  same
mechanisms showed a similar clearance half-time of 1 day.  The  percent of initial lung
particle deposition cleared during the  rapid phase by the guinea pig was 17%, compared  to
34% in the rat.
    The long-term clearance phase, however, was surprisingly very  different.  As shown in
Figure 2, almost no clearance was  observed in the lungs of the guinea pig  between day 14
and  day  105 post-exposure.   This  would indicate  a greater potential  for the guinea pig  to
accumulate diesel particles in the  pulmonary regions, compared to the Fischer rat.  This
seeming  lack  of  'normal' alveolar clearance in  the guinea  pig would account for the
relatively large amount  of diesel particles recovered from lungs of  guinea pigs exposed  to
                                          21                  13
diesel exhaust in a chronic inhalation study.   In  an earlier study,   the  clearance of diesel
particles to the lymphatic system was observed in  the Fischer rats, where a few percent  of
the initial deposition was measured in the  mediastinal lymph nodes as early as a few days
post-exposure.  These measurements have followed  the  accumulation  of  inhaled  diesel
particles in the lymph nodes of the guinea pig and the rat up to 105 days  post-exposure.
Clearance of  the diesel particles from the pulmonary regions to the lymph nodes occurred
in both species during the first few days with 2%  of the initial lung dose retained after 105
days  in both  the  rat  and guinea  pig.  Although this observation  is not uncommon  for
insoluble particles that have been known to reside in the lymphatic  system for long periods
of time, the transport  of diesel particles from the  lungs to  the lymph nodes was faster than
                                173

-------
                        10  20  30   40  50  60   70  80  90  100  HO
                                  Days Post—exposure

 Figure 2.  Differences in the clearance of  C-tagged diesel particles from the lungs of rats
 and guinea pigs after a single acute exposure.  At each measurement, N = 4.  From Chan
 and Lee20

                                                on
 that of fused-clay particles reported by Thomas.    Diesel particles of 0.15 urn  diameter
 could conceivably evoke faster lymphatic drainage, as compared to the larger 1.5 ym clay
 particles.  In this study, only the largest thoracic lymph nodes (hilar and mediastinal) were
 analyzed, so that the  absolute amount of diesel  particles in the entire lymphatic system
 might be slightly higher than the values computed.  Furthermore, the alveolar clearance of
 diesel particles in the guinea pig was extremely  slow,  with more than 80%  of the initial
 dose retained  after 105 days.   The  pulmonary  clearance half-time  for inhaled diesel
 particles in the guinea pig is estimated to  exceed 300  days which strongly contrasts with
 60-80  days in  rats (determined by fitting  experimental data collected so far  to two- or
 three-phase clearance models.)   The  differences observed in this study  demonstrate a
greater long-term retention of inhaled diesel particles in the guinea pig possibly caused by
slower clearance processes in the deep lung of  this species.  The actual biological dose to
the respiratory epithelium would also be different in both species. This clearly  indicates
the difficulty in comparing studies on  potential health effects of inhaled diesel  particles
among different species and in extrapolating experimental animal data to man.
Accumulated Dose Effect
   Chan et al    also measured the effect of accumulated particle mass in  the lung on the
kinetics of clearance of acutely inhaled diesel  particles.  Preliminarly results obtained on
                               174

-------
 groups of Fischer 344 rats pre-exposed to substantial doses of diesel particulate in the
                                   O                                      n
 large-volume chambers (6000 ug/m   for seven and  112 days, and 250 ug/ni  for  112 days)
 indicate that the clearance rate of   C-tagged diesel particles inhaled for 45 minutes is
 substantially altered by the condition of the  animals.  Preliminary results  were obtained
 from pooled measurements on 4 animals.   In the group exposed to  6000 ug/m  for 112 days
 and having an accumulated lung burden of 10 mg particles per gram of  lung at the time of
 exposure,  a retardation of clearance was noted which  persisted  over the first 49 days with
 only approximately 15% of the lung-deposited material being cleared in  this interval. Lung
 clearance in  animals exposed  to  the lower pre-loading doses  (approximately  0.5-1.0 mg
 particles per gram of lung) was also reduced, but to a lesser extent  as depicted in Figure 3,
 and compared  to  age-matched control animals, which  had  only  clean air exposure prior to
 inhaling the   C-tagged diesel particles. At this  time in the analysis, it  appears that an
 increase in the retention of  the tracer particles  has  occurred  as  a result of the  massive
 accumulation of  diesel particulate material  in  the lung,  and  that the  effect  is  dose-
 dependent.
                                                   a 6000 Mfl/m'—TO d
                                                   • 6OOO Mfl/m*— 7 d
                                                   o 250 n9/m>—112 d
                                                   • CONTROLS	
                                tQ            V>0
                                    Day*  Post—«xposur«
                                                                       20O
Figure 3.  The effect  of pre-exposure  to various doses of diesel particulate on  the lung
clearance rate  of Fischer 344  rats after acute  exposure  to  '''C-tagged diesel particles.
From Chan et al.23
PARTICLE TRANSPORT MODEL
Summary of experimental data
   Briefly, the experimental  results of the deposition and clearance studies and additional
information available  in the  literature  on the clearance of submicron  particles may be
summarized as follows:
                                175

-------
   About 8% of the inhaled particulate deposits in the head of rats and rapidly enters the
GI tract;9 another 6-8% deposits in the  tracheobronchial region of  the  lung  (ciliated
                                                                               9.13
airways) and is cleared to the GI tract with a clearance half-time of about 1 day, *  and
                                                             9 13
about 12% deposits in the pulmonary (alveolar) region of the lung. '
   The particulate which  deposits in the pulmonary region is fairly quickly engulfed by
pulmonary macrophages and possibly other cell types.  One estimate of the half-time of
                            24
phagocytosis in mice is 1 day.
   Some of the scavenger  cells containing particulate travel  to the ciliated airways and
are cleared to the GI tract.   Others travel through the  lymphatic system  to  the lymph
      25
nodes.    The overall clearance half-time from the pulmonary region was estimated to be
60-80  days  in rats.  ^  In another  experiment, the fraction  of lavageable macrophages
which  contained diesel particulate decreased after the exposure ended with a half-time of
40 days in rats and 37 weeks in guinea pigs.
   Particularly after excessive exposures,  some particle-laden macrophages do not carry
                                                                        24
particulate out of the lung, but  aggregate  in pulmonary connective tissue,   in lymphoid
                           25                                          26 27
foci in the lung parenchyma,   and in complexes near terminal bronchioles.  '
Model development
             28
   Soderholm   reviewed the data and developed a particle transport model with compart-
mental  divisions illustrated in Figure 4.   Each  compartment  represents  a significant
"reservoir"  in  which particles can  be  found.   Four lung  compartments  are included;
particulate in the tracheobronchial region (T), free particulate lying on deep lung surfaces
(F),  particulate in macrophages  or  other  mobile scavenger cells (M),  and sequestered
particulate (S). Two compartments are external to the lung:  gastrointestinal tract (G) and
lung-draining lymph nodes (L).
   The  model consists of a  set of differential equations specifying the rate of transport of
particulate mass among compartments. The kinetics are assumed to be first  order since it
is  the simplest assumption which  can be made and since it leads to exponential clearance
curves,  the type found experimentally. Because the kinetics are assumed to be first order,
the clearance  rate for  each pathway is proportional to  the amount  of  material in  the
compartment being cleared.  Each proportionality constant is written in terms of the half-
time of clearance.  The half-times are designated by four-letter symbols (HTxy) where the
third letter  is the compartment which the particulate is leaving, and the fourth letter is
the compartment which the particulate is entering. The other parameters in  the model are
the deposition  rate into the tracheobronchial compartment (RT) and the deposition rate
into  the free  particulate compartment  (RF).   Each of  these  deposition  rates can be
calculated from the  deposition efficiency of the  compartment,  the animal's respiratory
minute  volume, and  the  airborne concentration of particulate.   The product of  minute
                               176

-------
volume and  the deposition efficiency of a compartment  is called  the  specific deposition
rate of that compartment.
                           =  LNG
                           SiiiiiiiiiiiHimiimiiiiiiiiimmiimiiiiiiimiimiftmiE
Figure 4.  Compartments and parameters in  the  lung  participate transport model:  T =
tracheobronehial, F = free  particulate on the deep  lung surface, M  = macrophages, S =
sequestered particulate, L = lymph nodes,  G = gastrointestinal tract, LNG = total lung, RT
= tracheal deposition rate,  RF  3 deep lung surface  deposition rate, HTxy  half-time for
clearance from compartment x to y.

   In a "short exposure" experiment,  the  animals  are exposed to high concentrations for
only a short period, and their changing particulate lung burdens are measured  starting  at
the end of the exposure. In  a "continuous exposure" experiment, the animals are exposed to
high concentrations continuously or nearly continuously over a long period  and  particulate
lung burdens are periodically measured.   This model is a first approximation in that the
clearance rates of the  individual compartments are assumed to be  independent  of the
amount of accumulated particulate and constant over the period of observation, even for
long-term continuous exposure experiments.
   Before analyzing the experimental  results, it  will  be  illustrative to consider the general
transport kinetics derived from the model for the two illustrative cases of "short exposure"
vs. "continuous exposure."  Considering first the  change in the total lung burden with time
after a short  exposure, we note that at  the end  of the  exposure all of  the  particulate
                               177

-------
resides in the "tracheobronchial" and "free participate" compartments.  The material in the
"tracheobronchial" compartment quickly clears with an assumed half-time of 1 day. Since
in rats a significant  fraction of the initial total lung burden was assumed to have deposited
in the "tracheobronehial" compartment, its fast clearance out of the lung is evident in the
decrease in the total lung burden over the first day after the exposure.  Meanwhile, the
material in the "free participate" compartment is picked up by scavenger cells with an
assumed half-time  of  2  days, causing  movement  of  paniculate  mass  from  the"free
participate" compartment to the "tnaerophage" compartment.  After  about 10  days, the
"free participate" compartment is empty and  all the participate is  in the  "macrophage"
compartment.   It  clears with  an  overall half-time consisting of  contributions from
clearance  into  the  GI  tract  and  clearance  into the lymph  nodes.   The  effect of a
sequestering region  would be  to cause the total  lung burden  to decrease to some level
above zero, rather than continuing to decrease to zero, as with no sequestering.
   For the "continuous exposure" case, particulate continuously deposits directly into the
"tracheobronchial" and "free particulate"  compartments.  The "tracheobronchial" compart-
ment burden reaches a plateau after a  few days at a relatively low level equal to 144% of
the  deposition rate times the  clearance  half-time.  Similarly,  the  "free particulate"
compartment quickly plateaus, reaching  a  level of 144%  of its deposition  rate  times its
clearance half-time. After several clearance  half-times, the "macrophage" compartment
content also plateaus, causing no further  increase in the total lung burden with continuing
exposure without sequestering.  The effect of a  "sequestered  particulate" compartment
would be to cause the total lung burden  to increase linearly with time after the particulate
burdens in the other compartments have  all plateaued.  It may  be observed that after the
first few days of exposure, the lung burden of chronically  exposed animals is insensitive to
the parameters of the faster clearing compartments.
              13
   Chan et al   fit  the post-exposure  clearance data with a two-phase exponential curve
whose  coefficients  and half-times  were  interpreted as indicating that 34% of  the initial
lung burden was deposited in the tracheobronchial region and cleared with a half-time of 1
day, while 66% of the  initial lung burden deposited in the pulmonary  region and cleared
with a half-time of 62 days.  These results can be  translated into  values  for the model
parameters.  When the  calculated parameters were plotted in Figure 5 against the original
data, using clearance half-time of the free particulate compartment  abitrarily set to  2
days, the analysis showed that the variation of the half-time between  1 or 4 days did not
significantly change the curve  for the total lung burden.  The satisfactory fit to the long-
term clearance data by  the model  when  the  experimentally determined  deposition and
clearance half-times are used shows  that  the  model  is  compatible  with  the  approach
originally used to interpret the data.
                               178

-------
                                            TRACHEO-BRONCHIAL
             0                 35                 70
                                   Time (days)
Figure 5.  Comparison of acute clearance data of diesel particles in the rat (Chan, et al.13)
with the lung clearance curve computed from the transport model using the experimentally
determined half-times.  From Soderholm (1981)29

   It is interesting to  note, however,  that  if the parameters derived from  this acute
exposure to diesel exhaust are used in the model and  the analysis  is cross-checked with
                                                                               12 21
data obtained after chronic  exposure of animals  to non-radioactive diesel exhaust  '   a
significant mismatch occurs.  The parameters derived from the acute exposure of the long-
term  clearance  data predict  that  the  lung burden  of  animals chronically  exposed would
eventually plateau as shown  in Figure  6.  However, the data from the chronic  exposure
experiment showed that the lung burden continually increased over  the period of observa-
tion.  The continual increase was originally interpreted  as indicating that at least some of
                                                       21
the deposited particulate was cleared very slowly, if at all.
   A revised set of model parameters was derived which provided a better description of
the chronic  exposure data.  Figure 7 shows that the model calculations, using the revised
parameters, do  not agree with the acute  exposure  experimental data, but predict higher
lung burdens at longer  times than were  observed.  Further consideration  of the model
solutions and the experimental data  reveals that no single set  of  model parameters can
satisfy all the available data.  This is a significant result, since it indicates that there is an
actual unaccounted  for difference among the experiments in the  respiratory system's
reaction to particulate.   It  is conceivable, if not likely,  that the  respiratory system
responds differently during a short  nose-only exposure than during a long-term whole-body
exposure.
   The revised  set of parameters used to calculate the curve in Figure 7 are  not the only
set which would fit the chronic exposure data.  The parameters were obtained as follows:
                                179

-------
          t>

          o
          o
           . u
          0.
          X
          in
          in
          O
          o
          a.
                     TOTAL LUNG
                       .AC
                       EE"
 	TMCHE.Qr^RONg.H.IAL
' UMACRQPHAGE
 •  550
 *  750
 »  1500
 •  6000
                    13      26     39     52     63
                                  Time (weeks)
                                                      78
                                                             91
                                                                   104
Figure 6.  Comparison of chronic  exposure lung burden data normalized by the exposure
concentration with model calculations using  the  actual intermittent exposure schedule.
The model parameters were taken  from the original analysis by Chan et al13 of their data
which describe  clearance  after a short  exposure.   Despite  the  fact that  the model
adequately describes the clearance kinetics for low doses (Figure 5), a significant mismatch
occurs when the parameters are used to model clearance for higher  doses of participate.
From  Soderholm (1981)"
          c
          
          T) d-
          k.

          GO
          C =>
          3
                                           TOTAL
                                  LUNG
                                  :o-BR(
TRACHEO-BRONCHIAL
rwE"VN:"DEE""
MACRbPHAGE
                              35
                                  Time (days)
                                                70
                                                                 105
Figure 7.   Comparison  of  Chan et al,13  data with model  calculations of particulate
transport after acute exposure using the set of  parameters modified to fit the long-term
clearance  data.  In each case lung burdens were normalized by the total lung burden
following acute exposure.  The comparison illustrates that a significant mismatch and over-
estimation  of lung retention  occurs  when parameters derived  from high  lung  dose
experiments are used to predict clearance at low doses. From Soderholm (1981)29
                             180

-------
the deposition efficiencies in the "tracheobronehial"  and "free particulate" compartments
and the clearance half-time  of the "tracheobronchial" compartment were taken from the
acute exposure and clearance study using   C, the minute volume  was assumed to be 200
mL for  rats  weighing approximately  400  g,  the  overall  clearance half-time  of the
"macrophage" compartment  was  taken from  experimental data, and the half-time of
clearance from the "macrophage"  compartment to the 'lymph  nodes"  was set to 107  days.
This effectively lumps all clearance out of the deep  lung into  one  overall clearance  half-
time since there is  little data available now on  the separate rates  of clearance to the
lymph nodes and GI tract.  With these assumed values, the overall clearance half-time out
of the deep lung and the clearance half-time into the "sequestered particulate" compart-
ment would each have to be 80 days  in order to fit the  slope of the  chronic data. There
may be  little significance in the values  of these two  half-times because of the large
number of assumptions which went into their derivation.  The principle result is that no set
of parameters fits all the data.
   The  major conceptual difference between this  model  of particulate kinetics in  the lung
                 8 22
and previous ones '    is that the deep  lung was  assigned three compartments based on
microscopic observations of  the fate  of diesel particulate in the respiratory system.  This
approach allows quantitative measurements of the particulate burden in a single eompart-
                                                           9fi
ment, for example, measurements of lavageable macrophages,   to be treated  within the
                                 Q
model.  The ICRP clearance model   includes a nasopharyngeal compartment and empha-
sizes particle dissolution.  These features could be added to the present model, although
they were deemed irrelevant  to the diesel particulate  body burden data available.  Inclusion
of the "free particulate" compartment does not significantly change the fit to experimental
data, but serves as a reminder that particulate resides on  the lung surface for a short time.
   The  acute exposure and  long-term  clearance  data give the deposition rate  into the
tracheobronchial and deep lung regions and the clearance  half-time of the  tracheobronchial
region.   The  contribution of  the model to the analysis of that data  is to point out  that the
long-term clearance  phase  may be interpreted in two  ways.  It may  be clearance of
                                                                                 13
particulate out of the deep lung with a half-time  of 62 days  as originally  interpreted,  or
it  may be a  combination  of deep lung clearance  and sequestering.   Only data taken at
longer times showing whether the  lung burden  plateaus above zero can resolve which
interpretation is correct.
   The  buildup of particulate in the lungs of chronically exposed animals was interpreted
as indicating that after excessive exposures some  portion of the deposited particulate was
retained for  long  times and indicated  the presence  of  a compartment with a prolonged
                                                        27
residence of diesel particles.  Since morphological studies'   reported the formation of
alveolar  macrophage  aggregates, it may be that  the sequestering compartment  actually
corresponds to this physical mode of isolating the diesel particles from the functional cells
                              181

-------
of the deep pulmonary region. The contribution of this work to the analysis of that data is
to relate the  rate of buildup to the fundamental transport parameters of the system. To
date, there is insufficient data to precisely set the values of all the parameters. However,
the calculations have shown that no single set of model parameters fit the data available
from experiments and more data must be collected to identify the source of the apparent
discrepancy.

DISCUSSION AND CONCLUSIONS
   Data from a variety of inhalation exposure studies has been presented and  compared in
an effort to better understand the deposition, clearance, and long-term retention of diesel
particles in the lung.   In response to observations of the apparent linear increase of the
lung burden after inhalation of high concentrations  of diesel exhaust particulate (250, 750,
                    O
1500, and 6000 yg/m ) for a long period of  time,  a model was derived which included a
                            29
"sequestering" compartment.    Material in this compartment clears very slowly, if at all,
and  is thought to be associated with  the aggregations of macrophages which have been
                             on
described by White and Garg.   Analysis of the chronic exposure data revealed that for
the rat,  7% of the inhaled particulate (assuming a respiratory  minute volume of 200 ml)
appears  to be held in the sequestering compartment which has a clearance half-time in
excess of one year.  Using this model, the long-term retention parameters which had fit the
chronic  exposure data  were  used to predict  the long-term results  of  the post-exposure
                                      13
clearance  experiments of Chan et al.,   who used radioactively-tagged  diesel particles.
The results were seen in Figure 7, and  show the predicted lung burdens to be significantly
higher than the actual ones for long times.   It appears likely that the discrepancy would
have been even greater if the post-exposure clearance had been followed for longer times.
Further  model calculations show that no single set of model parameters (minute volumes,
deposition efficiencies, clearance half-times) could  be found which would predict both sets
of experimental data.
   Thus, it is clear that under conditions of high-level exposures,  the model does  not
accurately reflect  all  the  major lung  clearance  mechanisms, and the  changes in their
contributions  with exposure conditions.  The  model calculations have quantitatively shown
that long-term retention does not occur to nearly as great an extent after a single exposure
in which less than 10 yg of radioactive  participates  is deposited in the pulmonary region as
during chronic exposure, when  hundreds of  micrograms  up  to several  milligrams have
deposited. This is not too surprising, since the long-term retention during chronic exposure
to high  concentrations is assumed to  be associated with aggregates  of macrophages, a
feature  not seen in unexposed lungs, or lungs containing very small particulate burdens.
Thus, it  appears that there is a threshold effect in the amount of long-term retention of
diesel  particulate in rats.   The threshold might correlate with  mass loading, exposure
                                182

-------
concentration, or exposure duration.  Only further studies using lower concentrations of
participate in chronic exposure experiments, and different exposure conditions in  the short
exposure, tagged-particulate clearance experiments can help to refine our understanding of
the threshold effect in long-term clearance.
                                                     •51
   One paper on this subject by Lewis  and Coughlin    measured lung burdens of acid
insoluble material (soot) in man at autopsy.  They reported an apparent linear increase in
lung burden with age and claimed that a fraction of the inhaled soot was retained for long
times.  On the other  hand, studies  of carbonaceous particle  content  in the lungs of coal
                            32
mine workers by Stober et al   suggests that a very active clearance of particles exists in
the human lung with a half-time of approximately 5 years. Consequently, the relevance of
the available  inhalation data to real-life conditions must be carefully weighed, especially
when extrapolating long-term retention data for health effects evaluations.
   In conclusion, this  work has shown that  low doses of diesel particulate material are
rapidly  cleared by  the lung  defense  system  but  that  species differences do  exist in
clearance and retention mechanics.  Therefore, test animal models should be  carefully
chosen when studying the long-term  retention of diesel particles to assure that the kinetics
will  appropriately relate,  if possible, to those  of the human lung.  Furthermore,  the data
available to date indicates that clearance and retention rates determined with hign particle
concentrations and high associated  lung burdens, are not equivalent to those measured at
lower levels of  exposure and particle burdens.  It  is clearly  evident from  this work that
information obtained at high levels of exposure cannot be scaled down to predict effects at
expected  ambient levels, since above certain,  as yet  undetermined thresholds, the  lung's
defense system  deals with  the deposited particles in different ways than it does  at  lower
levels.  In particular, the data indicate that the particle retention rates that  have been
found after  excessive inhalation exposures  were  much  higher than those  occurring at
minimal exposures.   Consequently, the  dose of  inhaled particles  and  their  resulting
biological activity which  could be  responsible for potential  adverse  health  effects  are
disproportionately larger when the inhaled concentrations are high than when the exposures
are minimal.  Based on the projected ambient  levels of diesel particles, and the  expected
minute real-life lung particle accumulations,  which  may range several orders of magnitude
below  those of  this study, it is concluded that the particle dose administered to  the
respiratory  system  by the increased  penetration  of diesel  engines  into  the light-duty
vehicle  fleet will be well  within  the coping  capacity of  the  human  lung clearance
mechanisms,  and that  the  pulmonary  defense system  may  be expected to be highly
effective in protecting the lung against the inhaled diesel particles.
                                183

-------
REFERENCES
~17Tssues Concerning the Light-duty Diesel (1979), Dept. of Energy, Washington, D.C.
 2.  Schreck,  R.M.,  McGrath, J.J.,  Swarin,  S.J.,  Bering, W.E.,  Groblicki, P.J. and
     MacDonald, J.S. (1978) Paper  No. 78-33.5, 71st Annual Meeting of the Air Pollution
     Control Assn., Houston, TX.
 3.  Vuk, C.T. and Johnson,  J.H. (1975) The Combustion Institute Central States-Western
     States 1975 Spring Technical Meeting, San Antonio, TX.
 4.  McGrath, JJ., Schreck, R.M. and  Siak, J.S.  (1978) Paper No. 78-33.6,  71st Annual
     Meeting of the Air Pollution Control Assn., Houston, TX.
 5.  Huisingh, J.,  Bradow,  R-, Jungers,  R.,  Claxton,  L.,  Zweidinger, R., Tejada, S.,
     Bumgarner, J., Duffield, F.,  Water, J., Simmon, V.F., Hare,  C., Rodriguez, C. and
     Snow, L. (1978) in Application  of short-term bioassays in the fractionation and analysis
     of complex environmental mixtures, EPA 600/9-78-027.
 6.  Siak,  J.S.,  Chan,  T.L.  and  Lee, P.S. (1980) in Health  Effects of  Diesel Engine
     Emissions: Proceedings of an International Symposium, Pepelko, W.E., Banner, R.M.
     and Clarke, N.A. eds., EPA-600/9-80-057a, 245-262.
 7.  Pederson, T.C. and Siak, J.S. (1981) J. of App. Tox., 1(2): 54-60.
 8.  Task Group on Lung Dynamics (Morrow,  P.E., Chairman) (1966)  Health  Physics, 12:
     173.
 9.  Raabe, O.G.,  Yeh, H.C., Newton, G.J., Phalen, R.F., and Velasquez, D.J. (1977) Inhaled
     Particles IV, Walton, W.H. ed., Pergamon Press, New York, N.Y.: 3.
 10.  Chan, T.L. and Lippmann, M. (1980) Am. Ind. Hyg. Assoc. J., 41: 399-409.
 11.  Schreck,  R.M., Chan,  T.L. and  Soderholm, S.C.  (1981)  Inhalation Toxicology and
     Technology, Leong, B.K., ed., Ann Arbor Science, Ann Arbor, Mich., 29-52.
 12.  Schreck, R.M., Soderholm, S.C., Chan, T.L., Smiler, K.L., and  D'Arcy, J.B. (1981) J.
     Appl. Tox., 1(2): 67-76.
 13.  Chan, T.L., Lee, P.S. and Bering, W.E. (1981) J. AppL Tox., 1(2): 77-82.
 14.  Guyton, A.C.  (1947)  Am.  J. Physiol., 150: 70-77.
 15.  Crosfill, M.L. and Widdicombe, J.G. (1961) J. Physiology, 158: 1-14.
 16.  Mauderly, J.L., Tesarek, J.E.,  Sifford,  L.J. and Sifford, L J. (1979) Lab. Animal Sci.,
     29(3): 323-329.
 17.  Lee, P.S., Chan, T.L. and Bering,  W.E.  (1981) Presented at EPA 1981  Diesel Emissions
     Symposium, Raleigh, N.C.
 18.  Stahlhofen, W., Gebhart, J. and Beyder, J. (1980) Am. Ind. Hyg. Assoc. J., 41(6): 385-
     398.
 19.  Schum, G.M. and Yeh, H. (1980) Bull. Math. Biology, 42(1): 1-15.
 20.  Chan, T.L. and Lee, P.S. (1981) Presented at EPA 1981 Diesel Emissions Symposium,
     Raleigh, N.C.
 21.  Rudd, C.J. and Strom, K.A. (1981) J. Appl. Tox., 1(2): 83-87.
 22.  Thomas, R.G. (1972) Assessment of  Airborne Particles, Stober,  W.  et al, eds., Thomas,
     C.C., Springfield, 01., 405.
 23.  Chan, T.L., Lee, P.S. and Bering,  W.E. (1982) Abstract: Society  of Toxicology Meeting,
     Boston, Mass.
 24.  Sorokin, S.P. and Brain, J.D. (1975) Anat. Rec., 181(3): 581.
 25.  Vostal, J.J., Chan, T.L., Garg,  B.D., Lee, P.S. and Strom, K.A. (1980)  Bealth Effects of
     Diesel Engine Emissions:  Proceedings  of  an International  Symposium,  United States
     Environmental Protection Agency  Report EPA-600/9-80-057b. 625-648.
 26.  Strom, K.A. (1981) Presented at EPA 1981  Diesel Emissions Symposium, Raleigh, N.C.
 27.  White, B.J. and Garg, B.D. <1981) J. Appl. Tox., 1(2): 104-110.
 28.  Karagianes, M.T., Palmer, R.F. and Busch, R.H. (1981} Am. Ind. Hyg. Assoc. J., 42(5):
     382.
 29.  Soderholm,  S.C. (1981) Presented atEPA 1981 Diesel Emissions Symposium, Raleigh,
     N.C.
 30.  White, H.J. and Garg, B.D. (1981) Presented at EPA 1981 Diesel Emissions Symposium,
     Raleigh, N.C.
 31.  Lewis, G.P. and Coughlin, L. (1973) Atmos. Environ. 7: 1249-1255.
 32.  Stiber, W., Einbrot,  H.J. and  Klosterkotter, W.U965) Inhaled Particles and Vapours  n,
     Davies, C.N. ed., Pergamon Press, London, 409-418.
                               184

-------
A SUBCHRONIC STUDY OF THE EFFECTS OF EXPOSURE
OF THREE SPECIES OF RODENTS TO DIESEL EXHAUST

Harold  L.  Kaplant,   William  F.  MacKenzie*,  Karl  J.  Springertt,  Richard  M.
Schreck** and Jaroslav J. Vestal**
tDepartment  of  Fire  Technology,   Southwest  Research  Institute,  San Antonio,
Texas; *Department  of Comparative Medicine, University of Texas Medical School,
Houston Texas; ttDepartment of Emissions Research, Southwest Research  Institute,
San  Antonio, Texas;  **Biomedical  Science Department, General  Motors Research
Laboratories, Warren, Michigan
INTRODUCTION
   The  projected  increase  in  the use  of diesel-powered automobiles  for fuel
economy  has  led to considerable concern over potentially adverse health effects
from  exposure  to the emissions of these engines.  The principal concern is with
the particulate matter produced by diesel  engines.   Diesel  exhaust contains 30
to  100  times  more particulate than  exhaust  from a  catalyst-equipped gasoline
engine  of comparable performance .  These  participates  are  small,  readily res-
pirable  and  are  composed of a  carbonaceous  core to which  a  variety of toxic,
mutagenie  and   carcinogenic  chemicals  are  adsorbed.    These chemicals,  when
extracted  and  concentrated,  have been shown to be mutagenic and carcinogenic by
a variety  of in vitro and in vivo assays    .  However, carcinogenic effects have
not  been  demonstrated  in chronic  inhalation  studies with  experimental animal
models,  although the number  of such  studies  has been  limited until recently.
   Early  last  year,  a large inhalation exposure facility was  constructed at the
Southwest  Foundation  for Research  and  Education in order to  investigate the
potential  health effects of  exposure to  diesel  exhaust emissions.   A chronic
15-month inhalation exposure study involving three dose  levels of diesel exhaust
particulate and three species of rodents was initiated in June, 1980.  In prepa-
ration  for this study,  a subchronic pilot  study  was  conducted at the high dose
level  of  diesel exhaust particulate.   Primary emphasis of this  study  was on
potential  carcinogenic  effects,   alterations   in  pulmonary ultrastructure and
morphometry  and  proliferative  changes  within  lung  epithelium as a  result of
exposure  to  diesel  exhaust.   This paper  reviews  some  of  the results  of this
study.
MATERIALS AND METHODS
   Diesel Exhaust Generation, Monitoring and Control
   Diesel  exhaust was  generated by a 5.7  liter Oldsmobile  engine operated con-
tinuously  at 40 mph 20 hours per  day,  7 days per week.  Hydrocarbons, CO, C02,
NO  and  particulates were  monitored on  a  periodic   basis and dilution was ad-
                              185

-------
justed,  as  necessary,  to maintain  a  1500 Mg/m3 particulate level.  The details
of the generation, monitoring  and control system,  as well as results of analyses
of particulates and gaseous  components,  are  described by Springer  .
   Exposure Chambers
   Two 8-ft cubical inhalation exposure  chambers,  constructed of stainless steel
and  glass,  were  used  in  this study  (Figure 1).   Each chamber,  including  its
pyramidal top and bottom,  had a  volume of  14.5  m3 and was  configured  to hold
four  stainless  steel  racks.  Each  rack  (Figure 2)  was  designed to hold 14 com-
partmented  stainless   steel cages  at 7 levels.   The  cages were  equipped with
removable  feeders and the  chambers, racks  and  cages  were equipped with  the
components for an internal  automatic  watering system.  Rat cages (Figure 3) were
compartmented with  removable dividers into  12  units and hamster and mouse cages
were  divided  into 24  units.   In this study,  one  rack  each was assigned to rats
and  hamsters  and two  racks were  used to  hold  mouse cages.   With  this arrange-
ment,  each  chamber had  the capacity for 168  rats, 336 hamsters  and  672 mice.
Figure 1.   View  of four  8-ft cubical  inhalation chambers  with diluted diesel
            exhaust/air  inlet systems on pyramidal  tops of chambers.  Pyramidal
            bottoms  of  chambers project into  basement area.   Each chamber has a
            volume  of  14.5 m3  and  accomodates  four  cage   racks.   Access  to
            chamber is provided by double doors on both sides of chamber.
                              186

-------
Figure 2.   Stainless steel rack holding 14 compartmented rat cages at 7 levels.
            Each rack is equipped with an automatic watering system manifold and
            trays for collection of urine and feces.
                           187

-------
Figure 3.   Stainless steel  cage,  compartmented with  removable dividers to pro-
            vide  capability  for  12  individually-housed  rats.   Each cage  is
            equipped with a central  removable  feeder and 12 lixits for automatic
            water supply.
   Animals
   The  animals used  in this  study were male  Fischer  344  rats,  Syrian  golden
hamsters and  Strain A/J mice.  The rats and  hamsters  were obtained from Charles
River Breeding  Laboratories and the mice were  obtained from the Jackson Labora-
tory.   All animals,  at  arrival,  were  approximately  six  weeks of  age.   Upon
receipt, the  animals were quarantined for  a  minimum of two weeks prior to their
introduction into the exposure  chambers.  Standard Purina rodent laboratory chow
and water were available ad libitum during  all  phases  of the study.
   The  Fischer 344  rats  were  supplied  in one shipment  and were  randomly and
equally divided  into experimental and control  animals at the end of the quaran-
tine period.  Hamsters and mice were received in two and four shipments, respec-
tively, at  approximately  weekly intervals.   Therefore, hamsters were identified
in the study as Group 1 or 2  hamsters and mice  as Group 1, 2, 3 or 4 mice.  Each
                              188

-------
group was  randomly and  equally divided,  at the  end  of its quarantine period,
into control  and experimental animals which were  simultaneously introduced into
the. two exposure chambers.  Although exposure of each group  of hamsters and mice
began at different times, all groups of  each  species  were  exposed for the same
length of time prior to removal from the chambers.
   Experimental
   In one  of the  two exposure chambers,  the Fischer  344 rats,  Syrian hamsters
and Strain A/J mice were  exposed to diluted diesel exhaust containing 1500 pg/m3
of particulate  20  hours  per day,  7  days  per week.  In the  second chamber, con-
trol groups  of  the three animal species  were  exposed  under the same regimen to
the filtered  air used to dilute the  raw  diesel exhaust.  Temperature was main-
tained at  22  ±  2°C and relative humidity at 50 ± 10% within the chambers.  The
animals  were exposed to  the diluted  diesel  exhaust  or  air for  a  tot^l  of
90 days, with the  exception of the hamsters which were removed from the chambers
at the end of 86 days of  exposure.
   During  the exposure period,  rats and hamsters  were  caged individually.  The
mice, however,  did not  adjust satisfactorily  to  the  automatic  watering system
when  they  were  separated  during  the  second   week  of  quarantine.   Therefore,
dividers were removed from  the  compartmented cages and  the mice  were caged in
groups  of  three.   Difficulties in obtaining  water were not  subsequently ob-
served.
   Engine  operation and  exposure of animals continued  on a 12:30 PM to 8:30 AM
schedule during the exposure phase.  During the four-hour daily engine shutdown,
a number of  data recording and animal  care  activities were accomplished.  Upon
inactivation  of the  engine,  animal  racks  were  removed  to permit  washing of
chamber interiors  with pressurized hot water and  detergent.  Feeders were emp-
tied  and   fresh  feed  was added daily.    Antibiotic  impregnated  cardboard  for
collection of feces and  urine was  replaced  on  alternate days to reduce ammonia
formation  within the  chambers.  Cages  were rotated within  each  rack and racks
were rotated  within each  chamber at regular  intervals  to equalize the exposure
of animals.  Animals were rotated to clean cages and racks on a weekly schedule.
   Animals were  observed  daily and body weights were obtained biweekly.  Sickly
animals were identified for further evaluation  and dead animals were removed and
necropsied, unless tissue autolysis  was  too advanced.   The disposition of all
animals, including spontaneous  or  accidental deaths and scheduled removals, was
recorded.
   At the  end of  86  days  of  exposure,  the hamsters were  removed  from the two
chambers.   The mice and rats were removed after 90 days of exposure.  From these
                              189

-------
animals,  thirty control  and thirty  experimental animals  of each  species  were
randomly  selected,  sacrificed by  an intraperitoneal  injection  of pentobarbital
sodium and exsanguinated  by  severing the  femoral arteries.   Each animal received
a complete necropsy prior to removal of  organs.  Lungs were inflated in situ by
tracheal  instillation of  a  measured  quantity of  10  percent buffered  formalin
prior  to  removal.   Other  organs  collected  for histopathological  examination
included  the heart,  liver,  kidneys,  spleen,   squamous  and  glandular  stomach,
ileum,  colon,   duodenum,  pancreas,  urinary bladder,  testicle,  adrenal  glands,
lymph nodes, larynx,  tongue, salivary  glands  and external  ear.   Heads,  denuded
of skin and muscle, were  decalcified and  three coronal sections were made of the
nasal cavity and paranasal sinuses.  A coronal section was  also collected  that
included  brain  and middle  ear,  as  well  as bone, bone marrow,  skeletal  muscle
and, often, Zymbol's glands and pituitary.   Tissues  were processed into  paraffin
by  standard  methods, cut at 6 microns and  stained with hematoxylin  and  eosin.
   Additional  control and  experimental  animals of each  species were  randomly
selected  for study of pulmonary ultrastructure and morphometry and proliferative
changes  in lung epithelium.   Lungs  from these animals were  inflated with  Kar-
oo vsky's  fixative for examination by light and electron microscopy.
   The  remaining animals were  placed  in a  separate holding  facility for post-
exposure  recovery  studies.   During the  recovery period, animals were  observed
daily and weighed at monthly  intervals.
   When  the  Strain A/J  mice reached nine  months  of   age,  500   control and 500
experimental animals  were randomly  selected for study of the pulmonary  adenoma
response.  This animal model was  first  applied  as  a quantitative bioassay of
chemicals  for  carcinogenic  activity in  1940  by  Shimkin   and,  since then,  has
                                                                   8  9
been used to detect the  carcinogenicity  of  a  number  of  chemicals '.   The  mice
were sacrificed by  an intraperitoneal  injection of pentobarbital sodium and the
lungs were immediately  removed,  rinsed and  .placed in  Tellyesniczky's fluid for
fixation.  Prior to  counting of adenomas,   the  lobes  of the  lungs  were  severed
from the  primary bronchus  to facilitate their examination.  The  adenomas  were
readily detected as pearly-white,  discrete nodules,  on the  surface of the lungs,
visible to the  naked eye or  under the dissecting microscope (Figure 4).  Light
microscopy was  used to confirm a  sampling  of the nodules  as adenomas.   A posi-
tive control group  of Strain A/J  mice was  used  to  verify  the pulmonary adenoma
response  to chemical carcinogens.   These  animals were  injected intraperitoneally
with urethane at a dose of 1  mg/g  of body weight at two months of age and sacri-
ficed for measurement of  tumor response four months  later.
                             190

-------
   At  the  end of  six months  post-exposure,  thirty  control  and thirty experi-
mental  animals  of  each of  the three  species  were  sacrificed  for histopatho-
logical examination.   The  protocol was the same  as  that followed in the histo-
pathology study at the termination of the 90-day exposure.
Figure 4.   The  lungs  of a  control (left)  and  an experimental  (right)  Strain
            A/J  mouse  showing  an  adenoma  on  each  lung.   The  experimental
            animal was  exposed  to  diesel exhaust containing 1500 pg/m3 particu-
            late for 90 days.

RESULTS
   The results  of  the  studies of pulmonary  architecture  and morphometry and of
proliferative  changes  within lung  exithelium are  aot  included  in  this  paper.
These results will be published with the results of the chronic study at a later
date.
   Mortality
   Spontaneous  deaths  of animals during the  three-month  exposure and six-month
recovery periods  consisted  of  3 control and  one  exposed  rats,  35  control and
31 exposed  hamsters  and  10  control and 7 exposed  mice.    These  mortality data
indicate that  exposure to  diesel  exhaust containing  1500  M8/n>3 °f particulate
                              191

-------
did not result in increased mortality rates in any  of the 3 species.
   Growth Patterns
   The mean  body weights  of control and  exposed  animals  of the 3 species are
shown in Figures 5 through 11.  Separate growth curves are shown for each of the
two  groups of  hamsters and each of  the  four  groups of  mice  since these groups
differed  slightly  in age and  weight when  introduced  into the  chambers.  No
significant differences were  found  at the  0.05 significance level in any of the
species when the  average changes  in body weight between  successive weighing
dates of  exposed animals  were  compared  with  those of  control  animals, using a
two sample t  test  .
             COMPARISON OF RAT GROWTH  CURVES
                        2-Month Exposure and Recovery
          Sod-
       S 400
       o>
       I
       >>
       •o
       o
       CD
300
          200
          100
                   I.I.I
                                  I .-I
             0  2   4  6   8 10  12  14 16  18 20 22 24 26 28 30 32 34 36 38
                       Exposure ond Recovery Time (wks)
Figure 5.    Mean and S.D. of body weights of Fischer 344 rats exposed  (	)
            to  diesel exhaust  containing  1500  pg/m3 particulate and controls
            (.	.)  during  a three-month exposure and a  six-month recovery
            period.
                            192

-------
            COMPARISON OF HAMSTER  GROWTH CURVES
                                Group-I
                      3-Month Exposure and Recovery
                                                  I
               I . I . I . I . I . I  i I ,  I
              2  4  6  8  10  12  14 16  18 20 22 24 26 28 30 32 34 36 38

                      Exposure  and Recovery Time (wks)
Figure 6.   Mean  and S.D of body  weights of Syrian hamsters exposed (•
           to  diesel  exhaust  containing  1500  Mg/m3 particulate  and controls
           (• ---- — -.)  during  a  three-month exposure and a six-month recovery
           period.

             COMPARISON OF HAMSTER GROWTH CURVES
                              Group-2
                    3-Month Exposure ond Recovery
         I50r-
           "0  2  4  6  8  10  12 14 16  18 20 22 24 26 28 30 32 34 36 38

                     Exposure ond Recovery Time (wks)
Figure 7.   Mean  and S.D of body  weights of Syrian hamsters exposed (•
                                                                       )
           to  diesel  exhaust  containing  1500  |jg/m3 particulate  and  controls
           (• ------- •)  during  a three-month exposure and a six-month  recovery
           period.
                          193

-------
              COMPARISON OF MOUSE GROWTH CURVES

                                Group-I
                      3-Month Exposure and Recovery
               2   4   6   8   10   12   14   16   18   20   22   24  26
                       Exposure and Recovery Time (wks)


Figure 8.   Mean and S.D of body weights  of Strain A/J mice exposed (•	•)'
           to  diesel  exhaust containing 1500  pg/m3 particulate  and controls
           (.	.)  during a three-month exposure and a three-month recovery
           period.

              COMPARISON OF MOUSE GROWTH CURVES
                                   Group-2
                       3-Month Exposure and Recovery
                    1.1,1.1.1.1,1,1
                2   4   6   8  10  12  14  16  18  20  22  24  26

                          Exposure and Recovery Time (wks)

Figure  9.   Mean  and  S.D of body weights of  Strain A/J mice exposed (•	0
           to diesel exhaust  containing 1500 Mg/m3 particulate and controls
           (•	•) during a three-month  exposure  and a three-month recovery
           period.
                           194

-------
               COMPARISON  OF MOUSE GROWTH CURVES
                                 Group-3
                       3-Month Exposure and Recovery
            0   2   4   6   8   IO  12  14   16  18   20  22  24  26

                        Exposure and Recovery Time (wks)

Figure  10.  Mean and  3-D of body weights of Strain A/J mice exposed  (	)
           to diesel exhaust  containing  1500 Mg/m3  particulate and controls
           (•—	•) during a three-month exposure and a three-month recovery
           period.

               COMPARISON  OF MOUSE GROWTH CURVES

                                 Group-4

                        3-Month Exposure ond Recovery
           34r-
                        6   8   10   12   14   16   18   20   22   24   26

                        Exposure and Recovery Time {wks)
Figure  11.  Mean and  S.D of body weights of Strain A/J mice exposed
                                                      '
•)
           to diesel exhaust  containing  1500 (Jg/m3 particulate and controls
           (.	.) during a three-month exposure and a three-month recovery
           period.
                           195

-------
ID
ov
              TABLE 1


              EFFECT OF EXPOSURE TO DIESEL EXHAUST (1500 (Jg/m3 PARTICULATE) FOR THREE MONTHS ON

              ORGAN WEIGHTS AND ORGAN-TO-BODY WEIGHT RATIOS IN MALE FISCHER 344 RATS3
Body
Weight (g)
Control
a-26
Exposed
n=29
314
+19
315
±22
.0
.0
.1
.3
Liver
g
10.84
±0.76
10.65
±1.04
g/100 g
3.46
±0.22
3.42
±0.25
Kidneys
g
2.07
±0.14
2.09
±0.16
g/100 g
0.65
±0.04
0.66
±0.04
Spleen
g
0.61
±0.06
0.61
±0.07
g/100 g
0.20
±0.02
0.19
±0.02
Heart
g
0.88
±0.09
0.87
±0.12
g/100 g
0
±0
0
±0
.28
.03
.28
.04
              a Values are Mean ± S.D.

-------
TABLE 2

EFFECT OF EXPOSURE TO DIESEL EXHAUST (1500 yg/ra3 PARTICULATE) FOR THREE MONTHS
ON ORGAN WEIGHTS AND ORGAN-TO-BODY WEIGHT RATIOS IN "SYRIAN HAMSTERS3

Control
n=30
Exposed
n=30
Body
Weight (g)
121.1
±18.8
126.2
±18.6
Liver
g
4.
±0.
A
±1
.26
.88
.65
.11
g/100 g
3.51
±0.35
3.67
±0.60
Kidneys
g
0.95
±0.11
1.00
±0.19
g/100 g
0.79
±0.10
0.81
±0.13
Spleen
g
0.13
±0.05
0.15
±0.05
g/100 g
0.12
±0.03
0.12
±0.04
Heart
g
0.51
±0.08
0.53
±0.06
g/100 g
0
±0
0,
±0.
.42
.08
.43
.07
3 Values are Mean ± S.D.

-------
    Organ Weights/Organ-to-Body Weight Ratios
    Organ weights  (liver,  kidneys,  spleen  and heart)  and  organ-to-body  weight
 ratios  of control and  experimental Fischer  344 rats and  Syrian hamsters  at the
 termination of the three-month exposure period  are  shown in Tables 1 and 2, res-
 pectively.   There were no significant differences at the 0.05 significance level
 in any  of  the organ  weights or  organ-to-body weight ratios  in  either species
 when means  of exposed animals  were  compared with control  means using  a two
 sample  t test  .
    Lung weights  and lung-to-body weight ratios  of control  and exposed Fischer
 344 rats and  Syrian hamsters are shown  in Table 3.   In both  species,  both ab-
 solute  and  relative (related to  100 g body weight) lung weights  were  slightly
 higher   in  animals  exposed to diesel exhaust  than  in  control animals.  These
'differences were  not  significant at p £  0.05,  using  a two  sample t test  ,
 except  for the difference  in relative lung  weights between  control  and exposed
 rats.   Increased  lung  weights after exposure  to  1500 Mg/m3 particulate  diesel
 exhaust have been reported previously, and were not the  result of water accumu-
 lation11.

 TABLE 3
 EFFECT  OF  EXPOSURE  TO DIESEL EXHAUST (1500 M8/m3  PARTICULATE) ON LUNG WEIGHTS
 AND LUNG-TO-BODY  WEIGHT  RATIOS IN  MALE FISCHER  344  RATS AND SYRIAN HAMSTERS

RATSb
Control
n = 15
Exposed
n = 15
HAMSTERS0
Control
n = 15
Exposed
n = 15
Body
Weight (g)


307.4 ± 20.1

296.4 ± 21.7


121.9 ± 17.4

125.9 ± 13.4
Lung
Weight (g)


1.03 ± 0.08

1.07 ± 0.08


0.54 ± 0.07

0.59 ± 0.08
Lung/Body Weight
(g/100 g)


0.33 t 0.03

0.36 ± 0.03d


0.45 t 0.05

0.47 ± 0.06
•0 Values are Mean  ±  S.D.
c Exposure duration  13  weeks
d Exposure duration  10  weeks
  Significantly  different from control, p <  0.05
                               198

-------
   Strain A/J Mouse Pulmonary Adenoma Response
   The  results  of the Strain A/J  pulmonary adenoma response study  are  shown  in
Table 4.  Lungs  of 458 control  and  485  exposed  animals were  examined  for  adeno-
mas.  This  difference in the number of  control and exposed  lungs resulted  from
the loss of lobes  from several  lungs during their  storage  in  fixative.   The  mean
number  of  tumors  per  mouse was  0.38  in the  control  group and 0.45  in  mice
exposed  to  diesel  exhaust.   These  values  were not  significantly  different  at
p  £ 0.05 when  compared,  using a  two  sample t test  .   Both  the  control and
exposed  values  are  somewhat higher than the 0.28  mean number  of  spontaneous
tumors in untreated  Strain  A mice at nine months of age reported in a review by
                   q
Shimkin and Stoner .  The prevalence of adenomas in the control group was 31.4%
and 34.2% in  the  exposed group.  These  values were  not significantly different
at 0.25  < p < 0.75 when  compared,  using the Chi Square test  .   In contrast to
the control and  exposed  animals, the positive urethane control group had a 100%
incidence of adenomas and a mean number of 22.6 tumors per mouse.
   Pathology
   Results  of  macroscopic   and microscopic  examinations  of  lungs  of  animals
exposed for three  months to diesel  exhaust containing 1500 |Jg/m3 of particulate
were consistent  in the  three species.  Gross examination revealed gray to black
discoloration of the lungs and mediastinal  lymph nodes.  The pigment was deposi-
ted both diffusely  and  in  focal  accumulations,  producing  a  grayish  overall
appearance  of the  lungs  with scattered,  denser  black  areas.   The deeply black-
pigmented lymph  nodes indicated  at least  partial  clearance  of  the  particulate
from the lungs via the lymphatics to regional lymph nodes.
   Microscopic examination revealed  no anatomic  changes in the upper respiratory
tract (Figure 12).   There was no  deposition  of  particulate  and the mucociliary
border  was  normal in  appearance.    In the  lungs,  particulate was observed dif-
fusely deposited  throughout  (Figure 13).  Most  of the particulate was in macro-
phages  but  some  was  free  as  small  aggregates  on  alveolar  and  bronchiolar
surfaces.   The  particulate-laden macrophages were  often  in  large accumulations
near the entrances of the lymphatic drainage and respiratory ducts  (Figure  14).
Associated  with  the larger  accumulations,  there was a minimal  increase in the
thickness  of the alveolar  walls   but  the  vast  majority  of  the  particulate
elicited no response.
                              199

-------
TABLE 4

PULMONARY ADENOMA RESPONSE OF STRAIN A/J MICE EXPOSED TO
DIESEL EXHAUST (1500 Hg/m3 PARTICULATE) FOR THREE MONTHS
ro
0


Control
Exposed
Urethane
(1 mg/g)







Number of Mice According to Number
Number
of Mice
458
485
18

of Lung Tumors/Mouse
0123
314 116 24 4
319 133 26 5
0000

4
0
1
0

13 >13
0 0
1 0
0 18

Total Number
of Tumors
176
217
407

Mean Number of
Lung Tumors/mouse
0.38 t 0.03
0.45 ± 0.04
22.6 ± 1.90

Percent
Prevalence
31.4
34.2
100

  Mean ± S.E.

-------
Figure 12.  The  mucoliliary border  of  a  mouse nasal  turbinate  at  the  end of
            three-months  exposure to diesel exhaust.   The  cilia  are normal and
            there is no deposition of particulate in the border.
                             201

-------
Figure 13.  Diffuse deposition of particulate in the lung of  a  rat  at  the end of
            three-months exposure to diesel exhaust.  Most  of the particulate is
            in macrophages, some is free in small aggregates.
                              202

-------
                                                     \  .
Figure 14.  Large  accumulations of particulate-laden macropnages  in  a rat lung
            at  end of three-months exposure to diesel exhaust.  Macrophages are
            aggregated  near  entrances  of lympahtic  drainage  and  respiratory
            ducts.
   After six months of recovery, the lungs of all three species had considerably
less  pigment  and  the pigment was  largely in focal  accumulations (Figure 15).
Intervening tissue was usually pink and normal appearing.  The lymph nodes were
still deeply pigmented, indicating  the continuing process of clearance of parti-
culate by  the  lymphatics.   Microscopically,  it  was  evident  that much,  if not
most, of the  pigment had been  cleared from the lung.  Of the  three species, the
tiamster  appeared  to  have  the greatest  capability for  removal  of particulate
(Figure 16).  The  mouse appeared to have the least  capability, but, even in this
species, most  of  the particulate had been cleared during  the recovery period.
Another apparent species  difference was the tendency  of  the  rat  to form aggre-
gates of particulate-laden  macrophages beneath the pleural surface of the lungs
(Figure  17).   There  are  lymphatics on this surface  and,  possibly,  the accumu-
lation  of  macrophages  in   this  area  represents  the process   of  particulate
removal.
   Miscellaneous lesions were observed in a number  of organs in each species but
the incidence of these lesions was  approximately the same in control and exposed
animals.   There was  no  indication that exposure to  diesel  exhaust resulted in
pulmonary  pathology,   other  than  the  accumulation  of particulate,  or  in any
pathological changes  in other organs.
                              203

-------
Figure 15.  A  lung from a control  (left)  and an exposed (right) Fischer 344 rat
            at  six-months recovery  following  a  three-month  exposure to diesel
            exhaust.   Most of  the particulate was  cleared during the recovery
            period.

Figure 16.  The  lung  of  a  Syrian  hamster  at  six-months  recovery following a
            three-month exposure to  diesel exhaust.  The  very small amounts of
            remaining  particulate indicate the effective clearance capability of
            this  species.
                              204

-------
                              'O~V.  3*tiif--iri*?*;
                                                                 >^-i>
Figure 17.  Accumulations  of particulate-laden macrophages  beneath the pleural
            surfaces  of a  rat lung  at six-months  recovery following  a three-
            month  exposure  to  diesel exhaust.  Except  for  these accumulations,
            the  lung  has a  normal appearance.
                              205

-------
ACKNOWLEDGEMENTS
   The author  is  grateful to Douglas  Malsbury for his  technical assistance and
Laura Berger and Deaise Taylor for their clerical  assistance.


REFERENCES

 1.   Vostal, J.J. (1980) Bull, N.Y. Acad. Sci., 56,  914-934.

 2.   Claxton, L.D.  (1979)  in Health Effects  of  Diesel Engine  Emissions:  Pro-
      ceedings  of an  International Symposium,  Pepelko, W.E.,  Banner, R.M and
      Clark, N.A. ed., Cincinnati, Ohio, pp. 801-807.

 3.   Huisingh, J., Bradow, R., Jungers, R., Claxton,  L., Zweidinger,  R.,
      Tejada,  S.,  Bumgarner, J., Duffield,  F. and Waters,  M.   (1978) U. S. EPA
      Health Effects  Research Laboratory  Publication No. EPA-600/9-78-027, pp.
      1-32.

 4.   Nesnow,  S.  (1979)  in Health Effects of Diesel  Engine  Emissions:   Proceed-
      ings  of  an International  Symposium,  Pepelko,  W.E.,  Banner,  R.M.  and
      Clarke, N.A. ed., Cincinnati, Ohio, pp. 898-912.

 5.   Slaga, T.J.,  Triplett, L.L.  and Nesnow,  S. (1979)  in Health Effects of
      Diesel  Engine  Emissions:   Proceedings  of  an  International  Symposium,
      Pepelko, W.E.,  Banner, R.M.  and Clark,  N.A.  ed., Cincinnati,  Ohio,  pp.
      874-897.

 6.   Springer,  K.  (1981)  EPA  1981 Diesel  Emissions  Symposium, Raleigh,  Oct.
      5-7.

 7.   Shimkin, M.B. (1940) Arch. Pathol. 29, 239-255.

 8.   Mirvish, S.S. (1968) Advanc. Cancer Res.  11, 1-42.

 9.   Shimkin, M.B. and Stoner, G.D. (1975) Advan  Cancer Res. 21,  1-58.

10.   Kempthorne, 0.  (1952)  Design  and Analysis of Experiments.   John Wiley and
      Jones, Inc., New York, pp 1-631.

11.   Misiorowski, R.L., Strom, K-A-, Vostal, J..J. and Chvapil,  M.  (1980)
      in Health Effects  of Diesel Engine Emissions:   Proceedings of an Interna-
      tional Symposium, Pepelko, W.E., Danner, R.M. and Clark, N.A.  ed., Cincin-
      nati, Ohio, pp. 465-479.
                              206

-------
       PULMONARY FUNCTION TESTING OF RATS CHRONICALLY EXPOSED
                   TO DILUTED DIESEL EXHAUST FOR 612 DAYS


                                    K. B. Gross
                          Biomedical Science Department
                       General Motors Research Laboratories
                                 Warren, MI 48090


Diesel engine emissions contain particulate matter composed of a multitude of organic
compounds, and of a size that may be readily inhaled and retained by the lung. It thus
has the potential for interacting with the deep lung, and possibly altering the structure
and function of this organ. The study was designed to address the question of whether
the  chronic  inhalation of  diluted  diesel  exhaust  may affect  pulmonary function.
Twenty-five Fischer-344 rats were exposed to diesel exhaust, diluted with clean air at a
ratio of 1:15 (particulate concentration = 1500 ug/m ) for 20 hrs/day, 5-1/2 days/week,
for 612 days.   Twenty-five  control animals  were treated in a  similar manner,  but
exposed to clean filtered  air. Noninvasive pulmonary function testing, which produced
no apparent harmful effects,  were performed on the animals at mean times of 11, 23, 30,
38,  51, 65, and  87 weeks  on  the exposure regimen.  Animals were anesthetized
(Halothane, 5%), transorally intubated, and placed in a plethysmograph.   Measurements
of lung volumes and flows were made while the animals  were spontaneously breathing
and during forced expiratory  maneuvers.  Functional residual capacities were computed
using the  Boyles  law principle.  All data was normalized by each animal's own forced
vital capacity in order  to compensate for animal growth and interindividual variability.
Through the first year of testing, no significant  differences between the two groups
were found for  any of  the measured parameters.  During the second year of exposure,
seven of the eighteen  measured indices  displayed  a significant difference  as indicated
by analysis of variance.  The  normalized  functional  residual capacity  (FRO and  its
component volumes -  expiratory  reserve  (ER)  and residual volume  (RV)  -  maximum
expiratory flows at 40% (MEF.J and 20% (MEF™)  of the lung  volume remaining,  and
the forced expiratory volume in 0.1 sec  (FEV.i fwere all greater in the  diesel-exposed
animals.  The  normalized inspiratory  capacity  (1C)  was significantly  larger  in  the
control group, but  the  test  point of greatest difference for this parameter  does  not
exceed 4%, and at all other test points is 2% or less.  The statistical significance of the
1C is interpreted as being the result of the very low variability of this parameter, rather
than  a  result of  a clinically important  change in  the pulmonary  function.   The
significantly larger values for relative  normalized FRC, ER,  and RV  in the  diesel-
exposed animals could be indicative of chronic obstructive lung disease and are similar
to the changes seen in  other  studies in which an emphysema-like condition  was induced
in rats by intratracheal instillation of elastase, as well as in reported clinical data on
chronic obstructive lung disease.  However, this interpretation of the changes in  the
three lung volumes is contradicted by the MEF .Q, MEF20, and FEV 1 values. If chronic
lung disease was occurring,  these parameters would  oe  expected °to decrease in  the
                                       207

-------
 diesel-exposed animals compared to the controls.  The fact that they increased in the
 diesel group, suggesting an improvement in airway caliber, is not consistent with what
 would be expected based on  studies reported in the literature for  human pulmonary
 disease.   In conclusion,  the majority  of the  measured  parameters  did  not  differ
 significantly between the control and  diesel-exposed  groups,  and  while  one cannot
 exclude  the  possibility that the differences that were observed  in this  experiment
 between the diesel-exposed and clean air controls may be attributable to  the chronic
 inhalation of the diesel exhaust, the results are not consistent with documented clinical
 findings on chronic lung disease.
                                                             la
                      03-
                      02-
                      01
                                                          Legend
                                                          CONTROL
                                                          DIrtc-t)
                            100   200  300  400  300   «00  TOO
                           DAYS ON EXPOSURE REGIMEN
                                                             1b
                    •3
                    O

                    b.
                       8-
                                                          Legend
                                                           CONTROL
                        0   100   a»  300  400   900   600  TOO
                           DAYS ON EXPOSURE REGIMEN
Figure la   Functional  residual capacity (FRC) of diesel exposed (-	) and  control
(      ) animals as a fraction of each animal's forced vital capacity (FVC). The bottom
curve (	) is the difference between the means of the two  groups  (control -
experimental).

Figure Ib  Normalized maximum expiratory flow rate at 20% of vital capacity (MEFon).
Legend same as Figure la.                                                        20
                                       208

-------
                     PULMONARY FUNCTIONAL RESPONSE IN CATS
                FOLLOWING TWO YEARS OF DIESEL EXHAUST EXPOSURE
                     William J. Moorman and John C. Clark
             National Institute for Occupational Safety and Health
                 Division of Biomedical and Behavioral Science
                             4676 Columbia Parkway
                             Cincinnati, OH  45226

                                     and
                      William E. Pepelko and Joan Mattox
                     U.S. Environmental Protection Agency
                      Health Effects Research Laboratory
                             Cincinnati, OH  45268
INTRODUCTION:

Both NIOSH and the EPA have responsibilities to assess potential health
effects from exposure to diesel engine exhaust.  These responsibilities are
mandated under both the OSH and Clean Air Acts, respectively.  While EPA's
responsibility relates to the general population, NIOSH's responsibility is
to the worker, especially those at high exposure risk in the mining and
transportation industries.

The advantages of diesel power are manyfold; however, the prime motivation
for the use of diesels relates to the reduced cost of operation due to longer
engine life, decreased maintenance, cheaper fuel, increased specific energy
(2.4 X that of gasoline), and increased safety.  The increase in safety is
related to diesel fuel's lower volatility, flash point with a concomitantly
reduced explosive hazard when compared to gasoline.  As a result of our
proximal locations in Cincinnati, the NIOSH, DBBS Cardiopulmonary Laboratory
participated with EPA-HERL Laboratory to apply comprehensive pulmonary
function testing to the cats exposed to the controlled diesel atmospheres.

BACKGROUND:

Evaluation of pulmonary function responses to experimentally controlled
diesel exhaust is primarily confined to investigation of the past few years.
Previous studies are largely epidemiologic and while they are important
guides, they lack qualitative and quantitative characterization, essential
for analytical, dose-response toxicology.

At the last symposium on Health Effects of Diesel Engine Emissions, Gross
reported pulmonary function findings in Fischer 344 rats following 38 weeks
of exposure at 1500 ug/m3 for 5-1/2 days/week, 20 hrs/dayd) .  He found no
differences in mechanical properties, lung volumes, or dynamic ventilatory
performance.  In a later report presented at the Society of Toxicology meet-
ing (1980),  he found higher function residual capacity and maximum expiratory
                                     209

-------
flow rates at 20% of vital capacity in the diesel exposed group.  These re-
sults are not interpretable or consistent with dysfunction.  His conclusion
was that no clinically important alterations were observed after 65 weeks of
exposure(2).

In another report, O'Neil et al., reported on functional and morphological
consequences of diesel exhaust exposure in mice following 3 months of expo-
sure to a 1:18 dilution (diesel and air) of diesel exhaust with particulate
levels of 6.4 mg/m3.  He found no statistically significant findings in lung
volume or diffusing capacity(3).

In a third report on functional responses in rodents, Vinegar et al., de-
scribed significant decrements in vital capacity, residual volume, and
diffusing capacity in Chinese hamsters exposed to a dilution of diesel
exhaust with a particulate level of 6.4 mg/m3(4).

In the only larger animal exposure, we reported findings from one year of
exposure in cats(5).  The exposure was conducted at the Cincinnati EPA, HERL
laboratory.  A 1:18 dilution diesel exhaust with 6.4 mg/m3 was used.  Our
results indicated no response in mechanical properties, lung volumes,
distribution, diffusing capacity or ventilatory performance following one
year of exposure.

METHODS:

In an effort to enhance functional response characteristics, the diesel
exposure was increased by decreasing the dilution ratio from 1:18 to 1:9
after the first year.  A summary of exposure is provided in Table I.  During
this 'second year, 19 of the original 21 male cats were exposed to 11.7 mg/m
particulate, 4.37 ppm nitrogen dioxide, 5.03 ppm sulfur dioxide, and 33.30
ppm carbon monoxide with total hydrocarbons of 7.72 ppm.  Twenty male cats
served as controls.  All cats were young adult males obtained from Liberty
Laboratories and were born and maintained in a disease-free environment and
inbred for several generations.  They were of uniform size (3.63 ± 0.46 kg)
and within two weeks of the same age.

Prior to pulmonary function testing, the cats were fasted for one day.  The
testing followed 18-20 hours of no diesel exposure.  On a random schedule,
the cats to be tested were anesthesized with ketamine and acepromizine at a
dose of 42 mgAg.  Following induction of the anesthesia, an esophageal
balloon was placed in the lower' third of the esophagus and an 18-22 F
endotracheal tube  (largest possible) was inserted into the trachea with the
aid of a laryngoscope.  All testing, except compliance (CL) and resistance
(RL) was performed with the cats placed in the prone position in a variable
pressure plethysmograph.  CL and RL tests were performed with the cats re-
cumbent to facilitate measurement of transpulmonary pressure.  Figure I
shows the general pulmonary testing situation diagrammatically.

Pulmonary mechanics were obtained from simultaneous volume, flow and trans-
pulmonary pressure.  Dynamic compliance (CLdym) was measured from volume
and transpulmonary pressure at points of no flow.  Average flow resistance
(RLave.flow) was measured from change in transpulmonary pressure at equal
                                     210

-------
volumes, divided by the sum inspiratory and expiratory flow.  All mechanics
were obtained while the cats were spontaneously breathing  (15-25 breaths/min).
The pulmonary function tests requiring breathing maneuvers  [lung volumes,
forced expiratory flows (FEF% Vol) , diffusing capacity (DI^l^) , nitrogen
washout (A^) , and closing volume (CV)] were performed using a  variable
pressure plethysmographic chamber previously described(6).  The methods of
Brashear et al.(7) and Mitchell at al. (8) were combined to obtain values for
DLC180 and total lung capacity (TLC).  The calculations for DLC180 were per-
formed according to the methods described by Wagner at al., for  Cl8o(9).
All gas analyses were done using a respiratory mass spectrometer (Perkin-
Elmer MGA1100).  Distribution was studied using the single-breath nitrogen
washout and closing volume adapted from the human methods described by Buist
and Ross(10).  All data was tested statistically by nonparametric, Kruskal-
Wallis one-way rank analysis of variance.

RES COLTS:

Following the first year of exposure, no significant differences were found
in mechanical properties, diffusing capacity, uniformity of distribution or
ventilatory performance.  In contrast to the negative findings  following the
first year, we now have clearly defined responses at the end of two years.
Table II presents all parameters studied for control and exposed cats con-
trasting the values for one and two years.

The reduction in inspiratory capacity, vital capacity, and total lung
capacity with normal values for ventilatory function (mechanics of breathing)
indicates that a lesion is present which restricts breathing but does not
cause airway obstruction or loss of elasticity.  This restrictive disease
found in this study is compatible with a diagnosis of pulmonary fibrosis of
the interstitial or intraalveolar type.  Concurrent status may  include
chronic inflammation, interstitial edema, or vascular engorgement.  Addi-
tional support for the diagnosis of interstitial disease is the finding of
impaired diffusing capacity.  Distribution of this disease appears nonuni-
form as indicated by the significantly elevated nitrogen washout values for
the exposed group.

DISCUSSION:

Pathological description of pulmonary responses to diesel exhaust at similar
concentrations has been previously characterized(11,12) .  The observations
include:  (1) marked accumulation of black pigment laden macrophage in the
interstitum localizing around blood vessels and respiratory bronchioles; (2)
hyperplasia of the alveolar lining cells with focal thickening  of the inter-
stitium; (3) interstitial pneumonitis; (4) traces of, or no emphysema or
peribronchiolitis.  While the pathological examination of the cats' lung
not complete, the above description is consistent with our physiologic find-
ing of restrictive lung disease.
                                      211

-------
 REFERENCES:

 1.  Gross, K.B.  1980.  Pulmonary Function Tasting of Animals Chronically
     Exposed to Diluted Diesel Exhaust.  Presented at the Environmental
     Protection Agency International Symposium on Health Effects of Diesel
     Engine Emissions. Cincinnati, Ohio.

 2.  Gross, K.B.  1981.  Noninvasive Pulmonary Function Testing of Fischer 344
     Rats Chronically Exposed to Diluted Diesel Exhaust for Fifteen Months.
     The Toxicologist. Vol. 1, No. 1.

 3.  O'Neil, J.J. et al., 1980.  Functional and Morphological Consequences
     of Diesel Exhaust Inhalation in Mice.   Presented at the Environmental
     Protection Agency International Symposium on Health Effects of Diesel
     Engine Emissions. Cincinnati, Ohio.

 4.  Vinegar, A., et al. 1980.  Pulmonary Function Changes in Chinese Hamsters
     Exposed Six Months to Diesel Exhaust.   Presented at the Environmental
     Protection Agency International Symposium on Health Effects of Diesel
     Engine Emissions.  Cincinnati, Ohio.

 5.  Pepelko, W.E., et  al.  1980.  Pulmonary Function Evaluation of Cats After
     One Year of Exposure to Diesel Exhaust.  Presented at the Environmental
     Protection Agency International Symposium on Health Effects of Diesel
     Engine Emissions.  Cincinnati, Ohio.

 6.  Moorman, W.J., et al.  1975.  Maximum Expiratory Flow Volume Studies on
     Monkeys Exposed to Bituminous Coal Dust.  J. Appl.  Physiol. 39:444-448.

 7.  Brashear, R.E., et al.  1966.  Pulmonary Diffusion and Capillary Blood
     Volume in Dogs at Rest and With Exercise.  J. Appl. Physiol. 21:520-526.

 8.  Mitchell, N.M., et al.  1968.  Application of the Single-Breath Method of
     Total Lung Capacity Measurement to the Calculation of Carbon Monoxide Dif-
     fusing Capacity.  Am. Rev. Resp. Dis.  97:581-584.

 9.  Wagner,  P.O., et al.  1971.  Diffusing Capacity and Anatomic Dead Space
     for Carbon Monoxide [C18O].  J. Appl.  Physiol. 31:817-852.

10.  Buist, A.S.,  et al.  1973.  Quantitative Analysis of Alveolar Plateau
     in the Diagnosis of Early Airway Obstruction.  Am.  Rev.  Resp. Dis.
     108:1078-1087.

11.  Wiester, M.J., et al.  1980.  Altered Function and Histology in Guinea
     Pigs After Inhalation of Diesel Exhaust.  Environ.  Res.  22:285-297.

12.  Karagiones, M.T.,  et al.  1981.  Effects of Inhaled Diesel Emissions and
     Coal Dust in Rats.  Am.  Ind. Hyg. Assoc. J.  42:382-391.
                                      212

-------
     Table I.  EXPOSURE CHAMBER COMPONENT CONCENTRATIONS,  STUDY AVERAGES
Weeks #1-61
Dilution Factor
Particulate Mass
Nitrogen Oxides
Nitric Oxide,
(air:diesel)
, mg/m3

ppm



Nitrogen Dioxide, ppm
Sulfur Dioxide,
ppm
Total Hydrocarbons , ppn
Carbon Monoxide ,
Carbon Dioxide,
ppm
%
Table II. PULMONARY

DF
M
NOX
NO
NO 2
S02
THCcorr.


CO
C02
18.16
6.34

11.64
2.68
2.12
4.15
20.17
0.30
FUNCTION PARAMETERS
± 1.72:1
± 0.81

± 2.34
± 0.80
± 0.58
± 0.97
± 3.01
± 0.04
COMPARING
GROUP TO THE DIESEL EXPOSED GROUP AFTER 1 YEAR
Mechanical
Properties
^dyn.
^ave.flow
Lung Volumes
TLC
FVC
FRC
ERV
RV
RV/TLC%
1C
Ventilatory
Performance
FEV.5%
PEFR
FEF50
FEF25
FEF10
FEF40%TLC
Diffusion
DLCO
Distribution and
Closing Volume
%N2/25%/VC
CV

Cine*
[
1 Exposed
23.5 ±
10.7 ±

415 ±
348 ±
158 ±
69 ±
86 ±
20.3 ±
279 ±


84.3 ±
1016 ±
728 ±
490 ±
196 ±
486 ±

1.18 ±


0.32 ±
25.6 ±
7.2
4.6

56.0
43.5
35.6
24.6
36.9
6.9
44.8


8.4
185
196
186.8
107.4
252.6

.43


.20
13. 4**
YfflST

Control
23.7
10.3

449
368.9
165
67
104
22.7
301


81.5
1042
761
481
222
557

1.22


0.29
36.0
± 9.3
± 4.4

±74.5
± 42.1
±42.2
±19.0
± 37.7
± 5.9
± 49.6


± 6.4
± 174
± 160
± 199.
± 156.
± 248.

± .40


± .30
±16.1

Weeks 62-124
9.
11.

19.
4.
5.
7.
33.
37 ±
70 ±

49 ±
37 ±
03 ±
22 ±
30 ±
0.52 ±
1.13:1
0.99

3.80
1.19
1.03
0.85
2.94
0.04
THE CONTROL
AND 2 YEARS
Two Ye
1 1 Exposed
27.5 ±
5.6 ±

428 ±
369 ±
145 ±
79 ±
67 ±
15.6 ±
291 ±


86.9 ±
887 ±
802 ±
5 518 ±
8 223 ±
0 586 ±

0.39 ±


0.39 ±
27 ±
4.9
3.2

56.3**
42 . 34- *
26. 2+*
24.0
14.3
1.9
44. H-*


6.1
98 4-*
125
154
109
173

.27 4-*


.27 t*
17.6
flI-<=[

1
Control 1
26.2
5.7

484
410
163
83
80
16.4
328


86.9
952
864
574
234
625

1.01


0.21
25
± 7.1
± 2.3

± 68.3
± 57.6
± 36.9
± 34.5
± 28.2
± 4.5
± 58.6


± 5.9
± 110.7
± 121
± 153
± 102
± 213

± .14


± .181
± 19.3
*Statistically significant P < 0.05
                                     213

-------
            NEOPRENE
           DIAPHRAGM

     SPIROMETER
    (It. Dry roHIng seal)
           4-WAY
           VALVE
          ASSEMBLY
HYDRAULIC
    PUMP


7 '/2 H R

MOTOR
                                SLAVE
                               CYLINDER
                                                    EQUILIBRATION
                                                       VALVE
                                        LIMIT
                                        SWITCH
                STAINLESS
                 STEEL
                CHAMBER
                  FLOWNATE
          AUTO. MANUAL
          o o o o
          IN OUT IN OUT  '••
    MASS
SPECTROMETER
          CONTROL PANEL
                                 RECORDER
                                    TRANSDUCER
                                                          PNEUMOTACH
1 1
PRINTER



COMPUTER
A-D Converter
CPU
Disc Drive
Top* slorogA


                Diagram of Pulmonary Function Lab.

-------
      DEPOSITION AND RETENTION OF SURROGATE AND ACTUAL DIESEL PARTICLES
                                      by
       R. K. Wolff, L. C. Griffls, G. M. Kanapilly and R. 0. McClellan
              Lovelace Inhalation Toxicology Research Institute
                                P.O. Box 5890
                       Albuquerque,  New Mexico  87185


INTRODUCTION

    Data on deposition and retention of  diesel particles are an important
need in assessing their toxicological impact.  Deposition and retention
experiments were carried out with °'Ga aggregate aerosols (surrogate
diesel particles) in Beagle dogs to provide good estimates for human depo-
sition.  Experiments were also carried out with these particles in Fischer-
344 rats to provide comparative information in a small laboratory animal
and also to estimate lung burdens in chronic exposures to diesel  exhaust.
A method was developed to quantitate actual diesel soot burdens in rats
exposed to diesel exhaust.  These values were found to be comparable to
these predicted from the surrogate particle deposition experiments.

METHODS

    Aggregated particles of ^63203, 0.02 and 0.1  urn volume median
diameter (VMD), were produced using heat treatment of 6?Ga tetramethyl-
heptanedione using methods described previously (1).  Ten Beagle dogs from
the Institute's colony were exposed for 1/2 hr in  a nose-only exposure unit
equipped with a plethysmograph for pulmonary function monitoring.   Whole
body counting and gamma camera analysis were used  to measure total  amounts
of activity deposited and their regional distribution.  A total  of 144
Fischer-344 rats were exposed 5 hrs/day for either 1  or 3 days in  the same
style multi-tiered exposure chamber used in the chronic diesel exhaust ex-
posure study.   Whole body counting and also sacrifice and tissue  counting
methods were used to measure deposited radioactivity.

    The diesel  exhaust exposure system consisted of an Oldsmobile  5.71
diesel engine connected to a dynamometer with the  engine load and speed
determined by an analog control system.   The engine was operated  on a
7 mode urban cycle and the fuel was U.S. Department of Energy Reference
fuel  8007 (Phillips Chemical  Co.).   The entire engine exhaust was  diluted
in a large stainless steel  tunnel  and then further diluted in 3  stages with
filtered air which produced average particulate concentrations of  4150, 990
and 200 ug/m3.   The diluted exhaust was drawn through the 2.2 mj  ex-
posure chambers (Hazleton Systems,  Inc., Aberdeen, MD) at 560 1/min.  The
exposure schedule was 7 hrs/day,  5  days/week.
                                    215

-------
    Laboratory reared F1scher-344 rats were 12-13 weeks old at the initla-
tion of the study.  They were exposed for a cumulative period of 541 hours
over 18 weeks.  At the end of this period 8 rats from each exposure group
were sacrificed.  Lungs were removed, homogenized, and centrffuged to pro-
duce a cell pellet.  This tissue pellet was dissolved 1n 1 ml H£0 and
2 ml tetramethyl ammonium hydroxide.  The remaining "soot" particles were
suspended  1n 5 ml ^0.  Light absorbance at 690 nm was measured and com-
pared against standards prepared from known weights of dlesel particles
collected on filters from the dilution tunnel  of the Inhalation exposure
system.  For these collections the same engine cycle was used as for the
animal exposures.

RESULTS AMD DISCUSSION

    Table  1 shows the mean total and regional  deposition values measured 1n
the Beagle dogs for the 0.02 and 0.01 ym 676a203 particles.   Depo-
sition was higher 1n all compartments for the  0.02 urn particles.   Despite
the overall high variability 1n deposition 9 of the 10 dogs had higher
deposition at 0.02 nm than 0.1  wm and the difference was statistically
significant (P < .05).  Although most of the material was deposited 1n
the pulmonary region, deposition 1n the nasopharyngeal ,  and tracheobron-
chial regions was becoming increasingly significant as particle size
decreased.  Figure 1 shows pulmonary deposition of the 0.02 and 0.1  urn
particles  is in good agreement with the trends in deposition observed pre-
viously in humans (2,3) and Beagle dogs (4)  at larger particle sizes.   The
deposition values are lower than predicted by the ICRP Task Group on Lung
Dynamics (5) but are in good agreement with  predictions  by Yen and Schum
(6) and also Yu (7).
    Deposition of the 0.1 urn   5^3 particles was somewhat lower
in rats than had been found 1n dogs.  Lung deposition (bronchial  and pul-
monary) was estimated to be 15 ± 3$.  This estimate was  based on  measured
lung burdens and assuming minute volumes as measured in  free standing
Fischer-344 rats at this Institute.   Absolute lung deposition was found to
be 2.4 ug/hr for a particle mass concentration of 1000 ug/m3.  Using
these initial deposition values and a measured lung half-life of  75  days
for °'6a203, lung burdens could be calculated for various exposure
periods.  Table 2 shows the predictions for the exposure period and  concen-
trations experienced by the animals  1n the chronic dlesel  exhaust exposures.

    Measured diesel soot burdens following the 18 week exposure are  also
shown in Table 2.  The observed values were in good agreement at  the high-
est exposure level but overestimated deposition at the two lower  levels.
Either deposition was higher or clearance was slower at  the high  exposure
level compared to the lower levels.

    The degree of agreement between  predicted and observed burdens shows
that deposition and retention behavior of the surrogate  particles is simi-
lar to actual  diesel  particles.   The observed long-term  retention of 67Ga203
in rats was very similar to the 62-day half-time reported by Chan, et_aj_. ,
(8) following acute exposures to 14C-labeled diesel  particles.  These
observations give confidence to extrapolations made from observations with
                                    216

-------
0.1 urn 67Ga203 aggregate particles.  These data do show that pul-
monary deposition 1s relatively high for diesel particles,  and  retention
times are relatively long.
                              ACKNOWLEDGEMENTS

    Research performed under U.S.  Department of Energy  Contract Number
DE-AC04-76EV01013 and in facilities fully accredited by the American
Association for the Accreditation  of Laboratory Animal  Care.

                                 REFERENCES

1.  Wolff, R.K., G.M. Kanapilly,  P.B. DeNee, and R.O. McClellan.  1981.
         Deposition of 0.1  urn chain aggregate aerosols  in Beagle dogs.
         J. Aerosol Sci. 12:119-129.

2.  Lippmann, M.  1977.   Regional  Deposition of Particles in  the Human
         Respiratory Tract.  Handbook of Physiology, Section A:  Reactions
         to Environmental Agents,  The American Physiological  Society,
         Chaper 14, pp.  213-232.

3.  Chan, T.L., and M. Lippmann.   1980.   Experimental measurements and
         emperical  modelling of the regional  deposition of  inhaled
         particles in humans.  Am.  Ind.  Hyg. Assoc.  J.  41:399-409.

4.  Cuddihy, R.G.,  D.G.  Brownstein, O.G.  Raabe, and  G.M. Kanapilly.  1973.
         Respiratory Tract Deposition of Inhaled Polydisperse Aerosols in
         Beagle Dogs. J.  of Aerosol Sci., 5, 35-43.

5.  Task Group on Lung Dynamics.   1966.   Deposition  and Retention Models
         for Internal Dosimetry of the Human Respiratory Tract.  Health
         Physics, 12, p. 173-207.

6.  Yeh, H.C., and G.M.  Schum.  1980. Models of Human  Lung Airways and
         Their Application  to Inhaled Particle Deposition.   Bulletin of
         Mathematical Biology, 42,  pp. 461-480.

7.  Yu, C.P.  1978.  A Two-Component Theory of Aerosol  Deposition in Lung
         Airways.  Bulletin of Mathematical Biology, 40, p.  693-706.

8.  Chan, T.L., P.S. Lee,  and W.E.  Hering.  1981.  Deposition and clearance
         of Inhaled diesel  exhaust particles in the  respiratory tract of
         Fischer rats.  J.  Appl. Toxicol.  1:77-82.
                                    217

-------
                               PULMONARY DEPOSITION
            .6r

.4
§ '
H
35
2
a .2

N
-
•%
JTteory
(*h and Schum)
t
                            \  yThtofy f/C/fP}
                               \
                                                   Human and Dog
                                                       Data
                                                                \
                                                                  \
                 aos     o.t


                —Volume Median Diameter
                     10
Mass Median Aerodynamic-*]
        Diameter
Figure 1.     Comparison of mean pulmonary deposition (±  S.D.)  of
              0.1  urn Irregularly shaped polydisperse aerosols (I) with
              that of spherical  monodlsperse  aerosols.  The mid range of
              deposition data (///)  taken from human experiments (Lippmann
              1977 (2);  Chan and Lippmann, (3))  and dog experiments
              (Cuddihy et ^1_. 1973,  (4)) is shown.   Also  shown  are
              theoreticIT predictions for depositions in  humans by ICRP
              Task Group on Lung Dynamics (5) and also by Yeh and Schum (6).
                                     218

-------
Table 1.  Comparison of Total and Regional Deposition of
          Beagle Dogs
                                                                  Particles  in
Compartment
Nasopharyngeal
Tracheobronchial
Pulmonary


0.1 un
7%
7%
25%
TOTAL 39%
Particle Size
0.2 \fn
9%
12%
32%
53%
 Table 2.   Lung Burdens of Diesel  Soot in Rats One Day After 18 Weeks  Exposure
           to Diluted Exhaust
Average Aerosol
Concentration3
                                                  Lung Burden
                                    Predicted
                                                                 Observed
 200 ±   70
 990 ±  390
4150 ± 1460
                                        100
                                        500
                                       2100
                                                                 36 ±   8
                                                                224 ±  39
                                                               1926 ± 335
  ± S.D.  of average daily values.
                                      219

-------
                LUNG CLEARANCE OF RADIO-ACTIVELY LABELLED
                      INHALED DIESEL EXHAUST PARTICLES
                       P. S. Lee, T. L. Chan, and W. E. Hering
                           Biomedical Science Department
                       General Motors  Research Laboratories
                                 Warren, MI 48090
The fate of inhaled diesel exhaust particles  was studied in male Fischer 344 rats using
radioactive diesel particles generated from a single cylinder diesel engine and tagged in
the insoluble particulate  core with radioactive ^C.  The particle size, extractability
and ll*C distribution  in  the  diesel exhaust were characterized at various engine load
conditions. At full load, only 1% of the radioactivity of diesel particles was extractable
by dichloromethane and the mass median aerodynamic diameter of the particles was
0.12 ym [J. Appl. Tox., l(2):77-82, 1981].  Radioactive carbon dioxide was removed from
the exhaust by  a diffusion  scrubber prior to exposures  via a  "nose-only"  inhalation
chamber.  Rapid elimination  of the inhaled 1 "*C02 from the blood in the expired air, and
urine  of test animals indicated  that the correction for increased  radioactivity due  to
the inhaled carbon  dioxide was necessary  only for the initial deposition measurement.
The amount of blood and its contribution  of ll*COz  activity was accounted  for in the
excised organs.

Test animals were exposed to diluted  diesel exhaust at two particulate concentrations
with similar  total inhaled dose (7000 pg/m3 for 45 minutes, and 2000 yg/m3  for 140
minutes) and had comparable deposition efficiencies. After the exposure, the animals
were  housed  in  a clean air environment and the clearance of the  radioactively tagged
particles was determined  over an extended period of  time.  Up to the period of 28 days
after  exposure,  no significant difference in particle clearance has been observed.  This
indicates that thus far, the differences in the concentration of inhaled particles did not
cause any significant alteration in the alveolar clearance process after a single, short-
term inhalation exposure, at least within the  studied concentration range.

The retention of inhaled  particles in animals exposed to  radioactive diesel exhaust at
7000 yg/m3 particulate concentration has  been investigated, thus far, to 126 days after
the exposure.  The particle retention data (Figure  1),  analyzed by a  curve stripping
procedure, indicated  three components  with approximate half-times of 1 day,  8 days,
and 80 days, respectively.   The biological meaning  of these  components  can  be
understood in  terms of clearance mechanisms.  The first mechanism deals preferentially
with  particles  deposited  in the tracheobronchial  tree  and  represents their rapid
transport  by  the mucociliary escalator.  The particles are finally cleared through the
gastrointestinal  tract, and their elimination is clearly documented by the presence of
YltC activity in the feces.  The second mechanism  is interpreted as the transport of
material deposited in the most proximal  respiratory bronchioles,  where only  a short
                                        220

-------
distance is required for transferring the particulates to the mucociliary escalator.  The
third  mechanism  removes  the participate matter  from  the alveolar region,  and  the
clearance  mechanism  may involve  endocytosis,  passive and  active absorption, and
dissolution or metabolism.  The extended data base in continuing studies is expected to
provide further information on the presence of additional clearance phase(s) of inhaled
diesel particles and to assess their clinical significance.
         too
                                                        •  7000 w-J,  a*

                                                        A  3000 W/-J, 140-
     z
     o
     1
     5
          so -
                                  so
                                                                             ISO
                                      DAYS POST-eXPOSUKt
Clearance of inhaled diesel exhaust particles in Fischer  344 rats.  The vertical lines
represent standard deviations.
                                       221

-------
          COMPARTMENTAL ANALYSIS OF DIESEL PARTICLE KINETICS
              IN THE RESPIRATORY SYSTEM OF EXPOSED ANIMALS


                                  S. C. Soderholm
                          Biomedical Science Department
                       General Motors Research Laboratories
                                 Warren, MI 48090


One key element in assessing the potential health effects of diesel engine emissions is
determining to what extent inhaled particles  deposit in the respiratory system and how
long  they  remain  before being cleared.   The  deposition  and clearance of  diesel
particulate is also of interest as an example of a seldom studied interaction between
the  lung and  submicron  particles which are  assumed to  be  insoluble.   Several
investigators have  collected data  relevant to the deposition and  clearance of  diesel
particulate in animal models. The  development of a model of the kinetics of insoluble
submicron particles in the respiratory system should provide a basis  for  organizing,
explaining, and comparing these experimental data.

A review of  available information on  the  deposition,  transport, and  clearance of
insoluble submicron particles suggests a model with compartmental divisions illustrated
in Figure 1.  The model consists of a set of differential equations specifying the rate of
transport of particulate  mass among compartments.  The kinetics are assumed to be
first order, since it is the simplest assumption which can be made and since it leads to
exponential clearance curves, the type found experimentally.

The major conceptual difference between this model of particulate kinetics in the lung
and some previous ones [1,2]  is that the model describes  a wide range of experimental
inhalation conditions, and provides  compartments and parameters  that apply even to
exposures using nearly maximum tolerated doses.  Although the ICRP clearance  model
includes a  nasopharyngeal compartment and emphasizes particle dissolution,  these
features were not considered in the present model because they are irrelevant to the
character of insoluble diesel  particulates. Inclusion of the "free particulate" compart-
ment does not significantly  influence the fit to experimental  data, but serves as a
reminder that particulate resides on the lung surface for a short time.

General solutions of the model for two types of experiments are  presented.  In a Type I
("post-exposure") experiment, the animals are exposed for  only a short period, and their
changing particulate lung burdens are measured starting at the end of the exposure.  In
a "long exposure" experiment, the  animals are exposed continuously or  nearly continu-
ously over a long period, and particulate lung burdens are periodically measured.

The  major  contribution  of  the particulate kinetics model  to the analysis of  the
experimental  data is to  point  out  that the  long-term clearance  phase may be
                                       222

-------
interpreted in two ways: it may be clearance of participate out of the deep lung with a
half-time of 62 days as  originally interpreted, or it may be a combination of deep lung
clearance  and sequestering.  The amount  of  participate in  the lungs  of  chronically
exposed animals  suggested  a linear  build-up over  time  and  indicated  that  after
excessive  exposure some portion of  the deposited  particulates may be retained in the
lung for longer times.   Consequently, the new  model emphasizes that the slope of the
buildup  may  be related to  the  clearance half-time from the "macrophage"  to the
"sequestered participate" compartment, to the  overall clearance rate from  the "macro-
phage" compartment, and to the deep lung particulate deposition rate.

The application of the model has so far shown that no single set of model  parameters fit
all  the  data  available  from  all experimental approaches.   The differences may be
related to deficiencies in the model or to actual differences in the respiratory system's
response to particulate in the different exposure situations.

REFERENCES

1.     Task  Group on  Lung Dynamics (Paul  E. Morrow, Chairman), Deposition  and
      retention models for internal dosimetry  of the human respiratory  tract.  Health
      Physics, 12:173, 1966.

2.    R.  G.  Thomas, An  interspecies model for retention of inhaled particles.  Asse-
      ssment of Airborne  Particles, W. Stober et al,  eds., Springfield, IL.  Charles C.
      Thomas, p. 405,1972.
                                        223

-------
                               IHIIIHBItlllllllllllllllllllllHIHIIIIUHIIHtlllltnS
                          •IllllllllllllllimilllllllUIIIIIIIIIIIIIIIIIIIIIIIIIIIII
Figure    Compartments and parameters in  the model.
Compartments:
T » "tracheo-bronchial"
F • "free partieulate" on
    deep lung  surfaces
n • "maerophages" and other
    scavenger  cells

Parameters:
BT    deposition rate into t
RT    SDT « C
SOT   (peeifia deposition
      rate into T
SDT   DIT • VM
DCT   deposition efficiency
      of compartment T
C   » concentration of
      airborne partieulate
HTxy • clearance half-time from
                                             S   « "iMU««tar«d  partieulata"
                                             1   • "lynph node*" draining lung
                                             0   • BOI tract"
                                             LHO • 'total lung*
                                             »F  « deposition rate into F
                                             »F  • SDF • C
                                             SBF • specific deposition
                                                   rate into F
                                             SBF • OEF « C
                                             DEF • deposition efficiency
                                                   of compartment  F
                                             vn  • minute volume

                                            compartment "*" to compartment *y"
                                     224

-------
              RESPONSE OF PULMONARY CELLULAR DEFENSES TO
       THE INHALATION OF HIGH CONCENTRATIONS OF DIESEL EXHAUST


                                 Kenneth A. Strom
                          Biomedical Science Department
                       General Motors Research Laboratories
                                 Warren, MI 48090


Bronchopulmonary lavage was used to explore the  responses of the pulmonary phago-
cytic defense in rats exposed  to diesel exhaust  at  concentratons  of  250,  750, or
1500 yg/m  diesel particles  for 26  to  48 weeks.  Figure 1 shows the  quantities of
alveolar  macrophages,  polymorphonuclear leukocytes and  lymphocytes  obtained  by
lavage of the respiratory airways (millions of cells in 40 mL). There is no difference in
the number or kinds of cells obtained from control and from 250 yg/m   exposed animals
after either 6 or 11 months of exposure.  However, after exposure to 750 or 1500 yg/m  ,
the number of alveolar macrophages  increases by 25-33% and  100-150%, respectively.
When the inhaled participate concentration  exceeds 250 yg/m ,  the quantity of the
lavaged alveolar  macrophages shows dependence on the inhaled concentration of the
diesel particulate after both 6  and  11 months of exposure.  In contrast, polymorpho-
nuclear leukocytes, which are not  observed in lavage fluid from Control rat lungs, are
obtained  in high numbers from rats exposed to 750  or 1500 yg/m  for  11 months or 1500
yg/m for 6 months.  The quantities of lavaged cells are, therefore, dependent primarily
on the concentration of exposure and secondarily on  the length of exposure.

Since equal amounts of diesel particulate are delivered to the lungs during exposure to
1500 yg DP/m  for 6 months or to 750 yg DP/m  for 12 months, comparison of the
quantities and kinds of cells obtained by  lavage points out differences in the response of
the  phagocytic  defense  to  the inhaled concentrations.   When identical amounts of
participates are administered over twice the  length of time, the number of  alveolar
macrophages is increased by  25% above control values, compared to 150% after shorter
exposure  to a higher concentration.  The^ounts of polymorphonuclear leukocytes are 5.1
± 1.5 x 10  compared to 19.1  t 4.4 x 10  at the shorter time of higher level exposure.
This comparison  demonstrates that the pulmonary cellular systems  respond primarily to
the rate  of submicron particles  entering the lungs,  rather than to  the total amount of
particulate delivered to the lung. Lymphocytes which are not obtained from control rat
lungs are lavaged from animals  exposed for 11 months  to both 750 and 1500 yg DP/m  ,
but not to 250 yg DP/m  . The presence of the  nonphagocytic lymphocytes in  the lavage
fluid  may  represent a  slow  immune response  to  the  presence of large numbers of
polymorphonuclear leukocytes and particle-filled alveolar macrophages in the lung.  The
enzyme contents of the acid  phospjbatase and  8-glucuronidase in alveolar macrophages
lavaged from control and 250 yg/m  exposed animals were identical.  At higher inhaled
concentrations, the cellular enzyme content per mL of lavage fluid increased, but due
to the variability of the cell  counts and types with  the continuing  diesel  exposure, the
contribution of each cell type to the total enzyme activity was not determined.
                                       225

-------
Under the  exposure regime used in these experiments (20 hrs/day, 5-1/2 days/week), a
threshold rate for response of, the phagocytic defense occurs between inhaled concen-
trations of 250 and 750 yg/m  diesel particles. The response consists of compensative
immigration into the lungs, at  first, of alveolar macrophages and later on, also, of
polymorphonuclear leukocytes. After excessive exposures (11 months of exposure to 750
or 1500 ug/m  diesel particles), mononuclear leukocytes (lymphocytes) are also lavaged
in high numbers.   The probable cause  of the  observed effects is  the  continuous
phagocytosis of  the diesel particles by the macrophage,  which results  in the cellular
accumulation  of excessive  amounts of  "indigestible" carbonaceous  particle  nuclei.
When the macrophage renewal rate in the lungs is too slow to cope with the influx of
particles, the  macrophages eventually become  overloaded.  Release of humoral factors
from active macrophages may stimulate the recruitment of more alveolar macrophages.
Polymorphonuclear and  mononuclear leukocytes most probably respond to other  not yet
known humoral agents from degenerating particle-laden macrophages.
         LAVAOID CIU POPUL»TJO««
    40


    30


    20


    10


     0
ri   n   fl.
     30


     20


     10-


     0
fl   a
Figure  Lavaged cell populations. The proportion
of each cell type in lavaged cells was multiplied
times  the  total  number  of  cells obtained  from
each rat.   The numbers were  averaged for  each
exposure group (six  rats) and charted as millions
of cells in  the total  lavage volume vs. the diesel
exposure concentration.  The line shows the stan-
dard deviation range of the data.
               2*0  no   itoo
                                      226

-------
      THE EFFECT OF DIESEL EXHAUST ON CELLS OF THE IMMUNE SYSTEM


                                    D. Dziedzic
                           Biomedical Science Department
                        General Motors Research Laboratories
                                  Warren, MI 48090


Inhalation of diesel  engine  exhaust results in the deposition  of submicron carbonaceous
particles in the respiratory airways.  The particles  are phagocytized by the pulmonary
alveolar macrophages, and are cleared from the respiratory tract via the mucociliary
escalator or through lymphatic channels.  Lung  clearance via lymphatics  results in  an
accumulation of particles  in the regional lymph  nodes, and literature data suggest that
the presence of hydrocarbons  or  carbonaceous particles in high doses might  affect
immune functions [1,2].

Two approaches have been used in  the present work to determine whether inhalation of
diesel exhaust could be  immunotoxic.  In one series of experiments, dichloromethane
extract of diesel particles was injected in massive doses (10-50 mg/kg) intraperitoneally
over  a  7 day period into C,- Bl mice in two separate  protocols.   First, the splenic
lymphocytes were isolated and studied for ability to respond to polyclonal stimulation
of B or T cells by E. coli lipopolysaccharide or concanavalin  A, respectively.  Secondly,
dinitrofluorobenzene-induced  contact  hypersensitivity  reaction was  measured as  a
reflection of T cell  function by quantifying changes in ear thickness  after an irritative
challenge.   In both experiments, a small  deterioration in the immuno-defensive  ability
of lymphocytes from extract-treated animals  was observed.   In  mitogen  response
assays, lipopolysaccharide  response  (LPS)  was reduced  by about  20% compared  to
vehicle control groups.  Similarly, a 20-50% reduction was seen in concanavalin A (CON
A) stimulated cultures.  In T cell mediated contact  hypersensitivity reaction, all of the
treated animals showed decreased ear thickness  response.  In none of the  experiments,
however,  was a direct dose-response relationship observed.  Furthermore,  fluctuations
in liver weights  from   experimental animals  indicated  the  possibility  that hepatic
changes induced by the excessive doses of injected hydrocarbons may be involved in the
observed effects.

The approach used in this series of experiments is clearly limited, since 1) by use of high
doses of diesel particle extract, the question of bioavailability of  hydrocarbon from
particles  is  ignored; 2) the  large doses of  extract  may overwhelm  normal defense
mechanisms; 3) the route of exposure allows for system distribution  of  material  which
may not occur when inhalation  occurs; and  4) in the case of mitogen  responsiveness,
splenic  rather than lymph node lymphocytes were studied. Nonetheless, the possibility
that the diesel exhaust particle extract administered  in high  doses  may  potentially
affect the immune system  is at least tentatively raised, and should  be further verified
in inhalation studies.
                                       227

-------
The second approach was used to detect immunotoxicity of diesel particles by studying
lymph nodes,  blood and spleen from guinea pigs exposed to diluted diesel exhaust at a
participate concentration of 1500 ug/m .  In  this experiment, immune system  organs
were studied for shifts  in lymphocyte  subpopulations counts.   Alterations in this
parameter have  been observed  in several forms of human diseases,  including  active
forms of lupus nephritis or chronic glomerulonephritis, inflammatory bowel disease, and
other disorders.   In  addition, exposures to environmental toxins such as lead, poly-
brominated biphenyls, and cigarette smoke may also be  associated  with changes in
subpopulation proportions. The data from the present experiments, however, show that
no shift in subpopulation occurred in the mediastinal  lymph node, the site of primary
diesel particle deposition. In spleen and blood, small fluctuations  of no more than t 5%
were observed, which is well within the limits of variability described in the literature,
and no other significant biological effects were identified.

In summary,  small functional differences were seen  in lymphocyte responsiveness of
Ce-Bl mice after treatment with diesel particle extracts.  However, limitations  of the
experimental protocol preclude direct extrapolation to possible findings during inhala-
tion  exposure to diluted  diesel  exhaust.   In a test system  where the more  realistic
inhalation mode  of  exposure  was employed, no  major effects were  seen, and more
studies are needed to focus on lung immune system reactivity before the  immunotoxic
potential  of  diesel  participates deposited  in  the mediastinal  lymph nodes can be
definitely assessed.
 REFERENCES
 1.    D. E. Bice, et al, Drug Chem. Toxic., 2,1979.
 2.    A. Zarkower, Arch. Environ. Health, 26,1972.
          INCREASED EAR THICKNESS VS. TIME
                            (MEAN ± S.E.)
                 30
                                                  VEHICLE CONTROL

                                                   V
                                                   10MG/KGDOSE


                                                       MG/KG DOSE
                    0       24       48       72
                  HOURS AFTER DNFB CHALLENGE
                                       228

-------
             THE PARTICIPATION OF THE PULMONARY TYPE E CELL
           RESPONSE TO INHALATION OF DIESEL EXHAUST EMISSION:
                                 LATE SEQUELAE


                             H. J. White and B. D. Garg
                          Biomedical Science Department
                       General Motors Research Laboratories
                                 Warren, MI 48090


Among the earliest cellular responses of the lung to inhalation of diluted diesel exhaust
is  the focal  proliferation  of the alveolar  Type  n  cell, the cell  responsible for the
synthesis  of  surfactant.  This  reaction takes place relatively  within  the same time
frame as that for phagocytosis by the  alveolar macrophage, and  can  be  seen in rats as
early as  twenty-four  hours  post-exposure  to  a  diesel particle  concentration of
6000  yg/m   [1]. The proliferation is  focally within alveoli, sometimes showing several
cells in a line. The variation  in the staining of their nuclei supports the idea of a fairly
rapid proliferation, although mitoses have not been seen.  There is no evidence  in our
hands that this proliferative  response is of a reparative nature secondary to damage of
the Type I cell.  The Type H  cells release considerably increased amounts of surfactant,
which accounts for a  morphologic change in the phagocytic alveolar macrophage which
now takes  on a more  foamy  appearance.  TJie transition between the early  macrophage
(one-day exposure to diesel  at  6000 vig/m   to that of 6 weeks' exposure) reveals the
gradual accumulation of surfactant material. Some of the phagosomes  can be seen to
contain recognizable  myelin  figures  of Type  n  provenance  changing  to  a  more
filamentous form, suggesting an unraveling of the more compact tightly-wound phos-
pholipid. The mechanism of  accumulation of cholesterol is certainly not clear,  although
plate-like  crystalline  structures  can be easily identified  within  the  phagosomes,
suggesting cholesterol ester formations.  The excess  cholesterol  could well  have its
origin from the phagocytosed surfactant material which then at some later  stage is de-
esterified to  form the familiar elongated acicular structures of cholesterol. This  foamy
cell is apparently more sluggish, and tends to accumulate near the terminal bronchioles
where further Type n  cell activity is elicited.
                                                  3                3
With  prolonged exposures to 9  weeks at 6000 yg/m   and 1500  pg/m   for two  years,
cholesterol also begins to accumulate  within the phagosome.  Later, the crystals grow
and tend to become extracellular. This can be demonstrated by both light and electron
microscopy.  Build-up of the  cholesterol deposits in these late stages has been  found to
be associated with increased  collagenosis  of the septal  wall in which mast cells are also
present.  The reaction of mast cells to the release of  extracellular cholesterol is  also
obscure.  One could speculate that the heparin of the mast cell is involved in activation
of lipoprotein lipase in an attempt to clear the excess of lipid release.  In some  way, the
amine component of the mast cell is also released at this time  to provoke the  laying
down of collagen.  In addition, there is an apparent intimate association between laying-
                                       229

-------
down of  extracellular  cholesterol, lipids,  septal  mast  cells  and collagenosis;  the
collagenosis is, however, focal and the  integrity of the septal architecture is preserved.
Occasionally, focal cholesterol granulomas have been observed. These are quite similar
in appearance to those seen  after prolonged  inhalation exposure  to Sb«O.  [2]  and
marihuana smoke [3].  The phenomenon of  "benign" focal alveolar collagenosis seems
to be a consequence of high particulate burden that stimulates an increased production
of phospholipids, and not  directly of the effect of the  diesel particle per se.  The
observation  that diesel participates, sequestered in the thoracic lymph nodes for up to
two years, do not provoke a fibrotic reaction  and supports this contention.
REFERENCES

1.    White, H.J.  and Garg, B.D. (1981),  Early pulmonary response of the rat lung to
     inhalation of high concentrations of diesel particles. J. Appl. Tox. 1:104-110.

2.    Gross,  P.,  Brown, J.H.,  and  Hatch, T.F.  (1952),  Experimental  endogenous
     pneumonia. Am. J.  Path., 28:211-221.

