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
------- |