Un red States
Environmental Protection
Agency
OTdce of

Washington. DC 20460
EPA 560-6 84 004
Toxic Substances
Scientific Rationale for the Selection of
Toxicity Testing Methods
II. Teratology, Immunotoxicology,
and Inhalation Toxicology

-------
     Printed  in the United States of America. Available from
              National Technical Information Service
                  U S. Department of Commerce
        5285  Port Royal Road, Springfield, Virginia 22161
     NTIS price codes—Printed Copy: Al 2  Microfiche  A01
This report was prepared as an account of work sponsored by an agency of the
United States Government Neither the United States Government nor any agency
thereof, nor any of their employees, makes any warranty, express or implied, or
assumes any legal liability or responsibility lor the accuracy, completeness, or
usefulness of  any information, apparatus, product,  or process disclosed, or
represents that its use would not infringe privately owned rights Reference herein
to any specific commercial product, process, or service by trade name, trademark,
manufacturer, or  otherwise, does  not  necessarily  constitute or  imply  its
endorsement, recommendation, or favoring by the United States Government or
any agency thereof The views and opinions of authors expressed herein do not
necessarily state or reflect those of the United States Government or any agency
thereof

-------
                                      ORNL-6094
                                      EPA-560/6-84-004
SCIENTIFIC RATIONALE FOR THE SELECTION
      OF TOXICTTY TESTING METHODS
   II. TERATOLOGY, IMMUNOTOXICOLOGY,
       AND INHALATION TOXICOLOGY

                    Editors

                Michael G. Ryon
               Daljit S. Sawhney

                 Contributors

              Mary Lou Daugherty
                Robert H. Ross
                Michael G. Ryon

         Chemical Effects Information Group
      Information Research and Analysis Section
         Information Resources Organization
          Oak Ridge National Laboratory

                Project Officers

               Daljit S. Sawhney
              William H. Farland
                Diane D. Beal

             Office of Toxic Substances
        U.S. Environmental Protection Agency
              Washington, DC 20460

          Date of Issue - September 1985
     OAK RIDGE NATIONAL LABORATORY
            Oak Ridge, Tennessee 37831
                  operated by
  MARTIN MARIETTA ENERGY SYSTEMS, INC.
                    for the
        U.S. DEPARTMENT OF ENERGY
      under Contract No. DE-AC05-84OR21400

-------
                           DISCLAIMER

This report was prepared as an account of work sponsored by an agency of
the United  States Government.   Neither the  United States  Government
nor any  agency thereof, nor any  of their employees, contractors, subcon-
tractors,  or  their employees,  nor the  publisher, makes any warranty,
express or implied, nor assumes any legal liability or responsibility for any
third party's use or  the results of such use of  any information, apparatus,
product  or process disclosed in this report, nor represents that its use  by
such third party would not infringe privately owned rights.

This report has been reviewed by the Office of Pesticides  and Toxic  Sub-
stances,  U.S. Environmental Protection Agency, and approved for publica-
tion.   Approval  does  not signify  that the contents necessarily reflect the
views  and policies of  the U.S. Environmental Protection Agency or the
publisher, nor does mention of trade names or  commercial products consti-
tute endorsement or recommendation  of use.

-------
 DALJIT S. SAWHNEV
                        CONTENTS

FIGURES	  vii

TABLES	  ix

ACKNOWLEDGMENTS	  xi

ABSTRACT	xiii

1.  EXECUTIVE SUMMARY

   1.1  TERATOGENICITY	   1

   1.2  IMMUNOTOXICOLOGY	   2

   1.3  INHALATION TOXICOLOGY	   4

        1.3.1  General Test Design	   4
        1.3.2  Exposure Chamber Design	   5
        1.3.3  Generation of Test Atmospheres	   7

2.  TERATOGENICITY

   2.1  INTRODUCTION   	   9
   2.2  GENERAL EXPERIMENTAL CONSIDERATIONS ...   9

        2.2.1  Dosage—Number and Level	  10
        2.2.2  Dosage—Duration   	   10
        2.2.3  Positive Controls	   16
        2.2.4  Number of Species	  16
        2.2.5  Number of Test Animals Per Dose Group   	  24
        2.2.6  Administration Route	   25
        2.2.7  Fetal Examination	   25

   2.3  STRUCTURE-ACTIVITY RELATIONSHIPS	  25

   2.4  TERATOGENESIS AND TIME OF
        ADMINISTRATION	   26

        2.4.1   Preimplantation    	   26
        2.4.2  Organogenesis	   27
        2.4.3  Histogenesis and Fetal Period	  30

   2.5  BEHAVIORAL TERATOGENESIS	  30

        2.5.1   Historical Perspective  	  32
        2.5.2  Behavioral Testing Methodologies	  32
                             m