3.    Fleischman, R.W., Baker, J.R., and Rosenkrantz, H.  (1979), Pulmonary pathologic
     changes  in  rats exposed to marihuana smoke for  one  year.   Tox.  and Appl.
     Pharmacol., 47:557-566.
                                       230

-------
SECTION 4



PULMONARY TOXICOLOGY AND BIOCHEMISTRY
                           231

-------
RESPONSE  OF  THE  PULMONARY  DEFENSE  SYSTEM  TO  DIESEL  PARTICULATE
EXPOSURE

JAROSLAV  J.  VOSTAL, HAROLD J. WHITE, KENNETH A. STROM, JUNE-SANG SIAK,
KE-CHANG CHEN, AND DANIEL DZIEDZIC
Biomedical Science Department, General Motors Research Laboratories, Warren, Michigan
48090, U.S.A.
INTRODUCTION
   In 1977, the  U.S.  Environmental  Protection Agency reported  that organic  solvent
extracts of diesel particles were mutagenic  in  bacterial assays,  and indicated that the
increased penetration of the diesel engine into the fleet of light-duty vehicles could result
in an increased frequency of lung cancer in the U.S.  population during the years 1985-2000.
Several predictions of  the expected magnitude  of  lung cancer  excess have been  offered
during the last years (Table 1), but the wide differences in the quantitative estimates of the
expected  effects indicate  clearly the  high level of  uncertainty of the  theoretical
predictions. In sharp contrast, the negative epidemiological studies of the British Medical
Research  Council conducted on London bus garage workers between 1955 and 1974, as well
as the recently published results of long-term inhalation animal studies  seemed to assure
that the growth of the  passenger car population  equipped with diesel engines and of diesel
particulate concentrations  in  the urban air would not, according to the current state of
knowledge, threaten public health.
   The  apparent discrepancy  between  the theoretical  prediction based on the suggestive
evidence of the simplified bench  tests and the real life situation is difficult  to explain.
First, it is important to note that positive mutagenic  tests were observed only after all
adsorbed hydrocarbons had been stripped by powerful organic solvents  and assayed in the
test  in the form of extracts.  It has been questioned whether it is scientifically appropriate
to use  an organic  solvent to extract  hydrocarbons adsorbed on  the  core of the diesel
particles when the in vivo biological activity of the  entire particle in the living  organism is
to be assessed.  The capacity of such solvents to solubilize organic  matter  is many times
stronger than  anything found in the human body, and therefore cannot simulate mechanisms
by which biologically active compounds are released for interaction with sensitive  cells of
                     o
the respiratory system.
                                232

-------
       TABLE!.  ESTIMATED ANNUAL EXCESS IN U.S. LUNG CANCER FATALITIES
       DUE TO LIGHT-DUTY DIESEL VEHICLES (95% UPPER CONFIDENCE LIMITS)
                                                 Predicted Annual Increase
       Thorslund , EPA 1981

       Albert,4 R., EPA 1979

       Harris5, J., NAS 1980

       Cuddihy,6 R.G., LITRI1981
  1151

  455

< 30 - 45

  < 30
   Using  the  same  laboratory method, Siak  et  al   and Brooks et al  have  clearly
demonstrated that when extraction fluids were used which are compatible with the internal

environment  of  the human body,  mutagenic  activity was  significantly reduced  and
represented only a small fraction of the effects reported for organic extracts (Figure 1).
                      150
                      100
           Relative
          Mutagenic
           Activity
                        50
                                                     TA 98 — S9
                                             <**
Figure 1 .   Comparison of the mutagenic activities  of  diesel participate  extract DCM
(dichloromethane), DMSO (dimethyl sutfoxide), FCS (fetal calf serum), BSA (0.5% bovine
serum albumin), SLS (simulated lung surfactant), and SLN (saline). From Siak et al. (1979)7

-------
   Thus, while all the findings  obtained  by using organic solvent extracts from diesel
participates  may  be of significant  scientific  value,  they are  not valid predictors of
potential health effects of diesel particulates inhaled by the human respiratory system and,
unless the availability of the chemical compounds adsorbed on  the  surface of diesel
particulates  for the biological fluids of the  human body is considered in the  assessment
process, estimates of lung cancer excess due to diesel emissions will remain arbitrary and
           2
unrealistic.
   Second, the highest estimates of increased incidence of lung cancer [Thorslund 1981,
Albert 1979]  are,  unfortunately,  not  based on laboratory or epidemiological evidence of
diesel emission effects at all; they have been  derived  from unsupported and speculative
assumptions "that the carcinogenicity of diesel engine exhaust in units of particle-bound
organics, extracted from  the exhaust and  oven emissions have the  same potency per unit
      g
mass,"  or that  a ratio  of the relative  carcinogenic  potencies can be obtained from
comparative  laboratory tests.    The  apparent fallacy of the assumption lies mainly in the
fact that  it presumes the activity  of one  single representative of  the polycyclic organic
matter, namely benzofal pyrene, is a common denominator responsible for  the carcinogenic
aggressivity of both coke oven as well as diesel  emissions.  However, a simple comparison
of both  pollutants clearly indicates  that benzofa] pyrene concentrations in  the diesel
particulates are lower by  at least two orders  of  magnitude  (Table 2) and  that  a large
fraction of the mutagenic activity  of diesel particulates is attributable to the presence of
powerful nitroaromatic mutagens, rather than to benzo[a] pyrene alone. '   In addition,
the epidemiological data for coke oven  emissions are crude and a number of assumptions or
                                                                            13
corrections must be made before  they can be used in the risk assessment process.
       TABLE 2.   AVERAGE CONCENTRATIONS OF BENZOta] PYRENE IN SOLUBLE
       COKE OVEN EMISSIONS AND DIESEL PARTICULATE EXTRACT
       Coke Oven Emissions                       5,135 ± 2.019 yg B[a] P/g CTVP
         (Jackson et al., 1974)*
       Diesel Particulate Extracts                   98.71 1 143 yg B[al P/g Ext
         (from 16 FTP tests on 10 Oldsmobile
         engines - Williams, 1981)**
        •Jackson, et al., Amn
        ** Williams, R. (1981), personal communication

   Third, results reported from the long-term animal experiments  indicate clearly that
despite  massive exposures and excessive accumulation of particles in the lung, no increased
                                234

-------
risk for tumor formation was  found  in the reported studies and no significant changes in
pulmonary function and structure have been detected to date.
    Pulmonary defense  mechanisms were found effective in removing inhaled particulates
and preventing  their contact  with  the  sensitive  cells.   This may  provide  the final
explanation of the apparent discrepancy between the positivity of laboratory bench tests
and negative effects  in chronic exposures of large human populations  or animal inhalation
studies.    Experimental work   on  the mechanisms  responsible for  particle deposition,
retention  and clearance from the respiratory system  has been reported separately, but  the
role of the cellular defense  mechanisms, primarily of the pulmonary alveolar macrophage,
represents an equally important if not more potent process which may  defend the organism
against biological  activity of the deposited particles.  Research programs in our laboratory
concentrated, therefore,  on the  analysis  of  the pulmonary cellular  mechanisms as a
potential explanation of the existing discrepancy.

RESULTS
    Pulmonary alveolar macrophages  belong to  the line of  mononuclear phagocytes which
are  widely distributed in  the  body,  and constitute an important part  of the  defense
mechanisms of the respiratory  system by clearing the tissues and fluids of particles such as
bacteria or microscopic foreign bodies which penetrate into the organism.   After ingestion
of  particulate  material,  living matter of microorganisms  is degraded, but insoluble
particles  may be  retained  inside the phagocytic cell and, along with the macrophage,
cleared from the organism via the mucociliary escalator.
    LaBelle and  Brieger  demonstrated  that a  highly positive correlation  exists  between
the amount of the inhaled dust cleared from the lung and the number of phagocytic cells
counted in a  unit of lung tissue or bronchopulmonary lavage, thus indicating that the early
clearance of dust  is mediated by phagocytes. The inhaled particles evoke various responses
from  the  alveolar  macrophages.   Inhaled cigarette smoke  increases  the number  of
                                1C 1 *7                  10                         1Q
lavageable alveolar macrophages,  '   the cell diameter,   cellular protein content,  and
                                20
the activities of several enzymes.
    Strom21 described  the  results  of chronic  experimental exposure  of  rats  to diesel
particles at concentrations which range  from  a nominal value  of 250  to  6000 ug diesel
particles per cubic meter of air.  Male Fischer 344 rats, (Rattus norvegicus) six weeks of
age, were exposed to  diluted diesel exhaust at concentrations of 0, 250,  750, 1500 and 6000
Ug/m  diesel particulate  per  cubic  meter in  the diesel exposure  program at  the GMR
Biomedical Science department.   After 6 months  and 1  year of  exposure,  groups of 6
animals each were removed and anesthetized with  sodium pentobarbital (50 mg/kg body
weight) administered intraperitoneally.   The  lungs  were lavaged  in situ  using Hank's
balanced salt solution (HBSS) which  was prepared  without  calcium  and magnesium.  The
                                235

-------
trachea was exposed and cannulated using a 14 gauge Teflon catheter, held in place with
three annular ligatures.  Before lavaging the lungs, the abdominal aorta was transected to
exsanguinate the animal, and the diaphragm was  punctured  to  deflate  the  lungs.  The
accessible  airways  were lavaged seven times  with 6 mL  aliquots of HBSS  at 37° with
nominal recovery of 40 mL of solution.
   In the case' of  unexposed animals, all the  lavaged cells are  usually  alveolar  macro-
phages, with a yield of 5 to 7 million cells per animal.  However, the cells from exposed
animals varied  in  cell  type  and cell  counts.   Figure 2 shows the  populations of  cells
obtained as a function of diesel particle concentration during long-term exposures.
LAVAGED CELL POPULATIONS
40
30
„ 20
Z 10
50-


6 months


n

~. JQ
•
I 30
=
5
2 20-
10-
ft-

i
r1
n n 1
i
i
i
0 macrophagn
• nautrophils
• lymphocytes
12 months I




n

pi,
j.
-L
n rk
1
1

                              0   250   750  1600
                         Diesel Exposure Concentration
Figure 2.  Total cell counts obtained from individual lavaged animals in various exposure
groups. Mean ± standard deviation. From Strom, K.A. (1981)21
   It is important to point out that at the lowest exposure concentration (250 yg/m ), as
well  as  in the  control,  the cells from  the  exposed  animals were entirely  alveolar
macrophages, with no  difference  in the cell counts in the lavage fluid.  At the  higher
                                      o
concentrations of 750  and  1500  yg/m  , the counts of  alveolar macrophages increased
proportionally with  the exposure  concentration; the results were  identical for alveolar
macrophages at both 6 and 11 months of exposure. Obviously, in the long-term exposure the
particulate enters the lungs at a rate  proportional to the inhaled concentrations, and the
normal  turnover  of  alveolar  macrophages  in the lung can  handle  the entering mass of
                                 236

-------
particulates below concentrations not exceeding 250 ug/m  .  At higher mass influx rates,
the resident macrophages escalate the rate of phagocytosis.  It  is expected that humoral
factors are  produced and released during  phagocytosis, increasing further immigration of
mononuclear phagocytes or recruitment of additional phagocytic cells from local resources.
The cellular pulmonary defense is stimulated by higher inhaled concentrations to a larger
                                                      21
extent than when low  concentrations are inhaled.  Strom   observed that the response in
the form  of significantly increased macrophage counts is  much larger after  exposure to
          3                                              3
1500 yg/m  for 6 months than after inhalation of 750 ug/m   for 1 year, although the total
mass  of deposited  particulate burden was identical.  He postulated that the number of
lavaged macrophages is proportional  to the mass influx of particulate, rather  than  to the
actual diesel particulate burden in the lung.  This suggests that there may be a threshold
for the rate of  mass  influx of diesel particulate into the  lung  above which  there  is
increased  recruitment of alveolar macrophages.  Under the exposure regimen employed in
the experiments (20 hrs/day,  5.5 days/week), the threshold rate was between the inhaled
                                 o
concentrations of 250 and 750 ug/m  diesel particulate.
          250
          KX>
                                  469
                               Lung  Content, mg Part./ g
                                                                10
Figure 3. Dose-response curve of the macrophage counts in the bronchopulmonary lavage
fluid versus lung content of the diesel particles in long-term exposed animals.  Data from
Strom, K.A. (1981)21
                               Z-i/

-------
    Apart from the question of whether the primary stimulus for recruitment of alveolar
 macrophages comes from the  specific role of the exaggerated particulate influx or the
 presence of excessive particulate burden in the lung, the threshold character of the defense
 reaction is  obvious, even when the macrophage counts are plotted against  the amounts of
 participates accumulated in the lung with exposure (Figure 3).  The type of curve indicates
 a distinct dose-effect relationship,  and confirms clearly  that prolonged exposure to high
 concentrations provokes defense mechanisms which do not  exist  at lower levels of
 exposure.
            25
            20
        x
        I  15
        CL-
             5-
                                  468
                                Lung  Content, mg Port./g
                                                                 10
                                                                           12-
Figure 4.  Dose-response curve of the polymorphonuclear leukocyte count in the broncho-
pulmonary lavage versus lung content of diesel particles in long-term  exposed animals.
Data from Strom (1981)21

   In addition, a similar character of response occurred in the counts of polymorphonuclear'
leukocytes in the lavage  fluid (Figure 4).   Both responses occur relatively late  in the
exposure process, and must be differentiated from the similar increases in the number of
macrophages and polynuclear neutrophils described after intratracheal instillations of large
                     15 22
loads of participates.  '    However, the  character of the  primary response  to intra-
tracheal administration of participates is transient, and  the cell counts return  to normal
                         22
levels after several days.    In contrast,  the response to long-term high-level exposure
seems to be permanent, and is proportional to the inhaled concentrations and/or duration of
the exposure.
                                238

-------
         21
   Strom   also  offered the explanation that the primary  response  of phagocytic cells
including polymorphonuclear leukocytes in the acute insult is provoked by the inadequacy
of the macrophages to remove  the particles from  the  alveolar surface, and indicated that
this would  be an unlikely explanation of the polymorphonuclear response  in the chronically
exposed  rats for  three reasons.  In the first case, the pulmonary defense probably has
equivalent  capacity  to produce  more  macrophages as  well as  polymorphonuclears  in
response to inhaled participate. Nevertheless, after exposure to 750 \jg/m  for 6 months
and 1 year, the macrophage counts do not increase, but the  number of leukocytes does.
Second,  morphologically the polymorphonuclears appear in  the lung primarily among the
diesel  particle-laden clusters of macrophages, rather  than freely roaming the alveoli  as
observed after intratracheal administration of particulates.   In contrast, the leukocytes
seem  to  represent a specific response to the aggregated macrophages in the late phase of
the exposure, rather than  to the presence of particulates themselves.  Third, in long-term
exposures,  the  leukocytes do not rapidly decline upon cessation of the exposure as they  do
                                          o
after   the  acute insult.   After 750 ug/m   exposure  for 1  year, significant counts  of
polymorphonuclears can be obtained by lavage even 16 weeks after the exposure  has ended.
   It is important to note  that a similar biphasic response of the pulmonary defense system
                                                                    23
was also observed in the reaction of the alveolar cells. White and Garg   investigated the
                                                                                 2
lungs  of rats exposed to diluted diesel exhaust at concentration levels of  6000 ug/m  for
                                                                              14
periods from one day  to nine weeks  using exposure methods previously described.   They
observed a highly significant scattered increase of Type U cells  without any accompanying
necrosis  of the endothelial (Type I) cells after only 24 hours of exposure (Figure 5).  Later
on in  the exposure (4 weeks), the authors identified that many of the alveolar macrophages
with particles became foamy and aggregated in  alveoli near  terminal bronchioles, as well
as near other relatively immobile structures  such  as vessels  and the pleura.  By  9 weeks,
there  was  continued increase  of diffusely-placed macrophages and accentuation of the
aggregated formations of  fused phagocytes.  The septa of the alveoli  that made up these
complexes  were slightly thickened and showed a positive stain for reticulin which some
authors consider to be a precursor  form  of  collagen.  Specific stains for collagen were
negative, however.  At the same time, the Type n cells lining  the alveoli of the complexes
were  again  more  numerous  in apparent proportionality to the  amount  of  aggregated
macrophages (Figure 6).
   Obviously, like  the immediate response of phagocytic cells, the alveolar Type n cells
also react transiently to the initial insult of the participate influx into  the  alveolus.  Later
on, another  proliferation  of Type  n cells  occurs,  however, in response to  the formed
aggregates of the macrophages. This effect cannot be found in the unexposed subjects, and
apparently do not exist at the  lower levels of exposure.  In  the literature, similar "dust
macula"  or focal occurrence of alveoli with thickened septa has been  described in coal

-------
Figure 5.  Lung -  6000 ug/m  , 24  hour exposure.   Note several Type n cells (arrows).
Macrophages contain diesel  particles (arrow head).  Periodic-acid Schiff.  250x From
White, HJ. and Garg, B.D. (1981)23
                            n
Figure 6.  Lung - 6000 ug/m , 6 weeks, alveolus containing mass of fused macrophages,
single  macrophages(*).   Note marked  Type  n  proliferation  (arrow  head)  apparently
secondary  to fused macrophage  mass.  Hematoxylin-eosin. 160x  From White, H.J. and
Garg, B.D. (1981)23
                              240

-------
workers' pneumoconiosis.   '    The term  "macrophagic alveolitis" was used when referring
to the lesion in the lung of the guinea pig  exposed to pure coal dust.
   The  exposure-dependent appearance of these  complexes was confirmed in  the  second
     27
study   when  White and Garg examined the histologic changes in the lungs of animals that
had been exposed for greater lengths of time up to two years to  concentrations  below 6000
     3
yg/m ,  but still at levels more than  one hundred  times  higher  than  expected average
roadside concentrations in the year 2000.  In the chronic experiment, one hundred and fifty
adult Fischer  344  male  rats (Rattus  norvegicus)  were divided into  four  equal  groups.
Animals of group one were exposed to filtered room air, and rats  of groups 2, 3,  and  4 were
                                   3
exposed to  250, 750, and 1500 ug/m   of  diluted diesel exhaust,  according to the methods
                          14
described by Schreck et al.   Three rats  from each group were  sacrificed immediately at
the end of 5,  10, 15, 25, 35, and 45 weeks, one year, 18 months, and two  years of exposure.
                                          3
In addition, three  rats exposed to 750 ug/m   for one year were  sacrificed after 35 weeks
recovery.
   Again, a general increase in diesel particulate deposition was observed in the lungs by
light  microscopy that was roughly proportional to the product  of  dose  and  time.   Thus,
                             o
animals exposed to  250 yg/m   for  5 weeks showed the least accumulations, while those
                             3
animals exposed to  1500  yg/m  for  two  years showed  the  maximal deposition.   Shorter
periods  of time and lower levels of  exposure also produced a smaller number of distinct
foci  of  deposition formed by  the  local aggregation  of  macrophages with ingested diesel
particles.   In  general, the foci tended  to locate  near terminal bronchioles and other
relatively fixed structures, including vessels and the pleura.
   The  occurrence of aggregates of  macrophages can be described as another form of the
cellular defense system:  the macrophages which for some reason were unable to leave the
peripheral airways became immobilized,  but still contained and stored the  phagocytized
diesel particles, thus preventing their more intimate contact with the sensitive cells of the
respiratory  system.  The  reasons for the  augmentation of the Type n cells in the directly
adjacent  alveolar walls cannot be  explained at  this  time.    Perhaps the presence  of
macrophages with particles stimulates their proliferation in order to increase production of
lung surfactant.  It is completely  unknown  if  the presence  of  excessive amounts  of
surfactant in  the alveolus leads to  the  aggregation and  clustering of  the macrophages.
However, preliminary data on the increased concentrations  of surfactant phospholipids in
the lavage fluid which coincide with the presence  of aggregated  macrophages may indicate
                                              28                                    29
a specific role of the surfactant  in this reaction.    In this regard,  Papahadjopoulos et al
have reported  in vitro evidence of enhanced fusion of cells in the  presence of phospholipids.
   Certainly,  the  containment of diesel  particles inside macrophages can explain why -
except  for  the functional stimulation of the Type  H cell    we do not see any specific
reaction of  the alveolar tissue early in the exposure.  It was only after  excessive amounts
                               241

-------
of particles had been accumulated in the lung - or after the macrophage clusters containing
the particles had become stationary in subpleural or peribronchial alveoli for an extended
                                                                            27
period of time, that the first alveolar wall reaction could be seen. White et al  described
                                     o
that in animals exposed to 1500 ug/m   for one year, late changes in  the appearance and
structure of the clustered macrophages  can occur.  Their cell borders become less distinct,
diesel  participates become less contrasting, and  give the cytoplasm a more homogeneous
foamy appearance which is no longer PAS positive.  At approximately the same time, the
first changes  in the adjacent pulmonary tissue were also observed.  Not only was there
laying down of reticulin, but focal septal collagenization was  reported, as seen by Masson
stain  and birefringence under polarized light.  The picture,  while still focal, was more
obvious in animals exposed for  two years than in those observed after  one year (Figure 7).
Only by exception was there any loss of parenchymatous integrity; i.e., although there  was
focal septal thickening, the alveolar structures were not obliterated.
   Two other histologically distinct findings accompanied the septal fibrotic  changes  which
were  prominent only where intra-alveolar diesel material had accumulated in the macro-
phage clusters. Lipid-processed, Oil Red O-stained tissue was  positive  for neutral lipids in
these areas,  which also showed fine  granular  and needle-like structures on polarization.
Where such polarizable material was noted, the fibrotic septa  were found to contain mast
cells.  These cells are readily recognized by the presence of metachromatic purple granules
in the cytoplasm on staining with toluidine blue buffered to pH4 (Figure 8).
   As mentioned above, occasionally small foci of more complete fibrosis with a tendency
                                                                                3
to obliterate the alveolar pattern were seen in a  few animals exposed to 1500 ug/m   for 18
months or longer (Figure 9).  Similar changes  were seen in  a singular group of animals
                    O
exposed to 750 yg/m  for one year and allowed to recover for 32 weeks before sacrifice.
The lesions, in addition to their pronounced delimited fibrotic  aspect, also contained large
acicular  clefts resembling  those  seen in atheromatous  lesions,  birefringent  in polarized
light, and presumably representing  cholesterol.  Along with the Masson-positive collagen
fiber and cholesterol accumulations, mast cells and histiocytes were prominent in the areas
of these lesions.  When they occurred in the 750 ug/m  exposed group,  the rest of the lung
showed little, if  any,  fibrosis.   Later  on, the cholesterol  accumulation  became  quite
localized, usually in the subpleural areas where, in the extreme situation,  discrete focal
lesions were formed that appear identical with the so-called "cholesterol granuloma."
   Probably, the proliferation of Type n cells with increased production of surfactant and
exaggerated pouring  of phospholipids into the alveolus in  response to high exposures of
respirable particulates  is directly responsible  for the local accumulation of lipids.  Acti-
vated alveolar macrophages ingest not only the inhaled particulates, but also the released
phospholipids.  The authors further speculated that phagocytosis of the lipids  makes the
macrophages less  motile, causing them to aggregate.  At this  point, local Type n cells
                                242

-------

Figure 7.  Diesel participate exposure to 1500  yg/m  for 18 months.  Focal deposition of
needle-like crystalline material (arrow) resembling cholesterol in areas of diesel accumu-
lations.   Note  fibrotic  reaction around  cholesterol crystals  with  focal obliteration  of
alveolar pattern. Hematoxylin-eosin. 40x  From White, H.J. and Garg, B.D. (1981)27
Figure 8.  Diesel particulate exposure to 1500 ug/m  for  2 years. Portion of alveolus with
large extracellular acicular formation (short arrow), surrounded by foamy macrophages.
Note mast cells (long arrow),  and septal collagen (broken arrow).   Electron micrograph.
2400x  From White, H.J. and Garg, B.D. (1981)27
                              243

-------
react by cell proliferation resulting in further release of alveolar surfactant, and switching
the metabolism of  the macrophage  toward  the  precipitation  of lipids in the form of
cholesteryl palmitate.  The final release of free cholesterol, first intraceUularly, and later
extracellularly in  the alveolus promotes the penetration of  mast cells into the septa and
stimulates the local production of collagen within the septum, but not within the alveolus
                                                   27
(Figure 10).  It is this  condition  which White et  al   would  term a "benign fibrosis" as
opposed  to  those  conditions which injure the alveolar  lining cells  and provoke a  fibrotic
reaction within the alveolus ("fibrosing alveolitis" or  "malignant" fibrosis).

   The capacity of the macrophage response  in the phagocytosis of the invading particu-
                                                 30
lates  is otherwise overwhelming.  Rudd and  Strom   developed a  method  for the direct
measurement of the amount of diesel particulate in  tissue, and reported that in guinea pip
(Cavia poreeZZus) exposed to diesel particulate for 22 weeks, the amount  of particulate in
the lung rose proportionally to the exposure concentration.  In order to explain the results,
they  examined  the lungs  of  exposed  animals in  order  to discover the  reservoirs for
deposited participates,  and  again identified  that   the  main  sequestering site  of the
particulates  is the  alveolar macrophage.  The authors reported  that through  an  as yet
unknown mechanism, the dust-filled macrophages aggregate in or near terminal bronchioles
and the pleural surface, and remain there for as long as two and one-half years after the
exposure.  They further noted that  although this  mechanism  may be  advantageous in
clearing particle-laden  cells from most of the respiratory surface, it does not promote the
actual clearance of particulate from the lungs.
                                                31
   In  a complementary study,  Siak and Strom    addressed  the question of how the
phagocytes handle the biological activity of the ingested particulates. Young male Fischer
344 rats (Rattas norvegocus) were exposed to diluted diesel exhaust at a concentration of 6
mg/m  for three days (20 hrs/day).  Alveolar macrophages  were obtained by bronchopul-
monary lavage immediately after exposure and at 1,  4, and 7  days thereafter. Macrophages
from  forty  animals were pooled  for each data point,  sized and  counted.   The alveolar
macrophages were concentrated by filtration  on pre-washed fiberglass filters and dried at
room  temperature to constant  weight.  The filters  were extracted with dichloromethane in
a Soxhlet apparatus for 4 hours (20-25 solvent cycles).  The Salmon&Ba typhlmvnum strain
TA98 was used for mutagenicity  assay.  For  thin  layer chromatography,  Whatman LK6
plates were used and the developing solvent was toluenethexane (5:1).
   Figure llcompares the mutagenic response of macrophage extracts from exposed rats as
detected in the Ames  microbial  assay  with  the extracts from diesel particles collected
from  the  inhaled  air.   Although the macrophage  extract  obtained from  exposed  animals
immediately after exposure gave  a positive  response, the data  show that intracellular
material extracted from macrophages interferes with the mutagenic response of diesel
                                244

-------
                                                     ?&
                                                                 '
Figure  9.   Diesel particulate  exposure to 1500  pg/m   for  two years.  The  increased
collagen material in the alveolar septa polarizes.  Note that the process is focal.  Masson
stain with polarized light.  16x From White, H.J. and Garg, B.D. (1981)27
Figure 10.  Diesel particulate exposure to 750 ug/m  for one year - 32 weeks recovery. By
polarized light, alveolar collagen and intra-alveolar lipids are accumulating  in subpleural
area.   Note  accompanying septal mast  cell  reaction  (arrow).   Toluidine  Blue (pH4),
polarized light.  40x From White, H.J. and Garg, B.D. (1981)27
                              245

-------
               500
             „ 400
             I
             k
             CD
             5 300
                200
c
+
£
 »
                100
Diesel particle extract
Diesel particle extract
+ 800 M"i macrophage extract
Day 0
                                       -«-  Day?
                   0    0.2   0.4  0.6   0.8   1.0
                  Equivalent Diesel Particutate Mass (mg)
                                per Plate

Figure 11.  Salmonella typbimwdum(TA98) mutagenio  activity response of diesel particle
extracts from alveolar macrophages on day 0 and day 7 after exposure.  From Siak, J-S.
and Strom, K.A. (1981)31

particle  extract  in the  Ames assay (Figure 11).  However, in spite  of the  interference,
mutagenic  activity of macrophage extracts obtained  from the exposed animals was only
detected at day  0 and day 1 after exposure; after that, no mutagenic activity could be
detected (Figure 12).  The interaction  between alveolar macrophages and the extractable
mutagenic  components of inhaled diesel particles was further corroborated by chromato-
graphic analysis.  In a parallel way, the thin layer chromatography UV  fluorescence banding
patterns of the extracts,  which were prepared  from macrophages lavaged from exposed
rats immediately and one day after the exposure, were similar to extracts of particles from
inhaled air, whereas the  extracts of macrophages on days 2, 3, 4, and 7 after exposure have
lost the banding patterns through as yet unknown  mechanisms.
                               246

-------
    Alveolar  macrophages are  capable  of  metabolizing  polycyclic aromatic  hydrocar-
bons   '   and previous work in our laboratory demonstrated that mammalian liver enzymes
activate the bacterial mutagenic activity of 1-nitropyrene and of diesel particle extract -
under  specific laboratory conditions.34  Therefore,  an enzymatic transformation of the
extractable organic compounds of diesel particles by  macrophages may be one of the
                 MUTAGENIC ACTIVITY IN MACROPHAGE EXTRACTS
                        OF DIESEL EXHAUST EXPOSED RATS
          3
          »  120-
          u
         "5
          Q.
             80-
          §  60-)
          £  40-
          CO
          £  20-1
                                 1        2        3
                                 Days post exposure
                                                         4&7
Figure 12.  Changes in  the  mutagenic activity of diesel particle extracts obtained from
lavaged pulmonary alveolar  macrophages at different times after the exposure.  From Siak
and Strom (1981)3'

possible  mechanisms  involved.  Another possible  mechanism  is the solubilization  of the
extractable organics from diesel particles by phospholipids from the  lung surfactant  and by
other cellular components of the macrophage.  The soluble complexes  may diffuse into
other tissues, and/or bind to other cellular constituents which render them  unextractable
by the method employed. Further in vivo and in vitro experiments are required to provide
a better  understanding of the mechanisms  involved, but  the results thus far demonstrate
that the  insoluble particulates stored for  a  prolonged period  of  time in alveolar macro-
phages represent virtually an innocuous material  which may  have long  lost most  of  its
biological activity.
                               247

-------
     The lack of biological activity of diesel participates deposited in the respiratory tract
 was further documented by the work of several laboratories. Chen et al35 investigated the
 effects of long-term inhalation  of diluted diesel exhaust on aryl hydrocarbon hydroxylase
 activity and cytochrome P450 content in lung and liver microsomes in male Fischer-344
 rats (Rattus norvegicus) and compared them with the results obtained after intraperitoneal
 and intratracheal administration  of organic solvent extracts of  hydrocarbon from the diesel
 participates.  Surprisingly, a decrease instead  of an enzyme induction was observed in lung
 microsomal aromatic hydroxylase activity of animals after the  full 9 months of exposure to
 diesel  exhaust at the particulate concentration of 1500 ug/m3.  The observations were
 confirmed by other investigators.  '     In contrast, 1.4- to 9-fold increases in aromatic
 hydroxylase activity were  observed in liver  and lung microsomes of rats pretreated by
 intraperitoneal doses of  particulate extract, which  were 10-15  times  higher than the most
 conservative estimate of the deposited lung burden (25-125 mg/kg BW).35   Furthermore,
 direct  intratracheal administration of the diesel particle  extract38 required  doses as high
 as 6 mg/kg body weight before the activity of the  induced  enzyme in the lung was barely
 doubled (Figure 13).  The induction was slow and occurred selectively in lung only (Figure
 14), indicating that diesel particulate extract probably does  not absorb easily  into the lung
 circulation, and is not distributed to other organs.39 The data suggest that the absence of
                   468
                   DOM (mg/kg)
24   48   72   96   120 144
 Hour* After Administration
Figures 13 and 14.  Aryl hydrocarbon hydroxylase (AHH) in the lung and liver tissue after
intratracheal  administration  of diesel particulate  extract (DP-Ext) or  benzotal pyrene
(Btal P).  From Chen, K-C. and Vostal, J.J. (1981)37
                                248

-------
the enzyme  induction  in rat lung  exposed to  diesel exhaust  is caused  either  by the
inavariability  of hydrocarbons for  distribution  in  the  body or  by  their  presence  in
insufficient quantities for enzyme induction. Available data, therefore, indicate that the
inhaled diesel particles would not be  capable of inducing aromatic hydroxylase in the lung
unless  the  total deposited dose in the lung reaches  approximately 6-8 mg of the particle
extract per kilogram  of body weight.  Since the extractable portion represents only 10-15%
of the  total particulate mass, the required  pulmonary deposits of diesel particles in  a 70 kg
man would be excessive to become a significant step  in promotion of a potential neoplastic
process.
   Published data on a similarly negative immune response of the lymphoid tissues in the
respiratory system to the presence of deposited particles are also in good agreement with
the observation  of  the  lack  of biological activity of  the diesel  particles during the
prolonged  inhalation  exposures.   The inactivity of the sequestered particles  is in sharp
contrast  with the laboratory demonstrations,  that the  diesel extract, when administered
                                                                                    40
alone in  excessive doses, produces positive effects in the immune response.  Dziedzic
administered massive doses of dichloromethane extract of exhaust particles (10-50  mg/kg,
three times over 7 days, intraperitoneally) to mice (MUS  musculus), and measured  splenic
lymphocyte response to  the  mitogens  lipopolysaccharide or  concanavalin  A.  Mitogen
responsiveness was determined by isolating spleens, making a suspension of lymphocytes in
medium  RPMI 1640  plus  10% fetal calf serum  and antibiotics (penicillin  100 U/mL, and
streptomycin 100 yg/mL) and culturing cells in flat bottom microtiter wells in the presence
of an optimally stimulating dose of lipopolysaceharide or  concanavalin A.  The cells  were
pulsed  with tritiated thymidine, and the uptake of radioactivity was used as an index of
                  SPLENIC LYMPHOCYTE RESPONSE TO MITOGENS
                                   (MEAN iS.E.)
                            UPOPOLVSACCHARIDE   J CONCANAVALIN A

60
50
30
20
10
0
RESPONSE
300
-
-
-
1





1





,_





200


100


-
-
-

J.




RESPONSE




j-








1


vc 10 25 BO V.C. 10 25 SO
TREATMENT GROUP
Figure 15.   Splenic lymphocyte response to B-cell mitogen lipopolysaccharide or T-cell
mitogen conconavalin A after intraperitoneal injection of diesel particulate extract.  VC =
vehicle control; 10, 25, 50 mg/kg dose. Mean t S.E. From Dziedzic, D. (1981)"°
                                249

-------
response.  The trend toward decreasing responsiveness in extract-injected animals can be
seen in Figure 15.  In a separate experiment,  T cell responsiveness of mice similarly
•njected with extract to a contact hypersensitivity reaction was studied. In this
                    INCREASED EAR THICKNESS VS. TIME
                                 (MEAN ± S.E.)
                      S  30
                         2B
                        r
                        ; 20
                      gl10
                      2    B
  VEHICLE CONTROL
  <
  10 MO/KG DOSE
*^2B MO/KG DOSE
                            0     24     48     72
                           HOURS AFTER ONFB CHALLENGE
Figure 16.  Ear  thickness response to the sensitization challenge of dinitrofluorobenzene
(0.5%) after  intraperitoneaUy administered diesel particulate extract.  From Dziedzic D.
(1981)"°

experiment, groups of mice  were sensitized with a 0.5% solution of dinitrofluorobenzene
(DNFB) on a  previously shaved abdomen.  After four days, they were challenged on their
left ears with 35 pL of the same solution.  Right ears were treated with  vehicle alone. The
increase in ear  thickness at 24, 48,  and 72 hours after challenge indicated a decreased
ability to respond in the extract-treated animals (Figure 16).

DISCUSSION AND CONCLUSIONS
   What appears  to  be evident from these results  is  that in contrast with low level
exposures, inhalation  of diesel particulate at high exposure levels provokes two immediate
responses -focal proliferation of alveolar Type H cells and  increased  numbers of  phago-
cytosing macrophages.  When the exposure is excessive, foamy macrophages with partieu-
lates  aggregate  in  focal areas of the lung.  In  the weeks or  months  that follow, they are
accompanied by a localized secondary proliferation of Type n cells.  Both of the responses
can be identified as  an expression of the efforts by the  respiratory tract to localize and
neutralize  the invading participates.   The reaction  can therefore be  classified as an
effective pulmonary defense mechanism  which can prevent the contact of participates with
the parenchyma  of the pulmonary system.  Only  after extreme levels of  exposure, the focal
                               250

-------
accumulation of macrophages leads to an increased deposition of cholesterol, in the form
of extra-cellular intra-alveolar deposits, which provoke  local quasi-pathological  cellular
changes resulting from the exaggerated protective response.
   The alveolar  macrophage  not only  effectively prevents  more  intimate contact  of
inhaled particles  with  the sensitive  cells of  the respiratory system,  but  is capable  of
deactivating  the  biological aggressivity of the  chemical  materials  adsorbed on  their
surface.  Even if a prolonged storage of the inhaled particles would, therefore, occur in the
respiratory system, it would primarily represent deposits of relatively inert material, which
might be more an indicator of past exposure rather than an index of a clinically significant
biological hazard.
   The biological inactivity of  the particulate  deposits is well illustrated by the negative
response of the inhaled particulates in the induction of metabolizing enzymes as well as by
the minimal  immunological reaction and  lack of significant functional or structural effects
resulting from long-term animal  exposures to  high concentrations of diesel particulates.
This happens  in  spite  of positive  responses  observed  after direct  administration  of
materials obtained by stripping the adsorbed  hydrocarbons by powerful organic  solvent
from the surface  of diesel particulates  in the laboratory test tube.  The living organism
obviously has effective cellular defense mechanisms which can protect against the invasion
of foreign materials and prevent  the occurrence of adverse health effects in vivo  despite
the fact  that they may  have been predicted  by the  complex  chemical  analysis and
biological testing of the materials in vitro.
   It may be therefore concluded that the accumulated experimental data offer a plausible
explanation why  both  the epidemiological studies or  extended animal  inhalation  experi-
ments with high concentrations of diluted diesel exhaust did not reveal clinically significant
changes despite the reported positivity of the laboratory screening tests indicating  serious
concerns for adverse health effect hazard.  It is more important that the experimental data
indicated  a  substantial  protective  role  of the  pulmonary defense mechanisms  despite
conditions of extremely exaggerated  exposure. It can therefore be expected to be fully
capable of protecting the  general population against adverse health effects of the  wider
use of diesel  engines on  our  roads, particularly in  view  of the  expected much lower
particulate levels in the ambient air.
REFERENCES
 1.  U.S. EPA (1977) Precautionary notice on laboratory handling of exhaust  products from
    diesel  engines:  U.S.  EPA  Office  of Research  and  Development,  Washington, D.C.
    (November 1977).
 2.  Vostal, J.J. (1980), Bull. New  York Acad. Med., 56: 914-934.
 3.  Thorslund, T. (1981) in a  letter to U.S.  EPA  Mobile Source Air  Pollution  Control,
    February 2, 1981.
 4.  Albert, R., (1979) in U.S. EPA-CAG Initial Review  on  Potential Carcinogenic Impact of
    Diesel Engine Exhaust, Washington, D.C., June 11,1979.
                                 251

-------
 5.  Harris, J.E. (1981) in the Report to the Diesel Impacts Study Committee, Assembly of
     Engineering, National Research Council, National Academy Press, Washington, B.C.
 6.  Cuddihy, R.G., Seiler, F.A., Griffith, W.C., Scott, B.R. and McClellan, R.O. (1980) in
     Lovelace Inhalation Toxicology Research Institute Report LMF-82, UC-48.
 7.  Siak, J.S., Chan, T.L. and Lee, P.S. (1979), in the  International Symposium on Health
     Effects of Diesel Engine Emissions, EPA, Cincinnati, December 3-5, 1979.
 8.  Brooks, A.L., Wolff, R.K.,  Royer,  R.E., et al. (1979), in the International Symposium on
     Health Effects of Diesel Engine Emissions, EPA, Cincinnati, December 3-5.
 9.  U.S.  EPA  Carcinogen Assessment Group's  Initial  Review  on  Potential Carcinogenic
     Impact of Diesel Engine Exhaust, Washington, D.C., June 11,1979.
10. Huisingh,Jv  Nesnow, S.,  Bradow, R.,  and Waters,  M. (1979), in the International
     Symposium on Health Effects of Diesel Engine Emissions, EPA, Cincinnati, December
     3-5.
11.  Pederson, T.C. and Siak, J.S.(1981) J. Appl. Tox., 1:  54-60.
12.  Pederson, T.C. and Siak, J.S. (1981) in the U.S. EPA 1981 Diesel Emissions Symposium,
     October 5-7.
13.  Thorslund, T. (1981) in the  U.S. EPA 1981 Diesel Emissions Symposium, October 5-7.
14.  Schreck, R.M.,  Soderholm, S.C., Chan, T.L., Smiler,  K.L. and D'Arcy, J.B. (1981) J.
     Appl. Tox., 1: 67-76.
15.  LaBelle, C.H.W. and Brieger, H. (1961) in  Inhaled  Particles and Vapors, Pergamon
     Press, Oxford, pp. 356-368.
16.  Scharfman, A.,  Lafitte, J.J., Tonnel, A.B., Aerts, C., Sablonniere, B. and  Roussel, P.
     (1980) Lung,  157: 135-142.
17.  Warr, G.A., Martin, R.R., and Gentry, L.O. (1976) Bull. Int. Union Tuberc, 51: 569.
18.  Davies, P., Somberger, G.C., and Huber, G.L. (1977) Lab. Invest., 37: 297-306.
19.  Cohen, D., And, S. and Brain, J.D. (1979) Science, 204: 514-517.
20.  Martin, R.R. (1975) Amer.  Rev. Resp. Dis., 107: 596-601.
21.  Strom, K.A.  (1981) in the  UJS. EPA 1981 Diesel Emissions  Symposium, October 5-7,
     1981.
22.  Bowden, D.H. and Adamson, I.Y.R. (1978) Lab. Invest., 38: 422-429.
23.  White, H J. and Garg, B.D.  (1981) J, Appl. Tox., 1: 104-110.
24.  Duguid, J.B. and Lambert,  M.W. (1964) J. Pathol. Bacteriol., 88: 389.
25.  Gross, P. and Nau, C.A. (1967) Arch. Environ. Health, 14: 450.
26.  Gernez-Rieux,  G.  Tecquet,  A.,  Devulder,  B.,  Voisin,  C., Tonnel, A.,  Aerts,  C,,
     Policard, A., Martin,  J-C.,  Bouffant, L.L. and Daniel, H. (1972) Ann. N. Y. Acad. Sci.,
     200:  106-126.
27.  White, H.J. and Garg, B.D. (1981) in U.S. EPA Diesel Emissions Symposium, October
     5-7.
28.  Eskelson,  C., Chvapil, M.,  Barker, E.,  Owen, J.A. and Vostal, J.J. (1981) in U.S. EPA
     Diesel Emissions Symposium, October 5-7, 1981.
29.  Papahadjopoulos, D., Post, G. and Schaeffer, B.G.  (1973) Biochimica et Biophysica
     Acta, 323: 23-42.
30.  Rudd, C.J. and Strom, K.A. (1981) J. Appl. Tox., 1: 83-87.
31. Siak, J.S.  and  Strom, K.A.  (1981) in U.S.  EPA  1981 Diesel Emissions Symposium,
     October 5-7, 1981.
32.  McLemore, T.L., Warr, G.A. and Martin, R.R. (1977) Cancer Letters, 2: 327-334.
33. Palmer, W.G., Allen, T.J. and Tomaszewski, J.E. (1978) Cancer Res., 38:1079-1084.
34.  Pederson, T.C. and Siak, J.S. (1981) J. Appl. Tox., 1: 61-66.
35. Chen, K.C. and Vostal, J.J. (1981) J. App. Tox., 1(2): 127-131.
36. Navarro, C., Charboneau, J. and McCauley, R. J.  (1981) Appl. Tox., 1: 124-126.
37. Peirano, W.B. (1981) in U.S. EPA Diesel Emissions Symposium, October 5-7, 1981.
38. Chen,  K.C. and Vostal, J.J. (1981) in the Toxicology in Michigan Today Symposium,
    May 8.
39. Chen,  K.C. and  Vostal, J.J.  (1981)  in  the  U.S.  EPA Diesel Emissions Symposium,
    October 5-7.
40. Dziedzic, D. (1981) in the U.S. EPA Diesel Emissions Symposium, October 5-7.
                              252

-------
INVESTIGATION OF TOXIC AND CARCINOGENIC  EFFECTS OF DIESEL EXHAUST
IN  LONG-TERM INHALATION EXPOSURE  OF  RODENTS

U.HEINRICHX,L.PETERSX,W.FUNCKEX,F.POTTXX,U.MOHRXXXand W.STCBERX
 Fraunhofer-Institut far Toxikologie und Aerosolforschung,Munster,
Federal  Republic of Germany;XXMedizinisches  Institut far Umwelt-
hygiene , Dflsseldorf, Federal Republic  of Germany ,-XXXMedizinische
Hochschule,  Experimentelle Pathologie,Hannover,Federal Republic
of  Germany
INTRODUCTION AND OBJECTIVES
   Approximately 2 1/2 ago a long-term  study  using Syrian golden
hamsters was begun taking  as  its basis a five months inhalation
    '                                                            1
study  using  three different dilutions  of  Diesel  engine  exhaust .
   The  primary objective of this  life  time  exposure study was to
investigate  as to whether a carcinogenic  or a syncarcinogenic
effect  could be induced in the respiratory  tract of the golden
hamsters  by  inhalation of either  the diluted total exhaust from a
Diesel-engine or the same exhaust void of particulate matter. In
addition  to  the testing for a potential carcinogenicity of the
Diesel  exhaust, a variety of data on clinical chemistry and
hematology were collected at certain intervals.  Furthermore,
several tests of pulmonary function were  performed. They were
carried out  with rats which were  exposed  to the  exhaust along with
the hamsters.

EXPERIMENTAL FACILITIES AND STUDY DESIGN
   The  exhaust gas for our experiments  was  produced by a 2.4 liter
Daimler-Benz Diesel engine. The engine  was  operated at a constant
load of 16 kW and a uniform speed of 24oo r.p.m.  The fuel used in
the engine was a European Reference Fuel  with a  sulfur content of
o. 36 %.
   The  exhaust gas pipe was divided into  two lines. One line passed
the emissions  directly from the exhaust pipe to  the mixing chambers,
while the other line sent the exhaust  first through a centrifugal
seperator in order to remove all  particulate matter from the ex-
haust. Both,  the genuine exhaust  and the  non-particulate exhaust
were then diluted in the mixing chambers  with clean, dry,
                          253

-------
 refrigerated air  at  a ' ratio of  one  part exhaust  to  seven parts
 of  air.  The previously  reported 5-months study had  indicated that
 this  was an optimum concentration for a life time study that was
 to  avoid acute  toxic effects.
    The  inhalation chambers,  had a volume of about 25o  liters. They
 were  horizontally ventilated with approximately 5m of diluted
 exhaust  per hour. Special baffles and perforated diaphragms were
 placed  between  the  exposure  area and the inlet and  outlet  tubes,
 respectively. They  provided  a  uniform distribution  of  the  aerosol
 flow  in  the exposure chambers.  In each chamber, 24  hamsters were
 held  on  2  levels of wire  cages.
    The  actual concentrations  of a selected number of exhaust
 components were  either  measured continuously, or monitored in
 certain  intervals.  One  set of  measurements was made right  in the
 inhalation chambers and another set was obtained in the  conduits
 approximately 5  m upstream of  the chambers. Inside  the  inhalation
 chambers,  the average particle  concentration of the diluted ex-
 haust was  3.9 mg/m  .  The  mass  median aerodynamic diameter  of the
 exhaust  particles was o.l  ,um.
    The  concentration of several gaseous components  contained in the
 exhaust,  such as CO,  CO.,  S02»  total hydrocarbon, methane  and non-
 methane  hydrocarbons, NO,  NO  ,  N0_  and O_, were present  in compar-
 able  concentrations in  both,v the chambers  receiving total  exhaust
 and the  ones receiving  the exhaust  without particles (Fig.l).
 The measurements taken  within  and upstream of the chambers indi-
 cated that no major changes  in  concentration occurred on the way
 from  the mixing  chambers  to  the inhalation chambers. The concen-
 tration  of all gaseous  exhaust  components  measured  in these
 locations were sufficiently  low so  that SO  and NO.  concentrations
would not exceed 3  and  1  ppm,  respectively, and no  acute toxic
effects  on the experimental  animals  were expected.
   The sulfur content of  the soot particles in the exposure
chambers was measured by  way of x-ray fluorescence spectroscopy,
which gave a value  of 37  mg/g  of soot.  In  addition  to this measure-
ment,  quantitative  analyses  were carried  out for 14 specified
polycyclic aromatic  hydrocarbons (PAH)  which were found  to be
bound to the soot particles. No details will be given here about
the extraction of the PAH.These processes  have been  described in
                         254

-------
in some  earlier publications  of  our institute'

CO I PI")
C02 tVoll]
so2 ipp»J
C^lPP.)
CH4 tPP»J
C^Hj, - CH4 Eppal
NO Ipi-J
MO, tpf»)
NO, IPF-I
07 [Woi»]
PAmCLfS [•»/• 1
MEASUREMENT
IN THE CHAMBERS
A
18,0 (+ 4,4)

2,8 (i 1,7)
7,9 <+ 3,3)
2,6 (i 1,8)
5,0 (+ 2,5)
17,2 (+ 5,9)
19,2 (J 6,1)
1,0 (* 1,5)
20,0 (+ O,7)
-
B
18,5 <+ 4,9)
0,54 it 0-13)
3,1 (i 1,8)
9,3 (J 4,6)
3,0 (t 2,2)
5,9 (+ 3,0)
16,5 (* S,8>
18,6 (J 5,8)
1,2 (+1 ,7)
19,5 (J 0,6}
3.9 (+ 0,5)
BEFORE THE CHAMBERS
A
17,0 (+ 3,7)
0,47 (J 0,11)
4,2 ( + 2,9)
4,9 (+ 2,7)
1,0 (J 0,3)
3,9 (+ 2,4)
16,2 (+ 5,5)
17,8 (+ 6,4)
0,8 <± 1,2)
2O, 0 (+ 0,7)
-
B
19,0 (± 5,1)

3.8 (t 2,2)
4,9 (+ 2,4)
1 ,0 (+ 0,3)
3,3 (* 1,9)
16,5 (+ 6,4)
17,9 (+ 7,2)
O,3 (j 0,3)
19,1 (+ 0,5)
-
            At EXHAUST KITHWT PWTICLES
                              TOTAL EXHAUST
 Fig.  1.  Component concentrations of the diluted  Diesel  exhaust
   Figure  2  shows in the last  column the concentration  of  14  PAHs
extracted  from the diesel soot as  compared to the concentration  in
samples  of  airborne dust taken from 4  different  cities  in  the
Federal  Republic of Germany  (Fig.2). The PAH concentrations  given
in  ,ug/g of  dust or soot indicate  that,  except for fluoranthene
 (FLO)  and  pyrene (PYR), the  concentration of PAH is substantially
higher in  airborne dust than in  Diesel soot. However, the  PAH
concentrations in airborne dust  are average values over  a  collec-
ting period  of one year (1979/8o)  during which the collecting
filter was  changed every two weeks.  In case of the Diesel  soot,
on the other hand the collecting  period  was only about  8 hours  and
the concentrations of NO  and  SO   were observed  to be higher  in  the
dilute exhaust than  in the city  air. Thus,  on account of the
different sampling conditions, a  comparison of the PAH
concentrations in Diesel soot  and airborne dust  is of  rather
limited  value.
                          255

-------
^N^MITLE
cmr"--^
pur
PY«-
Btf-
Elf
W*
OK*
BR*
Bep-
BAP *
P0C
IP*
HA*
BwiP'
CM'
1. Tow
26.1
21.9
7,1
11.0
36.7
80.9
138.7
59.4
27.2
6.3
56.6
14.6
50.7
24.0
2. Tow
21.5
19.6
6.1
11,4
35.6
80,2
141.3
58.3
27.0
5.9
37.2
11.2
44.5
19.2
3. Tow
31,5
35.9
12.2
20.8
51.0
101,6
165,4
69.5
42.0
7.3
42.7
13,5
56.5
26.9
4. TOM
17.0
13.1
2.4
5.6
17.8
46.2
94.1
42.0
14.7
3.9
23.5
7.2
34.8
18.2
DIESEUW.
134.6
65.8
5.4
5.3
9.8
25,7
22.2
14.1
7.0
_
13.4
-
21.4
12.5
                  » • cMcmmmc. - - IW-IUHCIKOUNIC in MINU.
Fig.  2. The  PAH  content (14 PAH)  of airborne particulate  matter
and of Diesel  exhaust particles (in ,ug/g dust or  soot).

   Based on  the  PAH  concentrations measured in 'the  inhalation
chambers, an estimate can  be made  as to the inhalation  impact of
PAH on the experimental animals.-  Assuming that the  hamsters have
a respiratory  minute  volume of loo ml, then a daily exposure period
of 8 hours for 5oo days in 2 years would amount to  24 m  of inhaled
diluted exhaust.  Estimating the deposition rate of  inhaled par-
ticles at some 4o %  which  is conservatively high and taking the
average particle  concentration at  4 mg/m , then, at most, about
4o mg of soot  would  be  deposited  in the lungs of the experimental
animals.  With  reference to a measured benzo(a)pyrene (B(a)P)
concentration  of  7  ,ug/g of soot,  this would represent  a  B(a)P
uptake of about  27o  ng,  if the availability of the  B(a)P  bound
to soot particles is  not limited.  Apparently, this  estimate
                        256

-------
suggests  that  less than 1 ^ug of B(a)P and  similar  amounts of
most of the  other PAHs considered here may  be  deposited in the
lungs of  the hamsters. Therefore, it seems  that  the quantity of
carcinogenic PAHs inhaled by the experimental  animals  is much too
small, and  the maximum available latency period  of  about 2 years
based on  the normal life span of the experimental  animals is much
too short to make any negative result on carcinogenicity really
conclusive.  In other words, by way of inhalation of dilute Diesel
enginge exhaust alone, the animal experiments  will  most likely
not permit  a firm and definite conclusion with regard  to a
carcinogenic!ty in human beings.
   Since  it is  rather unlikely, that unsophisticated  straight-
forward inhalation tests with dilute exhaust gas will  ever give
statistically  significant proof for the existence  or non-existence
of a carcinogenic effect, our lifetime experiment  used  the follow-
ing animal  model: In order to observe subtle changes in tumor
incidence   rates, it is desirable to work with an  increased tumor
frequency.  An  enhanced basic tumor incidence rate  in the
respiratory  tract can actually be induced by intratracheal or
subcutaneous administration of carcinogenic dibenzo(a,h)anthracene
(DB(a,h)A)or,  for instance, diethyl nitrosamine  (DEN).  Then, in
case there  is  an actual additive or synergistic  effect of the ex-
haust, the  basic tumor incidence rate will  increase if  the Diesel
engine exhaust is inhaled simultaneously.Thus, without  using
extremely large numbers of animals, the probability of  detecting
a significant  change in the tumor frequency is greater,  if these
changes occur  along the steep slope of the  sigmoidal dose response
curve rather than along the shallow initial slope  of that curve.
This may  be  effected by having a sufficiently  enhanced  basic tumor
rate instead of the low spontaneous incidence  rate. In  the steep
slope range  of the curve, relatively small  changes  in  dose.should
result in relatively large changes in tumor frequency.
   The hamsters were kept in wire cages. Due to  space  limitations,
each compartment held 3 animals. The temperature in the inhalation
chambers was kept at 24-25°C and the humidity  was  55-65 %. The
animals were exposed for 7-8 hours/day and  5 days/week.  During the
exposure periods, all feed was removed from the  cages  in order to
prevent oral uptake of Diesel soot.
                         257

-------
    As  can be seen in the experiment protocol,  18  groups  of  at  least
 48  animals were used in the inhalation program  (Fig.3).
in. IMT. / i.e.
HITHWT
m<..h>, I m
IK I 0.3 M IN 0.02 «J
a(*.» * 2 M
(» * o.i m IN e.i* ML)
mm J M
(SO 1 0.1 Hi IN 0.1S Mj
DO '•» M/Ki HI
D« <•> M/KI M
CUM AI«
48
48
48
48
77
48
TOTAL EXHAUST
48
48
48
48
57
60
EXHAUST WITHOUT
PARTICLES
48
48
48
48
48
48
 Fig.  3.  Number of animals in the different groups with  and  without
 additional treatment (i.tr. - intratracheal, s.c. -  subcutaneous)
    There  were  three  exposure atmospheres: clean air, total exhaust
 and exhaust  without  particulate  matter.  For each exposure type,
 there  were 5 treated animal groups.  Two  of these groups received
 subcutaneous injections  of  1.5 or 4.5 mg DEN/kg body weight at the
 beginning of the  experiment. Two other groups received o.l and o.3
 mg  of  DB(a,h)A by intratracheal  instillation  once a week for 2o
 weeks. The control group corresponding to the animals treated with
 DB(a,h)A  was instilled once a week  for 2o weeks with pyrene, a non-
 carcinogenic PAH. Another group  received no additional treatment.

 RESULTS
 Median lifetime of the hamsters
   The experimental  animals  employed in  this  study were female
 Syrian golden  hamsters from  a breeding farm in  Frankfurt-Hoechst,
West-Germany.  At  the beginning of the experiment,  the animals were
approximately  8 weeks old.Inspite of an  apparently unfavorable
condition of keeping the hamsters in wire cages,  the median
                         258

-------
experimental lifetime,  that is the time where  5o % of the  animals
had died,  was still  72-74  weeks for both,  the  controls and  the
animals  exposed to the  emissions  (Fig.4).  This corresponds  to  an
actual average lifetime  of 8o-82 weeks. Contrary to expectations,
the longest average  lifetime did not occur among the untreated
controls,  but with the  hamsters treated with  DB (a,h)A.
              •I.
             100-
              75-
             50-
             25-
                                          — control grouptcUon air]
                                          	total t»haw*t
                                                •»nau$i
                24  32  41  49  57  66  74  82  90  98  106 £  122
Fig.  4.  Surviving hamsters at different  times after the  start of
the  experiment  (in  %).

Clinical chemistry  and  hematology
   For  the animals  not  receiving any  special pre-treatment,  several
hematological parameters  as  well as a number  of  enzyme activities
and  metabolic products  were determined  following the  29th,  the
42nd,  the 55th  and  the  75th week of the  experiment. The  different
values  at various points  in time represent in each case  an  aver-
age  value from  14-2o  animals. Lines were  drawn through these  data
points  to yield curves  for the individual parameters. The  graphs
of the  hematological  (Fig.5,6) and clinical chemical  data  (Fig.7,8)
of the  control  animals  and the animals  exposed to the total ex-
haust  give an example.
The  following hematological and clinical  chemical parameter were
determined;  erythrocytes  (ERY),  hemoglobin (HB), hemoglobin/ery-
throcytes (HB ),hematocrit  (HCT), mean  cellular volume  (MCV),
              E
                          259

-------
                                     AFTER aB.4Z.58 AMJ7B WEEKS
                                CCNTHOt. OWUP
Fig.  S.  Haematological  data  of the  control  group
                          HAEMATOLOQT AFTER 2542.56 AND 75 WEEKS
                                  TOTAL EXHAUST
                                                            MCHC,
                                                            •«*,
                                                            •HB,
Fig.  6. Baematological data of  the  group  inhaling  total  exhaust
                          260

-------
                                     OJNC&L O&fSTPY AFTER 39 42 S5 -WO 75 WtEKS
                                       CONTROL GBOUP
 Fig.   7.  Clinical  chemical  data  of  the   control  group
                                          CLINICAL CHfiMIBTKY AFTEfl K 42. S Af>C 75WEEKS
                                            TOTAL  EXHAUST
                                                                                  CHE'
Fig.  3.   Clinical   chemical  data  of  the  group  inhaling  total  exhaust
                                       261

-------
mean cellular  hemoglobin  concentrations (MCHC) , leucocytes  (LEUCO) j
glutamic-oxalacetic  transaminase (GOT), glutamic-pyruvic  trans-
aminase  (GPT),  cholinesterase (CHE), lactic dehydrogenase  (LDH) ,
alkaline phosphatase  (AP),  cholesterol (CHOL), gamma-glutamyl
transferase  (gamma-GT), alpha-hydroxybutyrate dehydrogenase  (alpha-
HBDH), blood urea, creatinine (GREAT).
For both,  the  controls  and  the exposed animals,  it was noticed
that several parameters,  such as CHE,  AP and alpha-HBDH, showed
values of  great  variations,  which did  not always go in the same
direction. At  the same  time,  from about the 3oth or 4oth experi-
mental week on,  the  controls  and the test animals, alike, showed
increasingly serious  degrees  of amyloidosis of the kidneys,
adrenals,  liver  and  spleen.  Usually the kidneys were affected the
earliest and most markedly.  In addition,  there were liver cysts
(enlargements  of the  bile ducts). Therefore, it appears justified
to attribute the observed fluctuations of the biochemical para-
meters mainly  to endogenous  organ changes .  In view of the fact,
that such  endogenous  organ  changes  were not detected during the
first 3o weeks of the exposure,  only the values for the 29th
experimental week are analysed here for potential changes due to
exposure.
  For the  animals exposed to  the exhaust emissions the results
indicate that, after  29 weeks, significantly enhanced values occur
for enzymes such as GOT, LDH  and AP. The same goes for gamma-GT
which was  not  included  in the graphs because of a different scale
(Fig. 9).  These  results give  reason to suspect that the inhaled
Diesel emissions may  have an  effect on the  liver.  The significantly
elevated values  for blood urea in the  animals exposed to Diesel
emissions  may  indicate  an impairment of the renal function, but
the low creatinine values do  not support this interpretation.
  Electron-microscopical studies employing  ultra-thin sectioning
techniques on  livers  of hamsters exposed to exhaust emissions show,
among other things, the degenerative changes of structure in the
mitochondria of  the hepatocytes  . The  increase in the mitochondria!
and microsomal enzymes, as  indicated by our GOT and gamma-GT values
may be considered a result  of this  type of  damage  to cells or
organelles.
  The hematological  studies in the  29th experimental week  revealed
                        262

-------
                                    OJNCAl OSMETHY AFTER 39 WSEKS
               330-


               300


               270-


               3*0-


               310-
CONTBOL

GASEOUS EXHAUST

TOTAL EXHAUST
                                                          UHSA4   GREAT'
Fig.  9.  Some  clinical  chemical  data  determined  for  2  exposed and
1  control  group  after  the  29th  week  of  the  experiment
                              HAEMATOLQGY AFTER 29 WEEKS
*,«! «£

90-
7O
BO-

50-
«,
30-

30-
10-
'rV gHO/aEr,5










T








TsT
r
^
p
^
'/ •
//
£
//,
1
I










T T T






It*





x^
i
^'
i
i
^
I
TI




















s
1



*







1



i
F







%
//^
1



1
ri
i























-i.











*
1
i









*


ncoMTO.
^ GASEOUS EXHAUT
[_J TOTAJ. etHALJST




S
J-VTX3-.





1
F''
i








••

"~1-




:

i
-, \s
^\








                             HCT1
                                           UICV3
                                                               LEUCD'
Fig.  lo.  Some  haematological  data  determined  for  2  exposed and
1  control group  after  the  29th  week  of  the  experiment
                                 263

-------
a  low  erythrocyte count which was accompanied  by an increase in
erythrocyte  volume and a reduced leucocyte  count in the animals
inhaling  Diesel  exhaust (Fig. lo).

Histo-pathology  of the respiratory tract
   The histo-pathological  examination of trie respiratory tract
yielded the  following findings:  A lung tumor was  found in only 2
out of all of  the hamsters used in the experiment.  One of these
animals belonged to the group that received a  total dose of  6 mg
DB(a,h)A  and was exposed to the particle-free  Diesel emissions. The
other  animal was found in  the group which had  inhaled the total ex-
haust  and had  received an  additional injection of  1.5 mg DEN/kg
body weight. These animals died after experimental  exposures of 75
and 67 weeks,  respectively.
   According to  the independent results of our co-author Friedrich
Pott,  intratracheal instillation of DB(a,h)A will  induce a high
degree of lung tumors  in the  golden hamsters which  are  available
from the Central Institute for the Breeding of Laboratory Animals
                         4
(TNO)  in the Netherlands .  By comparison, the  animals used in our
exhaust inhalation study,  which  were obtained  from  the  breeding
farm of the Hoechst Company,  West Germany,  appear  to  be  sub-
stantially less  sensitive  to  DB(a,h)A.  In view of  the fact that the
same dose and method of application was used,  it must be  concluded
that these two lines of hamster  show genetic differences  regarding
the sensitivity  to the carcinogenic action of  DB(a,h)A.  At present,
a suitable experiment  is in progress to check  this  conclusion. This
may explain why  we failed  to  produce any enhanced basic  tumor rate
in the hamsters  treated with  DB(a,h)A.

   A more common observation  were some discrete proliferative
changes in the lung,  60 %  of  which were described  as  adenomatous
proliferations.  If these cases are expressed in per cent of  the
total number of  investigated  animals per experimental group, one
arrives at the following result: In all experimental  groups  exposed
to the total exhaust emissions,  independent of any  additional treat-
ment, the number of animals  exhibiting definite proliferations in
the lung is significantly  higher than for the  corresponding  experi-
mental groups exposed  to particle-free emissions or clean air
                         264

-------
 (Fig.  11).  Furthermore, the percentage  of  hamsters exhibiting rro-
 liferative  growth in the lung is particularly high for all experi-
 mental  groups  receiving intratracheal  instillations.  Nevertheless,
 even  in case  of the hamsters treated by instillation, the highest
 incidence  of  proliferative growth  is always  observed  in those ani-
 mals  exposed  to the total exhaust  emissions.  In this  case, pro-
 liferation  is  seen much more frequently in  the experimental groups
 treated with  DB(a,h)A than in animals  instilled with  pyrene.
   With the DEN-dosages used in our study,  that is 1.5 and 4.5 mg/
 kg body weight, basic tumor rates  of 13.4  and 44.7 %, respectively,
 were  induced  in the larynx/trachea region  of  the control groups.
 These  tumors  proved exclusively to be  papillomas.  Papillomas  have
 also  been  observed in inhalation studies with pure B(a)P . In the
 case  of the higher DEN-dosage the  tumor incidence  increased
 significantly  to  66 % and to 7o.2  % by  inhalation  of  particle-free
 exhaust and total exhaust,  respectively. The  difference between the
 two increased  tumor rates however, is  not  significant (Fig. 12).
   The  animals which were treated  with  1.5  mg DEN/kg  bodyweight and
 exposed to  the exhaust emissions,  also  exhibited a tendency toward
 increased  tumor incidence rates. However,  there was no significant
 difference  in  the tumor incidence  among the  three  groups of
 different  exposure atmosphere. As  expected,  the tumor incidence
 curves  for  the smaller DEN dosage  groups are  substantially lower
 and first papillomas  were observed about 2o  weeks  later than  in
 case  of the high  DEN dosage.
   Among all the  experimental groups not treated with DEN, only
 5 animals were observed with papillomas  in  the larynx/trachea
 region.  These  5 animals all belong to  the  three experimental  groups
 treated with  (DB(a,h)A. No  animal without  special  carcinogenic
 treatment,  exhibited any tumor in  the  respiratory  tract,  no matter
what  the inhalation exposure was.
The high DEN dosage which induced a basic  tumor incidence  rate of
44.7  %  in the  controls, caused an additional  increase of the  tumor
 frequency in conjunction with the  total exhaust emissions  as  well
as with  the gaseous components alone.  Further inhalation studies
involving Diesel  and  gasoline engine emissions are  presently  under
way in  our  institute  to explore  in greater detail  as  to whether
this effect is  reproducible and significant  for a  co- or syn-
                         265

-------


1
1
1
1
1
1
p1


E
E
1
1
in









i;
1
EXHAUST ONLT |* PTHENE


i
i
i
i
i
Hi





I
1
1
1
1
2mg \-OSlv
| 	 1 CLEAN AIR
SEXH WITHOUT run
1 1°"IE*M 1
1





0
1
~
E
E
E
E
—
E

1
1
1 n
1 y
i
in I
^
—
~
—
E
=

)A Imt (.C6ta,hlA 6mg
1
1
1
1
1
1 E
Q| =
ra lin 153 =
. DEN I.Smg \- DEN i.Smg
Fig.  11.  Percent  of hamsters  with  focal proliferations  in the  lung.
There are  5 treatment  groups  (Pyren,DB(a,h)A,DEN)  and  1 group
without  additional treatment
             80-

             70-

             60-

             50-

             40

             SO-

             OT-

             10-
' Hamsters with papiHomas in larynx and/or trachea


    > DEN 45 , total exhaust
    °  "  "  , exhaust without particles
    o  "  "  , dean air
       "  1.5 , total exhaust
       "  "  , exhaust without part
       "  "  , dean air
                   1020304050607080
                                                   90
                                                        100  110   120
                                                            [weeks]
Fig.  12.  DEN-treated hamsters  with  papilloma in  larynx/trachea ex-
pressed as  a percentage  of the animals oer arouo at  the start of
the  experiment.  DEN 4.5  = 4.5  mg/kg  b.w.;DEN 1.5  = 1.5  mg/kg  b.w.
                            266

-------
carcinogenicity of Diesel  exhaust,  or may be obtained  with  any
other  inhalation burden  containing  some gaseous components  such
as  N02  and SC>2. It will  further  be  checked whether  the  effect will
also appear in combination  with  other carcinogens affecting the
respiratory tract. The basic  tumor  incidence rate of 13.4  %,  as
induced by the small DEN dosage,  may possibly not have  fallen into
the steep slope range of the  dose dependent tumor frequency curve.
This would explain why no  statistically measurable  change  in  the
tumor  incidence rate could  be  observed.

Pulmonary function tests
    The  pulmonary function  tests  have been conducted on  female Wistar
rats,  which were exposed to the  exhaust for up to 24 months under
the same conditions as the  hamsters.
    After an exposure period of approximately 18 months, these rats
were subjected to tests  in  order  to determine the lung  clearance
                                          59
rate for a hematite iron oxide aerosol (   Fe O )   The  iron  was
radioactively labelled and  the insoluble  aerosol was inhaled  for
a short period of time so  that a  fraction would deposit in  the
respiratory tract. At various  timesafter  the inhalation of  the iron
oxide  aerosol, the activity over  the thoracic lung  area was measured
and recorded as a percentage  of  the activity measured  on the  first
                                  59
day. After being exposed to the   Fed -aerosol the animals con-
tinued  to inhale the Diesel exhaust. As can be seen from the  de-
creasing exponential curves for  the activity measured  between the
15th and the 7oth day following  the iron  oxide inhalation,  the
long-term alveolar clearance  is  definitely impaired by  the  total
exhaust emissions and also, to a  somewhat lesser degree, by the
gaseous phase of the exhaust  (Fig.  13,14)   The biological halflife
of  the  iron oxide deposit  in  the  lungs of the animals exposed to
the non-particulate components of the exhaust is  about 4o % longer
than in case of the controls. For the animals exposed to the  total
exhaust emissions, the halflife  is  nearly twice as  long as  in the
controls.
    For  the rapid phase of  lung clearance,  the biological halflife
was not measured because the  short-term clearance which is
primarily  determined by the bronchial clearance mechanism,  is
practically completed within  15 days after iron oxide exposure  and
could not  accurately be measured  on account of the  small numbers
                          267

-------
          I"
          540
                          CLEARANCE OF S9f*&3 FROM RAT LUNGS
                                           50     6O
                             59
Fig.  13.  Lung clearance of   Fe-O-particles in rats  exposed to
total  Diesel exhaust.
                              5 9
Fig.  14.  Lung clearance of   Fe^O-particles in rats  exposed to
Diesel  exhaust without particles.
                          268

-------
of animals  involved. However,  it  can  be  seen from the distribution
of the  measured values within  the first 15 days that  animals  ex-
posed to  the  particle-free emissions  show  a better degree of
bronchial  clearance activity than  the  corresponding control ani-
mals. The  animals exposed to total  exhaust emissions exhibit  a
lower bronchial clearance activity  than  the controls.
   No significant changes of respiratory rate,  respiratory minute
volume  or  compliance and resistance,  as  measured with a whole body
piethysmograph,  were observed  in  any  of  the test animals. This
applies to  exposures for over  2 years  to the gaseous exhaust
components  as well as to the total  emissions (Fig.15)

brMtb/Bia.
TldAl VQlVM
ml
HlBOt* VOlOM
ml
lntr.pi«r^
em 1,0
CaBpliABC*
mi/ em H.O
BMUUDC*
<=• H.O/ml-W 1
qnotlmt
T «xp./ T tn*p.
flMrt r.t*
MlM/BiB.
Control
41.
(SO,
3,
(3,
127.
(133.
4,
(4,
0,
(0,
0,
(0,
* 1
+ 1


+ 3
* 3
*
*
3 +
1 ±
4 *
6 -
0,S3 *
(0,58 +
317 + 4
(333 * 3
; ,
, i
. 3)
,
. )
.77
,13)
,3S
,14)
.03
,03)
.14
.111
,5
.7)
°— "">in>«"
(44,
3,
(3.
1)8,
(131,
4,
(«.
o.
(0.
0,
(0,
0,
(0,
+ 21,6
* 21.2)
*
1
* 7
* 5
*
1
*

*

+
±
134 +4
(334 ± 4
,«1
,90)
,5
-8)
,1
.0)
,15
.31)
r 16
,1B)
,15
,25)
,3
,0)
Teul «xnau»t
(5

(
14
(13

(

(

(
,5 i 18.0
,4 + 17,81
,6 * 0,64
.5 * 0,52)
,5 * 39,5
,9 + 40,7)
,0 + 0,97
.1 + 1.0)
,84 * 0,27
,7B + 0.26)
,14 * 0,05
, 17 + 0,O»)
0,60 + 0,13
(0,56 * 0,16)
105 + 32.9
(309 + 31,5)
                    ntmtion -0.3%
?ig. 15. Pulmonary  function test of  anesthetised rats after ;
vears exposure  with total Diesel  exhaust  or gaseous exhaust
without particles.
                         269

-------
   Even when briefly  inhaling an acethylcholin  aerosol  prior  to
the measurements  of pulmonary function, the test  animals  did  not
react more sensitively  than  the control animals.  It  thus  may  be
concluded that, even  after  long term inhalation of dilute Diesel-
engine exhaust, a significant impairment of the mechanical pul-
monary functions  is not observed.

ACKNOWLEDGMENTS
   The work published in  this paper is part of the research'
activities of the working group "Investigations on the  carcino-
genic burden of humans  by air pollution" of the Umweltbundesact
Federal Republic of Germany.
The Diesel engine exhaust inhalation study was to some extent
supported by the Daimler-Benz AG.

REFERENCES
1.  Heinrich.U.,Stober,H.  and  Pott,F.(198o)  in Health Effects
   of Diesel Engine Emissions:  Proceedings of an International
   Symposium,Pepelko,W.E., Danner,R.M. and Clarke, N.A.,ed.,
   EPA-6oo/9-8oo-o57b,pp. Io26  - Io47.
2.  K6nig,J.,  Funcke,W., Balfanz,E.,  Grosch,B., Romanowski,T. ,
   Pott,F. (1981)  Staub-Reinhalt.  Luft, 41/3,  73 - 78.
3.  MeiB,R.,  Robenek,H., Schubert,M., Themann,H., Heinrich,U.  (1981
   Int. Arch. Occup. Environ.  Health 48, 147   157.
4.  Pott.F.,  Mohr.U., Brockhaus.A.  (1978) Abstract in Medizinisches
   Institut fur Lufthygiene und Silikoseforschung, Jahresbericht
   1978,  Vol. 11,  pp. 225 - 226, W.  Giradet, Essen, FRG.
5.  Thyssen.G., Althoff,J., Kimmerle,G., Mohr,U.(198o) in  VDI-
   Berichte  Nr. 358, pp.  329-333,  Verein Deutscher Ingenieure,
   Dusseldorf, FRG.
                         270

-------
MORPHOHETRIC ULTRASTRUCTURAL ANALYSIS OF ALVEOLAR LUNGS OF GUINEA PIGS
CHRONICALLY EXPOSED BY INHALATION TO DIESEL EXHAUST  (DE)
MARION I. BARNHART, STEVEN 0. SALLEY*. SHAN-TE CHEN AND HENRY PURO**
Departments of Physiology, Anesthesiology* & Pathology**, Wayne State
University School of Medicine, Detroit, MI  48201
INTRODUCTION
   Increased usage of diesel engines is expected in the future to meet energy
needs for industrial, home and recreational needs.  Concern about possible
health effects of diesel exhaust  (DE) is reasonable.  Most studies have focused
on cancer risk but there is presently no statistically significant difference
in tumor incidence in controlled  laboratory animal experiments .
   Diesel emissions have a gaseous phase similar to emission controlled gasoline
engines.  But, it is established  that diesel engines emit 50 to  80 times more par-
ticulates (DEP) than do gasoline  engines .  Most if not all of the particulates
are in the respirable range with  particulate size near 0.3 um  .
   In assessment of the health risks of DE, it seems especially pertinent to
understand the lung's response to a chronic burden of DEP.  The physiologic and
pathologic impacts of DE exposure on the lung can be defined by a systematic
ultrastructural approach.  Our group began such a study three years ago in coop-
eration with the General Motors Biomedical Laboratory.  We are studying the
effects of acute and chronic inhalation of DE (250 -6000 ug DE/m3) in Fischer
rats and Hartley guinea pigs exposed for periods up to 2 years.  Several publi-
cations are anticipated as necessary to adequately describe our results and
make interpretations of the massive data bank.  Preliminary reports have been
published   .  The present report documents the impact of DE on pulmonary ultra-
structure of alveolar lung of guinea pigs subjected to chronic exposure of DE
from 2 weeks to 2 years.  Morphometric approaches are used to quantify selected
lung parameters.

MATERIALS AND METHODS
Animals
   Male Hartley guinea pigs (about five weeks old) from Charles River Breeding
Laboratory, were housed at the General Motors Biomedical Research Laboratory
(GMBL), Warren, MI for an initial quarantine period of two weeks prior to enter-
ing the controlled exposure regimens at GMBL7.  Individually caged pigs were
                              271

-------
placed in either a  clean  air  environmental chamber or chambers receiving fresh-
ly diluted diesel engine  exhaust to achieve particle concentrations of either
250, 750 ug, 1500 ug or 6000  yg/m3.  Chamber temperature was 22 + 2°C with 56 +
6% relative humidity.  During cleaning periods the animals were rotated to en-
sure equal exposures.  Food and  water were always available.
   Fasted guinea pigs  (3  per  exposure group plus age matched controls) were de-
livered at Wayne State University for sacrifice, dissection and tissue process-
ing for the ultrastructural studies.   These 64 animals were assigned code num-
bers which were unknown to investigators.
   Exposure conditions and characteristics of DE.  The animals were exposed by
inhalation to either clean air or DE  air for 110 hr/week.
   The DE exhaust was emitted by a production 1978 Oldsmobile 5.7L, four cycle,
indirect injection  diesel engine.   It was  run at steady speed and load to simu-
late a 65 km/hr (40 mph)  cruise  situation.  Amoco type 2D federal compliance
fuel and Amoco 200  30W lubricating oil were used.  Well dispersed DE particu-
lates were delivered uniformly to the inhalation chambers in an air flow of
2.8L/min.  The mass median aerodynamic diameter of the DE aerosol was 0.19 +
0.03 urn with 83 + 5% of the mass in particles smaller than 1 ym .  Details of
exposure conditions and monitoring of particulate and gas concentrations have
been published .  The average particulate  mass concentrations were within 2% of
the target dose value.
   Animal and tissue preparation.   Guinea  pigs were anesthetized with an intra-
peritoneal injection of sodium pentobarbital.   A tracheotomy was done and a
plastic cannula (Abbott Butterfly -19 set  without needle) was secured in the
trachea.  After collapsing the lungs  by letting air enter the thoracic cavity,
the intact lungs were reinflated and  fixed in situ by instillation of pH 7.4
cacodylate buffered 1% glutaraldehyde of 300 mOsms.  This fixative fluid was
instilled at a pressure of 20 cm H-0  above the hilum to achieve a normal expan-
sion of the lungs.  The trachea  was tied to insure preservation of the intra-
pulmonary fixative  volume and the still intact lungs were removed from the
chest and placed in fresh fixative.  .Lung  volumes were determined after removal
of extrapulmonary structures,  using volumetry following the submersion method
          o
of Scherle .  Lungs remained  submerged in  glutaraldehyde for 18-24 hr prior to
selection of tissue samples for  microscopy.
   Tissue slices CO.5 x 1 cm)  were taken for the scanning electron microscopy
(SEH)  study which were used to establish the alveolar parenchymal lung volume .
Specimens were prepared according to  published procedure.
                             272

-------
   The transmission electron microscopy  (TEM) specimens  (cubes  0.2  cm  thick)
were cut from tissue slices of the right middle lobe and  left lower lobe.  An
admixture of cubes represented dorsal and ventral parts,  exclusive  of  pleural
edges.  At least 5 cubas from each animal were selected randomly and processed
by conventional procedure  .  Semi-thin sections (0.5 - 1.0 urn)  and  ultrathin
sections (about 60 nm) were cut using a diamond knife.  The semi-thin  sections
were stained with methylene blue and viewed in a light microscope for  orienta-
tion purposes; eg., to be  certain the specimen was mostly alveolar  lung and not
conducting airway or large blood vessels.  Ultra thin sections, doubly
stained with uranyl acetate and lead citrate, were examined and photographed
in an electron microscope  operated at 50 kV.
   Details regarding morphemetry.  Evaluations were made on micrograph data
banks developed from systematic photography of one nearly perfect section from
each of five randomly selected blocks of alveolar lung per animal (Fig. 1).  A
m-ln-twiiim of 100 photographs was taken per animal using two levels of initial
magnification (X 2000 and  X 4000) which were enlarged to  final magnifications
of X 5000 and X 10000 for  use in point counting, intersect counting and linear
measurements.
   For determinations of volume density and surface density of  the  major tissue
components a plastic overlay (Weibel coherent multipurpose test system of 168
           o
test points ) was placed over the X 5000 micrographs.  The distribution of
points falling on structures and the surface intersects with certain cell sur-
faces were counted and recorded using a Zeiss MOP 3.
   We used Hally's method  of relative standard error as a guide to  the minimum
number of points to count  for an accurate estimate of the volume density of the
cell type expected least often  .  To achieve a relative  standard error of 5%,
if the cell type of interest (eg., alveolar macrophages)  comprises  2%  of the
total cells, a total point count of 19800 is required.  But a 7600  total point
count should accurately reveal distribution when the cell type  is 5% of total
cells.
   Mean caliper diameters  for all nuclear types were calculated based  on the
modeling of cell nuclei as oblate spheroids.  The method of Cruz Orive   was
used for this purpose.  The major and minor axis of each  complete nuclear pro-
file was measured using the Zeiss MOP 3.  With this information, the mean cali-
per diameter (5) for each  type of cell nucleus was estimated by the formula:
                                                         -1
           D—    n
                                               0>-2'0)
                            273

-------
    where:
                 -  1
         -PI
           L«iJ
                                          ,th
M.J
                    the major axis of the i   elliptical profile

                    the minor axis of the iC  elliptical profile

                    the number of profiles measured
j 0.-2.0)
                                        .7T/2

                                        (1  - y2sin2er5/2(l  + 2y2sin29)de
                              v - y? j o

    The above integral for q was evaluated using four point Gaussian Quadrature.
Fig. 1. Low power view of an area of an electron  microscope  grid bearing a thin
section of alveolar lung.  Random photographs  are taken  to develop  the micro-
graph data base for morphometry.   Note that  neither  large blood vessels nor ter-
minal airway are present.  From about 40 grid  views  of this  perfect section of
lung, 15-20 low power micrographs are taken  at random per block of  tissue.
                                274

-------
    Harmonic mean thickness and diffusion capacity were determined using another
plastic overlay (square lattice of 6 horizontal and 8 vertical lines).  This was
placed over the X 10000 micrographs for direct linear measurements (Zeiss MOP
III) of the length of the intercept line crossing from the epithelial to the
endothelial surface and from the endothelial surface to an erythrocyte surface.
    Formulae used for morphometry calculations are listed.  Derivation of for-
mulae and complete discussion of rationale for applying these to analyzing lung
                   q
are given by Weibel .
    Volume Density (Vy.)   Pi/Pt
    Surface Density (JL.) ~ 2 I1/Lt
    Arithmetic Mean Thickness (T-) = Z'P.j/2-I.
    Harmonic Mean Thickness (Th-)1'l  - 3     1
                                  7hi  *"?-iri
    Diffusion Capacity (D, );1   1 + 1 + 1
                            \  \  ^  \
Identification of symbols: P, (number of points falling on item of interest),
P^ (total points); I.- (number of intersections with item); L^ (test line length);
Z (line length in Weibel multipurpose test system based on an equilateral rhom-
bus) ; L-  (length of intercept); D. (diffusion capacity tissue); D  (re plasma);
D  (re erythrocyte),
    Computations were run on  the  Wayne  State University  Computer  (Michigan
Terminal System Operating System), which was also used  for  raw  data  storage,
primary and  secondary parameter  computation, manipulation and storage.   Confi-
dence  limits  for quantitative evaluation procedures were computed  using  the
Student's t-test.

RESULTS
    Histopathology.  Even upon gross inspection of the  lungs  it was  apparent
which  animals had received DE exposure.  Such  lungs were  dusty  grey  in color
with punctate blacker regions.  By light microscopy the most  notable  sign of DE
exposure was  the scattering of pigmented macrophages which were occasionally
clustered at  ends of terminal bronchioles.  Pigmentation  was  also  observed in
lymphatics near bronchiolar-alveolar junctions.  There was an increase in such
pigment deposits related to both  duration and  exposure dose.  By 6 months at
750 yg DE there was beginning fibrosis  in regions of macrophage clusters and
evidence of focal epithelial type 2 cell proliferation.   Occasional  eosinophils
were present  although tissue reaction was not  prominent.  One incidental adenoma
                             275

-------
 occurred in an animal exposed for 24 months to 250 ug DE.   Bronchial and bron-
 chiolar epithelium appeared normal.
    Ultrastructure of  alveolar lung.   The general organization of the lung was
 not appreciably altered.   Alveolar epithelium retained tight junctions.  Two
 cell types  of the alveolar parenchyma (epithelial type 1 cells and macrophages)
 and one type of granulocytic leukocyte (eosinophils)  phagocytized DEP.  The DEP
 was stored  within phagosomes of  alveolar macrophages  (Fig.  2).   Occasionally a
 macrophage  (reactive  monocyte) with  multilobulated nucleus  was noted.

              .                                            •
             4-     • --v... - "•-,•-''-"^^ih... ^V/Stfiifl
Fig. 2.  Alveolar macrophage with DEP contained within phagocytic vesicles.
Also note (arrow) DEP in Epi type 1 cell from 3 month 1500 ug DE exposure.
                            276

-------
Fig. 3. Micrograph from 3 month  750 ug DE exposure has an eosinophil* in air-
space but the lacy constituents  around the specific granules are indicative of
granule degeneration.  Note Epi  1 and Epi 2 cell comprise alveolar wall and that
an endothelial cell  (EC) nucleus shows in right hand capillary.  Mononuclear (M)
cells within this capillary have phagocytized RBC.  Platelets  (arrow) also are
present.  The debris in upper right alveolar space is secretory product from
Epi 2 cells.
                             277

-------
    Eosinophils emigrated  into  the  alveolar airspace (Fig. 3).  Occasionally
 they contained DEP within phagosomes.   Increasing numbers of epithelial type 1
 cells  (Epi 1) contained DEP, illustrating a DE dose dependency (Fig. 2,4,5).
 Neither epithelial type 2 cells  (Epi 2)  nor endothelial cells took up DEP.
 Fig. 4. Epi 1 uptake of DEP (arrow) in a specimen  from 24  month 250  yg DE ex-
 posure.  Interstitial cell* also has DEP.  Symbols:   EC=endothelial  cell, 1=
-interstitium, C=capillary.
    Especially notable at DE exposures greater than 250 yg  was  the  increase in
 cellular composition of the interstitium in contrast  to the  age matched  controls
 (Fig. 2,4).  Fibroblasts, monocytes, eosinophils,  plasma cells and macrophages
 were identified more often within the perivascular and peribronchiolar inter-
 stitium than in alveolar walls.  Interstitial macrophages  sequestered DEP within
 membrane lined vesicles (Fig. 4); more often this  was seen in  the  higher dose/
 duration animals.  DEP was not seen scattered among the fibrillar  components of
 the interstitium.
                              278

-------
                                  Fig. 5. Uptake of
                                  DEP by Epi 1 in
                                  alveolar lung of
                                  a. 2 week expo-
                                  sure to 750 yg
                                  DE/m3.
   Hypertrophy  and  proliferation of Epi 2 cells vas observed in certain alveoli
walls  (Fig.  6)  even by 2  week exposure to 750 yg DE/m3.   With increasing DE
dose/duration,  Epi  2 cell clusters occurred in some alveoli.  Excessive secre-
tory products also  were noted in alveolar spaces (Fig.  3) but were particularly
prominent  after 1500 and  6000 yg DE.
   Physiologic  features of animal groups.  Experimental groups and their age-
matched, generally  concurrent, control groups had approximately equivalent body
weights and  lung volumes.   For example, for the 12 mon  sets, body weights were
1041 + 140,  1044 +  81,  1146 + 121 and 1083 + 85 g while lung volumes were 19.0
+1.9, 21.7+2.4,  23.4+2.4 and 23.7 + 3.4 for respectively the control, 250
yg, 750 yg and  1500 yg DE groups.  These physiologic parameters as well as cer-
tain of the  morphometrically determined lung parameters are age related.
   Harmonic  mean tissue thickness
not change appreciably for control guinea pigs of 2 mon to 52 weeks of age.
The mean value  of these controls was  0.552 + 0.09 um. Significant (p < 0.05)
changes in T.   occurred in the DE exposed animals primarily through 6 mon.
   Arithmetic mean  tissue  thickness (T )  of the air-blood barrier.   Exposure  to
750 yg DE/m3 resulted  in  a significant increase (p < 0.05) in T: t at all sacri-
fice times,  to  1 year  but  T  dropped  from a peak of 2.46+0.43 Mm to 1.67 +
0.18 ym.   In a  similar  way the 1500 yg DE set decreased  from a peak Tt of 2.56
+0.30 at  6 mon to  1.88+0.34 ym after 18 mon exposure.
   Morphometric Diffusion Capacity  (DT ) .  As  expected DT varied  in  relationship
      r                     '        L                 Li
to body weight and age of the  guinea pigs;  for example the 6 wk  old pigs DL =
0.09 + 0.13,  the 19-22 wk pigs DT = 1.66 +  0.18,  the 30-34 wk  pigs  D  = 3.2 +1.1,
     -                          Li                                   -!->
      T,  ) of the air-blood barrier.  The T,   did
279

-------
the 46-48 wk pigs D. - 3.84 +  0.46,  and 60 wk pigs DL = 3.7 + 0.46.  DE expo-
sure appeared to increase D.  about  37%.
Fig. 6. From an animal exposed for  12 months  to 750 pg DE, note proliferation
of Epi 2 cells with their microvilli and  secretory vesicles (EM preparative
artifact leached out the lamellar bodies).  On the right upper alveolar wall is
a transitional epithelial cell* which is  classed as an Epi 1 cell because it
does not show distinctive features  of Epi 2 cells.
   Nuclear diameters.  The mean caliper diameters were computed as specified
previously.  Four thousand nuclei were measured with 2090 controls and 2023
from DE exposed animals.  The dimensions  for  controls were 7.22, 7.17, 7.56,
7.60 and 6.18 ym for Epi 2, endothelial cells, Epi 1, alveolar macrophages and
interstitial cells respectively.  The 250,  750 and 1500 ug DE sets were not
significantly different through 1 year.   However, endothelial cell nuclei
appeared smaller in the 4 mon 6000  and 18 mon 1500 ug DE sets as well as in 21
and 24 mon animals.  From the mean  caliper diameters one can calculate absolute
volumes and numerical density for the parenchymal cell types.
                               280

-------
   Morphometrlc  studies on the alveolar lung parenchyma.   From point and inter-
sect counts the  fractional volumes, surface areas  and  numerical densities were
established for  each cell or tissue compartment  of the alveolar lung.  The
percentage differences between controls and the  750 ug DE/m3  sets are shown in
Tables 1-3.
                  TA8UE1
     EJTOTS OF DIESa DOWJST (750 *V&) EXPOSURE W
     mmoenuc PMMCTER OF cats m ALVEOLA* LUB
         TUl£2
fIFUOCE OF DIESa EXHAUST 1BHALAT10I
   OK 1KTEIBTITII* ( ALVEOLAR
AlVEOLA*
UK
CELL TTO
EttODCLJAL EFI I S

V (R/en3)
S (af-la?)
1 <*/o«J>

EPmejAL TffE 1
S (c»2/wJ>
II IMof)

EPmOJALTWE 2
V (Ju'oi3!
» . j. «.
S (cm/aP)
1 (f/oi3)
PEHBITASE
AFTD750
2 W 3m


• 45.7
•21.9
• 64.6*


•24.1
•31.9
• 44,4*


• 7S.M
•118.4
•H7.1A


•21.2
•22.2



•15.6
• 6.4
- 33.3»


•m.o
^S
QIAWE FROI coma.
»• DEP EXPOSURE
6 in 9w 12 m


0.0
•2S.D»
•4.6


•20
•14.9
-21.4


- 3.6
-28.4
-6.9


•15.1
•23,9
•X5.7A


-10.3
•29.9
•44.4


•46.41
•18.0


- 3.6
- 6.0
• B.CT


0.0
• 1.7
• 44.4*


•_6Z£
• 12.3
«im.nA


EXPOSURE
SPECIFICS
TI!e-0&« ^TI
2>*-750

Jn-750
3»-lSOO
*t-60on
ta-750
6n-1500
>-250
9K-750
12n-750
UK-1500
24W253
• 48.8*

• 26.1
-21.7
U8JI
• 13.3
•146.7*
• 33.3*
• 35.5»
^ 54.5*
t m.tt'
0.0
A*I STUOCKT'I
II THE raiMUIT




PERCEHTASE CHANGE TOB OTTW.
1NTFRSTIT1M ALVEOLAR ntCROPHACES
Vl VCI «C! "»N "«l
• 44.4*

• 29.6
-S8.8*
• 15.8
• 33.3
«HK.?A
• 75.0*
• 60.0
• 23.8
• 5?.;
- 4.3
'T* TUT
DATA AT
• 52.0*

• 21.0
0.0
•146. i*
- 4.2
•104 .7*
0.0
• 16.0
•108.0*
• 75.0*
- 3.3
•177 0*
"•' '^
• 6.b
- 2.2
•^4.;*
- 16.0
•ra.n*
• 2.3
• 12.8
* n.;«
,. 64.1*
- 10.2
EfTAM.ltld> SI0IPI
A • f * 0.05 AHO I
+ 85.1 * 132.0*

» 25.0 - 47.7*
• 77.8 • 19.3
•M.H* *141.1*
• 1.3 - 35.6
*Ti?.s* *m.o*
•isn.n* •is;.8*
• 14.3 *193.0
.175.0* •171.0*
•S67.0» •S95.0*
* 75.0 • 34.1
ICAHT DlfFEKMCES
• P < 0.01
 *<* Sraorr't *r* TOT KTMUIMI •imncMT DI
    niMirr DATA AT A • f « 0.05 AM I • P * 0.01.
   Even  as  brief a DE exposure as 2 weeks  increased tissue volume,  (V^,) , over
the age  matched controls.  For the 750  yg  DE exposure sets, V  increased 36%
after  2  weeks, 46% after 3 mon and remained  near 35% through the year's exposure.
The DE inhalation of 1500 pg resulted in a significant increase (p  -• 0.01)  in
V  to  112%  at 6 mon and still exceeded  controls by 81% after 18 months exposure.
Four months exposure to 6000 yg DE evoked  a  68% increase in V .  In contrast,
250 ug DE sets were not significantly increased at 9 or even 24 mon exposures.
   Subdivision of the tissue into the individual cellular components reveals
some interesting changes.  For the 750  pg  DE sets, volume density (V), surface
density  (S) and numerical density (N) were significantly increased  for Epi  2
cells  but a time related reduction occurred  (Table 1).  A similar though not as
spectacular change occurred with endothelial cells.  A different pattern of
response by interstitial cells occurred with the early increase followed by re-
duced  volume density from 3 through 9 mon  but significantly increased by 12 mon
exposure (Table 2).  The non-cellular interstitium (NCI) was not significantly
changed  during the exposure.
                               281

-------
    Considering the interstitial responses to DE further, the 1500  yg DE sets
 showed  increased volume  density after  6  mon followed  by a time related reduction,
 that after 18 mon exposure still exceeded control values (Table 2).   The 6000
 Ug DE/m3  exposure for 4  mon resulted in  a 69% increase in total interstitium
 (V  ) with the significant change being  an increase of 146% in the fractional
 volume  of cellular interstitium (VGI).   When nuclear  diameters were  taken into
 account,  this value translated into a  214% increase in numerical density of
 interstitial cells.
                                        TMU3
                               MRMocntc CHARMS m MJOR iwe cau:
                              OVMVItSOR V CORTMLS MTH THC 750 y« K SITS
ALVEOLAR LUNh
CUTNCHAl TTH 2
TOTAL •>. I 10*

AK. VDL.vM*'
AK. SURFACE AREApIT*
% TOTAL LUB COLS

% All. SURFACE COmCD

IRTCBT1TIAL CELLS
TOTAL «0. 1 10*

AK. WL../JT1
% TOTAL LURE CELLS

ALIEOUR RAODPHACCS
TOIAl W. t 10*

AK. ML../**3

% TOTAL LUNG CELLS

EWmKLIAL "t'«
TOTAL W. 1 10*
AK. fOL. • pfT
j
AK. SURFACE AftEA.pJI*
% TOTAL LUB CELLS

EPITHELIAL TTK 1
TOTAL RO. I 10*
AK. WL.V
AK. SURFACE AREApPT
% TOTAL LUB CELLS

XALV. SURFACE COHKED

Z IK
C K

11.4 .
51,1*
175*115 .
47**I01*
49* 1* .
107*19*
11.7
11.5
7
11.1

51.3 .
115.7*
417*113 .
K.I
35.5

4
9.3*
1175*535
535*591
t.(
2.9

*4.1
107.8*
54»* at .
470*572
573* 7)
41.4
33.7

"•° 26.3*
1361*117
641* 97
155*154'
11.5
1.4
93
M.9
3 P
C

11.6

1113*411
50* I)
12.3

6.1


46.5

419*124
30.9


1.1

1000*427

5.1


55.3
597*115
559*134
35.1


21.3
1502*325
61*4150
14.2

93.2

ON
K


37.4*
1B7.510
113* 72

20.1

13.4


41.6
473* 95

27.0


4.**

2391*1131
— ~— —
2.6

75.1*
533*1*2
6U*I4*

41.1

14.1*
2*24*492*
730* 99

7.1

16.6
C

29.0

9*6*239
95* 21
17.3

5.0


49.1

4*4*43
29.7


9.9

7*t*tlO

5.9


64.9
4*5* 61
•40*110
31.1


13.9
1!3S*2M
170* 92
1.3

90.1

•UN
OE


2*.l
1007*2*9
M* J>

17.2

1.4


41.5
445* 13

21. «


7.3

1055*3*4

4.1

61.5
445* 14
630* (9*

44.0

11.2
2*10*516'
740* 11

7.2

91.6
9 1
c

32.2

•5*216
W* IS
20.3

10.*


43.1

47t* 21
27.2


4.2

•1*344

2.*


•9.9
47** 21
77T* 62
44.0


9.3
3140*305
747* 60
5.9

19.4

*JH
DE


63.4*
647*137
105* 20

25.9

9.9


53.3
544*141

21.1


12.3

1301H11

5.0

10L1*
374* 39*
932* 67*

41.7

13.5
234«*t35*
952*62*

5.6

90.1
C

25.5

941*272
106*24
20.S

11.4


30.2

391*96
24.3


4.9

1203*503

5.6


52.6
525* 73
137* M
42.4


1.9
2921*513
Ul* 95
7.2

98.C

wn
OE


$l.y*
760*113
119. »

25.3

12.S


45,1*
545*163

22.1


11.7*

941*135

9.2

zui
3)3.34*
717.71

3S.I

. «•'*
2055.1211
63S*f(

(.3

87.S
A,B
  Stydwt'l 'f «lt 
-------
   The fractional volume and numerical  density  of  alveolar  macrophages  were ele-
vated over control values after most  DE exposure doses  and  durations  (Table 2).
The greatest increases occurred after 1500  and  6000  ug  DE exposures.
   Selected absolute changes in morphometric  parameters for the  responses  of
parenchymal cells to the 750 ug DE  challenge  are given  in Table  3.  A striking
feature for all cell types was the  increased  cellularity.   Since comparisons
were with age-matched controls these  findings appear DE and not  age related
changes although the latter are documented  in Table  3.   During DE exposure Epi 2
cells doubled or tripled the control  numbers  while their cell  volumes fluctuated
and they generally covered a greater  surface  area  of the alveolar wall  than con-
trol cells.  The total number of  Epi  1  cells/cm3 was not reduced by this expo-
sure regimen.  Interstitial cells,  a  heterogeneous group of different cell types,
were particularly increased after the 2 wk  exposure; eg., about  116 million/cm3
occurred to contrast with 51 million/cm3 in the matched control.  Endothelial
cells were increased in number and  had  significantly smaller average  cell  vol-
umes.
   Translation of the numerical density data  to absolute pulmonary cell numbers
per animal was done.  Graphic display of certain data from  the 750 yg sets  per-
mits contrast of controls and DE  exposures  for  1 year (Tig.  7).
   CHANCES IN TOTAL ALVEOLAR LUNG CELLS
    IN GUINEA PIGS EXPOSED CHRONICALLY
              TO 750ug DE/m?
z
(J •
*o
                                           Fig.  7. Age-related  changes are
                                           shown in  the  control curves for each
                                           cell  type;  eg. a = Epi  1, b = Epi  2,
                                           c = endothelial cells  (EC).  Effects
                                           of DE exposure are designated in the
                                           other 3 curves; eg.  d   Epi 1, e
                                           Epi 2 and f = endothelial cells  (EC).
              EXPOSURE MONTHS
                             283

-------
About 230 million Epi 1 cells were  in the guinea pig lungs regardless of age up
to one year.  The DE exposure did not reduce that number significantly.  Age
related changes occurred in Epi  2 and endothelial cells.  The youngest guinea
pigs had in their lungs about 220 million Epi 2 and 760 million endothelial
cells which numbers reached 660  and 1200  million respectively by 9-10 mon of
age but declined somewhat by 12  mon.   Cellularity for these two cell types was
increased 2 to 3 fold over controls at most  times; for example after 9 mon ex-
posure there were about 1200 and 1900 million Epi 2 and endothelial cells re-
spectively.  In certain higher 0E exposures, namely 1500 yg for 3 mon and 6000
pg DE for 4 mon,  the absolute numbers of Epi 1 were statistically reduced over
control values.  The percentage  change was -60% and -24% respectively (Table 4).

Comparison of pulmonary responses to  possibly equivalent DE doses
   Five sets of the DE animal series  had  similar total dose-duration exposures
although the actual doses and exposure intervals differed.   If one assumes
these paired sets of animals had equivalent  burdens of DEP  to respond to and
handle, one might expect similar magnitudes  of tissue responses in the paired
exposures.  Comparative data in  terms of  percentage change  in numerical density
are given in Table 4.  It is apparent,  if the macrophage responses are dis-
counted, that in 4 of 5 comparison  sets the  shorter duration higher doses eli-
cited statistically significant  (p  <  0.05) and greater tissue responses as in-
dicated by underlined values in  Table 4.   The paired sets of 18m - 1500 ug DE
and 4m - 6000 pg sets had a mixed pattern of increased cellularity.

Tolerance of the lung to the 250 pg DE/m   inhalation
   This concentration seemed relatively ineffective as a stimulus to elicite
appreciable alveolar tissue responses. After 9 mon exposure there were approxi-
mately 30S increases (p < 0.05)  in  Epi type  1, Epi type 2,  endothelial cells and
the interstitium over concurrent age matched controls (Tables 2,4).  By 24 mon
exposure, with the exceptions of the alveolar macrophage and Epi type 1 cell
responses, there were no significant differences from controls.  The alveolar
macrophage responses appeared to be reduced  as the exposure continued.  On the
other hand, the Epi type 1 cell  responses were maintained at an increased
numerical density near 30%.
   Throughout the inhalation exposure (from  3-24 mon), DEP was observed within
vacuoles of certain Epi type 1 cells (Fig. 4).  Interstitial macrophages also
sequestered phagosomal DEP  (Fig. 4).
                             284

-------
                                            TABLE 4
                        EVIDENCE OF ADAPTATION  IN LONGER DURATION EXPOSURES
                                   TO DIESEL EXHAUST
   EXPOSURE                       PERCENTAGE  CHANGE IN CELLS/OP:
  CONDITIONS                "E EXPOSURES COMPARED TO AGE MATCHED CONTROLS
               ENDOTHELIAL   EPITHELIAL       EPITHELIAL      INTERSTITIAL      ALVEOLAR
  TIME-DEuG    '    CELLS       TYPE 1           TYPE 2            CELLS        MACROPHAGES

   9«-250     +32.0         + 31.2+ 32.3+ 2.5+152Ji
   3w-750            + 35.8           + 33.8          +101.1B         +  1.5            + 17.1*

  21«-250     -12.2         + 31.5           + 10.9          - 8.5          + 32.9
   9«-750            * H5.6           + as.2          + 96.9s         + 23.7A           +192.8A

   5M-750     + 5.5         - 19.1           -  7.6          -16.3          - 26.3
   3H-1500           - 13.6B         - 60.1B          * 97.8B         -  2.6            * 19.3B

  12*-750     + 82.8        + 12.7           +101.2          +52.0          +171.0
   6»-1500           +188.0*         +107.2*          +255.2A         +268.5*           +3W.1A

  18w-1500    + 35.1B       +107^*          +113.8*         +65.0*         +591.1*
   itn-6000	+ 60.9*	- 23.8B          +115.8B	+213.9*	+139.3*

        A,»  STUDENT'S "T* TEST  ESTABLISHED SIGNIFICANT DIFFERENCES IN THE PRIMARY DATA
        AT A • p < 0.05 AND B • P * 0.01
DISCUSSION

    This communication  chiefly documents the lung's quantitative responses  to a

chronic burden of DEP.   Exposures  to  250  pg DE/m3  resulted in  little  or no

change in  morphometric  parameters  through 24 mon  exposure.
                                 285

-------
   Of course, DEP was seen within  the  alveolar nacrophages and Epl 1 cells in
the 230 yg DE seta but was more prevalent  at higher DE doses.  However, regard-
less of the carrier cell type, DEP remained sequestered within membrane-enclos-
ing vesicles, contrary to the report of Wiester's group that DEP was in the
macrophagic cytoplasm  .  Standard histologlc procedure and light microscopy
employed by Wiester's group are inadequate to resolve membrane detail of the
phagolysosomes.
   The finding that, the phagocytlzed DEP does not escape into the cytoplasmlc
milieu is a significant difference from events that occur when cytotoxic silica
                                                                   13—14
particles, asbestos fibers and fly ash are taken up by macrophages     .  More-
over if DEP was cytotoxlc to alveolar  macrophages,  one would expect their ab-
solute numbers to decrease whereas they increased;  eg., after 18 mon 1500 pg
DE/m3 the 5913 Increase over controls  (Table 4)  reflects 1300 mill ion macro-
phages versus 210 million in the control.   However, there may be some degree
of functional impairment for phagocytosis  according to Dr. Shan-te Chen's work,
in our laboratory, using broncholavaged macrophages '
   The question of whether or not  epithelial type 1 cells are injured by their
uptake of DEP at these concentrations  is not completely resolved.   Several
points support the view that the contained DEP is not  cytotoxlc.   1.   The epi-
thelial DEP remained within membrane-bounded vesicles.   2.  Tight  epithelial
cell junctions were maintained.  3. ' Neither the numerical density nor absolute
number of Epl 1 cells were reduced over age matched controls except in the 3
mon 1500 and 4 mon 6000 ug/m3 sets.
   The appearance of DEP within Epl 1  cells is probably a sign of  a particulate
overload which resident alveolar macrophages cannot adequately handle.   Thus
Epi 1 uptake of particles may represent a  second line  of defense against par-
tlculates.  The mechanisms by which Epi type 1 cells take up DEP is not yet
clear but probably involves both macropinocytosis of individual DEP (Fig.  5)
and phagocytosis of DEP aggregates.  More  Epi type  1 cells contained DEP as
dose and duration of the exposure  increased but  the actual magnitude of the
response needs to be established.   Also, the fate of the Epi type  1 cells con-
taining large vacuoles of DEP is not known.   Nor is it  known whether the DEP
is released from the Epi type 1 cells  to the interstitium.
                              286

-------
   Ultrastructural changes occurred as early as 2 weeks after exposure to 750
Ug DE/m3 illustrating how dynamic and responsive the normal lung can be to an
environmental stress.
   Thickened alveolar septa were noted for exposures > 250 ug DE/m3.  Increased
thickness was due to at least 2 components; 1). the increased mass of the inter-
stitium, largely the result of increased numbers of interstitial cells some of
which were macrophages and eosinophils and 2). increased numerical density and
hypertrophy of Epi type 2 cells.  Although arithmetic and harmonic mean tissue
thicknesses of the air-blood barrier were occasionally increased the morpho-
metric diffusion capacity was not adversely affected.
   The estimates derived from morphometric analysis lead us to suggest the fol-
lowing conclusions:
   1.  The 250 ug DE/m3 elicits insignificant tissue changes, except for alveo-
                <
lar macrophage uptake of DE, through 24 mon exposure.
   2.  The lung responds rapidly to DE challenges > 250 ug DE/m3 by increased
cellularity of all types of alveolar lung cells.
   3.  Ultrastructural changes are DE concentration dependent for a single
duration of exposure but are not for roughly equivalent doses of DE experienced
for different durations.
   4.  Partial adaptation occurs in normal guinea pigs during chronic exposure
to DE illustrating the normal lung's potential for repair and defense.

ACKNOWLEDGMENTS
   This work was partially supported by General Motors Research Laboratories,
Warren, MI and the Bargman Foundation Laboratory for Cell and Molecular Research,
Wayne State University.  The authors express their sincere gratitude for the
expert technical assistance of C. Becker, R. Blakeley, L. Dang, S. Khan, R.
Kraemer, D. Prokopchak, M. Potts and J. Thompson.

REFERENCES
1. Moore, W. , Orthoefer, J.G., Burlcart, J.K. and Malanchuk, M. (1978) in Proc.
   71st Annual Meeting of the Air Pollution Control Assoc., Houston, TX.
2. Springer, K.J. and Baines, T.M. (1977) Society of Automotive Engineers Paper
   770818.  Detroit, MI.
3. Breslin, J.A., Strazisar, A.J. and Stein, R.L. (1976) in R.I. 8141r.  U.S.
   Bureau of Mines, Washington, DC.
4. Barnhart, M.I., Chen, S. and Puro, H. (1980) in Health Effects of Diesel
   Engine Emissions:  Proc. Internat. Symposium.  Center for Environ. Research
   Information EPA, Cincinnati, OH  45268, pp. 649-672.
                            ~28/

-------
 5.  Chen,  S., Weller,  M.A.  and Barnhart,  M.I. (1980) Scanning Electron Micro-
    scopy.   3,  327-338.
 6.  Barnhart, M.I.,  Chen,  S.f  Salley,  S.O.  and Puro, H. (1981) J. App. Toxicol.
    1,  88-103.
 7.  Schreck, R.M., Soderholm,  S.C.,  Chan, T.L., Smiler, K.L. and D'Arcy, J.B.
    (1981) J. Appl.  Toxicol. 1,  67-76.
 8.  Scherle, W.F.  (1970) Mikroskopie 26,  57-60.
 9.  Weibel,  E.R.  and Bolender, R.P.  (1973)  in Electron Microscopy Morphometry.
    Principles  and Techniques  of Electron Microscopy, Hyatt, M.A. ed., Van
    Nostrand, Reinhold, New York,  pp.  237-296.
10.  Rally, A.D.  (1964) Q. J. Microsc.  Sci.  105, 503-508.
11.  Cruz Orive, L.M. (1976) J. Microsc. 107,  235-253.
12.  Wiester, M.J., Iltis, R. and Moore, W.  (1980)  Environ.  Health 22, 285-297.
13.  Allison, A.C.  (1975) in Air  Pollution and the  Lung, Aharonson, E.F., Ben-
    David, A. and Klingberg, M.A.  eds., Wiley and  Sons, New York, pp. 114-134.
14.  Aranyi,  C., Miller, F.J.,  Andres,  S., et  al.  (1979) Environ.  Res. 20,  14-
    23.
15.  Chen,  S., Weller, M.A.  and Barnhart,  M.I.  (1981)  in Abstract  Book, EPA 1981
    Diesel Emissions Symposium,  Raleigh,  NC.
                            288

-------
       BIOCHEMICAL ALTERATIONS IN BRONCHOPULMONARY LAVAGE FLUID AFTER
         INTRATRACHEAL ADMINISTRATION OF DIESEL PARTICIPATES TO RATS

                                     by

               C.D.  Eskelson, M.  Chvapil, E. Barker, J.A. Owen
             Department of Surgery, Division of Surgical Biology
                University of Arizona Health Sciences Center
                           Tucson, Arizona  85724

                                 J.J. Vostal
                        Biomedical Science Department
                    General Motors Research Laboratories
                           Warren, Michigan  48090


     Male Sprague Dawley rats weighing 180-200 g were given intratracheally
5 mg of diesel  particulates (DP)  in 0.75 ml saline.  Control rats were given
saline.  Five days after administering the DP and three hours before the rats
were sacrificed they were injected with a pulse of 20 uC of ll*C-acetate.

     The lungs  were intubated before they were removed and lavaged 3 times
with 5 ml of saline.  The combined lavage fluids were lyophylized.  The
lipids were extracted from the lyophylized lavage fluid with chloroform:
methanol (2:1)  and its cholesterol (C) and phospholipid (PL) content
determined.  The radioactivity incorporated into the lipids was determined by
separating the  lipids by a TLC method and scraping the spots of interest into
counting vials.

     Pulmonary  lavage fluid from the rats given DP contained 4 times more PL,
C, and protein  than in control rats (Table 1).  Total radioactivity
incorporated into lecithin was twice that of controls and was 3.5 times
greater than the radioactivity found in the other PL studied.  The lavaged
lungs from the  control and experimental rats were lyophylized and homogenized
in a chloroform:methanol (2:1) solution.  The lipid analysis showed no
difference in the PL and C levels between the control and experimental lungs
(Table 2).

     The fatty  acid (FA) profile of the lavage fluid determined by a GLC
method indicated a three-fold increase in palmitic acid and arachidonic acid.
Stearic, oleic  and linoleic acids were not significantly altered (Table 3).

     These studies imply that the lipid loading observed in lungs exposed to
5 mg of DP for  5 days are a result of increased deposition of pulmonary sur-
factant (extracellular lipids) and are not a result of intracellular lipids.

-------
              Table 1.  Analysis of Pulmonary Lavage Fluid from
         Rats Intratracheally Exposed to 5 mg of Diesel Particulate
Phospholipids

Experimental
Control
Student's t
mg
1.90
0.49
7.82
SD
0.48
0.18
P < .001
Cholesterol
mg SO
.539 .040
.151 .081
6.37 P < .001
Protein
mg
8.47
2.09
8.25
SD
0.33
1.04
P < .001
Results expressed as mg of lipids in the total  lavage fluid
        Table 2.  Analysis of Lavaged Lungs from Rats Intratracheally
                    Exposed to 5 mg of Diesel  Particulate
Experimental
Control
Student's t
Phospholipids
mg SD
34.9 1.34
31.30 10.6
.453 NS
Cholesterol
mg
15.9
13.1
1.18
SD
1.02
3.23
NS
Results expressed as mg of lipid per lung
                                    290

-------
           Table  3.   Fatty Acid Profile from Lung Lavage Fluid of
                   Rats Exposed to 5 mg Diesel Particulate
Experimental
SD
Control
SD
Student's t

C16
1.77
.07
0.541
0.195
8.50
P < .001
C18
0.115
.033
.097
.103
NS

C18:l
0.152
.016
0.152
.015
NS

C18:2
0.152
.030
0.225
.322
NS

C20:4
0.104
.015
.030
.021
4.61
P < .01
Results expressed as mg of the fatty acid methyl ester per total lavage fluid
                                      291

-------
                     LIPID CHANGES IN LUNG OF RATS AFTER
             INTRATRACHEAL ADMINISTRATION OF DIESEL PARTICULATES

                                     by

               C.D.  Eskelson, E.  Barker, M.  Chvapil, J.A. Owen
             Department of Surgery, Division of Surgical Biology
                University of Arizona Health Sciences Center
                           Tucson, Arizona  85724

                                 J.J. Vostal
                        Biomedical Science Department
                    General Motors Research  Laboratories
                           Warren, Michigan   48090


     Lung, liver and serum from 180-200 g male Sprague Dawley rats were
analyzed for various lipids 5 days after the rats were given intertracheally
in a saline solution 1  mg of diesel particulates (DP).  The lipogenic
activity occurring in these rats  was studied by giving i.p. 20 uC °f
:i*C-acetate one hour before they  were sacrificed.

     Phospholipids and  cholesterol content of the lungs were significantly
increased while that of triacylglycerols were not changes significantly (see
Table 1).

     Contrariwise to pulmonary lipids, hepatic phospholipids and cholesterol
levels were decreased in rats exposed to diesel dust while hepatic triacyl-
glycerol (TG) levels were not significantly  altered.  Accompanying the loss
of hepatic lipids are an increased phospholipid, cholesterol and TG specific
activity indicating increased hepatic lipogenesis.  To determine if the loss
of hepatic lipids were  due to their being mobilized to the serum from the
liver serum lipids were determined and were  found not to be significantly
altered.  However, the  specific activity of serum phospholipids, cholesterol
and triacylglycerols were all significantly  increased in rats intratracheally
given the diesel particulates.  A corresponding doubling of radioactivity in
pulmonary phospholipids and cholesterol was  also detected in these animals.

     The results obtained here are similar to those reported earlier for rats
intratracheally given silica dust (1-3) and  suggest that a particulate insult
to the lungs results in the lung  producing lipotrophic factors which
stimulate the liver to  increase lipogenesis  and lipid export to the blood.
The lung in turn picks  up the lipids from the serum arid remodels them to meet
pulmonary lipid need.  To further study this hypothesis, rat hepatocytes were
isolated and incubated  in a pH 7.0 phosphate buffer containing 2 uC
                                     292

-------
1[*C-acetate and several  cofactors.  To this hepatocyte suspension was added
lung slices and the system thence incubated for 2 hours at 37°C.

     Phosphatidyl  choline (PC) was isolated from each sample by TLC and the
PC spots from each TLC scraped into counting vials.  The amount of
radioactivity incorporated into the PC spots of the 5 samples for each
experimental manipulation was averaged and is presented in Figure 1.

     This study clearly demonstrated increased lipogenesis above that of the
sum of lung slice PCgenesis and hepatocyte PCgenesis.  The results from the
in vitro studies strongly support the concept that a pulmonary lipotropic
factor exists which stimulates lipgenesis in the liver and that these de novo
synthesized lipids are utilized in part to maintain lipid homeostasis TrT the
lung.
        Table 1.  Pulmonary Lipids from Rats Intratracheally Exposed
                        to 1 mg of Diesel Particulate
Phospholipids
ing SD
Experimental
Control
Student's t
51.7
30.8
6.15
4.64
4.66
P < .001
Cholesterol
mg
11.95
8.27
5.15
SD
1.11
0.63
P < .005
Triacylglycergls
~mg SD~
20.00
24.52
2.32
3.15
1.24
NS
Results expressed as mg of lipids per lung
        Table 2.  Hepatic Lipids from Rats Intractracheally Exposed
                        to 1 mg of Diesel Particulate
                               (mg/g of liver)
Experimental
Control
Phospholipids
riig SD
27.9 1.9
31.3 1.7
Cholesterol
mg
3.36
4.02
SD
0.10
0.11
Triacylglycergls
~nig SD~
8.57
9.07
1.43
0.88
Student's t
2.52
P < .05
8.60
.001
.53
                                    293

-------
      Figure 1.
                                  Lecithin Formation in Hepatocytes
                                  and Lung Slices from 14C-Acetate
                                           T P< 05
                             20 T
                             15-
                          X

                          Q.
                          0
                             10-
                              5-























g
2
Q
+
en
3
|

a.
a>
r
T








_c
03

®
s
,
1
Q
tf)
CD
1?
o
I
T






o
CO
Q
a>
e
s
0)
ii
(/3
0)
C
3
_J
C
1
I
—
01
o
0
I
X


p<.05
,„ F = 23.44
i





O
co
o
ID
o
CO
a
c
_j
+
01
u
o
3.
X


ul
2
I

o
i
Q
+
en
2
8
1
X
~x_







o
CO
a
T
uT1
O
Q.
^~
0
en
2

a,

-------
                  BIOAVAILABILITY OF DIESEL  PARTICLE BOUND
                        [G-3H-] BENZO(a)PYRENE  (3H-BP)
                      AFTER INTRATRACHEAL  INSTILLATION

                                      by

                  P.K. Medda, Sukla Dutta  and Saradindu Dutta
                   Wayne State University  School of Medicine
                           Detroit, Michigan    48201


     Recently, we have investigated whether  pulmonary prostaglandin dehydro-
genase (PGDH) activity is affected by long-term exposure to low doses of
diesel exhaust (1) and whether high doses  (6.0 mg/m3) for a short period (2-
8 weeks) can affect such biochemical functions  as (a) blood methemoglobln
level, (b) reduced glutathione  levels, (c)  angiotensin coverting enzyme
activity, and (d) mixed function oxidase activity (2, 3).  In general these
studies have shown that as far as the above  set of biochemical parameters is
concerned, there exists no particular adverse effect of the diesel exhaust in
the exposed animal.  Because of these findings  and the observation made by
Siak et_ ^1_., (4) that simulated biological fluids elute no significant mut-
agenic activity from the diesel particles, we have contended that the many
polycyclic hydrocarbons, such as benzo(a)pyrene (BP) which are known to be
present in the diesel particles (DP), probably  remain unavailable to the
pulmonary tissue.  In order to provide support for this contention attempts
have been made to determine the bioavailability of 3H-BP following intratra-
cheal administration of this agent bound to  diesel particles in an albumin-
saline suspension.

     In order to determine the bioavailability  of benzo(a)pyrene as bound to
diesel particles, 3H-BP (120 uCi/1.79yg/1.0 ml ethanol) was allowed to bind
diesel exhaust particles by adsorption.  After  removal of free 3H-BP by re-
peated suspension and centrifugation, nearly 90-95 yCi of 3H-BP remained
bound to 1 mg of DP.  Tightness of the binding was tested by continuously
washing 125 ug of 3H-BP bound DP with 6.0  ml either Krebs-Henseleit (K-H)
solution or K-H with 3.2% albumin solution or DMSO or dichloromethane for one
hour.  Results showed the least dissociation in K-H solution (2.1%) and the
highest in the presence of dicholoromethane  (72%).  For the bioavailability
experiments, female guinea pigs (Hartley) were  lightly anesthetized with ether
in preparation ^or instillation of 1 mg labelled DP.  When animals were appro-
priately anesthetized, diesel particles as suspended in 0.2 ml K-H solution
containing 3.2% albumin was introduced slowly into the trachea over a period
of 10-15 minutes by inserting a PE10 polyethylene tubing through the tracheal
ring and pushing deep inside the broncheal tree.  After intratracheal instil-
lation of labelled particles guinea pigs were divided into two groups
     Five guinea pigs of one group were immediately sacrificed within 20-25
                                    295

-------
minutes, while six guinea  pigs  belonging  to  the  second group were moved to
Nalgene(R) metabolic cages,  held  there  individually for collection of urine
and feces for 48 hours and then sacrificed for excision of selected organs
such as  liver, kidney, intestine  and  collection  of blood for measurement of
radioactivity.

     Results of these studies showed  that intratracheal  instillation of 1  mg
labelled DP did not produce  any obvious symptoms  in these animals.   Further-
more, except for individual  difference, no abnormality was noted in feeding
and excretory profiles of these guinea  pigs.  At  autopsy, lobar localization
of DP was clearly visible as distinct patches of  black marks of 1-2 cm
diameters.  Percentages of radioactivity  retained by guinea pig lungs  immedi-
ately after intratracheal instillation  of DP bound to 3H-BP, when calculated
on the basis of the administered  dose,  showed wide variability in that 77-95%
of the theoritical radioactivity  (90-95 uCi 3H-B(a)P/mg  DP) was actually mea-
sured in the lungs of these  guinea pigs that were sacrificed within 20-25
minutes of instillation.  As observed by  Henry and Kaufmann (5),  in the pre-
sent study also the discrepancy between the amount of dose actually intended
for delivery and the amount  actually  measured in  the lungs after intratracheal
instillation could not be accounted for any loss  due to  regurgitation  of the
suspension.

     Percentages of radioactivity retained by six guinea  pig lungs  at  48 hours
following intratracheal instillation  of DP bound  3H-BP, when calculated on  the
basis of the administered dose, showed a  mean disapperance of 42% ±6 of radio-
activity during the 48 hour  time  period.  However,  the observation  that actual
delivered dose was somewhat  less  than the dose intended for instillation meant
that the lungs might have lost much more  radioactivity during the 48 hours
time interval.  It was also noted that at 48 hours  radioactivity  had distri-
buted widely in that all the organs studied such  as  liver,  kidney and  intes-
tine showed about 1-2% as much 3H-BP  content per  gram in  comparison  to  the
radioactivity retained per gram of lung tissue.   Furthermore,  during 48 hours
25% +. 3 of the radioactivity was excreted in urine  and feces  in these  animals.

     In conclusion, these studies show that 3H-BP  dissociates  from  the labelled
diesel  particles upon instillation in the lungs and  appears  in  urine and
feces.   This rapid dissociation of 3H-BP  from the  diesel  particles  implies
that by the existing method of labelling  of DP by  adsorption  with 3H-BP we
may not have simulated the forces by which benzo(a)pyrene  binds  to diesel
particles under engine condition.
                                     296

-------
                                 REFERENCES
1.    Chaudhari,  A.,  R.G.  Farrer and S. Dutta.   1981.   Effect of exposure
      of diesel  exhaust of pulmonary prostaglandin dehydrogenase (PGDH)
      activity.   J.  Appl.  Toxicol. 1:  132-134.

2.    Chaudhari,  A.  and S.  Dutta.    1982.   Alteration in tissue glutathione
      and angiotensin converting enzyme due to inhalation of diesel  exhaust.
      J. Toxicol.  Envir.  Heal.  (In Press).

3.    Navarro,  C.,  J. Charboneau and R. McCauley.    1981.   The effect of
      in vivo  exposure to diesel  exhaust of rat hepatic and pulmonary
      microsomal activities.  J.  Appl. Toxicol. 1:  124-127.

4.    Siak, J., T.L.  Chaiti  and P.  Lee.   1979..   Diesel particulate extracts
      in bacterial  test system.  Presented at the U.S. Environmental
      Protection Agency Symposium on Health Effects of Diesel Engine
      Emissions.  Cincinnati, Ohio.

5.    Henry, M.C. and D.G.  Kaufman.   1973.   Clearance of benzo(a)pyrene
      from hamster lungs  after administration of coated particles.  J.
      Nat. Cane. Inst. 51:  1961-1964.
                                    297

-------
           THE POTENTIAL FOR AROMATIC HYDROXYLASE INDUCTION
                  IN THE LUNG BY INHALED DIESEL PARTICLES
                            K. C. Chen, and J. J. Vestal
                          Bio medical Science Department
                       General Motors Research Laboratories
                                 Warren, MI 48090


Diesel exhaust particles contain trace amounts of a wide spectrum of polyaromatic
hydrocarbons (PAH) adsorbed on the surface and, when extracted by an organic solvent,
produce  mutagenic  effects  in  short-term  microbial  laboratory  tests.   Since the
mutagenic  or carcinogenic  effects of  hydrocarbons are  frequently initiated  by  a
metabolite  rather  than  by  the  parent  molecule,  and  since the  activity  of the
metabolizing enzymes can be easily increased by preceding administration of powerful
inducers, the enzyme induction could theoretically predetermine the  potential for the
adverse health effects of inhaled diesel exhaust emissions.

The  effects  of  long-term inhalation of diluted diesel exhaust  on aryl hydrocarbon
activity  (AHH)  and cytochrome P450  content  in  lung and  liver microsomes  were
investigated in male Fischer-344  rats and  compared with repeated parenteral adminis-
tration of organic solvent  extracts of hydrocarbon adsorbed on the diesel participate
surface during the combustion process.   No significant effects of long-term inhalation
exposure were observed in live£3microsomal AHH activity.  The animals were exposed
to concentrations of 750 yg  m   or 1500 ug m~  of diesel particulates from a  5.7 GM
diesel engine at 20 hours per day,  5-1/2 days per week for up to 9 months, or treated by
repeated IP injections of diesel particulate extract dissolved in corn oil, from the same
engine at the several  dose levels for 4 days.  A decrease  in lung microsomal  AHH
activity  was found in rats following a, months of exposure  to diesel exhaust at the
particulate concentration of 1500  yg m   .In contrast, 1.4- to 9-fold increases in AHH
activity  were observed in liver and lung microsomes of rats pretreated by intraperi-
toneal doses 10-15 times larger (25-125 mg/kg BW) than the most conservative estimate
of the deposited lung burden [J. Appl. Tox., 1(2):27,19811.

Since the intraperitoneal injection of diesel particle extract may not fully represent the
activity  of PAH deposited on  the  inhaled diesel particles  in  the respiratory  airways,
direct  intratracheal instillation  (ITI) of  various  doses of  extract  was  used,  and
microsomal enzyme induction  was investigated in the lung as well as  in the liver in
order to detect the local and systemic  response to  hydrocarbons deposited in the
respiratory system.  Diesel particulate extract or  pure benzofa] pyrene,  dissolved in a
gelatin-saline solution  and used as a reference compound, were administered by ITI at
several  dose levels.  The  results show that direct  intratracheal administration of the
diesel particle extract  required doses as high as 6 mg/kg BW before the activity of the
induced enzyme in the lung was barely doubled (Figure 1). The induction was slow and
                                       298

-------
occurred selectively in lung only (Figure 2), indicating that diesel participate extract
probably does not absorb easily into the lung circulation, and is not distributed to other
organs.   The  data suggest  that the absence of  AHH activity induction in rat lung
exposed to diesel exhaust is due to the inavailability of hydrocarbons for distribution in
the body and insufficient  quantities for enzyme induction.  All data seem  to indicate
that the inhaled diesel particles would not be capable of inducing aromatic hydroxylase
in the lung unless the total deposited dose in the lung reaches approximately 6-8 mg of
the  particle  extract per  kilogram  of  body weight.   Since the  extractable portion
represents only 10-15% of  the total participate mass, the required pulmonary deposits of
diesel particles  in a 70 kg  man would be excessive to become a significant step in
promotion of a potential neoplastic process.
                     = 500
                     9
                     2
                     a-400
                     o>
                   >E
                   1 £300
                   « £
                   <\
                     o 200
                               Lung
                                     N =4
   CO

   ~Q

   a.
                       100
                           0
                     468
                     Dose (mg/kg)
                                                  DP- Ext
                                                  T—7
                                                        10   12
                               Liver
  .£  1000
  «
  o
  a  800
  en
>£
1 ^   600
                                       N =4
                                                    -Ext= 12 mg/kg
                                                B(a) = 5 mg/kg
^
o
m
~o
£
Q.
400

200

0
—

-

i i i i I I
               24    48   72    96   120
                Hours After Administration
                                                              144
                                        299

-------
         XENOBIOTIC METABOLIZING  ENZYME LEVELS  IN MICE  EXPOSED  TO
                 DIESEL  EXHAUST OR  DIESEL  EXHAUST EXTRACT
                          William Bruce Peirano
                    Health Effects Research Laboratory
                   U.S.  Environmental Protection Agency
                             Cincinnati, Ohio
     Diesel exhaust  (DE) contains  organics,  specifically polycyclic ar-
omatic hydrocarbons  (PAH),  which are mutagenicU)  and  also potentially
carcinogenic.  These PAH, for the most part,  must be metabolized via the
xenobiotic metabolizing enzymes  to  become active mutagens and/or carcino-
gens.  The primary functions of  these enzymes  are  to  detoxify and/or to
make these PAH more readily excretable.  Unfortunately, a  percentage of the
metabolites formed are reactive electrophiles, which can  bind to proteins,
RNA and/or DNA to cause mutations and/or cancer.   Moreover, the body can
respond to an environmental PAH assault by  increasing the levels of the
metabolizing enzymes; thus,  potentially increasing the formation of active
metabolites and the potential risk of cancer.  Therefore,  the effects of DE
on  the  xenobiotic metabolizing  enzymes levels  were determined as  one
approach to assessing the potential carcinogenic  risk from DE exposure.

     The inhalation  study  involved the  chronic exposure of  Strain   A/J
male mice 8 hours/day,  7  days/week for periods of  6 and 8 months to  clean
air or DE  diluted to 6 mg/rn^ of particulates  at the U.S.  Environmental
Protection Agency, Center Hill facility in Cincinnati, Ohio").  In this
study,  lung   and  liver microsomal  aryl  hydrocarbon  hydroxylase  (AHH)
activities and liver  microsomal cytochrome P448/450 levels were determined
using the slightly modified methods  of  Van Cantfort  (1977)  et  al.,  and
Omura and Sato (1964), respectively.  The results  (Table 1)  indicated that
there were no statistical differences in the liver microsomal cytochrome
P448/450 levels and liver microsomal AHH activites between clean air and DE
exposed mice both at  6  and  8 months.  Small differences were noted in the
lung microsomal  AHH activities, but these are believed to be artifactual
differences, due  to  increases in non-microsomal  lung protein present in
the microsomal preparations.  The only significant  differences found were
in the weight of the animals' lungs.  The  DE exposed mice were found to have
significantly increased wet  lung weights,  which  could be attributed, in
                                    300

-------
part,  to  the  deposition of diesel participates in the lungs and possibly
to lung physiological and biochemical changes caused by the DE insult on
the lungs.

     A follow-up study was conducted  to  assess  the  ability of extracted
diesel  particulate  organics  to  cause changes  in  the  levels of  liver
cytochrome P448/450-   This  was done to  see  if  changes  in these  enzymes
levels could  be  produced from  intraperi toneal  (i.p.) injections of diesel
exhaust extract  (DEE)  given  at  a  maximum tolerated dose  to  mice.   The
extract of the  diesel  particulates  was  used  because it  should be  fully
bioavai Table  to the body's systems, whereas there are still doubts  as to
the degree of  bioavai labi lity  of the OE organics  when adsorbed onto diesel
carbon  particulates  and  then   deposited  in  the  body.   The experimental
approach involved the i.p. injection  of  male  and female Strain A/J mice
with  DEE,  the positive enzyme inducing  compounds  phenobarbital  (PB,  a
cytochrome P45Q  inducer)  or 3-methylcholanthrene  (MC,  a cytochrome PWQ
inducer),  or  the appropriate vehicle controls for two days and sacrificing
the animals the  third day.  Liver  microsomal preparations were immediately
prepared after  sacrificing and the  liver microsomal cytochrome P448/450
levels were determined.  The total doses given  per  kg body weight were DEE
- 500 mg,  PB  - 160 mg, MC  - 40 mg,  and vehicle controls  (DMSO and  saline)
- 300 ul.   The DEE was derived from the 24-hour  soxhlet extraction,  without
cellulose thimble,  of  DE particulates collected  on teflon coated pallflex
TbOA20 type filters,  using methylene chloride as  an  elutant.  The  extract
was  then  made  up  to the desired  injectable concentration in DMSO via
solvent exchange using  a nitrogen  atmosphere  for methylene chloride
removal.   The results  (Table  2)  showed  that PB  and MC did result in the
induction  of  the  respective   ?448  and  P^Q  enzymes in  both sexes  as
expected.   The  injected  DEE  caused  significant  increases  in  the  liver
cytochrome P448/450  levels in male  but not  female mice.  This increase,
however, was  smaller  than  those seen in the PB  and  MC induced animals, and
the  male  DEE induced  enzymes were found spectrally between where the
cytochrome ?448 and ?45Q enzymes were found for PB  and MC, respectively.
     It is therefore concluded, from the chronic  inhalation and the  i.p.
injection studies, that enzyme  inducing  chemicals  are  present  in  DE and
that the  absence  of  enzyme changes found  in  the mice  exposed  to  DE via
inhalation may be due to 1) the enzyme inducing organics  associated with DE
were not bioavai Table to the  body system and/or 2) the inhalation dose was
not sufficient to elicit a detectable change in the enzyme levels.


                               REFERENCES
1.   Pitts, Jr., J.N., K.  Van  Cauweberghe,  A.M.  Winer,  and W.L.  Belser.
     1979.  Chemical  Analysis  and Bioassay of Diesel  Emission  Particu-
     lates.  U.S. Environmental Protection Agency Report of Contract No.
     R806042.
                                    301

-------
2.   Hinners, R.G.,  J.K.  Burkart, M. Malanchuk,  and  W.D.  Wagner. 1980.
     Animal Exposure Facility for  Diesel  Exhaust  Studies.   In: Generation
     of Aerosols.  K. Willeke, ed.   Ann Arbor Science Publishes: Ann Arbor,
     Mich., pp.  525-540.

3.   Van Cantfort, J., J.  DeGraeve, and J.E. Gielen.  1977.  Radioactive
     Assay for Aryl  Hydrocarbon Hydroxylase.  Improved Method and Biolo-
     gical Importance.  Biochem. Biophys. Res. Comm. 79: 505-512.

4.   Omura, T.,  and  R.  Sato.   1964.  The Carbon Monoxide  - Binding Pigment
     of Liver Microsomes.   J. Biol. Chem. 239: 2370-2378.
                                  302

-------
                                    Table 1.   Liver Cytochrome P448/450 Levels and Liver Aryl Hydrocarbon
                                               Hydroxylase Activity in Mice Exposed to Clean Air or Diesel Exhaust
                                   _ Cytochrome P448/450 Level
                          Values = X +; SEM in (nMoles/mg microsomal
                                              protein)
                          *n = sample size
     Aryl_Hydrocarbon Hydroxylase Activity
Values = X _t SEM (pMoles/min./mg microsomal
                   protein)
*n = sample size
OJ
o
Months
Exposed
6
8
Control
1.52
n
1.61
n
+ 0.073
^ 10
+ 0.067
= 9
Diesel Exposed
1.54 +
n =
1.62 +
n =
0.066
11
0.081
8
Control
48.41 +
n =
50.84 +
n =
2.40
10
2.57
9
Diesel E)
44.08 +
n =
49.04 +
n =

-------
                                          Table 2.   LIVER CYTOCHROME P448/450 LEVELS

                                           Values  =  X _+  SEM in  (nMoles/mg microsomal protein)

                                                n  =  Sample Size
CO
o
                     Saline (control)
                     100 Hl/30gm BW
   Phenobarbital
   160 mg/Kg BW
in 100 pi saline
DMSO (control)
300(il/30gm BU
 Diesel  Exhaust
   Extract        3-Methylcholanthrene
  500 mg/KG BU        40 ing/KG BU
in 300 H! DMSO      in 300 H 1   DMSO
Males

Females

1.386
n
1.302
n
+ 0.049
= 4
+ 0.067
= 4
2.426
n
3.013
n
+ 0.020
= 4
+ 0.127
= 4
1.096 +
n =
1.106 +
n =
0.059
8
0.056
7
1.346
n
1.186
n
+ 0.080
= 6
+ 0.066
= 5
1.524 +
n =
1.512 +
n =
0
6
0
6
.077

.056


-------
SECTION 5



MUTAGENESIS
                        305

-------
MUTAGENIC ACTIVITY OF DIESEL EMISSIONS
Joellen Lewtas
Genetic Toxicology Division, Health  Effects  Research Laboratory,
U.S. Environmental Protection Agency,  Research Triangle Park,
North Carolina
INTRODUCTION
   The initial report that the organics  extractable  from diesel particles
demonstrate mutagenicity in the Ames Salmonella  typhimurium assay  has
                                                      2—4
now been confirmed by many independent investigators.      The mutagenic
activity in bacteria is characterized as causing frameshift mutations without
requiring microsomal metabolism.  Recent studies have  attributed this
bacterial mutagenic activity to the presence of  nitrated polynuclear
aromatic (NO -PNA) compounds in diesel organic emissions. '    Certain NO.-PNAs
(e.g., 1,8-dinitropyrene) are unusually potent frameshift bacterial mutagens,
which do not require an exogenous microsomal metabolic activation system for
         7 8
activity; '  they appear, however, to be activated by  endogenous bacterial
nitroreductases.   Concern that bacterial mutagenesis  assays may "overestimate"
                                  ,9
the mutagenic activity of NO -PNAs  in diesel emissions points to the importance
of evaluating the mutagenic activity of  these emissions in eucaryotic organ-
isms, mammalian cells, and whole animals.
   Mammalian cell mutagenesis bioassays capable  of detecting gene mutations, DNA
damage, and chromosomal aberrations have confirmed the mutagenic activity of
diesel emissions.  '    Many of these assays performed with mammalian cell lines
(e.g., L5178Y mouse lymphoma cells, BALB/o  3T3 cells,  and Chinese hamster ovary
[CHO] cells) require the addition of a metabolic activation system containing
microsomal as well as other mammalian liver enzymes  to metabolize polynuclear
aromatic hydrocarbons (PAHs).  Very few  studies  have been published on the
activity of NO -PNAs or presence of nitroreductases  in these systems.
   The objective of this paper is to review the  mutagenic activity of diesel
emissions.   The organics extractable from diesel particles,  which may consti-
tute 5 to 50% of the mass of these submicron particles,  have been most exten-
sively examined in microbial and mammalian  cell  mutation assays.   This paper
compares the microbial mutagenicity, mammalian cell  mutagenicity, and mouse
skin tumorigenicity of these organics.  Whole diesel particles, gaseous
emissions,  and whole exhaust emissions,  examined in  several different in vivo
bioassays for both somatic and heritable mutagenic activity, will also be
reviewed.
                             306

-------
METHODOLOGY
Mobile source emissions samples
   The diesel and gasoline particle  emissions  used  in  the  microbial  and
mammalian cell mutagenesis studies reported  here  were  collected  using  dilution
tunnel sampling techniques.    The total  exhaust  from  passenger  cars or  a
portion of the exhaust from heavy-duty  engines was  diluted with  filtered air
(10:1) prior to collection on  20  in. x  20 in.  Teflon-coated  Pallflex T68-A20
filters.  The mobile sources,  fuels, and  test  conditions are shown in  Table 1.
The diesel samples were all obtained from vehicles  and engines operated  on the
same lot of No. 2 diesel fuel.  The vehicles were operated on a  chassis  dyna-
mometer, using the highway fuel economy test cycle  (HWFET) that  averages 48
miles per hour in 12.75 minutes over 10.24 miles.   The engines were  operated
on engine dynamometers at steady-state  operation.

TABLE 1
MOBILE SOCTRCE SAMPLES
                  Vehicle Description
                                                        Fuel
    Driving
     Cycle
  Diesel:       Cat   Caterpillar 3304             Diesel No. 2
             Nissan   Nissan Datsun 220C           Diesel No. 2
               Olds   Oldamobile 350               Diesel No. 2
          VW Rabbit   Volkswagen Turbocharged      Diesel No. 2
                        Rabbit
           Mercedes   300 D Mercedes               Diesel No. 2
Gasoline:   Mustang   1977 Mustang 11-302,         Gasoline
                      V-8 Catalyst and EGR         Unleaded
           Chev 366   Heavy Duty Chev 366          Leaded
           Ford Van   6  cylinder,  in  line  van      Leaded
   Mode IIa
   HWFET"
   HWFET
   HWFET
   HWFET
   HWFET
   Full-rated
   loadc
   HWFET
aMode II      • 2200 rpm steady  state,  85  Ib.  load
''HWFET        « Highway fuel  economy  cycle 10.24 mi, ave. 48 raph,
cFull-rated   - 2300 rpm, 1 00% load
12.75 min.
   The particle samples collected on 12 to  16 Teflon-coated filters were Soxhlet-
extracted with dichloromethane  (DCM) in a 2.3 liter side-chamber extractor for
48 hours.  The Soxhlet-extracted organics were  filtered using Teflon millipore
filters  (0.2 ym pore) to remove any remaining particles and concentrated by
rotary evaporation under reduced pressure.  Aliquots were evaporated to dryness
                              307

-------
under nitrogen and stored  frozen  in the dark.  The samples were dissolved in
dimethyl sulfoxide  (DMSO)  for  all of the bioassays except mouse skin tumori-
genesis, oncogenic transformation,  and mutation in BALE cells where acetone
was used as the solvent.
   The whole exhaust emissions employed in the in vivo mutagenesis bioassays
are described in the references cited for each of those assays.

Bioassays
   The mutagenesis bioassays applied to diesel emissions generally included
assays for which standard protocols have been developed and validated with a
series of individual chemicals.   The mutagenesis assays were selected to detect
gene mutations, DMA damage, and chromosomal aberrations, as outlined in Figure
1.  The bioassays were conducted  with coded samples at 5 to 7 doses or concen-
trations after a preliminary toxicity range-finding test.   For those assays
where a positive dose response was  obtained,  the slope of the dose-response
curve was determined by the linear  regression analysis, except for the £.
typhimurium plate incorporation assay,  where the non-linear model slope
was used.12'13

GENE MUTATION ASSAYS
Salmonella typhimurium bioassay
   The Ames S_. typhimurium assay  measures histidine reversion in a series of
tester strains.  The S_. typhimurium plate incorporation test was conducted as
                         14                                                  15
described by Ames et al.,   with  minor modifications as described by Claxton.
The modifications included adding the minimal histidine to the plate media rather
than to the overlay and incubating  for 72 rather than 48 hours. Claxton initially
reported the specific activity of five of these mobile source samples calculated
from the linear regression analysis at 100 yg of sample.    This data and that
from the additional samples have  been reanalyzed using the non-linear model slope
         12 13
analyses,  '   as shown in Table  2.
   The extractable organics from  the diesel particle samples were all mutagenic
without the addition of metabolic activation.  Comparison of the mutagenic
activity in all five tester strains   showed the diesel samples to be positive in
TA1538, TA1537, TA98, and  TA100,  and negative in TA1535.  The Caterpillar (Cat)
and Volkswagen  (VW) Rabbit samples  show increased mutagenic activity in the
presence of the S9 activation  system, whereas the Nissan and Oldsmobile (Olds)
samples show decreased activity with S9 activation.  The gasoline particle
samples were all less mutagenic in  the presence of S9 activation.
                             308

-------
 A.  MUTAGENESIS BIOASSAYS
     1 .  GENE  MUTATION  ASSAYS
         A.  Bacterial
             1 .  Salmonella  typhimurium
             2.  Escherichia coli

         B.  Mammalian  cell
             1.  Mouse  lyraphoma, L51 78Y
             2.  Mouse  embryo  fibroblasts,  BALB/C3T3
             3.  Chinese  hamster ovary, CHO

     2.  DNA DAMAGE  ASSAYS
         A.  Yeast
             1 •  Saccharomyces cerevisiae D3  recombinogenic  assay

         B.  Mammalian  Cell
             1 .  DNA strand  breaks  in SHE cells
             2.  Unscheduled DNA repair in  liver  cells
             3.  Sister chromatid exchanges in CHO  cells

     3.  CHROMOSOMAL ABERRATIONS
         A.  Mammalian  cells
             1.  CHO cells
             2.  Human  lymphocytes

 B.  CARCINOGENESIS  BIOASSAYS

     1 .  ONCOGENIC TRANSFORMATION ASSAYS
         A.  Chemical transformation
             1 .  Mouse  embryo  fibroblasts,  BALB/ c 3T3
             2.  Syrian hamster embryo, SHE

         B.  Viral enhancement of transformation
             1 .  SA7 virus enhancement in SHE cells

     2.  SKIN  TUMOR  INITIATION
 Fig. 1 .  Outline  of  the  bioassays  used  to  examine  the  extractable organics from
          mobile source particle  emissions.


   A significant difference was observed between  the particle emission rates
                                                                             I7
 (g/mi) and the percent organic extractable  matter for the different vehicles.
The diesel cars emitted approximately  100 times more particles per mile than
the unleaded gasoline car.  A direct comparision  of the mutagenic emission rate
for the cars is best  expressed as revertants/mile.  Claxton and Kohan   have
reviewed the mutagenic emission rates  for a number  of certification vehicles
and found that the diesel vehicles  emitted  45 to  800 x  as much mutagenic
activity per mile as  the gasoline catalyst  vehicles.
                             309

-------
 TABLE 2
 REVERSE  NOTATION IN SALMONELLA TYPHIMDRIOM
 Vehicle Description
                                                    Slope:  Rev/pg* in TA98
                                               -MA                        +MA
Diesel:



Cat
Nissan
Olds
VW Rabbit
0.30
20. 8
2.1
5.2
1.6
15.1
1.4
6.1
 Gasoline,  unleaded:
              Mustang
 Gasoline,  leaded:
             Chev 366
             Ford Van
       Benzo(a)pyrene
 1-Nitropyrene (>99%)
  1-Nitropyrene  (95%)
   2.1
   INC
  16.8
   NEC
 572
4234
                             8.6
  INC
 29.7
167.9
802
736
 ^Non-linear model slope,  revertants/ug.
Escherichia coli WP2 bioassay
   The E_. coli WP2 tryptophan reversion  assay is very similar to the S_.
                                     14
tyhimurium plate incorporation  assay  using McCalla's E.  coli WP2 tryptophan
                                                           ~"~18            19
auxotroph  (trp) with a. DMA repair deficiency mutation (uvrA).    Mortelsmans
found that the Mercedes diesel  sample elicited a. reproducible dose-related in-
crease in the number of trypotophan-independent revertants in the absence of
metabolic activation.  In the presence of metabolic activation, the Mercedes
diesel sample was non-mutagenic.
L5178Y mouse lymphoma mutagenesis  assay
                                                        20
   The L5178Y mouse lymphoma assay  of Clive and Spector   measures forward
mutation frequency at the thymidine kinase (TK)  locus.   The mouse lymphoma
assay was conducted according  to  the method of Clive et al.    by Mitchell
      22                        23
et al.   and Cifone and Brusick  in the evaluation of the mutagenicity of
a series of diesel and related environmental emissions.  Preliminary dose-
range toxicity assays were  conducted to select 10 concentrations of each
sample that resulted in cell survivals of 5 to 90% of the controls.  In the
mutagenesis assays, duplicate  samples were used for each concentration tested.
                            310

-------
In each assay, 6 x 10  L5178Y TK+/- cells were  treated  with  the  organic  extracts
in 10 ml for 4 hours while rotating in a roller drum  at 37°C.
   The mutation frequency was calculated by  dividing  the number  of mutant  cells
per ml by the number of viable cells per ml  at  each concentration.Concentrations
resulting in less than 10% total relative growth were not used in determining
the slope of the mutation response curve for each  emission sample, as shown in
Table 3.

TABLE 3
GENE MDTATION  IN MOOSE LYMPHOMA  L5178Y CELLS*
                                      Slope:   Mutation freq/10   cells/yg/ml
  Vehicle  Description
                                     -MA
            (P2)
                                                             +MA
             (r2)
Diesel: Cat
Nissan
Olds
VW Rabbit
Mercedes
0.25
4.19
1 .21
0.98
NEC
(.96)
(.88)
(.95)
(.89)

0.063
2.87
1 .28
0.72
1 .82
(.78)
(.86)
( .93)
(.64)
( .87)
   Gasoline,  unleaded:
                Mustang
      Gasoline,  leaded:
              Chev 366
              Ford Van
         Benzo(a)pyrene
   1-Nitropyrene  (95%)
0.38
1 .50
NEGb
NEC
NEC
           (.98)
           (.81)
 1 .09

 3.20
 5.60
 5.42
39.3
(.77)
( .90)
 5Assay performed  with 6x10° cells in 10 ml for 4 h.
 ^Highly  toxic  at  less than 10 yg/ml.
   All the diesel samples were mutagenic  in  the mouse  lymphoma assay, and except
for the Mercedes sample, all the diesel samples showed that  the mutagenic activ-
ity was greater in the absence of the metabolic activation system.  All the
diesel organic samples were also more cytotoxic in  the absence of metabolic
activation than in its presence.     The maximum increases in mutation frequency
(2 to 4 times the spontaneous frequency)  occurred at concentrations ranging
from 20 to 300 ug/ml.  The gasoline  catalyst Mustang sample  was more mutagenic
and cytotoxic in the presence of metabolic activation  than in the absence of
the activation system.  Polycylic aromatic hydrocarbons such as benzo(a)pyrene
(B[a]P) are not mutagenic in this assay without the addition of the S9 metabolic
activation system.  Preliminary evaluation of 1-nitropyrene  (95%) in this assay
                              Jl I

-------
suggests that it also requires  an  exogenous metabolic activation system for
activity.

BALB/c 3T3 mutagenesis assay
                                                                          24
   The BALB/c 3T3 mutagenesis assay was  developed by Schechtman and Kouri  to
measure simultaneously both mutagenic  activity and morphological transformation.
                                                       25                 6
Forward mutation is measured using ouabain resistance.    Cells (1-2 x 10 ) were
exposed in suspension for 2 hours  with increasing concentrations of the diesel
organics dissolved in acetone.  Curren et al.    assayed the Caterpillar, Nissan,
and Oldsmobile diesel samples,  and the Mustang gasoline sample in the BALB/c
3T3 mutagenesis assay.  Although several individual doses of the diesel sample,
did induce a significant increase  in ouabain-resistant mutants, none of the
samples induced a dose-dependent increase in mutation frequency.  A majority of
the concentrations tested appeared to  be above the limit of solubility as
evidenced by  insoluble material in the  assay.   This problem had not previously
been encountered when DMSO was  employed  as a solvent with a similar sample.
             26
Curren et al.   assumed that all seven doses tested may have been similar due
to the solubility limits.  They combined all of the mutant colonies observed
for a sample and divided it by  the total number of surviving cells to determine
a mutation frequency for the dose  range  tested.  Using this method of analysis,
both the Nissan diesel sample and  Mustang gasoline sample were highly mutagenic •'
(p<0.05) both without and with  metabolic activation.  The Oldsmobile diesel
sample showed approximately a twofold  increase in mutation frequency, which
was not significantly different from the solvent control, and the Caterpillar
diesel sample showed no increase in mutation frequency.
Chinese hamster ovary mutagenesis  assay
   The CHO assay measures forward  mutation at the hypoxanthine-guanine phos-
phoribosyl transferase  (HGPRT)  locus  using 6-thioguanine resistance.    The CHO
                                                       28
assay was conducted with modifications by  Casto et al.    to evaluate the
                                                                          6
mutagenic activity of the diesel and  gasoline samples.   Cells (1.5-2 x 10 ) were
treated with increasing concentrations of  the organics dissolved in DMSO for
24 hours.  These assays were only  conducted in the absence of metabolic activa-
tion in two to three separate  experiments.  Re-analysis of the combined data
with cell survivals above 10%  using linear regression analysis showed a
relatively weak to negative response  for the Caterpillar, Oldsmobile, and
Mustang samples.  The samples  that would be considered positive were the Nissan
and VW Rabbit with activities  of 0.16 (r =0.73)  and 0.091 (r =0.46) mutation
                            312

-------
 frequency/10  cells/pg/ml,  respectively.   Li and Royer29 also have reported
 that  the  extractable organics from a series of diesel cars was generally very
 low in  mutagenic activity in CHO cells with a slight increase in activity with
 metabolic activation.   Simultaneous treatment of the CHO cells with a co-mutagen
 (e.g.,  B[a]P)  caused significant enhancement of the mutagenic activity.29
   Cheshier  et al.    have shown that CHO  cells readily phagocytize whole diesel
 particles, which become closely associated with the nucleus.   Under these
 conditions,  100 ug/ml of diesel particles caused a  tenfold increase in mutation
 frequency above the controls.

 DNA DAMAGE ASSAYS
 Saccharomyces  cerevisiae D3  recombinogenic assay
   The  diploid yeast S.  cerevisiae D3 can be used to measure mitotic recombination
 by scoring for red  pigmented mutant colonies formed in the presence of adenine.
 The mutants  are generated from a recombinational event resulting from DNA
 breakage  and repair after exposure to DNA-damaging  chemicals.   Initial studies
 on the  diesel  and comparative samples reported that no reproducible or dose-
 related responses were observed.     Further studies of these  samples in  the S_.
 cerevisiae assay showed that two of the three diesel  samples  assayed,  the Nissan
 and VW  Rabbit,  did  result in reproducible dose-related increases in mitotic
 recombinants.   The  Nissan sample caused 62 mitotic  recombinants/10  surviving
 cells/ug/ml  (r =0.79)  without activation  and 46 (r  =0.64)  with activation.  The
 VW Rabbit caused 24 mitotic  recombinants/10  surviving cells/pg/ml (r =0.4)
 without activation  and 7.2  (r  =0.2)  with  activation.   The  response was greater
 and more  reproducible  in the absence of metabolic activation.   The Oldsmobile
 sample  was weakly positive in  the absence of metabolic activation,  and the
 gasoline  Mustang sample  did  not reproducibly increase  the  number of mitotic
 recombinants.   Polycyclic aromatic hydrocarbons (e.g.,  B[a]P)  do not induce
 mitotic recomb:
 in this assay.
mitotic recombination in S. cerevisiae,   and NO -PNAs have not been examined
DNA strand breaks in Syrian hamster embryo cells
   Damage to cellular DNA, which results  in the formation of DNA fragments, can
                                                    32
be measured directly by alkaline elution  techniques.    Casto has shown that
chemical induction of DNA damage in primary Syrian hamster embryo (SHE) cells
can be detected following centrifugation  on alkaline sucrose gradients.    The
diesel Caterpillar, Nissan, Oldsmobile, and VW Rabbit samples were tested at four
concentrations from 31 to 250 yg/ml.  None of these samples produced a significant
                            313

-------
change in the sedimentation profile  of DNA from the treated SHE cells.  The
gasoline Mustang sample did produce  a significant increase in DNA strand breaks
at the highest concentration,  tested  (250 ug/ml) .   In comparing several In vitro
tests to detect carcinogens in  Syrian hamster cells, Casto suggests that the
                                                                          28
DMA strand breakage assay  is  the  least sensitive  of the assays evaluated.

Unscheduled DNA repair in  liver cells
   The liver cell DNA repair  assay measures autoradiographic unscheduled DNA
synthesis in freshly isolated hepatocytes.   The Oldsmobile diesel sample was
evaluated in the hepatocyte primary  culture/DNA  (HPC/DNA)  repair assay by
Williams according to previously  published procedures."'   Unscheduled DM repair
was induced from 10 to 100 pg/rol  with an average  of 36.7 grains/nucleus at
100 yg/ml.  The response appeared to be dose-related;  however, insufficient
numbers of concentrations were  tested in any one  experiment to perform a
regression analysis.  Combination of the data from four separate experiments
resulted in a. slope of 0.325  grains/nucleus/pg/ml (r =0.78) .
Sister chromatid exchange assay in CHO  cells
   The sister chromatid exchange  (SCE)  assay  measures the increase in exchanges
between two chromatids of each chromosome  in  cells grown in the presence of
bromodeoxy-uridine  (BrdU) during  replication.   The increase in SCEs observed
after cells have been treated with chemical mutagens has been related to ra-
combinational or post replicative repair of DNA damage.     The diesel and gaso-
line emission samples were tested in  the SCE  assay using the CHO cell system
                     22
previously described.    This method,  uses  a 21.5-hour sample exposure period;
however, due to the cytotoxic effects of the  metabolic activation system, only
a 2-hour exposure period was used when  the samples were tested with metabolic
activation.  It is not possible,  therefore, to  compare directly the induction
of SCEs with and without metabolic activation.   The slope of the dose-response
regression analysis is shown in Table 4.
   All of the diesel and gasoline samples, except the Oldsmobile sample, induced
SCEs in the absence, of metabolic  activation.   In the presence of metabolic
activation, all of the diesel samples induced SCEs except the Caterpillar
sample.  The significantly lower  activity  in  the presence of activation is
presumably due in part to the much shorter exposure period.  The polycyclic
aromatic hydrocarbon  (B[a]P) tested in  this assay only induced SCEs when meta-
bolic activation wa,s added.  Prelimi-nary studies on 1-nitropyrene  (95% pure)
showed that it was weakly active  in the absence of the metabolic activation
system.
                             314

-------
 TABLE 4

 SISTER CHROMATID EXCHANGES IN CHO CELLS
                                            Slope:  SCE/cell/gg/ml
   Vahicla Description             -MA**         (r2)       +MAb         (r2}
Diesel: cat
Nissan
Olds
VW Rabbit
Gasoline, unleaded:
Mustang
Benzo(a)pyrene
1 -Nitropyrene (95%)
0.011
0.30
NEC
0.075

0.076
NEG
0.066
(.83)
( .93)

(.99)

(.99)

(.80)
NEG
0.071
0.017
0.030

NTC
1 .24
NEG

( .87)
( .46)
( .92)




 b+MA exposure 2 h.
 °NT - not tested.
CHROMOSOMAL ABERRATIONS
Chromosomal aberrations  in CHO cells
   Chromosomal aberrations that can be  detected  as  a  result  of treatment of
cells in culture  include both numerical and  structural  aberrations.  Scoring
of numerical aberrations,  however, is not  generally recommended  for this assay.
Structural aberrations include breaks,  deletions, gaps,  exchanges, or trans-
locations at chromosomal and/or chromatid  levels.   These aberrations are
generally observed between 6  to 24 hours after cell treatment.   In order to
determine the optimal time after treatment to observe aberrations, CHO cells
treated with the  Nissan  sample for 6 hours were  scored  for structural chromo-
somal abnormalities' at 12,  15,  and 21 hours.     A summary of those results is
shown in Table 5.  A dose-related positive response was  observed at all three
time periods.

Chromosomal aberrations  in human lymphocytes
   Human lymphocytes freshly  isolated from blood samples  taken from normal
individuals can be exposed to chemicals in vitro and  analyzed for chromosomal
aberrations.  The diesel Oldsmobile sample was exposed to lymphocytes from two
individuals at five doses  ranging from  0.1 to 100 ug/ml with and without an
39 metabolic activation  system.   Chromosome aberrations were scored by McKenzie
according to previously published criteria.    In the absence of metabolic
                             315

-------
 TABLE 5
 SUMMARY  OF CHROMOSOMAL ABERRATIONS IN CHO CELLS
Hours after Treatment

yg/ml
Nissan
0
20
40
60
80

Total
Cells
515
136
129
115
192
12
Percent
Aberrations3
1 .75
5.07
6.98
18.26
20.83

Total
Cells
276
152
59
79
Tb
15
Percent
Aberrations
4.06
10.53
28.80
62.03
T

Total
Cells
147
151
98
122
88
21
Percent
Aberrations
2.00
6.80
10.91
15.28
22.72
 Percentage of cells with all types of aberrations.
 kr - toxic.
activation, treatment of lymphocytes with  the  diesel  Oldsmobile sample resulted
in a four- to fivefold increase in the percentage  of  cells with chromosomal
aberrations over the dose range tested.  Chromosome and chromatid breaks and
aneuploidy were observed at 0.1 to 1.0 vig/ml.   Chromosomal fragments, dicentrics,
and endoreduplications were observed at doses  above 5 yg/ml.   Chromosomal
and chromatid gaps were only observed at 100 ug/ml.   In the presence of
metabolic activation, no increase in the total percentage  of  cells with
aberrations was observed, although an increase in  chromosomal fragments and
dicentrics was observed.

IN VIVO MUTAGENESIS BIOASSAYS
   The organics extractable from diesel particle emissions are mutagenic in
many microbial and mammalian cell assays,  as described above.  However, these
assays are not readily applicable to testing whole diesel  emissions nor can
they test for the heritability of mutations.   -For  these reasons,  plant
(Tradescantia), insect (Drosophila), and mammals  (mice and hamsters) have
been employed to evaluate the iri vivo mutagenic activity of diesel emissions,
as summarized in Table 6.
   The Tradescantia micronucleus test and  stamen hair gene mutation assays
both have been shown to detect the mutagenic activity of volatile and
                                                   •3 Q
gaseous chemicals and environmental emissions. Ma   reported that diluted
diesel exhaust induced micronuclei  (broken pieces  of  chromosomes)  in
                                                                  39
Tradescantia.  Whole diesel emissions were also shown by Schairer  to
induce gene mutations in the Tradescantia  stamen hair assay.
                             316

-------
TABLE 6

IN VIVO MUTAGENICITY OF DIESEL EMISSIONS
Bioassay System
       Endpoint
                          Reference
                  Whole                         Extractablea
                 Emissions  Gases8  Particles3   Organics
Tradescantia
Tradescantia
  mutation
Micronucleue test        Ma et al.38
Stamen hair gene assay   Schairer39
Drosophila
  melanogaeter
Sex-linked recessive
  lethal test
Schuler and
Nieraeir40! Nix41
 Mouse
 House
 Mouse  specific
   locus
 Mouse  dominant
   lethal
 Mouse  heritable
   translocation
   test
Micronucleus assay
Bone marrow SCE assay
Point mutation test

Chromosome damage test

Chromosome damage test
Pereira42
Pereira42
Russell et al.44

Russell et al.44

Russell et al.44
 Chinese hamster   Micronucleus  assay       Pereira42
 Syrian hamster    Lung cell  SCE assay      Rounds43
 Syrian hamster    Fetal liver SCE assay    Pereira42
 "Where bioassays have not been conducted,  no entry  in  the  table  is shown.
 b(+) = weakly positive.

-------
   The fruit fly, Drosophila melanogaster,  provides a well-defined genetic
test system to measure  inherited damage.   Two independent investigators  '41
have evaluated the mutagenicity of whole  diesel emissions using the D._
                                                     41
melanogaster sex-linked recessive lethal  assay.  Nix   also tested the gaseous
emissions from filtered exhaust.   Neither the whole nor filtered exhaust was
found to induce mutations  in this assay.
   Whole animal rodent  bioassays  using mice or hamsters provide the opportunity
to measure genetic damage  (e.g.,  induction of micronuclei or induction of
SCEs) in somatic cells  as  well  as heritable genetic damage.  Both mice and
                                              42           43
hamsters have been used in studies by Pereira   and Rounds   to measure
induction of micronuclei and SCEs in bone marrow,  lung cells, and fetal liver
after exposure to whole diesel  emissions.  In all  of these studies except the
lung cell SCE assay, the whole  emissions  were negative.  After exposure to
collected particles, the SCE assays were  positive  in both bone marrow and
lung cells.  All of these  genetic damage  assays in somatic cells were positive
when the animals were treated with the organics extracted from diesel
particles.  These studies  suggest that the  organics associated with diesel
particles are capable of inducing genetic damage in somatic cells in the  lung,
bone marrow, and fetal  liver.   However, under conditions where the animals
were exposed to high concentrations of whole diesel exhaust for several months,
only induction of SCEs  in  lung  cells was  observed.  These results suggest
that insufficient concentrations  of the mutagenic  organics would reach
the germinal cells to cause heritable mutations.
   Heritable mutations  in  mice  after exposure to diesel exhaust were assayed
                     44
for by Russell et al.   using the specific  locus,  dominant lethal, and heri-
table translocation assays.  The  results  in all the heritable mutagenesis
assays were negative.
   The in vivo mutagenesis studies further confirm the mutagenic- activity of
the organics associated with diesel particles,  while showing the lack of
transmitted genetic effects after animal  exposure  to whole diesel exhaust
emissions.  These findings suggest either that the mutagenic components do
not reach the gonads, or that the heritable genetic assays are insensitive
to the frameshift mutagens present in diesel emissions.  Polycyclic aromatic
hydrocarbons and other  frameshift mutagens  such as the NO -PNAs have not
been well studied in either the Drosophila or mouse heritable mutagenesis
assays.
                             318

-------
CARCINOGENESIS BIOASSAYS
Oncogenic transformation assays
   Chemically induced carcinogenesis  is currently considered to be a multistep
process that may involve DNA damage or mutation as an  initial step.  Oncongenic
transformation assays measure the induction of morphological transformations
that result in the formation of colonies of cells phenotypically similar to
malignant cells  (Type III foci) .  These transformed cells generally cause
tumors when injected into a syngeneic host.
   Several of the diesel and gasoline samples in Table 1 have been tested in two
                                                ^/-
oncogenic transformation assays.  Curren et al.,   using mouse embryo cells
 (BALB/c 3T3) found that all of the diesel and gasoline samples, except the
Caterpillar sample, induced some transformed foci.  Dose-related responses were
not observed, which may be due to the problems discussed above with the BALB/c
                                                                   28
mutagenesis assay that was conducted  simultaneously.   Casto et al.,   using
primary SHE cells, found that none of the diesel or gasoline samples caused
transformation in these experiments.  Unfortunately, lack of induction of
transformation by one of the positive controls and difficulties in obtaining
acceptable lots of serum for these assays prevented further testing.

Viral enhancement of transformation
   The viral enhancement assay measures the increased  sensitivity of cells to
virus-induced transformation.  Although this assay is  listed with the trans-
                              45
formation assays, Casto et al.   have reported the significance of DNA damage
and repair in the enhancement of viral transformation  by chemicals.  This
assay may be a measure, therefore, of DNA damage.  The viral enhancement of
the transformation assay of Casto     was employed in the evaluation of the
                                                  29
diesel, gasoline, and several comparative samples.     The transformation
frequency was determined  (number of transformed foci per 10  surviving cells)
in at least three separate experiments.  The dose response curves for selected
                                         28
experiments were reported by Casto et al.    The combined data from all experi-
ments have been re-analyzed to determine the slope of  the dose response.
Concentrations resulting in less than 10% survival were not used in determining
the slope of the transformation response, as shown in  Table 7.
   All of the diesel and gasoline samples, except the  Caterpillar sample, in-
creased the viral enhancement of transformation.  The  Oldsmobile and VW Rabbit
samples were very weakly active, and  the dose responses had low r" values, 0.68
and 0.25, respectively.  The variation in response between the three separate
                              319

-------
TABLE 7
ENHANCEMENT OF VIRAL TRANSFORMATION  IN
SYRIAN HAMSTER EMBRYO CELLS
                                           Transformation
  Vehicle Description                      Freguency/vg/ml              (r2)
Diesel : Cat
Nissan
Olds
VW Rabbit
Gasoline : Mustang
Benzo ( a )pyrene
NEC
0.328
0.021
0.059
0.33
351 .0

(0.76)
(0.68)
(0.25)
(0.18)
(0.85)
experiments was signif icant^, and even the Mustang gasoline sample,  which caused
a 0.33 transformation frequency/ug/ml, had  an  unacceptably low r value of
0.1S for the combined slope analysis.  The  Nissan sample caused a transformation
frequency equivalent to the Mustang  sample, with an r  of 0.76.

Skin tumor initiation
   Mice treated topically with chemical carcinogens produce both benign
(papillomas) and malignant  (squamous cell carcinomas)  tumors.   The  tumor-
initiating activity of a chemical can be determined when mice  are treated
with a single application of the chemical and  subsequently treated  with a
        &
strong tumor promoter (i.e., 12-0-tetradecanoyl  phorbol-13-acetate  [TPA]).
Tumor-initiating chemicals are thought to induce somatic mutations  as  a result
of covalent binding to DMA and other macromolecules.     Nesnow et al.   have
reported the detailed methods and results of skin tumor initiation  studies
on these diesel and gasoline extracts in SENCAR  mice.   The skin tumor-
initiating activity to produce papillomas of these samples is  shown in Table 8.
   Papillomas were induced with all  of the  samples except the  Caterpillar.
Complete analysis of the tumor initiation activity and a discussion of the
carcinogenic activity of these samples on mouse  skin is reported by Nesnow
    .  49
et al.
                              320

-------
TABLE 8
SKIN TUMOROGINESIS IN SENCAR MICE
                                         Slope:  Papillomas/
  Vehicle Description                        Mouse/rag*                (r2)
Diesel: Cat
Nissan
OldB
VW Rabbit
Mercedes
Gasoline, unleaded:
Mustang
Benzo ( a ) pyrene
1 -Nitropyrene (>99%)
NEG
0.52
0.14
0.30
INC

0.085
86.2
INC

(.99)
(.83)
(.53)


(.76)
(.99)

aAverage of males and  females.

 SUMMARY AND DISCUSSION
 Comparison of mutagenic and carcinogenic activity of extractable organics
 from diesel particle  emissions in various bioassgys
    The organics extractable from diasel particle emissions were found to be muta-
 genic in all three types of bioassays:  gene mutation assays, DMA damage assays,
 and chromosomal aberration assays.  The three mutagenesis assays that resulted
 in reproducible dose-response data and that have also been used to evaluate at
 least four organic emission samples are:  S_. typhimurium bacterial mutagenesis
 assay (Table 2), L5178Y mouse lymphoma mutagenesis assay  (Table 3), and the SCE
 assay in CHO cells  (Table 4).  The relative activity of the diesel and
 gasoline organic emission samples has been compared between these three
 assays and with the two short-term carcinogenesis assays, which resulted in re-
 producible dose-response data, enhancement of the viral transformation assay
 (Table 7), and mouse  skin tumor initiation assay  (Table 8).
    In order to evaluate whether the relative activity of these samples correlated
 between assays, the activity determined from the slope of the dose-response for
 each sample in one assay was plotted versus the second assay.  Linear regression
 analysis and confidence bands were determined as shown in Figure 2.  The correla-
 tions, as indicated by the r2 values  (Table 9) for the gene mutation assays
 when plotted versus all the other assays was very good  (r2>0.90) in the absence
 of metabolic activation, except for the mouse lymphoma and the skin tumorigene-
 sis versus SCE in CHO cells.  The addition of metabolic activation to the  5.
                              321

-------
                  t4.00
CO
ro
ro
                          SALMONELLA VS MOUSE LYMPHOHA       /

                          BOTH ASSAYS +MA                / FORD
                                         SLOPED. 181    ,
                                           RT2..0.63    /
                                                                            •2.00
                                                                            «l .
                           /
                                      REVERTANTS/UC (+S8I
                                      INON LINEAR OOOEL SLOPE I
                          0.00   10.50  +1 00
                                              |  58  f2.t
SALMONELLA VS MOUSE LYMPHOMA    /

BOTH ASSAYS -MA              /
              SLOPE=0.I87    ,
                R12=0 96   /
MUTATION

FREO
                                                                                     PER
          REVERTANTS/UG I-S8I
           INON LINEAR MODEL  SLOPE)
                                                                                     9.00    40.50
                                                                                                      .00    t|.50
                Fig. 2.  Linear regression analysis of the mutagenic activity of diesel and gasoline samples in the
                         S.  typhimurium mutagenesia assay versua the L51 78Y mouse  lymphoma assay.  Confidence bands  are
                         shown in dotted lines.

-------
typhimurium assay decreased its correlation with both viral  enhancement  and  skin
tumorigsnesis.  The viral enhancement  assay,  which  is thought  to  be  dependent
upon DMA breakage to allow increased frequency  of virus  insertion, correlated
highly  (r >0.96) with the mutagenesis  assays  in the absence  of metabolic
activation.  The mouse lymphoma assay  both with and without  metabolic  activation
correlated highly  (r =0.95) with the skin tumorigenesis  assay.


TABLE 9

COKRELATIOK OF  DOSE-RESPONSE  SLOPES
DIESEL  AND GASOLINE
       Bioassay  Comparison
                                                   Exogenous
                                                   Metabolic
                                                   Activation
 Salmonella  versus  Mouse Lymphoma
 Salmonella  versus  Mouse Lymphoma

 Salmonella  versus  SCT in CHO
 Salmonella  versus  SCE in CHO

 Salmonella  versus  Viral Enhancement
 Salmonella  versus  Viral Enhancement
 Salmonella versus Skin Tumorigenesis
 Salmonella versus Skin Tumorigenesis

 Mouse  Lymphoma versus SCE in CHO
 Mouse  Lymphoma versus SCE in CHO

 Mouse  Lymphoma versus Viral Enhancement
 Mouse  Lymphoma versus Viral Enhancement

 Mouse  Lymphoma versus Skin Tumorigenesis
 Mouse  Lymphoma versus Skin Tumorigenesis

 SCE  in CHO versus Viral Enhancement
 SCE  in CHO versus Viral Enhancement

 SCE  in CHO versus Skin Tumorigenesis
 SCE  in CHO versus Skin Tumorigenesis

 Viral  Enhancement versus Skin Tumorigenesis
-MA
+MA
-MA
+MA
-MA
+MA
-MA
+MA
-MA
+MA
-MA
•fMA
-MA
+MA
-MA
-I-MA
-MA
+MA
-MA
 Exogenous  metabolic activation (S9)  added to one  or both  assays  is
  shown  as +MA;  when no exogenous metabolic activation system  was  added
  to either  assay,  it is shown as -MA.
0.96
0.93

0.98
0.94

0.99
0.79

0.90
0.72

0.84
0.87

0.96
0.83
 .95
 .95
0.96
0.93
0.83
0.83
                    0.92
                               323

-------
  These  studies  suggest that there is generally good agreement both qualitatively
and quantitatively between the short-term rnutagenesis and carcinogenesis bio-
assays in which a dose-related response  is observed for the organics extractable
from diesel and gasoline emission particles.   Several assays (e.g., DMA strand
breaks and oncogenic transformation  in SHE cells)  do not detect activity in
these samples.  Other assays  (e.g.,  mutagenesis  and oncogenic transformation in
BALB/c 3T3 cells) did provide qualitative data to  indicate that these  organics
were active; however, reproducible dose-related  responses were not observed.
This result may be due to a lack of  increasing amounts of chemical reaching the
cell as the exposure concentration increased,  probably as a result of  solubility
problems with these complex mixture  samples.   More solubility problems arose in
those iri vitro assays where acetone  rather than  DMSO was used as  a solvent.

Conclusions
   The studies reviewed here were undertaken to  evaluate the mutagenicity of
organics associated with diesel particle  emissions in a battery of mammalian
cell bioassays.  These data provide  strong evidence that these organics are
mutagenic in mammalian cells.  Furthermore, the  relative activity of a series
of emission extract samples, which exhibit approximately one order of  magnitude
range in activity in the S_. typhimurium bacterial  mutagenesis assay, exhibits
a similar range in activity in mammalian  cell  assays.   These studies suggest
that bacterial mutagenesis assays are not overestimating the mutagenicity of
these organics compared to mammalian cells, nor  are they greatly  overestimating
the relative tumor-initiating activity in skin carcinogenesis studies.
   Since a significant portion of the bacterial  mutagenic activity appears to
be due to NO -PNA compounds, and particularly  mono-and di--nitrated pyrene, more
studies are needed to evaluate the activity of these compounds in mammalian
cells.  Preliminary studies reported at this symposium on the activity of
1-nitropyrene  (95% pure  and  contaminated with  dinitropyrenes)
suggest that these compounds are active in mammalian cells.   The  concentrations
of mono- and di-nitrated pyrenes   in the samples  tested here,  however,  can not
account for all of the "direct-acting" mutagenic activity observed in  mammalian
cells treated with diesel particle organics.   Furthur research is needed to
identify other mutagenic and potentially  carcinogenic constituents of  diesel
emissions.
                             324

-------
ACKNOWLEDGMENTS

   The author gratefully acknowledges  the  editorial  assistance of  Olga Wierbicki,

Northrop Services,  Inc., the  technical assistance  of Katherine Williams, D.S.

Environmental Protection Agency,  and the helpful review  comments on  early drafts

of this manuscript  of Larry Claxton and Stephen Nesnow,  U.S.  Environmental
Protection Agency.


REFERENCES

 1.  Huisingh, J.,  Bradow, R., Jungers, R., Claxton, L., Zweidinger, R.,
     Tejada, S., Bumgarner, J., Duffield,  F., Waters, M., Simmon, V.,
     Hare, C., Rodriguez, C.  and  Snow,  L.  (1978) in Application of Short-
     Term Bioassays in the Fractionation and Analysis of Complex
     Environmental  Mixtures,  Waters, M.D., Nesnow, S., Huisingh, J.L.,
     Sandhu, S.S. and Claxton, L. ed. ,  Plenum Press, New York, pp. 383-418.
 2.  Siak, J.S., Chan, T.L. and Lee, P.S.  (1980) in Health Effects of Diesel
     Engine Emissions, Vol. I, Pepelko, W.E., Danner, R.M. and Clarke, N.A.
     ed., EPA 600/9-80-057a,  D.S. Environmental Protection Agency, Cincinnati,
     OH, pp. 245-262.
 3.  Loprieno, N.,  DeLorenzo, F., Cornetti, G.M. and Biaggini, G. (1980)  in
     Health Effects of Diesel Engine Emissions, Vol. I, Pepelko, W.E., Danner,
     R.M. and Clarke, N.A. ed. , EPA 600/9-80-057a, 0.S. Environmental
     Protection Agency, Cincinnati, OH, pp. 276-308.
 4.  Lofroth, G. (1980) in Health Effects  of Diesel Engine Emissions, Vol.  I,
     Pepelko, W.E., Danner, R.M.  and Clarke, N.A. ed., EPA 600/9-80-057a, D.S.
     Environmental  Protection Agency,  Cincinnati, OH, pp. 327-358.
 5.  Tokiwa, H., Nakagawa, R., Morita,  K.  and Kamachi. (1979) Japan Abstr.  pp.
     18.
 6.  Schuetzle, D., Lee, F.S., Prater,  T.J. and Tejada, S.B.  (1981) Int.  J.
     Environ. Anal. Chem. 9,  93-145.
 7.  Mermelstein, R., Kiriazidea, D.K., Butler, M., McCoy, E.C. and
     Rosenkranz, H.S. (1981)  Mutation  Res. 89, 187-196.
 8.  Rosenkranz, H.S., McCoy, B.C., Sanders, D.E., Butler, M., Kiriazides,  D.K.
     and Mermelstein, R. (1980) Science 209, 1039-1043.
 9.  Rosenkranz, H.S., McCoy, E.C., Mermelstein, R. and Speck, T. (1981)
     Mutation Res.  91, 103-105.
10.  Lewtaa Huisingh, J. (1981) Bull.  NY Acad. Med. 57, 251-261.
11.  Huisingh, J.L., Bradow,  R.L., Jungers, R.H., Harris, B.D., Zweidinger,
     R.B., Gushing, K.M., Gill, B.E. and Albert, R.E. (1980) in Health Effects
     of Diesel Engine Emissions,  Vol.  II,  Pepelko, W.E., Danner, R.M. and
     Clarke, N.A. ed., EPA-600/9-80-057b,  D.S. Environmental Protection
     Agency, Cincinnati, OH, pp.  788-800.
12.  Hasselblad, V., Stead, A., Creason, J. and Kasica, V. (1980)  Users Guide:
     The Ames Test  Curve Fitting  Program,  EPA-600/2-80-1 84,  D.S Environmental
     Protection Agency, Research  Triangle  Park, NC, pp. 1-55.
13.  Stead, A.G. , Hasselblad, V., Creason, J.P. and Claxton, L.  (1981)
     Mutation Res.  85, 13-27.
14.  Ames, B.N., McCann, J. and Yamasaki,  E. (1975) Mutation Res.  31, 347-364.
15.  Claxton, L.C.  (1980) in Health Effects of Diesel Engine Emissions, Vol.
     II, Pepelko, W.E., Danner, R.M. and Clarke, N.A. ed., EPA-600/9-80-057b,
     D.S. Environmental Protection Agency, Cincinnati, OH, pp. 801-809.
16.  Claxton, L.D.  and Huisingh,  J.L.  (1980) in Pulmonary Toxicology of
     Respirable Particles, Dept.  of Energy Symposium Series 53,  CONF-791002,
     pp. 453-465.
                              325

-------
17.  Claxton, L.D. and Kohan, M.  (1981)  in Short-Tenn Bioassays in the Analysis
     of Complex Environmental Mixtures 1980,  Waters,  M.D., Sandhu, S.,
     Huisingh, J.L., Claxton, L.  and Nesnow,  S.  ed.,  Plenum Press, New York,
     pp. 299-318.
18.  McCalla, D.R. and Voutsinos, D. (1974) Mutation  Res.  26,3-16.
19.  Mortelmans, K.E. (Personal Communication).
20.  Clive, D. and Spector, J.F.S.  (1975)  Mutation Res.  31, 17-29.
21.  Clive, D., Johnson, K.E., Spector,  J.P.S.,  Batson,  A.B. and Brown, M.M.M.
     (1979) Mutation Res. 59, 61-108.
22.  Mitchell, A.D., Evans, EoL., Jody,  M.M.,  Riccio,  E.S., Mortelmans, K.E.
     and Simmon, V. (1980) in Health Effects  of  Diesel Engine Emissions, Vol.
     II, Pepelko, W.E., Danner, R.M. and Clarke, N.A.  ed., EPA 600/9-80-057b,
     0.S. Environmental Protection Agency, Cincinnati, OH, pp.  810-842.
23.  Cifone, M.A. and Brusick, D.J. (Personal  Communication).
24.  Schechtman, L.M. and Kouri,  R.E. (1977)  in  Genetic  Toxicology, Scott,
     D., Bridges, B.A. and Sobels, F.H.  ed. , Elsevier/North-Holland Biomedical
     Press, New York, pp. 307-316.
25.  Barker, R.M., Bumette, D.M., Mankovitz,  R.,  Thompson, L.B.,  Whitmore,
     G.F., Siminovitch, L. and Till, J.E.  (1974) Cell  1, 9-21.
26.  Curren, R.D., Schechtman, L.M., Kim,  C.M. and Kouri,  R.E.  (1980)
     in Health Effects of Diesel Engine  Emissions,  Vol.  II, Pepelko, W.E.,
     Danner, R.M. and Clarke, N.A. ed.,  EPA-600/9-80-057b, U.S.  Environmental
     Protection Agency, Cincinnati, OH,  pp. 861-873.
27.  O'Neill, J.P., Brimer, P.A., Machanoff, R., Hirsch, G.P.  and  Hsie, A.W.
     (1977) Mutation Res. 45, 91-101.
28.  Casto, B.C., Hatch, G.G. , Huang, S.I.., Huisingh,  J.L., Nesnow, S. and
     Waters, M.D. (1980) in Health Effects of  Diesel Engine Emissions, Vol. II,
     Pepelko, W.E., Danner, R.M. and Clarke, N.A.  ed., EPA-600/9-80-057b, U.S.
     Environmental Protection Agency, Cincinnati,  OH,  pp.  843-860.
29.  Li, A.P. and Royer, R.E. (In Press) Mutation  Res.
30.  Chescheir, G.M., Garrett, N.E., Shelburne, J.D. ,  Lewtas Huisingh, J. and
     Waters, M.D. (1981) in Short-Term Bioassays in the Analysis of Complex
     Environmental Mixtures, Waters, M.D., Sandhu,  S., Huisingh, J.L.,
     Claxton, L. and Nesnow, S. ed., Plenum Press,  New York,  pp. 337-350.
31.  Brusick, D.J. and Mayer, V.W.  (1973)  Environ.  Health  Perspect. 6,83-96.
32.  Swenberg, J.A., Petzold, G.L. and Harbach, P.R.  (1976)  Biochem Biophys.
     Res. Commun. 72, 732-738.
33.  Casto, B.C., Janosko, N., Meyers, J.  and  DiPaolo, J.A.  (1978) Proc.  ftm.
     Assoc. Cancer Res. 19, 83.
34.  Williams, G.M. (1980) in Chemical Mutagens, 6, de Serres,  F.  and
     Hollaender, A. ed., Plenum Press, New York, pp. 61-79.
35.  Perry, P. and Evans, H.J. (1975) Nature 258,  121-125.
36.  Huang, S.L. (Personal Communication).
37.  McKenzie, W.H., Knelson, J.H., Rummo, N.J.  and House,  D.E.  (1977)
     Mutation Res. 48, 95-102.
38.  Ma., T.H., Anderson, V.A. and Sandhu, S.S.  (1981) in  Short-Term Bioassays
     in the Analysis of Complex Environmental  Mixtures 1980,  Waters, M.D.,
     Sandhu, S., Huisingh, J.L., Claxton,  L. and Nesnow, S.  ed., Plenum Press,
     New York, pp. 351-358.
39.  Schairer, L. (Personal Communication).
40.  Schuler, R.L. and Niemeier, R.W. (1980) in Health Effects  of  Diesel Engine
     emissions. Vol. II, Pepelko, W.E.,  Danner,  R.M. and Clarke, N.A.  ed.,
     EPA-600/9-80-057B, U.S. Environmental Protection  Agency,  pp.  914-923.
41.  Nix, C. (Personal Communication).
42.  Pereira, M.A. (1981) Induction of Sister  Chromatical  Exchanges by Diesel
     Emissions, Paper presented at 1981   Diesel Emissions Symposium, Raleigh,
     NC.  (See this volume).
43.  Rounds, D. (1981) Final EPA Contract No.  68-03-2945 report.

-------
44.  Russell, L.B., Generoso, W.M. , Russell, W.L. and OaJcberg, E.F. (In Press).
     Evaluation of Mutagenic Effects of Diesel Emissions I.  Tests for
     Heritable and Germ-cell Effects in the Mouse.  U.S Environmental
     Protection Agency.
45.  Casto, B.C. (1976) Chemico-Biological Interactions 13, 105-125.
46.  Casto, B.C. (1973) Cancer Res. 33, 819-824.
47.  Miller, B.C.  (1978) Cancer Res. 38, 1479-1496.
48.  Nesnow, S., Triplett, L.L. and Slaga, T.J.  (1981) in Short-Term Bioassays
     in the Analysis of Complex Mixtures 1980, Waters, M.D., Sandhu, 5.,
     Buisingh, J.L., Claxton, L.  and Nesnow, S.  ed.. Plenum Press, New York,
     pp. 277-298.
49.  Nesnow, S., Evans, E., Stead, A.  , Creason, J., Slaga, T.J. and Triplett,
     L.L.  (1981) Skin  Carcinogenesis Studies of  Emission Extracts, Paper
     presented at  1981 Diesel Emissions Symposium, Raleigh, NC.  (See this
     volume).
50.  Nishioka, M.G., Peterson, B.A. and Lewtas,  J. (Unpublished Results).
                            327

-------
GENOTOXICITY OF DIESEL EXHAUST  EMISSIONS IN LABORATORY ANIMALS
MICHAEL A. PEREIRA
U.S. Environmental Protection Agency,  Health Effects Research Laboratory,
Cincinnati, Ohio, USA  45268
INTRODUCTION

      The number  of diesel powered  passenger cars  in the United  States  has
increased.  It has  been  estimated that by 1985 diesel powered passenger cars
will comprise ten to 25 percent of new cars.1  The exhaust emissions of diesel
powered passenger cars produce 30 to 100  times more particulate matter than a
comparable gasoline engine with  catalytic converter.   These  particulates  are
composed  of  carbonaceous  material  onto  which a  complex mixture  of  organic
chemicals  have  been adsorbed.    Some of  the  organic  chemicals  are  known
carcinogens and mutagens while many  have  not  been tested for  carcinogenic  and
mutagenic activity.2  The assessment of the human health hazard, if  any, due to
the  increased  number of diesel  powered cars  requires the evaluation of  the
genotoxic activity of the complex mixture of organic chemicals  adsorbed onto  the
particles.
    Is  this communication,  I shall describe  the work in progress at the Health
Effects  Research  Laboratory in  Cincinnati,  Ohio,  to  evaluate  the genotoxic
activity of diesel exhaust emissions.  The complex mixture of organics adsorbed
onto  diesel  exhaust  particles  is   being tested  in  laboratory rodents  for
mutagenic and  clastogenic  activity.   We are comparing the  following  three
different types of  exposure to these organics, 1)  intraperitoneal  (i.p.)  and
intratracheal  administration of methylene  chloride extract  of particles, 2)
i.p. and  intratracheal  administration of particles and 3) inhalation of  the
exhaust emissions. The methylene chloride extract should be the most genotoxic
of the types of exposure and inhalation of the particles the least.  Therefore,
the methylene chloride extract  was used  to determine, under optimal conditions,
whether the particle contain a  significant amounts of genotoxic chemicals to be
detected.  The inhalation  studies  were  performed  in  order  to determine  the
genotoxic effect under conditions by  which humans might be exposed.   Exposure to
the particles provided an estimate of the ability of the organics to  be released
in vivo from the particles  in order to exert a  genotoxic activity. The activity
of the particles  should  be intermediate between  the extract  and the exhaust.
                              328

-------
The  genotoxic  assays  performed  include,   1)  Ames  Salmonella  mutagenicity
with/without metabolic  activation of urine,  2)  micronuclei in  polychromatic
erythrocytes, 3) sister chromatid exchange in bone marrow cells, fetal hamster
liver exposed in utero and primary lung cultures from hamster exposed j^n vivo,
4) metaphase analysis in bone marrow cells and 5) sperm morphology and motility.

MATERIALS AND METHODS
Generation of diesel exhaust-^
   Diesel  emission was produced  by  one of two Nissan CN-6  diesel  6  cylinder
engines coupled to  a Chrysler  torque-flite  automatic  transmission Model A-727
and mounted on an Eaton-Dynamometer Model 758-DG.  The engines were operated by
Federal Short Cycle type driving modes.  The exhaust was diluted with air 1:9 so
that  it contained  about 12 mg/ra^ particulate matter.  In  earlier studies  the
exhaust was diluted 1:18  so  that  it contained  6-7  mg/m^.
Preparation of diesel particulate and extract
   The particles were collected  on teflon coated  pallflex T60A20 type filters
(Pallflex  Products  Corp.) and extracted  for 24 hr.  in  a soxhlet  extraction
apparatus using methylene chloride (Fisher Chemical  Co., Pittsburgh, PA) as  the
eluant.  The eluant was filtered through a fluoropore  filter  (Millipore Corp.)
backed by a microfiber glass disc  (Millipore Corp.).  The mass of the extract  was
obtained  by gravimetric  determination  of  an  aliquot of  the  filtrate  after
blowing off the methylene chloride with nitrogen.   The rest of the filtrate  was
made  up to  the  desired  concentration in dimethyl sulfoxide  (DMSO)  by solvent
exchange  using  a stream  of  nitrogen to  remove  the methylene chloride.   All
extractions and processing procedures required to obtain the diesel  extract in
DMSO  were performed under yellow  lights.
   Animals.   The animals were  maintained in accordance with the standards  set
forth in the "Guide  for the Care and Use  of Laboratory Animals" of the Institute
of Laboratory Animal Resources, National Research Council.  They received water
and Purina Laboratory Chow (Ralston Purina Co., St. Louis, MO)  ad libitum.  The
animals were exposed to diesel  exhaust emissions in chambers 8 hrs/day from 7:00
a.m.  to 3:00 p.m. and 7 days/week.
Micronucleus assay  '
   The animals were sacrificed by  cervical dislocation and the bone marrow cells
harvested by the method of Schmidt.4  Briefly,  the  intact femurs were removed
and cleaned of all  muscles.  For  the mice,  both proximal  and distal ends were
snipped off.    Due  to  the limited yield  of bone  marrow cells  from  Chinese
hamsters, the ends  of the femurs  were not snipped  off.   About  0.2 ml of fetal
                             329

-------
bovine serum  (FBS) was aspirated into a 1 ml tuberculin syringe fitted with a 23
gauge needle. The needle tip was carefully inserted into the proximal end of the
femur, and  after  the  femur was completely submerged  in FBS contained  in small
test tube, the bone marrow was gently aspirated.  The needle was removed from the
syringe, and  FBS containing the bone marrow cells was slowly injected back  into
the tube.  The syringe was  then inserted into  the distal end and the aspiration
repeated.    The bone  marrow  suspension from both  femurs was combined  and
centrifuge at 1000 rpm for  5 minutes.  The pellet was gently suspended in a drop
of FBS and  spread  with a coverglass on a slide.
   The  slides  were  air-dried and  stained  within   24 hours.   The staining
procedure included:  3 minutes  in 0.2% (w/v)  May Grunwald dissolved in methanol;
2 minutes in 0.2% (w/v) May Grunwald diluted with an equal  volume of H20; a rinse
in H20; 10 minutes in  Giemsa  (1:6, Gurrs Improved R66 Giemsa:H20);  and  a final
rinse in H20.  The slides were blotted dry, cleared in xylene for 5  minutes and
immediately covered with a cover  glass using  Pro-Texx.   All slides  were coded
prior  to evaluation.   For each  animal,  the number  of  micronucleated poly-
chromatic erythrocytes in  1000 such cells was determined  as recommended  by von
Ledebur and Schmid.6
Sister chromatid exchange  (SCE) assay^'^
   Bone marrow cells.   A slight modification  of the procedure  of Allen et al.9
for  in  vivo sister chromatid  exchange was used.   Briefly,  twenty-four hours
prior to sacrifice a 60 mg pellet of 5-bromo-2-deoxyuridine (BrdU) was implanted
under the skin  between the scapulae.  Two hours prior to  sacrifice  the animals
were injected intraperitoneally with colchicine (10 rag/Kg  bd. wt.).   The animals
were sacrificed by cervical dislocation and  the intact  femurs removed.  All
muscle was removed from the bone and the proximal end of the bone gently snipped
off.  The bone marrow cells were flushed from  the canal with 0.075 M KC1  and the
contents from both femurs  combined.
   The  cell  suspension  was incubated  in  a  37°C water bath,  for  30 minutes,
followed by centrifugation at 1,000 g  for   10  minutes.  The  supernatant  was
discarded.    The  cells  were  fixed in  ice-cold Carney's  solution  (3:1  v/v
methanol:glacial acetic acid) .  After 20 minutes at 4°C the fixative  was removed
by centrifugation. The fixation process  was repeated two more  times. The final
cell pellet  was resuspended  in a  small volume of the ice-cold fixative  and
dropped onto  cold, wet slides.   The  slides were air-dried  in a  dust free
atmosphere overnight.
   The Hoechst-Giemsa  black-light method described by Goto et  al.10  was  used to
stain the cells. The slides were rinsed in double distilled water, stained with
                               330

-------
Hoechst 33158 dye  (50 ug/ral) for 15 minutes, rinsed again  in water, and blotted
dry.   Next,  the slides  were  placed on a slide warmer  (50°C),  covered  with a
coverglass using Macllvaines buffer and exposed to black light for 22 minutes.
Then the coverglass was removed and the slides allowed to dry for at least 2 hrs.
Counter staining was done in 6% Giemsa for 10 minutes.  The coverglass was then
remounted onto the microscope  slide using Pro-Texx mounting  medium.
    All slides were coded prior to evaluation.  Twenty-five metaphases from each
animal were evaluated for the number of SCE.  The  mitotic  index was determined
by counting the number of dividing cells  per  1,000 cells.
    Lung cells
    Using  the  intratracheal  instillation procedure  of  Saffiotti  et  al.,11
animals  were  administered  either  1) diesel particles,  2)  dichloromethane
extract of diesel particles absorbed onto  carbon black or  3)  benzo(a)pyrene
absorbed onto carbon black.  The  samples were suspended in Hank's balanced salt
solution  (BBSS) containing  20% (v/v)  Emalphor El-620.
    The animals were sacrificed by cervical disolocation and the heart and lungs
quickly excised.   The  lung  tissue was finely minced with  sterile scissors  in
McCoy's  5A medium suplemented  with 10% fetal  bovine serum,  100  units  of
penicillin and 100 ug streptomycin/ml.  The minced  tissue was applied to a Petri
dish and incubated overnight in a 5% CO^ incubator at 37°C.  The unattached cells
and tissue  fragments were  removed,  washed  in HBSS  twice  in order  to  remove
erythrocytes and  cellular  debris, and then distributed onto  three additional
Petri dishes.  The  attached cells in the original dish were washed once with HBSS
to  remove cellular  debris  and  then  incubated with  complete McCoy's  medium
supplement with 10%  fetal bovine serum.
    When  the  cultures showed colonies containing  50 or more  cells,  they were
treated with BrdO (10 ug/ml) in subdued illumination.  The dishes were  incubated
in the dark for an additional  40 hrs.  At  that time,  the cultures were treated
with 0.05 ug/ml colcemid (Grand Island Biological  Co.)  for  four hrs.  The cells
were then suspended by trypsinization (0.05%  trypsin  in calcium  and  magnesium
free HBSS) and treated for  30  minutes at  37°C in a hypotonic  solution of 0.075
M KC1.  Finally  the cells were fixed in ice-cold Carnoy's solution and stained  by
the Hoechst-Giemsa black-light procedure  described above.
Fetal liver cells  exposed in utero
   On day 12 of gestation,  pregnant Syrian  hamsters were  treated with BrdU  by
implanatation of a pellet under the skin.  The diesel particulates, extracts  or
benzo(a)pyrene were administered  two  hrs. later by intraperitoneal injection.
Eighteen hours after  the  BrdU was given, the animals were sacrificed by cervical
                               331

-------
dislocation and the fetuses quickly removed.  Fetal livers from each litter (10-
14 fetuses) were pooled in 10 ml of calcium and magnesium free  BBSS.  After 20
minutes of incubation at 37°C, the  livers were gently crushed  with  a spatula
against the side of the flask to release the cells  from the  connective tissue.
The suspension was treated with colcemid for  2 hrs. at 37°C in a 2 incubator.
The cells were collected by centrifugation and resuspended in 5-6 ml of 0.075 M
KC1.   After  incubation  for 10 minutes  at 37°C,  the  cells were collected  by
centrifugation.  The collected cells were fixed with ice-cold Carney's solution
and stained by the Hoechst-Giemsa black-light method as  described  above.

RESULTS
Micronucleus assay
   The number of micronucleated polychromatic erythrocytes in mice and Chinese
hamsters  exposed  to diesel exhaust  emissions  (12  mg/m3  particulate) for  one
month or  administered 640 mg/kg bd. wt. diesel  particulate was not  different
from the corresponding controls of clean air and DMSO controls (Table  1). There
appeared to be a slight increase in micronucleated cells in mice but not Chinese
hamsters  administered an extract of diesel  particulate   (1000 mg/kg  bd.  wt.).
The clastogenic activity of the  organic chemicals  present on diesel exhaust
particulate would appear  to be below the sensitivity of the micronucleus assay.
Sister chromatid exchange in  bone marrow cells
   Exposure of mice to diesel exhaust emissions  (12 mg/m3 particulate) for one
month did not induce SCE in bone marrow cells (Table 2).   The administration of
either diesel particulates (300 mg/kg bd.  wt.) or their extract (800  mg/kg bd-
wt.) resulted in an increased incidence of SCE in mice sacrificed two days post-
treatment.  The increase  above DMSO  controls in  the incidence of SCE  was 5.64,
3.68 nd 7.82  SCE/metaphase for diesel particulate, diesel extract and benzo-
 (a)pyrene  (100 mg/kg bd. wt.) ,  respectively.  The dose of diesel particulate and
extract administered  to the mice was only three and eight times the dose  of
benzo(a)pyrene.  It would appear  that the activity of benzo(a)pyrene was only
4.2 fold more active than diesel particulate and  17 fold for diesel extract.  It
should  be recognized that  these  calculations  are tenuous  and  need  to  be
substantiated  by  comparison  of  the  dose-response  relationship  for  diesel
particulate and extract  to the  dose-response relationship for benzo(a)pyrene.
    Sister chromatid exchange  in lung cells.  While exposure of Syrian hamsters
to diesel exhaust  emissions  (6-7 mg/m3  particulate) for  three  months did not
increase  the  incidence  of  SCE   in lung  cells,  the  exposure  to  12  mg/ro
particulates  for  3.5  and 8.5  months  did  induced  SCE.   The  intratracheal
                              332

-------
TABLE 1

MICRONUCLEUS ASSAY IN MICE AND  CHINESE  HAMSTERS3
Experiment
Treatment
 Micronucleated Polychromatic Erythrocvtes
   Mice                       Chinese Hamsters
Experiment I

Diesel Exhaust
  a.  1 month

Clean Air

Water

Cyclophosphamide
  a.  1000 rag/kg  bd wt

Experiment II

Diesel Particulate
  a.  640 rag/kg bd wt

DMSO

Cyclophosphanide
  100 mg/kg bd wt

Experiment III

Diesel Extract
  a.  1000 mg/kg  bd wt

DMSO

Cyclophosphamide
  a.  100 mg/kg bd wt
0.59 + 0.07 (10)b

0.60 + 0.16(10)

0.54 + 0.13(5)


2.02 _ 0.43(5)




0.10 + 0.05(6)

0.43 + 0.07(6)


5.77 + 0.58(6)




0.80 + 0.20(6)

0.32 + 0.06(6)


3.83 + 0.19(6)
0.49 + 0.07(10)

0.48 + 0.10(10)
0.23 + 0.10(6)

0.20 + 0.04(6)


5.38 + 0.80 (6)




0.25 + 0.04(6)

0.12 + 0.03 (6)


5.58 + 0.50(6)
Preliminary  results  from M.A.  Pereira (U.S.  EPA,  HERL-Cincinnati,  OH)  and
  P.S.  Sabharwal (Environmental Health Research and Testing, Inc., Cincinnati,
_OH).  Mico (36C3F1) were  from  Harlan,  Indianapolis,  IN  and Chinese hamsters
 fron northeastern Univ.,  Boston, MA.
Percentage of  micronucleated  polychromatic erythrocytes presented  as the
  mean  +  SE for  groups containing the number of animals in parenthesis.
administration of either diesel particulate or diesel extract resulted in a dose

related  increase  in  SCE.  The activity of diesel extract was  approximately  10

times greater than the  activity of  diesel  particulate.   Since the  recovery  of

the mass of the particulate in the extract  was on the  average  23%,  it appeared

that  at  least  a  third  of  the genotoxic  material  adsorbed onto  the  in-

tratracheally  instilled particulate  was available.   The genotoxic chemicals

-------
adsorbed onto the particles in diesel exhaust emissions were available to exact
their activity  in lung cells  when administered by all three types of exposure
studied, i.e. inhalation,  particulate and extract.
TABLE 2
INDUCTION OF SISTER CHROMATID EXCHANGE IN BONE MARROW CELLS OF MICE3
Exper iment
                                N              SCE per Metaphase
Treatment
Experiment  I
Diesel Exhaust
  a.  6 month                   10               4.57 + 0.39(10)b
Clean Air                        6               4.17+0.42
Corn Oil                         6               5.87 + 0,47
Benzo(a)pyrene
  a.  100 mg/kg bd wt            6              13,16 + 1.02
Experiment  II
Diesel Particulate
  a.  300 rag/kg bd wt            6              11.31 +0.94
Diesel Extract
  a.  800 mg/kg bd wt            6               9.35+0.60
DMSO                             6               5.67+0.45
Benzo(a)pyrene
  a.  100 rag/kg bd wt            6              13.49 4- 1.01

Preliminary results from M.A.  Pereira  (U.S.  EPA,  HERL-Cincinnati, OH)  and
 P.S. Sabharwal  (Environmental  Health Research and Testing Inc., Cincinnati.
 OH). Mice  (B6C3F1) were from Harlan, Indianapolis,IN.
bResults are means + SE.
Sister chromatid  exchange in fetal  liver  exposed in utero.   The  exposure of
pregnant Syrian hamsters from day one of gestation to diesel exhaust emissions
(12 mg/m3) or to diesel particulate at the LD5o (300 mg/kg bd.  wt.)  on day 12 of
gestation, did not increase the  incidence of SCE in fetal liver when determined
on day 13 (Table 4) .  The administration of diesel extract on day 12 of gestation
resulted in a dose-dependent increase in SCE in the fetal liver on day 13.  This
would indicate  that once  the organic chemicals absorbed onto the particles in
diesel exhaust are  eluted, they  are  capable  of crossing the placenta and
exerting  a  genotoxic  effect.  Metabolic  activation,  if  necessary,  of the

-------
TABLE 3

INDUCTION OF SISTER CHROMATID EXCHANGE  IN  PRIMARY  LUNG CULTURES  FROM SYRIAN

EXPOSED IN VIVO3
Experiment
Treatment
 SCE per  Metaphase
Experiment I

Diesel Exhaust
  a.  6 rag/m3;  3 months

Clean Air
  a.  3 months

Experiment II

Diesel Exhaust
  a.  12 mg/m3; 3.5 months
  b.  12 mg/m3; 8.5 months

Clean Air
  a.  3.5 months
  b.  8.5 months

Experiment III

Diesel Particulates
  a.  0
  b.  44 mg/kg
  c.  87 mg/kg
  d.  130 mg/kg

Benzo(a)pyr ene
  a.  20 ug/hamster

Experiment IV

Diesel Extract
  d.  0
  b.  3.3 mg/kg
  c.  6.6 mg/kg
  d.  13.3 mg/kg

Benzo(a)pyrene
  a.  20 ug/hamster
11.86 + 0.47(5)b


11.52 + 0.58 (8)
19.41 + 1.03(12)
22.22 + 1.02(4)
10.94 + 0.43(12)
 9.03 + 0.38(5)
10.74 + 0.17(5)
12.95 + 0.41(5)
15.37 + 0.21(5)
18.18 + 0.65(5)
17.16 -I- 0.50(5)
10.47 + 0.32(5)
13.90 + 0.78(5)
16.18 + 0.21(5)
19.96 + 0.92(5)
16.34 + 0.47(5)
aResults from D.E.  Rounds  (Pasadena  Foundation  for  Medical  Research,  Pasadena,
 CA) Final  Report  Contract No.  68-03-2945 with  U.S. EPA HERL-Cincinnati,  OH
 (Project Officer:   John G.  Orthoefer).   Syrian hamsters were from Engle Lab,
 Farmersburg, IN.
bResults are means _+ SE  for groups containing the number of animals in paren-
 thesis.
                              335

-------
genotoxic chemicals could have occurred either maternally prior to crossing the
placenta or in the fetus.  It also appeared that sufficient  toxic  material was
eluted  from the  administered particles to cross the placenta  and  decrease the
mitotic index  in the  fetal  liver.

DISCUSSION
    The  genotoxic activity  of diesel  exhaust emissions was evaluated in the
micronucleus  and sister chromatid  exchange  assays.   The  micronucleus assay
measures the ability  of a substance to either cause pieces  of chromatin to be
severed  from  chromosomes or  to disrupt the  spindle  apparatus.5,6,12,13  particulate >
inhalation.   The Syrian hamsters  were  exposed  to diesel  exhaust  emissions
containing  12  mg/m3  particles  for  8  hrs/day,  7 days/week. A Syrian  hamster
weighing about 90 gm inhales  approximately 0.06 liters/min.,  so the exposure in
3.5  months  would  result in  the maximum  accumulation of  389 mg/kg  bd.  wt.
particulate.  The maximum accumulated dose induced an equivalent number of SCE
compared to an extract equivalent to 58 mg particulate/kg bd. wt. as calculated
with 23%  being  the recovery of the mass  of  the particulate  in  the extract.
Therefore, it would appear that at least 15% of the genotoxic material adsorbed
onto inhaled diesel particles was available to exert its activity in the lung.
The amount of available genotoxic material was probably higher, since  not all of
the inhaled particles  were deposited  in the lungs.
                             336

-------
TABLE 4
SUMMARY OF THE  EFFECT ON SISTER CHROMATID EXCHANGE AND MITOTIC INDEX IN
HAMSTER FETAL LIVER CELLS EXPOSED IN UTERO3
Treatment
Diesel Exhaust
Diesel Particulate
Diesel Extract
SCE per Metaphase
no effect
no effect
increase
Mitotic Index
no effect
decrease
decrease
 ^Results  are  from M.A.  Pereir-a (U.S.  EPA,  HERL-Cincinnati,  OH)  and P.S.
  Sabharwal (Environmental Health Research and Testing, Inc., Cincinnati,  OH)
  and  have been submitted for publication

    The  ability of  the  genotoxic material  in diesel exhaust emissions to  be
distributed  systemically was evaluated by the micronucleus  and  SCE assays  in
bone marrow calls.   The micronucleus  assay was not sensitive enough to detect
genotoxic activity  by  the organics  adsorbed onto  diesel  exhaust  particles
administered  either  by  inhalation or by intraperitoneal injection of particles
or  extract.   In the  SCE assay a six month exposure of mice  to  diesel  exhaust
emissions did  not increase the incidence of SCE in bone marrow cells, while the
intraperitoneal  administration of  diesel  particulate or  diesel extract did
induce SCE.   The  assumption  of a linear  dose-response relationship for diesel
particulate to the increased incidence of SCE  in bone marrow cells would appear
reasonable, since a  linear dose-response relationship was found in lung cells.
This linear extrapolation predicts that an  i.p. dose of 75 rag/kg bd. wt. diesel
particulate would have  been detected.  The six month exposure in mice that did
not increase  the  incidence of SCE  (assuming  that a 30 gm mouse  inhales  0.02
liters/min.)  would have resulted in the  maximum accumulation of 667 mg/kg bd.
wt. particulate.  This  dose  of inhaled particles was  ten times  the predicted
minimum single systemic dose  of  75  mg/kg  bd.  wt.  particulate to which  the SCE
assay in bone marrow cells would  tie  sensitive.  Therefore, a sufficient  dose  of
genotoxic  materials  was administered  by  inhalation to possibly  be detected  by
the SCE  assay  in bone  marrow cells. However,  under the  conditions  of the
experiment,  the  SCE assay in  bone  marrow was unable  to demonstrate that the
genotoxic  material  adsorbed  to  inhaled  particles was available  for systemic
distribution.  This  is  in contrast to direct  application of  the  particles  to
lung cells by intratracheal  instillation  and inhalation where  most  of the
genotoxic material was  available  to induce  SCE in lung cells.
                             337

-------
   The  induction of heritable mutations,  teratogenesis and  embryotoxicity by
the organic material adsorbed onto diesel particles requires that the material
be  distributed  systemically  and when  appropriate,  cross  the  blood-testes
barrier or the placenta.  The SCE assay in bone marrow cells  is  more sensitive
than the  specific locus assay20  used to determine the mutagenic hazard of a
chemical.   Therefore,  one would  predict that  the  genotoxic  activity  of a
substance  (inhalation  of  Diesel  Exhaust  Emissions)  which  was  below  the
sensitive of the SCE assay  in bone marrow cells, would be too low to be detected
in the specific locus assay as a mutagenic hazard.  Our evidence  would indicate
that an insufficient amount of the genotoxic material adsorbed to particles in
diesel exhaust emissions (as determined by the SCE assay in bone marrow cells)
was available for systemic  distribution to the reproductive organs to represent
a measurable mutagenic  hazard.
   When the genotoxic material was administered as an extract, the material did
cross  the placenta and  induce  SCE  in  fetal  liver.    Intraperitoneal  ad-
ministration of diesel  particulate and inhalation of diesel  exhaust emissions
did not  increase the  incidence of  fetal  SCE.    We were  therefore  unable to
demonstrate  that the genotoxic material  when administered adsorbed onto the
particles reached the fetus.  This distribution to  the fetus  would be required
for the genotoxic material  to represent a teratogenic hazard.  In  conclusion, we
were unable  to support  a  possible  mutagenic or  teratogenic hazard  for  the
exposure  in  laboratory animals of the particles  in diesel exhaust emissions.
Therefore, if a genotoxic hazard exists for this exposure it would appear to be
limited to the lung where  it might cause  cancer.

ACKNOWLEDGEMENTS
   The  author  gratefully acknowledges the  use of  preliminary results of the
collaborators  Drs.  Pritam  S.   Sabharwal,  Environmental Health  Research  and
Testing Inc.; Donald E. Rounds, Pasadena  Foundation for Medical  Research; and
John G. Orthoefer,  Health Effects Research  Laboratory, U.S.  EPA.

REFERENCES
 1. U.S.  Environmental  Protection Agency  (1978) Health Effects Associated with
    Diesel Exhaust  Emissions,  U.S. EPA-600/1/78-063.
 2. Huisingh, J., Bradow, R.,  Jungers, R., Claxton, L.,  Zweidinger, R., Tejada,
    S.,  Bumgarner,   J. ,  Duffield,  F., Waters,  M.,  Simmon, V.F.,  Hare,  C.,
    Rodriquez, C.,  and  Snow, L.  (1978) in Application of Short-Term Bioassays
    in  the  Fractionation   and  Analysis  of  Complex Environmental  Mixtures,
    Waters, M. , Nesnow, S., Huising, J. , Sandhu, S.  and Claxton, L. eds., Plenum
    Press, New York, pp. 383-418.
                             338

-------
 3. Hinners, R.G. ,  Burkart, J.K. ,  Malanchuk,  M. and Wagner,  W.D.  (1980)  in
    Health  Effects  of  Diesel  Engine  Emissions; U.S.  Prceedings of  an In-
    ternation Symposium, Vol. 2.  Pepelko, W.E., Danner,  R.M.  and  Clarke, N.A.
    eds.; EPA-600/9-80-051b, pp.  681-697.
 4.  Schraid, W.  (1977)  in Handbook  of Mutagenicity Test Procedure, Kilbey,  B.J.,
    Legator, M. Nichols, W.  and Ramel, C. eds. Elsevier, Amsterdam,  pp. 235-242.
 5.  Jenssen, G. and Ramel,  C.  (1980)  Mutation  Res.  75, 191-202.
 6.  Von Ledebur, M. and  Schmid, W.  (1973)  Mutation  Res.  19, 109-117.
 7.  Latt, S.A., Allen, J.W., Rogers, W.E.  and Jeurgens, L.A. (1977) in Handbook
    of Mutagenicity Test Procedures,  Kilbey, B.J., Legator, M., Nichols, w. And
    Ramel, C. eds. Elsevier, Amsterdam, pp. 275-292.
 8.  Latt, S.A., Schreck, R.R.,  Loveday, K.S. and Shuler, C.F.  (1979) Pharmacol.
    Rev. 30, 501-535.
 9. Allen J.W., Shuler,  C.F. and  Latt, S.A.  (1978)  Somat.  Cell Genet. 4, 393-
    405.
10. Goto, K., Maeda,  S., Kano,  Y. and Sugiyama,  T.  (1978)  Chromosoma 66, 351-
    359.
11. Saffioti, U. , Cefis, F.  and Kolb, L.M.  (1968) Cancer  Res.  28,  104-124.
12. Miller,  R.C.  (1973)  Environ.  Health Perspec.  6,  167-170.
13. Tsuchimoto, T. and  Matter,  B.E.  (1979) Arch.  Toxicol.  42,  239-248.
14. Wolff,  S., Bodycote, J.  and Painter,  R.B.  (1974)  Mutation Res. 25, 73-81.
15. Kato, H. (1980) Cancer  Genet. Cytogenet.  2,  69-77.
16. Perry,  P. and Evans, H.J.  (1975)  Nature 258,  121-124.
17. Carrrano, A.V., Thompson, L.H.,  Lindl, P.A.  and Minkler, J.L. (1978) Nature
    271, 551-553.
18. Bradley, M.O., Hsu,  I.C. and  Harris,  C.C.  (1979)  Nature 282,  318-320.
19. Kinsella, A. and Radman, M.  (1978) Proc. Natn. Acad. Sci. (U.S.A.)  75,  6149-
    6153.
20. Russell, L.B.  (1978) Environ. Health  Perspec. 24, 113-116.
                             339

-------
HUMAN CELL MUTAGENICITY OF POLYCYCLIC AROMATIC HYDROCARBON  COMPONENTS OF DIESEL
EMISSIONS

THOMAS R. BARFKNECHT+>++44 RONALD A HITEs""", ERCOLE L. CAVALIERS"1^", AND
WILLIAM G. THILLY+'
"toxicology Group,  Department of Nutrition and Food Science, Massachusetts
Institute of  Technology,  Cambridge, Massachusetts 02139; "^Department of
Chemistry, School of  Public and Environmental Affairs, Indiana University,
Bloomington,  Indiana  47405; +++Eppley Institute for Research in  Cancer,
University of Nebraska  Medical Center, Omaha, Nebraska 68105; ++"H"Current
address, Chemistry  and  Life Sciences Group, Life Sciences and Toxicology
Division, Research  Triangle Institute, P. 0. Box 12194, Research Triangle Park,
North Carolina  27709.
INTRODUCTION
   In an earlier report  by  Liber et al.  ,  it was shown that a methylene chloride
extract of automobile  diesel  exhaust particulate was significantly mutagenic to
both Salmonella typhimurium and  diploid  human lymphoblasts in the concentration
range of 50-100 yg/ml  when  treatment was in conjunction with an Aroclor-induced
rat liver microsome-containing postmitochondrial supernatant (PMS).  This die-
sel exhaust particulate  extract  also contained "direct-acting" mutagans for S.
typhimurium.  However, no "direct-acting"  mutagenic activity was detected by
the human lymphoblast  mutation assay.
   Fractionation of  the  methylene chloride extract was performed and the
seven resultant fractions were tested for  their mutation inducing ability in the
—" typhimurium 8-azaguanlne resistance forward mutation assay,  with and without
rat liver PMS. '   A hexane/toluene fraction, which represents 6.5% by weight of
the total methylene  chloride  extract, contained the polycylic aromatic hydro-
carbons (PAH) and was  found to be the most mutagenic fraction to _S_. typhinurlua
with PMS activation. '   Although others have reported that organic extracts of
diesel exhaust particulate  from  various  sources contain "direct-acting" mutagens
                   3 4
for mammalian cells  '  ,  we  have  placed an  emphasis upon determining what PAH nay
be responsible for the mutagenicity of our diesel exhaust particulate sxtract to
human lymphoblasts.  In  addition,  we have  initiated studies to determine the
mutagenicity of PAH  and  their derivatives  that are found in other organic ex-
tracts of emissions  collected from diesel  engines.
MATERIALS AND METHODS
   Diesel exhaust particulate  sample.   The diesel exhaust extract sample utili-
zed in these experiments was a gift of Dr. Morton Beltzer of Exxon Research and
Engineering Company, Linden, New Jersey.   The diesel exhaust particulate was
                               340

-------
 collected  from a 1978 Oldsmobile 350 CID engine burning blended commercial No.
 2  diesel fuel operated with repetitive hot start Federal Testing Procedures.
 The  exhaust  was passed through a dilution chamber and the particulate was col-
 lected  on  a  Pallflex type T60A20 filter.  The filter was Soxhlet extracted with
 methylene  chloride for 16 hr followed by solvent evaporation on a steam bath
 with nitrogen purging.
   The  methylene chloride extract of the diesel soot was then fractionated
 based on polarity for bacterial mutagenicity testing.  Details of the fractiona-
 tion procedure have been published in a separate paper.    The PAH containing
 hexane/toluene fraction was analyzed by gas chromatography/mass spectrometry
 (GC/MS) for  the identification of individual PAH and derivatives.2
   Source  of PAH.   Benz(a)anthracene and phenanthrene were purchased  from
 Eastman Chemical Co.,  Rochester,  NY.  Benzo(a)pyrene was purchased from Sigma
 Chemical Co.,  St.  Louis,  MO.   Chrysene,  triphenylene, 1-methylphenanthrene,
 2-methylphenanthrene and  1-methylpyrene were obtained from ICN Life Sciences
 Group,  Planviev,  NY.   Anthracene, 2-methylanthracene, 9-methylanthracene,  9,10-
 dimethylanthracene,  fluoranthene and pyrene were obtained from Aldrich Chemical
 Co.,  Milwaukee,  WI.   The  9-methylphenanthrene was obtained as a gift  from M. L.
 Lee,  Bringham Young University.   The fluoranthene 2,3-dihydrodiol was  provided
 by W. Rastetter,  Massachusetts Institute of Technology.
   Cyclopenteno(c,d)pyrene  and its derivatives,  cyclopentano(c,d)pyrene (CPAP),
 CPAP-3,4-oxide,  CPAP-3,4-trans-diol and  CPAP-3,4-cis-diol were synthesized in
 the  laboratory of  E.  L. Cavalieri.   A detailed report of the synthesis of
 cyclopenteno(c,d)pyrene (CPEP)  and its derivatives will  be reported elsewhere.
   Source  of  PAH metabolizing  element.   Liver microsome  containing postmito-
 chondrial  supernatant  (PMS)  from Aroclor 1254-induced male Sprague-Dawley  rats
was  prepared  by  the method  of  Ames et al.   and purchased from Litton Bionetics,
Kensington, MD.   The  PMS  had a concentration of  29.3  mg  protein/ml  as  deter-
mined by the  supplier  and was  utilized at  a final concentration  of  5%  v/v
 (1.5 mg protein/ml)  in the  human  lymphoblast mutation assay.   The  PMS  was
radiosterilized at -80°C  and maintained at this  temperature  until  being thawed
immediately before use.
   Human lymphoblaat mutation  assay.   The  diploid human  lymphoblast cell line
TK6,  a presumptive heterozygote at  the thymidine  kinase  locus, was  utilized to
                                                       7  $
select mutants resistant  to trifluorothymidine (F^tdR).  '    The  cells were rou-
tinely maintained in  suspension culture  at  37°C  in RPMI  1640  culture medium
supplemented to 10% v/v with heat  inactivated  horse serum, both  purchased
from Flow Laboratories, Inc., McClean, VA.
                             341

-------
   A minimum of 2 x 10   exponential phase cells were treated per culture and
duplicate cultures were  treated  per experimental point.  Treatment with either
the diesel exhaust particulate extract or individual PAH was in the presence of
                                           Q g
rat liver PMS and the necessary  cofactors.  '    Following treatment at 37"C for
3 hr, the cells were pelleted by centrifugation, resuspended in fresh medium,
and counted to determine  the initial cell density.  At this time a small ali-
quot was also taken, diluted and plated to  determine the toxicity of the treat-
ment.
   Expression of resistance to F,tdR was allowed to develop for a minimum period
of three days before mutant selection occurred.   This amount of time is more
than sufficient to allow maximum phenotypic expression of the induced F,tdR
                  8
resistant mutants.
   F,tdR resistant mutants were  selected by diluting the cultures to a cell
    J            5
density of 2 x 10 /ml and trifluorothymidine  was added to give a final concen-
tration of 2 lag/nil.  An aliquot  of  cells was  also diluted to 10 cells/ml and
plated to determine cloning efficiency.   Four 96 well microtiter plates (Linbro
Scientific Inc., Hamden, CT) at  0.2 ml of cell suspension/well were plated to
select the F,tdR resistant mutants  and 2 plates  were plated at low cell density
(2 cells/well) to determine the  cloning efficiency of each culture.  Plates were
incubated at 37°C in a 5% CO, in air incubator for 14 days before scoring of
clones took place.
   Calculation of mutational data.   The use of the Poisson distribution and
associated calculations to determine the cloning efficiency and F.tdR resistant
mutant fraction for the TK6 human lymphoblast mutation assay has been re-
       O 1 Q
ported. '    The minimal concentration of a treatment required to induce a sig-
nificant mutant fraction is determined by interpolation on the concentration vs.
mutation frequency curves to the historic upper  99% confidence limit of the
spontaneous background mutation  frequency.   The  concentration at this inter-
polation point is taken as the minimal concentration which would induce a
statistically significant mutant fraction.   The  historic upper 99% confidence
limit for the spontaneous background mutation frequency for the TK6 human
lymphoblast F.jtdR resistance mutation assay is 5 x 10~ .

RESULTS
   The abilityrof the total methylene chloride extract of our automobile die-
sel exhaust particulate sample to induce resistance to F,tdR in human lympho-
blasts is shown in Figure 1.  A  concentration of 70 yg/ml of the diesel soot
extract was required to induce a significant  mutant fraction with a. survival
                            342

-------
                  o
                « 2
                  Ifl
                 
-------
                  200
                                  10      20      30
                                  /j.M Concentration

Fig. 2.  Mutagenicity of benzo(a)pyrene  (•)  and  fluoranthene (•)  to  human
lymphoblasts.  The error bars  represent  the  95%  confidence limits.  The dashed
line represents the historic upper  99% confidence  limit  of the spontaneous
background mutation frequency.
represent a major proportion of  the PAH  associated with  diesel exhaust parti-
  ,    11-13
culate.
   Based on the above chemical analysis  of the PAH containing fraction of our
diesel exhaust particulate extract, we initiated testing of the available indi-
vidual PAH to determine which  ones  play  a role in  the  mutagenicity of the whole
extract to human lymphoblasts.   Eleven PAH found in the  diesel exhaust extract
have been tested to determine  their tnutagenic potency.
   Fluoranthene, one of the most abundant PAH found in our diesel  exhaust
particulate extract, induces a significant mutant  fraction at a concentration
of 2 yM and has approximately  50% of  the mutagenic potency of benzo(a)pyrene
(BaP), as shown in Figure 2 and  Table 2.
                             344

-------
TABLE 1

POLYC7CLIC AROMATIC HYDROCARBONS AND DERIVATIVES IDENTIFIED IN A METHYLENE
CHLORIDE EXTRACT OF DIESEL EXHAUST PARTICULATE AS DETERMINED BY GAS CHROMA-
TOGRAPHY/MASS SPECTROMETRY ANALYSIS
Compound
mechylfluoranea
dibenzo thiophene
phenan thr ene
C_- fluorenes
3-me thy Iphenanthr ene
2-methylpb.enanthrene
9- and 4-methylphenanthrane
1-methylphenanthrene
C -fluorenes
phenylnaph thai ene
C.-phenanthrene
fluoranthene
C.- fluorenes
A
pyrene
methylphenylnaphthalenes
C,-phenanthrenes
methyl fluoranthenes & methyl pyrenes
C.-phenylnaphthalenes
C,-phenanthrenes
C.-phenylnaphthalanes
benzo(ghi) fluoranthena
benz ( a) anthracene
chry sene / tripheny lene
nitropyrene
benzofluoranthenes
benzopyrenea
perylene
TOTAL
WT %
0.03
0.02
0.21
0.17
0.15
0.21
0.13
0.13
0.11
0.065
0.52
0.21
0.21
0.21
0.21
0.30
0.17
0.13
0.17
0.03
0.03
0.007
0.01
0.03
0.03
0.03
0.002
3.5
   Figure 3 presents the results obtained when phenanthrene and three of its
monomethyl derivatives were tested  for their mutation inducing ability in
human lymphoblasts.  Phenanthrene was weakly active requiring a concentration
of 100 yM to induce a significant mutant fraction.  2-Methylphenanthrene was
                              J45

-------
                 If
                 I
                    O.I

                    50
                _
                u
                .2  30
                o
               c
                
-------
                                 50       100     150
                                  fj.M Concentration
200
Fig. 4.  Toxicity and mutagenicity of benz(a)anthracene (•), chrysene (B)  and
tripheneylene (A) to human lymphoblasts.  Error bars are as in Fig. I.
   Other PAH components of our diesel soot extract that have been tested for
their mutagenic potency to human lymphoblasts are pyrene and 1-methylpyrene.
These two PAH were inactive up to the tested concentrations of 300 yM and 100
uM respectively.  Mutagenicity results for the 11 tested PAH that are present
in our diesel exhaust particulate extract are summarized in Table 2.
   Many other PAH and derivatives have been found to be associated with  parti-
culates produced by the combustion of diesel fuel, some of which are shown  in
Table 3.  Anthracene and its alkylated derivatives have been identified  as  com-
ponents of various organic extracts of diesel exhaust emissions,      which led
us to test several of these PAH for their ability to induce mutations in the
human lymphob last assay.
                             347

-------
TABLE 2

MUTAGENIC CONTRIBUTION OF INDIVIDUAL PdLYCYCLIC AROMATIC HYDROCARBONS TO AN AUTOMOBILE DIESEL EXHAUST
PARTICULATE EXTRACT FOR HUMAN LYMPHOBLASTS
Compound
benz (a)anthracene
chrysene
fluoranthene
phenanthrene
l-methylphenanthrene
2-me thylphenanthrene
9 -me t hy Iphenan t h r ene
pyrene
1-me thy Ipy rene
triphenylene
benz o ( a) pyrene
Component contribution
Total extract
Induced
Mutation
+ 9
+ 6
+ 2
+100
+ 5
-200
+ 4
-300
-100
+ 20
+ 1


Relative15
Mutagenicity
0.1
0.2
0.5
0.01
0.2

0.25


0.04
1


Weight
Percent
0.007
0.01
0.2
0.2
0.1
0.2
0.07
0.2

0.01
0.03
1.03
100
Concentration'
(pM)
0.04
0.06
0,9
1.3
0.7
1.0
0.3
1.0

0.06
0.01
5.4 pM
100 pg/ml
Mutation Contribution
Mutant Fraction x 106
0
0
1.3
0
0.4
0
0.2
0
0
0
0.03
1.93
4.3
 (+) indicates the induction of a significant mutant fraction at the indicated pM concentration or  (-) indi-
 cates the response was negative up to the pM concentration tested.

 Indicates the mutagenic potency relative to that of benzo(a)pyrene which induces a significant mutant
 fraction at a concentration of 1 pM in the human lymphoblast mutation assay.
CThe pM concentration of the individual PAH in 100 pg/ml of the methylene chloride extract of our dlesel
 exhaust particulate extract.
 The predicted amount of the mutant fraction contributed by the individual PAH to the mutagenicity of 10O
 US/ml °^ our diesel exhaust particulate extract after correction for spontaneous background mutant fractions.

-------
TABLE 3
OTHER POLYCYCLIC AROMATIC HYDROCARBONS AND DERIVATIVES IDENTIFIED IN VARIOUS
EXTRACTS OF DIESEL EXHAUST PARTICULARS
        Compound                                                Source
  alkylnaphthalenes                                             a,c
  alkylanthracenes                                              a
  cyclopenteno(c,d)pyrene                                       b,c
  methyl benz (a) anthracene,                                     c
  chrysene or triphenylene
  methylbenzofluoranthenes                                      c
  methyl benzo(a)pyrene,                                        c
  benzo(e)pyrene or perylene
           , C. F. ,  Fischer,  J.  B. ,  and  Johnson,  D.  E. ,  Health Effects of Diesel
 Engine Emissions,  Pepelko,  W.  E. ,  Danner,  R.  M. , and  Clarke, N. A. ed. , U.S.
 Environmental Protection Agency,  Cincinnati,  OH, pp.  34-48  (1980).
 Stenberg, D. , Alsbert,  T.,  Blomberg, L. , and  Wannman,  T., Polynuclear
 Aromatic  Hydrocarbons,  Jones,  P. W. , and Leber, P.  ed. ,  Ann Arbor  Science
 Publishers,  Ann  Arbor,  MI,  pp.  313-326 (1979).
CSchuetzle, D. , Lee,  F.  S.-C. ,  Prater,  T. J. ,  and Tejada,  S. B. , Int. J.
 Environ.  Anal. Chem.  9,  93-144 (1981).
   Figure  5 shows that anthracene  and 9,10-dimethylanthracene were  not  muta-
genic  to human lymphoblasts  up  to  the tested  concentrations  of  200  yM and 100
yM,  respectively.   2-Methylanthracene was only weakly  active as  a mutagen to the
human  lymphoblasts  inducing  a significant mutant fraction at the concentration
of 60  yM.  However, 9-methylanthracene  was  mutagenic to human lymphoblasts  at a.
concentration of  9  yM, which is similar to  the mutagenic  potencies  of BA and CH.
A summary  of  the  results for anthracene and its methylated derivatives  is pre-
sented in  Table 4.
   We  are  not only  interested in determining  what  role PAH play  in  the  muta-
genicity of diesel  exhaust emission to  human  cells, but also in how certain PAH
are  metabolized to  their ultimate  mutagenic forms  for  human  cells.  We  have car-
                                                                        -| £_1 Q
ried out such a study with the  carcinogen  cyclopenteno(c,d)pyrene  (CPEP)     ,
which  is a. major  component of gasoline  exhaust emissions  and is  associated  with
the  particulate of  diesel exhaust  as well.13'19"21   The parent  PAH, CPEP, in-
duced  a significant mutant fraction at  a concentration of 7  yM,  (Figures 6,
Table  4) which is similar to the results in a previous report utilizing a dif-
ferent human  lymphoblast cell line.9 An arene oxide at the  3,4-position of
CPEP (CPAP-3,4-oxide)  was highly mutagenic  to the human lymphoblasts without
metabolic  activation (Figure 6), inducing  a 3ignificant mutant  fraction at  a
                              349

-------
                    oo-
                     20 -
                                9,
                                  50      100      150
                                   fj.M  Concentration
200
Fig. 5.  Toxicity and mutagenicity  of  anthracene (0),  2-methylanthracene  (•),
9-methylanthracerie  (•)  and  9,10-dimethylanthracene (A).   Error bars are  as in
Fig. 1.
concentration of 0.4 yM.  These  data show that CPAP-3,4-oxide is an ultimate
mutagen for human lymphoblasts.   However,  in the presence of rat liver PMS
this arene oxide was mutagenically  inactive  (Figure 6).   Cyclopentano(c.d)-
pyrene (CPAP), which lacks  the 3,4-double bond of CPEP and therefore cannot
form the 3,4 arene  oxide, was significantly  less mutagenic to human lympho-
blasts compared with the activity of CPEP (Figure 6).   CPAP was approximately
6-fold less mutagenically active in the  human lymphoblast mutation assay  rela-
tive to CPEP as indicated by  the data  shown  in Table 4.
   The mutagenicity of  two  dihydrodiol derivatives of CPEP, CPAP-3,4-cis_-diol
and CPAP-3,4-trans-diol was also determined.   CPAP-3.4-trans-diol, which  is the
                                              22       ^~"~™"~
major rat liver microsome metabolite of  CPEP,   was not mutagenically active in
the human lymphoblast mutation assay up  to a concentration of 80 yM (data not
                             350

-------
                 s
              (T .2
                    .01
                    70
•g  6°
 I  50

 i-
 2  30
'*
 £-20
                    10

                     0
                            CP*P-3.4-Oua«
                            Ne CMS
              CMP- 3.4-Oud*»PMS
                 20
                                         40       60
                                       Concentration
80
Fig  6   Toxicity and mutagenicity of cyclopenteno(c,d)pyrene (•),  CPAP  (O),
CPAP-3,4-oxide, no PMS (A), CPAP-3,4-oxide, +PMS (A) and CPAP-3,4-cis-diol
(•).   Error bars are as in Fig. I.
shown).  However, the results presented in Figure 6 demonstrate that CPAP-3,4-
cis-diol has the same mutagenic activity as the parent, CPEP.  This interesting
result suggests that CPAP-3,4-cis_-diol is a proximate mutagen of CPEP.
   In addition to our studies with the possible metabolites of CPEP, we  have
tested the ability oftrans-2,3-dihydrodio1fluoranthene to serve as a proximate
mutagen of fluoranthene to human lymphoblasts.  In the presence of rat  liver
PMS, trans-2,3-dihydrodiolfluoranthene was metabolized to a mutagen that was  as
active as the parent, inducing a significant mutant fraction at a concentration
of 2 uM (data not shown).
                              351

-------
TABLE 4
MUTAGENICITY OF PAH FOUND IN OTHER EXTRACTS  OF DIESEL EXHAUST TO
HUMAN LYMPHOBLASTS
Compound
anthracene
2-methylanthracene
9-methylanthracene
9 , 10-dimethylanthracene
cyclopenteno(c,d)pyrene
cyclopentano(c,d)pyrene
CPAP-3,4-oxided
CPAP- 3 , 4-t rans-diol
CPAP-3,4-cis-diol
fluoranthene-2, 3-dihydrodiol
Induced )JM
Mutation Cone.
200
+ 60
+ 9
100
+ 7
+ 40
+e 0.4
80
+ 6
+ 2
Relative0
Mutagenicity

0.017
0.1

0.14
0.025
2.5
0.17
0.5
 (+) indicates the induction of a significant mutant  fraction or (-)  indi-
 cates no significant, mutagenic effect.
 The ]M concentration that induces a significant  mutant  fraction or the
 highest concentration tested in the case  of a  negative  result.
 Mutagenic potency of the individual PAH relative to  benzo(a)pyrene which
 induces a significant mutant fraction in  the human lymphoblast  assay at a
 concentration of 1 pM.
 CPAP is cyclopentano(c,d)pyrene.
 Tested without metabolic activation.
DISCUSSION
   The results presented in Figure 1 show  that  a  methylene chloride extract of
automobile diesel exhaust particulate is mutagenic to human lymphoblasts in-
ducing a significant mutant fraction at a  concentration  of 70 pg/ml when a meta-
bolic activation system was present.  These results are  similar  to those re-
ported by Liber et al.  who also found that our diesel soot extract was not
mutagenic to human lymphoblasts without metabolic activation. However,
                2"}
McConnick et al.   have found that extracts of  diesel engine particulate and the
whole particles themselves are significantly cytotoxic to normal human fibro-
blasts and xeroderma pigmentosum (XP) fibroblasts which  suggest  that  "direct-
acting" cytotoxic agents for human fibroblasts  are associated with diesel
exhaust particulates.  In addition, they  (J. J. McCormick, personal communica-
tion) have found that both the organic extracts of diesel combustion  particu-
latas and the whole particulates induce resistance to 6-thioguanine in both cell
                              352

-------
types.  Further, the XP fibroblasts had an approximately 9-fold higher induced
mutation frequency relative to the normal human fibroblasts, suggesting that
the DNA adducts formed by these "direct-acting" mutagens are excisible.
   Since our extract of diesel soot was mutagenic to the human lymphoblasts
only in the presence of a metabolic activation system, we have concentrated on
determining what promutagens are responsible for the mutagenicity of the diesel
soot extract to these human cells.  Results of fractionation experiments demon-
strated that the hexane/toluene PAH containing fraction of our diesel soot
extract was the most mutagenic to S_. typhimurium in the presence of rat liver
    1 2
PMS, '  suggesting that this fraction could be responsible for a major propor-
tion of the human cell mutagenicity.  Eleven of the 27 identified PAH have been
tested for their ability to induce mutations in human lymphoblasts.  The results
are summarized in Table 2.  Based on this mutational data and the chemical anal-
ysis presented in Table 1, it was possible to calculate the induced mutant frac-
tion contribution of each tested PAH to the total mutagenicity of 100 yg/ml of
the methylene chloride extract of our diesel soot sample (Table 2).  The data
suggest that as few as three PAH; fluoranthene, 1-methylphenanthrene, and 9-
methylphenanthene, may account for up to 44% of the total mutability of the
diesel emissions particulate extract to human lymphoblasts.  Fluoranthene
alone may be responsible for approximately 30% of the total mutagenicity of
our diesel soot extract to human cells.  Benzo(a)pyrene, which has routinely
been used as a standard indicator of PAH levels, is present in too small a
concentration to play any significant role in the mutagenicity of our diesel
exhaust particulate extract to human lymphoblasts (Table 2).
   It is our belief, that once the mutagenic potency to human lymphoblasts of
the other PAH present in our diesel soot extract has been determined, the PAH
alone will account for the entire mutagenicity of our extract to human lympho-
blasts.
   We have also carried out studies on the mutagenicity to human lymphoblasts of
PAH found in other diesel soot extracts.  Methylated anthracenes have been found
to be associated with the particulate emitted from several diesel engines.
Comparison of the data presented in Tables 2 and 4 demonstrates that 9-methyl-
anthracene has a mutagenic potency to human lymphoblasts similar to that of the
known carcinogens benz(a)anthracene, chrysene and cyclopenteno(c,d)pyrene.
Therefore, in light of the role played by alkylated phenanthrenes in deter-
mining the mutagenicity of diesel emissions extracts, one should not overlook
the role played by simple alkylated PAH in determining the genotoxic effect of
complex combustion mixtures.
                              353

-------
                                          1 f\ 1 R
   The carcinogen cyclopenteno(c,d)pyrene      (CPEP) is a major PAH component
of gasoline exhaust particulate  and is known to be associated with diesel
                             13 19—21
exhaust particulate as well.   '        CPEP appears to be converted to its ulti-
                                                                           79 O/
mate mutagenic form via  a  one step  epoxidation across the 3,4 double bond.  "'
Our results demonstrate  (Figure  6)  that the 3,4-oxide of CPEP is a potent
direct-acting mutagen to human lymphoblasts and are similar to those obtained
                                               25
with the L5178Y mouse lymphoma mutation assay.     Therefore, CPAP-3,4-oxide is
an ultimate mutagen of CPEP  to mammalian cells.  However, our results suggest
that there is more than  one  pathway for the activation of CPEP.  CPAP, which
lacks the 3,4 ethylene bond  of CPEP and therefore cannot form the 3,4-oxide,
still induced a significant  mutant  fraction in human lymphoblasts (Figure 6),
albeit at a higher concentration relative to CPEP (Table 4).  In addition,
                                                                            22
though CPAP-3,4-trans-diol,  the  major- rat liver microsome metabolite of CPEP  ,
was mutagenically inactive,  CPAP-3,4-cis-diol had the 3ame activity as CPEP.
                             22
Recently, Gold and Eisenstadt    have reported that a 9,10 "K-region" dihydrodiol
of CPEP is a minor rat liver microsome metabolite of the parent PAH.  Therefore,
it is possible that CPAP-3,4-cis-diol is oxidatively metabolized to CPAP-3,4-
^i£-diol-9,10-oxide which  serves as a second ultimate mutagen of CPEP.
   We conclude that there  are at least three pathways for the metabolic activa-
tion of CPEP; 1) a predominate pathway which proceeds via the epoxidation at the
3,4 double bond, 2) a pathway independent of the 3,4-ethylene double bond and,
3) a pathway specific to CPAP-3,4-cis-diol.   We caution the reader that this
third pathway of activation  may  not pose a genotoxic threat to man in that
the formation of cis-dihydrodiols of PAH are not known to be produced by mam-
malian cell metabolism.
   In our initial studies  to determine the metabolic pathway for the activation
of fluoranthene to an ultimate mutagen for human cells, we found that trans-2,3-
dihydrodiolfluoranthene  had  the  same mutagenic potency as the parent PAH with
metabolic activation.  This  result  suggests that trans-2,3-dihydrodiolfluor-
anthene is a potential proximate mutagen of fluoranthene for human lymphoblasts.
Furthermore, our unpublished data demonstrate that trans-2,3-dihydrodio1-1,10B-
epoxyfluoranthene is a potent direct-acting mutagen for Salmonella typhimuriua
indicating that this diol-epoxide of fluoranthene is an ultimate mutagen for
bacterial cells.
   In Table 5 we present a summary  of the PAH and derivatives that have been
tested for mutation inducation in other human cell mutation assays.   Several
epoxide derivatives have been  tested and with the exception of benz(a)anthra-
cene-5,6-oxide, all were mutagenic  to human fibroblasts.   This finding suggests
that other oxide derivatives of  PAH are mutagens for human cells.
                              354

-------
 TABLE  5
 POLYCYCLIC AROMATIC HYDROCARBONS AND DERIVATIVES TESTED FOR MUTATION INDUCTION
 IN  OTHER HUMAN CELL SYSTEMS
Compound
benz (a) anthracene
benz (a) anthracene-5 , 6-oxide
benzo(a)pyrene

benzo ( a) py rene-4 , 5-oxide
benzo(a)pyrene-7,8-diol-
9 , 10-oxide
dibenz ( a , c ) anthracene
dibenz (a, h) anthracene
dibenz (a ,h) anthracene-5 , 6-
oxide
7 , 12-dime thylbenz ( a) anthra-
cene-5 , 6-oxide
anthracene
chrysene
Cell
Type
Epithelial5
Fibroblast0
Epithelial/
Fibroblastd
Fibroblast
Fibroblast6

Epithelial
Epithelial
Fibroblast

Fibroblast

Epithelial
Epithelial
Mutation Genetic
Induction Marker3
DTR
8AGR
+/+ DTR/6TGR

+ 8AGR
+ 8AGR

DTR
DTR
+ 8AGR

+ 8AGR

DTR
DTR
 _  R                                    R                               R
 OT  - diphtheria toxin resistance;  SAG  - 8-azaguanine  resistance;  6TG  -
 6-thioguanine  resistance.
 Rocchi, P.,  Ferreri,  A.M.,  Borgia,  R. ,  and Prodi,  G. , Car cino genes is  1,
 765-767 (1980).
 CMaher, V. M. ,  McCormick,  J.J.,  Grover,  P.  L.,  and  Sims,  P.,  Mutation  Res.  43,
 117-138 (1977).
 dAust, A.  E. , Falahee,  K.  J.,  Maher  ,  V.  M. , and McCormick, J.  J., Cancer Res.
 40, 4070-4075  (1980).
 SMaher, V. M. and McCormick,  J.J.  (1978)  in Polycyclic Hydrocarbons  and Cancer,
 Gelboin,  H.  V.,  and T'so, P.  0. P., ed,,  Academic  Press, New York,  Vol.  2,
 pp. 137-160.
                 ?fi
   Rocchi  et  al.    found that  benz(a)anthracene and chrysene  were inactive  as
 mutagens in  their human epithelial cell mutation assay (Table 5) which is
 contrary to  our results with human lymphoblasts.  However,  Rocchi et al."  used
 a concentration of only 1  yM,  which  is  most  likely  too low  a  concentration  to
 induce a significant mutagenic response.

ACKNOWLEDGEMENTS
   The authors  wish to  acknowledge the  excellent  technical  assistance  of
Hans-Peter Bieman,  Beatrice  Brunengraber,  the  late  Robert Cuzick Jr. and  Iria
                              355

-------
Romano.  Research was  supported by DOE contracts No. DE-AC02-77EV-04267 and
DE-ACQ2-80EV-10449, NIEHS  program grant No. 5-P01-ES00597, NIEHS  center grant

No. 1-P30-ES02109 and  NIEHS grant ES02145.


REFERENCES
1.   Liber, H. L., Andon,  B.M., Hites, R. A. and Thilly, W. G.  (1980) In Health
     Effects of Diesel Engine Emissions, Pepelko, W. E. , Danner,  R.M. and
     Clarke, N. A. ed.,  U.  S. Environmental Protection  Agency,  Cincinnati, OH,
     pp. 404-412.
2.   Yu, M.-L. and Hites,  R.  A. (1981) Anal. Chem. 53,  951-954.
3.   Mitchell, A. D.,  Evans,  E. L., Jotz, M. M., Riccio, E. S., Mortelmans,
     K. E. and Simmon,  V.  F.  (1980) in Health Effects of Diesel Engine Emis-
     sions, Pepelko, W.  E., Danner, R. M. and Clarke, N. A. ed.,  U.S.
     Environmental Protection Agency,  Cincinnati, OH, pp. 810-842.
4.   Casto, B. C., Hatch,  G.  G. , Huang, S. L., Huising, J. L., Neshow, S.  and
     Waters, M. D. (1980)  in Health Effects of Diesel Engine  Emissions, Pepelko,
     W. E., Danner, R.  M.  and Clarke,  N. A. ed., U.S. Environmental Protec-
     tion Agency, Cincinnati, OH, pp.  843-860.
5.   Ames, B. N., McCann,  J.  and Yamasaki, E. (1975) Mutation Res. 31, 347-364.
6.   Barfknecht, T. R.,  Andon,  B.M., Bishop, W. W. and  Thilly, W. G.  (1981)
     Environ. Mutagenesis  in press.
7.   Skopek, T. R., Liber,  H. L., Penman, B. W. and Thilly, W. G. (1978) Bio-
     chem. Biophys. Res. Commun. 84, 411-416.
8.   Thilly, W. G., DeLuca, J.  G., Furth, E. E., Hoppe, H., IV, Kaden, D.  A.,
     Krolewski, J. J.,  Liber, H. L., Skopek, T. R. , Slapikoff, S. A., lizard,
     R. J. and Penman,  B.  W.  (1980) in Chemical Mutagens Vol. 6,  deSerres, F. J.
     and Hollaender, A.  ed.,  Plenum Publishing Corp., New York, NY pp. 331-364.
9.   Skopek, T. R., Liber,  H. L. Kaden, D. A., Hites, R. A. and Thilly, W. G.
     (1979) J. Natl. Cancer Inst. 63,  309-312.
10.  Furth, E. E., Thilly,  W. G., Penman, B. W., Liber, H. L., and Rand, W. M.
     (1981) Anal. Biochem.  110, 1-8.
11.  Choudhury, D. R.  and  Bush, B. (1980) in Health Effects of Diesel Engine
     Emissions, Pepelko, W. E., Danner, R. M. and Clarke, N.  A. ed., U. S.
     Environmental Protection Agency,  Cincinnati, OH, pp. 175-186.
12.  Rodriguez, C. F.,  Fischer, J. B.  and Johnson, D. E. (1980) in Health
     Effects of Diesel Engine Emissions, Pepelko, W. E., Danner,  R. M. and
     Clarke, N. A. ed.,  U.S.  Environmental Protection Agency, Cincinnati,  OK,
     pp. 34-48.
13.  Schuetzle, D., Lee, F. S.-C., Prater, T. J. and Tejada,  S. B. (1981)
     Int. J. Environ.  Anal. Chem., 9,  93-144.
14.  Christensen, H. E., Fairchild, E. J. and Lewis, R. J. (1976) Suspected
     Carcinogens. National Institute of Occupational Safety and Health,
     Cincinnati, OH, pp. 1-251.
15.  Dipple, A. (1976)  in  Chemical Carcinogens, E. E. Searle  ed., American
     Chemical Society,  Washington, D.  C., pp 245-314.
16.  Neal, J. and Trieff,  N.  M. (1972) Health Lab. Sci. 9, 32-38.
17.  Wood, A. W., Levin, W.,  Chang, R. L., Huang, M.-T., Ryan, D. E., Thomas,
     P. E., Lehr, R. E., Kumar, S., Koreeda, M., Akagi, H., Ittah, Y.,
     Dansette, P., Yagi, H.,  Jerina, D. M. and Conney,  A. H.  (1980)  Cancer
     Res. 40, 642-649.
18.  Cavalieri, E. L.,  Rogan, R., Toth, B. and Mungall, A. (1981) Carcino-
     genesis, 2, 277-281.
19.  Grimmer, G.  (1977)  IARC Sci. Pub. 16, 29-39.
                              356

-------
20.  Grimmer, G., Naujack, K.-W. and  Schneider,  D.  (1980)  in Polynuclear
     Aromatic Hydrocarbons, Bj^rset,  A.  and  Dennis,  A.  J.  ed.,  Battelle
     Press, Columbus, OH, pp.  107-125.
21.  Stenberg, U., Alsberg, T.,  Blomberg,  U  and Wannman,  T.  (1979)  in
     Polynuclear Aromatic Hydrocarbons,  Jones, P. W.  and Leber, P. ed.,
     Ann Arbor Science Publishers,  Inc., Ann Arbor,  MI, pp. 313-326.
22.  Gold, A. and Eisenstadt,  E.  (1980)  Cancer Res.  40, 3940-3944.
23.  McCormick,  J. J. , Zator,  R. M.,  DaGue,  B. B. and Maher, V. M. (1980)  in
     Health Effects  of Diesel  Engine  Emissions,  Pepelko, W. E.  Danner, R.  M.
     and Clarke, N.  A. ed., U.  S.  Environmental  Protection Agency, Cincinnati,
     OH, pp. 413-415.
24.  Eisenstadt, E.  and  Gold,  A.  (1978)  Proc. Nat.  Acad. Sci.  75,  1667-1669.
25.  Gold, A., Nesnow, S., Moore,  M. , Garland, H. ,  Curtis, G.,  Howard, B.,
     Graham, D.  and  Eisenstadt,  E.  (1980)  Cancer Res. 40,  4482-4484.
26.  Rocchi, P.,  Ferreri, A.M.,  Borgia,  R.  and Prodi, G.  (1980) Carcinogenesis
     1, 765-767.
                               357

-------
       CYTOTOXICITY, MUTAGENICITY AND COMUTAGENICITY  IN DIESEL  EXHAUST
         PARTICLE EXTRACTS ON CHINESE HAMSTER OVARY CELLS  IN VITRO

                                     by

         A. P. Li, R. E. Royer, A. L. Brooks, and R.  0. McClellan
            Lovelace Inhalation Toxicology Research Institute
                 P. 0. Box 5890, Albuquerque, NM  87185


Diesel exhaust particle extracts were found cytotoxic to Chinese hamster ovary
(CHO) cells.  Extracts from cars of different manufacturers had different
cytotoxicity.  The emission rates of cytotoxic chemicals were calculated for
the different cars using the cytotoxicity of the extracts, the  percentage of
extractable chemicals on the exhaust particles, and particulate emission
rates.  The ranking of emission rates of cytoxic chemicals for  the different
cars were found to be the reverse of the ranking of the cytotoxicity of the
extracts (1).  Our data indicate the need to include  emission data other than
the activities of the extracts, when the emission of  noxious agents from dif-
ferent vehicles are compared.

The cytotoxicity of diesel exhaust particle extracts  is antagonized by serum,
lung and liver cytosols, and sulfhydryl agents in vitro (2) [Figures 1, 2],
The detoxifying effects of the cytosols is enhanced further by  the addition
of cofactors (NADP and glucose-t-phosphate); therefore, suggesting enzymatic
detoxification in addition to protein binding.  Our data suggest that similar
detoxification of the toxic chemicals associated diesel exhaust particles may
occur j_n_ vivo.

All diesel exhaust particle extracts had low mutagenicity towards CHO cells.
This low activity was observed using different endpoints including sister-
chromatid-exchange and mutation at the hypoxanthine-guanine phosphoribosyl
(HGPRT) gene locus (3).  The mutagenicity was slightly enhanced by the ad-
dition of exogenous aroclor 1254-induced liver S9.  Although the extracts
had only low mutagenicity, they were found to have definite co-mutagenic
activities (4).  Treatment of CHO cells with a combination of a mutagen
(N-methyl, N'-nitro, N-nitrosoguanidine or benzo(a)pyrene) and diesel exhaust
extract yielded a 2-3 fold higher mutant frequency than that calculated by the
mutagenicity of the mutagen and the diesel exhaust extract alone [Table 1].
This co-mutagenicity was observed for all extracts tested, using three dif-
ferent endpoints:  mutation at the HGPRT gene locus, mutation at the Na+-Ka+-
ATPase gene locus, and sister-chromatid-exchange.  We have shown that diesel
exhaust particles a.re associated with chemicals with cytotoxic, mutagenic,
and co-mutagenic properties.  Engineering variables, biological detoxifying
molecules, and other environment mutagens/carcinogens, all could possibly
modify the health-effect of the diesel exhaust emission.
                                        358

-------
REFERENCES

1.  Li,  A.  P.,  R.  E.  Royer,  A.  L.  Brooks,  R.  0.  McClellan,  W.  F. Marshall, and
      T. M.  Naman.  Cytotoxicity of diesel  exhaust particle extract—a com-
      parison among five diesel passenger  cars of different manufacturers.
      Manuscript in preparation.

2.  Li,  A.  P.  1981.   Antagonistic effects of animal  sera,  lung and  liver
      cytosols, and sulfhydryl  compounds on the cytotoxicity of diesel ex-
      haust particle extracts.   Toxicol. Appl. Pharmacol.  57:55-62.

3.  Li,  A.  P., and A. L. Brooks.  1981.  Use of Chinese hamster ovary cells
      in the evaluation of potential hazards from energy effluents—applica-
      tion to diesel  exhaust emission.  Lewis, M. (ed.).   "Proceedings,  the
      International Symposium of Health Impact of Different Sources  of Ener-
      gy", jointly sponsored by WHO/UNEP/IAEA, Nashville,  TN, June 22-26,
      1981.  In press.

4.  Li,  A. P., and R. E. Royer.  1981.  Diesel exhaust particle extract  en-
      hancement of chemical-induced mutagenesis in cultured Chinese  hamster
      ovary cells:  Possible interaction of diesel exhaust with environmental
      carcinogens.  Mutat. Res.  In press.
                                      359

-------
      Table 1.  Co-mutagenicity of Diesel Exhaust Particle  Extracts
                  in the Presence of Exogenous Activation System3
                  (Li and Royer, 1981)
Treatment
                       Mutant Frequency
-B(a)P    +B(a)P (o.5 yg/ml)
               Expected*3
Observed
Enhancement0
Solvent (DMSO)
   Control           7 (A)
                                   56 (B)
Exhaust Extracts
(60 ug/ml)
Car A
Car B
Car C
Car D
Car E

28
6
11
9
25

77
56
60
58
74

218
170
194
184
229

2.8
3.0
3.2
3.2
3.1
 An Aroclor 1254 - induced rat liver cytosol/cofactors mixture was used
 for exogenous activation.


 Expected mutant frequency = mutant frequency (B(a)P alone) (B) + mutant
 frequency (diesel exhaust particle extract alone) - mutant frequency
 (DMSO alone) (A).


"Enhancement = observed mutant frequency -s- expected mutation frequency.
                                    360

-------
1.0-


IVE SURVIVAL
p
VI
5
at
0
t^^^^m





1
51



1 	 1




01











51



£
































OMSO OIESa HUMAN CALF CtSTEINE SERINE GLUTA- OXIDIZED MERCAPTO- ETHYLENE
SERUH SERUM THIONE 6LUTATHIONE ETHANOL aYCOL
1 1
                                         DIESEL
Figure 1 (Li,  1981).
Effects of animal sera, sulfhydryl compounds and their
non-sulfhydryl analogs, on the cytotoxicity of diesel
exhaust extract.
                                     361

-------
                                                 Liver S9
                                                 + Cofactors
                 Lung S9
                 + Cofactors
                    Lung S9
                    - Cofactors
0           100         200
 CONCENTRATION (^g/ml)
                                    0            100
                                     CONCENTRATION
Figure 2  (Li,  1981).
              Effects of lung (A) and liver (B) cytosols on the
              cytotoxicity of diesel exhaust extract.
                               Jbi!

-------
   MUTAGENIC ACTIVITY OF DIESEL PARTICLES IN ALVEOLAR MACROPHAGES
               FROM RATS EXPOSED TO DIESEL ENGINE EXHAUST


                              J-S. Siak and K. A. Strom
                          Biomedical Science Department
                       General Motors Research Laboratories
                                 Warren, MI 48090


Diesel engine  exhaust contains  submicron size carbonaceous  particles.   Dichloro-
methane  extracts   of  these  particles   collected  by   filtration  or  electrostatic
precipitation elicited mutagenic responses in bacterial  mutagenicity assay. Currently,
there are no data  on the mutagenic properties of  inhaled diesel particles  that are
deposited directly in the  lung.  The  purpose of this  experiment was to determine the
mutagenicity of the inhaled diesel particles and the interaction between the  particles
and alveolar macrophages.
                                                  n
Adult male Fischer 344 rats were  exposed to 6 mg/m  of diesel particles for 3 days (20
hrs/day).  Alveolar  macrophages  were obtained by  bronchopulmonary  lavage  imme-
diately  after exposure and at 1,  4, and 7 days  thereafter.  Macrophages from forty
animals were  pooled for each data  point, sized and  counted.   The  mass of  diesel
particles phagocytized in alveolar macrophages was determined  by  a spectrophoto-
metric  method  (Rudd  and Strom, J. Appl.  Tox.,  l(2):83-87,  1981).   The  alveolar
macrophages were concentrated by filtration on pre-washed fiberglass filters and dried
at room  temperature to constant  weight.   The filters were  extracted  with  dichloro-
methane in  a Soxhlet  apparatus  for 4 hours (20-25 solvent cycles).  The  resulting
extracts were oily, indicating  cellular lipids and surfactant were  extracted from the
macrophages.   The  Salmonella typhimurium strain TA98  was used  for  mutagenicity
assay.   For  thin  layer chromatography,  Whatman  LK6  plates  were  used  and  the
developing solvent was toluenethexane (5:1).

Table 1 shows the diesel particle mass recovered in  alveolar macrophages from exposed
rats.  The mass of diesel  particle  recovered from the lavage accounted for 45-50% of
the particles deposited in rat lungs.   The  extracts of diesel particles (DPE) collected
from  the exposure  chamber  by filtration  were used as reference for the thin layer
chromatographic  and  mutagenic  analysis  of  the  macrophage  extracts.   The TLC
fluorescence banding pattern of the samples from macrophages obtained immediately or
one day post-exposure were similar to that of chamber DPE.  However, the extracts of
macrophages recovered on the  fourth  and seventh day post-exposure lost their fluores-
cence patterns.  Figure 1 shows the results of the mutagenicity assay. The data indicate
that the cellular  lipids extracted  from macrophages mitigated the mutagenic response
of the airborne DPE, but a positive result  was still  detectable in the extracts of the
macrophages obtained immediately, and one day after  the  exposure.  In contrast, the
mutagenic activity  of extracts from  macrophages obtained on the fourth and seventh
                                       363

-------
day after the exposure  was undetectable.  The data indicate that:  1) inhaled diesel
particles contain extractable mutagenic compounds  - whether  they  are the same as
those found in the particles collected by other means has yet to be resolved;  2) alveolar
macrophages have  the ability to release or transform the fluorescent  and mutagenic
extractable hydrocarbons from  phagocytized diesel particles over a period of several
days and thus may significantly influence their biological activity in the respiratory
system.
                                      Table  1

                          DIESEL  PARTICLE MASS RECOVERY
                    IN ALVEOLAR MACROPHAGES FROM EXPOSED RATS
GROUPS
IMMEDIATELY
1 DAY-POST
4 DAY-POST
7 DAY-POST
DP ug/mL
Lavage Fluid
5.0
5.9
6.3
6.6
DP ug/106
Macrophages
41.3
34.3
25.5
25.8
Total Recovery
(mg)
8.4
10.3
10.6
11.1
Mutagenic activity  of  airborne  diesel
particle   extract   and   macrophage
extracts.

           Airborne  diesel   particle
           extract.

           Airborne  diesel   particle
           extract +  800  yg control
           macrophage extract.

           Macrophage  extract  from
           exposed  rats immediately
           after exposure.

           Macrophage  extract  from
           exposed  rats  7  days  after
           exposure.
                 04    10

fOUIVAUNT  OUSEL PMTICUWTt MASS (mjl

          HI PUT!
                                        364

-------
SECTION 6



CARCINOGENESIS
                        365

-------
SKIN CARCINOGENESIS  STUDIES  OF EMISSION EXTRACTS
S. Nesnow, C. Evans, A.  Stead and J.  Creason
Carcinogenesis and Metabolism,
Data Management and Biostatlstics Branches
U.S. Environmental Protection Agency
Research Triangle Park,  NC
and
T.J. Slaga and L.L. Triplett
Biology Division
Oak Ridge National Laboratory
Oak Ridge, TN

INTRODUCTION
   The incomplete combustion  of  fossil fuels results in the emission of parti-
culate and organic vapor-phase components to the atmosphere.  The particulate
phase of these emissions contains organic materials adsorbed onto the particulate
matrix.  These organic materials have been subjected to intense chemical analysis,
fractionation and characterization.       The characterization of the biological
activities of organic materials  emitted in the environment has been reported for
samples collected as whole condensates and as particles.  Emissions from gasoline
engines, collected as condensates, were tumorigenic when applied dermally to
     16—22
mice.       Extracts from particles collected from a gasoline engine were also
tumorigenic on mouse skin.
               24
   Kotin et al.   reported that  extracts of particles collected from diesel
engines were active in producing tumors on strain A mice, while Mittler and
         22
Nicholson   reported little such activity from diesel exhaust condensates.
   We have previously reported that extracts from particulate emissions from
coke oven, roofing tar, and several diesel and gasoline engines producedpapillo-
mas when applied to SENCAR mice.   '     These tumor initiation experiments
indicated that the tumorigenic activities of the diesel engines were dependent
upon the particular engineering  qualities of each engine  and that these emission
samples produced a wide range of activities.
   The SENCAR mouse has been  shown to be highly sensitive to chemical carcinogens
and useful in mechanistic studies of  carcinogenesis.  ~    This paper describes
the results of a systematic study of  the ability of extracts of particulate
emissions to induce benign and malignant tumors in SENCAR mice and their abilities
to act as tumor initiators, tumor promoters, or complete  carcinogens.
                              366

-------
 MATERIALS AND METHODS
 Sample Generation and Isolation
    The details of sample generation and isolation have been reported elsewhere.2
 Particulate emissions were collected from a 1973 preproduction Nissan Datsun,  a
 1978  Oldsmobile,  a prototype 1976 VW Turbo Rabbit, a 1977 Mercedes 300D and a
 1977  Mustang vehicle (Table 1), each which was mounted on a. chassis dynamometer
 with  a repeated highway fuel economy (HWFET)  cycle of 10.24 mi,  an average speed
 of  48 mph, and a running time of 12.75 min.  Another sample was  collected from a
 1972  heavy-duty Caterpillar 3304 engine mounted on an engine dynamometer at 2200
 rpm steady state with an 85-lb load.  The residential furnace sample was
 collected from a Day and Night Air Conditioning Model 125-OU-AC-A furnace,
 100,000 BTU,  0.9 gal/min firing rate,  operating at 20% lean and  using a cyclic
 mode  of 10 min on,  20 min off.  The Caterpillar,  Nissan,  Mercedes,  VW and
 Oldsmobile engines were fueled with the same  batch of No.  2 diesel fuel.   Parti-
 cle samples were collected with a dilution tunnel in which the hot exhaust was
 diluted,  cooled and filtered through Pallflex Teflon-coated fiberglass filters.
 Topside coke oven samples were collected from the top of  a coke  oven battery at
 Republic  Steel,  Gadston,  AL,  by use of a Massive  Air Volume Sampler.   Because  of
 the topside ambient location and local'wind conditions, an unknown portion of this
 emission sample contains particles from the local urban environment.   The coke
 oven  main sample  was collected from a  separator located between  the gas collector

 TABLE 1
 DIESEL AND GASOLINE SAMPLES
Source
1972 Caterpillar 3304
1973 Preproduction Nissan
Datsun 220C
1978 Oldsmobile 350
-1976 Prototype VW
Turbo Rabbit
1977 Mercedes 300D
Residential Furnace

1977 Mustang 11-302,
V-8 catalyst and EGR
Fuel
Diesel No.
Diesel No.

Diesel No.
Diesel No.

Diesel No.
Diesel No.

Unleaded
gasoline

2a
2

2
2

2
2



Driving
Cycle
Mode IIb
HWFETC

HWFET
HWFET

HWFET
10 min
20 min
HWFET







on/
off


aAll diesel fuel samples were from the same lot.
^Mode II cycle was conducted at 2200 rpm steady state with an 85-lb
 load.
cHighway fuel economy cycle (HWFET) was a 10.24-mi cycle averaging
 48 mph and taking 12.75 min.
                             367

-------
and the primary  coolers within the coke oven battery.   The  roofing tar emission
sample was collected from a conventional tar pot with  external propane burner.
Pitch-based tar  was  heated to 182° to 193°C and emissions were collected with a
1.8-m stack extension and Teflon socks in a baghouse.   It should be noted that
only one vehicle or  source was used for each sample, therefore each sample may
not be representative of a particular technology.
   All samples were  Soxhlet extracted with dichloromethane, which was then
removed by evaporation under dry nitrogen gas.

Animals
   SENCAR mice,  selected for their, increased sensitivity to carcinogens,    were
used in this  study.   These mice were derived by breeding Charles  River .CD-I
mice with male skin-tumor sensitive (STS)  mice that were originally derived from
Rockland mice.   Mice were selected for sensitivity to  the 7,12-dimethylienz(a)
anthracene-12-0—tetradecanoylphorbol-13-acetate (TPA)  two-stage  system of
tumorigenesis for  eight generations.  These mice were  obtained initially from
Dr. R. Boutwell  (HcArdle Laboratory for Cancer Research, University of Wisconsin,
Madison, WI)  and are now raised at the Oak Ridge National Laboratory, Oak Ridge,
TN.

Chemicals
   TPA was obtained  from Dr.  P. Borchert (University of Minnesota,  Minneapolis,
MN) and benzo(a)pyrene B(a)P from Aldrich Chemical Co.  (Milwaukee,  WI) .   All
agents were prepared under yellow light immediately before use  and  were applied
topically in  0.2 ml  of spectral-quality acetone.

Tumor Experiments
   Studies involved  80 7- to 9-week-old mice per treatment group  (40 of each
sex).  Animals were  housed in plastic cages (10 per cage) under yellow light
with hardwood chip bedding,  fed Purina chow and water  ad libitum, and maintained
at 22° to 23°C with  10 changes of air per hour.  All mice were  shaved with sur-
gical clippers two days before the initial treatment and only  those mice
in the resting phase of the hair cycle were used.  Under the tumor  initiation
protocol, all samples at all doses were applied as a single topical treatment,
except for the 10  mg dose,  which was administered in five daily doses of 2 mg.
One week after treatment,  2.0 yg of the tumor promoter  TPA was  administered
topically twice  weekly.   Under the complete carcinogenesis protocol, samples
were administered  weekly (or twice weekly for the highest dose  level) for 50
                            368

-------
to 52 weeks.   Under the  tumor promotion protocol,  mice were first initiated
with 50.5 ug of  B(a)P  and than treated weekly (or  twice weekly at the  highest
dose level) for  34  weeks with the sample,   skin tumor formation was  recorded
weekly, and papillomas greater than 2 mm in diameter and carcinomas  were  in-
cluded in the  cumulative total if they persisted for one week or longer.  The
number of mice with tumors,  the number of  mice  surviving,  and total  number  of
tumors were determined and recorded weekly.   At six  months the numbers of
papillomas per surviving animal were recorded for  statistical purposes.  Histo-
logical verification of  tumors as well as  histopathological identification  of
nondermal tumors will  be reported elsewhere.

Data Collection, Validation  and Storage
   A unique identifier was assigned to each treatment group. Treatment information
for each group and  raw data  were coded onto forms  for data processing.  Data
entered on the form were punched onto cards which  underwent 100% verification
against the forms.   Weekly group scores (numbers of  surviving mice,  carcinomas,
papillomas and mice with tumors)  and periodic individual animal scorings for each
group were validated by  a computer program,  corrected,  and updated on a cumulative
file maintained  on  a Univac  1110.
   The main data base  was sampled periodically  by  use of Dodge-Romig sample
inspection tables and  verified against the data entry forms.     An error rate of
less than 2% was maintained.   Following each update  of new data,  reports listing
treatment protocol, weekly tumor scores and individual papilloma and carcinoma
scores by animal group identifier were generated.  The program which generates
these reports  also  builds subset card image data files from the main data base,
which are input  to  various statistical analysis routines.

Statistical Analysis Methods
   Analyses were carried out  on tumor scorings  performed 24 to 26 weeks after
initiation.  Two types of statistical analyses  were  performed.   For  tumor
incidence data,  a probit model was fitted,  taking  into account the numbers  of
spontaneous tumors  occuring  in the TPA control  groups.   The probit formula  used
is
                           P  = 6   + (1-6 )  *  (Bj^ + 62 In x)

where P is the probit  proportion,  x is the dose applied and *  is  the standard
normal cumulative distribution function.     The model parameters  gQ, 8^ and  &2
were estimated from the  raw data  by maximum likelihood methods.   The dose which
produces tumors  in  50% of the surviving mice  over  the TPA-treated controls  was
                             369

-------
then estimated as a function of  the  estimated parameters.  The 95% confidence
limits were estimated using the  asymptotic variance-covariance matrix estimated
during the model-fitting process.  Chi-square goodness~of-fit and likelihood
ratio tests were also computed to  examine the appropriateness of the model and the
strength of the dose effect.
   The tumor multiplicity data were  analyzed by a nonlinear Poisson model:
                                           B  + (3  In (x.)
                               X.  =  6Q + e
where X. is the number  of papillomas per mouse, x.  the dose and 3 , g. and  fi
the model parameters.   Using maximum likelihood methods, the model parameters
were estimated from the raw data and used to calculate the number of papillomas
per mouse for a dose of 1 mg.  Asymptotic 95% confidence intervals for these
activities were obtained.  Tests for the Poisson assumption, adequacy of the
model, and strength of  the dose  response were also calculated.
RESULTS
   The mouse skin bioassay system  can be used to evaluate agents as tumor
initiators, tumor promoters, cocarcinogens and complete carcinogens.   The two
protocols that can be employed to  detect chemical carcinogens in the mouse
skin tumorigenesis assay are complete carcinogenesis and tumor initiation,  as
illustrated in Figure 1.  Multiple application of tha test agent for  up to  60
                                          PROTOCOLS
                                     TEST               TPA. 2 x WEEKLY
                                                    l                  I
                                                    |TSST AGENT, WEEKLYj
                                     B-A-P
           TUMOR INITIATION
           TUMOR PROMOTION            4    •   •    f    ?__?
                                     TEST  ' """"" T™"""•"  —1-°=      (
                                     AGENT          I TPA, 2 x WEEKLY   J
                                     B-A-P  p*V-j——p—Y™—= j    j  ' 1
           COCARC1NOGENESIS            t    •   f    «	fc    |    y    y
                                                    I                 1
                                                    JTEST AGENT, WSEKLY!
                                               T    r    «    r
           COMPLETE CARC1NOGENESIS     | _  f^	1. .„ ,j__L^__l.
                                                    I
                                            WEEK OF EXPERlMEfJT
                                                    i
                                                    I
                                                  SCORE             SCQRE
                                                   FOR               FOR
                                               PAPILLOMAS        CARCINOMAS
 Fig. 1.  Schematic diagram of tumor initiation,  tumor promotion, cocarcinogenesis
          and complete  carcinogenesis bioassay protocols.
                              370

-------
weeks will give rise primarily to malignant carcinomas of the skin.  This
protocol  for complete carcinogens is a test for agents exhibiting both tumor-
initiating and tumor-promoting activities.  The bioassay protocol for tumor
initiators is a single application of test agent followed one week later by
multiple  applications of the potent tumor promoter TPA.  The bioassay protocol
for  tumor promoters is initiation with a strong tumor initiator, B(a)P, followed
by weekly applications of the test agent.

Tumor Initiation           i
   Tumor  formation after application of B (a)P or Nissan extract began after a
7- to 8-week latency period and reached a plateau (Figure 2)  for both tumor
multiplicity and tumor incidence.
   The  results of the tumor initiation experiments on SENCAR mice for B(a)P and
for  topside coke oven, coke oven main, Nissan, roofing tar, 01dsmobile,VW Rabbit,
       i i i  i  I i  i i
                 I 10 11 » II H tt I117 « 10 20 21 22 23
                   wieu
     I I  I I I  II  I I  I |  | |  I I I  I I  I I I  I I
                                                        I W 11 II 13 W » M 17 IS It » II H 23 :• 2S
                                                          WEEKS
Fig. 2.  SENCAR mouse  skin  tumor initiation.   Male SENCAR mice were initiated with
         either a single dose  of B(a)P (50.5  pg)  or five daily treatments of
         Nissan extract  (2  mg).   Animals  were then treated biweekly with TPA
         (2 ug).  Left, B(a)P; right,  Nissan  extract.
                             371

-------
Mercedes,  Caterpillar,  residential furnace, and Mustang  extracts  are  shown in
Tables  2 to  12,  respectively.   Animals were scored at  six months  for  papilloma
formation  and  at one year for  carcinomas.  The carcinoma data represent the
cumulative number of carcinomas found in each treatment  group and therefore
include tumors on both  living  and dead animals.  The B(a)P,  topside coke oven,
coke oven  main,  Nissan,  and roofing tar samples produced an  89% or greater tumor
incidence  at the highest dose  level applied.  Tumor multiplicity  ranged from 5 to
6 papillomas per mouse  in the  roofing tar and Nissan samples  to greater than 7 in
the B(a)P, topside coke  oven and coke oven main samples.  These groups of animals
also produced  a  significant number of squamous cell carcinomas, ranging from 13 to
65% of  the mice  bearing  carcinomas at the highest dose evaluated.  In general,
samples which  produced a papilloma response of greater than five papillomas per
mouse at six months produced a carcinoma response of 0.15 to  0.65  carcinomas
per animal,  with 13 to 65% of  the animals bearing at least one tumor per year.
   The  Oldsmobile sample (Table 7)  produced a biphasic response in both male
and female animals;  the  highest activity occurred at 2 mg/mouse with 0.35  to
0.40 papillomas  per mouse and  20 to 40% of the mice bearing tumors.  Some
carcinomas were  observed after one year, but their numbers were not appreciably
above those  observed in  the TPA control animals (Table 2).
   The  VW  Rabbit sample  (Table 8)  produced dose-related  increases in papillomas
in both male and female  mice,  with the maximum activity  for each sex at 10 mg.
At this dose there were  0.34 to 0.47 papillomas per mouse, with 24 to 42%  of the
animals bearing  tumors.   Few carcinomas were scored at one year.
   The  Mercedes  sample  (Table  9)  was also a weak tumor initiator on SENCAR mouse
skin, producing  a response in  male mice at 10 mg of 0.47 papillomas per mouse
and a similar  response in female mice but at 1.00 mg.  As in  the Oldsmobile
sample, the  response in  female mice was biphasic.   Animals scored at one year
produced no  more carcinomas than the TPA controls.
   The  Caterpillar sample (Table 10)  did not elicit a dose response in either
sex of  SENCAR  mice and was marginally higher than the background TPA control.
Carcinoma  formation was  minimal.
   The  residential furnace sample (Table 11)  produced a maximal response at the
highest dose applied (10 mg/mouse)  of 0.29 to 0.38 papillomas per mouse with 21
to 25%  of  the  animals bearing  tumors.  Carcinoma formation was minimal.
   The  Mustang sample (Table 12)  was tested at doses from 0.1 to 3 mg/mouse due
to sample  limitations.   The response was maximal in the  female animals at  3 rag/
mouse and  activity plateaued at 2 to 3 mg/mouse in the male animals.  Twenty
percent of the female mice produced carcinomas at the highest dose tested.
                             372

-------
TABLE  2
 SENCAR MOUSE  SKIN  TUMORIGENESIS
 BENZO(a)PYRENE  - TUMOR INITIATION
Dose
( lag/mouse )


2.
2.
12.
12.
50.
50.
101
101
0
0
52
52
6
e
5
5


(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
No. Mice
Surviving
37
39
40
39
40
37
39
40
38
38
Mice with
Papillomas*
(%)
8
5
45
31
73
57
100
75
95
97
Papilloraas
per Mouse8
0
0
0
0
1
1
5
2
10
7
.08
.05
.50
.44
.8
.1
.8
.8
.2
.9
Mice with
Carcinomas Carcinomas
(%) per Mouseb
5
0
5
5
20
23
25
20
30
25
0

0
0
0
0
0
0
0
0
.05
0
.07
.05
.20
.23
.25
.20
.33
.25
 aScored at 6 months.
 ^Cumulative score after one year.
TABLE 3
SENCAR MOOSE SKIN TUMORIGENESIS
TOPSIDE COKE OVEN - TOMOR INITIATION
Dose
( yg/mouse )








10
10
100
100
500
500
1000
1000
2000
2000
,000
,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
No. Mice
Surviving
40
40
40
40
37
39
39
38
39
40
Mice with
Papillomasa
(%)
13
10
73
70
95
72
95
90
100
100
Papi llamas
per Mouse3
0.
0.
1.
1.
2.
2.
4.
3.
7.
7.
13
20
6
8
6
0
0
5
1
7
Mice with
Carcinomas53 Carcinomas
(%) per Mouseb
0
8
5
15
15
3
13
10
13
20
0
0
0
0
0
0
0
0
0
0

.08
.05
.15
.15
.03
.13
. 10
.15
.23
^Scored at 6 months.
^Cumulative score after one year.
                             373

-------
TABLE 4
SENCAR MOUSE SKIN TUMORIGENESIS
COKE OVEN MAIN - TUMOR INITIATION
Dose
( yg/mouse )
100
100
500
500
1000
1000
2000
2000
10,000
10,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)

-------
TABLE 6
SENCAR MOUSE SKIN TUMORIGENESIS
ROOFING TAR - TUMOR  INITIATION
Dose
( ug/mouse )








10
10
100
100
500
500
1000
1000
2000
2000
,000
,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
No. Mice
Surviving
40
39
40
39
39
40
39
38
39
40
Mica with
Papillomasa
(%)
10
^5
23
13
38
45
36
37
100
95
Papillomas
per Mouse*
0.
0.
0.
0.
0.
0.
0.
0.
6.
5.
13
21
35
15
41
80
62
45
4
7
Mice with
Carcinomas15
(%)
5
10
10
18
5
15
13
15
23
43
Carcinomas
per Mouse
0
0
0
0
0
0
0
0
0
0
.05
.10
.10
.18
.05
.15
.13
. 15
.25
.43
 aScored  at  6  months.
 ^Cumulative score  after  one year.
TABLE 7
SENCAR MOOSE SKIN TUMORIGENESIS
OLDSMOBILE - TUMOR INITIATION
Dose
( yg/mouse )






10
10
100
100
1000
1000
2000
2000
,000
,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
No. Mice
Surviving
40
40
39
40
40
40
39
39
Mice with
Papillomas3
(%)
13
15
26
18
20
40
21
10
Papillomas
per Mouse3
0
0
0
0
0
0
0
0
.13
.18
.36
.25
.35
.40
.21
.13
Mice with
Carcinomas
(%)
8
a
8
8
3
5
8
13
Carcinomas
per Mouse13
0.
0.
0.
0.
0.
0.
0.
0.
10
10
08
08
03
OS
08
13
aScored at 6 months.
^Cumulative score after one year.
                             375

-------
TABLE 8

SENCAR HOUSE SKIN TUMORIGENESIS
VW-RABB1T - TUMOR INITIATION
Dose
( yg/mouse )
100
100
500
500
1000
1000
2000
2000
10,000
10,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
No. Mice
Surviving
40
37
37
40
38
39
38
35
38
38
Mice with
Papillomas3
Papillomas
Mice with
Carcinomas13
(%) per Mouse3 (%)
18
14
14
5
21
18
21
14
24
42
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
18
14
14
05
21
26
24
17
34
47
0
0
0
0
3
3
5
6
5
10
Carcinomas
per Mouseb
0
0
0
0
0.
0.
0.
0.
0.
0.




03
03
05
06
05
10
aScored at 6 months*
^Cumulative score after one year.
TABLE 9
SENCAR MOUSE SKIN TDMORIGENESIS
MERCEDES - TUMOR INITIATION
Dose
(pg/mouse)
100
100
500
500
1000
1000
2000
2000
10,000
10,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
No. Mice
Surviving
39
38
29
39
40
39
40
38
38
40
Mice with
Papillomas3
13
5
21
3
23
21
5
11
37
15
Papillomas
per Mouse3
0
0
0
0
0
0
0
0
0
0
.13
.05
.21
.03
.25
.48
.05
.13
.47
.18
Mice with
Carcinomas Carcinomas
(%) per Mouseb
0
0
5
0
5
0
0
0
5
0
0
0
0.05
0
0.05
0
0
0
0.05
0
^Scored at 6 months.
"Cumulative score after one year.
                             376

-------
TABLE 10

SENCAR MOOSE SKIN TUMORIGENESIS
CATERPILLAR - TUMOR INITIATION



Do so
( tig/mouse )








10
10
100
100
500
500
1000
1000
2000
2000
,000
,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)

No. Mice
Surviving
39
38
39
36
39
40
38
39
40
39
Mice with
Papillomasa
(%)
10
0
10
a
IS
5
11
5
10
5


Papillomas
per Mouse3
0

0
0
0
0
0
0
0
0
. 15
0
.13
.08
.15
.05
.11
.05
.10
.05
Mice with


Carcinomas Carcinomas
(%) per Mouseb
5
0
0
0
3
3
3
0
0
0
0
0
0
0
0
0
0
0
0
0
.05



.03
.03
.03



aScored at 6 months.
^Cumulative score after one year
TABLE 11

SENCAR MOUSE SKIN TUMORIGENESIS
RESIDENTIAL FURNACE - TUMOR INITIATION
Dose
( yg/mouse )








10
10
100
100
500
500
1000
1000
2000
2000
,000
,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
No. Mice
Surviving
40
40
39
40
40
40
39
40
38
40
Mice with
Papillomas3
(%)
0
0
5
10
18
3
5
3
21
25
Papillomas
per Mouse3
0
0
0
0
0
0
0
0
0
0


.05
.10
.20
.03
.05
.03
.29
.38
Mice with
Carcinomas
(%)
8
3
3
8
3
0
3
3
5
8
Carcinomas
per Mouse
0
0
0
0
0
0
0
0
0
0
.08
.03
.03
.08
.03

.03
.03
.05
.08
aScored at 6 months.
^Cumulative score after one  year.
                             377

-------
TABLE 12

SENCAR MOOSE SKIN TOMORIGENESIS
MUSTANG - TUMOR INITIATION
Dose
( yg/mouse )
100
100
300
500
1000
1000
2000
2000
3,000
3,000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
No. Mice
Surviving
39
39
39
38
40
40
37
39
34
40
Papillomae*
(*)
5
13
13
18
18
10
22
21
18
23
Papillomas
per Mouse*
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
05
23
15
24
18
13
24
23
24
28
Carcinomas
(%)
5
13
0
10
5
10
15
13
5
20
Carcinomas
per Houseb
0
0
0
0
0
0
0
0
0
0
.OS
.13

.10
.OS
.10
.15
.13
.05
.20
 aScored at 6 months.
 ^Cumulative score after one year.

    The lack of a monotonic dose response across the  complete dose  range tested in
 the Oldsmobile, VW Rabbit, Mercedes, Caterpillar and Mustang samples may indicate
 a toxic response to these samples.  Damage to the  skin  epidermal cells will
 result in lower expression of the tumorigenic response  by these  complex mixtures.
 This is particularly clear with the Oldsmobile and Mustang samples, where a
 5-fold increase in dose (to 10 mg/mouse) results in  equal or lower tumor
 response.  In these examples the highest dose  (10  mg) was administered in five
 daily increments of 2 mg, which should have lowered  the toxic responses when
 compared to a single administration.
 Complete.Carcinogenesis
    Six agents were examined for their ability  to act as complete carcinogens
 in the SENCAR mouse skin system: benzo(a)pyrene,.coke oven main, roofing tar,
 Nissan, Oldsmobile, and Caterpillar extracts.  Benzo(a)pyrene (Table  13),
 when applied once per week produced greater than 93% carcinoma incidence at
 50.5 yg/week, with almost one carcinoma per mouse.  Higher doses did  not
 increase the tumor multiplicity.
    Coke oven main also produced a strong complete  carcinogen response in both
 male and female mice  (Table 14).  Male mice seemed to be more sensitive,
 as 98% of the mice bore approximately one  carcinoma, while only 75% of the
 female mice responded.  The roofing tar sample produced a significant response
 only at the highest dose applied  (4 mg per mouse per week) with 25 to 28% of
 the mice bearing tumors  (Table 14).
                              378

-------
   The diesel samples  (Nissan, Oldsmobile  and  Caterpillar)  were  essentially  in-
active as complete carcinogens at  the  doses  applied  on  SENCAR mouse  skin
(Table 15).  No tumors were observed in  the  negative control  animals (Table  13).
TABLE  13

SENCAR MOOSE SKIN TOMORI GENES IS
BENZO(a)PYRENE - COMPLETE  CARCINOGENESIS
Dose iig/mouse/week
12.6
12.6
25.2
25.2
50.5
50.5
101
101
202
202
0
0
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
Mice with carcinomas*
{%)
10
8
63
43
93
98
80
90
80
93
0
0
Carcinomas
per mouse*
0.10
0.08
0.63
0.43
0.93
0.98
0.83
0.98
0.80
0.98
0
0
aCumulative  score  after  one  year.
TABLE 14

SENCAR MOOSE SKIN TOMORIGENESIS
COKE OVEN MAIN AND ROOFING TAR - COMPLETE CARCINOGENESIS
Mice with carcinomas
Dose
( jig/mouse/ week )
100 (M)
100 (F)
500 (M)
500 (F)
1000 (M)
1000 (F)
2000 (M)
2000 (F)
4000 (M)
4000 (F)
Coke Oven
Main
5
5
36
30
48
60
82
78
98
75
Roofing
Tar
0
0
0
0
3
0
3
8
25
28
a
Carcinomas per mouse
Coke Oven
Main
0.05
0.05
0.36
0.30
0.55
0.60
1.00
0.78
0.98
0.85
Roofing
Tar
0
0
0
0
0.03
0.03
0.08
0.28
0.28
^Cumulative  score  after  one  year.
                            379

-------
TABLE  15

SENCAR MOUSE SKIN TOMORIGENESIS
NISSAN OLDSMOBILE AND  CATERPILLAR - COMPLETE CAPCINOGENESIS
   Dose
 (yg/mouse/
Mice with carcinomas3  (%)
Carcinomas per mouse3
week)
100
100
500
500
1000
1000
2000
2000
4000
4000
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
(M)
(F)
Nissan
0
0
0
0
0
0
0
0
3
5
Oldsmobile
0
0
3
0
0
0
0
0
0
0
Caterpillar
0
3
0
0
0
0
0
0
0
0
Nissan Oldsraobile
0
0
0
0
0
0
0
0
0.03
0.05
0
0
0.03
0
0
0
0
0
0
0
Caterpillar
0
0,03
0
0
0
0
0
0
0
0
"Cumulative score after  one  year.
Quantitative  Analysis
   The mouse  skin data were stored in computer files as described  in the
Materials and Methods section.   These data were subjected to computer modeling
and statistical procedures specifically designed for analysis of tumor multi-
plicity and tumor incidence data using an interactive computer terminal graphics
system.  Tumor incidence data were applied to a probit model with  background
correction.   From the model the dose which elicits tumors in 50% of the animals
over the TPA  control  rate (TID5Q)  was estimated.  An example of the data and
probit analysis for the Nissan  sample is shown in Figure 3a.  The  plot of mice
with papillomas vs dose applied shows the data points and the probit curve calcu-
lated from the data.   The TID5Q and associated 95% confidence intervals calculated
from the fitted parameters are  shown as well as the raw data.  Tumor multiplicity
data were analyzed by a. nonlinear Poisson model with a background  correction term.
The data were fitted  to the model and the model parameters were estimated; from
these values the  number of papillomas per mouse at 1 rag and the associated 95%
confidence intervals  were estimated.  An e.xampla of the graphics display is shown
in Figure 3b.
                             380

-------
TEST AGENT CODE: 9831 PROTOCOL: TI
TEST AGENT NAME: NISSAN DCH 1979
DOSE MICE *PAPS
.888 37 3.188
188.989 37 .988
598.898 38 26.315
1988.888 48 32.588
2988.888 35 £3.714
18888.988 38 89.474
PROSIT HODEL WITH BACKGROUND
ESTIMATES
SETA INITIAL FINAL ASYH UAR
9 .9811 .8448 .9883
1 -2.3283 -3.6797 .7488
2 .3871 .7671 .9133
TEST CHI-Sfl DF P
G-O-F 3.73 3 .1243
DOSE 192.92 2 .9888
ESTIMATE LOWER 93X UPPER
ED38 1322.93 1188.93 2196.73
TD38 1642.18 1289.64 2229.17
TEST AGENT CODE: 983 i PROTOCOL: TI
TEST AGENT NABE: NISSAN DCI1 1979
DOSE »«ICE *PAPS
.888 118 3.883
188.898 39 2.564
388.888 39 23.877
1888.988 38 39.474
2888.888 48 57.588
18888.988 38 97.363
HONLIN POISSON HODEL WITH BACKGROUND
ESTIMATES
BETA INITIAL FINAL ASYH UAR
9 .8393 .8479 .8883
1 -1.7472 -7.1132 .1688
2 .3138 .9624 .8822
TEST CHI -SO OF P
POISS 326.69 386 .1998
ADOCY 9.81 3 .8291
DOSE 674.83 2 .8888
PAPSxH ? 1 HG LOWER 93V. UPPER
SPEC .676 .373 .798
EXCS .628 .321 .737
STRAIN: s SEX: n WEEK; 26
START DATE: 972879
rtEAH S.D.
.881 .277
.888 .888
.342 .627
.373 .386
1.143 1.167
3.474 3.269
OSS I EXP
160
's,
u
I
T
H 59 —
P
A
P
S
:
<
r"
/
US DOSE
***

(a)
3888 18890
MICROGRAHS
STRAIN: s SEX: F WEEK: 26
START DATE: 872879
riEAN S.D.
.939 .271
.926 .168
.383 .847
.326 .762
1.698 1.892
5.658 3.636
Q8S I EXP US DOSE
5
P
A
P
/
H
0
U 2
S 2
E

/
^
&
a *«
/
/

ea 19
: ,(b)


889
                                                       niCROGRAHS
Fig. 3.  Samples of computer-generated analyses of tumor data:  (a) tumor
         incidence data and probit analysis for the Nissan sample, (b) nonlinear
         Poisson model analysis of tumor multiplicity data for the Nissan sample.
                             381

-------
Tumor Promotion
   The coke oven main and  roofing tar samples were applied weekly to mice pre-
viously initiated with a single dose of 50.5 yg of benzo(a)pyrene.  Coke oven
main produced a response equal to the positive control TPA but at 1/500 the dose
applied (Table 16).  The roofing tar sample was also active as a tumor promoter
(Table 16) and produced a  dose-related effect up to the highest dose applied.
Mice treated with only a single dose of B(a)F produced no tumors.
TABLE 16

SENCAR MOUSE SKIN TOMORIGENESIS
COKE OVEN MAIN AND ROOFING TAR -  TUMOR PROMOTION
Dose
( yg/mouse/

0 (M)b
0 (F)
100 (M)c
100 (F)
500 (M)
500 (F)
1000 (M)
1000 (F)
2000 (M)
2000 (F)
4000 (M)d
4000 (F)
TPA, 4 yg (M)e
TPA, 4 yg (F)
Mice with
«
Coke Oven
Main
0
0
3
10
26
38
53
68
84
85
100
100
86
97
Papillomas3
>)
Roofing Tar

0
0
0
0
5
0
20
16
23
13
55
30
100
100
Papillomas per mouse3
Coke Oven Roofing Tar
Main
0
0
0.02
0.10
0.44
0.83
1.2
1.2
2.5
3.1
8.2
8.8
3.1
5.9

0
0
0
0
0.05
0
0.27
0.36
0.32
0.15
1.2
0.6
5.2
7.2
"Scored at 34 weeks.
bMice initiated with B(a)P  (50.5  yg)  and subsequently treated weekly with
 acetone.
cMice initiated with B(a)P  (50.5  yg)  and subsequently treated weekly with Coke
 Oven Main or Roofing Tar.
^ice initiated with B(a)P  (50.5  yg)  and subsequently treated twice weekly with
 2 mg Coke Oven Main or Roofing Tar.
aMice initiated with B(a)P  (50.5  yg)  and subsequently treated twice weekly with
 2 yg. TPA.
                             382

-------
DISCUSSION
   The SENCAR mouse,  specifically bred for increased sensitivity towards two-
stage  (initiation-promotion)  carcinogenesis,  has  demonstrated its ability to
respond to carcinogens.   '     of  three mouse  strains and stocks  examined, the
SENCAR mouse was  the  most sensitive  to the initiating effects of B(a)P  (Table 17)
with the C57 Black  strain completely inactive.  The  exact nature of  theinability
of C57 Black mice to  respond  to B(a)P is  unknown,  but certain lines  of  evidence
indicate that there is a  lack of  promotion response  in their  skin epithelial
  11  36
cells.
   This study of  the  effects  of ten  complex mixtures and B(a)P on SENCAR mouse
skin is the most  extensive  to date and the results confirm the applicability
of this mouse strain  to the analysis of complex mixtures.  The qualitative
results from these  studies  as summarized  in Table  18 are based on decisions from
the following empirical rules:   (1)  a tumor initiation-promotion assay  is consi-
dered positive for  papilloma  formation if there is evidence of a dose response
and if at least two doses yield a. papilloma-per-mouse value equal to three times
the background value, and (2)  a tumor initiation or  complete  carcinogenesis assay
is considered positive for  carcinoma formation if  at least one dose produces  a
tumor incidence of  at least 20%.
TABLE 17
COMPARISON OF THE TUMOR INITIATING ACTIVITY OF
BENZO(a)PYRENE IN THREE MOOSE STRAINS AND STOCKS8
Strain
(Stock)
SENCAR



CD 1



C57 Black





B(a)P (ug)
50.4
25.2
12.6
2.5
50.4
25,2
12,6
2.5
404
202
101
50,4
25.2
12.6
Papillomas
per Mouse"
8.2
3.0
1.6
0.9
3.8
1.8
0.7
0.1
0
0
0
0
0
0
Mice with
Papillomasb
100
80
60
42
72
58
40
10
0
0
0
0
0
0
aData taken from DiGiovanni et al., (29), and Slaga and
 Nesnow (unpublished).
^Scored at six months.
                             383

-------
CO
Co
                   TABLE  18

                   SUMMARY


Sample
Benzol a Jpyrene
Topside Coke Oven
Coke Oven Main
Roofing Tar
Nissan
Oldsmobile
VW Rabbit
Mercedes
Caterpillar
Residential
Furnace
Mustang
Tumor

Papillomas
+A°
+A
+A
+A
+A
+A
+A
V-
-/-

-/-
+A
Initiation

Carcinomas
+A
-A
+A
+A
+A
-/-
-/-
-/-
-/-

-/-
-A
Complete Carcinogeneais

Carcinomas
+A
NDd
+A
+A
-/-
-/-
ie
ND
-/-

ND
ND
Tumor
Promotion

Papillomas8
*A
ND
V+
V+
ND
ND
ND
ND
ND

ND
ND
                   aScored-at 6 months.
                   "Cumulative score at 1 year
                   cMale/Feraale.
                   dND = Not Determined.
                   el = Incomplete.

-------
   Benzo(a)pyrena, coke oven main, and roofing tar  samples were positive in both
sexes as tumor initiators  (papillomas and carcinomas) , tumor promoters and complete
carcinogens.  In general,  those agents which produced a  strong tumor-initiation
papilloma response also produced carcinomas in the  same  animals when scored at one
year.  Four diesel samples were positive as tumor initiators  (Nissan, Oldsmobile,
VW Rabbit and Mercedes) as was the gasoline engine  sample  (Mustang).  Of all the
strong tumor initiators, Nissan extract was the only sample which was not a complete
carcinogen at the doses tested and presumably had no  tumor promoting  activity on
mouse skin.  In fact, none of the diesel samples evaluated was found  to be a com-
plete carcinogen in the dose ranges tested.  Benzo(a)pyrene,  coke oven main, and
roofing tar, all which were complete carcinogens, possessed tumor-promoting acti-
vity.  The lack of tumor-promoting activity in the  Nissan sample is probably a
function of the composition of the Nissan mixture. The skin tumorigenesis results
indicate that the coke oven main was a stronger tumor promoter than the roofing
tar sample.  Chemical fractionation and mutagenesis studies show that both the
chemical composition and genetically active components of diesel,roofing tar and
coke oven main samples are significantly different  (J. Lewtas,personal communica-
tion) .
   Tumor initiators on mouse skin may also possess  complete carcinogenic activity
when administered by other routes to mice and rats.   A review of the  literature
                       ^7                        op
indicates that urethane    and triethylenemelamine   are  both  probably pure
mouse skin tumor initiators:  repeated applications of these  agents on mouse
skin do not yield tumors.  However, urethane administered intraperitoneally,
subcutaneously or orally to mice produced a variety of lesions, including lung,
liver and lymphoid tumors.  Urethane administered orally to rats also produces
multiple tumors.  Triethylenemelamine produces lung tumors in mice after intra-
peritoneal injection and muscle tumors in rats after  subcutaneous injection.
   It is compelling to postulate that the B(a)P in  these complex mixtures
could account for their tumorigenic activity, since mouse sfcin is exquisitely
sensitive to this agent.   The results presented here  reveal that a single
application of less than 5 ug of B(a)P will yield a 50%  tumor incidence as
a tumor initiator.  However, the relationship between B(a)P content in each
mixture, and papilloma response for each mixture, is  not linear (Figure 4).
Probably none of the activity of the coke oven sample can be  explained by B(a)P
content, as the B(a)P-induced tumor response at the B(a)P level in the coke
oven sample is quite small.  Even the B(a)P level in  the Nissan sample  (11 ug/
lOmg extract) can only account for 20 to 30% of the papilloma response elicited
                             385

-------
CO
00
CT1
   10.0
    9.0
    6.0
 $  7.0
£  6.0
"g  5.0
 §•
Q.
          3.0
          2.0
          1.0
                                                      Coke Oven
                             Roof
                              Tor
            VW Rabbit
                           Denzo(a)pyrene
            0.01 Olds
                 Cat
0.10
                                         1.00
10.00
100.00
                                    Benzo(a)pyrene,
           Fig. 4.  Relationship of skin tumor initiating activities of complex mixtures
                 to their benzo(a)pyrene levels:  comparison to skin tumor initiating
                 activity of purre benzo (a)pyrene.

-------
by the Nissan sample.  Other components of  the mixtures may play an  important role
in their tumorigenic activities.
   Quantitative methods for the analysis  of tumor data are many and  employ
tumor incidence, tumor multiplicity, and  tumor latency data.  Statistical methods
have been employed using Poisson and other  distribution assumptions, as well as
both uni- and multivariate analytical  approaches.   '   '       We have chosen to
apply a nonlinear Poisson modal to  the papilloma incidence data. This model assumes
a Poisson distribution of tumors, that tumor multiplicity is related to dose,that
the response may be nonlinear, and  that there is a  background response.Estimates
from the models are presented only  if  they  are in the range from which the data
were obtained and if the observed data adequately fit the calculated model
(Table 19) .
   Results from the nonlinear Poisson  model suggest the following ranking:
topside coke oven > Nissan >. roofing tar  >.  VW Rabbit = Mustang, The values calcu-
lated are only estimates and in some cases  all the  assumptions made  to derive the
estimates are only partially fulfilled.
   A probit model has been chosen to evaluate the tumor incidence data.  The
probit model examines animals with  tumors (regardless of multiplicity)and animals
without tumors.  Results from the probit  analysis suggest the ranking:  B(a)P >
coke oven main >_ topside coke oven  > Nissan = roofing tar. These are not the only
models which can be applied to these data,  and although they appear  effective in
this case, more effort is being placed in improving statistical and  modeling
techniques.
   In addition to the tumorigenesis studies described above, detailed gross and
histopathological analyses of selected animals have been undertaken.  Further
results from these detailed pathological  studies on the formation of internal
tumors and the appearance of tumors with  longer latency periods will be
presented at a later date.

ACKNOWLEDGEMENTS
   The authors wish to thank R.L. Bradow, R.H. Jungers, B.D. Harris, T.O.Vaughn,
R.B. Zweidinger, K.M. Gushing, J. Bumgarner, and B.E. Gill for the sample collec-
tion, preparation, and characterization,  and C.J. Alden and J.L. Wilson for
assistance in preparation of the manuscript.  The research was sponsored by the
0.S. Environmental Protection Agency,  contract no.  79D-X0526, under  the Inter-
agency Agreement, U.S. Department of Energy no. 40-728-78, and the Office of
Health and Environmental Research,  U.S. Department  of Energy, under  contract
no. 7405 eng-26 with the Union Carbide Corporation.
                             387

-------
TABLE 19
SENCAR MOUSE SKIN TUMOR INITIATION
ESTIMATES FROM TWO MODELS BASED OH PAPILLOMA DATA AT 6 MONTHS3
Nonlinear Poisson
Papillomas/Mouse
at 1 mg
Benzo(a)pyrene M
F
Coke Oven Main M
OJ _
00 F
00
Topside Coke Oven M
F
Nissan M
F
Roofing Tar M
F
VW Rabbit M
F
Mustang M
b
NDC
2.2d
2.0d
0.49d
0.68d
0.38d
0.44d
0.21
0.17
0.17
95%
Confidence Intervals


2.60
1.90
0.38
0.57
0.30
0.35
0.14
0.11
0.12


- 2.40
- 2.20
- 0.63
- 0.79
- 0.49
- 0.55
- 0.30
- 0.25
- 0.24
Problt

Dose for 50% Papilloma 95%
Incidence (TID50), mg Confidence Intervals
0.0036
0.0091
0.079
0.19
0.30
0.42
1.60
1.50
1.8
2.1
e


0.0021
0.0057
0.027
0.14
0.22
0.31
1.2
1.1
1.2
1.5



- 0.0062
- 0.015
- 0.23
- 0.28
- 0.40
- 0.58
- 2.2
- 1.9
- 2.7
- 2.8



aEstlmates calculated from models according to Materials and Methods section.
^Not calculated since data were obtained at a lower dose range.
CND = not determined.
aThe distribution of tumors at all dose levels was not Poisson as the variances exceeded the means.
eNot calculated since tumor incidence did not equal 50%.

-------
REFERENCES

  1.  Huisingh, J., Bradow, R., Jungars, R., Claxton, L., Zweidinger, R.,
      Tejada, A., Bumgarner, J., Duffield,  F. , Waters, M., Simmon, V., Hare, C. ,
      Rodriguez, C. and Snow,  L.  (1979) Application of bioassay to the charac-
      terization of diesel particle emissions.   In:  Application of Short-term
      Bioassays in the Fractionation and Analysis of Complex Environmental Mixtures
      Waters, M.D., Nesnow, S., Huisingh, J.L.,  Sandhu,  S.S. and Claxton, L., eds.
      Plenum Press:  New York, pp. 383-418.
  2.  Huisingh, J.L., Bradow,  R.L., Jungers, R.H., Harris, B.D., Zweidinger, R.B.,
      Cushing, K.M., Gill, B.E., and Albert, R.E.  (1980)  Mutagenic and
      carcinogenic potency of  extracts of diesel and related environmental
      emissions:  study design, sample generation, collection, and preparation.
      In:  Health Effects of Diesel Engine  Emissions.  Proceedings of an Inter-
      national Symposium.  Pepelko, W.E., Danner,  R.M. and Clarke, N.A., eds.
      Vol 2, Washington, D.C., U.S. Govt. Printing Office (EPA Publication No.
      EPA-600/9-80-057b), pp.  788-800.
  3.  Pepelko, W.E., Danner, R.M. and Clarke, N.A.  (1980) Health Effects of
      Diesel Engine Emissions:  Proceedings of an  International Symposium, Vol. 1,
      Washington D.C., U.S. Govt. Printing  Office  (EPA publication No. EPA-600/
      9-80-057a), pp. 1-227.
  4.  Katz, M and Pierce, R.C. (1976) Quantitative distribution of polynuclear
      aromatic hydrocarbons in relation to  particle size  of urban particulates.
      In:  Carcinogenesis, A Comprehensive  Survey.  Freudenthal, R.I. and Jones,
      P.M., eds.  Raven Press, New York, pp. 413-429.
  5.  Thomas, R.S., Lao, R.C., Wang, D.T.,  Robinson, D. and Sakuma, T. (1978)
      Determination of polycyclic aromatic  hydrocarbons in atmospheric
      particulate matter by gas chromatography-mass spectrometry and high
      pressure liquid chromatography.  In:  Carcinogenesis, A Comprehensive
      Survey.  Jones, P.W. and Freudenthal, R.I.,  eds.  Raven Press, New York,
      pp. 9-19.
  6.  Daisey, J.M., Leyko, M.A. and Kneip,  T.J.  (1979) Source identification
      and allocation of polynuclear aromatic hydrocarbon  compounds in the
      New York City aerosol: Methods and applications.  In:   Polynuclear Aromatic
      Hydrocarbons, Jones, P.W. and Leber,  P., eds.  Ann Arbor Science Publishers,
      Ann Arbor, pp. 201-215.
  7.  Bjorseth, A. (1979) Determination of polynuclear aromatic hydrocarbons
      in the working environment.  In:  Polynuclear Aromatic Hydrocarbons.
      Jones, P.W. and Leber, P., eds.  Ann Arbor Science Publishers, Ann Arbor,
      pp. 371-381.
  8.  Greenberg, A., Yokoyama, R., Giorgio, P.  and Cannova,  F. (1980)
      Analysis of polynuclear  aromatic hydrocarbons on the airborne particles
      of urban New Jersey.  In:  Polynuclear Aromatic Hydrocarbons:  Chemistry
      and Biological Effects,  Bjorseth, A. and Dennis, A.J.  eds.  Battelle Press,
      Columbus, pp. 193-198.
  9.  Kaiser, C., Kerr, A., McCalla, D.R.,  Lockington, J.N.  and Gibson,  E.S.
      (1980) Mutagenic material in air particles in a steel foundry.  In:
      Polynuclear Aromatic Hydrocarbons:  Chemistry and Biological Effects.,
      Bjorseth, A. and Dennis, A.J., eds.  Battelle Press, Columbus, pp. 579-588.
 10.  Kites, R.A., Yu. M.L. and Thilly, W.G. (1981) Compounds associated
      with diesel exhaust particulates.  In:  Chemical Analysis and Biological
      Fate:  Polynuclear Aromatic Hydrocarbons.  Cooke, M. and Dennis, A.J.,eds.
      Battelle Press, Columbus, pp. 455-466.
 11.  Gold, A. (1972) Carbon black adsorbates:    separation and identification
      of a carcinogen and some oxygenated polyaromatics.  Anal Chem.
      47:1469-1472.
                              389

-------
12.  Waters, M.D., Nesnow, S., Huisingh,  J.L.,  Sandhu, S.S. and Claxton, L.
     eds. (1979) Application of Short-term Bioassays in the Fractionation
     and Analysis of Complex Environmental Mixtures.  Plenum Press, New
     York.
13.  Waters, M.D., Sandhu, S.S., Huisingh,  J.L.,  Claxton, L. and Nesnow, S.
     eds. (1981) Application of Short-term Bioassays to the Analysis of
     Complex Environmental Mixtures.  Plenum Press,  New York.
14.  Gold, A., Nesnow, S., Moore, M., Garland,  H.,  Curtis, G., Howard,  B.,
     Graham, D. and Eisenstadt, E.  (1980)  Mutagenesis and morphological
     transformation of mammalian cells by a non-bay region polycyclic cyclopenta-
     (cd)pyrene and its 3,4-oxide.  Cancer Res. 40:4482-4484.
15.  Menster M., and Sharkey,  Jr., A.G.  (1977) Chemical characterization of
     diesel exhaust particulates.  Oak Ridge, Tenn., ERDA Technical Information
     Center, PERC/RI-77/5, pp. 1-47.
16.  Wynder, E.L. and Hoffmann, D. (1962)  A Study of air pollution carcinogenesis:
     III.  Carcinogenic activity of gasoline engine  exhaust condensate.  Cancer
     15:103-108.
17.  Hoffmann, D. and Wynder, E.L.  (1963)  Studies on gasoline engine exhaust.
     J, Air Pollut. Contr. Assoc. 13:322-327.
18.  Hoffmann, D., Theisz, E. and Wynder,  E.L.  (1965)  Studies on the
     carcinogenicity of gasoline exhaust.   J. Air Pollut. Contr. Assoc.
     15:162-165.
19.  Brune,  H.F.K. (1977) Experimental results with  percutaneous applications
     of automobile exhaust condensates in mice.  In:   Air Pollution and Cancer
     in Man.  Mohr, U., Schmahl, 0., Tomatis, L. and Davis, W.,  eds. Lyon:
     International Agency for Research on Cancer.   IARC Scientific Publications,
     No. 16, pp. 41-47.
20.  Misfeld, J. and Tim, J. (1978) The  tumor-producing effect of automobile
     exhaust condensate and fractions thereof.  Part III:  Mathematical-
     statistical evaluation of the test results.  J.  Environ. Path, and Toxicol.
     1:747-772.
21.  Misfeld, J. (1980) The tumor-producing effects  of automobile
     exhaust condensate and of diesel exhaust condensate:  mathematical-
     statistical evaluation of test results. Ini Health Effects of Diesel Engine
     Emissions.  Proceedings of an International Symposium, Pepelko, W.E.,
     Danner, R.M. and Clarke, N.A. eds.   Vol 2.  Washington, D.C., U.S.  Govt.
     Printing Office (EPA publication No.  EPA 600/9-80-057b).  pp. 1012-1025.
22.  Mittler, S. and Nicholson, S.  (1957)  Carcinogenicity of atmospheric
     pollutants.  Ind. Med. and Surg. 26:135-138.
23.  Kotin,  P., Falk, H.L. and Thomas, M.  (1954) Aromatic hydrocarbons.  II.
     Presence in the particulate phase of gasoline-engine exhausts and  the
     carcinogenicity of exhaust extracts.   AMA Arch.  Ind. Hyg.  Occup. Med.
     9:164-177.
24.  Kotin,  P., Falk, H.L., and Thomas, M.  (1955) Aromatic hydrocarbons.  III.
     Presence in the particulate phase of diesel-engine exhausts and the
     carcinogenicity of exhaust extracts.   AMA Arch  Ind Health,  11:113-120.
25.  Slaga,  T.J., Triplett, L.L. and Nesnow,  S. (1980)  Mutagenic and
     carcinogenic potency of extracts of  diesel and  related environmental
     emissions:  two-stage carcinogenesis in skin tumor sensitive mice  (SENCAR) •
     In:  Health Effects of Diesel Engine Emissions,   proceedings of an
     International Symposium.  Pepelko, W.E., Danner,  R.M., Clarke, N.A. eds.
     Vol 2.   Washington, D.C.  U.S. Govt.  Printing  Office (EPA publication No.
     EPA-600/9-80-057b), pp. 874-897.
                              390

-------
26.  Nesnow, S., Triplett, L.L., and Slaga, T.J.  (1981) Tumorigenesis of diasel
     exhaust, gasoline exhaust, and related emission extracts on SENCAR mouse
     skin.  In:  Short-term Bioassays in the Analysis of Complex Environmental
     Mixtures.  Waters, M., Sandhu, S.S., Lewtas, J., Claxton, L. and Nesnow, S.,
     eds.  New York:  Plenum Press, pp. 277-297.
27.  Slaga, T.J., Gleason, G.L., Mills, G., Ewald, L., Fu, P.P., Lee, H.M.
     and Harvey, R.G.  (1980) Comparison of the skin tumor-initiating activities
     of dihydrodiols and diol-epoxides of various polycyclic aromatic
     hydrocarbons.  Cancer Res. 40:1981-1984.
28.  Iyer, R.P., Lyga, J.W., Secrist, III., J.A., Daub, G.H. and Slaga, T.J.
     (1980) Comparative tumor-initiating activity of methylated benzo(a)
     pyrene derivatives in mouse skin.  Cancer Res. 40:1073-1076.
29.  DiGiovanni, J., Slaga, T.J. and Boutwell, R.K.  (1980) Comparison of the
     tumor-initiating activity of 7,12-dimethylbenz(a)anthracene and
     benzo(a)pyrene in female SENCAR and CD-I mice.  Carcinogenesis 1:381-389.
30.  Hennings, H., Devor, D., Wenk, M.L. Slaga,  T.J., Former, B., Colburn, N.H.,
     Bowen, G.T. Elgjo, K., and Yuspa, S.H.  (1981) Comparison of two stage
     epidermal carcinogenesis initiated by 7,12-dimethylbenz(a)anthracene or
     N-methyl-N'-nitro-N-nitrosoguanidine in newborn and adult SENCAR and
     Balb/c mice.  Cancer Res. 41:773-779.
31.  Slaga, T.J., Fischer, S.M., Triplett, L.L.  and Nesnow, s. (in press).
     Comparison of complete carcinogenesis and tumor initiation in mouse skin:
     Tumor initiation promotion a reliable short-term assay.  J.  Environmental
     Pathology and Toxicology.
32.  Boutwell, R.K. (1964) Some biological aspects of skin carcinogenesis.
     Progr. Exptl. Tumor Res. 1964; 4:207-250.
33.  Dodge, H.F. and Romig, H.G. (1959) Sampling Inspection Tables:  Single and
     Double Sampling.  John Wiley and Sons, New  York.
34.  Finney, P.J.  (1971) Probit Analysis, Cambridge University Press, Cambridge.
35.  Stead, A.G., Hasselblad, V., Creason, J.P.  and Claxton, L. (1981) Modeling
     the Ames Test.  Mutation Res. 85:13-27.
36.  Reiners, Jr., J. , Davidson, K. Nelson, K.,  Mamrack, M. and Slaga, T.J.
     (in press)  Skin tumor promotion:  A comparative study of several stocks
     and strains of mice.In -. Organ and Species Specificity in Chemical Carcino -
     genesis, Langenbach, R., Nesnow, S. and Rice, J., eds.  Plenum Press, New
     York.
37,  IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to
     Man  (1974) International Agency for Research on Cancer, Lyon, France,
     Vol. 7, pp. 111-140.
38.  IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to
     Man  (1975) International Agency for Research on Cancer Lyon, France, Vol. 9
     pp. 95-105.
39.  Drinkwater, N.R. and Klotz, J.H. (1981) Statistical methods for the
     analysis of tumor multiplicity data.  Cancer Res., 41:113-119.
40.  Gart. J.J., Chu, K.C., and Tarone, R.E.  (1979) Statistical issues in
     interpretation of chronic bioassay tests for carcinogenicity.  J. Nat.
     Cancer Institute 62:957-974.
41.  Totter, J.R. and Finamore, F.J.  (1978) Dose response to cancerogenic
     and mutagenic treatment.  Environment International 1:233-244.
                             391

-------
DERMAL CARCINOGENESIS BIOASSAYS  OF DIESEL PARTICULATES AND DICHLOROMETHANE
EXTRACT OF DIESEL PARTICULATES IN  C3H MICE

LINVAL R. DEPASS+, K.C. CHEN""" AND LOT! G. PETERSON*
+Bushy Run Research Canter, Export,  Pennsylvania;  ++Biomedical Science Depart-
ment, General Motors Research Laboratories,  Warren, Michigan
ABSTRACT
   Diesel participates  (DP) and  dichloromethane (DCM) extract of DP were tested
to assess their potential as complete  carcinogens and as initiators or pro-
moters of carcinogenesis.  The test  agents  were applied as suspensions in ace-
tone to the dorsal skin of 40 male C3H mice per group at various concentrations
to obtain information on dose-response relationships.  Dosing was performed 3
times weekly in the initiation and complete carcinogenesis studies and 5 times
per week in the promotion studies.   Positive control groups received repeated
applications of benzo[a]pyrene (BaP) for  complete carcinogenesis, or a single
application of BaP followed by repeated applications of phorbol myristate ace-
tate (PMA) for the initiation and promotion studies.   The test agents were
applied in place of BaP in the complete carcinogenesis and initiation studies
and in place of PMA in  the promotion studies.   One tumor-bearing animal has
been observed at the highest dosage  of DCM   extract in the complete carcino-
genesis study.  In the promotion study, 1 and  2 tumor-bearing mice, respec-
tively, have been observed in the 2  highest dosage groups of DCM extract.  In
the initiation study, 3, 3, 2 and 1  tumor-bearing mice have been observed in
the groups that received DP, DCM extract  and 2  negative control groups re-
spectively.  The results of the  initiation  and  promotion studies suggest that
the test agents did not significantly  increase  tumor incidence compared to the
controls.  The results of the complete carcinogenesis study are equivocal be-
cause of the single tumor observed,  and the absence of tumors in the concurrent
and historical controls.  Since  the  studies were not  completed at the time of
manuscript preparation, final conclusions have  been deferred until all the data
are available.

INTRODUCTION
   In recent years, the need for improved fuel  economy has led to an increase
in the production and use of fuel-efficient diesel vehicles.  Because of the
considerably higher particulate  emissions of diesel engines compared with
                            392

-------
gasoline engines1 it has become  important  to  determine  the  toxicological  im-
pact of increased exposure of  the  general  population  to  dlesel  particulates.
Since chemical analysis of organic extracts of  diesel particulates  has demon-
strated the presence of polycyclic aromatic hydrocarbons, including benzo[a]-
pyrene   , the possible oncogenicity  of  diesel  emissions is an  important  public
health consideration.  In fact,  previous studies  hava shown that extracts of
diesel particulates are mutagenic  in  Salmonella assays'*'5 and oncogenic for
mouse skin.6'7
   With respect to oncogenicity, Kotin et  al6 reported  the  induction of skin
tumors in C57 and A strain mice  painted  with  diesel exhaust extract.  More
recently, Slaga et al7 reported  the initiation  of papillomas in SENCAR mice
treated with dichloromethane extract  of  particulates  from Nissan and Olds
diesel engines, although the latter extract had extremely low activity.   Cater-
pillar diesel extracts were negative  in  those studies.
   The present studies were designed  to  assess  more definitively the potential
of diesel emission extracts as complete  carcinogens and  as  initiators or  pro-
moters of carcinogenesis.

MATERIALS AND METHODS
   Test Substances.  Samples of  diesel particulates (DP) and dichloromethane
(DCM) extract of DP were supplied  by  General  Motors Research Laboratories to
the Bushy Run Research Center  on a regular basis  throughout the study.  The
samples were collected from a  GM Oldsmobile 350D  engine  with a  road load  con-
dition of 65 kilometers/hour.  The particulates were  collected  in a bag house
filter at a temperature of 100 ± 10°C.   The DCM extracts were prepared by a
Soxhlet continuous extraction  procedure.   The samples were  shipped  on dry ice
and stored in a freezer (-12°C)  except during preparation of dilutions.
   Phorbol 12-myristate 13-acetate (PMA) from PL  Biochemicals,  Milwaukee, WI
were used as the promoting agant in the  initiation studies.  Benzo[a]pyrene
(BaP) from Eastman Kodak, Rochester,  N.Y.  was used as the initiating agent in
the promotion studies and as the positive  control substance for the complete
carcinogenesis studies.  Acetone,  spectrophotometric  grade, from Fisher
Scientific Co., Pittsburgh, PA was used  as the  diluent for  preparation of
dosing dilutions and as the negative  control  substance.
   Animals and Husbandry.  C3H/HeJ mice  from  Jackson  Laboratories,  Bar Harbor,
Maine were used in these studies because of their low spontaneous skin tumor
incidence and our experience with  chemical induction  of  skin tumors in this
strain.  The mice were housed  5  per cage in stainless steel suspended cages
                             393

-------
located in Airo-Neg  Safety  Enclosures (Airo Clean Engineering Inc., Broomall,
PA).  The mice received  Zeigler  Block feed (Zeigler Brothers Inc., Gardners,
PA) and water from an automatic  watering system, both ad libitum.
   Experimental Design and  Procedures.   The mice were randomized into 18 groups
of 40 mice each such that the means and variances of the body weights were
statistically equivalent before  treatment began.  In the complete carcinogene-
sis studies, DP was  applied as either a 10% or 5% suspension in acetone.  DCM
extract of DP was applied as suspensions of 50%, 25%, 10% or 5%.  A positive
control group received 0.2% BaP,  and a negative control group received acetone
only.
   In the promotion  studies, a single initiating dose of 1.5% BaP was applied
followed after one week  by  repeated applications of one of the following:  a)
10% DP; b) 50% DCM extract; c) 25%  DCM extract; d)  acetone only; e) 0.0001% PMA
(positive control for initiation  and promotion studies).  An additional group
was untreated after  the  initiating  dose of BaP.
   In the initiation studies, a  single initiating dose of 10% DP, 50% DCM ex-
tract, acetone or PMA was followed  after one week by repeated applications of
0.0001% PMA.  The concentration  of  PMA was changed for the initiation and  pro-
motion studies after 8 months of  treatment to 0.01%.
   The test substances were applied with an Eppendorf automatic pipet set  to
deliver 25 microliters.  Animals  were treated 3 times per week in the complete
carcinogenesis and initiation studies and 5 times per week in the promotion
study.  All test substances were  applied to the skin of the back from which the
fur was clipped once each week.   All suspensions were prepared on a weight/
weight basis.
   Mice were observed frequently  for clinical signs and the appearance of
tumor-like growths.  Formal observation of each mouse for tumors was performed
monthly.  The studies were  designed to  last until the death of all animals.
Dosing was stopped   only if all  the mice in a group had malignant skin tumors
by gross observation.
   Necropsies were performed on dead mice as soon as possible after death.
Mice were sacrificed when found moribund.   All body cavities were examined and
suspect internal tumors  were fixed  in 10% neutral buffered formalin for histo-
pathologic examination.  The dorsal skin of all mice with or without neoplastic
skin lesions, was also fixed for  histopathologic examination.
                            394

-------
   Data Analysis.  The mortality  and  tumor  incidence  among  the various dosage
groups were compared using  the  Breslow8  and Mantel-Cox9  tests.

RESULTS
   Complete Carcinogenesis  Studies.   The results  of the  complete carcinogenesis
studies are presented in Table  1.  The number  of  animals alive, as shown  in
this and the other Tables,  was  accurate  as  of  the time of manuscript prepara-
tion.  The dosages were calculated based on measurements of the volume admin-
istered, the density of the solution  or  suspension, and  the concentration of
test substance.
   One tumor-bearing animal was observed at the highest  dosage of DCM extract
after 714 days of treatment.  The tumor  was in the treatment area and was
diagnosed as a squamous cell carcinoma.   Thirty eight tumot--bear ing animals
were observed in the positive control (BaP) group for an effective tumor
incidence of 100%, since the two  animals which did not have tumors died early
in the study.  No skin tumors were observed in the negative controls or in any
other dosage group.  Survival was not affected by treatment except for the
positive controls which died early with  skin tumors.
   Promotion Studies.  One  mouse  in the  50% DCM extract  group and one in the
25Z DCM extract group have  been diagnosed with squamous cell carcinomas with
pulmonary metastases (Table 2) .   A second animal  in the  25% DCM extract group
was alive with a grossly diagnosed papilloma..  In the positive controls, 19
tumor-bearing animals (11 papillomas, 8  carcinomas) were  observed, repre-
senting a statistically significant increase over the negative controls.
Survival was also significantly reduced  in  this group.  No tumors have been
seen in any other dosage group.
   Initiation Studies.  Three tumor-bearing mice  have been observed in each of
the groups initiated with DP or DCM extract (Table 3).  These include 2
papillomas and 1 carcinoma  ±n the DP  group, and 2 papillomas plus 1 fibrosar-
coma in the DCM extract group.  In the acetone-initiated group, 1 papilloma-
bearing mouse was observed.   In the PMA-initiated group, 1 carcinoma- and 1
papilloma-bearing mouse were recorded.   Statistical analysis of the tumor and
survival data revealed no significant differences.

DISCUSSION
   The results to date suggest  that DP and  DCM extract of DP have little, if
any, tumor-initiating or tumor-promoting activity under  the conditions of these
bioassays.  This conclusion is  based  on  the absence of a statistically signifi-
                             395

-------
cant increase in tumor incidence  (or  reduction in time to tumor) in any
treatment group.  The statistically significant tumor response in the 2
positive control groups clearly established  the susceptibility of the animals
to the induction of skin tumors.
   The above conclusion must be qualified because of the observation of a
carcinoma in the high dosage DCM  extract group of the complete carcinogenesis
study.  Although the presence of  a single tumor is clearly not statistically
significant, its importance must  be considered in the light of extensive his-
torical control data.  The C3H/HeJ strain has  been found to have an extremely
low spontaneous skin tumor incidence  in this laboratory.  Of 474 acetone-
treated controls, only a single mouse with a squamous cell carcinoma of the
eyelid has been observed.  No tumors  have been observed in the treatment area.
Thus, the tumor in the treatment  area of a DCM extract-treated mouse may have
toxicological importance.  Interpretation of this finding is further com-
plicated by the absence of a definite increase in tumor incidence in the 50%
DCM extract group of the promotion study in which the animals received  a larger
total dose than that in the complete  carcinogenesis  study,  following an initi-
ating dose of BaP.
   Although the final results and conclusions  of  these studies are not  yet
available, the results to date are not consistent with the  highly significant
tumor yield reported by Kotin et  al6.  The difference in results may be
attributed to differences in the  composition of the  diesel  emissions, which
itself is a function of engine speed, load and maintenance5.   The differences
in mouse strain and dosage (not clearly defined)  may also be important.
   The recent preliminary report7 of  positive  tumor-initiating activity by DCM
extracts from an Olds diesel engine is not definitive because of the "extremely
low" activity observed.  In addition, the differences between those studies and
ours include the source of test and control substances,  mouse strain, sex,
treatment regimen and specific response parameters.   More definitive conclusions
from both studies will be possible when the complete data are available.
                            396

-------
TABLE 1

RESULTS OF COMPLETE CARCINOGENESIS STUDIES
                       DP
                                      DCM EXTRACT
                                                           B(a)P
All mice received 3 days/week doses of the test agents
DP=»Diesel Particulates
DCM=Dichloromethane
BaP»Benzo[a]pyrene
Cp<0.001
ACETONE
Concentration (%)
Dosage (mg/day)
Tumor-Bearing
Animals
Number Alive
Time To First
Tumor (days)
Median Time To
Tumor (days)
Mean Survival
(days)
10
2.0

0
3

-

-
477

5
1.0

0
1

-

-
545

50
12.0

I
5

714

714
551

25
5.1

0
4

_

_
541

10
2.2

0
6

_


541

5
1.0

0
3

_

_
541

0.2
0.038

38C
0

175

252
311°

100
17.1

0
4

_


508

TABLE 2

RESULTS OF PROMOTION STUDIES

Concentration (%)
Dosage (mg/day)
Tumor-Bearing
An -final a
Number Alive
Time To First
Tumor (days)
Median Time To
Tumor (days)
Mean Survival
(days)
DP
10
2.0

0
7

-



523
DCM
EXTRACT
50
12.0

1
4

361

361

478
25
5.1

2
10

452

567

586
ACETONE UNTREATED
CONTROL CONTROL
100
17.1

0 0
8 6

-

-

552 562
POSITIVE
CONTROL (PMA)
(0.01)
1.5X10

19
0

354

465
h
452°
of test agent.
DP=Diesel Particulates
DCM-Dichloromethane

bp<0.01    cp<0.001
PMA=Phorbol Myriatate Acetate
                             397

-------
TABLE 3
RESULTS OF INITIATION STUDIES

Concentration (%)
Dosage (mg/day)
Tumor-Bearing Animals
Number Alive
Time To First Tumor
(days)
Median Time To Tumor
(days)
Mean Survival (days)
DP
10
2.0
3
0
319

528

516
DCM
EXTRACT
50
12.0
3
0
395

508

476
ACETONE
CONTROL
100
17.1
1
0
532

562

516
PMA
CONTROL
0.01
1.5X10
2
1
452

612

468
All mice received PMA 1.5  ug/day 3  times/week, after one initiating dose of
test agent.
DP-Diesel Particulates
DCM»Dichloromethane
PMA-Phorbol Myristate Acetate

ACKNOWLEDGMENT S

   The authors wish to recognize the excellent technical assistance of Daniel
Meckley in the performance of these  studies.   We  also thank Carrol Weil, Elton

Homan and Stephen Dempsey for their  helpful  suggestions in the preparation of

the manuscript and Florence Zaremba  for  excellent secretarial assistance.



REFERENCES

1.  Santodonata, J., Basu, D., and Howard, P.  (1978)  in Health Effects Asso-
    ciated with Diesel Exhaust Emissions, EPA-600/1-78-063.
2.  Lee, F.S.C., Pierson, W.R. and Ezike, J.  (1980)  in Polynuclear Aromatic
    Hydrocarbons:  The Fourth International  Symposium.   Ann Arbor Science,
    Ann Arbor, Michigan
3.  Choudhury, D.R. and Bush, B. (1980)  in Health Effects-of Diesel Engine
    Emissions:  Proceedings of an International Symposium,  Vol.1, EPA-600/
    9-80-057a.
4.  Clark, C.R. and Vigil, C.L. (1980).  Toxicol, Ap.pl. Pharmacol., 56, 110-115.
5.  Ohnishi, Y., Kachi, K., Sato, K., Tahara,  I., Takeyoshi, H.,  and Tokiwa, H.
    (1980).  Mutation Res., 77, 229-240.
6.  Kotin, P., Falk, H.L., and Thomas, M.  (1955). Arch. Ind. Hyg. Occup. Med.,
    11, 113-120.
7.  Slaga, T.J., Triplett, L.L., and Nesnow,  S. (1980)  in Health Effects of
    Diesel Engine Emissions:  Proceedings of  an International Symposium,
    Vol. 2, EPA-600/9-80-057b.
8.  Breslow, N.  (1970).  Biometrika 57, 579-594.
9.  Mantel, N.  (1959).  Cancer Chemotherapy Reports 50, 163-170.
                            398

-------
RESPIRATORY CARCMOGBriCITT Of DXESXL FUEL EMISSIONS
         RESULTS
Alan  M.  Shefner,  Bobby R. Collins, Lawrence Dooley, Arson Fiks, Jean L. Graf,
and Maurlina  N.  Preache
IIT Research  Institute, Life Sciences Research Division, 10 Wast 35th Street,
Chicago,  Illinois 60616
 IHTRCOOCTTCB
   An  experiment is in progress in which diesel engine emission particles  (DP),
 organic  solvent extracts of diesel particles (DE), extracts of roofing tar
 volatiles  (RT)  and coke oven  mains    (CO), and cigarette smoke condensate  (CS)
 are being  evaluated for their carcinogenic potential when administered by
 intratracheal  instillation to hamsters.   Appropriate control animal groups—
 untreated  colony controls (CC), solvent (SV), solvent plus ferric oxide (SF),
 benzo(a)pyrene  (BP) as a positive control, and gel-saline plus ferric oxide
 (OS)—are  included in the study.  Because of the number of hamsters being
 treated, the experiment was conducted in two replicates identical in design
 except for a gel-saline control included in the second replicate (Table 1).  At
 the time of this interim report, hamsters in Replicate 1 had been on test for 61
 weeks  and  those in Replicate 2 for 44 weeks. Histopathologlc findings are
 reported for a  subset of the animals from Replicate 1 that were sacrificed at 12
 month* of  age after being on test for approximately 9 months.

 MATERIALS  AID METHODS
 Test Materials
   Test materials were supplied through the courtesy of EPA and sample gen-
eration and collection has been previously described.
   Whole Particle M.eoel Hgfaanat Snapenaioaa.  The whole particle diesel exhaust
material was  received  as a dry powder that had been scraped from collection
filter substrates.   Microscopical examination of the powder as received showed
that the individual  submicrometer carbonaceous exhaust particles had
agglomerated  and aggregated during generation and after capture on the
collection filter to form large diameter hollow carbonaceous spheres (up to
70 Urn), and large thin flakes  (up to 150 um).  To produce particle suspensions
suitable for  intratracheal instillation, the whole particle diesel exhaust had
                              399

-------
TABLE 1.

Treatment Group
Diesel Particle
Diesel Particle
+ Fe203
Diesel Extract
+ Fe203
Coke Oven
Cigarette Smoke
+ F«2°3
Roofing Tar
+ Fe203
Benzo ( a ) pyr ene
Solvent
Solvent
Gelatin Saline
Control0
Colony Control
Total Replicate 1
Total Replicate 2
1 FOR BCH
Doseb
(mg/wk)
5.0
2.5
1.25
5.0
2.5
1.25
5.0
2.5
1.25
5.0
2.5
1.25
5.0
2.5
1.25
5.0
2.5
1.25
2.0

5.0




REPLICATE OF THE
Sex
H
H
M
M
M
M
H
M
M
H
M
H
H
H
M
H
H
H
M
M
M
H


Number of
Animals
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
125
50
100
1,225
1,275
CHRONIC t>i\IL»I"
12 Month Sacrifice
Number of Animals
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
25
10
20
245
255

Total
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
60
150
60
120
1,470
1,530
a The numbers shown are the number of animals  in  each group for each
  replicate of the study.
b Dose refers to the quantity of test material- or ^2°3  wnen only one
  component was present, and to the quantity of each component when test
  materials were admixed with Fe2°3'
c Gelatin Saline Control was included in only  the second replicate.
                             400

-------
to be reduced in size  so  that 90 percent by mass of the material was below 10 Urn
in size and suspended  in  physiological saline.
   Certain physical properties of the whole particle diesel exhaust such as its
low density, wide particle  size range, and high electrostatic charging tendency
prevented application  of  conventional dry grinding techniques.  The hydrophobic
nature of the powder and  the  requirement for sterility in the final suspension
made it necessary to achieve  particle size reduction and suspension
simultaneously.  Two different size  reduction-suspension preparation methods
were developed over the course of the instillation phases of the bioassay
experiments.  Before any  size reduction or particle suspension techniques could
be developed, however, a  nontoxic,  saline-miscible wetting agent for the diesel
exhaust particles had  to  be selected.  Reagent  grade propylene glycol was found
to wet the diesel exhaust particles  adequately  without dissolving (at room
temperature) the extractable  organics adsorbed  on the carbonaceous diesel soot
particles.  The reported  low  toxicity of propylene glycol,  its high viscosity,
and its miscibility with  saline rendered it a desirable material for use in
suspension preparation.   Gelatin was dissolved  in the saline to improve
suspension stability and  was  found  to be soluble in the propylene glycol as
well.
   For preliminary dose-response bioassay experiments, a simple ball milling
technique was selected to simultaneously reduce the diesel exhaust particles in
size and to suspend them  in the gelatin saline  suspending fluid.  The ball-
milling technique has  been  described in detail  elsewhere.   Briefly, the
technique involves wetting  the particles with propylena glycol before adding
them to a sterilized wide-mouth jar  containing  3 to 5 mm glass beads and the
gelatin-saline suspending fluid; the filled milling jar is placed on a roller
for a 10 day milling period.
   The ball-ailling method  was adequate for the small-scale, short term acute
toxicity experiments,  but was inadequate for the greater materials requirements
of the long term chronic  toxicity and carcinogenic potential evaluation
bioaasays.  Therefore, j  mechanical  mixer that  simultaneously generates both
mechanical shearing and ultrasonic  energies was investigated to determine if it
could perform both the particle grinding and the particle suspension operations
required.  The mixer—a Polytron® (Brinkmann Instruments) fitted with the
special purpose PT-35/4 probe generator—was tested and found to grind the
diesel particle aggregates  to the desired sizes and suspend them in the gelatin-
saline vehicle.  Simple probe generator designs such as the standard saw tooth
(ST) probes did not produce adequate mechanical shear energy to reduce the
                             401

-------
particle size sufficiently.   Propylene glycol was still required  as  a wetting
agent.  The concentration  of  propylene glycol used in these  suspensions was
slightly higher than  that  in  the ball-milled suspensions.  Gum arable was added
to the suspending fluid  (Table  2),  only to make the suspending fluid for the
diesel particles suspensions  and the diesel exhaust emulsions  identical in
composition.
      2.
COHPOSITiai OF WHOLE  FHRTXCLB DIESEL BXHAD5T SUSPENSION
                            Ball-Milled                  Polytron* Mixed
Dose Range          S,  3,  and  1  mg/0.2 ml           5, 2.5, and  1.25 mg/0.2 ml
Carrier Liquid      Saline with  0.5 percent w/v     Saline with  0.5 percent w/v
                    gelatin                          gelatin and  0.5 percent
                                                    w/v gum arable
Wetting Agent       Propylene  glycol—7 percent     Propylene glycol—10 per-
                    by volume                        cent by volume

Carrier Dust        Fa9°3—5'  3f  and 1 W3/0'2 "d.    Fe2°3—5' 2*5' and
                                                    1.25 mg/0.2  ml
   The method of preparing the Polytron* suspension was straightforward.  The
diesel particles were placed  in sterile pharmaceutical graduate cylinders and
the propylene glycol was  added, with hand-swirling of the cylinder, to wet the
particles.  The sterile saline, with gum arable and gelatin dissolved in it, was
then added.  The suspensions  were mixed for 2 min with the PT-35/4 at a
moderately high speed  (7  on a scale of 10) to complete the grinding-suspension
process.  For the  iron oxide  containing suspensions, the iron oxide was added
after the grinding-suspension was completed and was mixed into the suspensions
with the standard  PT-10ST generator since no additional particle size reduction
was desired at this point.
   Diesel RinauBt  Si U. act Suapenaiona (Baulaions).  Solvent extracts of whole
particle diesel exhaust were  submitted for the bioassay experiments as solutions
of the extracted materials in dichloromethane. Before any emulsion preparation
could be prepared,  the solvent had to be removed under a slow stream of pure,
dry nitrogen.  Upon removal of the solvent, the diesel exhaust extract was found
to be composed of  a light amber oily phase and a semi-solid, dark brown tarry
phase, neither of  which was appreciably soluble in, or miscible with, saline.
Thus, in order to  produce emulsions of the extract in the saline instillation
                              402

-------
carrier  fluid,  a wetting agent was required, which would be miscible with the
oily phase  and  saline and which would dissolve the tarry phase.  Propylene
glycol was  again found to fulfill the requirements of <. nontoxic wetting agent.
   Two different methods for diaael extract emulsion preparation were also
developed over  the course of the program.  The initial preparation method,
described in  detail elsewhere,3 was a very tedious hand-mixing method, and is
briefly  described here.   The extract solutions were pipetted into glass tissue
grinders and  the solvent was blown-off under nitrogen.  Propylene glycol was
added to the  solvent-free extract, and the mixture was heated to 50-65°C to aid
solvation of  the tarry phase.  SPM4-80*, a surface active agent, was then added
and mixing  with the glass pestle began.  Once the components had been thoroughly
emulsified, the gum-arabic-gelatin-saline fluid (also heated to 50°C) was slowly
poured in with  continuous mixing with the pestle.  Emulsification required 10-15
min of vigorous action of the pestle in the tissue grinder.  Iron oxide was
added once  emulsification was completed.
   A more mechanized method of emulsion preparation was obviously required for
the larger-scale chronic studies.  The Polytron® mixer, equipped with a standard
PT-10ST  probe generator was tested and found to produce an excellent suspension.
Emulsions were  prepared directly in the pharmaceutical graduate cylinders in
which the solvent removal step was conducted.  Warm propylene glycol was then
added and mixed with the Polytron* operating at moderate speed for 30-60 sec.
The preliminary emulsification step did not require the addition of a
surfactant.   With the mixer running, the warmed gum-arabic-gelatin-saline
mixture  was slowly added.  Once all the saline was added (1 min),  mixing at
moderately  high speed was continued for another minute.  Iron oxide was mixed in
at moderate speed for 30 sec after emulsification was completed.  The Polytron*
preparation technique simplified the preparation procedure, as well as the
composition of  the final product (Table 3).
   Other Extracts.  The extracts of coke oven  mains    and roofing tar
volatiles were  also received as solutions in dichloromethane.  Upon removal of
the solvent,  the coke oven extract was found to be a very viscous,  dark brown-
black, sticky,  tar-like material, while the roofing tar extract was a light
green, waxy material.  Both substances were highly odorous.
   The preparation of stable emulsions of these extracts in gelatin-saline
followed the  same general procedures described for the diesel engine exhaust
extract.  The order of addition of components and mixing steps  (with the glass
tissue grinder  emulsifying apparatus as well as the Polytron*}, were somewhat
different for these materials,  however.  For the coke oven   mains, after gently
                              403

-------
ABLE 3.
COHPOSITIOIl OP  DIESEL EXHAUST EXTRACT EMULSIONS
                    Tissue Grinder—Hand Qnulsified
                                      Polytron* Bnulsified
Dose Range
5, 3, and  1 ng/0.2  ml
5, 2.5, and  1.25 mg/0.2ml
Carrier Liquid
Saline with 0.5 percent w/v
gelatin and 0.25 percent w/v
gum arabic
Saline with 0.5 percent
w/v gelatin and 0.5
percent w/v gum arabic
Wetting Agent       Propylene glycol—10 percent by
                    volume  and sorbitan monooleate
                    (SPAN-80*) 0.1  percent by
                    volume
                                    Propylene glycol-10 per-
                                    cent by volume
Carrier Dust
Fe2°3  ~ 5'
Fe,03  - 5, 2.5, and
1.25 mg/0.2 ml
warming the extract plus  propylene glycol until the very viscous  tar essentially
melted, one-third of  the  required saline heated to 60°C was added to the mixing
container before emulsification was begun.  This initial mixing with the
Polytron® required 45-60  sec before, the remainder of the heated saline was
slowly added.  Final  emulsification required 2-3 min of mixing with the
Polytron* operating at  maximum speed.  For the roofing tar extract, after
addition of the propylene glycol followed by gentle heating,  the  initial
emulsification was conducted for 1-2 min with the Polytron* operating at top
speed.  The iron oxide  was then added and emulsified for 3 sec, and finally,
room temperature saline (plus the gelatin and gum arabic in solution) was added
with the Polytron* running at moderately high speed.  The final emulsification
required mixing at top  speed for 1-2 min.
   "The cigarette smoke condensate was received as a solution in acetone.  Removal
of the solvent left a viscous, dark brown, somewhat tarry residue. Emulsi-
fication was conducted  following steps identical to those used for the diesel
extract.
   Bento(a)pyrena-Ferric  Oxide Mixture.  The positive control material, a 1:1 by
mass mixture of benzo(a)pyrene and ferric oxide was prepared  by precipitating
the benzo(a)pyrene onto the iron oxide.  The iron oxide was suspended in 20
volumes of distilled  water and stirred constantly with a magnetic stirrer.  The
benzo(a)pyrene, dissolved in one volume of acetone, was added to  the iron oxide
suspension by slowly  pouring the solution into the vortex.  The benzo(a)pyrene
immediately precipitated  from the acetone solution upon impact in the water,
                              404

-------
thereby capturing  iron  oxida particles within the crystals formed.  The
benzo(a)pyrene-iron  oxide particles thus formed were filtered from the water
suspension and  dried under a nitrogen stream.
   Male Syrian  Golden  hamsters (LAK:LVG(SYR)) were obtained from Charles River
Breeding Laboratories,  Wilmington,  MA, at 6-8 weeks of age and were held in
quarantine until  one week before they were placed on test at 12-13 weeks of age.
The animals were  inspected for health status upon arrival and periodically
during quarantine.  For each replicate, 10-15 hamsters were randomly selected,
killed, and examined for pathogenic bacteria, mycoplasma, yeast, fungi,
endoparasites,  and  ectoparasites.   There were no problems with the health status
of the animals  during  quarantine other than the death of a small number of
animals, as is  consistent with shipping stress.
   The hamsters were maintained in  plastic, solid-bottom cages on a bedding of
hardwood chips  (Ab-sorb-dri®).  In  the first replicate, three hamsters were
housed per cage initially.  By Week 27 of the study,  however, the hamsters had
to be rehoused, two per cage,  separated by a stainless steel divider because of
fighting.  Hamsters in the second replicate were housed two per cage at the
start of the  study; dividers were added by Week 4.  Food (Wayne Blox, Locke
Erikson Labs, Melrose  Park,  111.) and tap water were  available for od libitum
consumption.
   During the week  preceding the first treatment, the hamsters were randomly
allocated to  treatment groups, as shown in Table 1.  Groups that received
different tast  materials were  housed in rooms separated from other test
groups.  Animals  in the solvent and solvent plus ferric oxide control groups
were together with  animals for which they were the control.  The 125 solvent
plus ferric oxide control animals in each replicate were distributed among the
appropriate five  test  rooms.   Colony control animals  were housed separately from
all other animals.  In  the first replicate the positive control group
(benzo(a)pyrene:ferric  oxide)  was housed separately,  whereas in the second
replicate, it was in a  room  with the gelatin saline controls.  Animal rooms were
maintained on a 12  hour light:12 hour dark cycle at a room temperature of 76"
±2°F, and humidity  was  controlled to avoid extreme excursions outside the range
of 40-70 percent  RH.
                             405

-------
Treataent
   Test and control materials were administered by the intratracheal
instillation method described by  Saffiotti and co-workers.*.S  Before each
intratracheal instillation  the  hamster was anesthetized with halothane dispensed
from an Airco Veterinary Anesthesia Machine,  Heedbrink, Model 960.6  When the
righting reflex was lost the animal was placed on a slanted board, its back on
the board and its mouth kept open by hanging the lower incisor teeth on a wire
hook, while the upper incisors  were retained by a tight rubber band  (Figure 1).
Figure  1.  Administration  of  diesel particles to hamsters by
intratracheal instillation.
A volume of  0.2 ml of  the  test material was delivered via a 0.25 ml tuberculin
syringe fitted with a  blunt  19 ga needle about 3 in. long and bant at a.  135°
angle at 45  irai from the  tip.   The tongue was pulled outward with forceps and the
rhythmic opening and closing  of the vocal cords observed. The blunt end  of the
needle was inserted into the  traches1 lumen past the open vocal cords and pushed
almost to the bottom of  the  trachea.  The suspension was gently injected and the
hamster was  retained on  the  board for approximately 1 min to make certain no
suspension was regurgitated.   Treatment, initiated at 12-13 weeks of aoe. was
performed once weekly  for  15  weeks.
                              406 '

-------
Observations
   The hamsters  were weighed weakly during the 15-week treatment period and
biweekly thereafter.  Physical examinations including palpation for tumors, were
performed at  the tine of weight determinations.  In addition, the hamsters were
observed daily for overt physical or behavioral signs of toxlcity or disease; a
second check  was performed each afternoon to allow identification of dead or
moribund hamsters.
   Extensive  necropsy examination was performed for all animals at the time of
death.  Moribund animals were sacrificed for immediate necropsy examination.  At
12 months of  age,  a subset of the test and control hamsters  (Table 1) were
randomly selected for interim sacrifice and necropsy.  Surviving animals will be
held up to  2  years of age, at which time they will be killed and necropsy
examinations  will be performed.  The necropsy procedure involves a thorough
examination of all external surfaces, body cavities, and orifices, with
approximately 35 tissues being examined and collected.  For animals killed at
the interim sacrifices,  the brain, heart, liver, spleen, kidneys, and testes
were weighed  at  the time of necropsy.  The eyes were fixed in gluteraldehyde and
the testes  in Bouin's solution for 24 hr and were then transferred to alcohol
for preservation.   The remaining tissues were fixed and stored in 10 percent
neutral buffered formalin.  Tissues were blocked in paraffin and 5-6 micron
sections were cut and stained with hematoxylin and eosin.
   All tissues collected were examined under a light microscope for all hamsters
killed for  the interim sacrifices and will similarly be examined for all control
hamsters and  hamsters in the high-dosage test groups whether death is
spontaneous or by moribund or terminal sacrifice.  For the remaining animals,
the respiratory  tract will be examined microscopically and the other tissues
will be saved for examination in the event treatment-related lesions are
identified  in the high-dosage test groups.  In addition to light microscopic
examination,  the lung and thoracic lymph nodes of hamsters sacrificed at 12
months were examined with a polarized light microscope.

Bata
   Data from each  replicate of the study were summarized separately.  Body
weight data were summarized by test or control group.  Single factor analysis of
variance tests  were  performed to determine whether there were significant
differences among  the  dosage levels of a specified test article and the solvent
or solvent plus ferric oxide hamsters housed within the same room.  If a.
significant F ratio  was obtained,  the test groups at different dosage levels
                             407

-------
ware individually compared  to the solvent control or appropriate subset of the
solvent plus ferric oxide control animals.  Organ weight data collected at the
interim sacrifices were  similarily analyzed*  Survival data were summarised as
the percent animals surviving at the end of the various test weeks.  Clinical
observations are continuously summarized based upon calculations and statistical
comparisons of the median time for each group between the first observance of a
specific sign and the death of the animal.  Necropsy observations were tabulated
as incidence of specific lesions within groups.  Histopathologic data were
summarized as the incidence and average severity of lesions for the different
test or control groups.
Teat. Material Characterisation
   The Polytron* prepared diesel  exhaust particle suspensions proved to be more
stable than the ball-milled suspensions  and contained smaller sized particles
(Table 4).  The Polytron* preparation method had the additional advantage of
eliminating of the glass fragments  fractured from the milling beads and vessels.
   The final suspensions were evaluated  microscopically for particle morphology
(Figures 2 and 3) and particle  size.   Samples were assayed by filtering measured
aliquots (diesel particles only)  and  by  ashing measured aliquots (suspension
containing diesel particles and iron  oxide).
TABLE 4.
PARTICLE SIZE ABALTSES Of •BOLE PARTICLE DIESEL EXHAO5T HID

Diameter, Urn
(Linear Dimension)
0.0
1.0
3.0
5.5
8.0
10.5
13.0
15.5
18.0
20.5
Cumulative Mass
As-Received
100.0
99.2
97.2
82.7
63.0
43.6
26.0
14.4
7.3
2.5
Percent Greater
Ball-Milled
100.0
99.6
95.7
76.1
48.1
25.3
12.4
4.7
1.5
0.0
Than Stated Size
Polytron* Milled
100.0
98.5
80.3
24.0
5.8
0.9
0.9
0.9
0.0
0.0
                              408

-------
                                               11 f 'f'^t.' •'•
                                               y-i^**£$.s
                                                      A-»—•
                                                    . *a
                                                    m.
.&f-
:^S^J>
       J*r->'&

       w&

'1^'&^&&'&*i ^-£%&£%, *>'*l
L, F^"'^^T'^t  ^tt'y.**?$£&• r'f *-"
 Figure 2.   Optical micrograph of whole particle diesel exhaust suspension after

 dispersion with the Polytron*; 502X.  The largest particle visible is

 approximately 5 I'm in  diameter.
Figure 3.  Electron micrograph of  whole particle diesel exhaust suspension after

dispersion with the Polytron*;  5000X.
                            409

-------
   The  Polytron* prepared diesal extract emulsions  also proved to be more stable
than the hand-mixed  emulsions (Figure 4).  The primary  improvement of the
emulsions introduced by the Polytron* preparation method was  the elimination of
the surfactant.
   No adequate assay method for determining the amount  of diesel extract present
in the  emulsions could  be developed because of the  complex nature of the
extracts.  The prepared emulsions were viewed microscopically to determine if
droplet sizes of both the tarry and oily phases were comparable (Figure  5).

Clinical Signs
   Almost all of the clinical signs exhibited by first  replicate hamsters during
the observation  period  preceding to the scheduled sacrifice were the direct
result  of fighting among cage mates.  One hundred percent of  the colony  control
animals developed lumbrosacral skin lesions secondary to fighting.   The  rest of
the test groups  exhibited an incidence rate that ranged  94-100 percent.   The
seriousness of the lesions varied widely among the  affected animals. Crusts,
necrotic tissue,  and one mass that was subsequently identified as an abscess
were observed.   All  of  the above lesions resolved after  the cage density was
reduced to two hamsters per cage and a stainless steel partition was placed in
the center of each cage, physically isolating each  hamster from his  cage mate.
   Keratitis was another clinical problem that was  observed secondary to
fighting. The lesions were reported in the colony controls  (1/20), benzo-
(a)pyrene (1/10), diesel particle (6/30), diesel particle  plus ferric oxide
(1/30), diesel extract  plus ferric oxide (3/30), cigarette smoke condensate
(3/30), and solvent  (1/10) treatment groups.  No cases of  keratitis  were
reported for the coke oven extract plus ferric oxide group  (0/30).   The  physical
isolation of each hamster effectively eliminated this problem from the study.
   Survival data to  date have indicated no treatment-related  differences.  In
general, the percentage of hamsters surviving at any given  time was  greater in
the second replicate than in the first.  This is undoubtedly  due to  a reduction
of deaths from causes related to fight wounds in the second replicate by earlier
separation of the hamsters.

Body Weights
   Hamsters treated  with the highest dose (5.0 mg/wk) of diesel particles had
significantly lower  body weights than solvent control hamsters in both parts of
the study (Tables 5  and 6).  A significant effect was first observed in  Test
Week 6 of the first  replicate and was sustained through  Test  Week 15, the last
                              410

-------
 Figure 4.  Diesel Exhaust Extract:  Left is  the extract itself after solvent
 removal; Right is the saline based emulsion  prepared at a concentration of
 2.5 mg/0.2 ml.
Figure 5.  Optical  micrograph of diesel exhaust extract emulsion.   (Dark field
illumination;  520X.)   The  droplets  are  all components  of the  diesel exhaust
extract.  The  largest  droplet is approximately 10 Mm in diameter.
                             411

-------
TABLE 5.

PATTKRH OF SIGHIFICftMT DIFFERENCES* FOR WEATED VS.  CONTROL OOMPAUSOBS Of BOOT
WZICfflTS BT TEST WEEK FOR EEPLICAIE 1
Test Groups'1
Test
Week
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
47
49
51
53
55
57
59
61
DP
_
-
F
-
-
-
H
H
H
-
-
H
H
H
H
H
.
-
-
-
-
-
-
-
_
-
_
.
-
-
-
-
-
-
-
-
-
-
-
DP+
.
-
-
-
_
H
R
H
H
B
H
H
H
H
H
H
-
-
-
.
_
-
_
_
_
_
_
_
_
-
-
-
-
-
-
-
-
-
-
DE+
Fe203
—
-
-
-
-
HM
HM
-
-
-
-
-
-
-
.
-
-
-
-
-
-
.
_
_
_
_
_
_
_
_
-
-
-
_
-
.
_
_
-
CS+
_
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
_
-
_
-
_
_
.
_
_
_
«.
_
-
.
-
_
-
-
_
_
-
CO+
—
-
-
-
-
-
-
F
F
F
F
F
-
-
-
-
-
-
_
-
_
_
_
_
_
_
_
_
•
H
HM
HM
HM
HM
HM
HM
HM
HM
-
RT+

_
.
.
_
_
.
.
-
-
_
_
-.
.
_
_
.
.
-
.
.
-
-
-
_
.
-
-
.
-
-
-
-
-
-
-
-
-
•
  A significance  level of p 0.05 was used.  F indicates that a  significant F
ratio was obtained  in  the analysis of variance but was due to difference among
the High, Middle, and  Low Dosages groups.  H, M, or L indicate  that  the High,
Middle, or Low dosage  group was significantly lower in body weight than the
appropriate  control group.
b DP » Diesel Emission Particles, DE <• Diesel Emission Particle  Extract,
  CS =- Cigarette  Smoke Condensate, CO * Coke Oven Emission Extract,  and
  RT - Roofing Tar  Extract.
                               412

-------
TABLE 6.
PATTERN
BOOT WEI
Test
Week
0
1
2

3

4
5
6
7
8
9
10
11
12
13
14
15
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
OF SIQUFICAHT DIFFERENCES* TOR TREATED VS. CCBTROL CQMPARISOHS OP
GHTS BY. TEST WEEK TOR REPLICATE 2

DP+
DP Fe2°3
_ —
H
H

H

HHL
H
H
H
H c
H c
H
H
H
H
H
H
H
- -
-
H
a
H
H
H
H
H
-
.
-
-
-
Test Groupsb
DE+ CS+ CO+
F«2°3 Fe2°3 F«2°3
"





HH
HML
HM
HH
= C c
c c c
HH
HML
HHL F
HHL
HHL
HHL
HHL
HHL
HHL
HHL
HHL
— — _
— — _
F
-
«> — _
_
_
_
_
-

RT+
Fe2°3


~
p

F
F
F
F
F
c
c
F
F
F
H
M
H
_
_
_
-

_
—
_
-
_
_
_
_
-
-
a A significance level of p 0.05 was used.  F indicates that a significant F
  ratio was  obtained in the analysis of variance but was due to differences
  among the  High,  Middle, and Low dosage groups.  H, M, or L indicate that the
  High, Middle,  or Low dosage group was significantly lower in body weight than
  the appropriate  control group.
b DP a Diesel  Qnission Particles, DE » Diesel Emission Particle Extract,
  CS =» Cigarette Smoke Condensate, CO - Coke Oven Emission Extract, and
  RT • Roofing Tar Extract.
c Body weights were  not determined at the scheduled times during these weeks.
                             413

-------
week of treatment.   In the second replicate,  the  effects of the highest dose of
diesal particles  on  body weight were apparent earlier,  sustained longer,  and
involved  larger differences.  In Replicate 2, with  the  exception of Test Weeks
18 and 20, body weights of the hamsters treated with  5.0 mg/wk diesel particles
were significantly lower than control weights for all weeks between Test  Weeks 1
and 34.   The maximum effect in Replicate 2 was seen during the latter part of
treatment and early  posttreataent weeks when  differences from the control
weights averaged  15-20 g.
   Body weights of hamsters treated with the highest  dose of diesel particles
plus ferric oxide were lower than those of their  solvent plus ferric oxide
control during the treatment period of Replicate  1  (Table 5).  Kote that  the
duration  of effects  closely approximates that seen in the high dosage diesel
particles group of Replicate 1.  In Replicate 2,  body weights of hamsters
treated with the  high dose of diesel particles plus ferric oxide were somewhat
lower than those  of  their controls during the latter  part of treatment and
following the treatment period.  This difference, however,  which ranged 8-12 g
between Weeks 12  and 20,  failed to achieve statistical  significance.
   During the latter weeks of treatment in the first  replicate and up to  Test
Week 21,  body weight means for hamsters treated with  diesel extract or their
solvent plus ferric  oxide control we're distributed in a dose-related fashion:
the control mean  was the greatest, and the low, middle,  and high dosage means
were consecutively lower.   However, statistically significant differences from
control means were limited to Test Weeks 5 and 6  and  involved only the high and
middle dosage groups (Table 5).  In the second replicate,  body weight means for
all groups of hamsters treated with diesel extract were significantly lower than
control values during the latter part of the treatment  period and up to Test
Week 24.  Effects in the high and middle dosage groups  were observed as early as
Test Week 4 and in the low dosage group in Test Week  5.   During the period when
significant effects  were obtained in all three dosage groups, the means for
hamsters  treated  with the high dose Were slightly (4-5  g)  but consistently lower
than those of the low and middle dosage groups.   The  means for the latter two
groups, however,  were often numerically (within 1 g)  as well as statistically,
equivalent.
   To date, no significant reduction in body weight gains has been observed for
hamsters  treated  with cigarette smoke condensate.  During the first replicate,
hamsters  treated  with the high dose of this test  material appeared to show
reduced body weight  gains.  This was not seen in  the  second replicate, where the
body weight curve for the high dosage group was virtually superimposable  on the
control curve for much of the period in question.
                              414

-------
   For hamsters  treated with coke oven emission in the first replicate, no
significant differences between body weight means for treated and control groups
were observed before  Test Week 43.  Beginning that week for the high dosage
group and in Test Week 45 for the middle dosage group, body weights for these
groups were significantly lower than those of the solvent plus ferric oxide
controls for all determinations through Test Week 59.  At the time of the last
determination  (Test Week 61), body weights of the high and middle dosage groups
still averaged approximately 15 g less than those of the control group, but this
difference is not statistically significant.  There have been no indications of
reductions in body  weight gains by coke oven emissions in the second replicate.
   These differences  in body weight between the high and middle dosage groups
and their controls  cannot be attributed to body weight loss in these test
groups.  Bather  the mean body weight of their solvent plus ferric oxide controls
showed a large increase beginning at about Week 35.  This increase in mean body
weight coincided with an increase in deaths of animals in the control group.
The removal from the  experiment of lower weight animals that had died could have
resulted in higher  calculated mean body weights of the surviving control
animals.
   Through Test  Week  61, there were no indications in Replicate 1 of treatment
related effects  on  body weights of hamsters treated with roofing tar extract.
In the second replicate, body weights of hamsters in the low dosage group tended
to be greatest and  those in  the middle dosage group lowest; the high dose and
solvent plus ferric oxide control values intermediate between the two.  This
distribution of  body  weight  means resulted in significant F values for the
analysis of variance  during  Test Weeks 2-15.  Through Week 12, however, the
significant values  were attributable to differences between the low and middle
dosage groups.   Thereafter,  in Weeks 13-15, the body weight means of the middle
dosage group were lower than those of both the low dosage group and the control
group.  The lack of a meaningful dose-response relationship in these results and
the absence of similar findings in the first replicate suggest that these
results cannot be attributed to the treatment with roofing tar extract.
   A gel-saline  plus  ferric  oxide control was included in the second replicate
as the appropriate  control for the benzo(a)pyrene positive control.  In this
replicate mean body weights  of the benzo(a)pyrene group are lower than their
respective control  during Weeks 20-40, though body weights of animals from both
groups are virtually  identical during the treatment period.
                               415

-------
Groas necropsy Cbaarvationa for Ani»als Sacrificed  at 12 Months of age
   (First Replicate)
   The most  common gross findings among the sacrificed animals were related to
the respiratory system.   Mottled black lungs were frequently observed in
hamsters treated with diesel particles or diesel particles  plus ferric oxide.
This occurred  slightly more frequently in the high  and middle dosage groups in
which 60% of the hamsters examined showed this effect.   Mottled blade or mottled
brown lungs were also characteristic of hamsters treated with the  high (6/10)
and middle  (6/9) doses of coke oven extract, but this  was not observed at the
low dosage for coke  oven,  and only rarely for hamsters  at any dose level for
diesel extract,  cigarette smoke condensate, or roofing tar  extract.   Of the 25
solvent plus ferric  oxide control animals examined  at  the interim  sacrifice, 3
had black or brown mottled lungs.  Gray mottled lungs  were  observed  in a few
hamsters treated with diesel particles plus ferric  oxide, coke  oven  extract, and
cigarette smoke condensate,  as well as in two of the solvent plus  ferric oxide
control animals.  Red mottled lungs was even more common but the distribution of
this finding,  including  its presence in 25% of the  colony control  hamsters,
suggested that it was not a treatment related effect.
   Black material was present in the trachea of some hamsters at all dosage
levels for diesel particles and diesel particles plus  ferric oxide.   All diesel
particle groups and  the  middle dosage group for diesel  particles plus ferric
oxide included hamsters  with black material present at  necropsy in the
respiratory  lymph nodes  and this was observed occasionally  but  not consistently
for all other  test materials except roofing tar extract.
   No grossly  observable masses were detected in any of the  sacrificed animals,
however, two of ten  hamsters in the diesel particle plus  ferric oxide low dosage
group had a nodule in the lung.  A diverse variety  of  other  gross  lesions was
sporadically observed in the liver, kidney, adrenal gland,  spleen, intestines,
and other organs.  Considering the low frequency and/or distribution of these
across treatment groups,  none of these could be attributed  to effects of  the
test articles.
   Group mean  organ  weights of brain, heart, liver, spleen,  kidneys,  and testes
of test hamsters from the first replicate interim sacrifice  showed no sig-
nificant effects of  test materials as compared with their solvent  controls.
   In summary,  the gross lesions observed at the interim sacrifice for the first
replicate indicate no effects of the test articles  other than those  which are
consistent with the  presence of residual test materials  in  the  respiratory and
lymphoid systems.
                              416

-------
Long Pathology
   Lung lesions  of  significance among colony control animals consisted of
adenomatous hyperplasia  of  the respiratory epithelium lining terminal bron-
chioles and/or respiratory  bronchioles.  The hyperplastic lesion occurred
independent of or in close  association with small aggregates of mononuclear
macrophages located principally in alveolar spaces surrounding the terminal
airway structures  (terminal bronchioles,  respiratory bronchioles, alveolar
ducts).  The  macrophages had a foamy cytoplasm or contained a brown, granular
pigment.  The above lesions affected approximately one half of the colony
control animals  and they were primarily focal and minimal in severity.  Similar
lesions were  also seen among animals in the solvent control groups of the
respective test  materials;  however, the incidence and relative severity were
slightly higher  in  most  instances.  With respect to these animals, small amounts
of test materials were also present in alveolar macrophages (except for the DP
Solvent group) and  associated with a minimal to mild, focal subacute alveolitis
in some animals  at  the site of the macrophage response.  The inflammatory
response (alveolitis)  appeared in lung sections from solvent control animals
that contained particles of test material.
   Among the  various test groups, lesions of adenomatous hyperplasia associated
with an alveolar macrophage response, phagocytosis of the test material nad
subacute alveolitis were more prevalent and severe among animals in DP and
DP:Fe203 groups  with a dose-response relationship.  These lesions were most
severe in the DP:Fe203 group.  Lesions of intermediate severity occurred among
animals of the COtFe203  group.  Similar lesions of lesser severity occurred
among animals in the DE:Fe2O3, CS:Fe203,  RT:Fe2
-------
   Focal lesions of  chronic pleurltis, accompanied by  particles of test material
at the reaction site,  occurred among a few test animals of  the  various treatment
groups except the CS:Fe203 group.   Similar lesions were observed among a few
animals in some of the solvent control groups when particles  of test material
were also present within  the lung.   These lesions were absent among the colony
control animals.  However,  focal lesions of subacute pleuritis  were present
among a few animals  in the colony control and some of  the solvent control and
test groups.  These  lesions were focal and of minimal  severity.  In view of the
above, this lesion appeared to be naturally occurring  and unrelated to the test
materials.
   Other pulmonary lesions of lesser significance among all groups included
congestion, recent hemorrhage and peribronchial, peribronchiolar or perivascular
lymphoid infiltrates.   These lesions did not appear to be compound related and
were ascribed to the method of sacrifice and spontaneous disease.
   Neoplasms of the  respiratory tract, classified as adenomas,  were observed
among single high-dose animals of the DP and DE:Fe203  groups.   Advanced lesions
of adenomatous hyperplasia graded as moderate to marked (Grade  III to IV) were
usually multifocal and rather extensive in their development.   However,  they
were regarded as proliferative lesions in response to  chronic irritation rather
than neoplastic.  The  exact pathogenesis of this lesion was not established but,
it could be characterized as an extension of the terminal bronchiolar epithelium
into the region of the respiratory bronchioles, alveolar ducts,  and alveoli.
However, a metaplastic origin of this lesion could not be excluded.  The
epithelial cells were  cuboidal with cilia on their apical border and they
appeared to undergo  both  hypertrophy and hyperplasia.  These  cells also assumed
an adenomatous to papillary pattern depending upon the relative severity of the
lesion.  The presence  of  moderate to large aggregates of particle-laden
macrophages was a relatively constant feature of the lesion.  Plugs of mucinous
secretory product were also present at the site of some of  these lesions.
   Treatment-related lesions were also observed in tissues  other than the lung
and consisted of the following.

Trachea and larynx
   Particles of test material were present in the submucosa of  the trachea
and/or larynx of some  animals of all treatment groups with  a  dose  response
relationship with regard  to the severity of the lesion.  There  was no tissue
reaction to their presence.  This lesion was also seen in some  solvent control
animals where particles of test material also appeared in the lung.
                              418

-------
thoracic  Lyaph Soda
   Particles  of test material, within macrophages, were observed in the thoracic
lymph node  of animals in all teat groups and was dose-related in severity.

»ontrejt»ent-«alat«d Leaiona
   A number of spontaneous and age-related non-neoplastic lesions were observed
among animals in the colony control, solvent control, and all test groups.  The
lesions most  commonly observed consisted of thyroid cysts, vacuolation of the
pituitary,  subcapsular cortical cell hyperplasia of the adrenal gland,
mononuclear cell infiltrates in the liver with nuclear hypertrophy and/or
nuclear inclusions, and degenerative changes in the kidney (nephropathy)
accompanied by mineralized foci.
   Tumors in  various organ sites were also found upon microscopic examination of
tissue sections of interim sacrifice hamsters from the first replicate.  These
included  adenomas of the thyroid, kidney and adrenal, melanomas of the eye, and
spleen hemangiomas.  These tumors did not appear to be present as a response to
treatment but rather aa background common to hamsters of this age range.

Conclusions
   Only  limited conclusions can be drawn prior to the completion  of this
lifetime  toxicity/carcinogenicity evaluation of diesel engine emission particles
and the  other  materials on test.  Treatment effects on body weight gain were
most pronounced in the case of hamsters treated with diesel exhaust particles
and with  diesel exhaust extract.  Reductions in body weight gain were generally
dose related and diesel particle treated hamsters in the first replicate gained
weight rapidly following cessation of treatment.  Effects of treatment in the
second replicate were longer lasting and greater in degree.  Admixture of ferric
oxide with  the diesel exhaust particles did not increase the effect on body
weight gain.   Treatment with diesel exhaust particle extract produced
substantial and prolonged decreased weight gain in the second replicate and
lesser but  significant effects in the first replicate.  Decreases in body weight
gain in other  test groups were not as severe and recovery was rapid following
the end of  treatment.   Thus it appears that toxicity as measured by decreased
body weight gains was  test material and dose-related and reversible in nature in
that recovery  generally occurred rapidly once treatment ended.
   At the time of the  12-month interim sacrifice a six month period had elapsed
since the last intratracheal instillation of test material.  Even after this
considerable period of time phagocytosis of diesel particles by alveolar
                              419

-------
macrophages was not  complete.   Some of the hamsters in  these  test groups  still
showed the presence  of  free  particles in extracellular  spaces.
   All particles  tested induced marked responses of alveolar  macrophages  and
extensive phagocytosis.  Particle laden macrophages were  found  within  the lumen
of the terminal airway  structures and in the thoracic lymph nodes of animals
from all particle  test  groups.   The severity of the response  was  both  dose and
test material related.   Adenomatoua hyperplasia was most  severe in diesel
particle and diesel  particle plus ferric oxide test groups, intermediate  in
response in the coke oven group,  and least severe in the  diesel particle
extract, cigarette smoke condensate and roofing tar groups.   Two  lung  adenomas
were found on microscopic examination; one in a high dose OP  hamster and  the
other in a high dose DE animal.
   Thus good correlation was observed between the severity of test material
effects on body weight  gain  and the response of lung tissue to  the
administration of  specific test substances.  It cannot  be determined at this
time whether the hyperplasia, metaplasia and other pathologic findings induced
by test material exposure are recoverable in nature or  whether  they are
indicative of future deleterious  processes.

Acknowledgements
   Dr. Donovan E.  Gordon performed independent diagnosis  and  review of
histopathologic findings on  first replicate interim sacrifice hamsters.   Ms.
Maria Hawryluk provided editorial review and aided in preparation of this
manuscript.  We wish to thank Drs. Donald Gardner and Judith  A. Graham, Health
Effects Research Laboratory, EPA, North Carolina, for their advice  and
assistance during  the course of these studies.  This work was supported by EPA
Grant No. R806929-01 and EPA Cooperative Agreement No.  CR806929-02.
1.  Huisingh, J.L., et.  al.,  (1980)  in Health Effects of Diesel Engine
    Emissions, Proceedings  ,  EPA-600/9-80-057b, November 1980, pp.  788-800.
2.  Windholz, M.,  Budavari,  S,,  Stroumtsos, L.Y., and Fertig, M.N., Eds.  (1976)
    The Merck Index.   9th Edition,  Merck 5 Co., Inc., Rahway, N.J., p.  1017.
3.  Graf, J.L. (1980)  in Health  Effects of Diesel Engine Emissions, Proceedings,
    EPA-600/9-80-057a, November  1980, pp 82-92.
4.  Stafiotti, n., Cefis, F.,  and Kolb, L.H. (1968) Cancer Res.,  28,  104-124.
5.  Stafiotti, 0.  (1969) Prog. Exp.  Tumor Res., II, 302-333  (Karger,  Basel).
6.  Smith, D»M., Goddard, K.M.,  Wilson, R.B., and Newberne,  P.M.  (1973) Lab
    Animal Science, 23,  869-871.
                             420

-------
              CARCINOGEN1CITY OF EXTRACTS OF DIESEL AND RELATED
            ENVIRONMENTAL EMISSIONS UPON LUNG TUMOR INDUCTION  IN
                               STRAIN  'A'  MICE
                  R. 0. Laurie, W. B. Peirano, W. Crocker,
                  F-  Truman,  J.  K.  Mattox  and  W.  E.  Pepelko
                     Health Effects Research Laboratory
                    U.S. Environmental Protection Agency
                              Cincinnati, Ohio
 INTRODUCTION

     The predicted increase in the use of diesel engines has resulted in a
 regulatory need for data assessing the relative carcinogenicity of diesel
 exhaust. Since cigarette  smoke,  roofing  tar  and coke  oven  emissions  have
 been shown  to  be carcinogenic,  a matrix of  experiments was designed to
 compare  the relative  potency of  these  pollutants with  diesel  exhaust
 particulate. The present study is one of several,  including skin painting
 of  Sencar  mice,  intratracheal  instillation  in  hamsters  and jn_  vitro
 testing, designed to provide  such a  comparison.

 METHODS

 Compounds

     Nissan diesel  particulate matter was collected  with a high volume
 sampler using Pallflex  T60A20 (teflon coated) filters. The samples  were
 collected  from  a large mixing chamber  containing  exhaust diluted  with
 about 9 parts clean air to produce a particulate matter concentration  of 12
mg/m3. Exhaust was produced with a 6 cylinder, 90  cubic inch displacement
 Nissan  diesel  engine run  on  the Federal  Short Cycle. For  details see
 Hinners et al  (1979). The Oldsmobile sample differed from the Nissan sample
 in that it  was generated  at a steady  state  (40 mph). Both samples  were
Soxhlet extracted with  dichloromethane.  Cigarette smoke condensate  (CSC)
was  supplied  by  the Chemical  Repository and  Tobacco Smoke  Chemistry
Division of the  Tobacco   and  Health Research  Institute,  University of
                                    421

-------
Kentucky. The CSC was produced from Kentucky Reference 2RI Cigarettes. For
details  of collection  and  generation  see  Patel  (1977).  The coke  oven
emissions were collected using massive  volume  samplers  from  the  top  of  a
coke oven battery at  Republic Steel  in  Godsden,  Alabama.  The roofing tar
emissions  were  collected using  a baghouse  filter fitted  with  special
nonreacting filter  bags.  Details of  collection procedures for both  coke
oven  and roofing tar  have  been  described  in  detail  by  Huisingh et al
(1979).  Most  samples  were dissolved  in DMSO: for  the Nissan particulate
matter 5% of  the DMSO  solution was EL620.

Animals  Used  and Experimental Design

     Strain  A/Jax  mice approximately  8  weeks  of  age,  were  randomly
assigned  to  9 groups:  1) uninjected controls,  2)  vehicle  controls, 3)
positive  controls,  4)  Nissan  generated  particulate  matter,  5)  Nissan
particulate extract, 6) Oldsmobile particulate extract,  7)  cigarette smoke
condensate, 8) coke oven emissions, and 9) roofing tar.  The doses  for  each
group are listed in  Table 1.  Because of limited  availability of Oldsmobile
extract  only  males  were injected. Due to high mortality rates among  mice
injected with Nissan  particulate the dose level was halved in the  second
experiment. The  experiment  was  carried out  in 2 parts  because of  the
limited  availability  of animals  and  manpower.  The mice were injected 3
times weekly for 8 weeks with the test substances. The injection volume was
50 micro!iters, delivered by Hamilton syringes  fitted with 26 G  needles.
The urethane positive controls received only one injection as  the  start of
the experiment.

Collection and Analysis of Data

     The  mice were  sacrificed at  9  months  of  age with  an  overdose of
nembutal. The  lungs  were removed  and  placed in buffered formalin. After 2
weeks the lobes were detached from the bronchi  and  the number of adenomas
visible on the surface  counted.  Questionable  areas were examined micro-
scopically. Analysis of variance compared the number of tumors per mouse
among groups;  Chi  Square  analysis compared  the frequency of mice with
tumors.
RESULTS

     From the  data  presented  in Table 1  it  is  clear that a significant
increase in number  of  lung  adenomas  per mouse  and  percent  of  mice with
tumors occurred in the positive controls  (urethane  injected). In experi-
ment 1 a significant increase in number of tumors per mouse was noted in
males injected with Nissan diesel  extract  compared with controls or those
injected with  Oldsmobile extract.  In  experiment  2, a significant increase
in  lung  tumor  rates  was  detected   in females  injected with  coke oven
emissions. After  combining the data from both experiments,  no statis-
tically  significant differences  were  noted,   except  for  the  positive
controls.
                                    422

-------
DISCUSSION

     There  were  no  consistent,  statistically  significant  differences
between the exposed and control  groups  with respect to either number of
adenomas  per  mouse or  number  of mice  with adenomas. The  response to
urethane was equivalent to  the  numbers  predicted from numerous previous
studies indicating  that the  animals were  sensitive  to  tumor induction
(Shimkin and Stoner, 1975). It was concluded that either the carcinogens
present in  the  test substances  were very  weak  or  that  an insufficient
concentration reached the lungs to produce  a positive result.
REFERENCES

Hinners, R.G., J.K. Burkart,  M. Malanchuk and W.D. Wagner. Animal exposure
     facility for diesel exhaust studies. Health Effects of  Diesel Engine
     Emissions:  Proceedings  of an International Symposium,  Vol. 2: 681-
     697, 1979.

Huisingh, J.L.,  R.L. Bradow, R.H.  Jungers,  B.D.  Harris, R.B. Zweidinger,
     K.M. Gushing,  B.E. Gill and  R.E. Albert. Mutagenic and carcinogenic
     potency  of  extracts  of diesel and related  environmental emissions:
     Study design,  sample generation, collection and preparation. Health
     Effects  of  Diesel  Engine  Emissions: Proceedings of an  International
     Symposium.  Vol. 2: 788-800,  1979.

Patel,  A.R.    Preparation  and monitoring  of  cigarette smoke condensate
     samples. In Report No.  3,  Toward less hazardous cigarettes.  The third
     set of experimental cigarettes, G.B. Gori,  Ed., DHEW Publication No.
     (NIH) 77-1280.

Shimkin,  N.B. and  G.D. Stoner.    Lung  tumors   in  mice:  Application to
     carcinogenesis bioassay.  Adv.  Cancer  Res.  21:1-58, 1975.
                                   423

-------
Table L - Strain A Mouse  Data  for  Induction of
          Adenoma* by Environmental Mixtures
Percent of Av. Number

Group Sex
Uninjected M
Controls F
DMSO * M
5Z EL520 F
Ore thane M
F
Nissan Diesel M
Particulata F
Nissan Diesel M
Extract F
Olds. Diesel M
Extract F
Cigarette Smoke M
Condensate F
Coke Oven M
F
Roofing M
Tar F

Injected M
Controls F
DMSO * M
51 EL620 F
Ure thane M
F
Nissan Diesel M
Particulata F
Nissan Diesel M
Extract F
Olds. Diesel M
Extract F
Cigarette Smoke M
Condensate F
Coke Oven M
F
Roofing M
Tar F

Mice with
Tumors
40
47
40
47
100A
100A
40
20
73
55
37
-
59
54
54
31
41
44

13
26
44
26
95A
100A
27
32
33
30
33

30
21
31
65
44
22

^ Significantly different from
^ Significantly different from
(p < 0.05).

of Tumors
per Mouse
0.6 + 0.2
0.6 + 0.2
0.9 + 0.5
0.7 7 0.2
22.5 + 1.9A
21.8 2 1-5A
0.4 * 0.1
0.5 + 0.1
1.4 + 0.3B
1.0 T 0.3
0.4 » 0.1
-
0.8 + 0.3
l.l J 0.2
1.2 + 0.3
0.5 + 0.2
0.7 + 0.2
. 0.7 + 0.3
Experiment I
0.2 + 0.1
0.4 * 0.2
0.5 + 0.2
0.3 + 0.1
7.3 + 0.7A
11.3 + 0^
0.3 + 0.1
0.3 + 0.1
0.4 * 0.1
0.7 + 0.1
0.4 * 0.1
-
0.4 + 0.1
0.2 * 0.1
0.4 * 0.1
0.9 «; 0.2A
0.7 + 0.2
0.3 * 0.1
Experiment 2
uninjected and
Percent
Surviving
75
85
50
75
75
S5
33
33
30
67
63
-
73
80
87
87
73
47

100
100
80
95
100
96
68
42
60
71
69
-
77
80
83
89
91
91


Dose
-
-
0.05 ml/injection

20 mg/mouse

4 mg/injection

1 mg/injection

I mg/injection

0.20 mg/injection

0.02 mg/injection

0.02 mg/injection


..
-
0.05 ml/injection

10 mg/mouse

2 mg/injection

1 mg/injection

1 mg/injection

0.20 mg/injaction

0.02 mg/ injection

0.02 rag/injection


injected controls (p*0.05).
uninjected controls and Olds


. diesel extract
_
                       424

-------
             THE INFLUENCE OF INHALED DIESEL ENGINE EMISSIONS
                UPON LUNG TUMOR  INDUCTION IN STRAIN'A1  MICE
                                    by
         William E. Pepelko, John G. Orthoefer, W. Bruce Peirano,
                     Wai den Crocker, and Freda Truman
                    Health Effects Research Laboratory
                   U.  S.  Environmental  Protection  Agency
                             Cincinnati, Ohio
     The  Strain  'A'  mouse was  one of  several  animal models chosen by  the
 Environmental Protection Agency to assess the carcinogenic risk of exposure
 to  diesel  engine emissions.  The  Strain 'A1 mouse was  selected because: it
 has  been  one of  the most  extensively  used models for assessment of lung
 tumor  induction;  the test is  well  validated;   exposure times are fairly
 short for a cancer assay; the test is relatively straight forward to perform
 and  evaluate and  finally;  it  is one  of  the most sensitive lung  tumor
 bioassays  available   (1).  The results presented are from experiments in
 which mice  were exposed  to exhaust with a  particulate concentration of 12
 mg/m^.  These studies are  a continuation of earlier work in which exposure
 levels were about one half those  used  in the present experiments (2).

     Details  of  the exposure conditions and experimental procedures have
 been published previously (2,  3),  Briefly, Strain  'A' mice obtained from
 Jackson or  Strong Laboratories were exposed 8 hrs/day, 7 days/week from 6
 weeks to either 9 or 12  months  of  age.   The mice were housed in wire cages
 and  exposed in  100 cubic feet  stainless steel  chambers.   Exhaust was
 produced by a 6  cylinder, 90 cu  inch  Nissan diesel engine.   In  order to
 simulate city  driving  conditions, the  engine  load and  speed were varied
 cyclically using the Federal Short cycle. After completion of exposure, the
 animals were sacrificed  and the lungs fixed and observed for the presence of
 pulmonary adenomas.

     Three experiments were carried out. In the first, 360 animals,  180 of
 each sex, were exposed to  clean  air or diesel  exhaust.   One  half of each
 group received a single  intraperitoneal injection  of 5 mg urethane prior to
 the  start  of  exhaust exposure.   In  the second study,  115 males  and 143
females were exposed to  diesel  exhaust, while 108 males and 142 females were
                                    425

-------
exposed to clean air.   Conditions  differed from that of the first study in
that the exposure occurred during the dark  portion  of  the  daily cycle when
the animals were presumably awake, active and respiring at  a greater level.
In the final study in which only males were  used, the  mice were sacrificed
at 12 months instead of 9 months of age. The number of  mice  with tumors were
compared using Chi  Square testing.  Results  are shown in Table 1.  Exposure
of mice to diesel exhaust until 9 months of age resulted in a significant
decrease in the mean tumor incidence in females, and  in males and females
combined (P <.05).  Decreases were also noted in males,  but  differences were
not significant.  In comparing groups treated with  an  initiating dose of 5
mg urethane prior to exposure,  a decreased tumor incidence was again noted
in the exhaust exposed mice, (P < .10) for  males,  (P   <.001) for females,
and (P < .001) for males  and females combined.  Exposure to diesel exhaust
during the dark  portion  of  the  daily cycle   also resulted in a decreased
tumor incidence compared with clean  air controls (P  < .01)  for  males,  (P<
.001) for females, and (P< .001) for males and females combined.  Finally,
exposure of males to diesel exhaust until 12 months of  age again resulted in
a decrease in lung tumor  incidence  (P  < .01).

     Overall, there was no indication that  exposure of Strain 'A1  mice  to
diesel engine emissions resulted in an increase in lung tumor incidence.  On
the  contrary,  the  studies  consistently showed  that  tumor   rates  were
decreased in exhaust exposed mice.  Such a decrease following exposure to a
potentially carcinogenic  pollutant is rare,  but is not  unknown.  Nettesheim
et al., (4) reported that inhalation of a combination of  ozonized gasoline
and ferric oxide particles  inhibited the tumorigenic  effects of  injected
diethylnitrosamine on  the respiratory tract.  Kotin and Folk (5) also showed
that exposure of  C57BL mice for  a  lifetime  to  an  atmosphere of  ozonized
gasoline resulted in a significantly lower incidence of malignant lymphomas
and hepatomas compared to mice breathing clean air.  Finally, Pereira (6)
found fewer gamma glutamyl transpeptidase positive liver  islands  in  rats
following exposure to diesel exhaust than  in clean  air controls.

     An  explanation  of  the present results must  await  further  study.
Possibly,  diesel  exhaust  inhalation inhibits the  induction of  enzymes
responsible for  converting procarcinogens  to their  active  forms.  The
immunocompetence of  the animals could also have been altered as a result  of
the inflammatory reaction  to  deposited exhaust  particulate.   The  results
cannot be explained by increased  mortality  of mice susceptible   to tumor
induction  since survival  rates  were  not   significantly  altered  by the
exposure positions.
                                   426

-------
                                REFERENCES
1.   Shimkin, M.B.,  and C.D.  Stoner.  1975.   Lung  Tumors In mice:   Ap-
     plication to Carcinogenesis Assay.  Adv. Cancer. Res. 21:1-58.

2.   Orthoefer,  J.G., W.Moore, D. Kraemer,  F. Truman, W. Crocker, and Y.Y.
     Yang.  1980.  Carcinogenicity of Diesel Exhaust as Tested  in Strain 'A'
     Mice.  Presented at the U.S. Environmental Protection Agency Interna-
     tional Symposium on Health  Effects  of  Diesel  Engine  Emission.  Cin-
     cinnati, Ohio.

3.   Hinners, R.G.,  O.K.  Burkart, M.  Malanchuk, and W.D. Wagner.   1980.
     Facilities  for  Diesel  Exhaust Studies.    Presented at  the  U.  S.
     Environmental Protection Agency  International  Symposium on  Health
     Effects of  Diesel Engine Emissions.  Cincinnati, Ohio.

4.   Nettesheim, P.,  D.A. Creasia, and  T.J.  Mitchell.  1975.  Carcinogenic
     and Co-Carcinogenic Effects  of Inhaled Synthetic Smog  and Ferric Oxide
     particles.   J. NCI. 55:  159-169.

5.   Kotin,  P.   and  H.L.  Falk.    1956.    The  Experimental   Induction  of
     Pulmonary Tumors and  Changes in the Respiratory Epithelium  in C578L
     Mice Following Their Exposure to an Atmosphere of Ozonized Gasoline.
     Cancer.  11: 473-481.

6.   Pereira, M.A., H. Shinozuka, and B. Lombardi.   1980.  Test of  Diesel
     Exhaust Emissions In the Rat Liver Foci Assay.   Presented at the U. S.
     Environmental  Protection Agency  International  Symposium on  Health
     Effects of  Diesel Engine Emission.  Cincinnati, Ohio.
                                    427

-------
                               Table  1.    Effects of  Inhaled Diesel Engine Emissions Upon Lung Tumor
                                                   Incidence in  Strain  'A' Mice
                                                                                  Number  of  Mice

                                                                                   with Tumors
ro
oo
Age
Treatment
Clean Air
Diesel Exhaust
at Sacrifice
(Months)
9
9
Illumination
During Exposure
Light
Light
Sex
M
F
M+F
M
F
M+F
Number of Surviving
Mice
10/44
11/43
21/87
5/37
4/43
9/80
P Values

NS
< .05
< .05

Clean Air
+
5 mg Urethane
Diesel Exhaust
+
5 mg Urethane
9
9
Light
Light
M
F
M+F
M
F
M+F
32/38
34/37
66/75
26/39
16/36
42/75

< .10
< .001
< .001

Clean Air
Diesel Exhaust
12
12
Light
Light
M
M
22/38
11/44

< .01

Clean Air
Diesel Exhaust
9
9
Dark
Dark
M
F
M+F
M
F
M+F
28/97
31/140
59/237
13/111
9/139
22/250

< .01
< .001
< .001

-------
OBJECTIVES AND  EXPERIMENTAL CONDITIONS OF A VW/AUDI
DIESEL ENGINE EXHAUST INHALATION STUDY

UWE HEINRICHX,  FRIEDRICH  POTTXX,
WERNER STC5BER ,  HORST KLINGENBERGXXX
 Fravmhofer-Institut fur  Toxikologie und Aerosolforschung,
3ooo Hannover,  Federal Republic of Germany,XXMedizinisches Institut
xxx nmwelthygiene'  4o°o Dusseldorf,  Federal  Republic of Germany,
   Volkswagenwerk AG,  318o Wolfsburg,  Federal Republic of Germany
INTRODUCTION AND  METHODS
   It has been  known for  some time that exhaust emissions from
passenger car engines contain polycyclic aromatic hydrocarbons (PAH)
a number of which has been shown to be carcinogenic in various ani-
                                         •
mal experiments.  However,  it has not been  possible to date to ade-
quately explore the  extent to which inhalation of automobile ex-
haust may increase the risk of cancer in human beings. This
situation presents a particular problem in so far as automobile ex-
haust, in comparison to many other emissions from combustion proc-
esses, contains only a relatively small amount of PAH but is closer
to the man in the street.  So the actual impact on human health is
difficult to evaluate.,
   In the absence of well  defined exposure groups and control
populations,it  cannot be expected  that  epidemiological  studies  will
provide reliable  conclusions as to whether a causal relationship or,
even more remote, a  dose-effect relationship does indeed exist bet-
ween automobile exhaust and lung cancer. On the other hand, long-
term experimental exposure of human beings would be impossible for
purely ethical  reasons.
   However, i t seems to  be possible to devise an adequate experi-
mental inhalation study on laboratory animals where precisely de-
fined exposure  conditions  may be maintained over a period of years
so that pathological reactions of  the test  animals  can  be  correla-
ted with exposure time and exhaust  concentration  level. The extra-
polation of the results of s.uch animal experiments to human beings,
may be limited, and  conclusions may be confined to qualitative
statements, but,  to a limited degree  an assessment  may  be  possible
provided the experiment is closely tuned to the problem to be in-
vestigated. It  is easy to  realize that negative findings in a

                              429

-------
carcinogenicity  study  of  exhaust emissions  on  a  limited number of
rats, hamsters and  mice cannot be definitely conclusive for  a
population of human beings,  numbering in  the millions,  that  are
exposed to exhaust  emissions.  This drawback applies  even to
studies involving several hundred experimental animals, because.
the amount of carcinogens inhaled is relatively  small and the
available latency period  of  about 2 years is relatively short. The
use of higher concentrations of exhaust in  an  animal experiment is
only to a limited degree  capable of compensating for the
substantially higher total exposure time  and latency period  in
human beings, because  there  are certain limits to an increase of
exposure concentrations.  For instance, the  concentrations in
animal experiments  must remain sufficiently low,  in  order to
                                         •
avoid acute  toxic effects that might substantially reduce the
life expectancy  of  the test  animals. In addition,  the very complex
ambient environment to which man is exposed may  modify  the effect
of inhaled automobile  exhaust, and this is  not taken into account
in animal inhalation experiments that simply make use of diluted
exhaust emissions.
  The purpose of this  inhalation study is to test the emissions
of a VW Diesel engine  in  an  experiment which is  designed to  provid
more detailed results  than in  case of a straightforward in-
vestigation  merely  on  the inhalation of dilute exhaust. In this
study, which is  financed  by  the Volkswagen  Company,  the emissions
are inhaled  by laboratory animals which have been pretreated with
various carcinogens in order to induce an elevated basic tumor
incidence rate in the  respiratory tract.  The probability of
observing a  statistically significant syncarcinogenic effect by
inhalationexposure  to  Diesel exhaust appears to  be substantially
greater, if  the  change in tumor incidence rate occurs within the
steep slope  of the  sigmoidal dose response  curve.  This  is to be
achieved by  establishing  an  enhanced basic  tumor incidence rate in
the laboratory animals. Along the flat and  linear initial section
of the curve of  the dose-response relationship,  it is obviously
necessary to bring  about  a relatively large change in dosage in
order to obtain  a statistically significant additional effect.
However, in  the  steep slope  of this curve a substantially smaller
increase in  dosage  would  be  sufficient. This may possibly be the

                              430

-------
case for the range  of  concentrations feasible in our experimental
studies of exhaust  emissions.  Initial results obtained by using
this experimental animal  model are described elsewhere in these
proceedings.
   The aim of the VW project is to extend the existing results of
the new animal model beyond the scope of a pilot study and provide
a firm data base. In addition, it is intended to test the
reproducibility  of  the data.

FACILITIES
Inhalation laboratories and exposure chamber
   The inhalation study on  the exhaust emissions of the VW Diesel
engine is presently in progress at our new facilities for
environmental hygiene  and inhalation tox'icology, which is part of
the Fraunhofer Institute  of Toxicology and Aerosol Research in
Hannover, West Germany. Special laboratories are now. available
which permit long-term inhalation studies designed for about 4ooo
rats, hamsters and  mice to  be  conducted under barrier conditions.
   The only access  route  for the personnel to enter this area is
via special lock chambers with shower units. Supplies and all other
materials needed can be brought into the restricted area through
large-capacity autoclaves and peracetic acid or formaldehyde lock
chambers, which  are integrated into the wall shown on the slide.
The SPF animals  intended  to be used in this experiment are brought
into the barrier area  by  a  special animal entry unit which can be
hooked to the lock  chamber  wall.
   The stainless steel inhalation chambers have a volume of about
12 m . They are  integrated  into the inhalation laboratory design
and are horizontally ventilated by the dilute emissions.
Depending on the animal species, the chambers will house 3oo to
600 animals.
Hamsters and rats are  kept  individually, and mice are held in pairs
in stainless steel  wire cages. A uniform distribution of the
horizontal flow  of  the exposure aerosol in the inhalation chambers
is established by the  use of special baffles and perforated plates
covering the walls  of  the inlet and outlet ducts.
The bottom panel of the door frames of the two-wing chamberdoors
swing up. This allows  a smooth insertion and withdrawal of the

                             431

-------
cage racks into and out  of  the chambers.
   The inhalation  laboratory  is hermetically sealed off from the
area for the measuring equipment behind the chambers. This techni-
cal section is not incorporated into the barrier area.
The ceiling and the rear wall of the chambers are part of the
hermetic seal separating the  animal area from the measuring
equipment. This division of space permits us to take measurements
in the inhalation  chambers  without the inconvenience that the
technical personnel and  their instruments have to enter the
barrier area housing  the animals. The air pressure outside the
exposure chambers  is  adjusted so that the inhalation chambers have
a slightly lower pressure than the room housing the animals and a
sli-ghtly higher pressure than the measuring station.
   Above the space for animal handling there is a working stage
accessible from outside  where additional supply and exhaust lines
for emissions and  clean  air are installed.  Each of the i inhalation
chambers used in this experiment may be connected.to its own mixing
box located directly  above  the inhalation chamber. It facilitates
the adjustment to  any desired dilution of the exposure aerosol.
   Engine bench
The exhaust emissions are produced by a VW Diesel engine connected
by an automatic transmission  to a fly wheel with an eddy-current
brake. The engine  is  continuously computer controlled to simulate
the US-72 Federal  Test Procedure Cycle.

EXPERIMENTAL PROGRAMME
   As in our pilot study ,  the total exhaust emissions and the
emissions after removal  of  the particles are investigated in a
long-term study. The  experimental animals are exposed for about
18 hours/day and 5 days/week.
   A preliminary experiment is scheduled to begin this month. It
is designed to reveal the subchronic effects of two or three ex-
haust dilutions containing  about 4, 8 or 16 mg of particles/m  for
the three species  of  animals. After an exposure period of 2 - 3
months, a series of tests on  clinical chemistry and hematology will
be conducted as well  as  an  investigation of some lung lavage
liquids. Furthermore, tests on pulmonary function and histo-
pathology will be  made at that time.

                              432

-------
 The subsequent  long-term study is to be conducted with an
 appropriate dilution  of the  exhaust emissions so that, on one hand,
 it has as high  a  particle content as possible, but on the other
 hand, it does not substantially reduce the natural life span of the
 experimental animals.  This particular dilution of exhaust emissions
 will be determined when the  results of the preliminary experiment
 are available.
    The long-term  study will  be conducted with a total of 39 test
 groups consisting of  96 animals each (Fig.l)
TEST6ROUP
HABTER
MUSTER « OU. S.C.
HABTER • J(A)P. I.TR.
HAIBTER '
RAT
RAT - on. s.c.
RAT •
IVUSE
IWSE • i(A>p. s.c. i. USE
mSE • I(A)P. S.C. 2. DOSE
lOSE • KA1P. I.TR. 1. DOSE
HUE • I1A1P. I.TR. !, OOSE
muz •
TOTAL EXHAUST
96
96
96
96
96
96
96
96
96
96
96
96
96
EXH. WITHOUT
PARTICLES
96
96
96
96
96
96
96
96
96
96
96
96
96
CUA« ii,1
96
96
%
%
36
9C
96
%
96
%
96
96
96
tOt DimraamoUMiMt
DPfl oinvTvuitno««iiM
* ran tioo«jiiCAL 
-------
   According to the  design  of our new animal test model, a number
of the animal groups will be treated with a known carcinogen in
addition to the exhaust  exposure. This will induce a specific basic
tumor incidence rate in  the respiratory tract. In case of the
hamsters, the tumors are induced by subcutaneous injection of
diethylnitrosamine  (DEN) or intratracheal instillation of benzo(a)-
pyrene (B(a)P).  In  the  rats, this is achieved by subcutaneous
injection of dipentyInitrosamine (DPN) and in the mice by sub-
cutaneous injection  or intratracheal instillation of B(a)P. A
systemic carcinogenic effect after subcutaneous injection of B(a)P
in newly born mice with  the lung as the primary target organ is
well documented. The latency period in this case is only about 6
months.
   The purpose of the increased basic tumor rates in the
respiratory tract of three  animals species induced by various
methods  is to reveal as  to  whether a syncarcinogenic or co-
carcinogenic effect  of Diesel exhaust, as it was observed in our
pilot study, can be  reproduced on a larger scale. Furthermore,  the
study should show whether the same effect can also be observed in
the other animals species and under the influence of other
carcinogens. Finally, the question is to be answered whether the
additional effect is related to carcinogenic properties of the ex-
posure aerosol or can be brought about also by non-carcinogenic
inhalation burdens.  This problem will be investigated in connection
with concurrent studies  on  the inhalation of gasoline engine exhaus'
and the  effluents of domestic coal-burning stoves.
REFERENCES
1. Heinrich,U., StSber.W.  and Pott,  F.  (198o)  in Health Effects
   of Diesel Engine Emissions:  Proceedings of an International
   Symposium, Pepelko,  W.E.,  Danner,  R.M.  and Clarke,N.A., ed. ,
   EPA-6oo/9-8o-o57b,pp.  Io26 - Io47.
2. Seidenstucker,R.,  Pott,F., Huth,F. (198o),  Abstract in Medizi-
   nisches Institut fur Omwelthygiene,  Jahresberichte 198o,Vol.l3,
   W. Giradet, Essen,  FRG.
                              434

-------
SECTION 7



EXPOSURE AND RISK ASSESSMENT
                       435

-------
POTENTIAL HEALTH RISKS FROM INCREASED  USE  OF  DIESEL  LIGHT DUTY VEHICLES
RICHARD G. CUDDIHY, ROGER 0. McCLELLAN, WILLIAM C. GRIFFITH, FRITZ A. SEILER
AND BOBBY R. SCOTT.
Inhalation Toxicology Research Institute, Lovelace B1omed1cal  and
Environmental Research Institute, P.O. Box 5890, Albuquerque,  New Mexico 87185
IMTRODUCTION
   The potential health risks for people who may be exposed to  Increased
levels of dlesel light duty vehicle emissions 1n the future are generally
expected to be similar to the risks from exposures to other combustion
products In the atmosphere.  They Include risks of developing respiratory
functional diseases, cancers of the respiratory tract and cancers of other
organs.  Extensive toxicology research programs are currently attempting to
determine 1f light duty dlesel vehicle emissions have physical or chemical
properties that would make them significantly more toxic than other combustion
products in the environment.  To date, however, no unique compounds have been
identified in dlesel emissions that add substantial new concerns to those
already raised by existing levels of these pollutants 1n urban air.  There-
fore, the goal of this report is to identify the potential contribution of
future light duty dlesel vehicle emissions to existing atmospheric pollutant
levels and to estimate an upper limit for their potential health risks.

DIESEL VEHICLE MARKET FORECAST
   The projected Increased use of dlesel light duty vehicles was stimulated by
the Federal  Corporate Average Fuel Economy requirement for 1985.   By 1985
manufacturers of light duty vehicles must attain an overall fleet average of
27.5 miles per gallon of fuel.  Vehicles equipped with dlesel  engines cur-
rently achieve more miles per gallon of fuel than those with gasoline engines
because of their higher efficiencies and because diesel  fuel contains about
15% more energy than an equal volume of gasoline.  Piesel vehicles have also
been popular with consumers because diesel fuel costs have traditionally been
less than those for gasoline.
   In the future, however, increased demand for dlesel fuel or other middle
distillate products including heating oil could raise its cost relative to
gasoline.  Current refinery processes produce about twice as much gasoline per
barrel  of crude petroleum than dlesel fuel.  Significant changes in refinery
                              436

-------
processes  aimed at producing more dlesel  fuel  relative  to  gasoline are likely
to  result  1n cost penalties.2  The efficiency  of  dlesel engines is also
likely to  be affected  by  the application  of  emission control technology.
Between  1965 and 1975  emission control  technologies reduced the efficiency of
gasoline engine vehicles  by  about 25%.3  Similar  losses of efficiency may
occur with dlesel  vehicles as emission  control  devices  are added.  However,
new engineering advances  and reductions in vehicle sizes are currently
improving  fuel  efficiency.
    In addition  to fuel  economy, fuel  supply  and vehicle costs, other factors
will Influence  future  use of diesel light duty  vehicles.  These include con-
sumer experiences with routine maintenance,  frequency of repairs and
convenience of  operation.  All  of these factors will Influence the market
forecast,  but because  adequate information is  not available, we have assumed
that dlesel  fuel  supply limitations will  restrict dlesel light duty vehicle
use to 20% of the total light duty vehicle fleet.
    After 1995,  the total  annual  distance  traveled by light duty vehicles in
the United States  has  been projected  to be about  3 x 10   km.   Therefore,
diesel vehicles are expected to travel  about 6  x  10   km per year.

EMISSIONS  FROM  DIESEL  VEHICLES AND OTHER  SOURCES
    Diesel  vehicle  exhaust contains five categories of pollutants for which
national air quality standards have been  promulgated.  These include total
suspended  particles, sulfur  oxides, nitrogen oxides, hydrocarbons and carbon
monoxide.   A summary of these emissions as measured in the exhaust of current
light duty diesel  vehicles is given in  Table 1.  The ranges of diesel vehicle
emission rates  include both  small  and large  automobiles that were driven on a
variety  of test cycles.   The total projected emissions for the entire diesel
light duty vehicle fleet  after 1995 are also given.  Values for the projected
fleet emissions given  in--parentheses  result  from  using the current proposed
federal  emissions  standards  applicable  to light duty vehicles instead of the
mid-ranges of the  measured values.  These values  are given whenever adherence
to  the emissions  standards would result 1n lower  estimates than those pro-
jected from  the measured  values.   Also  listed in Table 1 are the current EPA
estimates  of emissions  into  the environment during 1977 from all diesel
engines and  from all area  sources  and point sources.    The projected diesel
light duty vehicle emissions for 1995 are 20% to  60% of the current diesel
engine emissions and they  are  less than 3% of the current emissions of these
pollutants  from all sources.
                             437

-------
TABLE 1
CURRENT AND PROJECTED EMISSIONS OF  REGULATED  POLLUTANTS FROM DIESEL VEHICLES
AND FROM ALL AREA AND POINT SOURCES IN THE UNITED  STATES.5"11
                     Diesel Light Duty Vehicles
                 EPA Estimated 1977 Emissions
                  Current Vehicle
                   Emission Rates
                        (g/km)
Projected Fleet    All Diesel    All Sources
Emissions: 1995     Vehicles9    (thousand
 (thousand tons) (thousand tons)    tons)
Particles
Sulfur Oxides
Nitrogen Oxides
Hydrocarbons
Carbon Monoxide
0.1 -0.6 (0.12)b
0.01-0.5
0.5 -2.0 (0.62)
0.1 -0.6 (0.25)
0.3 -1.5 (2.10)
200 ( 90 )b
100
700 (350)
200 (150)
500
350
430
3500
530
1900
15000
30000
25000
30000
noooo
alncludes heavy duty vehicles, off-highway vehicles and railroad engines.
 Projected values assuming emission control advances are made to achieve
 conformance with proposed future federal emission standards.
   Because the emissions of particles, vapors and gases from dlesel light duty
vehicles are expected to be only a few percent of current emissions, they are
not likely to produce measurable changes in the levels of these pollutants in
the environment.  Mathematical modeling studies of their dispersion in the
atmosphere also support this conclusion.    The modeling studies were done
with two computer simulation models; one was used to project typical urban
concentrations of photochemical reactant gases and the second model was used
to project dlesel particle concentrations in cities of different sizes and
population densities.
   The modeling studies of photochemical reactant gases were done by Joyce
Penner, Michael MacCracken and John Walton at the Lawrence-Livermore National
Laboratory using the LIRAQ computer model.    Pollution sources, topography
and weather patterns typical of the San Francisco Bay Area were used.  When
20% of the light duty vehicle emission source term was changed to represent
diesel emissions, no significant changes were projected in the air
concentrations of nitrogen oxides, carbon monoxide, sulfur oxides, ozone or
hydrocarbons.  These results even applied for simulated weather conditions
that favored production of high atmospheric oxldant levels.
                             438

-------
   Modeling studies of the atmospheric dispersion of dlesel exhaust particles
1n urban environments were done with a computer model based upon Gaussian
plume atmospheric dispersion characteristics that was extended to represent
area sources.12  Results of these studies, projected for 20% dlesel light
duty vehicles after 1995, Indicated that the average concentration of dlesel
particles 1n U.S. cities would be about 0.2 ug/m3. A histogram of the
projected air concentrations of diesel particles for cities with more than
25,000 people is shown in Figure 1. These studies also projected that the
largest cities would average about 2 ug/m . At the present time, air
concentrations of particles In large cities average about 100 pg/m so
that the small projected increase of 2 ug/m due to dlesel particles is
consistent with their small projected contribution to the total particle
emissions in the United States as shown in Table 1.
   300
 CO
 Hi
    200
DC
LU
§
D
    100
                   CITY AIR CONCENTRATION
                  (weighted  by population density)
                      median DO. 15
                          mean DO. 19
         WEIGHTED AIR  CONCENTRATION  (yg/rrn
Fig. 1.  Projected distribution of average air concentrations of particles
from dlesel light duty  vehicles in cities having populations of more than
25,000.
                          439

-------
    The average concentrations of  diesel  particle  emissions 1n  urban  environ-
ments projected from the computer modeling studies do  not  reflect  the
potential high concentrations that could  occur  1n  congested urban  areas where
there 1s restricted air circulation.  Urban street canyons are  known to have
higher concentrations of automblle emissions than  less congested areas,
although few quantitative studies have been done to determine human exposure
levels near central metropolitan streets.  Some studies that have  been done
above streets 1n downtown Manhattan, Nashville, San Jose and St. Louis were
                                                                2
summarized In a report of the U.S. Department of Transportation.   A mathe-
matical relationship was also developed for projecting air concentrations of
non-interacting pollutants based upon the measured relationships between
levels of carbon monoxide in automobile exhaust and street canyon  air
concentrations.  The overall model equation for projecting  these air
concentrations 1s the following:
                 Pollutant Concentration  (ug/m ) • * x EF  x TC
  where; * = pollutant concentration Index calculated  as the concentration in
             yg/m  divided by the product of vehicle emission rates (g/mi)
             and the traffic count (veh/hr),
        EF = pollutant emission factor (g/m1),
        TC = street canyon traffic count  (veh/hr).
Empirical  values for * at street level, 1n the units defined above, were
between 0.04 and 0.06 for typical meteorological conditions and between 0.08
and 0.18 for unfavorable conditions.  Assuming that diesel  light duty vehicles
will meet future emissions standards, the projected air concentrations of par-
ticles are those given in Table 2 for a vehicle traffic count of 2000 vehicles
per hr.
    Gasoline vehicles burning unleaded fuel have only  about U of the particle
emission rates of diesel vehicles, although, both types of vehicles emit
similar amounts of nitrogen oxides and other gases.  Having 2Q% light duty
diesel vehicles is likely to raise current street canyon atmospheric particle
concentrations from 100 ug/m  to 120 ug/m .  If more than 20% diesel
vehicles are used 1n urban transportation or 1f their  particle emission rates
are greater than 0.12 g/km, then they could add up to  100 ug/m  to the air
1n street canyons.  Because diesel vehicles and gasoline engine vehicles emit
similar amounts of nitrogen oxides and other gases, little effect of increased
diesel vehicle utilization 1s likely to be observed in the  levels of these
pollutants.
                             440

-------
TABLE 2
PROJECTED AIR CONCENTRATIONS OF PARTICLES IN URBAN STREET CANYONS AND PARKING
GARAGES OF MAJOR METROPOLITAN AREAS (ASSUMING 20% OIESEL LIGHT PUTY VEHICLES)

                                              Total      Diesel Light Duty
                                            Particles    Vehicle Particles
                                              ug/m3            ug/m3
        Ambient Urban Background               100                2
        Urban Street Canyon3
             Typical Meteorology                                  5
             Unfavorable Meteorology                             20
        Urban Parking Garage3                                    30
aUrban  background  concentrations must be added to the projected street
 canyon concentrations to project total concentrations at street level.   These
 totals may  also provide the  background air concentrations to be added to the
 projected urban parking garage emissions.

   Particle  concentrations  in urban  garages were also modeled in the U.S.
Department of Transportation  report.   Their model consisted of a simple box
configuration vrith vehicles as pollutant sources inside of the box.  The
garage  volume, ventilation  rate and  vehicle activity were the important model
parameters.   Results of their studies indicated that for 20S diesel vehicles,
particle  concentrations inside of garages could average 30 ug/m  above
that of outside air and during peak  traffic, particle concentrations could
increase  to  500 wg/m , Table  2.

ESTIMATION OF LUNG CANCER  RISKS
   Concern for the potential  health  risks to people exposed to diesel vehicle
emissions was stimulated by observations that diesel particle extracts were
mutagenic to bacteria cells grown  in culture.  Chemical analyses of diesel
particle  samples revealed  the presence of known carcinogenic compounds
including polycyclic aromatic hydrocarbons.  However, previous studies of
London  transit workers who  were exposed to high levels of diesel bus emissions
between 1950 and 1974 failed  to  show any  increased risk of developing lung
cancers or other health effects.     Unfortunately, the results of  these
studies v/ere confounded by  the absence of smoking information in these
                             441

-------
populations, the mobility of workers, their changing ethnic  background and the
lack of follow-up after retirement.  To date, no direct observations of
Increased cancer risks 1n people exposed to dlesel engine emissions are avail-
able.  However, several very diverse attempts have been made to Infer upper
limits for these risks.  These attempts have made use of bacteria and mam-
malian cell mutagenesls assays, skin painting studies and studies of Inhaled
or Instilled dlesel particles 1n laboratory animals.  Many studies have
attempted to determine the relative potency of dlesel particle extracts as
compared to other surrogate combustion products for which human health effects
have been documented.  These Include cigarette smoke, coke oven emissions and
atmospheric participate pollution.
   The Diesel Impacts Study Committee of the National Research Council
recently completed a review of many of these studies and their review is an
important source of biological effects information used in this cancer risk
evaluation.    Much of the data summarized 1n the report describes studies
of the mutagenic potential of dlesel particle extracts as compared to
benzo(a)pyrene, cigarette smoke condensate, coke oven emissions, roofing tar
and gasoline engine particle extracts.  Although the different 1n vitro test
systems all provided a measure of the relative ability of these agents to
transform cells genetically, no quantitative relationships can be developed
from these studies alone that would predict their carcinogenic potential  in
human exposures.  Quantitative relationships can not be developed because of
the difficulties in extrapolating between effects on cells in vitro and human
carcinogenesls.  These Include the following.  Extracts of the particle
samples were usually obtained with organic solvents for these tests.  When
biological fluids or surrogates were used for the extraction, the mutagenic
activity of the test substance was markedly decreased.  Therefore, mutagens
associated with the particles may not be readily available for contact with
lung cells after Inhalation and deposition.  Also, different test systems
showed different measures of cell transforming ability for the extracts.   Some
of the mutagenesis tests with different substances depended upon activation by
added biological enzymes, but others did not.  The relative potency of some of
the samples depended upon the types of engines and fuels that were used and
upon the mechanism required for cell transformations.
    The National Research Council report reviewed studies that  used skin
painting, inhalation or  intratracheal instillation; many are still  in
progress.  Results of the skin painting studies reported by  Slaga et al.,
showed that chemical compounds,  known to be  in diesel particle  extracts,  are
                             442

-------
complete carcinogens.   However,  none  of  the inhalation studies reported to
date  have  demonstrated that diesel  emissions  are carcinogenic to laboratory
animals.
   One  conclusion  that can  be  drawn from the  National Research Council report
is that diesel  emissions  have  not been shown  to be more mutagenic or carcino-
genic than cigarette smoke  or  coke  oven  emissions on a unit particle mass
basis.  Diesel  particle emissions were also shown to be less than 1% as muta-
genic as benzol a)pyrene.  Although  not discussed in the National Research
Council report, another study  has shown  that  soot collected from urban air was
only  about 2% as nutagenic  as  benzol a)pyrene.17  The urban soot, like diesel
particles, contained both direct acting  and indirect acting mutagenic com-
pounds.  Therefore,  urban soot appears to be  similar to diesel  particle
emissions  in  its mutagenic  potency.
   In the  following  evaluation of lung cancer risks in people exposed to
diesel  vehicle  particle emissions,  it is assumed that the concentrations of
particles  in  rural and urban air, in  air near coke ovens and in cigarette
smoke can  be used  as an index  of lung cancer  risks in the exposed popula-
      12
tions.     The information displayed in Figure 2 shows the reported annual
lung  cancer risks  for groups of  smokers  and non-smokers living  in rural and
urban areas and for  coke  oven  workers.   It also indicates their exposure air
concentrations  of  particles averaged over all of the air breathed by an
individual.  These were obtained by calculating the total amount of particles
inhaled from cigarette smoke,  urban air, rural air or air near  coke ovens
during  one year and  dividing by  the total amount of air breathed in that
year.   Also shown  in Figure 2  is the projected range of exposures to future
light duty diesel  vehicle emissions for  people living in the United States.
   The  annual lung cancer risk for each  population was divided  by the average
exposure air concentration  to  obtain  the cancer risk factors shown in
Table 3.   Assuming that the exposures of people to airborne particles are
reasonable indices of their lung cancer  risks, the risk factors show that
smokers have lower cancer incidences per unit mass inhaled than nonsmokers and
coke  oven  workers.   Although large quantities of particles are  inhaled by
cigarette  smokers, they are more likely  to deposit in the upper airways and to
be cleared more quickly than the particles inhaled by nonsmokers and coke oven
workers.   For the purpose of estimating lung cancer risks for people exposed
to diesel  particles,  an annual lung cancer risk factor of 150 cancers per
100,000 persons per  mg/m3 of air over the lifetime of individuals was taken
as an upper estimate of the risk.  Using this risk factor along with the air
                             443

-------
concentration Information shown 1n Figure 1 and urban  population statistics
for the United States, we estimated that dlesel light  duty vehicle particle
exhaust 1s not likely to result 1n more than 30 lung cancers per year after
1995.  For reference  purposes, about 100,000 lung cancer deaths currently
occur 1n the United States each year.  Thus, on the average, the lung cancer
risk Is increased by  0.5X per ug/m3 of dlesel  particles.
 0.
 8
 Q_
    10*-
 03
 £
 0>
 O
 CO
O
 o>
 c
 3
 CO
 3
 c
 c
    10"
      ,-1
10"
                                           Coke Oven Workers
                                                          . Smokers
                           • Urban Non-Smokers
                        Rural Non-Smokers
                       Potential Level of
                       Diesel Emissions in
                       1995 for 20% Diesels
                io2
10"
                                    1
10'
10*
     Average Ambient Air Concentration (mg particles /m )
F1g. 2.   Measured annual  lung cancer risks for various populations  compared to
average  air concentrations of particles.
                            444

-------
TABLE 3
SUWARY OF POPULATION EXPOSURES TO AIRBORNE PARTICLES AND ANNUAL CANCER RISKS
Average Air
Study Concentration
Population mg Particles/in3
Rural Nonsmokers 0.03a
Urban Nonsmokers 0.1
Smokers (cigarettes/day)0
1- 9 2-16
10-19 18-35
20-39 36-71
40 +•
Coke Oven Workers 3d
Annual Lung
Cancer Risk Per
100,000 People
3b
7

26
47
80

400*
Annual
Cancer Risk Per
100,000 People
Per mg/m3
100
70

3
2
2

130
W8
bHaenszel et al.19
cSurgeon General  , Hammond  ,
dSm1th et al.23
eRedmond et al.24
Kahn
    22
   Two additional analyses of the carcinogenic risks from exposures to  diesel
emissions were reported during the past year.  The first analysis  was con-
tained 1n a report by Harris to the Diesel Impacts Study Committee of the
National Research Council.25  This analysis was mainly based upon  studies of
London bus garage workers.  Their lung cancer incidences were compared  to
other transit workers including engineers, bus drivers, conductors and  subway
motormen.  Although the studies failed to show a definite increased lung
cancer risk in the garage workers, Harris calculated an upper limit for their
lung cancer risks based upon statistical considerations.  The upper 95% con-
fidence limit on the increased lung cancer risks was 1% per yg/m
particulate exposure.
   A second study has been completed by DuMouchel and Harris   that esti-
mates lung cancer risks from diesel emissions based upon the results of
laboratory studies of mutagenesis and viral cell transformation produced by
diesel particle extracts.  The relative potency of the diesel particle
extracts in these test cell systems was estimated as compared to roofing tar
vapors and coke oven emissions.  The results of epidemiology studies were  used
                             445

-------
 to estimate the absolute lung cancer  risks  per  unit  of  exposures  to roofing
 tars and coke oven emissions.  By this technique DuMouchel and Harris
 estimated the upper 95% confidence limit  for exposures  of people  to dlesel
 particles to be "\.8% Increase 1n lung cancer risks per vg of partlcles/m
 of air over the entire lifetime.  As  shown  1n Table  4,  all of the estimates of
 lung cancer risk to people exposed to dlesel particle emissions are reasonably
 similar.
   The above calculations of lung cancer  risks  apply to the entire population
 of the United States.  It 1s also Important to  consider the Individual risks
 for people who might have unusually high  exposures in or near city street
 canyons, such as police officers, utility workers and some office workers.
 Garage workers may also be exposed to high levels of dlesel emissions in
 enclosed work areas.  Using the models for street canyon pollutants, it
 appears that some of these people could be exposed to dlesel exhaust particle
 concentrations on the order of 20 ug/m  of particles.  The Individual  risk
 of lung cancer for various projected exposure levels 1s shown in Table 5.
 Although the added risks from dieseT emissions appear to be small  for most of
 these groups, it should be noted that nonsmokers working or living near
 heavily polluted street canyons could double their risks of developing lung
 cancer.

 TABLE 4
 SUMMARY OF THE PROPORTIONAL INCREASED RISK OF LUNG CANCER

                                  Proportional  Increased
                                  Risk of Lung Cancer Per
	Data Sets Used	ug/m3 Particul ate	Reference

 Cigarette Smokers, Urban Residents
   and Coke Oven Workers                  0.5*a              Cuddihy,  et al.
                                              h                    25
 London Garage Workers                     l.OS               Harris
 London Garage Workers, Roofers
   and Coke Oven Workers                  1.8X               DuMouchel and
	Harris26

 aLargest estimate
 Upper 95% confidence limit
                             446

-------
 TABLE 5
   Population
Measured Annual
Lung Cancer Risk
 (cancers/year)
                Estimated Added
Diesel  Particle   Annual  Risk
 Concentration    from Diesel
    (ug/m3)      (cancers/year)
Nonsmoker Smoker3
Street Canyon
residents
Highest city 7 x
average exposure
Average city
exposure '
Rural 3 x
1 t
lO'5 8 x 10-4 2



ID'5 I

0.2

0.02
1-3 x 10"5
1-3 x 10"6

1-3 x 10'7

1-3 x 10"8
 Average U.S.  smoker of 1  1/2 packs per  day.

 SCOPE  OF RISK  ANALYSIS
    This analysis  considered  the  potential  lung cancer risk from increased
 exposure to  diesel  particles  and  projected  a small increase in the risk for
 U.S. residents.  It should be kept  in mind  that this risk estimate was based
 upon the assumption that future diesel vehicles will meet the proposed emis-
 sion standard  for  particles of 0.12 g/km and that consumers will continue to
 maintain their vehicles to meet this standard.  The current measured diesel
 vehicle  particle emissions have often been  several times this level.  Should
 this proposed  standard be  exceeded  in the future, the calculated risk would be
 proportionately higher.
    The  analysis did not directly consider  the possible interactive effects
 between  increased exposures to diesel exhaust particles and other occupational
 or environmental pollutants that  could influence the risk of lung cancer.
People who were included in the epidemiology studies that formed the basis  of
this analysis  also had exposures  to a multitude of pollutants in the environ-
ment, in homes and  in workplaces.  Therefore, this analysis assumes that their
exposures are  typical of future population exposures.  Gaseous emissions of
diesel  vehicles are  similar to those of gasoline engine vehicles, however,
these could change with use of certain emissions control  options that might
increase emissions  of oxides of nitrogen while decreasing particle emissions.
                             447

-------
    It should also be kept in mind that  the potential  health  risks  associated
with gaseous emissions alone may exceed  the risk related to particulate emis-
sions.  This is especially true for respiratory function impairment.  Most
assessments of the health risks from gaseous emissions have assumed threshold
values for the health-effect relationships with the  threshold levels being of
the same order of magnitude or higher than current ambient levels of these
pollutants.  Thus, the models do not provide a basis for predicting possible
health risks from small  incremental increases in emissions of exhaust gases.

SUNMARY
    Diesel light duty vehicles are expected to comprise 2Q% of the total light
duty vehicle fleet in the United States within 15 years.  Their use should
grow because of a perceived advantage over gasoline engine vehicles in fuel
costs, but it will be limited by the capacity to produce sufficient diesel
fuel economically.  Chemical analyses of diesel emissions have identified
potentially toxic gases including nitrogen oxides and carbon  monoxide along
with particles that contain known carcinogenic compounds including poly-
aromatic hydrocarbons.  Diesel particle extracts have been shown to be
mutagenic to cells in culture, to cause cell transformations  and to induce
tumors in the skin of mice.   Further studies have also shown  that diesel
particle extracts are not markedly different from extracts or condensates of
cigarette smoke, coke oven emissions, or urban soot in their  ability to cause
these biological effects per unit mass.
    Atmospheric concentrations of particles emitted by diesel   vehicles were
projected for major cities in the United States using a computerized atmos-
pheric dispersion model.  This model predicted that with 20%  diesel  light duty
vehicles the average city particulate pollution levels would  increase by about
0.2 yg/m  over existing levels.  The largest cities would be  expected to
have an overall increase of 1 to 2 ug/m  but in urban street  canyons and
parking garages diesel particles may add 20 to 30 ug/m  to the existing
particulate concentrations.
    By using the results of previous epidemlological studies  of smokers, coke
oven workers and urban residents, we obtained an upper estimate of lung cancer
risk that would be expected in people exposed to diesel exhaust particles.
The risk estimator was taken to be 0.0015 cancers per year per mg/m  life-
time exposure to diesel  particles.  Combining this risk factor with projected
future air concentrations of diesel particles in urban environments, we
                             448

-------
 estimated that less than 30 lung cancers per year  could  be  related to the

 projected Increased use of dlesel  light  duty vehicles In the United States.


 ACKNOWLEDGMENTS


     This research was performed under U.S.  Department of Energy Contract
 Number  DE-AC04-76EV01013.


 REFERENCES


 1.   Code of Federal  Regulations  (1980)  49CFR Part 533.
 2.   Forrest,  L.,  Lee,  W.B.  and  Smalley,  W.M.  (1980) Assessment of Environ-
     mental  Impacts of Light-Duty Vehicle Dieselization.  U.S. Department of
     Transportation DOT-TSC-NHTSA-80-5, Washington, DC.
 3.   Considine,  D.M.   (1977)  Energy Technology Handbook.  McGrawHill,  New
     York,  Sect. 3.
 4.   U.S.  Department of Energy   (1979)  Environmental Development Plan for Light
     Duty  Diesel Vehicles.   DOE/EDP-0042.
 5.   Hare,  C.T.  and Baines,  T.M.   (1979)  Characteristics of Particulate and
     Gaseous  Emissions  from  Two  Diesel Automobiles as Functions of Fuel  and
     Driving  Cycle.   Society of  Automotive Engineers Technical Paper Series
     #790424,  Warrendale, PA.
 6.   Williams,  R.L.  and Swarin,  S.J.   (1979)  Benzo(a)pyrene Emissions from
     Gasoline  and  Diesel Automobiles, Society of Automotive Engineers Technical
     Paper Series  #790419,  Warrendale, PA.
 7.   Department of Transportation   (1980) Transportation Systems Center
     internal  report.
 8.   Bartlesville  Energy Technology Center   (1980) internal report.
 9.   Springer, K.J.  and Baines,  T.M.   (1977) Emissions from Diesel  Versions of
     Production Passenger Cars.   Society  of Automotive Engineers Technical
     Paper  Series  #770818, Warrendale, PA.
 10.  Braddock, J.N.  and Gabele,  P.A.   (1977)  Emission Patterns of
     Diesel-Powered Passenger Cars  - Part II.  Society of Automotive  Engineers
     Technical Paper  Series  #770168, Warrendale, PA.
 11.  U.S. Environmental  Protection Agency  (1980)  1977 National  Emissions
     Report:   National  Emissions  Data System of the Aerometric and Emissions
     Reporting System,  EPA-450/4-80-005.
 12.  Cuddihy,  R.G.,  Seller,  F.A.. Griffith, W.C.,  Scott,  B.R. and McClellan,
     R.O.   (1980)  Potential Health  and Environmental Effects  of Diesel  Light
     Duty Vehicles.   Inhalation  Toxicology Research Institute, LMF-82,
    Albuquerque,  NM.
13.  MacCracken, M.C.,  Wuebbles,  D.J., Walton, J.J., Duewer,  W.M.  and Grant,
    K.E.  (1978) The Livermore Regional Air Quality Model:   Concept  and
    Development.  J. Appl. Meteor, 17, 254.
14. Waller, R.  (1979) Trends in Lung Cancer in London in Relation to  Exposure
     to Diesel Fumes, EPA International Symposium on the Health Effects of
    Diesel Emissions,  December 1979,  Cincinnati.
15. Report of the Health Effects Panel of the Diesel  Impacts Study Committee,
    National Research Council   (1980) Health Effects  of  Exposure  to  Diesel
    Exhaust, National Academy Press,  Washington,  DC.
                             449

-------
16. Slaga, T., Trlplett, L. and Nesnow, S.  (1979) Mutagenic and Carcinogenic
    Potency of Extracts of Diesel and Related Environmental Emissions:
    Two-Stage Carc1nogenes1s 1n Skin-Tumor Sensitive Mice (Sencar), EPA Sym-
    posium on the Health Effects of Diesel Engine Emissions, December 1979,
    Cincinnati.
17. Toklwa, H., KltamoH, S., Takahashl, K. and Ohn1sh1, Y.  (1980) Mutagenic
    and Chemical Assay of Extracts of Airborne Partlculates, Mut. Res. 77,
    99-108.
18. Bond, R.G. editor  (1972) Handbook of Environmental Control, Vol. 1: A1r
    Polutlon, CRC Press, Cleveland, OH.
19. Haenszel, W., Loveland, D.B. and Slrken, M.G.  (1962) Lung Cancer Mor-
    tality as Related to Residence and Smoking Histories I. White Males, J.
    Nat. Cancer Inst., 28, 947.
20. Surgeon General  (1979) Smoking and Health, A Report of the Surgeon
    General, U.S. Department of Health, Education and Welfare,  Washington, DC.
21. Hammond, E.C. (1968) Quantitative Relationship Between Cigarette Smoking
    and Death Rates, Nat. Cancer Inst.  Monogr. 28, 3.
22. Kahn, H.A.  (1966) The Dorn Study of Smoking and Mortality Among U.S.
    Veterans:  Report on Eight and. One-Half Years of Observation, Epidemio-
    loglcal Approaches to the Study of Cancer and Other Chronic Diseases,
    Haenszel, W., editor, Nat. Cancer Inst.  Monogr.  19, Washington,  DC., 1.
23. Smith, D.L., Johnston, O.E. and Lockwood, W.T.  (1979) The efficiency of
    Respiratory Fibers in a Coke Oven Atmosphere, Am. Ind. Hyg. Assoc.  J. 40,
    1030.
24. Redmond, C.K., Stroblno, B.R. and Cypess, R.H.  (1976) Cancer Experience
    Among Coke By-Product Workers, Occupational  Carc1nogenes1s, U.  Saff1ott1
    and O.K. Wagoner, editors, Annals of the New York Academy of Sciences,
    271, 102.
25. Harris, J.E.  (1981) Potential Risk of Lung Cancear from Diesel  Engine
    Emissions, Report to the Diesel  Impacts Study Committee, Assembly of
    Engineering, National Research Council, National  Academy Press,
    Washington, DC.
26. DuMouchel, W.H. and Harris, J.E.  (1981) Bayes and Empirical Bayes Methods
    for Combining Cancer Experiments 1n Man and Other Species,  Technical
    Report No. 24, Department of Mathematics, Massachusetts Institute of
    Technology.
                            450

-------
    HEALTH EFFECTS OF EXPOSURE TO DIESEL FUMES AND DUST IN TWO TRONA MINES

                                      by

                       M.D. Attfield and Aremita Watson
           Appalachian Laboratory for Occupational Safety and Health
             National Institute of Occupational Safety and Health
                           Morgantown, West Virginia

                                      and

                                  G.W. Weems
                     Mine Safety and Health Administration
                               Denver, Colorado


     The industrial environment is often a useful situation in which to study
the effect of health hazards, as workers usually receive higher exposures than
does the general population.  This is particularly so for miners exposed to
diesel fumes underground, since the restriction on ventilation acts to
concentrate the fumes.  This study involves 700 workers engaged in Trona
(Na2CC>3 •  NaH C03 • 2H£0) mining.  These miners were given chest radiographs,
asked questions on chest symptoms, smoking and work history, and given
spirometric tests.  In addition, comprehensive industrial hygiene surveys were
undertaken at the two mines which were studied.  The data availabe from these
surveys is being explored for dose-response relationships between health
indices and measures of diesel engine-related pollutants.  This paper reports
on some preliminary results.

     The two mines were quite similar in character.  Although one had opened in
1949 and the other in 1967, they were similar in size and employed just over
100 diesel units underground each.  Total dust levels were high in 1976
(13 mg/m3), but NO? levels were low (0.1 ppm), probably because of the high
ventilation velocities and generally low horsepower of the units.  Diesels had
been used  underground for up to 10 years at the two mines.

     About 680 white males were studied overall.  Table 1 shows statistics of
age and exposure.  Although some workers had remained at work in the mines for
many years, the predominant duration of exposure was low.  This indicates a
rapid turnover of workers; one cause of this may have been ill health arising
from exposure to dust or diesel fumes.  Mechanisms such as this can bias or
obscure dose/response relationships in epidemiological studies.

     In order to explore the possible effect of N02 on lung function, the data
were analyzed separately by age group, first overall, and again with the
omission of those with other dust exposure (359 workers).  To do this, linear
                                    451

-------
least squares models were fitted to forced vital capacity (FVC), forced
expiratory volume in one second (FEVj) and flow at 50% of VC (FEFsg).  In the
older group, both with and without those with other exposure, lung function
decline with age was unusually and significantly great in all smoking groups
(-0.50 or worse, liters/yr for smokers).  Despite this, no clear deleterious
relationship between lung function and either dust of N02 exposure could be
detected.  In the young group, about the only variable to be significantly
related to lung function was height.

     Despite the superficially negative nature of these findings, it is
believed that caution is necessary in the interpretation of these results.
This is advised, not only because of the high rates of decline in lung
function, but also because the short duration of tenure indicates the
possibility of a powerful 'healthy worker' effect.  Further analysis needs and
will be undertaken on these data; this may show whether there is a problem in
these two mines, and whether that problem is dust or diesel. exhaust.
             Table 1.  Age and Exposure Statistics of Mine Workers


                                  Age < 25  (S.D.)          Age > 25  (S.D.)


Number                               481
Age (years)                           38     (12)              22       (2)
Dust exposure (years)                  5     (16)               2       (1)
Dust exposure (mg years/m3)           74    (104j              24      (23)
Diesel exposure (years)                3      (2)               2       (1)
N02 exposure (ppm years)               0.4    (0.4)             0.2     (0.1)
Other dust exposure3 (years)           6      (8)               1       (3)

dPrincipally in coal mining.~
                                     452

-------
          MUTAGENICITY AND CHEMICAL CHARACTERISTICS OF CARBONACEOUS
                PARTICULATE MATTER FROM VEHICLES ON THE ROAD
                                      by


        William R. Pierson, Robert A. Gorse, Jr., Ann Cuneo Szkarlat,
         Wanda W. Brachaczek, Steven M. Japar, and Frank S.-C. Lee*
                  Engineering & Research Staff - Research
                             Ford Motor Company
                               P.O. Box 2053
                          Dearborn, Michigan 48121

                   Roy B. Zweidinger and Larry D. Claxton
                    U.S. Environmental Protection Agency
                Research Triangle Park, North Carolina 27711


     Two experiments were conducted in the eastbound tunnel of the Allegheny
Mountain Tunnel of the Pennsylvania Turnpike in 1979 to evaluate the bacterial
rautagenicity of the organic solvent extracts of particulate emissions from
heavy-duty Diesels and from (predominantly light-duty) gasoline-powered vehi-
cles in highway operation.  Filter samples (PTFE-Teflon-impregnated glass
fiber and PTFE membrane) collected during periods dominated by Diesel traffic
as well as periods dominated by gasoline-powered vehicles were Soxhlet-extract-
ed with dichloromethane (CH^Clz) followed by acetonitrile (CH3CN).  Concur-
rently collected ambient-air samples (in the ventilation intake fan rooms and
at a tower on the mountaintop) were treated the same way in order to distin-
guish between vehicle and ambient contributions to the mutagem'c activity of
the tunnel samples, and also to compare mutagem'c activity of vehicle and
ambient particulate-matter extracts.  Total tunnel air flow, traffic volume,
and traffic composition were monitored to permit calculation of emission rates
per unit distance driven, for Diesels and for gasoline-powered vehicles (e.g.,
Fig. D.

     Mutagenicity was determined by the Salmonella typhima-iwn plate incorpo-
ration assay [Ames test (1)] using several tester strains, with and without
microsomal activation by S9 rat-liver homogenate.  The number of revertant
colonies per km travelled was calculated for each sampling run and plotted
against traffic composition (e.g., Fig. 2) to obtain revertants/km averages
for gasoline- and Diesel-powered vehicles separately.  High performance
liquid chromatography (HPLC) profiles were obtained on the CH2C12 and CHsCN


*Present address:  Amoco Research Center, Standard Oil Company of Indiana,
 P.O.  Box 400,  Naperville, Illinois  60566
                                     453

-------
extracts.  Gas-chromatographic (GC) molecular-weight distributions (retention-
time distributions) were obtained on the CH2C12 extracts and resolved as above
according to vehicle type.

     The main findings are as follows:

(1) The Diesel-produced aerosol  in the Allegheny Tunnel  is similar to that
    encountered in dilution tubes, with respect to all  criteria, viz., per-
    centage extractable into CH2C12 (24 £ 3%), Ames mutagenicity in revertants
    per km travelled or revertants per ug of CH2Cl2-extracted material, HPLC
    fluorescence profile, and molecular-weight distribution.
(2) Expressed as revertants per ug of CH2Cl2-extracted  material, the mutagenic
    activities of the Diesel-produced aerosol  in the Allegheny Tunnel are of
    the same order of magnitude as the mutagenic activities of the ambient
    aerosol in the vicinity at Allegheny.
(3) Expressed as revertants per km travelled,  the mutagenicity of the CH2C12
    extract of the particulate emissions from  Diesels is several times that
    from gasoline-powered vehicles.
Some of the mutagenicity results are summarized in- Tables 1 and 2.

                                 REFERENCES


1.  Ames, B. N., J. McCann, and E. Yamasaki, 1975.  Methods for determining
      carcinogens as mutagens with the Salmonella/mammalian microsome muta-
      genicity test.  Mutation Research 31:347-364.
                                     454

-------
Table 1.  Mutagenicities, Thousands of TA98 Revertants per Kilometer
          Travel! ed;CH2Cl 2 Extracts, Allegheny Mountain Tunnel  1979.


Without S9
With 59


May/ June
Aug/Sept
May/June
Aug/Sept
Gasol ine-
powered
Vehicles
39+24b
19+10
26+14
12+4
Diesel
Trucks
(a)
211+113
80+20
181+40
51+7
     Average gross weight approximately 35 tons.
    "'Error  quoted  is  the  standard  deviation.
     Table 2.  Mutagenicities, TA98 Revertants per Microgram of
               CH2C1 2-extracted Material, Allegheny Mountain Tunnel
               1979.
                                   Vehicles
Ambient Air
Gasoline- Diesel

Without S9

With S9-


May/ June
Aug/Sept
May/ June
Aug/Sept
powered
3+2
4+3
2+1
2.4+1 .6
Trucks
1 .1+0.6
0.4+0.1
0.9+0.2
0.27+0.04
Over-all
1 .3
0.6
1 .0
0.4
Fan
Rooms
0.9
0.6
0.4
0.4

Tower
-
0.2
-
0.08
                                   455

-------
                      200

                      180

                      160

                      I4O

                      120

                      100

                       ao

                       60

                       40

                       20
                            10  20  30  4O  50 6O  70  80  90  100

                                 % ga«olln«-powered ("IOO-X1
Figure  1.   Plot of mg/km emission  rate of CH2Cl2-extractable  particu-
            late matter  vs. traffic composition,  Allegheny Mountain
            Tunnel August/September 1979.  Intercept at 0% gasoline-
            powered vehicles is the emission rate (186+^ mg/km)  from
            Diesels.
   60000


   50000


-  40000

i
I  30000
a

*  20000
a:

   10000
                                     »«SI 133-402.95
                                     r« - 0.868          ,
                                     Spark-rtgnltlon-dlTiaaixiO rw/km
                                     Di«Hl**(Sl.ll7.S)xiO m/km
                         10  20  30  40  50  60  70  30  90  100

                                % ga»olln»-pow«r»d (•100-X)
Figure 2.   Plot of revertants/km v^s_.  traffic composition, CH2C12  ex-
            tracts, tester strain TA98 with microsomal  activation  (+S9),
            Allegheny Mountain Tunnel  August/September  1979.
                                   456

-------
    EMISSIONS  OF GASES AND PARTICULATES FROM DIESEL TRUCKS ON THE ROAD (2)

                                      by

                                Ralsaku Kiyoura
                  Research Institute of Environmental  Science
                       4, 5-Chome, Kojimachi, Chiyoda-ku
                                 Tokyo, JAPAN


     (A)  Experiments have been conducted to estimate the diesel  trucks emission
rates at  the Nihonzaka Tunnel.  The Tunnel is a 4-lane dual  tunnel  (2 eastbound
lanes through  one tube, 2 westbound lanes through the other), 2  km long,  on  a
slight grade upward (2.5%, 1.2 km), then downward (-1.84%, 0.8 km toward  the
east), located 170 km west of Tokyo.  The vehicle traffic in the tunnel  is  very
high with an average 1,107 - 1,148 cars/h; the percent of diesel trucks  is
30 - 85%.  Magnetic counts of eastbound traffic are by car length of 6 ±  0.5 m.
(Over 6 m is almost diesel-truck.)  The 90% of the diesel trucks is 6.5  - 22 t
car weight.   The 53% is 10 - 22 t.  Intake fans above each of the tunnels force
ventilation air into the tunnel through overhead louvers at  303  m3/s.  Air  is
drawn in  also  through the vehicle entrance portal by the ramming action  of  the
traffic.   All  of the air leaves via the vehicle exit portal, at  volumes
averaging 380  m3/s.  Truck speed was 80 km/h.  Sulphur content of fuel oil  was
0.4%.  Measurement procedures are almost similar to the Allegheny tunnel  study
by William R^  Pierson and Wanda W. Brachaczek (1).  The preliminary study was
done in 1972;  the present study was started in 1979, and will continue to 1981.


     (B)   Average emissions rates of diesel truck were found as  in Table  1  and
Table 2.   Particulates are ~0.03 un by electronmicroscope.

     (C)  The overall sulphur dioxide conversion to sulphate in the Tunnel was
2% (1980), 3%  (1979).  The measurements of the ambient are on the way.
Sulfuric  acid  particulates of 2 - 10 micron spheres were observed on the  thymol
blue dye  coated films exposed in the ambient 40 meters distant from the Tunnel
portal, when relative humidity was +90%.


REFERENCES

1.   Pierson,  W.R., and W.W. Brachaczek.  Particulate matter associated_with
       vehicles on the road.  Automotive Engineering Congress and Exposition,
       Detroit, MI, Feb. 23-27, 1976.  Paper No. 760039.
                                      457

-------
          Table 1.  Diesel Gaseous Emission Rates on the Road  (g/km)a


Year                           1980-Oct.                            1979-Oct.
NO
N02
NOX
CO
S02
T-HC
CH4
NM-HC
6
0
7
5
1
1
1
0
.42
.62
.03
.04
.27
.73
.08
.63
±
£
±
±
±
±
±
±
9.
15
8.
17
20
9.
9.
17
7%
%
8%
%
%
7%
2%
%
5
0
5

1



.02
.77
.79

.14



+
±
±
--
+
—
—
— ™
7
6
7

27



.4%
.5?
.3%

%



aln 1980, numbers of measurements:  n»36.
In case of S02 measurement, n=12.  Pearson's correlation coefficient, p < 0.05.
In 1979, n=24.  In case of S02 measurement, n=8, p < 0.05.
        Table 2.  Diesel Partlculates Emission Rates on the Road (g/km)


Year                           1980-Oct.                            1979-Oct.


Total partlculates3           1.03  ±  4.9%                       0.92  ±  5.4?
Sulphate partlculates         0.041 ± 25  %                       0.051 ± 19  %
Nitrate particulates          0.003 ± 37  %                       0.003 ± 15  %
Ammonlate particulates        0.005 ± 44  %                       0.004 ± 33  %

dPart1culates of under -10 \tn were measured by high-volume air sampler.
In 1980, numbers of measurement:  n=36.  Pearson's correlation coefficient,
p < 0.05.
In 1979, numbers of measurement:  n=24, p < 0.05.
                                       458

-------
                         DIESEL BUS TERMINAL STUDY
          EFFECTS  OF  DIESEL  EMISSIONS ON AIR POLLUTANT LEVELS
                                     BY
               Robert M.  Burton,  Robert Jungers,  Jack  Suggs
              Environmental  Monitoring  Systems  Laboratory
                 U.S. Environmental  Protection  Agency
                Research  Triangle Park,  North Carolina
 INTRODUCTION
     The New York City  Bus Terminal  Diesel  Study was  designed  to measure,
collect, chemically characterize,  and  bioassay  diesel  exhaust  as it exists
after becoming resident  in the  ambient atmosphere.  Concentration  levels of
size fractionated particle mass, organic  vapors and inorganic  vapors were
determined.  Gram amounts of  size  fractionated  particle  samples were col-
lected for detailed chemical  analysis  and  bioassay  (1) screening.

STUDY DESIGN

     A semi-enclosed area of  the NYC bus  terminal in  which approximately
14QC buses operate daily was  used  for  the  diesel exhaust  collection site
(indoor, Site =H ); a second site (outdoor,  Site #2) on 9th Avenue  upwind
of the terminal, was used as  a  particle and  organic vapor background sampl-
ing and collection site.  The background  site was used for ambient control
and for comparative chemical  and physical  characterization of  the  terminal
aiesel enriched pollutants and  the unaffected outdoor  pollutants.  Back-
ground inorganic gaseous data were obtained  from nearby  ambient monitoring
stations which were also unaffected  by the  terminal pollutants.

     The terminal site was selected  so that  the majority  of the ambient
pollutant loading would be specifically concentrated  diesel exhaust origi-
nating from diesel vehicles operating  in  a  specified  area.  Street canyons
have the disadvantage of being  traffic pattern  and meteorology dependent
and are confounded by stationary sources  and gasoline  powered  vehicles.

     The traffic count of buses passing by  the  indoor  site remained^
consistent througnout the study with the weekday daily count averaging
1,415 (1,336 minimum; 1,440 maximum) and  the weekend  daily count averaging
615 (579 minimum; 643 maximum).  Monitoring  and sample collection  began
July 16, 1979 and ran daily on  a 24-hour  schedule through July 30, 1979.
                                     459

-------
     TSP hi vol (2), Dichotomous  (2), and massive  air  volume (3)  particle
samples were collected; volatile  organic compounds were  collected on Tenax
cartridges; and primary pollutant inorganic vapors were  measured  by con-
tinuous sensors.

     The Tradescantia plant system  (4) developed by  Brookhaven  National
Laboratories for on-site detection  of toxic air pollutants,  was also
operated at the indoor site.  The Tradescantia system  allows ambient air
to be screened for the presence of  mutagenic chemical  vapors.

     The majority of buses using  the terminal were equipped  with  Detroit
two-stroke heavy diesel engines.

RESULTS AND DISCUSSION

     Results of aerometric measurements reveal that  some pollutants  were
elevated considerably at the indoor site, while others there remained at
background level.  Particle measurements revealed very high  levels  of fine
particulate matter being generated  in the terminal building.

     The average Total Suspended  Particulate (TSP) 24-hour average  level in
the terminal was 325 ug/m  while  the same TSP outdoors was 120 ug/m  .
Dichotomous measurements gave a 240 ug/m3 24-hour average for indoor parti-
cles less than 2.5 microns in diameter compared to 60 ug/m   at outdoor
Site r2.  The 24-hour average coasse fraction of the dichtomous sample
(2.5-15 microns) averaged 46 ug/m   indoors and 20 ug/m  outdoors.  Weekend
particle mass levels dropped considerably more indoors than  outdoors. Re-
sults of 24-hour TSP sulfate, nitrate, and lead levels indoors were  at the
same approximate concentration as at background Site #2.  The outdoor levels
were influencing the indoor air for the three pollutants.  Ratios of sulf-
ate, nitrate, and lead to TSP were  all lower indoors than at  the  outdoor
background site.

     The massive air volume size fractionating particle collectors were
operated at both sites throughout the study.  The amount of  material
collected is shown in Table 1.

     Like the size distribution of  the particle mass collected by the
dichotomous samplers, the massive volume samplers show the majority  of
the elevated particulate mass at the indoor site to  be aerosol in the
fine size range.

Comparison of Daily Maximum Hourly  Averages for the  Gaseous  Primary  Pol-
lutants
     Real time continuous measurements of the gaseous primary pollutants
revealed peaks to occur at 8.a.m. and 5 p.m. at the  indoor site each day
when bus traffic was at a maximum.  More variation in maximum hourly
averages occured at the indoor site than outdoors for all gases measured.
A summary of gaseous pollutant measurements follows:
                                    460

-------
     Sulfur dioxide (SO.,).  Based on  15-day  averages,  there  was  no  difference
(statistically) between indoor and  outdoor maximum  hourly  averages.   Both  the
Mable Dean Bacon School (outdoor) site  and the  Central  Park  (outdoor)  site
SCL levels are equivalent to those  found  inside the  terminal,  thus  indicat-
ing very low SO, emissions from  the buses inside  the  terminal.   There  is no
significant difference between weekday  and weekend  maximum hourly values
inside the terminal.

     Nitrogen dioxide  (NO,,).  For N02  levels, the data  indicate  more  varia-
tion in maximum hourly averages  for the  indoor  site  than outdoors.  The
maximum hourly average indoor values  (mean of 1.36  ppm) are  on the  average
significantly higher than those  for outdoors (mean  of  .09  ppm) based  on
15-day average of maximum hourly values.  Weekend maximum  hourly values
indoors are significantly lower  compared  to  weekdays.   The NCL mean of 1.36
ppm is 10 times higher than the  maximum  24-hour level  of the National Ambi-
ent Air Quality Standard.

     Nitric oxide (NO).  NO maximum hourly values indoors  throughout  the
study were significantly much higher  than those outdoors.  This  is  consist-
ent with other diesel  exhaust products measurements where  NO has been shown
to be emitted at high  concentration levels.

     Ozone (0.).  There were no  detectable 0., levels  indoors at  the bus
terminal duriflg the study.  Outdoors,  the maximum hourly 0.,  values  ranged
from 0.0 to .12 ppm with an average of  .04 ppm  over  the sample period.  With
the NO levels exceeding 7 ppm and NO-,  levels exceeding  1.0 ppm,  it  is safe
to assume all  of the 0-, at indoor Site =H was reacting with  NO to form NO/,.
                       J                                                  L.
     Carbon monoxide (CO).  There was  significantly more day-to-day variation
in the maximum hourly  CO values  for the  indoor  site as  compared  to  the out-
door site,  "his was apparently  due to  lower weekend  values  for  indoor Site
=1 maximum hourly averages.  Averaged  over the  sampling period,  the maximum
hourly average for indoors (17.79 ppm) was significantly higher  than  the
outdoor 10-day average of 2.6 ppm.

     Tota; hydrocarbons (THC).   No  ambient THC  data were available  for corn-
pa risoT17rEh"Tndoor~TnEI~?ir~1eve!s.   Peak hourly maximum averages for in-
door Site ?1  was at the 10 ppm level.

Comparison of Indoor Diurnal Patterns  for the Gaseous  Primary  Pollutants

     The indoor generated gaseous pollutants were also  analyzed  for diurnal
variation of concentration.  Weekday  (Monday-Friday)  hourly  averages  and
standard deviations for the study period were computed; weekend  (Saturday-
Sunday) nourly averages with standard  deviations were  also computed.  As
described below for each gaseous pollutant,  the diurnal concentration levels
for weekdays  (1400 buses) were always  considerably  higher  than for  weekends
(600 buses) for all gaseous pollutants except S02-

     Sulfur dioxide (SO.,).  Indoor  S0? diurnal  patterns are  similar for
both weekends  and weekdays.  Both wer&  influenced by  peak  hour traffic
                                   461

-------
activity occurring at 8 a.m. and  5  p.m.  each  day.   The indoor average 5
p.m. value was significantly higher during  the  weekday as compared to week-
end levels.  The ambient background SCL  levels  did not experience 8 a.m.
and 5 p.m. peaks.  As mentioned earlier,  there  were no significant differ-
ences between indoor and outdoor  peak daily hourly averages  for the dura-
tion of the study.

     Nitrogen dioxide (NOp).  Diurnal patterns  indicate significantly
higher average levels for weekdays  compared to  weekends for  the 8 a.m.  and
5 p.m. indoor hourly averages.  For weekdays  the hourly averages range  from
.06 ppm during early morning hours  (3-4  a.m.) to 1.55  ppm during 5 p.m.
rush hour.  For weekends the hourly averages  range from .08  ppm during
morning hours (3-5 a.m.) to 0.45  ppm during 5 p.m.  averaging time.   Had
more ozone been present in the terminal,  much higher l^ levels may have
been expected.

     Nitric oxide (NO).  Trends indicate  a higher  indoor NO  average level
during peak hours 8 a.m. and 5 p.m. for weekdays compared to weekends.  The
values for weekday peaks were beyond the  range  of  the  instrument but are
estimated to be approximately 10  ppm.

     Carbon monoxide (CO).  Indoor  weekday 8 a.m.  and  5 p.m.  hourly aver-
ages were significantly higher than weekend averages during  the same aver-
aging  times for weekends.  During  weekdays, hourly  averages were  slightly
elevated compared to weekend averages for hours before  8 a.m.  to after 5
p.m.  This cannot be concluded about other gaseous  pollutants  examined in
the study. The differences for CO were not significant  on  an hour-by-hour
basis during this time period, even though the  hourly averages  at  8 a.m.
and 6 p.m. were elevated above the  rest of the  hours in  the  day.

     Total hydrocarbons (THC).  Diurnal patterns indicate  elevations  in
indoor hourly averages during rush  hour activity (8  a.m.  to  5  p.m.)  for
weekday measurements.  The 5 p.m.  measurement averaged  10.3  ppm during
weekdays and is significantly higher than the weekend average  of 3.6  ppm.

Effect of Buses on Particle Levels

     A statistical analysis for describing the  relationship  between the bus
activity and particulate levels both indoors and outdoors  was  performed.  A
paired t-test was used to statistically examine the  difference  between the
indoor and outdoor sites for TSP, dichotomous (fine, coarse,  and total),
sulfate, nitrate, and lead.  The  inside measurements were  on the average
significantly higher (<* = .05) than outside for TSP, (0-2.5y)  and  (0-15u)
particles.  These were the only significant differences.   There was np_
significant difference between indoor and outdoor  (2.5-15y)  coarse  fraction
particles.

     Correlation coefficients between daily bus activity and pollutant
levels were calculated.   Several  correlations were  determined  to be signifi-
cantly different from zero.  These  were  specifically TSP indoors .84; (0-
                                   462

-------
2.5)  indoors,  -.7£; (0-15y) indoors, .87; sulfate indoors, -.65_; sulfate
outdoors,  -.65.

     An important observation is that the bus activity does not correlate
very highly with indoor sulfate, nitrate, and lead.  Indoor levels of
these pollutants are essentially the same as outdoors.  Obviously the
bus emissions  contribution to sulfate, nitrate, and lead are lower than
levels of these pollutants already  resident in the atmosphere.

CONCLUSIONS

     Sulfate,  nitrate, and lead emissions from the buses were at a low
level.  Sulfur dioxide from the bus emissions were also at a low level,
since no significant difference between the indoor and outdoor SCL levels
was found.  Small  particles below 2.5u aerodynamic diameter, and the gaseous
pollutants of NO,  N09, THC, and CO  were all emitted at high levels from the
buses.  The indoor stte was somewhat shielded from ultraviolet radiation,
and its absence could have an effect on the organic exhaust products found
in the atmosphere  (5).  Ozone was below detectable limits due to its use
in the production  of NOT
                                    463

-------
TABLE 1.  Amount of Size-Separated Ambient Particles Collected
          by the Massive Air Volume Sampler
STAGE I (20-3.5y)        Stage II (3.5-1.7y)      Stage III (1.7-Oy)

Site #1  7.67 gm              1.72 gm                  61.89 gm
Indoors
Site 32  6.06 gm              1.18 gm                  14.68 gm
Outdoors
                                 464

-------
                                REFERENCES
1.    Huisingh,  J.  et.  al.   "Application of Bioassay to the Characterization
       of Diesel  Particle Emissions."  In:  Application of Short-Term
       Bioassays  in the Fractionation and Analysis of Complex Environmental
       Mixtures,  M. Waters, et. al.  eds.  Plenum Press, New York, 1979.

2.    Rodes,  C.   "Inhalable Particulate Network Operations and Quality
       Assurance  Manual,"  U.S. EPA Office of Research and Development,
       Environmental Monitoring Systems Laboratory, Research Triangle Park,
       N.C.  27711 ,  May 1980.

3.    Mitcnell,  R.I., et. al.   "Massive Volume Sampler for Gram Quantities
       of Respirable Aerosols."  APCA Proceedings, June 22-24, 1977, Toron-
       to, Canada.

4.    Scnairer,  L.A., et. al.   "Measurement of Biological Activity of Ambi-
       ent Air Mixtures Using  a Mobile Laboratory for ln_ Situ Exposures:
       Dreliminary Results from the Tradescantia Plant Test System."
       pp. 419-440 in Application.

5.    Claxton, L.  and H.M. Barnes.  "The Mutagenicity of Diesel Exhaust
       Exposed to Smog Chamber Conditions as Shown by Salmonella Typhimu-
       rium," submitted to Mutation Research for publication.
                                   465

-------
                          DIESEL BUS TERMINAL STUDY:
           CHARACTERIZATION OF VOLATILE AND  PARTICLE  BOUND ORGANICS

                                      by

                   Robert H. Jungers and Joseph  E.  Bumgarner
                     U. S. Environmental Protection Agency
                 Research Triangle Park, North Carolina  27711

                  Charles M. Sparacino and Edo D.  Pellizzari
                          Research Triangle  Institute
                 Research Triangle Park, North Carolina  27709
     The  New  York City Bus Terminal was selected  as  a  unique source of heavy
duty  diesel  bus  engines.   To evaluate  emissions from  these bus  engines in
comparison to  ambient air a  semi-enclosed  site was  selected  inside the ter-
minal where  approximately 1400 diesel buses  operate  daily and  a  second site
was  selected  outside and upwind of the terminal.  Volatile organic compounds
were  collected  on  Tenax cartridges,  recovered  by  thermal  desorption and
introduced into  a high resolution gas chromatographic  column for  separation.
Characterization  and quantification of these compounds were  accomplished by
mass  spectrometry  measuring  total   ion  current  and  mass  fragmentography.
Volatile  chemicals  selected  for quantitation were benzene;  toluene; xylenes;
ethylbenzene;  1,1,1-trichloroethane;  trichloroethylene;  tetrachloroethylene;
benzaldehyde;  n-octane;  and  n-hepane.   These were  selected  to  represent
chemicals  associated with diesel  engines,  i.e.,  octane,  heptane,  benzalde-
hyde,  chemicals  not associated with  diesels, i.e.,  chlorinated hydrocarbons
and general aromatic chemicals which could  be found in  the  atmosphere.

     The  quantitation  of the  volatile  organic  compounds  collected over a
two-week  period  showed that  the chemicals  which had  consistently  higher con-
centration inside were  n-octane,  n-heptane,  1,1,1-trichloroethane,  xylenes,
and  toluene.   The chemical which was  higher  outside was tetrachloroethylene
while  benzaldehyde,  trichloroethylene, benzene,  and  ethylbenzene  were  about
the same  concentration.  Week day (1400 buses/day) concentrations  were  higher
for all ten chemicals than on weekends (600 buses/day).

     Air  particles were collected inside and  outside  (at  the  same  site  as the
volatile  samplers).   Total suspended  particulate  (TSP) measurement was  done
by the  standard  Hi-Volume sampler method (1) and  size  fractionated particles
were collected using a massive air volume  sampler (MAVS) which  separates the
particles  into three size ranges.   Table 1 presents  data on  samplers,  mass,
organic extractable and benzo-ot-pyrene analysis.
                                     466

-------
     The  smallest size  range  (1.7  urn and  below)  of  all  samples collected
inside the terminal  were combined into a  single sample.   This procedure was
also  followed  for the  outside samples.  This was  done to insure sufficient
quantity of sample for chemical and biological analysis.

     The  air  particle concentration,  organic concentration  and  the percent
organic extractables were, generally, considerably  higher  inside than outside
the terminal  while the benzo-ot-pyrene  concentration  was  higher outside than
inside the terminal.

     The  air  particle  samples were subjected to a  fractionation procedure to
yield six  fractions  of various chemical properties and polarities.  The acid
fraction  contains both  weak  (e.g.,  phenols)  and strong (e.g.,  carboxylic
acids)  acids.   The  base  fraction contains  organic,  Bronsted  bases  (e.g.,
amines).  The  neutral  fraction is subdivided into  three main fractions based
on compound polarity.   The non-polar neutral (NPN) fraction  is comprised of
compounds  less  polar than  ~  naphthalene.   Paraffinic materials  are charac-
teristic of this  fraction.  The PNA fraction contains  compounds of intermedi-
ate polarity,  and is  selective for  condensed  ring  aromatics.   All  neutral
materials  with polarities  greater  than  PNA hydrocarbons are found  in the
polar neutral  (PN)  fraction.   Prior to the  subfractionation of  the neutral
fraction,  the latter must  be dissolved  in  cyclohexane.   All components are
not soluble in  this  solvent.  The insoluble material   is collected as a sepa-
rate  fraction (CI),  and  is comprised of  intermediate and highly polar com-
pounds.

     Spillover of various compounds  into all fractions  is a  natural  feature
of  solvent  partitioning processes.  Polar neutral  material was removed from
the PNA fraction  by  silica gel chromatography.  The PNAs were chromatographed
utilizing  HPLC such that  a  fraction  containing  only PNA  hydrocarbons was
obtained  (PNA-1).   Other fractions  (PNA  2-4) were collected that contained
compounds of intermediate to high polarity.

     Most fractions  were  directly analyzed by capillary GC/MS.  The fractions
enriched in polynuclear aromatic hydrocarbons (PNAs) were  further purified by
column chromatography, and the collected subfractions  were analyzed by GC/MS.
A portion  of  each sample, after fractionation,  was prepared  for bioassay by
removal   of  the  fractionating  solvent  and addition of dimethylsulfoxide
(DMSO).

     Comparison of the mass distribution of each chemical  fraction inside and
outside the bus  terminal  showed several significant differences.   The organ-
ics from  the  outside air contained  a  higher percent   mass of one  of the PNA
subfractions  (PNA-3),  the polar  neutral  fraction  as well as  the acids and
bases.  The non  polar neutral fraction was  present at a  higher percent mass
inside the bus  terminal.   This appears to be due to higher concentrations of
alkanes   from  unburned  fuel.   Bioassay analysis  of   the  non  polar  neutral
fraction (2)  suggests  that  this fraction may contain substantial  amounts of
polynuclear aromatic hydrocarbons.
                                      467

-------
REFERENCES
     Code  of  Federal  Regulations.   1980.   Title 40,  part 50,  appendix  B.
          Reference Method  for  Determination of Suspended Particulates in the
          Atmosphere  (High Volume  Method).   General  Service Administration:
          Washington, DC.   pp.  531-535.

     Huisingh,  J., R.  Bradow,  R.   Jungers, L.  Claxton,  R.   Zweidinger,  S.
          Tejada,  J.  Bumgarner,   F.   Duffield,  V.F.   Simmon, C.  Hare,  C.
          Rodriguez,  L.  Snow, and  M. Waters.   1978.   Application of bioassay
          to  the  characterization  of  diesel  particle  emissions.   Part  I.
          Characterization  of Heavy Duty  Diesel  Particle Emissions.   Part II.
          Application  of  a  mutagenicity bioassay to monitoring light  duty
          diesel  particle emissions.  Application  of  Short-term Bioassays  in
          the  Fractionation  and  Analysis of Complex  Environmental  Mixtures.
          M.D. Waters, S. Nesnow, J.L.  Huisingh, S.S.  Sandhu, and L.  Claxton,
          eds. Plenum Press:  New York.   pp. 381-418.
                                    Table 1

SAMPLER
particle
size range
Hi-Vol
(0-50 urn)
MAVS I
(3.5-20 urn)
MAVS II
(1.7-3.5 (jm)
MAVS III
(0-1.7 |jm)
Particle
Concentration
ug Parti cle/M3
I
325.0
58.5
6.5
214.5
0
120.0
24.0
6.0
66.0
Organic
Concentration
jag Organics/M3
I
14.95
7.37
0.84
60.49
0
4.08
0.98
0.31
9.57
% Organic
Extractable
ug Orgam'cs/
100 ug Particle
I
4.6
12.6
13.0
28.2
0
3.4
4.1
5.2
14.5
BaP
ug BaP/g
Particle
I
15.4
5.9
11.9
0.8
0
32.0
25.2
34.3
5.9
I = Inside
0 = Outside
                                     468

-------
  DIESEL BUS TERMINAL STUDY:  MUTAGENICITY OF THE PARTICLE-BOUND ORGANICS AND
                               ORGANIC FRACTIONS

                                      by

                 Joellen Lewtas, Ann Austin, and Larry Claxton
                      Health Effects Research Laboratory
                     U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina


     Emission testing of both heavy-duty diesel engines and light-duty diesel
cars using tunnel dilution and filtration to collect diesel particles has shown
the organics associated with these collected particles to be mutagenic in the
Ames Salmonella typhimurium assay (1).  The mutagem'city of these organics has
been shown to be dependent on fuel quality (2) as well as engine type (3).  The
organics associated with particle emissions from heavy-duty diesel  engines have
generally been less mutagenic than the organics from light-duty cars.
Fractionation and bioassay studies suggest this is due to a greater
concentration of nonmutagenic aliphatic compounds emitted from unburned fuel.
Over 90% of the mutagenic activity has been observed in the polar neutral
fractions not requiring metabolic activation (1).  Mutagem'city studies of the
organics associated with urban ambient air particles have also reported
mutagenic activity in the organics extracted from particles (4,5).

     In order to evaluate the impact of emissions from heavy-duty buses on the
mutagenicity of ambient air, this study was designed to compare the mutagenic
activity of the total extractable organics from size-fractionated air particles
and chemical class fractions both inside and outside a diesel  bus terminal.

     The mutagenicity of the ambient air inside and outside the New York Port
Authority Bus Terminal was compared using a microbial mutagenesis bioassay.
Approximately 1400 diesel buses operate daily in the semi-enclosed  bus
terminal.  Air particles were collected simultaneously using both the Massive
Air Volume Sampler (MAVS) (6) and the standard Hi-Volume air sampler (Hi-Vol).
The dichloromethane-extractable organics from these air particles were
bioassayed in the Salmonella typhimurium plate incorporation assay  (7) in TA98
with and without metabolic activation with minor modifications (3).  The slope
of the dose-response curve (rev/ug) was determined using a nonlinear model (8).
The air particle concentration inside the bus terminal was nearly 3 times the
outside concentration based on the Hi-Vol TSP.  Comparison of the Hi-Vol and
MAVS data showed the increased concentration of particles inside was due
primarily to increased concentrations of particles less than 1.7 micron in
size.   These small  (less than 1.7 micron) particles inside the terminal had a
higher concentration of extractable organics than the small particles outside
the terminal.  Although both the small particle and organic concentrations
                                    469

-------
were lower outside the terminal, the mutagenldty of the organlcs  from outside
was significantly greater (nearly 10 times) than inside the terminal.  Both
direct-acting and indirect-acting mutagens were detected in these  samples.  The
mutagenic activity of the air 1n revertants per cubic meter provides a direct
comparison of the mutagenlcity of the Inside and outside air.  Using data from
either the Hi-Vol samples or the smallest particles from the MAVS, the outside
air was approximately twice as mutagenic as the air Inside the bus terminal.

     Fractionation and mutagenesls bloassay of the organlcs from the less-than-
1.7-micron particles were conducted to compare the chemical composition inside
and outside.  The mutagenldty (rev/jig) of each fraction and the mass
percentage of each fraction were used to calculate weighted mutagen1c1ties.
The percent of the total mutagenldty attributable to each chemical fraction
was determined and compared inside and outside the terminal.  The diesel
emissions inside the bus terminal Increased the concentration of aliphatic
hydrocarbons found in the non-polar neutral fraction.  The higher mutagenicity
in the outside ambient air appears to be due to higher concentrations of
organic acids and direct-acting moderately polar neutral compounds.  The highly
polar neutral fraction showed more direct-acting mutagenic activity Inside the
terminal.


REFERENCES

1.   Huislngh, J., R. Bradow, R. Jungers, L. Claxton, R. Zweldinger, S. Tejada,
       J. Bumgarner, F. Duffield, V.F. Simmon, C. Hare, C. Rodriguez, L.  Snow,
       and M. Waters.  1979.  Application of bloassay to the characterization
       of diesel particle emissions.  Part I.  Characterization of heavy  duty
       diesel particle emissions.  In:  Application of Short-term Bioassays in
       the Fractionation and Analysis of Complex Environmental Mixtures,
       Environmental Science Research, Vol. 15.  M.D. Waters, S. Nesnow,
       J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.  Plenum Press:   New
       York.  pp. 382-400.

2.   Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R. Zweidinger, S. Tejada,
       J. Bumgarner, F. Duffield, V.F. Simmon, C. Hare, C. Rodriguez, L.  Snow,
       and M. Waters.  1979.  Application of bioassay to the characterization
       of diesel particle emissions.  Part II.  Application of a mutagenicity
       bioassay to monitoring light duty diesel particle emissions.  In:
       Application of Short-term Bioassays in the Fractionation and Analysis of
       Complex Environmental Mixtures, Environmental Science Research,  Vol.  IB.
       M.D. Waters, S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
       Plenum Press:  New York.  pp. 400-418.

3.   Claxton, L.D.  1980.  Mutagenic and Carcinogenic Potency of Diesel and
       Related Environmental Emissions:  Salmonella Bloassay.  EPA Report
       EPA-600/9-80-057b, U.S. Environmental Protection Agency:  Research
       Triangle Park, NC.  pp. 801-809.
                                     470

-------
4.   Toklwa,  H., H. Tak1yosh1, K. Morita, K. Takahashi,  N.  Soruta,  and
       Y.  Ohnishi.  1976.  Detection of mutagenic activity  in urban air
       pollutants.  Mutat. Res. 38:351-359.

5.   Lewtas Huisingh, J.  (in press).  Bioassay of particulate organic  matter
       from ambient air.  In:  Short-term Bioassays in the  Analysis of  Complex
       Environmental Mixtures.  1980.  Michael D, Waters,  Shahbeg S. Sandhu,
       Joellen Lewtas Huisingh, Larry Claxton, and Stephen  Nesnow,  eds.  Plenum
       Press:  New York.

6.   Jungers, R., R. Burton, L. Claxton, and J. Lewtas Huisingh.   (in press).
       Evaluation of collection and extraction methods for mutagenesis  studies
       on ambient air particulate.  In:  Short-term Bioassays in  the Analysis
       of Complex Environmental Mixtures, 1980.  Michael D. Waters,
       Shahbeg S. Sandhu, Joellen Lewtas Huisingh, Larry Claxton, and Stephen
       Nesnow, eds.  Plenum Press:  New York.

7.   Ames, B.N., J. McCann, and E. Yamasaki.  1975.  Methods for  detecting
       carcinogens  and mutagens with the Salmonella/mammalian-microsome
       mutagenicity test.  Mutat. Res. 31:347-364.

8. Stead, A.G., V.  Hasselblad, J.P. Creason,  and L. Claxton.  1981.  Modeling
       the Ames test.  Mutat. Res. 85:13-27.
                                       471

-------
       NITRO DERIVATIVES OF POLYNUCLEAR AROMATIC HYDROCARBONS IN
                    AIRBORNE AND SOURCE PARTICULATE
                                  by
                           Thomas L. Gibson
                   Environmental Science Department
                 General Motors Research Laboratories
                           Warren, Michigan
INTRODUCTION
     Direct mutagenic activity is observed in the Ames Salmonella bioassay
of organic extracts from ambient airborne particulate and source emissions.
Nitro derivatives of polynuclear aromatic hydrocarbons (nitro-PNA), some
of which are strong direct-acting mutagens, are considered to be probable
contributors to this activity.

SAMPLING AND ANALYSIS

     Particulate samples were collected on dexiglas filters and extracted
with benzene-ethanol (80:20 v/v) with a Soxhlet apparatus.  Automobile
exhaust samples were obtained using a chassis dynamometer and dilution
tube.  An analytical method for nitro-PNA was recently developed which
involves reduction of these compounds to the corresponding amino-PNA and
their determination by HPLC with fluorescence detection.1  With HPLC
methods, the concentrations of 1-nitropyrene, 6-nitro-BaP, pyrene, and
BaP were measured in samples of the various particulates.  In all of
these samples, part of the nitro-PNA may have resulted from the reaction
of oxides of nitrogen found in emissions and in ambient air with PNA
bound to the particles on the filter.  For example, an increase in
nitro-PNA was measured when diesel exhaust particulate was exposed to
filtered diesel exhaust gases, suggesting a strong likelihood that nitro
artifacts are formed during filter sampling.1  The direct mutagenic
activity was determined by the Ames Salmonella bioassay (Litton Bionetics,
Kensington, MD) using tester strain TA-98 without metabolic activation.
Each particulate extract was dissolved in DMSO and tested at five doses
using triplicate plates with equal numbers of bacteria from the same
starting culture.  The slope of the initial linear part of the dose
response curve was considered as the specific activity.
                                    472

-------
AMBIENT AIRBORNE PARTICULATE

     Ambient particulate collected at an urban and a suburban site
during Spring and Summer contained 0.2-0.6 ppm (i.e., ng/mg particulate)
of 1-nitropyrene corresponding to airborne concentrations of 0.016 to
0.030 ng/m3, and also contained 0.9-2.5 ppm 6-nitro-BaP corresponding to
0.04-0.28 ng/m3.  As shown  in Figure 1, the fluorescence spectra of the
nitro-PNA were scanned to determine their identity by comparison with
authentic standards.  The direct mutagenic activity of the suburban
samples in the Ames bioassay was 0.15-0.56 revertants/microgram of
particulate corresponding to 7-20 revertants/m3 airborne mutagenicity.
Based on its reported specific activity, the 1-nitropyrene in the particu-
late could account for less than 0.3% of the total activity.

AUTOMOBILE EXHAUST PARTICULATES

     Samples were collected from a few of the numerous sources of particu-
late emissions including a  1981 2.5-L 4-cylinder catalyst car, a 1980
4.3-L 8-cylinder car with no catalyst using unleaded gasoline, a 1974
5.7-L 8-cylinder car with leaded gasoline, and 5.7-L 1980 8-cylinder
diesel cars made by General Motors divisions.  The catalyst car gave
particulate with a lower concentration of 1-nitropyrene (0.63 ppm) than
the noncatalyst cars (3.9-4.3 ppm) and production diesel car (8.0 ppm).
The catalyst car particulate had 0.21 ppm 6-nitro-BaP, the noncatalyst
cars 17-33 ppm, and the diesel car less than 0.4 ppm.  On a per mile
basis also, the catalyst car emitted less of the PNA, nitro-PNA, and
direct mutagenic activity than the other vehicles.  The level of direct
acting mutagenicity seemed  to be directly related to the total nitro-PNA
concentration of the particulate and not to PNA concentrations.

     An experimental (noncommercial) low emission diesel car was also
tested.  Compared to the production model of the 1980 diesel, the low
emission diesel gave much lower particulate, PNA, nitro-PNA, and mutagenic
emissions.

STATIONARY SOURCES

     Particulate samples from a wood-burning fireplace did not contain
levels of the nitro-PNA above the minimum detection limits (less than
0.1 ppm).  The concentrations of pyrene and BaP found in these samples
were low compared to automobile particulates and depended on how the
samples were collected:  averages of 3-4 ppm in particles collected from
the raw fl'ue gases with an  EPA Method 5 sampling train (heated filter
and impingers) compared to  30-60 ppm when collected from emissions
diluted 25-fold with air.   The increased levels of PNA in particulate
from cool, diluted fireplace emissions suggests that much of the organics
remain in the vapor phase in emissions sampled by the EPA method.
Particulate emissions measured by this method should not be compared to
vehicle emissions determined by the dilution tube method and may lead to
                                    473

-------
erroneous estimates of emissions from stationary sources (sampled from
raw flue gases).  Similarly, a particulate sample from a coal-fired
boiler, collected from hot, undiluted flue gases, showed low or undetect-
able levels of PNA (less than 0.5 ppm) and nitro-PNA (less than 0.02
ppm).

CONCLUSIONS

     Nitro-PNA are found in ambient airborne particles and various
source emissions.  Because of the very limited data from only a few of
the possible sources and the complicating effects of differences in
sampling methods, filter artifact formation, and atmospheric reactions,
source allocation for PNA derivatives in ambient particulate is not
feasible at the present time.

1.   T. L. Gibson, A. I. Ricci, and R. L. Williams, "Measurement of
     Polynuclear Aromatic Hydrocarbons, Their Derivatives and Their
     Reactivity in Diesel Automobile Exhaust," in Polynuclear Aromatic
     Hydrocarbons:  Chemistry and Biological Effects, A. Bjorseth and A.
     J. Dennis, Eds., Battelle Press (in press), presented at the 5th
     International symposium on Polynuclear Aromatic Hydrocarbons,
     Columbus, OH, Oct. 28, 1980.
                                    474

-------
                1-AMINOPYRENE
    c;
    O
                                  EXCITATION
EMISSION
   LU
   oi.
                           250   300   350   400        400   450   500
                                          WAVELENGTH (nm)
                      10
                TIME
Figure 1.   HPLC chromatogram of ambient participate extract after treatment
           with a reducing agent (Conditions, see reference1) Fluorescence
           Detector — excitation 365 nm, emission 430 nm.  Stop-flow
           scanning  gave the emission and excitation spectra shown, which
           match those of authentic 1-aminopyrene.
                                    475

-------
                      RISK ASSESSMENT OF DIESEL EMISSIONS

                                      by

                                   R. Albert
                      Institute of Environmental Medicine
                       New York University Medical Center
                              New York, New York

                                      and

                                 T. Thorslund
                     U.S. Environmental Protection Agency
                       Washington, District of Columbia


     The observations made a number of years ago, which have since been
verified, to the effect that diesel exhaust particulates are mutagenic and
contain agents that are recognized carcinogens, established the position that
diesel  particulates are likely to be carcinogenic in humans.  However, the
unanswered question Is how potent are these particulates and what is the
magnitude of the cancer hazard to the general population.  In view of the
absence of any direct animal experiments or epldemiologic data, an approach to
risk assessment several years ago which seemed reasonable was to use the
available epldemiologic data that involved exposure to combustion products
having similarities to diesel particulates and to compare the relative potency
of these materials with diesel exhaust particulates.  The epldemiologic studies
that were chosen involved cigarette smoking, coke oven emissions, and roofing
tar.  An extensive series of studies including mutagenesis, cell
transformation, skin painting, inhalation, and intratracheal intubation have
been undertaken to compare these materials with diesel particulates.  The
present status of the carcinogen risk assessment in terms of the epidemiologic
and laboratory studies will be presented.
                                     476

-------
SECTION 8
POSTER PRESENTATIONS
                       477

-------
           MUTAGENICITY OF PARTICLE-BOUND ORGANIC CHEMICAL FRACTIONS
                     FROM DIESEL AND COMPARATIVE EMISSIONS

                                      by

                 Ann Austin, Larry Claxton, and Joellen Lewtas
                          Genetic Toxicology Division
                      Health Effects Research Laboratory
                     U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina


     A variety of mobile and stationary sources emit particle-bound organics
that have demonstrated mutagenicity.  The objective of this study was to
measure the mutagenicity of the chemical class fractions derived from the total
extracted organics for diesel and several comparative emission sources.

     The four sources of combustion organics were diesel engine exhaust
particles, a cigarette smoke condensate, a coke oven main sample, and roofing
tar emissions.  The diesel exhaust particles were collected from a 1978
Oldsmobile 350 diesel vehicle operated on the highway fuel economy test cycle
(HWFET) with No. 2 diesel (Union 76) fuel.  The particles were collected on
Pallflex T60A20 filters and the organics were removed by Soxhlet extraction
with methylene chloride as previously described (1).  The 2RI Kentucky
reference cigarette smoke condensate was generated according to the method of
Patel (1977) at Oak Ridge National Laboratory (1).  The coke oven main sample
was collected from a separator located between the gas collector main and the
primary coolers within a coke oven battery at Republic Steel in Gadsden, AL,
about 60 miles northeast of Birmingham.  The roofing tar sample was generated
and collected using a conventional tar pot containing pitch-based tar, enclosed
within a chamber and heated to 360° to 380°F, a normal temperature for
commercial use.  The evaporative emissions were collected using a small bag
house fitted with Teflon filter bags (1).  The solvents used to extract or
condense the organics from each of these samples were removed by evaporation
under nitrogen.  The total extracted organics were class-fractionated into
organic acids, organic bases, cyclohexane insolubles, polar neutrals, non-polar
neutrals, and polynuclear aromatics (PNA).  The PNA fraction was further
fractionated chromatographically using gradient elution on high pressure liquid
chromatography (HPLC) such that a purified fraction containing PNA hydrocarbons
was obtained (PNA1) by elution with 2% dichloromethane (DCM) 1n hexane.
Elution with more polar solvents resulted in three additional fractions
(PNA2-4) that contained compounds of Intermediate to higher polarity.  Each
class of organics was chemically characterized using GC/MS (2).
                                       478

-------
     The  chemical  fractions for each emission source were bioassayed using the
Salmonella  typhimuriurn/microsomal plate incorporation test as developed by Ames
et al.  (3).Due to the large number of samples to be bioassayed at one time in
this study  and the limited amounts of some of the samples, only the mutant
strain  (TA98) of Salmonella typhimurium was used.  The protocol described by
Ames et al.  (3)  was followed with minor modifications (4).  The data was
analyzed  using a non-linear model (5) to determine the slope of the
dose-response curve.   Weighted mutagenlcities were determined for each fraction
based on  the mutagenicity model slope (rev/pg) and the percent of the total
mass recovered from each fraction represented.  The weighted mutagenicities
were then used to determine the percent of mutagenicity attributed to each
chemical  fraction.  Based on this data as summarized in Table 1 the following
summary can be made:

     1.   Olds Diesel.  Although the non-polar neutral fraction (NPN)
          represented the greatest percent of the total mass recovered upon
          fractionation, it accounted for very little (< 2%) of the mutagenic
          activity.  From 45 to 50% of the mutagenic activity was found in the
          polar neutral fraction (PN).  Polar neutral compounds having limited
          solubility in cyclohexane would appear in the cyclohexane insoluble
          fraction (CI), which contained from 15 to 31% of the mutagenic
          activity.  Both of these fractions (PN and CI) contained
          direct-acting mutagens.

     2.   Cigarette Smoke.  The cyclohexane insoluble fraction (CI) represented
          the greatest percent of the total mass recovered upon fractionation.
          The purified polynuclear aromatic fraction (PNA1) was the most active
          fraction; however, when the model slopes were weighted according to
          the percentage each fraction represented of the total organic sample,
          the basic fraction (BASE) accounted for the majority of the mutagenic
          activity in the presence of metabolic activation (57%), and a polar
          neutral PNA contaminant (PNA4) accounted for the majority of the
          mutagenic activity in the absence of metabolic activation (87%).

     3.   Coke Oven Mains.  The greatest percent of the total mass recovered
          after fractionation was represented by the cyclohexane insoluble
          fraction (CI).  The basic fraction (BASE) and the cyclohexane
          insoluble fraction (CI) contained the largest percentage of the
          mutagenic activity in the presence of metabolic activation (41% and
          34%, respectively).  A polar neutral PNA contaminant (PNA4) accounted
          for the majority of the mutagenic activity in the absence of
          metabolic activation (76%).

     4.   Roofing Tar.  The chemical fractions representing the greatest
          percent mass were the non-polar neutrals (NPN) and a purified PNA
          fraction (PNA1).  Although mutagenic activity was associated with
          several of the fractions, the cyclohexane insoluble fraction (CI)
          accounted for the majority (> 50%) of the mutagenic activity both in
          the presence and absence of metabolic activation.
                                       479

-------
     This study demonstrated significant biological differences among the four
emission sources.  Within each source, the relative mutagenicity of each
fraction was significantly different in the presence and absence of metabolic
activation.  The two sources which showed some similarities were the cigarette
smoke and the coke oven mains.  These sources had similar profiles in the
percent of mutagenlc activity attributed to each fraction both with and without
metabolic activation; however, chemical characterization showed significant
differences in the compounds identified 1n these two sources (2).  Further
chemical characterization of the constituents of each fraction is required to
determine which specific chemicals are biologically active within a single
source.


REFERENCES

1.   Huisingh, J.L., R.L. Bradow, R.H. Jungers, B.D. Harris, R.B. Zweidinger,
       K.M. Gushing, B.E. 6111, and R.E. Albert.  1980.  Mutagenlc and
       Carcinogenic Potency of Extracts of Diesel and Related Environmental
       Emissions:  Study Design, Sample Generation, Collection, and
       Preparation.  EPA Report EPA-600/9-80-057b.  U.S. Environmental
       Protection Agency, Research Triangle Park, NC.  pp. 788-800.

2.   Sparacino, C.M., R. Williams, and K. Brady.  1981.  Fractlonation and
       characterization of the organics from diesel and comparative emissions.
       Presented as a poster abstract at the U.S. Environmental Protection
       Agency Diesel Emissions Symposium, Raleigh, NC.

3.   Ames, B.N., J. McCann, and E. Yamasaki.  1975.  Methods for detecting
       carcinogens and mutagens with the Salmonella/mammalian-microsome
       mutagenicity test.  Mutat. Res. 31:347-364.

4.   Claxton, L.D.  1980.  Mutagenlc and Carcinogenic Potency of Diesel  and
       Related Environmental Emissions:  Salmonella Bloassay.  EPA Report
       EPA-600/9-80-057b.  U.S. Environmental Protection Agency, Research
       Triangle Park, NC.  pp. 801-809.

5.   Stead, A.G., V. Hasselblad, J.P. Creason, and L. Claxton.  1981.  Modeling
       the Ames test.  Mutat. Res. 85:13-27.
                                      480

-------
Table 1.  Percent of Mutagenlc Activity Attributed to Each Chemical Fraction
          from Comparative  Sources  (Reported  as Percent of Weighted Slope)3
                TA98  Without  Activation
TA98 With Activation

Fractions
Acid
Base
PN
NPN
PNA1
PNA2
PNA3
PNA4
CI
Olds
Diesel
9.56
4.63
44.90
0.00
0.00
9.44
0.74
0.21
30.53
Cigarette
Smoke
2.22
0.00
0.00
0.00
0.00
0.00
10.84
86.94
0.00
Coke
Oven
0.00
0.00
0.00
0.00
0.00
0.00
24.14
75.86
0.00
Roofing
Tar
0.00
0.00
4.67
0.00
0.00
6.85
0.03
0.07
88.38
Olds
Diesel
3.88
3.66
49.94
1.26
0.19
15.87
9.67
0.30
15.23
Cigarette
Smoke
0.46
57.39
0.00
1.99
0.20
0.00
0.18
8.37
31.41
Coke
Oven
0.15
40.89
8.99
7.04
4.31
0.00
2.90
1.84
33.87
Roofinq
Tar "
1.09
5.91
17.22
4.21
4.79
7.84
0.14
0.96
57.83
dPercent of mutagenic activity (% M) determined by:weighted  mutagenicity  of
 each fraction (model slope [rev/ug]  x % mass of fraction)  x  100 * total
 weighted mutagenicities of all the fractions.
                                      481

-------
            SCANNING ELECTRON MICROSCOPY OF TERMINAL AIRWAYS  OF
           GUINEA PIGS CHRONICALLY INHALING DIESEL  EXHAUST  (DE)

         Marion I. Barnhart, Fatraa Mohamed and Ahmet Kucukcelebi
                         Department of Physiology
                 Wayne State University School of Medicine
                             Detroit, MI  48201

     The structural physiology of airways near gas  exchanging alveoli was
documented to establish any changes induced by DE exposure.   Preliminary
findings are published on effects of DEP inhalation on alveolar macrophages
(1,2).  Here scanning electron mfcroscopy was used  to reveal  cell  interrela-
tions and to resolve distribution of DE partfculates CDEP)  along the terminal
airway.  Thirty guinea pfgs inhaled either 0, 250,  750 or 1500 yg  DE/m3 for
110 hr/week. for 2 weeks, 3 and 12 raon while fifteen rats were exposed for 10
weeks 6000 pg DE, 6 mon 750 pg DE and 12 mon 1500 yg DE.  Peripheral airways
were selected for study and photography wfien they were of sufficient length
to provide structural information from terminal bronchiolus to alveolar out-
pockets.  Airways were evaluated without knowledge  of the animal's history.
The relative amount and distribution of deposited particulate, was graded on
a scale of 1 to 5+,  Decoding was done later followed by final interpretations.
At least 10 terminal airways/animal were extensively photographed.  DEP was
identified as free individual particles, 0.1 + 0.03 ym, (Fig. 1).  DEP was
adherent to epithelium and irregular patches of particles were prominent at
airway bifurcations (Fig. 1AJ..  Proximal airway, characterized by an epithel-
ium of secretory and ciliated cells, had even larger agglomerates of particu-
lates especially in 12 mon 1500 and 10 week 6000 ug DE sets.  These agglomer-
ates consisted of various sized particles, only some of which had sizes appro-
priate for DEP.  Quite likely some of the admixture was secretory granules and
congealed proteins; which tend to be larger and-more irregular than particles
suspected of being DEP (Fig. IB).  Terminal bronchioles often were crowded
with macrophages and granulocytes, exiting the lung.  Surface domes were
prominent on Clara cells which may be increased in  number in  DE sets.  More
pneumocyte II cell clusters occurred at bronchiolar-alveolar  junctions in
heavily exposed than in age-matched control animals.  Eroad expanses of ter-
minal airway in DE exposed animals appeared relatively clean, but still con-
tained more particulate than companion controls.  Alveoli opening off termin-
al bronchioles had more particulates than other alveoli.  The morphology and
0.1 urn size of the spherical particles and relative sparcity  of such in con-
trols suggests that this is a visualization of the DEP burden  but is insuffi-
cient for absolute identification.  However the highest DE  exposure conditions
were associated with the dustier terminal airways.  (This study was aided in
part by General Motors Research Laboratory, Warren, Michigan).
                                     482

-------
                               REFERENCES
Barnhart, M.I., Chen, S. and H. Puro.  1980.  Impact of diesel engine exhaust
  (DEE) particles on the structural physiology of the lung.  Health Effects
  of Diesel EngineEmissions: Proc. Internet. Symp., Vol. 2, pp 649-672.
  Center for Environ Research Information EPA, Cincinnati, OH.
Barnhart, M.I., Chen, S.,Galley, S.O. and H, Puro.  1981.  Illtrastructural
  and morphometry of the alveolar lung of guinea pigs chronically exposed
  to diesel engine exhaust:  Six month's experience.  J. App. Tox. 1: 88-103.
                                           Fig.-1 A,-Terminal airway of guinea
                                           pig'exposed to 1500 ug DE fori.12.
                                           mon.  Note patches of particles
                                           whose individual size is 0.1 vim
                                           and could be DE deposits.

                                           Fig. IB. Terminal bronchiole ad-
                                           jacent to alveolus in rat exposed
                                           to-6000 yg DE for 2 mon.  Small
                                           dust particles are 0.1 ^m in
                                           diameter and probably DEP.
                                    483

-------
            EMISSION OF  DIESEL PARTICLES AND PARTICULATE  MUTAGENS
                         AT LOW AMBIENT TEMPERATURE

                                      by

                               James N. Braddock
                 Environmental Sciences Research Laboratory
                     U.S.  Environmental Protection Agency
                   Research Triangle Park,  North Carolina

     Gaseous and particulate exhaust emissions from two  1980 model diesel-
powered  passenger cars, an Oldsmobile 98 and a Volkswagen  Rabbit, were
measured over the urban dynamometer driving schedule of  the Federal Test
Procedure (FTP) as a function of ambient temperature (23°F-84°F).  Gaseous
emissions analysis included total hydrocarbons (HC), carbon monoxide (CO),
nitrogen oxides (NO),  and fuel economy (MPG).  All were  measured as a
function of the 3 individual test phases (i.e., cold transient-phase 1, stabi-
lized-phase 2, and hot  transient-phase 3)  of the FTP.  Particulate emissions
analysis included total particulate matter, particulate  organic emissions,
and molecular weight distribution measurements (also measured as a function
of FTP individual test  phase), polynuclear aromatic hydrocarbon measurements
including benzo-a-pyrene  (BaP), nitropyrene, and pyrene, and Ames Salmonella
bioassay of the particulate organics in strain TA-98.  See figure detailing
diesel emissions analysis scheme below:


                            DIESEL EMISSIONS ANALYSIS SCHEME
                                                             •HC (g/mi)
                          DIESEL EXHAUST	*  GASEOUS EMISSIONS  <
                                                              "Ox
                                                               -
                              i
                        PARTICULATE EMISSIONS
                               J
                    TOTAL PARTICULATE MATTER (mj/nii)
                                 EXTRACT WITH CH2CI2
                                 (DISCARD INERT CARBONACEOUS MATERIAL)
                             •PARTICULATE ORGANIC EMISSIONS P        NITROPYRENE     PYRENE
                          (nj/rng, m/mi)     (ng/nn./jj/mi)     (nj/mg.
            •MEASURED AS A FUNCTION OF FTP PHASE 1,2,3 AND COMPOSITE FTP.
                                         484

-------
     The data trends may be conveniently divided into two general categories,
gaseous emissions and participate emissions.  Gaseous emission trends: HC,
CO, and NOX emissions increased slightly (0.10 < r2 < 0.75) with decreas-
ing test temperature.  Oldsmobile HC emissions ranged from 0.20 to 0.29 g/mi,
CO emissions from 1.1 to 1.4 g/mi, and NOX emissions from 1.2 to 1.4 g/mi;
Volkswagen HC emissions ranged from 0.26 to 0.32 g/mi, CO emissions from 1.0
to 1.2 g/mi, and NOX emissions from 1.0 to 1.2 g/mi.  Fuel economy was more
temperature dependent (r2 > 0.80) with the Oldsmobile's decreasing from 21.6
to 18.0 mpg and the Volkswagen's decreasing from 45.7 to 41.0 mpg.

     Particulate emission trends:  total particulate matter and particulate
organic emissions (i.e., the CH2C12 extractable particulate matter)
increased with decreasing test temperature.  Comparison of overall FTP data at
82°F (median high temperature) versus 45°F (median low temperature)
indicates that Oldsmobile total particulate emissions increased from -567 to
739 mg/mi (+30%) and particulate organics  increased from  -94 to -153 mg/mi
(+63%).  For the Volkswagen, total particulate emissions  increased from -361
to "423 mg/mi (+17%) and particulate organics increased from -72 to  -101
mg/mi (+40%).  Molecular weight distributions of the particulate organ-
ics in the C-|2-C38 carbon number range, determined by gas chrornatography,
indicated that much of the organic matter  associated with the particulate
appears to be uncombusted diesel fuel.  This is evident when comparing a
lower temperature FTP to a higher temperature FTP.  Using the Oldsmobile for
example, at 32°F, 61% of the overall FTP particulate organic emission rate
of 127 mg/mi is attributable to Cn-C22 while at 82°F, only 40% of the
particulate emission rate of 83 mg/mi is attributable to C-|3-C2?-  This
uncombusted diesel fuel effect is even more pronounced in the FTP test phase
1 molecular weight distributions:  at 32°F, 68% of the Oldsmobile organic
emission rate of 137 mg/mi is attributable to C-|^-C22 while at 82°F, only
35% of the particulate emission rate of 105 mg/mi is attributable to Ci3~
C22-

     Polynuclear aromatic hydrocarbon (PAH) analysis of selected FTPs indicat-
ed decreasing (r2 -0.60) BaP, nitropyrene, and pyrene emissions (in ng/mg
extract) with decreasing FTP test temperature. Overall Oldsmobile BaP emis-
sions ranged from 0.5 to 1.1 yg/mi (average 0.7  ±0.2); nitropyrene emissions
ranged from 6.6 to 13.6 ug/mi (average 9.8±  2.2); pyrene emissions ranged
from 82.1 to 133.7 vq/mi (average 106.6±   7.4).  Overall Volkswagen BaP
emissions ranged from 15.5 to 20.2 yg/mi (average 18.2* 1.5); nitropyrene
emissions ranged from 4.2 to 12.6 yg/mi (average 8.3± 2.8); pyrene emissions
ranged from 260.4 to 354.5 ug/mi (average  293.9* 33.7).

     Ames activity levels, in terms of revertants per microgram of particu-
late organic emissions, correlated moderately (r2 = 0.73) with FTP test
temperature indicating decreasing mutagenic activity with decreasing test
temperature.  Activity levels also correlated moderately  (r^ - 0.73) with PAH
emissions indicating decreasing mutagenicity (rev/ug) with decreasing PAH
emissions (ng/mg extract).  Mutagenic activity was greater without metabolic
activation (-S9).  For the Oldsmobile, activity with metabolic activation
ranged from 1.1 to 1.6 rev/yg (average of  1.3± 0.1) and without metabolic
activation from 2.4 to 3.8 rev/ug (average of 3.4* 0.9).  Corresponding
                                       485

-------
Oldsmobile rev/mi x 103 ranges and rates were 105 to  185  (average  of 152  ±
29) with S9 and 297 to 475 (average of 383 ± 66) without  S9.   The  Volkswagen
displayed slightly greater mutagenic activity than the Oldsmobile.   Activity
with metabolic activation ranged from 1.8 to 2.9 rev/yg (average of  2.2 ±
0.3)and without metabolic activation from 3.2 to 5.7  rev/ug  (average of 4.5 ±
0.80.  Corresponding Volkswagen rev/mi x 103 ranges and rates  were 163 to
226 (average of 192 ± 23) with 59 and 341 to 426 (average of 385 ± 28)
without S9.

CONCLUSIONS

1.  The regulated gaseous emissions (HC, CO, and NOX) of  diesel-powered
passenger cars were slightly temperature dependent with decreasing FTP test
temperature slightly increasing HC, CO, and NOX emissions. Fuel economy was
moderately temperature dependent with decreasing FTP  test temperature de-
creasing fuel economy.

2.  Total particulate matter and particulate organic  emissions were  moder-
ately temperature dependent.  Decreasing FTP test temperature  increased
total particulate and particulate organic emissions.

3.  Increases in particulate organic emission rates at low FTP test  tempera-
tures may be primarily attributed to uncombusted diesel fuel.

4.  There appeared to be moderate correlation between polynuclear aromatic
hydrocarbon emissions and FTP test temperature with PAH emissions decreasing
with decreasing test temperature.

5.  There appeared to be moderate correlation between Ames test mutagenic
activity (without metabolic activation) and FTP test  temperature with muta-
genic activity decreasing with decreasing test temperature.  There also
appeared to be moderate correlation between mutagenic activity and PAH emis-
sions with mutagenicity decreasing with decreasing PAH emissions.
                                      486

-------
           THE  DESIGN OF THE CCMC'S LONG-TERM INHALATION PROGRAM TO
               INVESTIGATE THE POSSIBLE TOXICOLOGICAL EFFECTS OF
                 DIESEL AND GASOLINE ENGINE EXHAUST EMISSIONS

                                      by

             J.  Brightwell, R.D.  Cowling, X. Fouillet, R.K.  Haroz,
                         H. Pfelfer, and J.C. Shorrock
                     Center for Toxicology and Biosciences
                                   BATTELLE
                            Geneva Research Centres
                                  Switzerland


     The health effects program of the Committee of Common Market Automobile
Constructors (CCMC) on diesel and gasoline engine emissions is presented  in
another poster.   Part of this program, the long-term inhalation study,  is being
carried out by  Battelle-Geneva, and in this poster we present the design  of the
equipment and the experimental protocol.

     Four different types of emissions—diesel (D), filtered diesel (DF),
gasoline (G), and gasoline with converter (GC)--are generated by three  engines
(VW Rabbit 1.5  litre diesel and two Renault R18 1.6 litre gasoline) running on
the FTP (US-72  hot start cycle).   Two species of animals (Syrian hamsters and
Fischer-344 rats) will be exposed for up to 24 months, 16 hours per day,  5 days
per week.

     These emissions can be diluted to three dose levels:  high (H), medium
(M), and low (I).  From this 4x3 matrix of exhaust types and dose levels,
three have been omitted as being of little potential interest.  An indicative
matrix of dose  levels selected for exposure are shown in Table 1.


                        Table 1.   Matrix of Dose Levels


   Dose Levels                 D          DF          G          RC


        H                    8.3        8.3         3.6        3.6
        M                    2.8        2.8         1.2        1.2
        L                    0.92
                                     487

-------
     The figures given in the above table are the calculated mean exhaust gas
concentrations (% volume) when the exhaust gases from one cycle are diluted by
the same volume of air for all engines.  In this example the dilution air
volume per cycle for the high dose levels is 300 Nm3 (equivalent to 40
Nm3/mile).  The M and L dose levels have been set at 1/3 and 1/9 of the H
levels.

     Such dose levels are considered to be directly comparable across the table
since they are based on equal mileage.  This comparison takes into account the
different internal dilution taking place 1n the engines and also,the different
fuel effidences of the automobiles being compared.

     In Table 2, the high dose levels, HD and HG, are expected to correspond to
the following concentrations of the regulated components (running the gasoline
engine at lambda =1).


          Table 2.  High Dose Level and Concentration Correspondence

Components
Exhaust gas in air
Particulate matter
CO
Nox (NO? equivalent)
THC

Units
%
mg/m3
ppm
ppm
mg/m3
Diesel
(HD)
a. 3
5.5
20
15
9.2
Gasoline
(HG)
3.5
-
203
49
37
     The figures given in these tables are all calculated from the data
supplied by the automobile manufacturers.  They are currently being evaluated
in our system and, if necessary, the flow rate of air to the dilution tunnels
will be modified with a view to keeping concentrations of the biologically
critical components in HD and HG at as high a level as is considered compatible
with a chronic study.

     Although slight differences exist in the distribution systems for each
engine, the basic principle remains the same and is described below for one
engine.

     The exhaust gases are injected from the tail pipe directly into a dilution
tunnel where they are mixed with a constant flow of conditioned air (filtered
and dried to a water content of 7 g/kg air).  The air is dried to compensate,
at least partially, for the high water vapour content of the gasoline engine
emissions so that condensation does not take place in the dilution tunnel, and
to ensure that the relative humidity in the inhalation chambers is not too
high.
                                      488

-------
     A fraction of the diluted exhaust gas is drawn off from the mixing tunnel
into a buffer tank where it is held for about one minute.  The need for this
tank is being investigated, and it will be removed from the system if not
required.   Its role is to attenuate the high peak values of NO? and CO that are
produced by the engines over short periods during the US-72 cycle to a level at
which they do not appreciably affect the breathing pattern of the animals.

     The dilution tunnels are run at the high dose level.  The medium and low
dose levels are achieved by further diluting the high dose level with air.  The
dose levels are computed directly from flow rate measurements using rotameters.

     The high dose level streams will be continuously monitored for CO and NOX
for safety purposes.  The concentrations of the regulated components and
certain non-regulated components will be checked in the inhalation chambers at
reqular intervals.

     BatteHe-Northwest designed Hazleton-1000 inhalation exposure chambers
will be used, each chamber housing one treatment group of 72 male and 72 female
rats or 156 male and 156 female hamsters.  The control groups (fresh air) will
contain 288 rats and 624 hamsters.

     Initial and interim sacrifices will be made on 8 animals of each sex from
each group after 0, 6,  12, and 18 months of exposure.  These animals will be
used for respiratory physiology, haematology, urinalysis, and blood chemistry
investigations.

     Complete autopsies will be carried out on all animals in the study and the
animals in the highest  dose levels and the control groups will be subjected to
a histopathological examination of the respiratory tract.  Any anomalies
detected during autopsies will also be subjected to a histopathological
diagnosis.  Other organs will be stored in formalin and be available for
further examination if  required.

     System testing of  the experimental facility and  some short-term exposures
at  the high dose levels were carried out up to August 19R1.  The plan to  start
up  the long-term exposure  using staggered  intakes of  animals during the months
of  September to December 1981 has unfortunately been  delayed by several months
due to an accident.  Rebuilding and testing of part of the facility is now
necessary before the long-term exposure can be started.
                                      489

-------
                  CHRONIC INHALATION ONCOGENICITY STUDY OF
                       DIESEL EXHAUST  IN  SENCAR MICE
                                     by
            K.  I.  Campbell,  E.  L. George, I. S. Washington, Jr.
                      P.  K.  Roberson,  and R. D. Laurie
                     Health  Effects Research Laboratory
                   U. S. Environmental Protection Agency
                              Cincinnati, Ohio
     A large number of Sencar mice were used in an investigation to assess the
long term potential inhalation oncogenicity of automotive diesel  emissions.
After exposure of parental mice from before mating  on through  gestation, the
offspring continued in exposure 8 hours daily  for 15  months  to an  atmosphere
of diluted automotive diesel engine exhaust. Exhaust  dilution  was  controlled
so as  to provide a particulate  level  of about 6 mg/m^ in  exposure chamber
atmosphere through the parental phase and until the young were mature (10 wks.
of age),  at  which time this concentration was adjusted to about  12 mg/m^.
Subgroups for testing initiation  (A), promotion (B), and whole  carcinogen (C)
potentials of  diesel  exposure  were administered, respectively, weekly i.p.
injections  of  promoter  (butylated hydroxytoluene)  for  about  1  year,  an
initial single i.p. injection of  initiator  (urethan),  and neither promoter or
initiator. Matching  controls  were  exposed to purified  air.  Each  subgroup
initially numbered 260, equally  divided by sex.  Additional negative, posi-
tive, and vehicle  control groups were used.

     Over all groups,  survival was 13 percent  less  in diesel-exposed than in
control mice (75 vs 65 percent),  the  initiation  test group males  being the
most affected.  Survival was least in initiation-test  mice, due apparently in
large part  to  consequences of frequent  i.p.  BHT-in-oil injections. In all
subgroups except A females  survival was greater  in  control than  in diesel-
exposed mice.  In general, males were more  susceptible than females to reduced-
survival effects of diesel exposure, and survival  effects were least severe in
mice receiving diesel  exhaust only  (i.e.,  not receiving promoter  or initia-
tor) .
                                    490

-------
     In periodically weighed sample  groups, body weight gain was depressed in
diesel-exposed mice of both sexes in all  subgroups compared  to corresponding
clean air controls, the effect ranging from 11 to 24  percent of control mean
weight gain. Similarly, mean terminal  body weights  taken for all survivors
showed lower values  (ranging 7  to  17%)  for  both sexes in all subgroups of
diesel-exposed compared to  control mice.

     Histological  results  showed a  small but not statistically significant
overall  increase  in  lung   (alveolar  bronchiolar)   tumor  rate  (primarily
adenoma) in surviving  diesel-exposed compared to control mice. However, for
several  types  of  respiratory lesions there  were  consistenly  and  greatly
increased  incidences  in  diesel-exposed  compared to  control mice.  These
lesions  included:  alveolar  macrophages,  black  alveolar  pigment  material,
perivascular and  peribronchial mononuclear cells, focal  fibrosis,  alveolar
interstitial thickening, rhinitis (females), and black pigment  in mediastinal
lymph nodes. Predictably,  serositis  was a common lesion in mice of the BHT-in-
oil  injected groups  of both control  and diesel atmospheres.

     The efforts  of  Dr. J.  E. Proctor and others of Experimental Pathology
Laboratories,  Inc.,  who provided all  pathology support, are acknowledged.
                                     491

-------
             SPECIES DIFFERENCES IN DEPOSITION AND CLEARANCE
                    OF INHALED DIESEL EXHAUST PARTICLES
                              T. L. Chan and P. S. Lee
                           Biomedical Science Department
                        General Motors Research Laboratories
                              Warren, Michigan 48090


Deposition of  inhaled  diesel particles  and their  subsequent clearance from the gas
exchange regions of the respiratory tract may play important roles in  the  question of
potential health impact of diesel emissions.  The initial particulate deposition in the
lungs depends  on  the  physical  characteristics  of  the  particles  and the  airways
morphometry may also affect the regional deposition within the lungs.  For example,
the  narrow  nasal  passages can shift  the  particle  deposition  in  the nasal region
proximaUy for  particles larger than a micron  in small experimental animals, and the
final deposition patterns in the lungs will be significantly different from those of larger
species or man. Although diesel particles are not large enough to deposit by impaction
in the upper  respiratory tract, species differences can still exist in alveolar clearance
mechanisms, clearance pathways and  kinetics.  Table 1 compares the estimated initial
particulate  dose to  the lungs in different species  exposed to  0.1 pm particles at
250  yg/m for an hour. Although the absolute particulate burden by weight is highest
in man, the immediate local dose to the lung tissues is expected  to be five times higher
in the dog and  guinea pig.  The relative  dose is even  higher  in rats and hamsters by at
least a factor of ten.

Male Hartley guinea pigs and  Fischer 344 rats were exposed in a nose-only inhalation
chamber to radioactive diesel exhaust particles. The  particles,  tagged  in the insoluble
carbonaceous core with  ^C, were generated by combustion of (1-  C)-n-hexadecane in
a single cylinder diesel engine  operated at full load [1].  The lf*C activity in the lungs
and lymph nodes were determined for groups of exposed animals  sacrificed immediately
after the 45-minute exposure and others at scheduled  intervals  for an extended period
of time.  Although the  initial lung deposition efficiencies and mucociliary clearance
half-times were  comparable  in  both  species,  the amount of inhaled  diesel particles
cleared from the upper respiratory airways in the guinea pig  accounted  for only 17% of
the  initial lung  burden, compared  to  34% in the rat.   Furthermore, the  alveolar
clearance of  diesel particles in the guinea pig was  extremely slow, with more  than 80%
of the initial dose retained after 105 days (Figure 1).  The  pulmonary clearance half-
time for inhaled diesel particles in the guinea pig is estimated to exceed 300 days which
strongly contrasts with 60-80 days in  rats (determined  by  fitting experimental  data
collected so far to two- or  three-phase clearance models.) The  differences  observed in
this  study demonstrate a greater long-term retention of inhaled diesel particles in the
guinea  pig possibly  caused by slower clearance  processes  in  the deep lung of  this
species.   The  actual  biological dose  to  the respiratory epithelium   would also be
                                         492

-------
different in both species.  This clearly indicates the difficulty in comparing studies on
potential health effects  of  inhaled diesel particles  among different species  and in
extrapolating experimental animal data to man.
            TABLE 1.  ESTIMATED INITIAL LUNG DEPOSITION DOSE IN
               DIFFERENT SPECIES AFTER INHALATION OF 0.1 um
                     PARTICLES AT 250 ug/m  FOR 1 HOUR

Species

Body
Weight
(g)
Minute
Volume
(mL)
Deposition
Efficiency
(%)
Lung
Wt
(g)
Particulate
Burden
(ug)
Particles/g
Lung Tissue
(yg/g)
  Man          70K   7000       25       1000      26           0.025
  Dog          12K   3100       27         80      12           0.15
  G. Pig       400     125       20        3.0     0.4           0.15
  Rat         250     150       17        1.5     0.4           0.25
  Hamster      92      61       20        0.4     0.2           0.50
  REFERENCES

  1.    T. L. Chan, P. S. Lee, and W. H. Hering, Deposition and clearance of inhaled
       diesel exhaust particles  in the respiratory tract of Fischer rats.   J. Appl.
       Tox., 1:77-82, 1981.
                                       493

-------
 Lung Retention of Inhaled Diesel Particles
   100
c
•2
(O
QC.

"c
a>
      0   10  20  30 40  50  60  70 80  90  100 110

                Days Post-exposure
                     494

-------
                      SPECIES COMPARISONS OF BRONCHOALVEOLAR
                LAVAGES FROM GUINEA PIGS AND RATS EXPOSED IN VIVO
                             TO DIESEL EXHAUST (DE)

                       Shan-te Chen, Mary Ann Weller and
                               Marion I, Barnhart
                            Department of Physiology
                   Wayne State University School of Medicine
                               Detroit, MI  48201

     This ultrastructural, biochemical  and functional study involved 93 Hartley
guinea pigs and 63 Fischer rats divided  into control and DE exposed sets of 3
animals.  Preliminary  reports on certain aspects of this study have been pub-
lished (1-3).  Clean air or DE exposures of 250, 750, 1500 and 6000 ug DE par-
ticulates/m3 were inhaled for a few days or chronically up to 18 mon.  Broncho-
lavage with Dulbecco's phosphate buffered saline provided free cells, which in
controls, were 89-100% macrophages.  Viability according to dye exclusion was
90-99% in all specimens.  The absolute  number of alveolar macrophages increased
1.4 to 2 times control values (7.63 ±1.11 million cells) by 6 and 12 mon ex-
posure to 750 and 1500 ug DE in guinea  pigs.  In a short term study in rats
after 2 mon 6000 ug DE macrophage number was 5 times controls.  Macrophage size
also increased, excepting 250 ug DE sets.  Maximum diameters and surface areas
were measured on scanning electron micrographs of cytocentrifuged lavage speci-
mens.  Calculated macrophage volume increased up to 4 times the controls. After
the in vivo DE exposure there were ultrastructural signs of macrophage activa-
tion.  DE particulates were phagocytized without cytotoxic effects since phago-
somal membranes remained intact and macrophage lactic dehydrogenase activity
(a signal of cell lysis) was not in lavage fluids.  Cytofluorometry revealed
partial blockade of macrophage phagocytosis of ex vivo fed latex.  Macrophages
from 12 mon 750 and 1500 ug DE/m exposures had reduced staining for acid phos-
phatase while the cell free lavage fluids showed 2-5 fold increases in that
enzyme.  Comparison of equivalently dosed animals revealed similar magnitudes
of change in macrophage number and granulocyte recruitment,  Upon DE exposure
granulocytes became a significant percent of the free cell population by 1  mon
at 6000 ug DE, 2 mon at 750 and 1500 ug DE and 12 mon at 250 ug DE.  In rat
lavages neutrophilic granulocytes appeared contrasting with eosinophilic granu-
locyte mobilization in guinea pigs.  Lymphocytes appeared in lavages of both
species after 2 mon.   While rats appear to be less responsive to the DE chal-
lenge, both species show elevations of albumin, IgG and total  protein in the
longer duration and higher exposure sets (Table 1).  Both granulocyte emigration
and elevations in proteins at the high DE doses are features of a classic in-
flammatory response,  but the 250 ug DE exposures even to 18 mon elicited few and
generally insignificant changes over controls.  Defense capabilities of these
healthy rodent species appear adequate to cope with chronic DE challenges at
the tested concentrations.
                                       495

-------
                                   REFERENCES
    Chen, $., Weller, M.A. and M,I. Barnhart,  1980.  Effects of diesel engine
      exhaust on pulmonary alveolar macrophages.  Scanning Electron Microsc,
      3:327-338.
    Weller, M.A,, Chen, S. and M.I. Barnhart.  1981,  Acid phosphatase in al-
      veolar macrophages exposed in vivo to diesel engine exhaust.  Micron 12:
      89-90.
    Barnhart. M.I,, Chen, S, and H. Puro.  1980.  Impact of diesel engine ex-
      haust (DEE) particles on the structural physiology of the lung.  Health
      Effects of Diesel Engine Emissions:  Proc. Internet. Symp., Vol. 2, pp.
      649-672.  Center for Environ. Research Information EPA, Cincinnati, OH.
    Table 1.  Comparison of Dose-Duration Effects
         on Protein & Enzyme Content of Acellular
                                                of Diesel Exhaust Exposure
                                                Broncholavage Fluids
  ANIMALS (#) &
   CONDITIONS
                   TOTAL PROTEIN
                       mg/ml
                 ALBUMIN        IgG           ACID
                  mg/ml         mg/ml       PHOSPHATASE
                      	n M/hr/mg prot
GUINEA PIGS (58)
  CONTROLS (16)
    2 WK-18 MON

  250 ug DE (15)
    2 WK & 2,4,6 MON
    12 & 14.5 MON
    18 MON

  750 yg DE (14)
    2,6 & 8 MON
    12 MON

  1500 yg DE (13)
    2 WK, 2 & 6 MON
    12 MON

RATS (38)
  CONTROLS (12)
    2-18 MON

  250 ug DE (11)
    2 MON
    12 & 14.5 MON
    18 MON

  750 ug DE (9)
    2, 5 & 8 MON
    12 MON
                     4.78 ±1,35    1.55 ±0.66   0.14 ± 0.10    7.97 ± 4.58
                     4,82 ± 0.93
                     4.74 ± 1.01
                    10.57 ± 4.19


                     7.89 ± 2.64
                    11.50 ± 3.77


                     8.55 ± 0.80
                    17.88 ± 5.64
               1.27 ± 0.65
               1.41 ± 0.66
               3.95 ± 2.59


               2.19 ± 0.94
               5.00 ± 2.61


               2.17 ± 0.17
               4.75 ± 0.90
0.12 ± 0.10   10.32 ± 5.07
0.26 ± 0.17   12.60 ± 6.74
0.72 ± 0.19   11.54 ± 5.85
0.39 ± 0.27
0.54 ± 0.21


0.25 ± 0.07
0.92 ± 0.43
16.41  ± 3.25
29.18 ± 7.94

13.72 ± 3.07
61.76 ±23,48
                     1.94 ± 0.21    0.34 ± 0.19   0.05 ± 0.01   19.10 ±12.34
1500
    2 MON
    12 MON
          DE (6)
1.99 ± 0.45
1.84 ± 0.18
2.75 ± 0.08


2.38 ± 0.15
3.55 ± 0.41


2.06 ± 0.23
7.26 ± 1.09
                                    0.28 ± 0.15
                                    0.28 ± 0.16
                                    0.33 ± 0.08
0.04 ± 0.02
0.06 ± 0.01
0.11 ± 0.03
                                    0.41 ± 0.18   0.07 ± 0.01
                                    0.89 ± 0.26   0.08 ± 0.0
                                    0.59 ± 0.41
                                    4.18 ± 2.67
0.04 ± 0.03
0.27 ± 0.02
29.24 ±15.58
28.30 ±10.88
23.40 ± 3.56


27.26 ± 6.95
42.14 ± 6.7


43.07 ± 2.98
70.93 ±10.82,
ACKNOWLEDGEMENTS:
   Laboratories.
                  This  work  was  partially  supported by General  Motors  Research
                                      496

-------
CHEMICAL CHARACTERIZATION OF MUTAGENIC FRACTIONS OF DIESEL PARTICIPATE EXTRACTS

                                      by

                              Dilip R. Choudhury
                             Toxicology Institute
                     Division of Laboratories and Research
                      New York State Department of Health
                               Albany, New York


     Projected increased use of diesel-powered automobiles has stimulated
considerable interest in research on health effects of the particulates and
identification of deleterious compounds adsorbed to the particulates.  Diesel
particulates are highly respirable and may present significant inhalation
health hazard.  It is now well recognized that organic extracts of diesel
emission particulates exhibit significant mutagenicity as detected by Ames
Salmonella bioassay and several other short-term mutagenicity assays.  A great
deal of effort has been directed to identification of known as well as hitherto
unrecognized mutagens in the particulate extracts.

     Ue have applied Ames Salmonella reversion assay to determine mutagenic
potencies of diesel particulate extracts and to aid in isolation of mutagenic
fractions for in-depth chemical characterization.  In this presentation I will
discuss the chemical characterization of mutagenic fractions employing a
combination of complementary analytical techniques.  Extracts of particulates
collected from three vehicles run on a chassis dynamometer-dilution tube have
been examined.  On-line high performance liquid chromatography-mass
spectrometry (HPLC-MS), HPLC-ultraviolet spectroscopy, and gas chromatography-
MS provided definitive characterization of a number of compounds in mutagenic
fractions.  Several polar derivatives of polycyclic aromatic hydrocarbons
(PAHs) including some nitrated PAHs have been identified.  Some N02~PAHs are
presently known to be bacterial mutagens.  However, it is likely that several
polycyclic carbonyl compounds detected in the extracts may also be mutagenic.
                                      497

-------
  PRELIMINARY REPORT OF SYSTEMIC CARCINOGENIC STUDIES ON DIESEL AND GASOLINE
              PARTICULATE EMISSION EXTRACTS APPLIED TO MOUSE SKIN

                                      by

        N.K. Clapp, M.A. Henke, T.L. Shock, T. Triplett, and T.J. Slaga
                               Biology Division
                         Oak Ridge National Laboratory
                             Oak Ridge, Tennessee

                                      and

                                   S. Nesnow
                     Cardnogenesls and Metabolism Branch
                     U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina


     Emission particulates may constitute a potential health hazard to persons
constantly exposed.  We are determining if emission components given by skin
application might cause carclnogenesis 1n other organs.  To the skin of SENCAR
mice, we applied dichloromethane extracts from particulates collected by
filtration of cooled diluted emissions from Oldsmobile (OLDS) (1 mg/mouse),
Nissan (10 mg/mouse), and Volkswagen (VW) (10 mg/mouse) diesel, and Mustang  V-8
(3 mg/mouse) gasoline fueled engines.  Appropriate controls including
benzo(a)pyrene (BP) (0.054 mg/mouse), 12-0-tetradecanoylphorbol-13-acetate
(TPA), and aging untreated mice (CONTROLS) were maintained.  In each treatment
group, 40 male and 40 female 6-week-old mice were treated for 52 weeks
following one of two protocols:  1) A single initiation dose of the compound
was followed twice weekly by applications of TPA (2 ugh or 2) the test
compound was applied twice weekly (OLDS only, 4 mg/mouse/week).  Surviving mice
were killed 52 weeks after initiation and examined grossly; tissues from
20 different organs were routinely taken for histologic examination.  The doses
chosen were those that gave maximal tumor-initiating activity (1,2).

     Survival and tumorigenesis for the experimental groups are shown in
Table 1.  The numbers of surviving mice were significantly reduced by TPA alone
and all experimental groups given TPA reflected this treatment with decreased
survival as compared with controls.  The only group that was further reduced in
survival by the test compound was Nissan + TPA, which had the lowest survival.
Lung tumor incidences varied with treatment groups but were not different from
controls and TPA lone.  Tumors of other organs were observed randomly in the
treated groups but showed no consistent increased incidences associated with
treatments; tumors were found in the mammary gland, uterus, pituitary gland,
cervix, and liver.  No evidence of leukemogenesis was seen in killed animals,
although 35% (6/17) of dead males had leukemias in BP + TPA group.  In mice
                                      498

-------
that died prior to the kill date significant numbers exhibited squamous cell
innnc ?naNicLnh^sk1n l^u 10 *? 25% "wtastases to regional lymph nodes and
lungs in Nissan mice.  We have also observed high incidences (40 to 60*) of
amyloldosis, primarily in the spleen and liver, and pyelonephritis and
papillary necrosis in the kidneys of mice given TPA with or without test
compounds as initiators.  The relationships and pathogenesis of these diseases
are now under investigation.  The sex effect, which shows a difference in
survival as well as tumorigenesis, is not consistent between treatment groups,
and its relationship to the tumor process is unclear at this time.  When OLDS
was given repeatedly over the 52-week treatment period (protocol  2), tumor
incidences were not different from untreated controls; it was not effective as
a complete carcinogen with this dose and protocol.

     Further analysis of remaining treatment groups and complete of
observations on mice through 24 months of age will provide information about
temporal advancement and tumor incidence modifications by various treatments.

     (Research jointly sponsored by the EPA under Interagency Agreement
40-728-78, and the Office of Health and Environmental Research, U.S. Department
of Energy, under contract W-7405-eng-26 with Union Carbide Corporation.)


REFERENCES

1.  Slaga, T.J., L.L. Triplett, and S. Nesnow.  1980.  Mutagenic and
      carcinogenic potency of extracts of diesel and related environmental
      emission:  Two-stage carcinoegenesis  in skin tumor sensitive mice
       (SENCAR).  In:  Health Effects of Diesel Engine Emissions.  Proceedings
      of an International Symposium, Vol. 2.  W.E. Pepelko, R.M. Danner, and
      N.S. Clarke, eds.  EPA-600/9-80-057b.  U.S. Environmental Protection
      Agency:  Cincinnati, OH.  pp. 874-987.

2.  Nesnow, S., L.L. Triplett, and T.J. Slaga.  (in press).  Tumorigenesis of
      diesel exhaust and related emission extracts on SENCAR mouse skin.  In:
      Short-Term Bioassays in the Analysis  of Complex Environmental Mixtures,
       1980.  Michael D. Waters, Shahbeg S.  Sandhu, Joellen Lewtas Huisingh,
      Larry Claxton, and Stephen Nesnow, eds.  Plenum Press:  New York.
                                      499

-------
Table 1.  Survival  and Tumor Induction 1n SENCAR Mice Surviving  1 Year Given Diesel and
          Gasoline  Partlculate Emission Extracts to the Skin
Tumors



No.
surviving

Treatment
Untreated
controls
TPA

BP 4-TPA

OLDS + TPA

Mustang +• TPA

Nissan + TPA

VW + TPA

OLDS (no TPA)

F » female; M

Sex*
F
M
F
M
F
M
F
M
F
M
F
M
F
M
F
M
- male.
mice
(X)
100
97
67
71
81
56
74
95
80
71
47
46
68
75
97
97


Skin




T-pos1t1ve
(paps)
(X)
0
0
19
4
74
100
27
42
31
12
94
89
54
50
0
0

No. paps/
No. surviving
„
~
0.23
0.04
3.10
4.70
0.38
0.67
0.46
0.12
4.50
3.40
1.00
1.00
. —
--

T-pos1t1ve
carcinomas
0
0
0
4
0
14
3
3
9
0
11
5
8
3
0
0

Lung
(X)
3
0
4
4
6
4
14
19
3
8
0
5
0
3
0
3

M1SC.
(?)
0
0
0
4
13
0
3
0
0
0
5
(1
4
0
3
3

Leukemlas
(X)
3
0
0
0
0
0
0
6
0
0
0
0
0
0
0
0

                                        500

-------
         INFLUENCE OF  DRIVING  CYCLE  AND  CAR  TYPE  ON  THE  MUTAGENICITY
                     OF  DIESEL EXHAUST PARTICLE EXTRACTS


                                      by


                  C.  R. Clark,  A.  L.  Brooks,  R. 0. McClellan
              Lovelace Inhalation Toxicology Research  Institute
                     P. 0.  Box  5890,  Albuquerque,  NM  87185

                                      and

                       T.  M. Naman and D.  E.  Seizinger
    U. S. Department of  Energy, Bartlesville Energy  Technology Center
                   P.  0.  Box 1398, Bartlesville,  OK  74003


Solvent extracts  of  particles  collected  from the  exhaust of diesel trucks
and automobiles are  known  to be mutagenic  in bacterial test systems.  To
reasonably predict the potential  health  hazard of diesel exhaust emissions,
differences in toxicity  likely to be produced by  different cars, or under
various driving conditions, were  studied.  Extracts  of exhaust particles
collected from Oldsmobile,  Peugeot,  Fiat,  Mercedes,  Audi and Volkswagen die-
sel automobiles were evaluated for mutagenicity in standard and dinitro-
pyrene-resistant  Salmonella tester strains.

EXPERIMENTAL

Diesel exhaust particles were  collected  from the  exhaust of cars operated on
a climate-controlled chassis dynomometer at  the Bartlesville (Oklahoma) Ener-
gy Technology Center.  Cars were  acquired  new from the dealer or on loan from
the manufacturer  and allowed a 4000  mile break-in period before testing.  All
test vehicles were operated according to the EPA  Federal Test Procedure (FTP).
To study the influence of  driving cycle, exhaust  samples were also collected
while driving the Oldsmobile on the  Highway  Fuel  Economy Test (HFET) and the
New York City Cycle  (NYCC).  The  exhaust was diluted in  a tunnel sized to
cool the air/exhaust mixture to below 125°F,  and  the particulate portion col-
lected on 40 x 40 inch Pallflex T60A20 filters.   All tests were conducted
with the same standard #2  diesel  fuel.

The organic material associated with  the exhaust  particles was extracted by
ultrasonication in dichloromethane and the extract evaporated to dryness
under nitrogen.  A portion of  the extract  was fractionated by high pressure
liquid chromatography  (HPLC) on a Biosil A column and eluted in a gradient
from 94% isooctane to  100% dichloromethane.  Only the gamma-1 fraction (I)
                                      501

-------
was evaluated in these studies.  The extracts were evaluated  for mutagenicity
in Salmonella strains TA 100, TA 98, or TA 98-1,8-DNPR, without  the  addition
of a liver enzyme homogenate.  Five concentrations of  each  sample were  tested
in triplicate, and the results reported as the  slope of the dose-response
curve (revertants/yg) calculated by linear regression  analysis.   Since  re-
sults of mutagenicity testing reflect only the  genetic toxicity  of the  mater-
ial extended from the filters, the results were normalized  for differences in
the amount of extractable material associated with the particles (extractable
fraction) and particle emission rates.  This provided  an estimate of the
amount of mutagenic activity emitted from the exhaust  per mile of vehicle
operation (revertants per mile).


RESULTS

Influence of Car Type - Extracts of exhaust particles  collected  from all six
cars demonstrated direct, dose-related increases in mutagenicity in  TA  100
(Table 1).  The amount of dichloromethane extractable  material associated
with the exhaust particles (extractable fraction) produced  by the cars  varied
markedly, and was inversely related to mutagenic potency of the  extracts.
The particle emission rates varied by about 3-fold in  the six cars.  Normal-
izing the mutagenic potency for extractable fraction and particle emission
rates yielded "revertants per mile" values of a different ranking than  that
shown by the revertants per ug value.

Influence of Driving Cycle - Operating the Oldsmobile  on highway,  urban and
congested urban driving cycles did not markedly influence the mutagenicity of
exhaust particle extracts but dramatically changed the particle  emission rate
and extractable fraction (Table 2).  Increasing the extent  of stop and  start
driving increased the particle emission rate but decreased  the amount of ex-
tractable material associated with the particles.  Thus estimates of mutagen-
icity emitted from the exhaust (revertants per  mile) were similar for the
three driving cycles.

Mutagenicity in a Nitroreductase-Deficient Strain of Salmonella  - Because of
the occurrence of nitro-substituted polycyclic  aromatic hydrocarbons (PAH) in
diesel exhaust extracts (1) and the extreme potency of nitro-PAH in  Salmonel-
la, presumably due to their high nitroreductase activity (2), the extracts
were evaluated in a strain shown to be resistant to the mutagenicity of some
nitro-PAH.  Mutagenicity of the gamma-1 HPLC fractions of the extracts, pre-
viously shown to contain nitro-PAH (1), were markedly  lower (20-60%) in TA
98-1,8 DNPR than the standard TA 98 tester strain.


CONCLUSIONS

The similar mutagenic potencies of extracts of  particles collected from six
different diesel cars operated on the same fuel suggest that engine  design
has very little influence on mutagenicity of the particle associated organic
materials.  Driving cycle also did not significantly alter  mutagenicity of
the particle extracts.  The large variability in extractable fraction and
particle emission rates between different cars, and in the  same  car  operated
on different driving cycles, emphasizes the need to include these variables
                                      502

-------
when estimating the quantities of potentially hazardous materials emitted
from the exhaust.  The lower response of nitro-PAH containing fractions of
the extracts in a nitroreductase deficient Salmonella strain is difficult  to
interpret since it is not known if the enzymes which activate nitro-PAH to
mutagenic metabolites in Salmonella are unique to bacteria.  (Research per-
formed in part under DOE Contract Number DE-AC04-76EV01013.)

REFERENCES

1.  Scheutzle, D., F. S. C. Lee, T. J. Prater, and S. B. Tejada.  1981. The
      identification of polynuclear aromatic hydrocarbon (PAH) derivatives in
      mutagenic fractions of diesel particulate extracts.  Int. J. Environ.
      Anal. Chem. 9:93-145.

2.  Mermelstein, R., K. K. Demosthenes, M. Butler, E. C. McCoy and H.  S.
      Rosenkranz.  1981.  The extraordinary mutagenicity of nitropyrenes in
      bacteria.  Mutat. Res. 89:187-196.
                                       503

-------
     Table 1.  Influence of Car Type on Mutagenicity of Diesel Exhaust
                 Particle Extracts
Test Vehicle
Fiat 131
Peugeot 504
Audi 5000
Oldsmobile D-88
VW Rabbit
Mercedes 300
Rev/ug
Extract
(TA-100)
6
13
13
17
16
15
Extractable
Fraction
(«)
71
29
43
20
26
13
Particle
Emissions
(g/mi )
.34
.21
.51
.38
.17
.26
Revertants
per mile
(x 103)
1500
800
2900
1300
700
500
Federal Test Procedure, hot start used for all tests
     Table 2.  Influence of Driving Cycle on Mutagenicity of Diesel
                 Exhaust Particle Extracts
Driving
Cycle3
HFET
FTP
NYCC
Average
Speed
(mph)
50
20
7
Rev/ug
Extract
(TA-100)
15
16
13
Extractable
Fraction
(%)
35
21
12
Particle
Emissions
(g/mi)
.22
.33
1.23
Revertants
per mile
(x 103)
1100
1100
1900
^Oldsmobile Detla-88 used in all tests
                                     504

-------
                   CCMC'S HEALTH EFFECTS RESEARCH PROGRAM


                                     by
           the Members of the Emissions Research Committee of the
          CCMC (Committee of Common Market Automobile Constructors)
                           Brussels - Belgium («)
1.  GENERAL

         In the past decade much progress has been achieved in cleaning
    the air by a concerted effort of the governments and the automotive
    industry.  With more health effect data becoming available, further
    legislative measures involving more stringent car emission standards
    become more and more questionable.  Before stricter standards  for
    emissions of motor vehicles are legislated, resources for research
    programs on health effects must be deployed on an increased scale.

         Already in 1974 in the U.S. the National  Academy of Sciences (NAS)
    established that there were virtually no results available relating  to
    the effects of automobile emissions.  Since then the situation has not
    altered greatly.  It is only with the concern about cancer with regard
    to diesel-powered cars that larger research programs have now  been
    started in the U.S.

         Although a limited number of research investigations have shown
    no carcinogenic effects of human exposure to diesel-engined vehicles'
    exhaust gases up to now, the European car manufacturers have also
    initiated a large scale research project on "an investigation  into
    possible toxicological and carcinogenic effects of diesel and  gaso-
    line engine exhaust emissions".  The project is sponsored by the
    Committee of Common Market Automobile Constructors (CCMC).
(»)   Mrs.  Chevrier (Renault),  A. Henriet (Peugeot S.A.),
     H.  Klingenberg (VW),  H. Metz (BMW),  0.  Montabone (Fiat),
     N.  Pelz (Daimler-Benz),  A. Piccone (Alfa Romeo),
     S.  Mailman (Volvo),  J.H. Weaving (BL)
                                   505

-------
2.  OBJECTIVES

         This program is considered to be a first step in what will
    probably be a continuing investigation by the European industry on
    the potential hazards associated with the emissions from automobiles.

         The objectives of this first CCMC program are

         -  to compare potential toxic and carcinogenic effects
            of diesel and gasoline engine emissions,
         -  to check out the presumed beneficial effect of the
            catalytic converter of gasoline engines,
         -  to check out the often assumed relative harmlessness
            of the gaseous fraction of the diesel emissions,
         -  to investigate the mutagenic properties of diesel
            and gasoline engine particulates and condensates.
3.  PROGRAMS

         The project is divided into two parts:

    (1)  a "long term" inhalation exposure of rats and hamsters
         to gasoline and diesel engine exhausts;

    (2)  a "short term" program of in vivo and in vitro tests on
         the effects of particulate extracts and condensates from
         gasoline and diesel engine exhausts.


    3.1  Long Term Program

              The "long term" inhalation program has been contracted
         out by the CCMC to the Geneva Division of the Battelle
         Memorial Institute.

              With about 6000 hamsters and rats the combined effects of
         different concentrations of total exhaust from gasoline engines
         with and without catalyst, and a diesel engine with and without
         particulate matter, removed by filtration, operating according
         to the U.S. FTP cycle, are being investigated to determine the
         dose/response relationship, and to make a carefully controlled
         comparison of diesel and gasoline engines, with respect to
         toxicity and carcinogenicity.  The running time of the project
         is 3 years and costs will amount to 4 million U.S. dollars.
                                   506

-------
          The program is described in detail in another paper in
     this Symposium.
3.2  Short Term Program

          The "short term" program, including three main approaches,
     is being conducted in separate laboratories in England, France
     and Italy.

          In different short term in vivo and in vitro tests with
     bacteria, mammalian cells, rats and mice, the effects of partic-
     ulate extracts and condensates from gasoline and diesel engine
     exhaust, collected at Fiat, Italy, are being investigated in an
     endeavour to identify the mutagenic components.

          The running time of this project is three years and the
     costs will amount to 0.73 million U.S. dollars.
          Three main approaches were taken:

     3.2.1  "in vitro" assays - detection and identification of
            mutagens (Microtest. University of York, United Kingdom)

                 The four basic preparations (particulate and con-
            densate, both diesel and gasoline) in total  and addi-
            tionally fractionated in accordance with the EPA pro-
            cedure, are subjected to a modified AMES procedure
            (bacterial mutation assay) using five strains of sal-
            monella typhimurium (TA 1535, TA 1537, TA 1538, TA 98
            and TA 100).  The preparations are also tested for the
            ability to elicit unscheduled DNA synthesis  (DOS) in
            cultured human fibro-blast,  (DNA repair assay using
            He!a cells in culture) and in vitro transformation in
            rodent cells (mammalian cell  mutation assay  using mouse
            lymphoma L 5178 Y cells).  All these assays  have been
            combined with a liver monoxygenase enzyme fraction.

                 The tests are also carried out with extracts of
            the particulate fractions.

     3.2.2  skin painting tests - detection and identification of
            promoters (Institute of Scientific Research  on Cancer
            (CNRS), Paris, France)	

                 In the short term work sebaceous gland  and hyper-
            pi as ia  tests are performed with the different exhaust
            preparations.  Investigation  is being undertaken to sepa-
            rate potential  cancer initiators and cancer  promoters.
                               507

-------
            Female 45-day old Swiss mice housed individually and
       randomised between the different groups receive, on well
       delineated skin areas previously clipped (3 days before),
       0.05 ml of acetone solution of the studied substance.
       The treatment is repeated on alternate days up to a total
       of 3 skin applications (more applications may be necessary),
       Eight days after the first treatment, the mice are killed
       and the areas of treated skin fixed, sectioned and stained
       for histological examination.  The thickness of the
       epidermis and the number of sebaceous glands present are
       determined in 2 to 4 microscopic fields of each of 6 sec-
       tions that are cut from each skin specimen.  The micro-
       scopic examination is carried out on code-numbered slides
       which carry no details of treatment given.

            For each term or treatment, 25 to 30 mice are being
       used.  To control the model and to evaluate the relative
       activities of the test compounds, adequate positive
       (benzo-a-pyrene and forbol ester-tpa) and negative
       (solvent) control groups will be provided for each assay.
3.2.3  in vivo tests - possible extraction of mutagens from
       particulate matter in the whole animal (bioavailability)
       (University of Naples, Department of Biochemistry, Italy)

            An important aspect of the CCMC short term project
       is the bioavailability of any mutagen or possible car-
       cinogen and although inhalation is the preferred test
       method, due to the long term nature of this process,
       additional short term tests are being conducted with
       animals.

            These tests consist of assimilating the sample
       (diesel particulates) into an appropriate medium, such
       as corn oil, bovine serum albumen and human blood serum
       and injecting this into the peritoneum and subsequently
       subjecting the urine to appropriate AMES assays to test
       for mutagenicity.  These tests are being carried out
       using male and female Sprague-Dawley rats.  Three
       animals for each dose are being used.
                           508

-------
             FRACTIONATION AND IDENTIFICATION  OF  ORGANIC  COMPONENTS  IN
                            DIESEL EXHAUST PARTICIPATE

 Mitchell  D.  Erickson,  David L.  Newton,  Michael C.  Saylor,  Kenneth B. Tomer,
                               and E.  D.  Pellizzari
                            Research Triangle  Institute
                                  P.  0.  Box 12194
                         Research  Triangle Park,  NC   27709

                      Roy B.  Zweidinger  and Sylvestre Tejada
                      Mobile Source Emissions  Research Branch
                          Environmental  Protection Agency
                            Research Triangle  Park,  NC

      Diesel  exhaust  particulate,  generated using production model passenger
 car engines  on  a  chassis dynamometer, was extracted from Teflon-coated glass
 fiber filters with methylene chloride and fractionated using either a solvent
 partition  scheme  partition scheme or low pressure liquid chromatography
 (LPLC).   The solvent partition  schemed)  generated  two acid, two base, a
 cyclohexane  insoluble,  polar neutral, non-polar  neutral, and PNA fractions.
 The LPLC  scheme generated fractions  which elute  from a silica gel (Lobar )
 column with  10% CH2C12/90% hexane (Fraction Fl and  part  of F2), 50% CH2C12/
 50% hexane (part  of  F2  and F2A),  100% CH2C12  (F3 and F4),  10% CH3OH/90%
 CH2C12 (F5 and  F6),  50%  CH,OH/50% CH?C12  (F7  and F8) and 100% CH3OH (F9 and
 F10).  A  hexane-insoluble (HI)  fraction  was analyzed using the same prepara-
 tive  LC technique.   The  fractions generated by both  schemes were analyzed by
 normal phase HPLC, glass capillary GC/MS/DS (electron impact [El], chemical
 ionization,  and negative ion  chemical ionization [NICI]), direct probe
 NICIMS, direct probe El  high  resolution  MS, FTIR and (GC)2/FTIR.

      Including isomers,  52  polycylic aromatic hydrocarbons and alky! deriva-
 tives, 35 PNA-ketone and di-ketone  derivatives,  20  aromatic aldehydes and
 cyclic anhydrides, 26 nitrogen-containing  PNAs (including nitro PNAs) and 30
 other compounds (including  alkanes  and  some possible background contaminants)
 were  identified.

      The compounds of particular  interest  are the various PNA ketones (e.g.,
 fluorenone) and the  nitro  PNAs(e.g., nitropyrene) and their alkyl-substituted
 homologs.   These compounds are  listed in  Tables  1 and 2.   The nitro PNAs are
of interest in light of  recent  findings  that  some of them are highly mutagen-
 ic.(2-4)   The nitro PNAs were found in the more  mutagenic fractions.  Thus,
 it may be speculated that  these compounds  are contributing much or possibly
most of the mutagenic activity  to  these fractions and therefore to the
diesel exhaust particulate  itself.
                                    509

-------
REFERENCES

1.   Hughes, T.J., L.W. Little, E.G. PelUzzari, C.M. Sparacino, G. McCue,
       L. Claxton, and M.  Waters.  Mutation Res.. 76., 51-83 (1980).
2.   Schuetzle, D., J.S.-C.  Lee, T.J. Prater, and S.B. Tejada, Int. J.
       Environ. Anal.  Chem., 9, 93-144 (1981).
3.   Lofroth, G., E.  Hefner, T. Alfheim, M. Mtfller, Science, 209. 1037-1039
       (1980).
4.   Rosenkranz, H.S., E.G.  McCoy, D.R.  Sanders, M. Butler, O.K. Kiriazides,
       R. Mermelsteim, Science, 209, 1039-1043 (1980).
                                    510

-------
            T»ble 1.   POHCTC1IC KKTONES AND DIONES  IIOTIFIED  M NISSAS DIESEL EXHAUST PARTICULAR
Identified Compound
naphtlwquinone
9-fluorenone or C1,H-01'
isoaer
•ethyl fluorenone isoawr
or C,,H,-0 isoaer
antnrone or ph«nanthrone Isoaer
C,-alkyl fluorenone Isooers or
C,*al kyl-f 1 uorenone iso«rs
Or C)SHU0 iso««rs (tent)
C.-alkyl fluorenone iso»ers
3r C1?H160 isoaerj (tent)
xanthont (tent)
antnraguinone
4H-cyclopenta(def)ph«rr
anthrene-4-one (t«nt)
oenzanthrone isoners
NuoCer of
Isoaen
Identified
1
1
3
1
4
4
2
1
1
1
3
nettiyl-lH-cyclopenta(def)- 2
phenanthren«*4-on« isoo«r (tent)
benzof 1 uorenon* 1so«*rs (tent)
C18H120 xetone isoners (tent)
4
C^H^O, dion« isoaer (tent)
6H-benzo(cd)pyrenone isoaers
2
2
1
3

GC/EIMS
x
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Analysis
GC/NICIMS HRMS GC/FTIR Other ^^IdentUled9
F4
x F4;F3;F2
x F4;F3;F2
F2
F3;F2
x F3;F2
x H1;F2
F3
?3 F3;F2
F3;F2
F3;F2
F2
F2
x F2
x F2
x F2
 or C19H1Q0 isoMr (Unt)

C,-alky1-4H-cyclopenU
 (def)phenanthren-4-one
 isomer (tent)

C,-alkyl fluorenone isoner
    C18H18° iso**r 
-------
            Table 2.  NITROGEN CONTAINING AROHATICS IDENTIFIED IN NISSAN DIESEL EXHAUST PARTICULAR
Identified Compound
N-pheny 1 naphthyl ami ne
isomer
C.-alkyl-N-phenylnaphthyl-
9*1 ne isoner (tent)
benzo(c)cinnol1ne
methyl benzo(c)c1 nnol 1ne2
isoners
C13H90 Isoner3 (tent)
nitroanthracene isoner or
Number of
Isonrs
Identified
1
1
1
3
1
1

SC/EIHS
X
X
X
X
X
X
Analysis
GC/NICIK HUMS GC/FT1R Other Fr'^lju
x F4;F2;F5
F5
F5-.HI
F5
F6
x X F1G;F2
(s) Containing5
nd Identified
;F6;F3;F8;F1G





mthylnltroanthracene or
 nethylnitrophenanthrene
 isomers (tent)

C.-alkyl nltroanthracene or
 C.-alkyl nltrophenanthrene
 iSoaers (tent)

C,-alkyl nltroanthraeene or
 C,-alky] nltrophenanthrene
 iSoners (tent)

nltropyrene isaner or

 C16H9N02 1soMr

methylnitropyrene isoenr or
 nltrobenzof1uorene Isooer (tent)
C18HllN02 is
F1G;F2



FIG



FIG



F2


F1G;F2


F2
i S«« text for fraction identifications.
  It is possible that these are polycyclic ketones  of the foneulas  C..H.O and  C-.H^O.   However, their mass
  spectra more closely resembled those for benzo(c)c1nno!1ne in standard spectra*  These coopounds were
  also later eluting than 9-fluorenone and its alkyl  hoaologs.   Further elucidation of  these compounds
  is currently underway for fraction F5 of the refractionated HI saople by Mans  of GC/FTIR and HRHS to
, determine whether these are indeed benzo(c)cinnol1nes.
  This eluant gave a mass spectrum similar to that  of acridlne or benzoquinoline, but only a trace
. quantity of the compound was present.
, Possible iso«wrs include nitrochrysene,  nltronaphthacene,  and nltrotrlphenylene isoaers.
  Fraction Fl was further fractionated to  yield subfractlons F1A through FIG.
                                                       512

-------
      EFFECT  OF  CHRONIC DIESEL EXPOSURE OF PULMONARY PROTEIN SYNTHESIS
                                   IN RATS
                                      by


                   R.  G.  Farrer,  Sukla Dutta and S.  Dutta
                  Wayne State University School  of Medicine
                          Department of Pharmacology
                            Detroit, Michigan 48201


     There  is  evidence that when rats are subjected to  acute  exposure  to  cig-
arette  smoke,  hepatic protein synthesis is inhibited and  the  extent  of inhi-
bition  is positively  correlated  with the dosage of smoke.   The  present study
has been undertaken to determine the effect of  diesel  smoke on  pulmonary  pro-
tein synthesis.   For  this study, male Fischer rats  have been  exposed to
diesel  exhaust (6 mg/m^)  for 2,  4 and 8 weeks.   At  the  end  of the  desired ex-
posure  periods,  the lungs have been removed respectively  from the  exposed and
time-matched control  rats and placed in an isolated lung  apparatus.

     The apparatus was devised by modifying the design  of Fisher et  aj_.,  (2)
for the perfusion of  lung excised from rats.  Briefly,  for  perfusTon of each
lung, the animal  was  anesthetized with pentobarbital  intraperitoneally and
trachea was cannulated and connected to a Harvard respirator.   At  this point,
by means of a  pressure transducer placed in the in-flow route of the respira-
tor, we recorded  in situ  ventilation pressure for 3-4 minutes in an  eight
channel recorder.  This  allowed  us to compare in vivo  tracheo-bronchiolar air
resistance under  inspiratory pressure of 10 aiTR^O  with that  of the  resist-
ance when the  lungs would be under ex vivo condition.   Following this  proce-
dure, thoracotomy was conducted  and~the lungs were  excised  by carefully sepa-
rating  pulmonary  artery  from the aorta.   Once the pulmonary artery was clear-
ly dissected out  from the other  mediastinal  structures, heparin (0.5 units/g)
was injected through  this.   After a few minutes of  circulation  of  heparin,
the pulmonary  artery  was  separated from the right ventricle and the  open  end
was cannulated for delivery of perfusate at 15  ml/min by  means  of  a  Harvard
peristaltic pump  from the reservior.   While perfusate was going through the
lungs, a small incision was made in the left artrium so that  the perfusate
might flow  freely and wash blood out of the pulmonary  vascular bed.   When
the lungs were cleared of blood,  they were transferred  to a water-jacketed
perfusion chamber  maintained at  35°C.   During the transfer, we  kept  respiring
the isolated lung  while  interrupting the perfusate  only for a few  seconds.
The peristaltic pump  was  then  switched on and upon  perfusion  of the  isolated
lungs the perfusate freely  drained into  the perfusion chamber and  from there
by means of Tygon  tubing  back  to the reservoir.   Thus,  a  fixed  volume  of
                                     513

-------
perfusate continuously recirculated  through  the lungs for the entire duration
of the experiment.  Moment  to moment  performance of the preparation was
monitored by recording continuously- pH  and  pC>2 of the perfusate before it
entered the lungs and the pressures required  for tidal  ventilation and for
delivery of perfusate at 15 ml/min, through the pulmonary artery.

     One exposed and one control  Fischer  rat  was tested in one day, and two
such runs were conducted each week.   It required approximately two hours to
complete one perfusion experiment which consisted of 30 minutes of equilibra-
tion followed by one hour of ^H  leucine incorporation.   During the equilibra-
tion period, the lungs were perfused  by recirculation with  50 ml  Leibovitz
solution and watched for any change in  pH,  p02, ventilation and perfusion
pressures.  If all these parameters looked  normal,  which was the case  for all
the 24 experiments needed to complete the series,  the perfusion medium was
changed to fresh 50 ml Liebovitz  solution containing approximately 1 yCi/y
mol/ml of 3H-leucine.  Thus, the  perfusion  technique made it possible  to
provide a constant supply of substrate, ^H  leucine,  with a desired specific
activity all throughout the incorporation period.

     At the end of ^H leucine incorporation period,  samples were taken for
autoradiography and, then,  the rest of  the  lungs were subjected to homogeni-
zation in 0.02 KH P04 and TCA precipitation of protein  which was washed'with
polar and non-polar solvents.  The washed protein  residue was dissolved in
4.0 ml  of 2.0 N NaOH at 50°C.  The protein was determined by the  method of
Sedmak and Grossberg (3).  Total  pulmonary  DMA was  determined by using the
modified diphenylamine technique  of Burton  (4)  following precipitation of
ONA.  Radioactivity obtained from incorporated 3H-leucine was measured in 1.0
ml samples of the dissolved protein using a Beckman  LS-100 counter.  Because
there were differences in quenching between the diesel  exposed (very dark,due
to presence  of diesel particles) and control  samples,  internal  standard (3H
tolune) was used to correct the observed CPM  before  expressing the results in
DPM which was converted to nmole  ^H-leucine by  using the known specific acti-
vity.  Results were normalized on the bases of mg  protein as well  as mg DNA
as obtained per gram of lung tissue.  The results  of -H leucine incorporation
as shown in table 1 revealed that 8 weeks of  exposure of male Fisher rats to
6.0 mg particulates/m3 of diesel  engine exhaust had  no  significant effect on
the lungs to incorporate 3H-leucine into the  TCA insoluble protein.  Similar
results were obtained after 2 and 4 weeks of  exposure to diesel  exhaust in
comparison to respective time matched controls.  Also,  electron microscopic
autoradiographic grain counts as  obtained from  the  diesel  exposed  vs.  control
rats showed no particular difference  among  various  groups.

Table 1.  Incorporation of 3H-leucine by the  perfused lungs as obtained from
          rats after 3 weeks of exoosure and  their  time matched controls.
Experiment
8 weeks of air
exposure
8 weeks of diesel
* Means + S.E.
nmoles leucine
oer mg protein
3.53 + 0.23*
3.30 + 0.48
nmoles leucine
per ma DNA
41.3+3*
46.3-i 6
Protein/ONA
11.8 + 0.4*
13.9 + 1.1
                                     514

-------
                                 REFERENCES
1.    Garrett, R.J.B., and M.A. Jackson  1979.   Effect of acute smoke
      exposure on hepatic protein synthesis.  J. Pharm. Expt.  Therap.
      209:   215-218.

2.    Fischer, A.B., C. Dodia and J. Linadk  1980..   Perfusate composition
      and edema formation in isolated rat lungs.  Expt. Lung Res.  1:  13-21

3.    Sedmark, J.J. and S.E. Grossberg  1977.   A rapid, sensitive  and
      versatile assay of protein using Coomassie Brilliant Blue G250.
      Anal. Biochem. 79:  544-522.

4.    Burton, K.  1955,   The relation between the synthesis of deoxyribo-
      nucleic acid and the synthesis of protein in the multiplication of
      bacteriophage T2-  Biochem. J. 61: 473-483.
                                      515

-------
    THE EFFECT OF EXPOSURE TO DIESEL EXHAUST ON PULMONARY PROTEIN SYNTHESIS

                                      by

                  C. Fillpowitz, C. Navarro, and R. McCaulev
                          Department of Pharmacology
                   Wayne State University School of Medicine
                               Detroit, Michigan


     Previous work performed 1n collaboration with the B1omed1cal Research
Division of General Motors Corporation had Indicated that exposure of rats to
diluted diesel exhaust for periods of up to one year did not induce the
activity of mlcrosomal benzo[a]pyrene-ox1d1z1ng enzymes in lung tissue.
Several explanations for this observation, Including the possibility that
exposed animals are unable to respond to inducing agents, have been suggested.
In this report, we will discuss the ability of animals which have been exposed
to exhaust in the concentration of 6 mg/m3 of diesel particles to synthesize
pulmonary proteins as judged by in vivo 3H-leucine incorporation and to respond
to 3-methylcholanthrene by the induction of pulmonary oxidative metabolism of
benzo[a]pyrene.

     (This research was supported by a grant from General Motors Corporation,
Warren, MI.)
                                      516

-------
                 THE RAPID ANALYSIS OF DIESEL EMISSIONS USING
               THE TAGA 6000 TRIPLE QUADRUPOLE MASS SPECTROMETER

                                      by

                    J.E. Fulford, T. Sakuma, and D.A. Lane
                                  SCIEX, Inc.
                          Thornhill, Ontario, Canada


     Because of the increasing consumption of diesel fuels by cars, trucks,
heavy equipment, and industry, the environmental scientist is concerned with
the atmospheric loading of toxic combustion products and their detrimental
biological effects.

     The conventional analysis of diesel fuel combustion products is very
time-consuming and difficult, since the quantity of toxic pollutants such as
polycyclic aromatic hydrocarbons (PAH) and their nitro derivatives is low,  and
since they are often associated with other contaminants (for example, unburned
diesel  fuel).  The analysis entails sample trapping, extraction, substantial
clean-up, and determination by capillary column gas chromatography combined
with mass spectrometry.

     Using MS/MS, the particulate extract can be directly analyzed for
nitropolycylic aromatic hydrocarbons without sample clean-up.  The sample is
deposited unto a temperature program direct insertion probe operated at
atmospheric pressure, and is desorbed over a period of twenty minutes.  In  the
negative mode, parent ions which yield a daughter ion of m/z 46 [N02~] can  be
scanned, or target compounds can be quantitated by integrating the desorption
curve or the m/z 46 daughter ion in the multiple ion monitoring mode.
Calibration plots for nitropyrene in spiked samples of diesel particulate
extract are linear (r = 0.99), and the extrapolated detection limit for
nitropyrene in the diesel extract is in the ppb range.

     Since the vapor phase emission of diesel engines are also of interest, the
exhaust has been sampled directly by the TAGA 6000 MS/MS.  In this study,
emissions of a diesel-powered vehicle were transported through a heated (150°C)
Teflon  pipe at 2 L/sec, and a small portion of this flow was admitted directly
into the atmospheric pressure ion source.  The mass spectrometric analysis  was
based on:

       (i)  the identification of particular compounds in the exhaust gas;
      (ii)  the rapid screening of the exhaust gas for nitro compounds;
     (iii)  the analysis of the exhaust gas for specific nitro-PAH.
                                      517

-------
            PREPARATION OF DIESEL EXHAUST PARTICLES AND EXTRACTS
                        AS SUSPENSIONS FOR BIOASSAY
                                     by
                                Jean L. Graf
                           IIT Research Institute
                      Fine Particles Research Section
                            10 West 35th Street
                           Chicago, Illinois 60616


     A bioassay program is being conducted at IITRI to evaluate the acute
toxicities and carcinogenic potentials of diesel engine exhaust components,
cigarette smoke condensates and organic solvent extracts of roofing tar
volatiles and coke oven emissions.  The test materials were administered to
hamsters by the intratracheal route.  Administrations in both the acute
toxicity and carcinogenic potential bioassay experiments have been completed.

     The test materials were supplied through the U.S. EPA Biomedical Research
Branch.  The diesel engine exhaust components supplied were a whole particle
exhaust consisting of carbonaceous soot with adsorbed liquid and gaseous
species, and a dichloromethane extract of the whole particle exhaust.  The
cigarette smoke condensate was supplied as a concentrated solution in acetone.
Both the roofing tar and coke oven emission extracts were supplied as
dichloromethane solutions.

     The intratracheal administration route required preparation of stable
suspensions and emulsions of the test materials in fluids compatible with the
hamster respiratory tract fluids.  Saline was the obvious suspending fluid to
be used  but additional ingredients were required to enable suspension of the
particles and the solvent-free extracts.  Examination of the as-received
whole particle exhaust revealed that very large (up to 150 ym) diesel parti-
cle aggregates were present.  These large particles hindered suspension
preparation and were not suitable for intratracheal suspension.  Therefore,
a research program was conducted to develop methods for preparing saline
suspensions of the whole particle exhaust in particle size ranges amenable
to intratracheal instillation and saline emulsions (liquid-liquid suspensions)
of the various extracts.

     For the short-term acute toxicity studies, a simple wet ball milling
                                     518

-------
method was developed which both reduced the size of the diesel engine exhaust
particles and suspended them in saline.  The hydrophilic nature of the diesel
particles required that they first be wetted with propylene glycol in the
glass milling jar before the saline suspending fluid and the glass milling
beads were added.  Gelatin was also added to the saline to serve as a pro-
tective colloid.

     Simple hand emulsifying techniques were used to prepare the diesel  en-
gine exhaust extracts and the other extracts condensate as stable emulsions
in the gelatin-saline fluid for the short-term acute toxicity studies.  Stan-
dard glass tissue grinders proved to adequately emulsify the four types  of
organic extracts, once the solvent had been removed and they were wetted with
propylene glycol and a surface active agent.  To maintain the emulsion sta-
bilities for more than 30 minutes, gum arabic was also added to the saline
to provide a stronger protective colloid.

     The greater materials requirements for the larger scale carcinogenic
potential bioassay experiments required development of a semi-mechanized
method to prepare the emulsions of the test materials.  Trial emulsion pre-
parations with a Polytron  tissue homogenizer proved successful and a unit
was purchased.  The Polytron  js a high speed mixing device which employs
both mechanical, shear action and ultrasonics to accomplish homogenization
of liquid samples.  Various types of homogenizing probes are available and
provide a wide range of shear and ultrasonic energies.  One probe design
provides sufficient shear energy to reduce particle sizes of soft solid  mater-
ials such as the diesel engine exhaust particles.

     Thus, for the long-term carcinogenicity bioassay experiments, development
of protocols to prepare the suspensions and emulsions with the Polytron   were
conducted.  The whole particle diesel engine exhaust suspensions were easily
prepared as stable suspensions in gelatin-saline, once the particles had been
wetted with propylene glycol.  Particle concentrations as high as 25 mg/ml
were attainable.  The primary advantages of the Polytron  milling over ball
milling to prepare the particle suspensions were shorter preparation times
(one hour versus 10 days elapsed time), elimination of glass milling con-
taminants, and reduction of the reagglomeration tendency after milling was
completed.  The diesel engine exhaust extract was also easily prepared as an
emulsion in gelatin-saline once the extract had been wetted with pro^ylene
glycol (after solvent removal).  The use of the high energy Polytron  to
emulsify the diesel extract as well as the other organic extracts in saline
eliminated the necessity of adding a surface active agent.  However, gum
arabic was still required as an additional protective colloid to maintain
emulsion stability.

     Assay and characterization methods for the prepared suspensions were
also developed.  While the emulsions could be characterized microscopically,
no practical assay methods could be developed.
                                      519

-------
           COMPOUNDS IN CITY AIR  COMPETE  WITH 3H-2,3,7,8-TETRACHLORO-
                  DIBENZO-p-DIOXIN FOR BINDING TO THE  RECEPTOR

           J.-A. Gustafsson, R. Toftgard,  J.  Carlstedt-Duke, G. L3froth
                  Dept. of Medical Nutrition  and Pharmacology,
             Karolinska institute, S-104  01 Stockholm, and Dept. of
            Radiobiology, University  of Stockholm, S-106 91 Stockholm


     It is well known  that filter collected urban particulate matter contain
compounds which are mutagenic  in  the  Ames' Salmonella  assay in the absence  of
rat liver metabolic activation, showing that  these compounds are different
from conventional polycyclic aromatic hydrocarbons (PAH).  In the present
study, we have shown that such particulate matter also contains compounds with
an affinity for the rat liver  receptor protein which specifically binds
2,3,7,8-tetrachlorodibenzo-p-dioxin  (TCDD). The content  of conventional PAH
cannot account for the degree  of  binding.
     Urban particulate matter  used in the  investigation  was collected on glass
fiber filters by high  volume sampling at  roof top levels in central Stockholm
and a suburban site. Collection,  Soxhlet  extraction  with acetone, extract
preparation in dimethyl sulfoxide (DMSO)  and  mutagenicity  testing of the ex-
tracts in the Salmonella assay have been  described elsewhere. The affinity  of
the particulate fraction of the air samples to the rat liver receptor protein
was measured by competition for the binding of  H-TCDD,  as described below.
     Liver cytosol was prepared and diluted to 3.5 A---,  -.-/ml 2-2.5 mg
protein/ml). Each experiment consisted of  a series of  ten  incubations of one
ml of cytosol with 1.5 nM  H-TCDD. Two experiments were  carried out in the
presence of 20 pi of DMSO (total  binding). Two incubations contained 150 nM
radioinert 2,3,7,8-tetrachlorodibenzofuran (TCDBF) (non-specific binding).
The remaining six incubations  contained varying amounts  of the extract of the
particulate fraction of an air sample.  2.5-20 yl of  the  extracts, dissolved
in DMSO, was added to  the incubation  mixture  and the volume of DMSO was then
adjusted to 20 yl/incubation.  After incubation at 0°C  for  60 min, the incuba-
tions were treated with dextran-coated charcoal and  the  amount of  H-TCDD
bound to the receptor  in each  of  the  ten  incubations was measured by iso-
electric focusing carried out  in  nolyacrylamide gel  as described previously.
     The non-specific  binding  of  H-TCDD was  obtained  from the two incubations
in the presence of 150 nM radioinert  TCDBF and was subtracted from the total
binding in each of  the other  incubations  (»  specific  binding). The specific
binding of  H-TCDD in  the presence of extracts of the  air  samples was express-
ed as a percentage of  the specific binding in the control  incubation ( H-TCDD
+ DMSO). The relative  binding  affinity for the individual  air samples was cal-
culated from log-logit plots of the competition for  TCDD-binding, where logit
6 = ln(6/l-6). The log-logit plots were calculated using linear regression
analysis. Each extract was analysed 3-4 times at a suitable dilution.
     An unused air filter (blank  filter) was  extracted in  the same manner as
                                      520

-------
the other samples and  tested  for  competition for receptor binding.  The  results
were expressed as an equivalent volume  of  air.  Benzo(a)pyrene  and  TCDBF were
dissolved in dioxane and  the  competition for binding to  the  receptor measured.
     The results are shown  in Table  1,  in  which are  given the  mutagenic effects
in Salmonella TA 98 and TA  100 in the  absence of mammalian metabolic activation
and the relative binding  affinities  expressed as the concentration of air sample
extract that competes  for 50% of  the TCDD-binding to the receptor  (ED,n; m
air/ml cytosol). Log-logit  plots  for the different air samples indicated
that they competed for the  same binding site. There  seemed to  be a gross
correlation between the binding affinity and the pollution level as measured
by mutagenic effects.  Samples collected in the  summer had higher ED  ~s than
samples collected in the  winter and  spring.
     Several compounds are  known  to  bind to  the receptor protein including many
PAH present in urban particulate  matter. Some of the particulate samples were
analysed for PAH. Assuming  that all  PAH present at concentrations  above 0.1
ng/m  have the same binding affinity for the receptor protein  as benzo(a)pyrene
(B(a)Pl, it can be calculated that the  ED  Q  value should be  1.7, 15.3 and
0.07 m  of air for sample 148, 173 and  S-253, respectively.  The  observed
values were 0.035, 0.14 and 0.015 m  ,  i.e. known PAH may account for about
2.4, 0.8 and 22.4% of  the binding, indicating that other types of  compounds
are of major importance.
     TCDD and TCDBF have  the  highest binding affinities  for  the  receptor among
investigated compounds. Two samples  have been analysed for TCDD  and TCDBF, and
the concentrations were below the detection  levels of 2  pg/m . Consequently,
these compounds cannot account for the  binding.
     The affinity of a compound for  the TCDD receptor is well  correlated to the
magnitude of aryl hydrocarbon hydrorylase  (AHH) induction caused by that
compound. Strains of mice with high  AHH inducibility are more  susceptible to
pulmonary cancers caused  by 3-methylcholanthrene than strains  with low induci-
bility, indicating a link between AHH  activity and appearance  of pulmonary
tumors. The rat lung has  a  high content of receptor  protein, and the capability
of the human lung to metabolise B(a)P  indicates the  presence of  the receptor
protein also in this tissue.  Although  TCDD has  been  shown to be  a potent
carcinogen in chronic  feeding studies,  it  is apparent that TCDD  shows no or
very little mutagenic  activity in rn vitro bacterial test systems  such  as
the Ames' test and a very low covalent  binding to DNA in_ vivo. In a recent
study, however, it was shown that TCDD  is  a  potent promoting agent for
hepatocarcinogenesis initiated by diethylnitrosamine.  TCDD and possibly also
other compounds with an affinity  for the same receptor may thus  better  be
described as cocarcinogens  and tumor promoters  rather than carcinogens. The
presence of compounds  with  affinity  for the  receptor in  urban  particulate
matter may be of importance with  regard to the  health implications of urban
air pollution. This type  of compounds may  or may not be  similar  to the
components that are mutagenic in  the absence of mammalian metabolic activation
in the Ames' Salmonella assay.

Conclusions
     Acetone extracts  of  filter-collected  ur^an atmospheric  particulate
matter contain compounds  which can displace   H-2,3,7,8-tetrachlorodibenzo-
p-dioxin from the rat  liver receptor protein. The concentration of conven-
tional polycyclic aromatic  hydrocarbons or chlorinated dioxins and dibenzo-
furans cannot account  for more than  0.8-22%  of  the displacement.
                                       521

-------
Table 1. Competition for TCDD receptor-binding and mutagenicity in Salmonella
TA 98 and TA 100 in the absence  of mammalian metabolic activation by extracts
of filter-collected urban particulate matter. Samples were collected at roof
top levels for 24 h starting about 6 a.m.  Samples 148 and 174 were collected
at a suburban site 22 km NNW Stockholm  and the others in central Stockholm.
The ED-0 is the concentration of competitor that competes for 50% of the spe-
cific Binding of  H-TCDD. The ED5Q~s  (in nM)  for 2,3,7,8-tetrachlorodibenzo-
furan (TCDBF), benzo(a)pyrene (B\a)P),  3-naphthoflavone (BNF), 3-methylchol-
anthrene (3-MC) and benz(a)anthracene  (BA) are given for comparison. The muta-
genic response is highest in the absence of mammalian metabolic activation and
decreases by addition of S9 from rat liver; the decrease is dependent on the
amount of S9 added. Samples 258,  264 and 268 were tested for mutagenicity both
prior to (Feb. 1980) and after  (Sept. 1980) the completion of the TCDD recep-
tor analyses; there were no detectable  changes in the mutagenic response.
Sample
149
148
173
174
S-258
T-262
S-264
T-268
Blank
filter
TCDBF
B(a)P
BNF
3-MC
BA
Receptor affinity3
(m air /ml cytosol)
0.049 -
0.035 -
0.137 -
0.302 -
0.015 -
n.d.
0.039 i
0.049 -
2.18 -
2.69 -
18.21 -
7.0 nM
2.9 nM
3.8 nM
0.014
0.017
0.073
0.147
0.007

0.010
0.026
0.640
1.91 nM
9.88 nM



Mutagenic response Sampling date
(revertants/m3) md site
TA 98 TA 100
f
19 11 79 04 10 inner city
14 8 79 04 10 suburban
3 n.d. 79 07 05 inner city
1 n.d. 79 07 05 suburban
59 79 80 02 04 inner city
92 102 80 02 05 inner city
30 20 80 02 06 inner city
9 7 80 02 07 inner city
<0.2 <0.4





 values represent the means -  standard  deviation (three to four determina-
 tions) .

 Blank filter extracted with the  same volume  of  acetone.  The competition is
 expressed as an equivalent volume  of air.
n.d. Not determined.
                                      522

-------
     GC/MS AND MS/MS  STUDIES OF DIRECT-ACTING MUTAGENS IN DIESEL EMISSIONS

                                       by

            T. R. Henderson, J. D. Sun, R. E. Royer and C. R. Clark
               Lovelace Inhalation Toxicology Research Institute
                            Albuquerque, New Mexico

                          T. M. Harvey and D. F. Hunt
                            Department of Chemistry
                            University  of Virginia
                           Charlottesville,  Virginia

                J. E. Fulford, A. M. Lovett, and W. R. Davidson
                                  Sciex, Inc.
                                Toronto, Canada


    While the direct-acting mutagens  in diesel  emissions  have not been un-
equivocally identified, evidence  has accumulated showing  that nitro-PAHs
 [polycyclic aromatic  hydrocarbons) may  be a  major source  of mutagenic activ-
ity (1,2).  We have found that diesel fuel-PAHs are devoid of mutagenic
activity in Salmonella test strains.  After  reaction with N02, the nitro-
fuel-PAH mixtures are direct-acting, frameshift mutagens  (200-1000 revert-
ants/ug) in TA98, but were not active  in TA1535, a base-pair substitution
indicator strain  (2).  The cytotoxicity to CHO  cells also increases after
reaction with N02(2).  These biological responses are very similar to
those observed with diesel soot extract fractions although the activity of
diesel extracts is much lower.

    The fuel-nitro-PAH mixtures have been found useful as a positive control
and as a reference mixture for MS/MS (triple stage quadrapole mass spectrom-
etry) analyses.  Although the fuel-nitro-PAHs are not resolvable by capil-
lary GC/MS, the fuel-PAHs before  reaction with  N0£ are readily separated
and identified by GC/MS.  Since the fuel-PAHs extracted with DMSO (Me2SO)
are readily identified by GC/MS,  correlations can be made with the nitro-PAH
masses observed in MS/MS spectra  of nitro-PAH mixtures.  These are used for
MS/MS interpretation  since MS/MS  does not differentiate between isomers of
the same molecular weight in most cases.  When  possible, these interpreta-
tions should be confirmed by independent methods (1).

    MS/MS was carried out to compare the relative intensities of ions in
samples of widely different mutagenic potencies.  In this way, it may be
possible to estimate which of the nitro-PAHs detected in diesel  soot ex-
tracts (1)  may make major contributions to the total  mutagenic activity.
                                    523

-------
    The MS/MS analyses Involved:  R-NOg* R-N02-H* (lonization under
isobutane chemical 1on1zat1on conditions) + R-N+0 (collislonally Induced
dissociation by collision with N2 1n quadrupole #2).  The MS/MS Instrument
was a modified F1nn1gan 3200, which has been described prev1ously(3).  The
first quadrupole was scanned from 80 to 350 amu with a scan time of
1 .3 sec.  The second quadrupole was operated (with RF voltage only) as a
collision chamber.  The third quadrupole was scanned the same as #1, but
17 amu behind 1t.  In this way, only Ions which lost 17 amu 1n passing
through the collision chamber were detected.  The Instrument was tuned with
l-n1tropyrene (parent-H+ 1on m/z 248) and the N£ pressure adjusted for
maximum m/z 231  daughter 1on (m - 17).  Extract samples (40 ug) were vola-
tilized into the source using a thermal desorber (temperature programmed
from 50 to 350°  C 1n 10 minutes).
    MS/MS analyses were done on selected samples with an APCI/MS/MS (triple
quadrupole MS/MS with an atmospheric pressure chemical 1on1zation source)  at
the laboratories of Sdex, Inc., Toronto, Canada.  The reaction monitored
was:
                        -
    R-N02 + 02 — * R-N02  -* R + N02


The first quadrupole was scanned over a 100 to 300 amu range or single ion
monitoring was done at m/z 247".  The third quadrupole was set for con-
stant daughter scans at m/z 46".  The corona discharge current was
6 uamp constant current at 4 to 6 kV and an ion energy of 65 eV.   Thus 1n
this reaction, only negative Ions of 100 to 300 molecular weight were de-
tected which yielded N02 on collision with argon gas.

    Diesel soot extracts were fractionated with DMSO to yield aliphatic,
aromatic and polar fractions for MS/MS analyses.  This fractionation method
has been described previously (4-6).  The aromatic fraction from a single-
cylinder Swan diesel engine was used for comparison with an Oldsmobile ex-
haust aromatic fraction because of differences in mutagenic activity.  The
aromatic fraction recovered by DMSO fractionation of diesel soot  extracts
typically contains 50-80% of the direct mutagenic activity of different ex-
tracts and is concentrated 5- to 10-fold in specific activity.  The aromatic
fraction of diesel exhaust is very similar to the Y fraction isolated by
Schuetzle .et .§J_.  (1) 1n containing mononitro-PAHs and dinltro PAHs of 2 ring
PAHs.

    The mutagenic activities of the extracts in TA98 (no S-9) under standard
Ames bioassay conditions (7) were:  n1tro-fuel-PAHs--435 rev/vg;  Swan
diesel soot aromatic fraction— 57 rev/no; Olds diesel soot aromatic frac-
t1on«2l  revertants/vg.  These samples, differing by about 20-fold in
mutagenic specific activity, were compared by MS/MS.  The unfractionated
extracts were low in the intensity of certain (M - 27) Ions, the  polar frac-
tions contained only a minor part of the total mutagenicity (<  20%) and
the aliphatic fractions contained very low intensities of n1tro-PAH ions.

    Figure 1  shows that the even-mass ions (dinitro-PAHs) appeared to cor-
relate with increased mutagenic activity of the three extracts, while odd-
mass ions (mononitro-PAHs) negatively correlated with mutagenic activity
                                   524

-------
with the exception of m/z 171.  The even mass ions of particular signifi-
cance and their tentative identification were:  m/z 252, dinitrophenan-
threnes; m/z 256, dinitromethylbiphenyls; m/z 230, 244 and 258, dinitro-
naphthalenes containing 2, 3  and 4-methyl groups.  These compounds might be
particularly important in the total mutagenicity of diesel soot extracts.
Nitropyrenes (m/z 231) did not significantly correlate with differences in
the mutagenicity of these samples, although these compounds may account for
5-10% of the total mutagenicity(4).

    Figure 1 also shows that  the same nitro-PAH (M - 17) masses were present
in all three types of samples.  The fuel-nitro-PAHs, having been treated
with excess N02, appeared to  be lower in mono-nitro-PAH type masses and
higher in the even masses (dinitro-PAHs) discussed above.  Since most of the
nitro-PAH ions present in diesel exhaust are also present in fuel  aromatic
fractions treated with N02 and these masses are relatable to the PAHs
present in the fuel burned by the two engine types, this suggests that the
fuel PAHs contribute to the exhaust nitro-PAHS.  One possible mechanism is
reaction of unburned fuel PAHs with NC^.

    Studies involving the addition of pyrene to the fuel for the single
cylinder diesel engine was performed to further test the hypothesis that
fuel PAHs may contribute to the formation of nitro-PAHs in exhaust soot ex-
tracts.  It was found that addition of 0.01 to 1.0% w/v pyrene to diesel
fuel (less than 0.01% pyrene  by GC/MS) resulted in increased pyrene/phenan-
threne ratios in soot extracts.  With no addition of pyrene, this ratio was
less than 2, but increased to 12 with 1% pyrene addition to the fuel.  The
mutagenicity in TA100 increased 2 to 3 fold by 0.01 to 0.1% pyrene addition,
but decreased somewhat with 1% pyrene addition.  MS/MS analyses of these
extracts by APCI/MS/MS showed increased ion intensities of mono- and
dinitropyrenes in the soot extract for pyrene additions up to 0.1% in the
fuel, but the intensities of  dinitropyrenes decreased at 1% oyrene addition
to the fuel.  (Supported in part by U.S. Department of Energy under DOE
Contract No. DE-AC04-76EV01013.)
                                    525

-------
                                  REFERENCES

1.  Schuetzle, D., T. Riley, T. J. Prater, T. M. Harvey and D. F. Hunt.
      1981.  The identification of nitrated derivatives of PAH in diesel
      participates.  Anal. Chem., in press.

2.  Henderson, T. R., A. P. Li, R. E. Royer and C. R. Clark.   1981.
      Increased cytotoxldty and mutageniclty of diesel fuel  after reaction
      with N02-  Environ. Mutag. 3:  211-220.

3.  Hunt, D. F., J. Shabanowltz and A. B. Giordani.  1980.  Collision
      Activited Decompositions of negative ions in mixture analysis with a
      triple quadrupole mass spectrometer.  Anal. Chem. 52:  386-390.

4.  Henderson, T. R., R. E. Royer and C.  R. Clark.  1981.   MS/MS
      Characterization of Diesel Emissions.  Proceedings of 29th Annual
      Conference of Mass Spectrometery and Allied Topics (in  press-extended
      abstract).

5.  Henderson, T. R., C. R. Clark, R. L.  Hanson and R. E.  Royer.  1980.
      Fractionatlon of environmental  organic extracts with dimethylsulf-
      oxide.  Applications to diesel  exhaust particulates.  Proceedings of
      28th Annual Conference on Mass  Spectrometry and Allied  Topics, p.
      243-244.  (extended abstract).

6.  Henderson, T. R., C. R. Clark, T. C.  Marshall, R. L. Hanson  and C. H.
      Hobbs.  1981.  Heat degradation studies of Solar Heat Transfer
      Fluids.  Solar Energy (1n press).

7.  Ames, B. N.  1979.  Identifying environmental  chemicals causing
      mutations and cancer.  Science  204:  587-593.
                                  526

-------
     10°
     10'
   « I04
   111
     10
           i
           o
 o
 o
 I
1 7 1
        °p°
      a  a
        a A
                          230
                                     i 8
                                     o
                                             0   244
                                 209
            A
            I
           195
                                                                  o

                                                                  a
       ISO
170
190
2IO       230

MASS (M/Z)
250
270
290
                                   FIGURE 1

Figure 1.  MS/MS Comparisons of Nitro-PAHs from Diesel  Fuel  and Exhaust
           Filter Extracts.  —&-- Mitro-fuel-PAHs; —a— Aromatic
           Fraction from Swan Engine Exhaust Soot Extracts;  --0-- Aromatic
           Fraction, Olds Exhaust Soot Extracts.   The masses represent the
           (M - 17) Ions, the parent nitro-PAH being 17 amu  more.   The ion
           intensity is the total  ion counts summed over the entire run,
           usually 300 scans.
                                    527

-------
               RESEARCH  PLANS  FOR DIESEL HEALTH EFFECTS STUDY

                                      by

                      Hironari  Kachi  and Tadao Suzuki
                          First  Research Department
                 Japan  Automobile Research Institute, Inc.
                    Yatabe-cho,  Tsukuba-Gun,  Ibaraki-ken
                                  305  Japan
1) PAST RESEARCHES ON AUTOMOBILE EXHAUST IN JAPAN AUTOMOBILE  RESEARCH
   INSTITUTE, INC. (JARI)

JARI was granted research contracts from the Japan Automobile Manufacturers
Association, Inc. (JAMA) for studying the reduction of automobile exhaust
and their health effects around 1970 when automobile exhaust  were a wide
public concern in Japan on the grounds that they may cause atmospheric pollu-
tions typically exemplified by the phenomenon of photochemical  smogs.

JARI started research on these contracts in 1971 including field surveys
using a mobile smog chamber, tracing of photochemical reactions using a sta-
tionary smog chamber and chemical analysis of emission components.  A behav-
ioral assessments on animals were started in 1975, and basic  researches on
health effects of NOX and 03 were started in 1976.  Preliminary researches on
diesel emissions were started around this time.  Small-scale  animal inhala-
tion system with modified Rochester type inhalation chambers  were designed
and tested.  Studies on the health effects of diesel emissions  to rats were
done preliminary for one month and then three months.  Investigations on
respiratory system revealed that morphological changes in early stage of
exposure are attributed mostly depend to the gaseous components, and it seems
that particulate matters amplify the changes depending on the particulate
concentrations.

The facility for Ames test was completed in 1980.  A preliminary Ames test
was conducted on extracts from diesel particulates using Salmonella Typhimu-
rium TA100 and TA98.  It was found that PAHs contained as neutral fraction
components showed a relatively high mutagenicity.  Examinations of sampling
methods which would allow to obtain artifact-free diesel particulates are
currently conducted comparing results of Ames tests and those of chemical
analysis.
                                    528

-------
2) HEALTH EFFECTS RESEARCH PROGRAM  (HERP)

In recent years, the health effects of  diesel  emissions has become an impor-
tant subject in Japan.  However,  the research works  in this field is still in
an embryonic stage.  Under these circumstance, JAMA referred this subject to
some researchers of automobile engineering, medical and chemical fields   As
a result, HERP has been drafted  in  1981.

An outline  of the HERP is presented below.

     Period : From 1981 to 1985

     Organization (Committee of  HERP)

Chairman  :  Dr. Atsushi WATARI
            President of the JARI
            Prof. Emeritus of Univ.  of Tokyo

Chairman  of Steering Committee  : Dr. Noburu ISHINISHI
                                 Prof,  of  Hygiene & Public Health
                                 Kyushu University

Subcommittees : 1. Diesel Exhaust Generation and Sampling
                2. Analysis and  Custoday
                3. Inhalation Studies
                4. Small Animal  Experiment
                5. Culture Cell  Experiment  (I)
                6. Culture Cell  Experiment  (II)
                7. Mutagenicity  Test using Microorganisms
                8. Miscellaneous including environmental assessment

Secretariat : JARI Officier

     Research items  :

Facilities  for particulate generation,  sampling and analysis and a full-scale
facility  for the inhalation experiment  will be built at JARI.  Preparations
are underway to complete these facilities  in 1982.

It is planned that JARI will participate principally sampling, analysis,
storage and delivery of diesel tars and the inhalation experiment.   The
various in  vitro and in vivo tests  on diesel emission samples will  be con-
ducted in~~some research institutes  and  laboratories of the national and
private universities.

Two types (large and small) of diesel engines will  be used in the project
considering that properties of emission materials might depend on diesel
engines.

Chronic toxicity tests and carcinogenicity tests will be conducted by the
inhalation  for over two years.   Intratracheal  instillations, skin painting
and other tests of the extracts  from diesel particulates will be conducted
                                    529

-------
on small animals.  Mutagehic test of  the  extracts  from diesel  particulates
will be conducted on culture cells and microorganisms (such as the Salmonella
Typhimurium).                                                         "~
                                     530

-------
              NEURODEPRESSANT EFFECTS OF UNCOMBUSTED DIESEL FUEL

                                      BY


                              Robert  J.  Kainz, Sc.D.
                  Environmental  Industrial Safety Consultants
                              Fredrick, Maryland

                            LuAnn E. White, Ph.D.
      Tulane University School of Public Health and Tropical Medicine
                 Department of Environmental Health Sciences
                           New Orleans, Louisiana


INTRODUCTION

Studies were conducted to characterize  the short term neurotoxic effects of
the inhalation of uncombusted diesel fuel vapors.  Since diesel fuel contains
indentified harmful hydrocarbon    constituents which may exert neurological
effects, the studies  in this  research effort were designed to screen for
neurotoxic effects of the uncombusted diesel vapors.  Mice were exposed at
concentrations of 0.204 mg/1, 0.135  mg/1 and 0.065 mg/1 of uncombusted
diesel vapor for 8 hours/day, 5  consecutive days.

METHODS

Three groups of mice were maintained throughout the experiment:  an exposure
chamber group, a control chamber group, and a vivarium control.group.  Ten
mice each were in the exposure chamber  and the control chamber groups;
five mice were in the vivarium control  group.  Conditions of the exposure and
control chamber groups were identical except for the presence of the
uncombusted diesel vapor.

Five tests were selected to identify changes in performance which are
related to interference of the nervous  system.  These tests were:  the square
box activity test, used to evaluate  activity of the mice by suggesting either
depression or stimulation of  activity;  the rota rod test which indicates
alterations of the integrity  of  neuromuscular junctions and coordination; the
inclined plane test, which serves to evaluate neuromuscular junction inte-
grity of neuromuscular strength  or paralysis; the cornea! reflex test, used
to screen for spinal cord depressant activity; and the hot plate test, a test
of analgesic response.  General  observations were made during testing and for
30 minutes after mice were returned  to  their cages.  The tests were adminis-
tered to exposure and control chamber groups 24 hours prior to the first day
of exposure, after completion of each day of exposure, and 24 hours after
the last day of exposure.  Results of the tests were compared between the
exposure chamber and control  chamber groups and expressed as percent of
control within standard error.   Exposure was conducted using an inhalation
chamber which exposed mice primarily via the respiratory system with minimal
ingestion and cutaneous exposures.   Vapor generation was directly from complex
liquid state to the vapor state  and  varied +_ 10% for the duration of the study.
                                     531

-------
RESULTS

Square Box Activity Test:  Mice exposed to 0.204 mg/1 showed 50 to 90% less
activity than the control chamber mice.  Those mice exposed to 0.135 mg/1
demonstrated activity comparable to that of the control chamber mice, while
mice exposed to 0.065 mg/1 had increased activity as high as 150 percent as
that of the control chamber mice.  Comparison 24 hours after removal from
exposure to the diesel vapor resulted in similar values between exposure and
control chamber groups.

Rota Rod Test:  When compared to the control chamber group, mice exposed to
0.204 mg/1 initially demonstrated a slight increase in performance which
drastically deteriorated as exposure continued.  Mice exposed to 0.135 mg/1
had a slight increase in performance followed by a slight decrease in
performance when compared to control chamber mice.  Those mice exposed to
0.065 mg/1 showed no relative difference in performance between exposure and
control chamber mice.  Comparison after 24 hours of recovery from last
exposure indicated little change between exposure and control chamber mice.

Hot Plate Test:  Results of the exposure to 0.204 mg/1 indicated an initial
increase in heat sensitivity followed by tolerance to heat.  At 0.135 mg/1 a
slight increase in heat sensitivity was observed for the entire exposure.
Exposure to 0.065 mg/1 identified a substantial increase in heat sensitivity
for the exposure mice as compared to the control mice in the chambers.  No
relative difference could be observed between exposure and control chamber
groups 24 hours after termination of exposure.

Corneal Reflex Test;  No difference in response was noted between groups for
any concentrations.

Incline Plane Test:  Both the exposure and control chamber groups had
negative results at each concentration.

All test result variations are depicted in Figure 1.

General Observations:  All mice exposed to 0.204 mg/1 displayed severe
discoloration of the tail indicating  vasodilation after three days of
exposure.  Severe dehydration was observed in all mice.  Grooming habits
deteriorated after day two of the exposure.  While walking, five (50%) of
the mice displayed tremors through day three.  Half of the mice with tremors
died at day three or sooner; the rest continued to have tremors but recovered.
A weight loss of 30% was observed in the exposure group.  The mice in the
exposure chamber group were generally less active than the control chamber
group when returned to their cages.

At an exposure of 0.135 mg/1, five (50%) of the mice displayed tail discolora-
tion after 3-4 days of exposure; slight dehydration was also apparent.
Tremors were evident in three (30%) of the mice while in motion; however, no
deaths occurred.  Grooming was poor but less so as compared to the 0.204 mg/1
exposure group.
                                    532

-------
In mice exposed to the 0.065 mg/1 concentration no appreciable difference
were displayed between the exposure and control chamber groups.

CONCLUSIONS

The exposure of the mice  identified general trends in the effect of uncombus-
ted diesel vapor on the nervous  system.  Comparison of exposure to control
chamber groups suggests a positive central nervous system involvement.
Exposure appears to be concentration and duration dependent.  At the
concentration 0.204 mg/1  depression of the inhibitory neuron occurs followed
by extensive depression of the stimulatory neuron.  Concentrations of 0.135
mg/1 appear to have little effect when compared to controls and suggest
depression of inhibitory  neuron  and slight depression of the stimulatory
neurons while the  concentration  of 0.065 mg/1  causes depression in the
inhibitory neuron.
                                     533

-------
en
CO
                 
                 O
                 O
                 O
O

LU
O.
O-


CO
                      175
                      150
                       125
     100
                       75
                       50
                        25
                                                                    Activity Test
                                                                    Rota Rod Test
                                                                    Hot Plate Test
                                                                    Questioned Data
                                                                       Point	
                             0123456
                            Concentration 0.204 mg/1
                                                                0123
                                                                Concentration
                                                                                                J	L
                                                                               456
                                                                               0.65 mg/1
                                     0123456
                                    Concentration 0.135 mg/1
                                    DAYS OF EXPOSURE
Figure 1. Results of Screening Tests as Percent Control for All Three Concentrations of Uncombusted
                                  Deisel  Vapor

-------
 EVALUATION OF THE RELEASE OF MUTAGENS AND 1-NITROPYRENE FROM DIESEL PARTICLES
              IN THE PRESENCE OF LUNG MACROPHAGE CELLS IN CULTURE

                                      by

             Leon C. King, Silvestre B. Tejada, and Joellen Lewtas
                     U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina


     Diesel particles have been shown to contain organic components which are
mutagenic in short-term bioassays (1).  Nitroaromatics including 1-nitropyrene
now appear to account for a portion of the mutagenic activity observed in these
organics (2).  It is important to determine whether these potentially
carcinogenic components are released from particles under physiological
conditions.  We have shown in a previous study (3) using the Ames Salmonella
typhimurium plate incorporation assay that serum and lung cytosol were very
effective in removing mutagenic activity from diesel particles (3).  This was
determined by measuring the activity extractable from the particles before and
after treatment of the particles with serum or cytosol.  Direct addition of
serum or lung cytosol to these organics reduced the mutagenic activity
detectable by 80 to 90%.  The objective of this study was to evaluate the
removal and release of mutagenic activity and 1-nitropyrene from diesel
particles in the presence of lung cells in culture.  The lung cells used in
these studies were alveolar macrophages obtained from rabbits by lung lavage.
The diesel particles used in this study were obtained from a Datsun Nissan
220C, 4-cylinder passenger car previously described (3).  The particles  were
sonicated in tissue culture medium (M199 with Hank's Salts) at 37°C for  30 min
to deagglomerate and disperse the particles.  Particle size analysis showed the
majority of the particles to be 2.0 to 2.5 un after this treatment.

     Rabbit alveolar macrophage (RAM) cells were lavaged and cultured with the
diesel  particles according to previously published procedures (4) except that
the final serum concentration was reduced to 10% and the culture time was
increased to 40 h.  In brief, RAM cells were added to individual wells of
cluster dishes containing suspensions of the diesel particles from 15 to
1,500 ug/ml.  Under these conditions, at concentrations of diesel particles
above 75 yg/ml, over 95% of the particles were phagocytized after 20 h of
culture.  After 40 h of culture, cells were harvested by trypsinization.  The
final exposure conditions for the mutagenesis and 1-nitropyrene analyses were
selected to maximize cellular exposure to the diesel particles while minimizing
cellular toxicity.  The final concentration of diesel particles selected was
375 ug  of particles/ml.  At this concentration, RAM cells engulf over 95% of
the particles and after 40 h of exposure, less than 7% cell lysis was observed
and cell viability was 63% of the control cultures.
                                      535

-------
     In order to evaluate the effect of the lung macrophage cells on the
removal of mutagens and l-n1tropyrene from dlesel particles, particles were
exposed to the culture medium at 375 ug/ml with and without lung macrophages
and cultured for 40 h.  After incubation the medium control treatment dishes
were combined and the medium was separated by centrifugation from the particles
which were washed once with water.  The harvested macrophages were separated
from the culture medium and sonicated to release engulfed particles.
Dichloromethane and methanol (DCM:MeOH, 1:1) were used to extract the medium,
free particles, engulfed particles, cell sonicate, and water washes.  Each of
these fractions was analyzed for 1-nitropyrene by LC/fluorescence (5) and
compared to untreated particles extracted with DCM:MeOH.

     Each of the fractions was also assayed for mutagenicity in the Ames
Salmonella typhimurium plate incorporation assay in TA98 as previously
described (3).The media, cells, and washes were assayed without extraction
and due to a high background of activity in both the medium and cells, no
detectable activity was observed as a function of the treatment group.
Mutagenicity was detected in the DCMrMeOH extract of the macrophage-engulfed
particles; however, it was only 2% of the mutagenicity originally present on
the particles.  Comparison of this activity with that of the media control
particles, in which 6% of the mutagenicity was recovered, showed that the
presence of macrophages decreased the mutagenicity 62%.

     Nitropyrene analysis of the medium control group showed 96% recovery of
the 1-nitropyrene with 31% of the nitropyrene found in the medium and washes
and 66% remaining on the particles.  Significantly less 1-nitropyrene was
recovered in the macrophage treatment group (76%).  Since over 95% of the
particles were phagocytized and only 27% of the original 1-nitropyrene was
recovered from the engulfed particles, it appears that the macrophages may have
metabolized the 1-nitropyrene to a nondetectable form.

     Comparison of the recoveries shows a greater loss of mutagenicity than
1-nitropyrene suggesting that other compounds detected in these particles may
contribute more to the mutagenicity than the 1-nitropyrene alone.


REFERENCES

1.   Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R. Zweidinger, S.  Tejada,
       J. Bumgarner, F. Duffield, M. Waters, U. Simmon, C. Hare, C. Rodriguez,
       and L. Snow.  1978.  Application of short-term bioassay to the
       characterization of diesel particle emissions.  In:  Application  of
       Short-Term B-ioassays in the Fractionation and Analysis of Complex
       Environmental Mixtures.  M.D. Waters, S. Nesnow, J.L. Huisingh,
       S.S. Sandhu, and L. Claxton, eds.  Plenum Press:  New York.
       pp. 383-418.
                                     536

-------
2.   Nishioka, M., B. Peterson, and J. Lewtas.  1981.  Comparison of nitro-PNA
       content and mutagenlcity of diesel emissions.  Presented at the
       U.S. Environmental Protection Agency Diesel Emissions Symposium,
       Raleigh, North Carolina.

3.   King, L.C., M.J. Kohan, A.C. Austin, UD. Claxton, and J.L. Huisingh.
       1981.  Evaluation of the release of mutagens from diesel particles in
       the presence  of physiological fluids.  Environ. Mutagen. 3:109-121.

4.   Garrett, N.E.,  J.A. Campbell, H.F- Stack, M.D. Waters, and Joellen Lewtas.
       1981.  The utilization  of the rabbit alveolar macrophage and Chinese
       hamster ovary cell for  evaluation of the toxicity of particulate
       material.  Environ. Res. 24:345-365.

5.   Tejada,  S.B., R.B.  Zweidinger, J.E. Sigsby, Jr., and R.L. Bradow.  1981.
       Identification and measurement  of nitro derivatives of PAH in diesel
       exhaust particulate extract.  Presented at the Chemical Characterization
       of  Diesel  Exhaust Emissions Workshop,  Dearborn, MI.
                                        537

-------
     BACTERIAL MUTAGENICITY OF A DIESEL EXHAUST EXTRACT AND TWO ASSOCIATED
           NITROARENE COMPOUNDS AFTER METABOLISM AND PROTEIN RINDING

                                      by

                         M1ke Kohan and Larry Claxton
                      Health Effects Research Laboratory
                     U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina


     Previous work has demonstrated that nitroarenes are associated with
organic extracts from diesel exhaust (1,2,3).  This study was designed to
characterize two nitroarenes and to determine whether these compounds
demonstrate the same type of mutagenic response as diesel exhaust organics.
An extract of the partlculate from exhaust of a VW diesel automobile,
2,7-dinitrofluorenone, and l-n1tropyrene were tested 1n the Salmonella
mutagenicity assay using strain TA98 and two nitroreductase-deficient strains
(TA98FR1 and TA98/l,8DNPs) to establish a proper dose for a multivariant
experiment. Each of the three samples was tested with and without a 9000g liver
homogenate (S9) prepared from Aroclor 1254-induced rats.  In the multivariant
experiment, the following treatments (both with and without the
NADPH-generating system) were used:

     (a)  no activating system (-S9),

     (b)  mlcrosomes derived from the original S9 by centrifuglng for
          90 min at lOO.OOOg,

     (c)  the cytosol fraction of the original S9, and

     (d)  boiled S9.

     In addition, the samples were tested 1n the presence of both boiled and
unboiled bovine serum albumin (BSA).  Salmonella tester strains TA98 and
TA98FR1 were used in the multivariant study.THese treatments were used to
determine the similarities of the three samples 1n the presence and absence  of
treatments with differing enzymatic and protein binding capabilities.  This
work is an extension of similar efforts by Wang et al. (4) and Pederson et
al. (5).

     The results using the diesel exhaust sample are seen in Table 1.  When
compared to the untreated (no activating system) situation, all of the
treatments with one exception gave reduced mutagenic activity.  Only the
cytosol fraction gave an increase in activity.  In comparison to the
non-activated situation the cytosol activation gave a relative increase to 111*
                                      538

-------
for strain TA98 and to 164% for strain TA98FR1.  In addition, it was noted that
TA98 was the strain most responsive to the diesel exhaust extract, followed by
TA98FR1, and then TA98/1,8DNP6.  1-Nitropyrene activity was decreased by half
in TA98/1,8DNP6 but almost abolished in TA98FR1.

     As in the diesel exhaust particulate sample, the mutagenic activity of
2,7-dinitrofluorenone was reduced  in the majority of the different treatments.
An increase to 170% of the untreated situation was seen when using the cytosol
fraction with strain TA98FR1.

     A reduction in the mutagenic  activity of 1-nitropyrene was observed in all
of the treatments without the generating system.  However, in contrast to
2,7-dinitrofluorenone and the diesel exhaust particulate sample, a twenty-fold
increase in the mutagenic activity of 1-nitropyrene was seen with the microsome
treatment when using the generating system and strain TA98FR1.  Smaller
increases in activity were also observed with the S9 and cytosol fractions when
using this same strain and generating system.  The mutagenic activity of
1-nitropyrene also was increased by the microsome fraction with the generating
system in strain TA98.  This microsomal activation of 1-nitropyrene, in
addition to the response pattern obtained with the nitroreductase-deficient
tester strains, indicates that 1-nitropyrene may not be the major mutagenic
component in this sample of extractable organics from VW Rabbit Diesel particle
emissions.


REFERENCES

1.   Pederson, T.C., and J.S. Siak.  1981.  The  role of nitroaromatic compounds
       in the direct-acting mutagenicity of diesel particle extracts.  J. Appl.
       Toxicol. 1:54.

2.   Claxton, L.D., and J.L. Huisingh.  1980.  Comparative mutagenic activity
       of organics from combustion sources.  In:  Pulmonary Toxicology of
       Respirable Particles.  Proceedings of the Nineteenth Annual Hanford Life
       Sciences Symposium at Richland, WA.  P.L. Sanders, F.T.  Cross,
       G.E. Dable, and J.A. Mahaffey, eds.  pp.  453-465.

3.   Lofroth, G.  1980.  Salmonella/microsome-mutagenicity assays of extract
       from diesel and gasoline-powered motor  vehicles.   In:  Health Effects  of
       Diesel Engine Emissions:  Proceedings of  an International Symposium.
       EPA-600/9-80-057a, Vol. 1.  U.S. Environmental Protection Agency:
       Cincinnati, OH.  pp. 327-342.

4.   Wang, Y.Y., R.E. Talcott, D.A.  Seid, and  E.T. Wei.   1980.  Antimutagenic
       properties of liver homogenates, proteins, and glutathione on diesel
       exhaust particulates.  Cancer Lett.  11:265-275.

5.   Pederson, T.C., and J.-S. Siak.   1981.  The activation  of  mutagens  in
       diesel particle extract with  rat liver  S9 enzymes.  J. Appl. Toxicol.
        1:61-66.
                                       539

-------
       Table 1.   Mutagenic Activity of Diesel  Exhaust  Particle  Extracts
                 Using Different  Treatment  Conditions
                   Without Generating System          With  Generating System3
Treatment
Untreated
S9
Microsomes
Cytosol
Boiled S9
BSA
Boiled BSA
TA98
1039 (100)b
467 (45)
781 (75)
687 (66)
595 (57)
354 (33)
343 (33)
<"NADPH-generating system: NADPH,
bAverage net revertants per plate
TA98FR1
788 (100)c
242 (31)
488 (62)
332 (42)
369 (47)
264 (34)
232 (29)
G-6-P, MgCI_2» and
at 100 pg organic
TA98
1001 (100)
696 (70)
826 (82)
1110 (111)
577 (58)
KCL.
extract.
TA98FR1
680 (100)
649 (9B)
557 (82)
1118 (164)
305 (45)

Relative response to untreated  sample  expressed  as  percent.
                                     540

-------
                  CHARACTERIZATION OF PARTICIPATE EMISSIONS
                  FROM IN-USE GASOLINE FUELED MOTOR VEHICLES
                                      by
                   John M.  Lang, Roy A.  Carlson,  Linda Snow
                           Northrop Services,  Inc.
                    Research Triangle Park,  North Carolina


               Frank M. Black,  Roy Zweidinger,  Silvester Tejada
                     U.S.  Environmental  Protection Agency
                    Research Triangle Park,  North Carolina
One  of  the  primary tasks of those concerned with the  study of environmental
quality is  estimating population exposure to air pollutants and  determining
the  risk associated with exposure.  Mobile sources generally contribute
significantly to the population's exposure to hydrocarbons (HC),  carbon
monoxide (CO),  oxides of nitrogen (NO ),  and fine particulate matter.
Much of the mobile source data available  in the  literature has been
obtained from well-maintained engineering test vehicles.  However,  emissions
from consumer-operated vehicles can vary  considerably from we11-maintained
vehicles.

Because  of  growing interest in diesel power for  light-duty motor  vehicles,
it has  become necessary to study particulate emissions comprehensively
to determine  the potential impact of dieselization on the public  health.
Assessment  of exposure and risk requires  knowledge of the emission rates
and  composition of diesel particulate matter and of the gasoline  particu-
late matter being replaced by diesels.

Our  study examined particulate emissions  from twenty  consumer-operated,
light-duty  gasoline fueled cars and trucks.  The  emissions characteristics
observed were compared with those previously reported by  Gibbs,  et al.,
for nineteen  consumer-operated,  light-duty diesels (1).

A test  fleet  of twenty light-duty gasoline passenger  cars and trucks was
obtained from local  residents and rental  agencies.  Four of the vehicles
were fueled by  leaded  gasoline;  the remaining sixteen fueled  by unleaded
                                     541

-------
gasoline. Vehicles from model years 1970 to 1981 were represented  and
ranged in mass from 907 kg (2,000 Ibs) to 2,268 kg  (5,000  Ibs)  and
accumulated mileage from 565 km (353 mi.) to 130,490 km  (81,559 mi.). A
variety of emission control configurations were represented,  including
oxidation catalysts, oxidation catalysts with air pumps, and  three-way
catalysts. All vehicles were tested as received.

Three gasoline fuels were purchased locally for the program;  two leaded
and one unleaded. Road load simulation was achieved with a Clayton
CTE-50-0 direct-drive, water brake dynamometer. The vehicles  were  tested
with a daily routine involving a cold-start Federal Test Procedure (FTP)
cycle followed by repetitive Highway Fuel Economy Test cycles (HWFET).
The FTP simulates city driving after the car has not been  started  for at
least twelve (12) hours. The HWFET simulates highway driving  after the
car has been warmed up.

A positive-displacement pump Constant Volume Sampling (CVS) system
collected the exhaust gases and allowed dilution air to flow  through the
sample train in order to maintain the dilution exhaust temperature at or
below 52° C (125° F). Before entering the CVS system, the dilution
exhaust flow passed through three 20 x 20 inch filters for collection
of particulate matter  (2). Pallflex T60A20 PTFE glass fiber filters were
used for  collecting the particulate.

The 20 x  20 filters were  soxhlet extracted for eight hours with dichloro-
methane  (DCM) to remove the soluble organic fraction  (SOF). The extracted
organics  were examined for nitro-pyrene  (NO.-P), benzo(a)pyrene (BaP),
and pyrene  (Py) content by high pressure liquid chromotography  (HPLC).and
bioassayed with Ames Salmonella strain TA-98 for mutagenic potency. The
activity  (revertants/ microgram SOF) was determined as the slope in the
linear portion of the dose-response curve. Five organic doses,  0  (solvent
blank) to 100 ug, were used to define the dose-response curve.

A review  of the regulated emission rates indicate that the test fleet
included properly functioning vehicles and vehicles with a variety of
emission  control malfunction conditions. HC emission rates ranged  from
0.07 to 24.S g/mi., CO had a range of 0.08 to 60.95 g/mi., and  NO   was
from 0.37 to 6.43 g/mi.. Comparison with the diesels indicate tha£ the
gasoline  fleet average emission rates exceed the diesel rates and  a
broader range was seen in the gasoline data.

The total particulate emission rates from leaded gasoline vehicles were
2.7 times greater during the HWFET than the FTP. On the other hand,
HWFET total particulate emission rates from unleaded gasoline vehicles
were 81% of the FTP value. Overall, the leaded vehicles emitted more
particulate than the unleaded vehicles during both cycles, 3.2  times
more during the FTP and 10.8 times more during the HWFET.

The light-duty diesel total particulate emission rates reported by
Gibbs, et al., compared to the gasoline vehicle rates as follows:
     (FTP) diesel ^5.9 x leaded rate; 19.1 x unleaded rate.
     (HWFET) diesel -—1.3 x leaded rate; 13.5 x unleaded rate.
                                     542

-------
Fleet average DCM soluble particulate organic emission rates during the
FTP and HWFET compared with the  light-duty diesels as follows:
     (FTP) diesel~ 5.9 x leaded rate; 8.6 x unleaded rate.
     (HWFET) diesel~ 3.4 x leaded  rate;  6.9 x unleaded rate.

Nitro-pyrene emission rates were similar  for leaded and unleaded gas-
oline vehicles. Diesels emit  about  20 to  30 times as much NO.-P as
gasoline vehicles.                                          ^

Gasoline vehicle BaP  emissions were greater during cold-start FTP driving
than during either hot-start  FTP or HWFET driving. Functioning catalyst
systems appeared to effectively  reduce polynuclear aromatic hydrocarbon
(PAH) emissions. During the cold-start FTP, BaP  emission rates from
leaded vehicles exceeded  BaP  rates  from diesels; BaP emission rates from
unleaded vehicles were similar to the diesel results. During the HWFET,
BaP emission rates from both  categories of gasoline vehicles were less
than the rates from diesels.

Generally, the Ames TA-98 mutagenic activity of  the gasoline particulate
organics was higher with metabolic  activation than without metabolic
activation. Ames TA-98 revertant per mile levels were substantially
higher for the leaded gasoline vehicles than for the unleaded gasoline
vehicles under both FTP and HWFET conditions.

The higher total particulate  and particulate organic emission rates of
diesels are compensated somewhat by lower Ames TA-98 mutagenic activities
when compared to gasoline vehicles.  Relative activities (without metabolic
activation) were as follows:
     (FTP) diesel'*- 0.6 x leaded activity; 0.5 x unleaded activity.
     (HWFET) diesel ~- 0.4 x leaded  activity; 0.4 x unleaded activity.

Fleet average Ames TA-98 revertant  per mile levels (without metabolic
activation) for FTP and HWFET conditions  compared with light-duty
diesels as follows:
     (FTP) diesel"- 3.4 x leaded rate; 12.1 x unleaded rate.
     (HWFET) diesel *- 1.5 x leaded  rate;  9.1 x leaded rate.

In conclusion, replacing the  gasoline passenger cars represented by the
test fleet of this program with  diesel passenger cars would decrease HC,
CO, and NO  population exposures, and increase total particulate and
mutagenic particulate organic exposures (as indicated by Ames Salmonella
strain TA-98).
                                   543

-------
01
-Fa
                                         Table  1.  Characterization  of Particulate Emissions,
                                                         Diesel  versus Gasoline

Diesel *
FTP HWFET
Total particulate, mg/mi.
Dichlorome thane soluble organics, mg/mi.
Benzo-a-pyrene, ug/mi.
Nitro-pyrene, ug/mi.
TA-98.-S9, rev/mg
TA-98,+59, rev/ug
TA-98.-S9, rev/mi. (x 10"3)
TA-98.+S9, rev/mi. (x 10~3)
607(1)
124(1)
4.5(3-7)
7.4(8)
4.1(1)
**
509(1)
**
345(1)
79.7(1)
2.7(3-7)
6.8(8)
3.0
**
239
**
Leaded
FTP HWFET
103
21.1
14.5
0.20
7.31
12.5
152
258
276
23.5
0.89
0.39
8.55
10.6
163
232
Unleaded
FTP HWFET
31.7
14.4
3.3
0.24
7.57
13.4
42.1
79.3
25.6
11.5
0.61
0.16
7.39
7.43
26.4
25.2

               * Number in parenthesis  indicates reference.
               ** Information unavailable at  this  time.

-------
                                  REFERENCES

1.  Gibbs, R.E., J.D. Hyde, and S.M. Byer. Characterization of particulate
   emissions from in-use diesel vehicles. SA£ paper number 810081,
   February, 1981.

2.  Killough, P., and J. Watson. Filter-type, high volume particulate
   samples for automotive diesel emission studies. ES-TN-79-13. Northrop
   Services, Inc., December 1979.

3.  Huisingh, J.L., and R.L. Bradow, R.H. Jungers, B.D. Harris, R.B.
   Zweidinger, K.M. Gushing, B.E.Gill, and R.E. Albert. Mutagenic and
   carcinogenic potency of extracts of diesel and related environmental
   emissions: study design, sample generation, collection and preparation.
   Health Effects of Diesel Engine Emissions: Proceedings of an Inter-
   national Symposium, EPA-600/9-80-OS6b. November 1980.

4.  Kraft, J., and K.H. Lies. Polycyclic aromatic hydrocarbons in the
   exhaust of gasoline and diesel vehicles. SAE paper number 810082.
   February 1981.

5.  Williams, R.L.,and D.P. Chock. Characterization of diesel particulate
   exposure. Health Effects of Diesel Engine Emissions: Proceedings of
   an International Symposium. EPA-600/9-80-057a. November 1980.

6.  Hare, C.T., and T.M. Baines. Characterization of particulate and
   gaseous emissions from two diesel automobiles as a function of fuel
   and driving cycle. SAE paper number 790424. February 1979.

7.  Williams, R.L., and S.J. Swarin. Benzo(a)pyrene emissions from
   gasoline and diesel automobiles. SAE paper number 790419.
   February 1979.

8.  Tejada, S.. Particulate NO -pyrene emissions from a 1980 Oldsmobile
   and a 1980 VW diesel, FTP and HWFET cycles. Unpublished data. June
   1981.
                                      545

-------
                SURFACE REACTIVITY OF DIESEL PARTICLE AEROSOLS


                                      by
                        Magnus Lenner, Oliver Lindqvist
                             and Evert Ljungstrcin
                       Department of Inorganic Chemistry
                          University of Gothenburg and
                       Chalmers University of Technology

                          Inger Lundgren and Ake Rosen
                             ^folvo Car Corporation

                               Gothenburg, Sweden
INTRODUCTION


     In 1979 a research project, concerning exhaust emissions from passenger
cars, was initiated by Volvo Car Corporation in collaboration with the
University of Gothenburg and Chalmers University of Technology. The first
report  (1), which appeared in September 1979, comprised of investigations
of exhaust particulates both from gasoline powered and fron diesel powered
passenger cars, as well as chemical analyses by several methods, of gaseous.
and particulate matter in exhaust samples frcm Volvo cars.

     The present work deals with the influence of diesel particles on the
oxidation of nitric oxide to nitrogen dioxide at different conditions of load,
temperature and dilution. Secondly, spectroscopic investigations of how the
chemical composition of diesel particle surfaces may be modified in the
atmosphere have been performed. A full report (2) will be published in October
1981.
EXPERIMENTAL


Determination of Conversion Rates

     To determine the effect of diesel particles on the conversion NO -* iSD2 /
two series, each comprising of eight rate constant determinations were made in
bag samples of exhausts from a Volvo passenger diesel. The samples were
                                      546

-------
analysed for nitrogen oxides concentrations  at  intervals during ~24 h after
sample collection. A Mbnitor Labs.  8440 Nitrogen Oxides Analyser was used.

     For the first series of samples the  engine of the car was run at
approximately 1500 rpm idle, while  for the second series the car was driven
at 40 km/h with a road load of  12.5 horse powers in  a chassis dynamometer.
The parameters temperature  (0°  or 23'  C), dilution rate  (~1/60 or~1/120)
and presence/absence of diesel  particles  were varied. The samples were
collected in Tedlar bags contained  in  an  aluminum barrel, which could be
evacuated to make the bag extract the  appropriate volume of exhaust gases.
Dry aix from a gas cylinder was used for  dilution. Diesel particles were
removed by a filter for the particle-free samples.

Spectroscopic Methods

     For infrared spectroscopy, samples were collected with an Electrical
Coroma Sampler  (3) on gold covered metal plates and analysed by a
reflection-IR method, with a Nicolet MX-1 FTIR  instrument.

     ESCA samples were collected  on Millipore filters and the electron
spectra were recorded with a Hewlett-Packard 5960 A  electron spectrometer.
RESULTS AND DISCUSSION


Conversion Rates

     The formation of NO.  from NO obeys  the  relationship:- ^|p = k |NC]"2 [O ] .
The rate constant k  is  conrcnly given in either of the dimensions
1/ [k]   per second or 1/ppm per minute.

     The values for  k  ( |M]    x sec.   ),  calculated for the respective  16
experiments, are summarized  in Table 1.  The  rate constant has a negative
temperature dependence, especially at higher dilution rates. The reaction is
enhanced by the presence of  diesel particles.  The latter effect, though, is
not as strong as the catalytic effect of street and wall surfaces, reported
by Lindqvist et al.  (4).

Spectroscopic Results

     ESCA measurenents  were  made on  three kinds of samples of diesel
particles, namely unexposed  samples, samples which had been exposed to 2 ppm
NO  for 48 h and finally samples which had been exposed  simultaneously to
NO:? and UV-light for 6  h.  Nitrogen (1s)  responses were obtained only fron
the latter kind of sample. Signals at 400 eV and at  402  eV were assigned
to emanate respectively frcm NX and  NH4  by comparison with the  results of
Chang & Novakov  (5).

     The infrared spectra  obtained by reflection of  IR-light through diesel
particles precipitated  on  a  gold film showed absorption  peaks at 12 90/cm
and at 860/on, corresponding respectively to C - N stretch in primary
                                      547

-------
aronatic amines and to N - 0 stretch in aronatic nitro compounds. Uhexposed
samples gave the same kind of spectra as samples which had been exposed to
N02 and UV-light.
REFERENCES
 1.  Lundgren, I., Rosen, A. and Lindqvist, 0. 1979. Unregulated pollutants.
       Measurements and analysis of exhaust gas and particulates from Volvo
       light-duty vehicles. Volvo Car Corporation:  Gothenburg.  80 pp.

 2.  Lenner, M., Ljungstrcm, E., Lindqvist, O., Lundgren, I., and Rosen, A.
       1981. Reactivity and catalytic activity of diesel particles. Studies
       of NO  and particle emissions from a Volvo passenger diesel. Volvo
       Car Corporation:  Gothenburg.  63 pp. In press.

 3.  Van de Vate, J. F., Plomp, A., de Jong, C. and Vrins, E. L. M. 1977.
       A battery-operated portable unit for electrostatic and impaction
       sampling of ambient aerosols for electron microscopy. Presented at the
       5th Conference of the Gesellscnaft fur Aarosolforschung, Karlsruhe,
       W. Germany.

 4.  Lindqvist, 0., Ljungstrcm, E. and Svensson, R. 1981. Low temperature
       thermal oxidation of nitric oxide in polluted air. Atm. Environment.
       In press.

 5.  Chang, S. G. and tfovakov, T. 1975. Formation of pollution particulate
       nitrogen compounds by NO-soot and NH.,-soot gas-particle surface
       reactions. Atm. Environment 9:495-505.
                                     548

-------
  Table 1. Calculated Bate  Constants,  k is the slope of the function
              1/ JND] t ~  V [NQl 0  = kt,  calculated from measurements of
              jfccjj  at intervals  after  the start of an experiment at
             t = 0. The values  have been multiplied by 10-4.
No.
      Particles

Temp. CO  Oil, rate
                               1500 rpn idle
No.
   No particles

Temp.(°C)  Oil. rate
1
2
3
4
0
0
23
23
1/60
1/120
1/60
1/120
2.11
2.39
1.81
1.82
5
6
7
8
0
0
23
23
1/60
1/120
1/60
1/120
1.74
2.04
1.56
1.57
            Particles
                 _40_tan/h.  Road load;  12.5 hp

                                           No particles
No.
9
10
11
12
Tenp. CO
0
0
23
23
Dil. rate
1/60
1/120
1/60
1/120
k
2.18
2.43
1.97
1.92
No.
13
14
15
16
Temp. CC)
0
0
23
23
Dil. rate
1/60
1/120
1/60
1/120
k
1.75
1.99
1.59
1.53
                                    549

-------
       EFFECTS OF OZONE AND NITROGEN DIOXIDE PRESENT DURING SAMPLING OF
     GENUINE PARTICULATE MATTER AS DETECTED BY TWO BIOLOGICAL TEST SYSTEMS
               AND ANALYSIS OF POLYCYCLIC AROMATIC HYDROCARBONS

                                     by

                                 G. Ldfroth
                         Department of Radiobiology
                          University of Stockholm
                             S-106 91 Stockholm
             R. Toftgird, J. Carlstedt-Duke and J-A. Gustafsson
              Departments of Medical Nutrition and Pharmacology
                            Karolinska Institute
                             S-104 01 Stockholm
                 E. Brorstrom, P. Grennfelt and A. Lindskog
             Swedish Water and Air Pollution Research Laboratory
                            S-402 24 Gothenburg


     Urban particulate matter was collected in wintertime at -5-0 °C during
24 h periods on glass fiber filter with two simultaneously operating high
volume samplers. One of the samplers was equipped with an ozone or nitrogen
dioxide dosage system enhancing the concentration with about 200 ppb ozone or
960 ppb nitrogen dioxide. Filters were Soxhlet-extracted with acetone and the
extracts analyzed with respect to eight polycyclic aromatic hydrocarbons
(PAH), mutagenicity in the Salmonella/microsome assay and ability to displace
2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) from the rat liver receptor
protein.

Polycyclic Aromatic Hydrocarbons
     A part of or the whole acetone extract of each filter was subjected to a
clean up procedure with respect to PAH and eight components were quantified
by gas chromatography on a Carlo Erba equipment with a SE-54 glass capillary
column. The range of concentrations detected is given in Table 1.
     Exposure to ozone had very little effect on the concentrations of the
PAH. Significant degradation occurred only in one of the experiments. At this
occasion the concentrations of nitric oxide and nitrogen dioxide in ambient
air were rather high and it seems likely that the degradation may have been
caused by nitrogen oxides formed by ozone oxidation of nitric oxide.
     Statistical analysis between the nitrogen dioxide exposed samples and
the simultaneously collected non-exposed samples showed a significant degra-
dation for pyrene, benz(a)anthracene and benzo(a)pyrene being on the average
about 20, 40 and 60 %, respectively.
                                     550

-------
Salmonella Mutagenicity

     Portions of the acetone extracts were reduced in volume, but not to dry-
ness, at  <40 °C and then diluted with dimethyl sulfoxide (DMSO). Each of the
DMSO samples was assayed at at least three different occasions for mutageni-
city by the Salmonella plate incorporation method with bacterial cultures
fully grown overnight. Tests were performed with the strains TA 98 and TA 100
in the absence and presence of the microsome containing liver supernatant
from Aroclor 1254-induced male rats (S9) and with the nitroreductase defi-
cient strains TA98NR, TA98/1,8DNPg and TA10DNR in the absence of S9. The
S9 was used at a level of 20 ul per plate and was added together with neces-
sary co-factors.

     The  response expressed as revertants per cu.m  was calculated from the
linear or approximately  linear dose response curves. The results for assays
in the absence of S9 are given in Table 2. The addition of S9 either de-
creased or did not change the mutagenic response.

     Exposure to ozone did not generally alter the mutagenic response except
for an increase in the experiment which also resulted in degradation of PAH.

     Exposure to nitrogen dioxide increased the mutagenic response in nitro-
reductase proficient as  well as nitroreductase deficient strains. The average
enhancement found in the three investigated experiments was a 3-4-fold
increase.
Affinity  to the Rat Liver TCDD-Receptor Protein

     Assays for the ability to displace TCDD from the.rat liver receptor
protein were performed by adding different amounts of the DMSO samples to
one ml of rat liver cytosol containing tritium-labeled TCDD and determining
the amount of TCDD which remained bound to the receptor. After correction
for non-specific binding, the relative specific binding was calculated from
log-log it plots and was  expressed as the amount that competes for 50 % of
the specific TCDD-binding, EC50, cu.m per ml cytosol.
     Two  experiments from the nitrogen dioxide and one from the ozone expo-
sure were analyzed and the results are given in Table 2. Neither nitrogen
dioxide nor ozone altered the affinity.

Conclusions
     High volume sampling of ambient particulate matter on glass fiber filter
in the presence of a high level of ozone does not cause a significant degra-
dation of PAH, alter the mutagenic response which is detected by the Salmon-
ella assay or change the affinity to the TCDD-receptor protein.
     Sampling in the presence of a high  level of  nitrogen dioxide causes a
significant degradation  of reactive PAH  and increases the mutagenic response
which is  detected by the Salmonella assay, but does not change the affinity
to the TCDD-receptor protein.
     Simultaneous sampling of genuine particulate matter without and with en-
hanced concentrations of reactive gases  may be the  best method for studying
artifact  reactions during sampling. Further studies are in progress with
nitrogen  dioxide, nitrous acid and nitric  acid.
                                      551

-------
Table 1. Concentration  ranges of analyzed PAH components; ng/m3.
PAH component
Phenanthrene
Fluoranthene
Pyrene
Benz(a)anthracene
Chrysene / Triphenylene
Benzo ( b&k ) f 1 uor anthenes
Benzo(e)pyrene
Benzo(a)pyrene
8 samples
non-exposed
0.42-2.2
0.64-22
1.1 -8.0
1.1 -5.2
0.8 -4.0
1.6 -7.8
0.39-3.6
0.17-2.4
3 samples
0,-exposure
200 ppb
0.66-2.0
1.7 -3.6
1.5 -3.6
0.93-2.0
1.1 -2.1
2.0 -3.5
0.48-1.0
0.16-0.66
5 samples
N02 -exposure
960 ppb
0.39-1.9
0.81-4.5
0.81 -5.8
0.82-3.8
1.1 -4.0
2.2 -7.6
0.36-2.2
0.07-1.0
Table 2. Mutagenic response  in  the absence of mammalian metabolic activation
         and affinity to the TCDD-receptor protein of extracts of particu-
         late matter collected  without and with enhanced concentrations of
         nitrogen dioxide or ozone; n.d. not determined.
Sample
                           Revertants per m
               TA 98   TA 98 NR
        TA 98/
       1,8DNPg
         TA 100  TA 100 NR
                   Receptor
                   affinity
                   EC5Q±s.d.

                     (n-4)
                     m3/ml
800225 AMB


800226 AMB


800228 AMB


800326 AMB

       °3
800327 AMB

       °3
800331 AMB
       0,
                 62
                 87

                 14
                 37

                 11
                 76

                  4.3
                  4.6

                 28
                 31

                 10
                 16
 37
 48

  8
 24

  6
 35

n.d.
n.d.

 11
 12

n.d.
n.d.
 28
 32

  3.2
  6.5

  5
 20

n.d.
n.d.

  4.5
  4.4

n.d.
n.d.
 76
142

 14
 41

 20
110

  2.9
  3.3

 21
 25

  8.4
 14
 36
 62

  5
 20

  5
 41

n.d.
n.d.

  5
  7

n.d.
n.d.
   n.d.
   n.d.

0.17±0.04
0.09 ±0.04

0.07 ±0.02
0.10±0.03

   n.d.
   n.d.

0.08 ±0.02
0.06 ±0.04

   n.d.
   n.d.
                                     552

-------
                  ALUMINA COATED METAL WOOL AS A
         PARTICDLATE  FILTER FOR DIESEL POWERED VEHICLES
                                by
                   M.  A.  McMahon,  W.  T.  Tierney,
                    K.  S.  Virk,  and C.  H.  Faist
Pending  federal  regulations will probably require  1985  and  later
diesel  powered  vehicles  to be  equipped  with aftertreatment de-
vices  to control particulate emissions.  Filtering  devices  which
employ  alumina coated steel wool  as the filter matrix  show pro-
mise  of  being  a  practical  means  for removing  submicron  size
particles from diesel  exhaust.   Our  diesel exhaust  filters (DBF's)
consist  of  two  alumina  coated  stainless-steel-wool filled car-
tridges  inserted into an outer  housing about the same  size as a
conventional  muffler  (Figure  1).   The alumina coating, which  is
applied  to the steel wool by a proprietary process,  characteristi-
cally  retains  soot particles  which  come  in contact  with it.

In our design, metal wool packing  density, alumina  coating thick-
ness  and the  physical dimensions  are  the  dominant performance
determinants   for  the  collection  efficiency  and  exhaust  back
pressure.  The surface area of the alumina in the DEF, which is a
function  of  all of  these,  correlates well with collection  effi-
ciency.

Our tests have  demonstrated that  particulate removal in the 70%
range in FTP testing can be achieved  at acceptable  backpressures.
Measurements also  show that trapping  efficiency  is  constant over
a gas flow range of 0.7 to  20 ft/sec.   The trapping  efficiency of
a given  size  filter  can  be increased above 80% but  only  at an
increased backpressure penalty.    Satisfactory  filters  have  been
made  for diesel  engines  varying  in  size  from  2  to 14 litres.

Generally, filters  capable  of  trapping soot  with efficiencies in
excess of 70%  have a pressure drop  of 3 to  4  inches of  water per
inch of  filter  at  gas velocities of 15  ft/sec.   For example, a
filter  mounted  on a  vehicle  equipped with a  1980 Oldsmobile
engine, having a trapping efficiency  in this  range  demonstrated a
pressure  drop  of 31 inches of water at 40 mph level road condi-
tions.
                                553

-------
In addition to  removing soot from diesel  exhaust,  alumina coated
steel wool filters remove significant  amounts  of  hydrocarbons  and
sulfates.   Importantly, a  significant percentage  (  50%) of  the
polynuclear aromatics,  as  indicated by benzo  (a) pyrene  measure-
ments,  is  removed.   Noise  is  attenuated to levels equivalent to
those observed using conventional mufflers.

Since the  volume of soot generated  by current diesel engines is
quite large, soot collected on DBF's must  be periodically removed
for the filter to trap  effectively for extended mileage.   Burning
at  temperatures  above  1000°F  is the most feasible  method  for
doing this.  Since  diesel  exhaust temperatures at  moderate  loads
are  generally  below  1000°F,   consideration  is  being  given  to
mounting torches  in the exhaust system to  increase  exhaust  gas
temperature when  regeneration  is required.  In that conventional
torches  present  difficult  operating  problems when  employed  in
exhaust  systems,  a  catalytic  torch  is  being developed in our
laboratories.   With this  kind  of torch,  hydrocarbons  injected
into  the  exhaust  are  catalytically  oxidized  to  increase  the
exhaust gas to regeneration temperature.   The  catalytic converter
used  was  made  by  applying a   noble  metal catalyst  to  alumina
coated  steel  wool.   Although  the torch performed  satisfactorily
using either propane or diesel  fuel,  propane  was used in most of
the development work because of  experimental convenience.

To date, about  3,000  miles have been  accumulated on a 1980  Olds-
mobile  vehicle  equipped with  a diesel  exhaust  filter.   During
this  test,  the  filter  was  regenerated at 150-200 mile intervals
with  our catalytic  torch  and  a trapping  efficiency  of  greater
than  65%  was  maintained throughout  the entire  test.   Since  the
amount  of  heat  released  from  burning  soot  in  the  limited DEF
volume  can be  large, the regeneration interval is determined by
this  factor  rather than by the potential  for high backpressure
due to filter loading.

In  addition  to  the durability  experience achieved  with diesel
engines, it is worth noting that alumina coated mesh filters have
undergone  millions of  miles  of  durability  testing  as  a lead
particulate  filter  in  the exhaust   of  a  variety of   gasoline
engines.
                                554

-------
                                     ALUM IN A-COATED METAL WOOL

                                           SUBSTRATE
                      INSULATION
en
en
en
                 INLET

                    GAS
                  OUTLET

                   GAS
PERFORATED
BAFFLES AND
RETAINERS
                                   GAS
                                   SPREADER
                 FIGURE 1 - Typical Texaco Diesel Exhaust Filter Design

-------
             ISOLATION AND IDENTIFICATION OF MUTAGENIC NITROARENES
                        IN DIESEL-EXHAUST PARTICULATES

               >                       by

              X.B. Xu*. Joseph P. Nachtman. Z.L. Jin*, E.T. Wei,
                             and Stephen Rappaport
          Department of Biomedlcal and Environmental Health Sciences
                            School of Public Health
                    University of California, Berkeley, CA

                                      and

                                A.L. Burlingame
                           Space Sciences Laboratory
                  University of California, Berkeley, CA, and
                    Department of Pharmaceutical Chemistry
                              School of Pharmacy
                  University of California, San Francisco, CA


     Particulate matter emitted from diesel engines contains chemicals which
are active in the Ames Salmonella typhimurium assay.  The major portion of this
mutagenic activity is liver enzyme independent and thus indicates that diesel
exhaust contains a different class of mutagens than unsubstituted polynuclear
aromatic hydrocarbons, which are liver enzyme dependent.

     Diesel exhaust particulates were collected on glass fiber filters from
heavy-duty engine test apparatus.  A total of 14 filters were extracted for 24
hours with 6 L dichloromethane in a Soxhlet apparatus, the extract filtered and
concentrated by rotary evaporator.  The combined extracts contained
approximately 225 g of organic matter which yielded 0.46 net TA98 revertants/ug
(1.0 x 108 net TA98 revertants).

     The CH2C12 extract was further fractionated on a preparative silica column
with successively increasing solvent strength:  hexane, chloroform, and
methanol.  The intermediate polarity fraction had the highest specific
mutagenic activity and was further separated on gel permeation, high
performance normal and reverse phase chromatography.  Mutagenic activity was
detected in virtually all fractions, so that fractions containing the highest
specific activity were selected for further analysis.
*Members of the Institute of Environmental Chemistry, Chinese Academy of
 Sciences, Beijing, People's Republic of China.
                                     556

-------
     High resolution mass spectrometry (HRMS) was performed on selected
subtractions from the reverse phase separation.  Each sample was evaporated
under N2 in the probe which was Inserted into the electron Impact source.
Identification of nltroarenes was based upon accurate mass determinations  (± 15
ppm at a resolution of 9000) of the molecular ion and of fragment ions
corresponding to losses of the neutral fragments NO and/or NO?.   Some
nitroarenes tentatively identified are listed in Table 1.  Those with an
asterisk (*) were found to have the same HRMS spectrum and high  pressure liquid
chromatography (HPLC) retention volume as their corresponding standard.

     In this study, about 50 nltroarenes have been tentatively identified  by
HRMS.  The variety of nltroarenes in dlesel exhausts is extensive and, because
only a few of the mutagenic fractions were examined, it is likely that more
nitroarenes will be characterized.  This complexity is not surprising if one
considers the numerous PAH substrates available for aromatic ring nitration:  a
chemical process which will readily occur in the presence of even low
concentrations of nitrogen oxides.  Positive identification of each nitroarene
1s made difficult by the small quantity of nitroarenes relative  to co-eluting
oxygenated and some sulfur-containing PAH and also by the absence of synthetic
standards.  Comparison of mass spectra and retention volumes with available
standards, however, support the identifications which have been  suggested.

     A number of nltroarenes are potent mutagens in the Ames Salmonella assay,
because nitroreductases 1n the tester strains facilitate reduction of
nitroarenes of electrophilic intermediates which, in turn, react with nucleic
acids.  This raises the question of whether nitroarenes in diesel exhaust  or
any other environmental source pose a significant human health hazard.  From a
qualitative viewpoint, nitroarenes are hazardous because compounds such as
2-nitronaphtalene and 2-nitrofIuorene are carcinogenic to animals.  However,
from a quantitative viewpoint, issues of the human dose and the biologic
potency of nitroarenes have not yet been established.  It is also not known if
nitroarene emissions are independent of or bear a reciprocal relationship to
the amount of unsubstituted PAH that is emitted.  The development of convenient
techniques to separate and quantify nitroarenes and metabolites  may expedite
the acquisition of data to answer these questions.
                                      557

-------
                       TABLE 1




 Mas a	Possible Compound
161.048          Nitroindene
  >

197.048          Nitroacenaphthylene


199.063          Nitrobiphenyl»


211.063          Nitrofluorene»


213.079          Nitro-methylbiphenyl


223-063          Nitroanthracene*


225.079          Nitro-methylfluorene


247.063          Nitropyrene*


261.079          Nitro-methylpyrene


287.095          Nitro-methylohrysene


225.043          Nitro-fluorenone*


241.074          Nitro-hydroxymethylfluorene


253.038          Nitro-anthraquinone


256.048          Dinitrofluorene*


340.143          Dinitro-(Cg)alkylfluorene


348.111          Dinitro-(C^)alkylpyrene


371.112          Trinitro-(Cg)alkylfluorene
                       558

-------
     COMPARISON OF NITRO-PNA CONTENT AND MUTAGENICITY OF DIESEL EMISSIONS

                                      by

                   Marcia G. Nishioka and Bruce A. Petersen
                        Battelle Columbus Laboratories
                                Columbus, Ohio

                                      and

                                Joel!en Lewtas
                      Health Effects Research Laboratory
                     U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina


INTRODUCTION

     The increasing number of automobiles and light-duty trucks powered by
diesel engines has generated concern over the emissions associated with these
engines.  The diesel engines have a higher particulate emission rate than
gasoline catalyst engines.  The extractable organics from both diesel and
gasoline particle emissions have been found to be mutagenic and
carcinogenic (1).  Recently, nitro-substituted polynuclear aromatic
hydrocarbons (nitro-PNA) have been identified in diesel particle extracts (2).
Several of these nitro-PNAs are very potent bacterial mutagens (3).

     A study was carried out to identify and quantitate nitroaromatic compounds
in the extract of particulate material from three different diesel engines
(Datsun Nissan diesel 220C, Oldsmobile diesel 350, and VW Rabbit diesel) and
one gasoline engine (Ford Mustang II).  The operating and sampling conditions
have been described elsewhere (4).  Mutagenic assay data was also collected on
these extracts using the Salmonella typhimurium TA98 bioassay.  The results of
these two studies were compared to determine whether the amount of
nitroaromatics detected can fully account for the direct-acting mutagenic
activity indicated by the bioassay data.


METHODS AND RESULTS

Chemical

     Two separate methods were developed to analyze the emission extracts:
1) combined high performance liquid chromatography/mass spectrometry (HPLC/MS)
with positive chemical ionization (PCI); and 2) on-column injection high
resolution gas chromatography/mass spectrometry (KRGC/MS) with negative
                                      559

-------
chemical 1on1zat1on (NCI).  The total dichloromethane extract was screened
first by the PCI HPLC/MS method, which consisted of a Supelco normal phase HPLC
column (Supelcosil LC-Si, 5 uti) interfaced to a Finnigan 4000 mass spectrometer
via a microbore capillary and a polyimide moving belt.  Nitropyrene and
nitrophenanthrene/anthracene were detected by this method in all of the diesel
engine extracts, but not in the gasoline engine extract.

     The extracts were fractionated on silica gel by open-bed liquid
chromatography into four compound class fractions:  1) hexane - aliphatic
hydrocarbons, 2) hexane:benzene - polycyclic aromatic hydrocarbons and
mononitroaromatics, 3) methylene chloride - moderately polar neutrals,
including di- and trinitroaromatics, and 4) methanol-highly polar neutrals,
primarily oxygenated compounds.  Over 94% of the total extract was recovered
for all four engines.

     The three fractions expected to contain nitro-PNA (2,3,and 4) were
analyzed by the NCI HRGC/MS system, which consisted of a JaW DB-5 bonded fused
silica capillary column interfaced to a Finnigan 4000 mass spectrometer.  The
quantitation of nitroaromatics was based on response factors for eight standard
nitroaromatics relative to the internal standard dy-nitronaphthalene calculated
over a concentration range of 100.  More than twenty nitroaromatics were
detected 1n the diesel engine extracts but only 1-nitropyrene was detected in
the gasoline engine extract.  In all cases the l-n1tropyrene was the
nitroaromatic detected in greatest quantity and Its concentration in the
extracts is shown in Table 1.  Quantitative data on the other nitroaromatics
will be presented.

     At masses higher than nitropyrene, the mono nitro derivatives of the
molecular weight 228 (benz[a]anthracene, chrysene) and molecular weight 252
(B[a]P, B[e]P, perylene, benz[o]fluoranthenes) PNAs were also detected.  Two
nitropyrenone isomers were tentatively identified in the Nissan and Oldsmobile
samples and three dlnitropyrene Isomers were Identified in the VW sample.

     Most of the compounds detected 1n the methanol fraction of the Nissan and
VW samples were quinones.  The methanol fraction represented 30% and 17% of the
total organic mass for the Nissan and VW extracts, respectively.


Biological

     The total dichloromethane extractable organics from each of the emission
samples were bioassayed in the Salmonella typhimurium plate incorporation assay
with minor modifications as reported elsewhere (5).The slope of the
dose-response curve (rev/vg) for each of these samples with and without S9
activation was determined using a non-linear model (6) and is shown in Table 1.
The emissions from the gasoline catalyst (Mustang II) differ from the diesel
emissions in that these organics are significantly more mutagenic in the
presence of the S9 activation system.
                                      560

-------
DISCUSSION

     Both the PCI HPLC/MS and NCI HRGC/MS methods are capable of detecting
nitroaromatic compounds.  However, the greater chromatographic resolution and
lower detection limit of the NCI HRGC/MS method favor the use of this method
over the PCI HPLC/MS method.  The limit of detection by NCI HRGC/MS is
approximately 0.05 ng for the mononitroaromatics, but only about 100 ng for
nitropyrene by the PCI HPLC/MS method.

     The concentration of 1-nitropyrene detected in each of the samples in not
highly correlated with the direct-acting mutagenic activity.  The total nitro-
aromatic content does account for a substantive portion of the direct-acting
mutagenicity of the Olds diesel, VW Rabbit diesel, and Mustang II gasoline
vehicles.  However, the nitroaromatics detected cannot account for the signifi-
cantly higher mutagenic activity associated with the Nissan diesel extract.

     It is possible that the quinones detected in the Nissan and VW extracts
are the oxidation productions of nitroaromatics in the stored extracts, as
similar extracts were shown to increase in toxicity with time (6).  A greater
concentration of quinones in the Nissan sample may indicate that a greater
concentration of nitroaromatics may have been originally present in the
extract.  This is consistent with the fact that the Nissan extract originally
had higher mutagenic activity than the VW extract.


REFERENCES

1.   Nesnow, S., and J.L. Huisingh.   1980.  Mutagenic and carcinogenic potency
       of extracts of diesel and related environmental emissions:  Summary and
       discussion of the results.  In:  Health Effects of Diesel Engine
       Emissions, Vol. II.  W.E. Pepelko, R.M. Danner, and N.A. Clarke, eds.
       EPA-600/9-80-057b.  U.S. Environmental Protection Agency:  Cincinnati,
       OH.

2.   Petersen, 8.A., C. Chuang, W. Margard, and D. Trayser.  1981.
       Identification of mutagenic compounds  in extracts of diesel exhaust
       particulates.  Proceedings of  the 74 annual APCA Meetings, Philadelphia,
       PA.

3.   Rosenkranz, H.S., E.C. McCoy, D.R. Sanders, M. Butler, O.K. Kiriazides,
       and R. Mermelstein.  1980.  Nitropyrenes:   Isolation, identification and
       reduction of mutagenic impurities in carbon black and toners.  Science
       209:1039-1043.

4.   Huisingh, J.L., R.L. Bradow, R.H. Jungers, B.D. Harris, R.B. Zweidinger,
       K.M. Cushing, B.E. Gill, and R.E. Albert.   1980.  Mutagenic and
       carcinogenic potency of extracts of diesel  and related environmental
       emissions:  Study design, sample generation, collection, and
       preparation.  In:  Health Effects of Diesel Engine Emissions, Vol.  II.
       W.E. Pepelko, R.M. Danner, and N.A. Clarke, eds.  EPA-600/9-80-057b.
       U.S. Environmental Protection  Agency:  Cincinnati, OH.   pp. 788-800.
                                      561

-------
     Claxton,  Larry D.   1980.   Mutagenlc and carcinogenic potency of dlesel  and
       related environmental  emissions:   Salmonella bioassay.  In:  Health
       Effects of Diesel  Engine Emissions, Vol. II.  W.E. Pepelko, P.M. Danner,
       and N.A. Clarke,  eds.   EPA-600/9-80-057b.  U.S. Environmental Protection
       Agency:  Cincinnati, OH.  pp. 801-807.

     Huisingh, J., R.  Bradow,  R. Jungers, L. Claxton, R. Zweidinger, S. Tejada,
       J. Bumgarner, F.  Duffield, V.F. Simmon, C. Hare, C. Rodriguez, L. Snow,
       and M.  Waters.   1979.   Application of bioassay to the characterization
       of dlesel particle emnissions.  Part II.  Application of a mutagenicity
       bioassay to monitoring light-duty diesel particle emissions.  In:
       Application of Short-term Bioassays in the Fractionation and Analysis of
       Complex Environmental  Mixtures, Environmental Science Research, Vol.  15.
       M.D. Waters, S.  Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
       Plenum  Press:  New York.  pp. 400-418.
     Table 1.  Quantification of l-N1tropyrene in Engine Exhaust Extracts
               and Mutagenic Activity of the Extracts
                                                     Mutagenic Activity, rev/ug
                        Cone. l-N1tropyrene
      Sample                  wt ppm                 TA98/-S9          TA98/+S9


Nissan diesel                  407                     20.8              15.1
Olds diesel                    107                      2.1               1.4
VW Rabbit diesel               589                      5.2               6.1
Mustang II gasoline              2.5                    2.1               8.6
                                      562

-------
1-NITROPYRENE EMISSIONS FROM FIVE PRODUCTION
MODEL DIESEL VEHICLES AND THE EFFECT OF
DAMPING VALVE ON THE EMISSION
              OCTOBER 1981
           NISSAN MOTOR CO.,  LTD,
                   563

-------
Background

      At the EPA Symposium on Health Effects of Diesel  Engine held
   in Cincinnati in December 1979, EPA and other organizations re-
   ported that a particulate exhaust sample obtained  from  a Nissan
   car had a very high BaP concentration.  The cause  of this high
   BaP level was investigated by Nissan and reported  to EPA. (A copy
   of the report is attached to this handout.)

      In March 1981, at CRC meeting, General Motors reported a very
   high 1-nitropyrene concentration in a sample from  a  Nissan diesel
   car.  This sample was collected by EPA and its extract  was dis-
   tributed to various laboratories.  The sample is sometimes called
   NI or NI-1 and is believed to be the same sample that was discuss-
   ed in the last Symposium.

      This is a preliminary report on 1-nitropyrene   emissions from
   Nissan .and other manufacturers' recent production  vehicles.  Possi-
   ble cause of high 1-nitropyrene emission is also'investigated.


Experimental Method

      Exhaust particulate samples from five diesel vehicles were col-
   lected using a chassis dynamometer and a dilution  tunnel on 20" x
   20" Pallflex T60A20 filters.  Samples were extracted with
   dichloromethane.

      The extract was treated with a reducing agent to  convert nitro-
   pyrene to aminopyrene.  Liquid chromatograph separation and fluo-
   rescence detection and measurement of the aminopyrene were per-
   formed by comparison with standard solutions.


Results

      Figure 1 shows 1-nitropyrene emissions of five  test  vehicles
   under Highway Fuel Economy Test cycle.  Datsun 810 and  Datsun
   Pickup are 1981.5 MY production models for the U.S.  Federal market.
   Car A is a 1979 Federal production model.  Cars B  and C are 1979
   production models for Japanese market.  Cars A,B and C  are not
   Nissan's product.  Datsun Pickup was also tested by  the South-
   west Research Institute and the data is included in  the figure.

      Sample NI-1 was collected by EPA from a Nissan  prototype car
   equipped with a 2.2iengine.  The car was sent to  EPA in 1973 and
   the sample was collected sometime in 1979.  This sample was ana-
   lyzed by GM and the result was reported in the CRC  Diesel Exhaust
   Emission Workshop in March 1981.  The same sample  was also measured
   by Nissan.

      Sample NI-1 shows a very high 1-nitropyrene concentration  by
   both GM and Nissan measurement.  Compared to NI-1,  samples from
   five production models show lower levels.  Although  Datsun Pickup


                                564

-------
   is equipped with the same type engine (SD-22)  as for the proto-
   type car provided to EPA, its 1-nitropyrene emission is the lowest
   among five vehicles.

      As in the case of BaP, secondary injection was suspected to
   be a cause of high 1-nitropyrene emission.  Secondary injection
   or injector "bounce" makes extra fuel injected late in the com-
   bustion _process and results in high hydrocarbon emission.  Damping
   valves installed in the fuel line are known to eliminate the sec-
   ondary injection.

      Figure 2 shows the effect of damping valve on 1-nitropyrene,
   BaP and HC emissions.  As expected, all three emissions decreases
   both in FTP and HFET when damping valves are installed.  As shown
   in Figure 1, Datsun 810 and Pickup are equipped with damping valves
   whereas the prototype car provided to EPA had no damping valves.
   Car C which has damping valves shows low emission.   Car A which
   has no damping valves gives highest 1-nitropyrene emission.  How-
   ever, it is too early to say that damping valves are indispens-
   able to reduce 1-nitropyrene because Car B shows a  modest level
   without damping valves.  And when damping valves were taken off
   from the Datsun Pickup, 1-nitropyrene emission did  not increased
   to the level of sample NI-1.

      Other factors to increase secondary injection such as larger
   diameter injection tube and fuel injectors with carbon deposit
   are worth examining to reproduce the 1-nitropyrene  level of NI-1.
   At the same time, sampling conditions such as gas temperature,
   N02 concentration and total gas volume passing through the filter
   should be investigated.


Summary

   1.   1-nitropyrene emissions from five diesel powered production
     vehicles were measured and compared with sample NI-1 obtained
     from EPA.  Samples taken from all five vehicles show consider-
     ably lower 1-nitropyrene concentration" than NI-1.  From these
     data NI-1 is not believed to be a proper sample to represent
     the production model diesel vehicles, even less it represents
     Nissan's current products.

   2.   Damping valves installed in a fuel line have an effect to
     reduce 1-nitropyrene emission with a certain model of engine.
     Further investigation  is necessary to clarify the cause of
     unusually high 1-nitropyrene concentration in EPA sample NI-1.
                               565

-------
               Figure 1
en
01
o-i
       1.2
     | 1.0
     X
     LU



     I 0.8
       0.6
     O)
     l_
    f 0.2
        0
                         efy™** X^K
                         no
              ssjon
           from  Diesel-Powered Vehicles
            Test Fuel: Type 2D

                   (C.I.45)

            * D.V. : Damping Valve
HFE

I.T.
7°B1DC
2o°ErrDC

3°ATDC
0°BTDC
7°BTDC

,ve
U
a
^
T



i

\
^
\





EGR
W
w/n
/ vy
yv/o
%>
^v/6

X '•'•'•'•















•X



\
\






D.V
W
W

y%
%
W





.











I.W.
3,500lbs
3000

2,250
2,750
2,750


JSL.
























i











           Datsun
Datsun

Pickup
A
    NI-1

   by   by

Nissan  GM

-------
             Figure 2
    Effect of Damping Vatve on l-Nitrojpyrene, BaP & HC
en
cr,


o
X




Q_
o
CD



c
l_
CL*
~z.
1

~ 0.6
1 0.4
O^
- 0.2
FTP
ap 	
4»
i»

0.08
~ 0.06
u
2 0.04
X
5 0.02
0

^

^




*
D

»
A


P
• ///
y^




0.03r
r< 0,06
UJ
^ 0.04
^ 0.02
n
>

o*
9*
	 {7777/1








••;•;.'
• V •











' *f . • .




















0.6
0.4
0.2
   0
0.08-
006
0.04
0.02
   0
0.08
0.06
0.04
0.02
                                                      Vehicle :  Datsun Pickup
                                                      I.W.   :  3.0001bs
                                                      Engine :  SD22(Acyl.2.2i)
                                                      Injection . 2
                                                      Timing
                                                      Injector : OSD193(New)
                                                      Test Fuel  : Type 2D
                                                              (C.I.45)
                                                       Of
                                                      _>
                                                       a
                                                       (D
                                                       CU
                                                             D.V.
                         Secondary Injection
                                 \
                                                           Crank Angle
                W D.V. vy6 D.V.
         D.V. vv/o D.V.

-------
ANALYSIS OF THE FACTORS AFFECTING UNUSUALLY
HIGH BAP EMISSION FROM A NISSAN SD-22 DIESEL
ENGINE VEHICLE OBSERVED AT EPA
             OCTOBER 1981
       NISSAN MOTOR CO,, LTD.
    NISSAN DIESEL MOTOR CO,, LTD,

                  568

-------
Background
                                                           -»
                Diesel  Vehicle mounted with a SD-22, 4 cylinder
                                                extracts also scored


      This experimental vehicle was sent to the EPA in reply to "the
   request by Mr. E. Stork in 1973.  The vehicle was equipped with
   an engine which had the fuel injection system modified from the
   1973 Japanese domestic specifications.

      We were very concerned about the EPA test results, and Nissan
   engineers visited the EPA ESRL at Research Triangle Park shortly
   after the Symposium in order to study the vehicle.  This vehicle
   was sent back to Nissan, and the causes of these problems were
   investegated by our laboratory in Japan.


Objective

      The objective of our study was to determine the causes of. the
   high BaP emission level.


Experimental Results

      The study described in this paper was done using the 2-D fuel
   nade by Nippon Sekiyu.  Table 1 shows the specifications of the
   test fuel in comparison with those of the EPA requirement and
   the EPA test fuel made by Union. Oil.

      BaP analysis was performed using the High Performance Liquid
   Chromatography (HPLC;  Hitachi Model 635) .

   (1)  Emission Confirmatory Tests

      After the SD-22 vehicle was returned to Nissan, emission con-
   firmatory tests were conducted as received condition.

      As shown in Table 2, the test results indicated that:

       • BaP and EC emissions at Nissan showed a large difference
         from the results of EPA test.

       • The BaP level at Nissan was within the range of other com-
         panys'  vehicles presented at the EPA Symposium in Cincinnati.
                               569

-------
(2)  Relationship between BaP and HC emissions

    The tests were conducted with parameters of new and aged in-
 jectors and with and without damping valve.

    It can be said that there is a direct proportional relation-
 ship between BaP and HC emissions, as shown in Figure 1.

(3)  Factors Influencing on BaP Emission

   (i)  Effect of Injection Timing

       The injection timing of the vehicle as received was 18°
    3TDC against the original setting of 20-°.  Our test results
    showed there was no significant difference in BaP emission
    within the range of 20 i 2° BTDC.

   (ii)  Effect of Aged Injector

       A comparison study of new and aged injectors was conducted.
    HC and BaP emissions with the aged injectors implied the exis-
    tence of deterioration.

       The obtained results are shown in Figure 2.  Although par-
    ticulate emission did not increase by the aged injector,  a
    marked increase in BaP emission was observed.

       It is thought that the high BaP emission was caused by dete-
    rioration of injector which led to the increase of the secon-
    dary injection.

   (rii)  Effect of Damping Valve

       Testing with damping valve resulted in a large scale re-
    duction of both HC and BaP.  As shown in Figure 2, the damp-
    ing valve is more effective to reduce HC and BaP when the aged
    injector is used.  For particulate emission, however, no sig-
    nificant diference was seen by use of the damping valve.

       Figure 3 and 4 show the fuel injection rates for without
    and with damping valve respectively.  The large amount of
    secondary injection was seen when the damping valve was not
    used.   From these figures it is found that the application
    of damping valve eliminates the secondary injection.

       The drawing of damping valve is shown in Figure 5.

   (iv)  Effect of Injection Tube Inner  Diameter

       The effect of injection tube inner diameter on the secon-
    dary injection rate and emissions was studied using the dia-
    meter of 2 mm and 3 mm.  The 3  mm  tube was equipped in the EPA
    tested vehicle.
                            570

-------
   tube ?nner d?am^,   %Sf °ndary  in^ction rate  for the injection
   rate Jas observ  H       I™'   Extremely h^  secondary injection
   in Fia»r* 7  4,   ^ Sntlre  6ngine  0Peratin9  conditions.  As shown
   r^l fn ^K   - ^    • WaS a marked  reduction of secondary injection

   of 3 ™ diLet™ tubl! dlameter  ^^ ^ COI^1S°n With the
      HC and BaP emissions  seen  in Figure 2 decreased significantly
   when the 2 mm diameter tube was used.  In addition, it is found
   that the system with  damping  valve and 2 mm inner diameter injec-
   tion tube does eliminate  the  high BaP problem even when    aaed
   injectors are used.


Summary of the Study

   •   HC and Bap emission  levels of the SD-22 vehicle tested in
     Nissan were much lower  than results of EPA testing.

   •   This BaP level at Nissan  was the same level as other compa-
     ny's vehicles  presented at the EPA Symposium in Cincinnati.

   •   The direct proportional relationship between HC and BaP emis-
     sions was observed.

   •   BaP emission increased significantly when clogged injectors
     were used, due to their secondary injections.

   •   The system with damping valves and smaller inner diameter
     injection tubes is  effective to reduce the secondary injection
     and does eliminate  the high BaP emission problem even when clog
     ged injectors were  used.


BaP Emission from Nissan's Current Diesel Vehicles

   (1)  Comparison of Fuel Injection System

       Table 3 shows the comparison of fuel injection system between
   the  EPA tested vehicle and the improved specification vehicle.
   This improved specification,  which is being used in Nissan's cur-
   rent U.S.  models, includes damping valves, new type injectors,
   and  2 mm diameter injection tubes.

   (2)  BaP Emission Level
       The emission test results of a vehicle with the improved fuel
   injection system are shown in Table 4.

       As a result, the current improved system eliminates the high
   BaP emission problem, even when aged injectors are used.
                               571

-------
Conclusions

      It is concluded that the high BaP emission rate of  the  SD-22
   diesel engine is not a problem inherent in all  SD-22 diesel en-
   gines, but rather is peculiar to this engine alone.

      The problem should be considered in light of the following
   factors;

       (1)   The vehicle in question was originally sent  to the EPA
          for testing as far back as 1973.

       (2)   The high BaP emission rate seems to be caused by the
          secondary injection of the injectors-.

       (3)   The details of the engine maintenance performed  by EPA
          are unknown to us, but the injectors probably clogged up
          and this promoted the occurance of secondary injection.

       (4)   The fuel injection system of the 1973 model  was not equip-
          ped with damping valves, since they were not available at
          that time.

       (5)   In view of the high HC and BaP levels,  Nissan's  current
          models adopting the improved fuel supply system eliminate
          them.
                               572

-------
         Table 1



Test Fuel Specifications
Specific Gravity 15/4 °C
Viscosity cst @ 37.8 °C
Cetane Index
Sulfur wt %
Distillation °C
IBP
10%
50%
90%
EP
Aromatics vol %
Olefins vol %
Saturates vol %
EPA Requirement
0.8393 - 0.8597
2.0 - 3.2
45 - 50
0.2 - 0.5

171 - 204
204 - 238
243 - 282
288 - 321
304 - 349
> 27
-
-
2-D by
2-D by Union Oil* Nippon Sekiyu
0.847 0.845
2.3
48 44.3
0.16 0.25

182 202
219 224
262 249
316 291
339 ' 322
32.5 27
1.3
66.2
 * used at EPA RTP ( SAE Paper 790422 )

-------
tn
~^J
-p.
                                          Table  2
                  BaP Emission  from  a  Nissan  Diesel  Vehicle  (  Test Results at Nissan )
                                                 VEHICLE AS  RECEIVED FROM EPA

                                                 I. W.    :   3,500  Ibs

                                                 Engine   :   SD-22  (  4 cyl.,  2.2 liter )
Mode
LA- 4 Hot
HFET
HFET *
( EPA Data )
Part
(gpra)
0.25
0.33
0.33
%
Extract
16.5
13.9
5.3
BaP **
(^g/rai)
2.87
0.87
20.5
HC
(gpm)
0.22
0.13
0.85
CO
(gpm)
1.51
0.80
1.31
NOx
(gpm)
1.68
1.48
1.08
mpg
28.6
30.5
32.7
                                                                              (  n = 2 )
                         **
 From the data presented at the EPA Symposium and
BaP emission is calculated by Nissan


 Sampled using 8" x 10" Pallflex T60A20 filter and
analyzed by HPLC ( Hitachi 635 )

-------
                Figure 1
       Re[alion5^ijD_between HC&BaP
       Test Mode;HFET
       Engine ^
       I. W.: 3,500 Ibs
Test Fuel: Type 2D
       (Cetene Index 44.3)
Injector :CSD 211
       New a Aged(23,500 km)
Damping Valve: w/ £
u
 J-  100
-o
N

~o
c
 CD-
CL
 n^
CD
              0.5      1.0      1.5     2.0

                 H C  (9/miie)
                   575

-------
                    Figure 2          •
Effect of Mileage Accumulation of Injectors
and Damping Valve on  HC. Part.& BaP
         Test Mode : HFET
         Engine : SD22
         I.W. : 3,500 Ibs
                            Test Fuel :Type 2D
                            Injector   : OSD 211
                            Injection Timing : 20°BTDC


1 n-
I.U
S "a)
X c
£ 5^
<-»' n c
T3 	 U.O
b o
Ul -I-
s x
2 n.









C^Wy
••'•v.V-:-
"-: ."';•;.
* -:•*..
*v **•*•"-
v.v /
ifi













v"vi:^
••-' •*"•'
"*""* V***
ft-\'- •'•'
,'V.-*tV
.* ,""••.•"•'







•
Y/A














I 	 1
      0.5r
   ^ O.A
q

CL
i

^
•o
a
N
J>>
O
<
      0.2
      o.i
       •o
       20
0
0
            17-^-T-"
                     EPA Observed Level (20.5/4a/mile)
                     ii range of data
                        (n = 3)
                             i = 2)
                                       range of
                                       (n=3)
               data
            ;n = 2)
           ._3_CD!ILL2mm_[_2_mm	;_  2mrn ; 2mm
           Injection Tube Inner  Pi a.
                D.v.
                          QV.
            New  Injector
w/o D.V.  w/ D.V.
                              Aged Injector
                                                   576

-------
                           Figure 3

"i
i _ l_ L_ i_. I. -I jl_-'
—r
i
^
f --
Jl
^s^^~~" /
_i
1
5^"
K ."<•
11
%l
_ _ \
/Av.
?\V

5 ^ : 20 :!30 I
;
i
i
i
J
:j
i
l
or
w^r
-.."7
--V_.V
~>r • r
N
[5
r>,
.x-^ ^-
\_ j V^J 1
'T
/
	 \
— k r- <-** r-
C
i -^
C
C
(
•^cor
•-i
i
1
icli
0 f60 '70 80'
i
i"*™***
i — ^
?r
"9
M ii. U..
/ Injc
! -~ 1 '
I InjectionRa
i , •'
2400 rpm
• 	 ~* j
i .
ii
te. 1,
.— •-- -v 	 1'^

^ctioni;
.'.NeedteLift _

0 160 deg.
Injector
Ini Ti ih(=>
                   -Main Injection
                               -r.r
     Inj.TubeDia.3mm
     Wjjjnout Damping
               Valve
20
30
,il
J I
 I,-
40
                          i-^J--'^^ . b
50     60
IjlJi ii..^-.*!^^^ .wa^t, ^

70   . 80
90     100  deg,

-------
                       Figure 4
1 • •-•
i ......
1
, i
J— 	 "~;
1 ' ' !
r- f • f— r~* r~~*
• i
r
t
s~

-,f~
y
J
,.-.•• -
,,
/
^/
L$
-' / » i
~X . IL
^
77
/ '"!
/ •;. ,
rjr::.-1(.
	 —
^
^
•i.l V k, •
1

t
I
i
i
I
i
y
•-^
i
-^
r"
"~>->-.-
r<
air
. V.. <
~r™ i

. . . • 1
i
i In:




ect
» £-<• r*» >
•
Tp i'' ' • -r-
•'. . 1. . ft. • . 1- 	 	
•>r !; ; i. ' i:
-, r- ! ! i !, | :
X Fuell.Injection Rate- J^— — J'
!: r ;! i' r ;'
>f :' .- ' :.:
<& . :; i: 2/>OOrpm
H !': i' 0^20 mm/£t.
•
       20  30  AO  50 60 70 deg.  '
Fuel Spigy	__,,,,,,	,,...,	

                    (V
     \\.
       20
30
 ..
50
                                                 Injector OSD193
                                                 Inj.Tube Dia. 3mm
                                                 With Damping Valve
60-.    70"deg; •

-------
           Figure 5
The Structure of Delivery Valve
  Damping Valve
   Damping J/ajye.
 mm%zz.'.¥z'-
                      £/
                 579

-------
                    Figure 6
Mop of Secondary Injection Rate
   Inj.TubeDia.  3mm       Secondary Inj. Rate = -fi-"100%
   Without Damping Valve                      w
(J
^ 50
—4 — •
^30
O
S20
i — t
- 1 n
CD IU
fc °
injector- u^u zn Lift






^>
2



k
5-30

•


/o \




22-2
\ (
j
o
^
i; <•/
O/o
20-2?


\
^ \
/
;
C7

75 \
? ^
30
V i
/
- y. ..\ /rr-nf
/ix

^^~-




^-




^^




^
x^i— -
I 3 tr 0/
J5/0

10 12 14. 16 18 20 22 24 26 28 30 32
Engine Speed (rpm) xlOO

-------

• 'x
E
JE
 QJ
|30
 C
o
QJ
OJ
£ o
                    Figure 7
Mop of Secondary Injection Rate

   Inj. TubeDia.1. 2mm
   Without Damping Valve
   Injector : OSD 211
                                                    h
                                 Secondary Inj. Rate= ~-^00%

        0
        10
                             10
                              20
                                  0
     10   12  14
               16   18  20  22  .24   26   28  30  32
                Engine Speed (rprn)               xlOO

-------
              Table 3
Comparison of Fuel Injection System
Item
Inject!
Injector
on Pump-
Type
Clearance*
in DIG.
Injection Tu be Dia.
EPA i es ted
Vehicle
WiihoutD/V
OSD211
31-41^
3 mm
Improved
Spec,
WithD/V
OSD193
28-34;j
2 mm
           \
Clearance-
                        Needle
                   7
                   V   Body
                    Carbon Deposit
                  582

-------
                                           Table  4
           Emission Test  Results  of  the  Vehicle with Improved Fuel Injection System
                                         TEST MODE :  HFET
in
Co
OJ
Injector
New Injector
Aged Injector*
Part. .
(gpm)
0.27
0.23
0.25
0.24
BaP
(pg/mi)
0.73
0.53
0.44
0.41
HC
(gpm)
0.11
0.08
0.07
0.06
CO
(gpm)
0.48
0.47
0.47
0.46
NOx
(gpm)
1.54
1.58
1.52
1.49
mpg
34.0
32.8
32.4
32.3
                                  *  Aged injector :  AMA  50,000  miles

-------
        CAPILLARY COLUMN GC/MS CHARACTERIZATION OF DIESEL

                    EXHAUST PARTICULATE EXTRACTS1

                                     by

                     T. J. Prater, T. Riley and D. Schuetzle
                        Analytical Sciences Department
                        Scientific Research Laboratory
                           Dearborn, Michigan 48121

INTRODUCTION

Previous studies have shown that most of the direct-acting Salmonella typhimurium
mutagenic activity £50%) in diesel exhaust particulate  extracts is  concentrated in
chemical fractions which contain compounds of moderate polarity (2).  Analytical-
scale normal phase high performance liquid chromatography(NP-HPLC) (3), packed-
column  GC/MS and high resolution  MS analysis were used in that work to determine
that these moderately-polar fractions consisted primarily of polynuclear aromatic
hydrocarbon (PAH)  derivatives.   The objective  of  this study was to  extend  the
previous work  by developing preparative scale NP-HPLC fractionation followed by
fused silica capillary GC/MS analysis in order to improve component resolution.

EXPERIMENTAL

Light duty  diesel exhaust particulate samples  were  collected on T60A20 Pallaflex
filters using a dilution tube and a chassis dynomometer test facility.  Filter samples
were soxhlet extracted with dichloromethane (DCM).

HPLC analysis was performed on a Varian Model 5600 LC equipped with a 7.8 mm i.d.
x 30 cm long  Microporasil  10 n  column.  The solvent  program consisted of 100%
hexane for 5 min, then 1% DCM/min for 5 min, followed by a linear gradient to 100%
DCM in the next 25  min, 100% DCM for 10 min, then  10% acetonitrile/min for 10
min, and  a final  10 min  flush  with  100% acetonitrile.   The  chromatographic
separation was monitored by UV at 254 nm and by fluorescence  at 254/320  nm.
Further details of this technique are presented elsewhere (3).

GC/MS  analyses were performed on  a VGMM ZAB-2F mass spectrometer equipped
with a 30 m long x 0.25 mm i.d. SE54 fused silica capillary column interfaced directly
to the mass spectrometer source.   Samples were injected  directly on-column  and
temperature programming was 80° to 270   at  4°/min.  Electron  impact ionization
techniques were used.
                                       584

-------
RESULTS AND DISCUSSION

This investigation emphasized the analysis of HPLC fractions containing nonpolar and
moderately  polar PAH  derivatives.   Nonpolar aliphatic  and highly polar  HPLC
fractions were not characterized by GC/MS. The nonpolar and moderately polar PAH
derivatives which were identified are listed respectively in Tables I and  H.   Many of
the  compounds  identified  have a  large  number  of isomers as  indicated by  the
parentheses in the tables.   This is illustrated by the mass chromatograms  in  Fig. 1
which show the increasing  complexity of the isomer series as  methyl substituents are
added to anthracene and phenanthrene.  There are  probably  more isomers for the
methylated anthracene and phenanthrenes than we were able to resolve even with the
high-re solution fused-silica capillary column.   It would be difficult and of limited
utility to  identify every specific isomer present in these fractions.  For this reason,
synthesis  of  standards and identification of isomers are  only being  undertaken  on
those groups of  PAH and PAH-derivatives which  yield  a relatively high  level of
direct- or indirect-acting Ames activity.  This has been found to be the case for the
nitrated-PAH  derivatives  which  tend   to  show  high  levels  of  direct-acting
mutagenicity compared to  other PAH and PAH derivatives  in these samples.

The combination of normal phase  HPLC fractionation followed  by  capillary GC/MS
analysis proved to be a  very  useful approach to the qualitative characterization of
diesei exhaust paniculate extracts.  The quantitative analysis  of these  diesel extracts
by  capillary  GC/MS  is complicated by  the  labile nature  of  some  of the  PAH
derivatives.   The quantitation technique found to be most  accurate and  presently
being used in our laboratory employs the deuterated analog of the PAH derivative of
interest which is  added to filter samples prior to extraction. The deuterated standard
exhibits the  same chemical characteristics as  the  native compound, but it can  be
distinguished  mass spectrometricaily.  Therefore, losses which occur during sample
workup and analysis  can  be accounted for.

1.   Prater,  T.  J.,  T.  Riiey, and  D.  Schuetzle.  1981.   Capillary column GC/MS
     characterization of diesel exhaust particulate extracts.  Presented at the 29th
     Annual Conference on Mass Spectrometry and Allied Topics, Minneapolis, MN.

2.   Schuetzle,  D.,   F.S.-C.  Lee,  T.  J.  Prater,  and  S. B. Tejada.  1981.   The
     identification  of  polynuclear  aromatic   hydrocarbon  (PAH)  derivatives  in
     mutagenic fractions of diesel  particulate  extracts.   Intern.  J.  Environ. Anal.
     Chem. 9:93-144.

3.   Schuetzle, D.,  and J. M. Perez.  1981.   A  CRC cooperative  comparison  of
     extraction and HPLC  techniques for diesel particulate emissions.  Presented at
     the 74th Annual  Meeting of  the Air Pollution  Control Association,  Paper #81-
     564, Philadelphia, PA.

4.   Levine,  5. P., and L.  Skewes. 1981.  High performance  semi-preparative liquid
     chromatography of  diesel engine emission  particulate extracts. J. Chromatogr.
     In preparation.
                                         585

-------
Table I. Nonpolar PAH Identified in Diesel Exhaust Particulate Extract

dibenzothiophene
anthracene and phenanthrene
methyl dibenzothiophene isomers(3)
methyl (phenanthrene and anthracene) isomers(4)
dimethyl dibenzothiophene isomers(7)
dimethyl (phenanthrene and anthracene) isomers(13)
fluoranthene and pyrene
trimethyl dibenzothiophene isomers(9)
BaP, BeP, perylene.and isomersO)
trimethyl (phenanthrene and anthracene) isomers(15)
tetramethyl dibenzothiophene isomers(12)
tetramethyl (phenanthrene and anthracene) isomers(16)
benzotgjhji) fluoranthene
benz(a)anthracene, chrysene, benzo(c)phenanthrene,
  triphenylene  isomers(2)
methyl benz(a)anthracene isomers(^)
pentamethyl dibenzothiophene isomers('f)
dimethyl benz(a)anthracene isomers(2)
methyl (fluoranthene and pyrene)isomers(7)

Table n. Moderately Polar PAH  Derivatives Identified in Diesel Exhaust
                Particulate Extract

benz(a)anthracenedione
methyl (anthrone and phenanthrone) isomers
thiozanthone isomer
dimethyl (anthrone and phenanthrone) isomers
pyrenone
trimethyl (anthrone and  phenanthrone) isomers
methyl thioxanthone
dimethyl thioxanthone isomers(2)
benz(d,e)anthrone and isomersO)
1-nitropyrene
1,1'biphenyl-ol
9-fluorenone
(pyrene or fluoranthene) carboxaldehyde
dibenzofuran carboxaldehyde
phenanthrone
anthrone isomer
9-xanthone
xanthene carboxaldehyde
(anthracene or phenanthrene)dione
dibenzothiophene carboxaldehyde
methyl (anthracene or phenanthrene)dione
phenanthrene carboxaldehyde
anthracene carboxaldehyde
methyl (anthracene and  phenanthrene)
  carboxaldehyde isomers(9)
dimethyl (anthracene and phenanthrene)
  carboxaldehyde isomers(8)
                                       586

-------
Figure 1.  Mass chromatograms of phenanthrene(P) and anthracene(A) (178),
          methyl-A and -P  (192), dimethyl-A  and -P (206),  and trimethyl-A
          and -P (220)
         M/Z
          1T8


         M/Z
         192


         M/Z
         206


         M/Z
         220
                   25
                INTENSITY
                     9168
                    11040
                    18860
                     7672
30            36
TIME (MIN)
                                   587

-------
      RESPIRATORY HEALTH EFFECTS OF EXPOSURE TO DIESEL EXHAUST EMISSIONS
         (Bus Garage Mechanics; Salt, Potash, Metal, and Coal Miners)

                                      by

                                  R.B. Reger
                     Epidemlologlcal Investigations Branch
             National Institute for Occupational Safety and Health
                           Appalachian Laboratories
                           Morgantown, West Virginia


     A comprehensive research program has been mounted relating to chronic and
acute respiratory health effects of diesel emissions exposure.  Special
attention has focused on occupational groups exposed in enclosed spaces.  This
study involves over 5,000 workers engaged in various types of mining
occupations as well as bus garage mechanics.  These subjects were given chest
radiographs, asked questions on respiratory symptoms, smoking and occupational
histories, and given spirometric tests.  These data have been coupled with
industrial hygiene information to evaluate relationships between selected
health parameters and component measures of diesel exhaust emissions.  This
paper reports the results for each group of workers separately.
                                     588

-------
         PHYSICO-CHEMICAL  PROPERTIES  OF  DIESEL  PARTICULATE MATTER


                                     by


                     Mark  M.  Ross  and Terence H.  Risby
                    Division  of  Environmental Chemistry
                Department of Environmental Health Sciences
     The  Johns  Hopkins  University School  of Hygiene and Public Health
                        Baltimore,  Maryland 21205

                              Samuel  S.  Lestz
                  Department  of  Mechanical Engineering
                     Pennsylvania  State  University
                   University Park, Pennsylvania  16802

                              Ronald  E,  Yasbin
                        Department  of Microbiology
                      Pennsylvania  State University
                   University Park, Pennsylvania  16802


     Numerous studies have dealt with the identification and quantifi-
cation of the compounds sorbed  onto  Diesel particulate matter.  The
ultimate  environmental  significance  of  these sorbed species depends upon
the relative bioavailabilities  which, in  turn, depend upon the nature
and strength of interactions  prevailing at the gas-solid interface.
This research focused on  the fundamental  adsorptive properties and sur-
face characteristics of Diesel  particulate matter.  In addition, the free
radical nature of  the particles and  the associated reactivity with
selected stimuli was  investigated.

     This study was  carried  out using a graphitized carbon black, Spheron
6, as a "reference"  solid.   Two Diesel  particulate samples were used.
The first, DPM-PSU,  was collected  from  a  single-cylinder engine operated
with a prototype fuel and a  lubricant free of trace inorganic compounds.
The second, DPM-EPA, was  collected from an Oldsmobile 350 engine operated
with a commercial  fuel and lubricant at the US  E.P.A.

     Electron micrographs and elemental  compositions revealed the common
spherical carbon particle structure  of all the samples.   Yet, the Diesel
samples had lower  bulk densities and higher external surface areas, as
calculated from mean particle diameters.  Nitrogen B.E.T.  surface areas
                                    589

-------
were measured in order to determine internal plus external surface areas.
The surface area of DPM-EPA was found to be dependent on outgassing
pretreatment.  The surface area accessible to nitrogen increased from
41 M2/g at 50°C degassing to 112 M2/g at 400°C degassing.  DPM-PSU was
measured to have a surface area of approximately 104 M2/g and showed
little change upon increasing activation temperature. oNitrogen porosity
experiments revealed the existence of pores of 100-200A in diameter in
both Diesel samples.

     Isosteric heats of adsorption of a variety of organic compounds on
the particulate samples were measured with gas-solid chromatography.
The heats on DPM-PSU were consistently greater than those on graphitized
carbon blacks (gcb).  The variations of adsorption energies with adsor-
bate surface coverage were determined by measurement of adsorption
isotherms at different temperatures.  DPM-PSU exhibited adsorption
characteristics similar to those of gcb and the few differences are
attributed to a more polar and energetically heterogeneous DPM-PSU sur-
face.  DPM-EPA was determined to have markedly different properties due
to the increased quantity of presorbed material.  The DPM-EPA surface
was found to be relatively non-polar and homogenous.  Removal of the
presorbed species caused the surface to become more active and similar
to that of DPM-PSU.  The significance of these results is that compounds
close to the carbon surface will be more difficult to remove than those
adhering to presorbed layers.  Sorbed compound bioavailability and sur-
face properties of Diesel particulate matter are dependent upon the
nature and amount of presorbed material.

     A related study of the reactivity of Diesel particulate matter with
respect to atmospheric stimuli was performed using electron paramagnetic
resonance (EPR) spectrometry.  The EPR signals of the three samples were
monitored after selected heat and evacuation treatments, gas (0-, NO,
NO-) exposures, and ultraviolet/visible irradiation.  The sample signals
differed with respect to line widths but all signals narrowed upon sample
evacuation and heat treatment and broadened upon exposure to oxygen and
nitric oxide.  The Diesel particulate sample signals were extremely
sensitive to nitrogen dioxide and irradiation.  Exposure of N0_ caused
an increase in the free radical concentration in the Diesel  samples.
Irradiation effects were varied depending on sample conditions but the
greatest signal increase occurred with evacuated Diesel  particles.   With
EPR, the existence of free radicals in Diesel particulate matter and the
reactivity of these species with respect to selected treatments were
demonstrated.  The results provide evidence of potential photochemical
reactivity of airborne particulate matter.
                                   590

-------
            SOME  FACTORS  AFFECTING  THE  QUANTITATION OF AMES ASSAYS


                                     by


                     Irving  Salmeen and Anna Marie Durisin
                       Engineering and Research Staff
                                   Research
                              Ford  Motor  Company
                              Dearborn, Michigan


     The  simple  theory of bacterial  mutation  expertments of Luria and
Delbruck  (1)  starts  with the  assumption
                                                                        (1)

where m  is the  number  of mutants,  n  is  the number of wild type, and u is the
mutation rate coefficient.   We  have  shown [2)  that, in the absence of cell
killing, this equation  predicts  for  the Ames assay

                                 M = aCN                                (2)

where M  is the  number  of revertants  per plate, a is the mutation rate per
concentration,  C, of mutagen and  N  is  the total number of histidine auxo-
trophs in the background lawn.   Thus the dose-response function is linear,
but the  slope is proportional to N.  In the Ames assay, N ~ n^P, where n0
is the initial  inoculum and  P is the average number of bacteria per background
colony of histidine auxotrophs.

     In a series of experiments  we determined  dose-response curves as a
function of ng; estimated n0 by  counting background colonies in photo-
micrographs (lOOX) of  the background lawn; and estimated the volume of
individual background  colonies which is proportional to P.  We found that P
depends nonlinearily on n0;  P is much larger at low np than it is at high n0
presumably because at  lower  n0 there is less competition among the colonies
for the trace histidine.  We observe that N decreases by about 1/3 when n0
is decreased from 10^  to 10? bacteria/plate.   N is roughly independent of n
when n0 is less than about 5x10° bacteria/plate.  For merely detecting
mutagens, neglect of the dependence of  slopes  on N, while conceptually
incorrect, may  have no  serious effect.  For quantitative experiments, such
as determining  the contribution  of a compound  to the mutagenicity of a
mixture, failure to take into account the dependence of the slopes on N may
                                     591

-------
cause errors by factors of 2 to 3.

     When the test compound causes killing, then equation 0) becomes:


                            {j£- un(t) - kbm                            (3)


where k& is the killing rate coefficient for mutants and, now, n(t) must
include a specific killing term for the histidine auxotrophs.  If we assume
that
                         n(t) = noexp[(.Y - kjt]                        (4)

where y is the growth rate coefficient and k=, is the killing coefficient for
histidine auxotrophs, and if we also assume that killing and mutations are
independent events and that the killing coefficients of auxotrophs and
revertants are equal and proportional to the concentration of mutagen, i.e.,
ka=kD=kC, then the number of revertants per plate is of the general functional
form
                             M ^ ctCN1 expC-kC).                          (5)

where N' is the total number of histidine auxotrophs in the background
population when kC«l .  The concentration which yields the maximum in the
dose-response function is C
     We have obtained dose-response functions using diesel particulate
extract which have a maximum followed by a monotonic decrease to zero.  We
determined the killing coefficient k in three ways:  0) from the value of
the concentration corresponding to the maximum of the dose-response function;
(2) from a classical dilution-plating killing curve; and C3) from a killing
curve developed from counting colonies in photomicrographs of the background
lawn.  The killing rates determined from these 3 methods agree to within
about 15%, suggesting that this simple model is a good approximation to the
mutation-killing kinetics, at least for these samples.

     This model is free of adjustable parameters in the sense that the two
parameters can be determined directly from the data.  Functional forms
similar to equation (.5) have sometimes been used ad hoc in statistical curve
fitting routines to describe Ames assay data; the~"above derivation provides
theoretical support for use of this function.  We will show data to illustrate
that the apparent slope of the initial approximately linear portion of non-
linear dose-response functions obtained with several diesel -particulate
extracts can over estimate, by as much as a factor of 2, the actual mutation
frequency in the Ames assay.


                                 REFERENCES
1.  Luria, S. and Delbruck, M. , Genetics 28, 491 1943.

2.  Salmeen, I. and Durisin, A., Mutat. Res. 85, 109 1981.
                                     592

-------
           CHEMICAL AND MUTAGENIC CHARACTERISTICS OF DIESEL EXHAUST
                     PARTICLES FROM DIFFERENT DIESEL FUELS
                                      by
                 D.  S.  Sklarew,  R.  A.  Pelroy and S.  P.  Downey
               Pacific  Northwest Laboratory operated by Battelle
                                 P-  0.  Box 999
                          Richland,  Washington  99352

                          R.  H.  Jungers and J.  Lewtas
                     U.  S.  Environmental  Protection  Agency
                 Research  Triangle  Park,  North Carolina  27711
     A potential  for  increased  use  of a  wide  variety  of  diesel  fuels  has  in-
creased the  importance  of  studies to  determine  whether the  nitrogen content
of different  fuels  affects the  chemical  and mutagenic characteristics of  the
particles  produced  during  combustion.   In  this  study, the exhaust particles
from five  diesel  fuels  with various nitrogen  and  aromatic hydrocarbon contents
were examined.  The fuels  included  minimum quality  #2 (Min. Qua!.), jet fuel
(JP-7), shale  diesel  fuel-marine  (DFM),  a  base  fuel plus heavy  aromatics  and
hexylnitrate  (BF+HAN+HN),  and the base fuel plus  isoquinoline (BF+IQ).  Table 1
indicates  the  aromatic  and nitrogen contents  of these fuels.  Particles used
in the study were generated by  Southwest Research Institute for EPA using a
Mercedes 240D  vehicle driven through  five  consecutive Highway Fuel Economy
Tests.

Experimental

     The particle samples  were  soxhlet extracted  with methylene chloride  and
these extracts and  the  fuel  samples were fractionated by acid-base and silica
gel column chromatography  methods.  Six  fractions resulted from each material:
base, acid, aliphatic hydrocarbon,  aromatic HC, moderately polar neutral, and
highly polar neutral.   Several  fractions were analyzed by selective detector
gas chromatography  and  by  gas chromatography-mass spectrometry.  A 30m fused
silica Carbowax column  was  used for the  moderately  polar and highly polar
fractions, and a  30m  SE-52  column was  used for  the  aromatic hydrocarbon frac-
tion.

     All  fractions  were assayed for mutagenicity  in the Ames histidine rever-
sion test with Salmonella typhimur^                          All samples were
                                     593

-------
tested with and without metabolic activation with Aroclor  induced  rat  liver
(S9) homogenates.

     Because of the current interest  in nitroaromatics  in  diesel particles
(1,2,3), an experiment was done to estimate the  recovery of  smaller  ring  ni-
troaromatics from filters.  2-Nitrofluorene was  added to one of  the  filter
samples (BF+HAN+HN) and this "spiked" filter as  well as a  control  filter  of
the same fuel were extracted, fractionated, and  assayed in the Ames  test  with
TA98 minus S9.

Results and Discussion

     The amount of material extracted with DCM from the diesel particles
ranged from 9% to 16%.  Recovery of material in  the fractionation  procedure
ranged from 82% to 105%.  The weight distribution in the six chemical  class
fractions did not appear to correlate with either the aromatic or  nitrogen
content of the parent (uncombusted) fuel.  Nitroaromatic standards distributed
between the aromatic hydrocarbon and moderately  polar neutral fractions.

     Gas chromatography with a nitrogen selective detector was used  to compare
the nitrogen-containing components from the five particle  samples.   The mod-
erately polar fractions showed similar patterns  in all five  samples, as did
the highly polar fractions (Fig. 1).  However, particles from the  diesel  fuel
marine and minimum quality fuels appeared to contain more  nitrogen compounds
than did the other fuels.  There did not appear to be a correlation between
the major nitrogen-containing peaks in the fractions and the nitrogen content
of the fuel.

     The aromatic hydrocarbon fraction of the shale diesel fuel-marine parti-
cles was analyzed by GC-MS.  It contained numerous polycyclic aromatic hydro-
carbons including phenanthrene, anthracene, fluoranthene,  pyrene,  benz(a)an-
thracene, chrysene, benzofluoranthenes, and benzpyrenes.   The moderately polar
fraction of the particles from minimum quality #2 contained  a number of car-
boxylic  acids,  carboxylic acid methyl esters, phenanthrenequinone,  alkylated
phenols, fluorenone, alkylated fluorenones, and  possible alkanones.  Compounds
tentatively identified in the highly polar fraction of particles from the base
fuel plus isoquinoline include a number of phenols, carboxylic acids, benzoic
acid, quinoline, xanthenone, and benzofuranone.  Another peak with a probable
molecular weight of 179 is tentatively identified as an anhydride, possibly,
a nitrogen-containing phthalic anhydride.


     The fractions from the parent fuels showed  no mutagenic  activity in  the
Ames assay.  The mutagenic activity of the fractions from  the diesel particles
is shown in Figure 2.  After fractionation mutagens were recovered in four
fractions:  acid, moderately polar neutrals, strongly polar  neutrals, and
aromatic hydrocarbons.  Recoveries of mutagenic  activity in  the diesel frac-
tions were low; however, standard reference mutagens, both direct  and indirect,
did not appear to be destroyed by the fractionation procedure.  In particular,
in the experiment with the 2-nitrofluorene spike, much of  the mutagenic activ-
ity was recovered in the aromatic hydrocarbon and moderately polar neutral
fractions.  It is of interest to note that ^50%  of the recovered mutagenic
                                      594

-------
activity from the samples  is not  in  the fractions in which the 2-nitrofluorene
was concentrated but instead is in the highly polar neutral fraction.

     The three most mutagenically active crude DCM extracts were derived from
the diesel fuels that contained relatively  high concentrations of aromatic
compounds.  No correlation was observed between increased mutagenicity and
nitrogen content of the parent diesel fuels.


References

1.  Wang, Y. Y., S. M. Rappaport, R.  F. Sawyer, R. E. Talcott, E. T. Wei.
      1978.  Direct-acting mutagens  in automobile exhaust.  Cancer Letters 5:
      39-47.

2.  Yu, M. L., and  R. A.  Hites.   1981.  Identification of organic compounds
      on diesel  engine soot.  Anal.  Chem. 53:951-954.

3.  Schuetzle, D.,  T. Riley, T. J. Prater,  T. M. Harvey, D. F. Hunt.  1981.
      The identification  of  nitrated derivatives of PAH  in diesel particu-
      lates.  Anal. Chem.,  in press.
           Table 1.   Aromatic and Nitrogen Content of Diesel  Fuels
                                    Aromatics          Nitrogen
                                     (vol  %)            Ippm]

                DFM                   29.9                 5
                M1n.  Qua!.             34.6               240
                BF+HAN+HN             30.8               718
                BF+IQ                  6.6               930
                JP-7                   2.7                <1
                                      595

-------
 DIESEL FILTERS

 MODERATELY POLAR NEUTRAL FRACTIONS

 NPD
                       ra   I •
                             iL __ /"V
       iiLiuLj-
w

                                                                              01  10
                                                          •1- i.
                                                          C 4-
                                                                             i o
                                                                             * n.

                                                                             E >>
                                                                              I—
                                                                             
-------
     DIESEL FILTERS
*
i  ,0
             1A98-S9
             TA98 S9
      OFM

      MIN QUAL.

 Hi BF < HAN i HN

 I   1 fftlQ

      JP-»

                                           1

             U P
         CRUDE EXTRACT
  2
ACIDS
                                                ALIPHATIC HC    AROMATIC HC
                                                                          MOD POLAR
                                                                           NEUTRALS
                                                HIGHLY POLAR
                                                  NEUTRALS
             Fig.  2.   Ames  mutagenicity data of  fractions  from five diesel  filters.

-------
               FSACTIONATION AND CHARACTERIZATION OF THE  ORGANICS
                      FROM DIESEL AND COMPARATIVE EMISSIONS

                                       by

                       C. Sparaclno, R. Williams, K. Brady
                           Research Triangle Institute
                    Research Triangle Park, North Carolina

                                   R. Jungers
                         Environmental Protection Agency
                    Research Triangle Park, North Carolina

     Semi-volatile materials were analyzed from various media including
diesel soot, tar from coke oven residues, roofing tar and cigarette
smoke condensate (CSC).

     The diesel exhaust particles were collected from a 1978 Oldsmobile
350 diesel vehicle operated on the Highway Fuel Economy Test Cycle
(HWFET) with No. 2 diesel (Union 76) fuel.  The particles were collected
on Pallflex T60A20 filters and the organics were removed by Soxhlet
extraction with methylene chloride as previously described (1).

     The 2RI Kentucky Reference cigarette smoke condensate was generated
according to the method of Patel (1977) at Oak Ridge National Laboratory
(1).

     The coke oven main sample was collected from a separator located
between the gas collector main and the primary coolers within a coke
oven battery at Republic Steel in Gadsden, Alabama, about 60 miles northeast
of Birmingham.

     The roofing-tar sample was generated and collected using a conventional
tar pot containing pitch-based tar, enclosed within a chamber and heated to
360°-380°F, a normal temperature for commercial use.  The evaporative emissions
were collected using a small bag house fitted with Teflon filter bags
(1).

     The solvents used to extract or condense the organics from each of
these samples was removed by evaporation under nitrogen.

     The nature and number of organic compounds associated with samples
of the type addressed in this program render them among the most complex
of environmental samples.  At the present time no direct determinant
approach, regardless of the resolving power, is available for routine
characerization.  The analytical problems can be minimized by pretreatment
of the sample in order to distribute the sample compounds into fractions
of similar chemical or physical properties.  This class division provides
useful information regarding the sample's overall make-up, yields conven-
ient and meaningful subfractions for biotesting purposes, and lessens
the analytical burden for ultimate characterization.  This procedure is
                                     598

-------
based on work by Novotny et  al.,  (2) which, after significant alteration,
was used for this study.  The entire scheme is depicted in Figure 1.

     Each sample was subjected  to  this  fractionation procedure to yield
six fractions of various chemical  properties and polarities.  The acid
fraction contains both weak  (e.g., phenols) and 3trong (e.g., carboxylic
acids) acids.  The base fraction contains organic, Bronsted bases (e.g.,
amines).  The neutral fraction  is  subdivided into 3 main fractions based
on compound polarity.  The non-polar neutral (NPN) fraction is comprised
of compounds less polar than ^  naphthalene.  Paraffinic materials are
characteristic of this fraction.   The PNA fraction contains compounds of
intermediate polarity, and is selective  for condensed ring aromatics.
All neutral materials with polarities greater than PNA hydrocarbons are
found in the polar neutral (PN) fraction.  Prior to the subfractionation
of the neutral fraction, the latter must be dissolved in cyclohexane.
All components are not soluble  in  this  solvent.  The insoluble material
is collected as a separate fraction (CI), and is comprised of intermediate
and highly polar compounds.

     Spillover of various compounds into all fractions is a natural
feature of solvent partitioning processes.  Polar neutral material was
removed from the PNA fraction by silica gel chromatography.  The PNAs
were chromatographed using gradient elution such that a fraction containing
only PNA hydrocarbons was obtained (PNA-1).  Other fractions (PNA 2-4)
were collected that contained compounds of intermediate to high polarity.

     Most fractions were directly  analyzed by capillary GC/MS.  The
fractions enriched in polynuclear  aromatic hydrocarbons (PNAs) were
further purified by column chromatography, and the collected subtractions
were analyzed by GC/MS.  A large portion of each sample,  after fractionation,
was submitted to the EPA for biotesting.  The remainder of each sample
was used for all analytical  work.

     Approximately 1 g of each  sample was partitioned.  Recovery of
material after application of the  fractionation scheme was generally ca.
80%.  Overall mass balances  are shown in Table 1.  The recovery for CSC
was uncharacteristically low (47.6%).   Extensive emulsions were not
formed during partition; the formation  of insoluble material upon dissolution
in methylene chloride prior  to  fractionation may account for the low
figure.

     The mass distribution for  each sample is shown in Table 2.  These
results represent approximate quantities since any solvent partition
process is a rough separation method.   The cyclohexane insoluble (CI)
fraction contains significant proportions of material in some samples.
This fraction is a measure of the  amount of sample, after acid/base
removal, that is not soluble in cyclohexane, and is therefore presumably
polar neutral material.  The CSC and coke oven samples both contain
major quantities of such material.
                                      599

-------
      The CSC  sample  contained  a  significant amount of organic bases,  while
 the acid fraction was more  important  for the diesel soot extract.   The large
 amount  of non-polar  neutral material  associated with the diesel soot  is not
 unexpected; aliphatic hydrocarbons  are known to constitute  a major proportioi
 of  such mixtures  (3).  The  roofing  tar sample is  notable for the size of the
 PNA fraction.

      The partition scheme proved effective  in providing  some fractions
 that were amenable to direct GC/MS  analysis.   Acid fractions require
 derivatization before comprehensive GC/MS analysis can be carried  out.
 The bases, NPN, PNA  and CI  fractions  can be successfully approached via
 GC/MS,  although clean-up is required  for some fractions.  The PN fraction
 represents a  very difficult analytical problem that probably requires
 LC/MS,  derivatization, further fractionation,  etc.   The  actual extent to
 which any fraction can be comprehensively analyzed by any GC technique
 is  unknown.   More work is clearly required  in this area.

      The partition scheme fractions showed,  after  GC/MS  analysis, more
 or  less the expected compositions.  The acid fractions were  shown  to
 contain phenolic materials; carboxylic acids  such  as  fatty acids were
 detected infrequently (derivatization required).   The non-polar neutral
 fraction was, for all sample types, highly  enriched in saturated and
 unsaturated aliphatic hydrocarbons.   The compounds covered a molecular
 weight  range  corresponding  to  ca. C,0-C_, paraffins.  The CSC fractions
 contained several plant natural  products.   The PNA fraction  from CSC was
 too small to  permit both adequate bioassay  and complete  analysis.   The
 PNA fraction  from all other samples showed  the presence  of PNAs and
 methylated PNAs containing  2-5 rings.   Other  compounds found were dibenzo-
 thiophene, dibenzofuran and oxygenated fluorenes.

      The preparative chromatography of the  PNA fraction  provided separation
 of  PNA  hydrocarbons  from the more polar contaminants  of  that  fraction.
 The amount of material recovered in the chromatographic  fractions  (PNA
 2-4)  was usually quite small;  only  phthalate  esters and  unknown species
 were indicated.  Based on similar chromatographic  schemes (4), nitro-
 arenes,  oxygenated PNAs and non-basic nitrogen containing PNAs  (e.g.,
 carbazoles) would be expected  constituents  of  these fractions.

      The polar neutral fractions showed, for  all samples, significant
 spillover of  PNAs.   Some oxygenated PNAs (e.g., benzanthrone,  anthraquinones,
 hydroxyaromatics) were also identified in this fraction.

References

1.   Huisingh, J.  L., R.  L.  Bradow,  R. J. Jungers, B. D.  Harris, R. B.
     Zweidinger,  K.  M.  Gushing, B.  E.  Gill,  R. E.  Albert. 1980
     Mutagenic and Carcinogenic Potency of Extracts of Diesel and Related
     Environmental Emissions:  Study Design,  Sample Generation, Collection,
     and Preparation.  EPA Reports  (EPA Report EPA-600/9-80-057b, pp.
     788-800.
                                    600

-------
2.

3.
4.
Novotny, M., P.L. Lee, K.D. Bartle.  1974.  J. Chrom. Sci. 12:606.

Rodriguez, C.P., J.B. Fisher, and D.E. Johnson.  1980.  Health Effect of
  Diesel Engine Emissions:  Proceedings of an International Symposium,
  Vol. 1.  EPA-600/9-80-057a.  U.S. Environmental Protection Agency:
  Cincinnati, OH.
     Erickson, M.D., D.L. Newton, K.B. Tomer.  1980.  Analytical
       Charactierization of Diesel Exhaust Particulate Extracts.
       Report.  EPA Contract No. 68-02-2767.
                                                             Third Annual
   Table  1.  Mass Balance Results from Fractionation of Comparative Samples


Amount fractionated (mg)
Fractional totals (mg)
Mass balance (%)
Diesel
Vehicle
803.0
696.9
86.3
Cigarette
Smoke
913.8
435.0
47.6
Coke
Oven
936.8
783. 1
83.6
Roofing
Tar
1071.2
895.9
83.6
          Table  2.   Percent  of Total  Mass  Recovered Upon Fractionation
                    of  Comparative  Samples
Fractions
Acid
Base
PN
NPN
PNA1
PNA2
PNA3
PNA4
CI
Diesel
Vehicle
3.6
1.0
7.6
74.2
1.4
1.7
0.9
1.6
7.8
Cigarette
Smoke
1.0
12.1
11.5
0.6
0.02
0.0
0.1
0.9
73.7
Coke
Oven
0.4
4.3
7.1
15.2
4.9
0.0
4.0
0.7
63.3
Roofing
Tar
2.3
2.9
13.1
40.4
31.8
1.2
0.02
0.3
8.0
                                      601

-------
       Figure i.   SOLVENT PARTITIONING FRACTIONATION  SCHEME

                              SAMPLE
                                 DCM (CHjO^) EXTRACTION
                                 ACID WASH
          BASIFY (pH 10)
     ORGANIC
     BASE
    BASE
CYCLOHEXANE
SOLUBLES

NON-POLAR
NEUTRALS
  NPN
                                           BASE WASH
       ACIDIFY (pH 3)
   ORGANIC
   AGIO
                                                  CYO.QHEXANE
                                             NEUTRALS

                                               MeOH WASH

MaNOj
Me
SC
WASH
r
MeC
SOL
1
LUBLES POL
NEL
PNA *
POLAR NEUTRALS
                 HPLC
                                    aaOHEXANE
                                    INSOLUBLES

25 OCM
IN HEXANE
PNA-1

so/ so
HEXANE OOI
DCM
PNA-2 PNA-3 1 PN;
                                                                   10S MeOH
                                                                   IN OCM
                               602

-------
                       SwRI-SFRE DIESEL HEALTH EFFECTS
                              EXPOSURE FACILITY
                                     by
                             Karl J. Springer
                     Department of Emissions Research
                       Southwest Research Institute
                             6220 Culebra Road
                            San Antonio, Texas
     The  future  of  the  fuel  efficient  automotive  diesel passenger car engine
has been  clouded by the possibility  that  its  exhaust particulates have car-
cinogenic  properties.   To  determine  whether this  is a possibility, auto
manufacturers  and the federal  government  are  seeking to determine the pos-
sible long  term  health  effects  of  diesel  exhaust  particulates.  What the
National  Research Council  called the most comprehensive effort in this area
is being  conducted  by Southwest Research  Institute and its sister organiza-
tion, Southwest  Foundation for  Research and Education.  The project is
sponsored  by General Motors  Corporation.

     The  facility is the largest and most advanced of its kind.  The build-
ing consists of  three rooms.   The  engine  room houses the diesel engine and
mechanical  equipment for conditioning  the dilution air.  The exposure levels
are monitored  and the operation of the experiment is handled from the con-
trol room.  The  four large exposure  chambers  are  located in the chamber
room and  are in  close proximity to the diesel engine exhaust, yet isolated
by a soundproof  wall.

     Each  chamber is 8  ft wide  by  8  ft long by 8  ft high.  They are large
enough to  handle  about  1250  rats,  mice, and hamsters for long term exposure
to diesel  exhaust for a total of about 5000 animals.  We wished to simulate
levels of  exposure  not  unlike  those  that  might be experienced on the street.
From previous  experience we  know that  the maximum dose one might experience
behind a  city  bus is about 1 part  of exhaust  in 120 parts of air.  So, one
chamber is  operating at that level.  Another  chamber is operating at twice
that dose,  or  1  part of exhaust to 60  parts of air.  A third chamber is
operating at about  one-third that  level,  1 part of exhaust to 360 parts of
air.  A fourth chamber  receives no diesel  exhaust at all, just purified air
and therefore  is  the control group for comparison to the other three.

     Two Oldsmobile 5.7 liter diesel engines  are  mounted on individual
stationary  dynamometers.  One is operated and the other is a back-up when
                                    603

-------
necessary.  The engine operates at an equivalent 40 mph cruise condition
(1,350 rpm and 71 ft Ibs torque) for 20 hours each day.  Individual samples
of  exhaust are directed to the top entrance of each of the three pyrimidal
shaped chambers for subsequent dilution and mixing.  The engine operates
on a type 2D emissions test fuel.  The engine room also houses the air
conditioning and air filtration equipment used to condition and purify the
dilution air.

     The control room contains the automatic controls to maintain engine
speed and power output,  the environmental controls for maintaining the
chamber temperature at 74°F and relative humidity at 50 percent with a
slight negative pressure of 0.5 in FLO.  Monitoring of the gaseous emissions
is on a semi-continuous basis.  Each chamber is automatically monitored for
10 minutes each hour for hydrocarbons by heated FID, CO and C02 by non-dis-
persive infrared analyzers, and for NO/NO  by chemiluminescence analyzer.
These and pertinent engine and chamber environment data are recorded each
10 minutes.

     The most important measurement is that of total particulate in each
chamber.  This is performed by collecting a sample of the chamber atmosphere
on a 47 mm diameter plastic coated fiberglass filter media.  The weight
gain is used with the sample volume to compute the particulate concentration
within each chamber.  These measurements are made at least once each day.
The amount of exhaust sample admitted into the mixing area prior to the
chamber is adjusted as required to maintain the concentration of particu-
lates as close to specification as possible.

     The entire facility was designed for long term, continuous, trouble-
free operation.  Redundancies are provided in terms of backup engine
dynamometer and controls, backup power for emergency air conditioning,
backup pumps and air conditioning units and backup controllers etc., that
may malfunction and result in an emergency condition.  The totally inte-
grated design located equipment items to simulate exhaust exposures as
close to that in the field as possible.  The system typically operates 20
hours each day (4 hours for animal and cage hygiene) and on a 7 day per
week basis.
                                    604

-------
     POST-EXPOSURE DIESEL PARTICLE RESIDENCE IN THE LUNGS OF RATS
      FOLLOWING INHALATION OF DILUTE DIESEL EXHAUST FOR 6 MONTHS


                             K. A. Strom and B. D. Garg
                           Biomedical Science Department
                       General Motors Research Laboratories
                                 Warren, MI 48090


Due to its submicron size, 15-17% of inhaled diesel particulate deposits in the airways of
the lung.  In  the alveoli of the lung, the alveolar macrophages  scavenge the diesel
particles, phagocytize  them and diesel  particulate-laden  macrophages  were  found in
lung lav.age fluid even 90 days following an exposure of  16 days to diesel emissions at
6 mg/m  [1,2] .  This indicated that the diesel particles may have a long residence time
in the macrophages within the lung, rather than being rapidly eliminated via the ciliated
airways or lymphatics.  The studies describe the results  of biochemical, morphological
and physical measurements on the alveolar rnacrophages,  as well as the histology of  the
lung after exposure of rats to 250  ug DP/m   for 6  months, and serial sacrifice up to 16
months post-exposure.

Male Fischer 344  rats (COBS CDF F-344/CrlBR)  were exposed to diesel  exhaust
particulate  concentration  of  250  ug/m    for  20  hrs/day,   5-1/2 days/week for 6
months.   Control animals were exposed to  clean air.   Lungs of exposed rats were
lavaged  In situ  with Hank's  Balanced Salt Solution  (without  calcium or  magne-
sium).    Differential cell counts  and assays of  the  enzymatic activities  of acid
phospnatase and beta-glucuronidase were performed on the lavaged cells.

Light microscopic  investigations  of the lung  revealed  that  immediately after  the
exposure,  diesel-laden   macrophages  were  diffusely   distributed  throughout   the
pulmonary  alveoli and  also  focally aggregated in  some alveoli.   Some macrophages
containing  diesel  particles  were  also observed  in the  bronchus-associated lymphoid
tissue (BALT) and in the lymphatics.  Subpleural pigmentation  consisted  primarily
of aggregations of alveolar macrophages containing diesel particulate.  Initially, ninety-
five percent of the lavaged macrophages  were completely pigmented with phagocytized
diesel particles,  such that only the nucleus of the cell  was visible and the  cytoplasm was
opaque.    The  percentage  of these  macrophages declined  with time,  showing   an
exponential decay with a 6 week  half-time.  The  percentage of lavaged macrophages
which were free of diesel particulates rose linearly at a  rate of 2.5 percent per week.
The  rest  of  the  lavaged  macrophages   contained some  diesel  particulate-filled
phagosomes within  the  cytoplasm.   The  macrophages  were  obtained  in the  same
numbers  as those from  control animals, and  had  comparable cell size  and activities of
the lysosomal enzymes, beta-glucuronidase and acid phosphatase.
                                        605

-------
With increasing time post-exposure, diesel-laden macrophages were no longer observed
diffusely distributed throughout the lung, but remained even after 16 months (life span)
in focal accumulations in the alveoli  (some of which had thickened alveolar  walls),
particularly noted in the  subpleural regions, also in the SALT and in the  lymphatics.
Initially, the lung tissue contained 0.663  ± 0.075 mg (n=2) diesel particulate.  After 16
months of  clearance, diesel particulate in the lung had declined to 0.250  ±  0.090  mg
(n=5) with only 0.014 mg (n=10) in the regional lymph nodes.

The rapid  decline of the diesel-laden  macrophages which was much faster than the
overall diesel particulate removal from the lung, suggests  that  under conditions of
prolonged exposures to high  concentrations disappearance of particulate-laden macro-
phages from the lavageable pool of cells seems to be due to the formation of aggregates
of alveolar macrophages rather than transport out of the lung.  The overall clearance of
diesel  particulate  after extensive exposures seems to be slow  and proceeds by  as yet
unknown mechanisms.  After 16 months post-exposure, alveolar macrophages containing
small amounts of diesel  particulate can still be  identified  in the lavage fluid.  In
addition, polymorphonuclear leukocytes are present among the aggregated macrophages
suggesting  that the incoming  alveolar macrophages and polymorphonuclear leukocytes
may be involved in or contribute to the breakdown of the macrophages aggregates.

REFERENCES

1.    S. D. Lee, K. I. Campbell, D. Laurie, R. G. Hinners, M. Malanchuk, W. Moore, R.
     J. Bhatnagar and I. Lee, Toxicological assessment of diesel emissions.  Abstract of
     presentation to Air  Pollution Control Assoc., 71st Annual Meeting, Houston, TX,
     25-30 June 1978.

2.   W. Moore, J. Orthoefer, J. Burkart, and M. Malanchuk, Preliminary findings on the
     deposition and retention of automotive diesel particulate in rat lungs.  Abstract of
     presentation to Air  Pollution  Control Association, 71st Annual Meeting, Houston,
     TX. 25-30 June 1978.
                                        606

-------
                     co
                     CNJ
FIGURE LEGENDS
                     3
Fig. la     250 pg/m   for  twenty-five  weeks and  8 weeks  post-exposure:   speckled
           appearance of the exposed lung.
                     3
Fig. Ib     250  ug/m   for twenty-five  weeks  and sixty-nine  weeks post-exposure.
           Diesel-laden  macrophages are still present in association with the pleura!
           surface region.
                     o
Fig. Ic     250  ug/m   for  twenty-six  weeks and  forty-five  weeks post-exposure.
           Scanning electron micrograph of macrophage aggregation in a pleural region
           from an area shown in Figure Ib.
                                         607

-------
                         TRAPPING GASEOUS HYDROCARBONS
                                      by
                                  Fred Stump
                  Environmental  Sciences Reserach Laboratory
                     U.S.  Environmental  Protection Agency
                    Research Triangle Park,  North Carolina
     The gas phase mutagen research was initiated at the request of the Office
of Mobile Source  Air Pollution Control for  the  development of a procedure to
trap gas  phase  hydrocarbons  (HC)  and  then to use  the method  in  a series of
studies with the  objective  of comparing  the  Ames Bioassay activity of the gas
phase with the particle-bound vehicle emissions.

     Several of  the hydrocarbon absorbents  in  use by  other  researchers  were
procured  for  evaluation  as to  applicability for HC collection  in the diesel
fuel range.   A thorough  examination  of the  known  physical  characterises of
the  mediums  resulted  in  the  selection  of  coconut charcoal   and  XAD-2  resin
(divinylbenzene polystyrene polymer) as the best possible candidates.

     Preliminary  evaluations  consisted  of making  injections  of  diesel  fuel
into a  small  dilution  tunnel and  trapping  the  hydrocarbons  with subsequent
recovery  determinations  proving  both  to  be excellent  HC  absorbers  but  the
XAD-2 was the better release medium.  Although, several different solvents and
solvent combinations were used  as extraction  agents to  improve recovery
efficiency from the charcoal, a recovery greater than 60X could not be achiev-
ed and  work with  charcoal  was discontinued.   Further recovery work showed the
XAD-2 to  be quantitative  by both  chromatographic and gravimetric analysis for
diesel   fuel  range  hydrocarbons.   The recovery  and qualification experiments
were performed  using A.D.  Little  (ADL)  developed  odor  traps,  3/8"  O.D.  x 2"
length, filled with the absorbents.

     With completion of  the  qualification  tests  several   samples  were  taken
from both  a 1978 Oldsmobile  and  a 1978 Datsun  220C  diesel powered passenger
car with  the resin  filled  ADL traps.   An extended  sample time in the order of
5-10 hours  was necessary,  due to  the  low  trap  flow rate  and  limited  resin
capacity, to obtain  sufficient materials for Ames Bioassay.  This long sample
time not  only tied  up  equipment  and personnel excessively but also posed  some
                                      608

-------
question  as  to  the possibility of  artifact  formation  on  the resin.   In order
to help  clarify the artifact question,  an  experiment  was conducted  by making
injections  of  diesel  fuel and  NOX on  resin filled traps  (ADL)  and  then
submitting  the  extract  for Ames  testing using  the TA  98  strain.  The  test
results  indicated  that considerable  artifact,  sample level in some situations,
could  be generated  under  these long  exposure conditions and  further  hydro-
carbon collecting  with these tubes was discontinued.

     To  eliminate  the  low  flow  and the long sample time problems, a larger 2"
x 2" trap  with  about 100 times  the resin capacity of the AOL tubes was fabri-
cated  and  carried through  the  same qualification experiments  as  the  smaller
traps with comparable recovery efficiencies  (greater than 95%).


     These large  traps  were then used to sample  a series  of different  cycles
(FTP's,  NYCC, HWFET's)  from a VW Rabbit  diesel passenger  car.   These  samples
were submitted  for Ames  testing  with the results  showing  the gas phase  mater-
ials to  be  active but this activity was  only  9-16%  (depending  on  test  cycle)
of  the   particle-bound  activity.   Although,  artifact  formation  is probably
still present the  trap  activity  data indicates that  it had been substantially
reduced.

     These tests  completed  the diesel  studies.  Since  the gasoline system had
little   N02  present,  with possible minimum  artifact  formation,  an XAD-2
trapping  system,  20" x  20"  x  2" bed, having  100 times (due to  low gasoline
emissions) the  resin capacity of the 2"  x 2"  traps  was fabricated to  collect
the  hydrocarbons.   A series  of  three gasoline powered vehicles were  tested:
(1)  a  1972  Chevrolet Impala using unleaded  fuel;  (2)  a 1981  Dodge van  with  a
light duty  catalyst  system also using unleaded fuel;  and  (3)  a 1970 Ford van
using commercially available leaded fuel.

     A sample from each  of the vehicles  was then submitted  for Ames Bioassay
testing  with the XAD-2 trapping results   being  quite different from  that
observed in the  diesel  vehicle.   The  activity of  the gaseous materials
trapped  by  the  XAD-2  was  at background  level for  all  three  vehicles.   The
Dodge van particle-bound HC,  without S9  activation, had  about  twice the  act-
ivity in reverents/microgram as the Impala,   and four times the activity of the
Ford, with  the  higher  Dodge  activity density probably due to  the  oxidative
properties of the emissions control  system  since the catalyst  was  the major
parameter difference between  the Dodge  and  Impala.   The  Ford  had the  lowest
activity  of  the three vehicles  and  this could possibly be  attributed  to the
lower fuel aromatic content (44.4% unleaded   and 27.8% leaded).

     A  comparison of  the  diesel  and  gasoline vehicles  activity  (reverents/
microgram) indicates that for the HWFET cycle,(the only cycle tested  common to
all   vehicles) without  activation the diesel  and  1972  Chevrolet are  about the
same activity,   the Dodge has twice the diesel activity, and the Ford  about one
half the diesel  activity.   Although  the gasoline emissions  have activities
comparable on a revertent/microgram  basis, when the  emissions  are  observed on
a reverent/mile basis,  without  activation,   the worst  of  the gasoline  vehicle
emitters  (Ford)  has  an  activity  that is  only 11.8% and, with activation,  only
4.5% of the diesel particle-bound activity.
                                      609

-------
     These trapping  studies  have clearly indicated the  low level of activity
associated with  the diesel  gas  phase  hydrocarbons  and the  extremely  low or
background levels present in gasoline gas phase emissions.

     The ability of XAD-2 to effectively collect diesel fuel range HC has been
well demonstrated  and  studies are  currently in progress  to  characterize the
XAD-2 collectability of gasoline fuel range hydrocarbons.
                                     610

-------
             ANALYTICAL METHODS FOR NITROAROMATIC  COMPOUNDS

                                   by

                          Silvestre B.  Tejada
                   Mobile Source Emissions  Research  Branch
                    U.S. Environmental  Protection  Agency
                     Research Triangle  Park, N.C.  27711

     A number of methods have been used to  detect  and/or measure nitroaromatic
compounds  in  environmental  samples  (l-5).   The  analysis  usually  involves  a
combination of  fractionation schemes  -  solvent-solvent extraction,  thin
layer  chromatography  (TLC),   open   column  chromatography, high  performance
liquid chromatography  (HPLC)  - and  followed by  analytical  finish  using  TLC,
HPLC  with  UV and  fluorescence detection, gas chromatography (GC), and  a
variety  of  mass spectrometric  (MS)  techniques.   Most  of  these methods  are
labor  intensive  and some  are  plagued  by  poor  sensitivity  and  interference
problems.

     We  have  developed  a reverse phase HPLC-fluorescence  method  using water-
methanol  solvent for the detection, identification and measurement of selected
nitroaromatic compounds with sensitivity at low and sub-nanogram levels.   The
detection  technique  is  based  on  on-column catalytic  conversion of  the  non-
fluorescent nitroaromatic compounds  to the highly fluorescent amine analogs.
Compound  selectivity is  achieved by  appropriate choice  of  wavelengths  for
fluorescence measurements.  Stop-flow  techniques  and spectral  scanning of the
trapped  peaks  were  used  to establish  chemical   identify  by  comparison  with
spectra of  standard samples.

     The  heart of our analytical  system  is a platinum-rhodium catalyst column
(maintained at 60-80 degrees  Celsius)  between  two reverse phase  ODS columns.
Initial  separation  is  achieved  in  the first   ODS  column,  reduction  to amine
analog is  immediately accomplished  in the  catalyst column and the final
analytical  separation of the aminoaromatic  compounds from  interfering compon-
ents  is  achieved in  the second ODS  column.  By allowing only selected amino-
aromatic  peaks through the  second ODS  column,   we  have managed to conveniently
eliminate  the  tedious  sample  clean-up prior   to  analysis.   Figures  1 and  2
illustrate the  use of this  technique  in  the  analysis  of nitro-pyrene.
Precision of ± 3%  at 1 nanogram  level   is  routinely obtained for nitro-pyrene
analysis.   Minimum  detectable quantity  of  nitropyrene  under our  present
analytical  configuration is  about 20  picograms.
                                      611

-------
The catalyst has been observed  to  reduce  nitro compounds to the corresponding
amine  reproducibly under  fixed conditions of  flow rate,  temperature  and
solvent composition.  The following nitroaromatic compounds are converted with
better  than  99% conversion efficiency:  nitro-naphthalenes,  nitro-anthrance,
nitro-fluorene,  nitro-chrysene, nitro-BaP,  dinitro-pyrenes  and  nitro-fluor-
enones.  We have obtained fluorescence spectra of the amine analogs of most of
the  nitroaromatic  compounds  available to  us.   The  amines,  especially  the
diamines,  were observed  to be unstable  under UV  light.   Adjustment  of  the
solvent pH to about 8 with  NaOH helped to stabilized the diamines adeqately to
make reproducible spectral  scans of the trapped peaks.

The present analytical system has been used to measure nitro pyrene in complex
matrices  such  as diesel  exhaust  particulate extracts,  leaded  and non-leaded
automotive exhaust  particulate extracts,  gas trap  extracts,  fly ash  extracts
as well  as biological  extracts.  Samples dissolved  in  DMSO intended  for Ames
tests  are likewise amenable  to analysis  without additional  sample clean-up.
Other  nitroaromatic compounds  can be  detected  and  measured  by  appropriate
choice of  chromatographic elution windows coupled with the optimum wavelengths
for fluorescence measurements.
1.   Jager, J.,  "Detection  and  characterization of nitro  derivatives  of some
polycyclic  aromatic  hydrocarbons by  fluorescence  quenching  after  thin  layer
chromatography:  Application  to air  pollution  analysis",  J. Chrom.  152,
575-578 (1978).                                                          	

2.  Schuetzle, D., Lee, F.S. -C., Prater, T.J., Tejada, S.B., "The Identifica-
tion of polynuclear aromatic hydrocarbon derivatives in mutagenic fractions of
diesel particulate extracts", Intern. J. Environ. Anal. Chem. 9_, 1-53,  (1981).

3.   Gibson, T.L., Ricci,  A.  I., Williams, R.L.,  "Measurement  of Polynuclear
Aromatic  Hydrocarbons,   Their  Derivatives  and  Their  Reactivity  in  Diesel
Automobile  Exhaust"  in "Chemical Analysis  and Biological  Fate:   Polynuclear
Aromatic Hydrocarbons", Cooke, M. and Dennis,  A.J., eds.,  Battle Press,  1981,
p.707.

4.  Schuetzle,  D.,  Riley,  T., Prater,  T.J.,  Harvey, T.M.  and  Hunt,  D.,  "  The
Identification  of Nitrated  Derivatives of  PAH in  Diesel  Particulates",  in
press.

5.   Rozenkranz,  H.J.,  McCoy, E.C.,  Sanders,  D.R.,  Butler, M.,  Kiriazides,
O.K.,  Mermelstein,  R.,"Niropyrenes:   Isolation, identificatin,  and  reduction
of mutagenic  impurities in  carbon  black  and  toners",  Science  202,  515-519
(1978).                                                           	
                                      612

-------
                        Figure 1
Neat
            extract (25 ug) through CDS oolum only ( - ) ,-
        CDS and catalyst coluons (— )-.  Note peak
enhancements due to formation of aminoconpourels.   Detection
              Eadtation (360 nn) ,  Bni^sicm (430  ran), 0V (254 nn) .
                        Figure 2

 Neat dicsel extract (25 uj)  through CDS-Catalyst-CDS oolonns.
 Oily the nain anunopyrene peak was injected into  the second
 iXE coluon. (	)  Ov and fluoresesnce profiles of the ample
 through CDS-Catalyst oolums.  Note removal of interf errant
 peaks after passage through  second CBS ooluim.
                            613

-------
                       TOTAL LUMINESCENCE SPECTROSCOPY
                        OF DIESEL EXHAUST PARTICULATE
                                      by
                            Gregory Wotzak, Ph.D.
                         Cleveland State University
                               Cleveland, Ohio

                             Robert Whitbv. P.E.
           New York State Department of Environmental Conservation
                               Division of Air
                              Albany, New York


     Total Luminescence Spectroscopy (TLS) analysis of organic extract
material from diesel exhaust particulate matter has been previously described
by Wotzak et aid).  TLS is the determination of the luminescence intensity
as a function of all accessible excitation and emission wavelengths.  TLS
data are typically obtained from between 50 and 200 emission spectra, each
taken at a specific excitation wavelength.  Contour mapping of the points of
equal luminescence intensity on an. excitation vs. emission wavelength grid
has been chosen as a convenient means of representation for TLS data.  TLS
analysis thus encompasses the computerized data acquisition, manipulation,
display, and interpretation of such luminescence data.

     The New York State Department of Environmental Conservation, Automotive
Emission Evaluation (AEE) unit has been studying the utilization of TLS in the
characterization of diesel particulate organic extract using a Baird Corpora-
tion SFR-100 Ratio Recording Spectrofluorometer.  Data acquisition and scan
control for this instrument is locally provided by a Baird MP-100 micropro-
cessor controller which are linked to a host Data General Nova 3 computer
system with 128K words of core and a 10 megabyte  disk.  Software was pro-
vided by Baird Corporation and modified by AEE computer personnel to include
data smoothing algorithms.  Contour plots are produced on a Houston Instru-
ments Complot™ X-Y recorder.

     More recent advancement in raw data reduction has been achieved by wave-
length correction and digital smoothing using Fast Fourier Transform methods.
Extract samples, fractionated by column chromatography, have been analyzed,
generating TLS contour maps for each fraction.  These fraction contours may
be added by computer software routines, using appropriate weighting by cut
recovery factors, and compared with spectra for the original bulk sample.
                                      614

-------
This reconstitution work was performed as an  Internal consistency test for
both the acidic, basic and neutral  cuts  of  the raw extract, as well as the
sub-cuts of the neutral fraction,

     A variety of  tasks have been performed in order to obtain a general in-
dication of the utility of this  relatively  new analytical procedure.  Se-
quential dilutions of extract  sub-fractions were performed in order to
determine  the extent of internal absorption of fluorescent radiation.  TLS
spectra of  the diesel fuel and lubricant were subtracted from appropriate
spectra in  order  to  facilitate further analysis, and determine the partition
of unburned fuel  and  lubricant among  extract  sub-fractions. The neutral
fraction was  analyzed qualitatively  and quantitatively for several known
compounds.


(^  Wotzak, G.,  R.  Gibbs, and J. Hyde,  1980, A Particulate Characterization
     Study  of  In-Use Diesel Vehicles.  In:  Health Effects of Diesel Engine
     Emissions:   Proceedings  of an  International Symposium, Volume 1.
     W.E.  Pepelko,  R.M. Danner,  and N.A. Clarke, editors.  U.S. Environmental
     Protection Agency, EPA-600/9-80-057a.  pp.  113-137.
                                      615

-------
   EVALUATION OF THE METABOLIC REQUIREMENTS OF DIESEL AND COMPARATIVE SOURCE
        SAMPLES IN THE SALMONELLA TYPHIMURIUM PLATE INCORPORATION ASSAY

                                      by

                     Katherine Williams and Joel!en Lewtas
                          Genetic Toxicology Division
                      Health Effects Research Laboratory
                     U.S. Environmental Protection Agency
                    Research Triangle Park, North Carolina


     The mutagenic activity of a mobile source sample (Nissan diesel exhaust
extract) and three comparative source samples (coke oven mains, cigarette smoke
condensate, and roofing tar extract) were examined in the Ames plate
incorporation assay using strain TA98.  The mutagenic and carcinogenic activity
of these complex mixtures had previously been determined in a bioassay test
matrix (1).  The comparative sources were less mutagenic in the Ames assay
compared to the mobile source sample.  However, they were more active in
mammalian cell mutation and mouse tumor initiation bioassays.  The objective of
this study was to determine whether modifications in the S9 activation system
would alter the mutagenic activity in the Ames bioassay.  The modifications
examined included altering both the species from which the liver S9 was
prepared and the concentration of S9 on the plate.

     Both Aroclor 1254-1nduced and -uninduced S9 were prepared from CD rats and
Syrian golden hamsters.  Livers were pooled from groups of at least 6 male
animals for each preparation.  The protein concentration of each S9 was
determined by the method of Lowry et al. (2).  Four doses of S9 were run:
0.31, 0.63, 1.25, and 2.5 mg/plate, chosen to encompass the usual concentration
in the Ames test (ca 1.5 mg/plate).  Aliquots were adjusted to contain the
required dose in a 0.5-ml volume as used in the Ames test.

     Three doses of each source sample (a low, medium, and high range) were
tested at all four S9 concentrations for each S9 preparation.  The dose varied
between samples and were selected to be below the toxic level for each, yet
high enough to have mutagenic activity.  For Diesel Nissan and cigarette smoke,
the doses were 30, 100, and 300 yg/plate.  For coke oven mains and roofing tar,
they were 5, 50, and 100 ug/plate.

     Experiments were run in duplicate, on different days, using triplicate
plates for each point.  In any experiment all 3 doses of 2 samples were tested
with all 4 concentrations of both uninduced and induced S9 from one species.
For each experiment fresh dilutions were prepared from aliquots of the 39
                                      616

-------
preparations, which were held at -80°C.  The time between duplicate experiments
ranged from 3 to 9 days.  The data is presented as the mean ± SE for duplicate
experiments.

     The optimum S9 dose for all samples, with the exception of the Diesel
Nissan which did not require metabolic activation, was either 1.25 or
2.5 mg/plate for both induced and um'nduced rat and hamster S9, as shown in
Table 1 for the induced rat liver S9.

     Recently, several studies have noted that induced hamster S9 is more
effective than induced rat S9 in activating such compounds as aromatic
amines (3) phenacetin (4) and diethylnitrosamine (5), while Aroclor-induced rat
S9 is more effective than hamster S9 with polycyclic aromatic hydrocarbons  (3).
This study showed no difference in effectiveness between hamster S9 and rat S9
in activating the Diesel or the comparative source samples.

     The mutagenic activity of the comparative source samples was higher when
Aroclor-induced 39 was the metabolic activator as compared to um'nduced S9,
whether rats or hamsters were the source of the S9.  The Diesel Nissan sample
was again the exception; the mutagenic activity was higher with um'nduced S9
than with induced.

      In conclusion, for the three comparative source samples, all of which
require metabolic activation for maximum mutagenic activity in the Ames test:
the optimum S9 dose is 1.25 to 2.5 mg/plate; Aroclor-induced S9 is more
effective as an activator than uninduced S9 regardless of species; and rat  and
Syrian golden hamster S9 are equally effective in activating these complex
mixtures.  These results would suggest the lower mutagenic activity of the
comparative source samples in the Ames test as compared with other mutagenicity
or carcinogenicity bioassays was not due to the exogenous metabolic activation
system.


REFERENCES

1.    Nesnow, S., and J.L. Huisingh.  1980.  Mutagenic and carcinogenic potency
       of extracts of diesel and related emissions:  Summary and discussion of
       the  results.   In:  Health Effects of Diesel Engine Emissions.
       Proceedings of an  International Symposium, Vol. 2.  W.E. Pepelko, R.M.
       Danner, and N.A. Clarke, eds.   EPA-600/9-80-057b.  U.S. Environmental
       Protection Agency:  Cincinnati, OH.  pp. 898-912.

2,    Lowry, O.H., N.J. Rosebrough, A.L.  Farr, and R.J. Randall.  1951.  Protein
       measurement with the golin  phenol reagent.  J. Biol. Chem. 193:265-275.

3.    Raineri, R., J.A. Poiley, R.J.  Pienta, and A.W. Andrews.  1981.  Metabolic
       activation of carcinogens  in  the  Salmonella mutagemcity  assay by
       hamster and rat liver S9 preparations.  Environ. Mutagen. 3:71-84.
                                      617

-------
4.   Nagao, M., T. Suglmura, and T. MatsusMma.  1978.  Environmental mutagens
       and carcinogens.  Ann. Rev. Genet. 12:117-159.

5.   Prival, M.J., V.D. King, and A.T. Sheldon.  1979.  The mutagenlcity of
       diabyl nltrosamines in the Salmonella plate assay.  Environ. Mutagen.
       1:95-104.
         Table 1.  Effect of Metabolic Activation Dose 1n Mutagenlcity
                   of Diesel and Comparative Source Samples9
                                          Revertants/plateb
    Sample
0.31C
0.63C
1.25C
2.5=
Diesel Nissan
Coke oven mains
Cigarette smoke
Roofing tar
816 ± 26
492 t 57
68 ± 0
59 ± 1
625 ± 39
727 ± 8
83' ± 4
86 t 2
491 ± 50
861 ± 21
74 ± 6
94 t 3
334 ± 31
874 ± 27
64 ± 3
113 ± 9
"Salmonella typhimurium TA98.Samples at 100 ug/plate.
"Mean ± SE of two experiments with triplicate plates.
cRat Aroclor 1254-induced S9, mg/plate.
                                     618

-------
                MS/MS  CHARACTERIZATION OF DIESEL PARTICULATES
                                      by

                                Karl  V.  Wood
                Fuels  Analysis  Laboratory-Chemistry  Building
                               Purdue University
                        West Lafayette,  Indiana   47907

                      James  D.  Ciupek and  R. Graham  Cooks
                            Department of  Chemistry
                               Purdue University
                        West Lafayette,  Indiana   47907

                               Colin  F-  Ferguson
                       School of Mechanical  Engineering
                               Purdue University
                        West Lafayette,  Indiana   47907
 INTRODUCTION
     Analysis of  the organic  constituents  adsorbed on diesel exhaust particu-
lates has become  increasingly important with  the  increase in the number of
diesel engine automobiles.  Extraction followed by GC/MS has been the usual
analytical method of choice for  these unknown  organic constituents.  Mass
spectrometry/mass spectrometry (MS/MS) offers  a possible means of direct
analysis with either minimal  or  no  separation.  This technique enables a
fast characterization of the  organic constituents of the whole diesel par-
ticulates.  Selectivity and specificity of particular compounds or classes
of compounds of interest using MS/MS can be improved by the appropriate choice
of the chemical ionization reagent  gas as  well as the choice of positive or
negative ion detection.

EXPERIMENTAL

     The MS/MS experiments described in this  study were run using a Finnigan
triple stage quadrupole mass  spectrometerJ   The diesel particulate sample
is introduced into the source with  the direct  insertion probe which is
heated in steps to obtain temperature profiles of the organics adsorbed on
the particulates.  The sample is  ionized using the chemical ionization
technique.  The ion of interest  is  mass selected by quadrupole 1, undergoes
collisionally induced dissociations in quadrupole 2 with the resulting
                                      619

-------
fragment ions being mass analyzed with quadrupole 3.  Typcial ion axial
energies into quadrupole 2 are 20 eV relative to the source.  The collision
gas used for these studies was argon at a pressure ca_. 2.2 mTorr.

     The diesel engine employed in this study was an AVL model^ 520.005
naturally aspdrated single cylinder direct injection engine.  The diesel
exhaust particulates are sampled using a mini dilution tunnel system.3  As
this system was designed both the dilution ratio and the temperature of the
particulate filter can be varied.

RESULTS

     The initial emphasis of this study was to identify constituents af
diesel exhaust particulates by direct analysis using MS/MS.  This identifi-
cation is done by comparing the MS/MS spectrum of a particular ion in the
diesel exhaust particulate sample with the MS/MS spectrum of the correspon-
ding ion of a standard reference compound.  For example, the MS/MS spectrum
of the m/z 143 ion from a diesel exhaust particulate sample has two major
fragment ions, m/z 128 (100%) m/z 115 (15%) besides the main beam ion m/z
143 (20%).  These ion ratios are nearly identical to that found in the
MS/MS spectrum of 2-methylnaphthalene, suggesting its presence in the
diesel exhaust particulate sample.  It is not possible to say which methyl-
naphthalene is present, if only one is, or the relative concentration of the
two in a probable mixture of both.  The case of the m/z 143 ion suggesting
the presence of predominately only one constituent and its positional isomers
is not unique.  However, as would be expected the MS/MS spectra of many ions
are suggestive of the presence of more than one type of structure.  An
example of this is the ion at m/z 139 in the diesel exhaust particulate
sample which is strongly indicative of the presence of hydroxybenzoic acid
isomers through comparison with standard reference compounds.  However, there
are other relatively intense fragment ions in the MS/MS spectrum which can-
not be resulting from hydroxybenzoic acid.  One other compound that may
account for the remaining fragment ions is decalin.

     Along these lines is the investigation of the selectivity and specifi-
city of the MS/MS technique as it relates to the direct analysis of diesel
exhaust particulates.  For example the comparison of positive and negative
ion isobutane chemical ionization of a diesel exhaust particulate sample can
yield information about specific classes of compounds.  An example of this is
the identification of carboxylic acids in the diesel exhaust particulate
samples.  In positive ion chemical ionization a carboxylic acid will be
protonated as will an aromatic hydrocarbon at the same nominal mass.  How-
ever, in negative ion chemical ionization the carboxylic acid will lose a
proton to give a (M-H)~ ion whereas the aromatic hydrocarbon will be ionized
by electron transfer to give a M" ion.  This technique was used to confirm
the identification of the components in the m/z 139 ion MS/MS spectrum
discussed previously.  Thus the use of negative ion chemical ionization allows
the separation of different classes of compounds for MS/MS identification in
the direct analysis of diesel exhaust particulates.

     Another example of selective ionization to allow a more accurate MS/MS
identification to be made is the use of differing reagent gases.  While-
                                      620

-------
isobutane chemical ionization is a general protonating agent, ammonia can
be used to protonate relatively basic compounds, like amines or aza compounds.
While this technique has been used successfully in the analysis of coal-
derived liquids^ it has not been as useful with diesel exhaust particulates.

     Besides these studies, two variables associated with sampling  the
diesel exhaust particulates have also been investigated.  These variables,
dilution ratio and particulate filter tmperature were utilized to gain a
better insight into the complex problem of particulate sampling.

CONCLUSION

     MS/MS provides a means for the rapid direct analysis of diesel
exhaust particulates.  The use of selective  ionization techniques further
enhances the positive identification when different types of compounds are
present at the same nominal mass.

REFERENCES

1.  Slayback, O.R.B. and M.S. Story. 1981. Chemical Analysis Problems Yield
      to Quadrupole MS/MS. Industrial Research  FEB; 129-134.

2.  Pischinger,  R. and W. Cartellieri.  1972.  Combustion System Parameters
      and Their  Effect Upon Diesel Engine Exhaust  Emissions. SAE Paper
      720756.

3.  MacDonald, J.S., S.L. Plee, J.B. O'Arcy,  and R.M. Schreck. 1980.
      Experimental Measurements of the  Independent Effects  of Dilution
      Ratio and  Filter Temperature on Diesel  Exhaust  Particulate Samples.
      SAE Paper  730834.

4.  Zakett, D.,  V.M. Shaddock and R.G.  Cooks. 1979. Analysis of Coal
      Liquids by Mass Analyzed  Ion Kinetic  Energy  Spectrometry. Anal.
      Chem. 51:1849-1852.
                                      621

-------
SECTION 9



PERSPECTIVES
                     622

-------
PERSPECTIVES  ON DIESEL EMISSIONS HEALTH RESEARCH

NORTON  NELSON
Institute  of  Environmental Medicine,  New York University  Medical Center,
550 First  Avenue,  New York,  New York   10016
   I will  not  attempt to summarize this very excellent,  information-packed
symposium.   Rather  I will make some very personal comments as to what  I have
learned  and  what  I've concluded.
   The problem before us is one of major social importance.  I think we can
conclude quite straightforwardly  that a major increase in tha Diesel fleet is
not going  to produce a disastrous epidemic of lung cancar.  I think at the
other extreme  that  ve are probably not in a position now,  today,  to reach  a
meaningful judgment as to the quantitative impact of such aa expansion*  It may
be that  it's going  to be negligible — that's quite possible.  Howevar,  it may
be that  the  impact  will be at a level that will require  some difficult social
decisions.   I  don't know.
   We've learned  an enormous lot  in the last three years; this is attributable
to the very  intensive and very fine work on the part of  EPA, on the part of
industry and on the part of independent universities and institutes.  The
science  that we've  seen here in the last three days has  been very impressive,;
not all  of it  has been equally elegant, not all of it has been as sharply
focused  as it  might have been, but the great bulk has been sound and to the
point.   We've  moved ahead a great deal in the last three years,  but we are not
quite where  we should be.
   I start with the premise that  I can't really see how  there can be a serious
doubt in anyone's mind at this stage that Diesel exhaust is potentially
carcinogenic for  humans.  We know from chemical analyses that there are
carcinogenic chemicals present in the Diesel exhaust.  We know from a  vast
amount of  study in  simple systems — revertant studies,  cell transformation -—
that they are  mutagenic in a variety of ways; we know that mutagenicity is
highly correlated in this class of compounds with carcinogenieity.  We have
biological data on  whole animals  which, although sometimes borderline, are
sometimes clear and decisive and  which support these findings.  We know,
therefore, that there is a potentiality for carcinogenicity.
   We further  know  that the extractable materials are (at least largely)
bioavailable.   This has been debated intensely; but as far as I'm concerned,
the data presented  in this meeting yield clear evidence  that the material  in  or
                              623

-------
on the carbon  particles  is  bioavailable.
   This brings us  to  a much more  complicated question, that is, recognizing
that the carcinogenic materials are there and that they are leachable, are they
leached in sufficient quantities  in exposed persons to reach the intracellular
biochemical unit (DM) that is  important  for an outcome of malignancy?  This
question involves  a whole series  of issues about which we are in varying
degrees still  somewhat ignorant.   It involves the pharmacokinetj.es of movement
from where the chemicals are lodged to where they are active. Are they
effectively sequestered, what happens  in  macrophages or in the lymph nodes?   Do
these chemicals reach the nucleus  of the  cells in the epithelial lining of the
lung in still potentially active  form in  sufficient quantity to initiate
malignant changes.  These issues  are largely measurable;  we still have large
uncertainties as to the  critical  steps determining dosages to the nucleus  of
the cell.
   Let me back up  for just  a moment to comment on the beginning of the problem,
that is the actual human exposure.   Characterization of the emission of the
particulates starts with their collection and analysis.  To what extent do
artifacts in sampling or analysis  disturb the outcome? To what degree are
nitropyrenes artifacts of sample  collection or,  on the other hand,  of genuine
human concern?  Basic is what happens  to  the particles once they leave the
exhaust pipe and reach the  human"in respired air.   From what I've heard here,
I'm not sure that we  really  have  satisfied ourselves on this score.   One of  the
central issues is what happens to  the  particulates and the vapor phase PNA's
between the time they leave the manifold  and the time they are ready for
inhalation by man; I'm not  sure we  know.   There  is NO  outside; there is also,
I believe, some evidence, perhaps  only suggestive,  that photochemical reactions
can participate in nitrating the  PNA's.   This may  be a factor which affects  the
actual potency of  the particles in  ambient air as  inhaled.  Such issues  need to
be resolved.
   Now reverting to the biological  aspects,  getting a positive response from
known carcinogens  in  lifetime studies  is  sometimes a very difficult procedure.
I spent a good part of my career  coping with that  problem.  If we were to
depend on inhalation  studies of cigarette smoke  in laboratory  animals to decide
that cigarettes are carcinogenic, we would give  them essentially a clean bill
of health.  It took ten  to  fifteen  years  of intensive work to confirm in the
laboratory that the known human carcinogens,  chromium chemicals, were
carcinogenic for the  lung.   We've known for years  that inhaled arsenic
compounds are human pulmonary carcinogens; only  recently  is there a promising
                             624

-------
positive experiment.   I  caution,  therefore,  to beware of putting too much
weight on negative  experiments of this sort.   Dosage is extremely  important;
the conduct  of  the  experiments is extremely  important.   The strain chosen,  the
species chosen,  are important.  I want to commend our visitors from Germany  for
having recognized what we  have not always recognized, namely  that  a simple
arithmetic calculation shows that to get a positive lung cancer outcome  in
Diesel particulates,  even  if they have significant carcinogenicity,  will take
ingenious experimental design if  we're interested in levels that are anything
less than 10,  15 or 20 percent incidence.  That is, if  ona  wants to get  data at
socially important  levels,  one must use special approaches  in designing  such
experiments.
   I have no patience with the view that experimentation in this area must be
relevant to  field conditions in terms of concentrations, times of  exposure and
so on.  It just  won't work.   If one wants to  detect socially  significant  levels
of carcinogenicity  with  what is an intrinsically insensitive  system,  one  needs
to devise experimental methods capable of detecting low but important cancer
potencies.
   The work  that's  been  done on isolated systems is extraordinarily  impressive,
and very elegant.   The work that's been done  chemically is  again very
impressive.
   One issue that has been dealt  with very little, if at all,  that may be of
the deepest  importance,  is the issue of interaction and promotion.  The  view
that one is  concerned with a one-for-one outcome between a  single  carcinogen
and a human  malignancy is  rarely  correct.  In almost every  case interaction
with one or  several factors is involved.  This is especially  true  where we're
dealing with what is  clearly in this instance something less  than  a high
frequency occurrence  of  malignancy.
   Passing from  cancer for just a moment, there were data suggesting
non-malignant histological changes and,  in particular,  fibrosis; we need to
pursue such  issues.
   Finally,  to summarize let me suggest a few things amongst  those that  nave
already been mentioned which I think need resolution.  First,  I think,  is the
issue having to  do  with  the actual state of  the particles in  the air as
breathed by  people.   Are there important changes between their departure from
the exhaust  pipe and  breathing point of man?   This ought to be a reasonably
straightforward  and attackable problem.
   It would seem to me that  it would be worthwhile attempting to see whether a
small number of  index PNA's  could be identified.  It is clear that
                             625

-------
benzo(a)pyrene  it by  oo means  the beet index agent.  Pyrene is cot the best
index agent.  But i»  it possible that tvo or three or four (not more than four)
could be  identified that  could be useful indicators for perhaps an equal number
of classes of PHA's.  Another  alternative would perhaps be some kind of
refinement of the HPLC fingerprint.   Another possibility would be a small
standardized set of revertant  tests.   The point is that work would go ahead
much more rapidly if  there  could be  developed simple, straightforward
techniques for  relatively quick identification and quantification of major
classes of PHA's.
   It would seem to me that  a  major  advance (and challenge) is to move
forcefully towards improving the utility of short term tests for quantitative
estimates of potency  for  humans; the  group that's worked here is eminently
qualified to do this.  There is no question but that the revertant and cell
transformation  tests  have all  been extremely powerful in this research.  They
are still, however, weak  tools for quantitative estimates of potency for
humans.
   We need, as  I mentioned,  more ingenuity in the design of some of our long
term experiments.  We must  refine of  our understanding of the kinetics of
movement of the important chemicals and  their metabolites from the inhaled
particles to the nucleus  of  the potentially responding cell,  by which I mean
the basal cells in the respiratory tract.
   Now all of this, as far  as  I'm concerned, should have one objective, and
that is estimation of the risk for man.   What w« now need in order to move on
with extending  our fuel,  and meeting  our transport requirements,  is to make an
estimate of whether the problem is trivial or socially significant.  Thus, risk
assessment should be  the  ultimate goal and should be given the highest
priority.  There have been  starts; I've  been not totally impressed with the
state of art.   The weakness  is not in mathematics.  The mathematics are,  by and
large, relatively straightforward. A quantitative understanding of the
relevant biological processes  is much more needed now than is improvement in
the mathematics.  An  important contribution to an improved biological
understanding could come  from  a better knowledge base on low dose-response
relationships in laboratory  animals with and without tumor promoters.
   I have only  to add when  controls have been developed and the decision is
made to move ahead, we will  need to maintain monitoring to assess the
importance of changes in  fuels,  and  in engines and engine designs.  Procedures
developed in the present  studies should  be extremely useful in monitoring such
changes.
                             626

-------
   The last word  I would  like  to  leave  is  the urgency  of moving ahead  to  secure
the data to assess the  human  impact  of  the expansion of  Diesel  usage  in  such  a
way as to permit  us  to  make rational and clearheaded decisions.
                               627

-------