-------
2.6  SPECIES COMPARISONS   	   33
     2.6.1   General Aspects	    	       37
     2.6.2   Rat     	   37
     2.6.3   Mouse ...      	   ....    ...       •        38
     2.6.4   Rabbit    	39
     2.6.5   Hamster	40
     2.6.6   Nonhuman Primates    	        41
     2.6.7   Other Species       	42
            2.6.7.1   Dog	  42
            2.6.7.2   Cat    	             42
            2.6.7.3   Pig	   43
      2.6.8  Results of Testing Some Human Teratogenic
            Chemicals in Animal Models     ....        	  43
             2.6.8.1    Aminopterin	  44
             2.6.8.2   Methotrexate	  44
             2.6.8.3   Thalidomide	  44

  2.7  RECENT APPROACHES IN TERATOLOGY TESTING   46

      2.7.1   In Vivo Methods	  46
      2.7.2  In Vitro Methods       	  47
             2.7.2.1   Whole mammalian embryo culture    ....  47
             2.7.2.2   Embryonic limb bud organ culture    ....  48
             2.7.2.3   Avian embryonic cell	  48
             2.7.2.4   Ascites tumor cell assay	  48
             2.7.2.5   Drosophila embryo cell assay	  49
             2.7.2.6   Poxvirus  morphogenesis     	    49
             2.7.2.7   Neuroblastoma cells	          ..     49
            2.7.2.8    Hydra attenuata system 	        50
            2.7.2.9    Planarian assay    	        	  50
            2.7.2.10  Frog embryo assay	      	  51
 2.8  CONCLUSIONS AND RECOMMENDATIONS
     FOR FURTHER RESEARCH	              . ...   51
     2.8.1   Conclusions ..        	          	     51
                . 1   General experimental considerations ...       5 ]
                .2   Structure-activity relationships	       51
                .3   Teratogenesis and time of administration      52
                .4   Behavioral teratogenesis	    	     52
                .5   Species comparisons .      ....          . .   52
                .6   Recent approaches in teratology testing        53
     2.8.2  Recommendations for Further Research.     ...        53
2.9   LITERATURE CITED...        	    ....     .     54
2.8.
2.8.
2.8.
2.8.
2.8.
2.8.
                               IV

-------
3.   IMMUNOTOXICOLOGY

    3.1  INTRODUCTION	69
    3.2  TESTS TO DETECT ALTERATIONS IN THE
        IMMUNE RESPONSE	  71

        3.2.1   Assessment of Cell-Mediated Immunity	72
               3.2.1.1  In vivo tests	  75
               3.2.1.2  In vitro tests    	   82
        3.2.2   Assessment of Humoral Immunity	87
               3.2.2.1  Assays for local production of antibody  ...  88
               3.2.2.2  Measurement of circulating antibody and
                      immunoglobulins	    93
               3.2.2.3  Other tests of B-cell function	  99
        3.2.3   Assessment of Indirect Parameters of Immunity   ...100
               3.2.3.1  Macrophage functions	  100
               3.2.3.2  Host resistance to infection	101

    3.3  ALLERGIC REACTIONS TO ENVIRONMENTAL
        CHEMICALS	102

        3.3.1   Allergic Response to Inhalants	106
        3.3.2   Allergic Response to Dermal Sensitizers	106
        3.3.3   Tests to Detect Sensitizing Potential
               of Chemicals   	  106
               3.3.3.1  Skin sensitization	  107
               3.3.3.2  Respiratory sensitization    	  109

    3.4  TIER TESTING   	110

        3.4.1   Moore and Faith (1976)   	110
        3.4.2   Luster and Faith (1979)    	Ill
        3.4.3   Speirs and Speirs (1979)	112
        3.4.4   Deanet al. (1979b)    	113
        3.4.5   White et al. (Unpublished Observations)    	113
        3.4.6   Luster et al. (1982b)  	116

    3.5  SUMMARY	  117

    3.6  GLOSSARY	  118

    3.7  LITERATURE CITED	  134

    3.8  GENERAL REFERENCES    	  148

4.   INHALATION TOXICOLOGY	149
    4.1  INTRODUCTION    	  149
    4.2  GENERAL TEST DESIGN    	  150
        4.2.1   Introduction	   150
        4.2.2   Test Species	   150

-------
    4.2.3  Duration and Mode of Exposure ...
    4.2.4  Dosage	   ^
    4.2.5  Observations for Effects	
           4.2.5.1  Clinical observations	^
           4.2.5.2  Biochemical and hematological tests         158
           4.2.5.3  Pathological observations	^9
           4.2.5.4  Respiratory function tests   	     162
4.3  EXPOSURE CHAMBER DESIGN	167
     4.3.1  Introduction	167
     4.3.2  Types of Exposure Systems	167
           4.3.2.1   Static systems	167
            4.3.2.2  Dynamic systems 	169
            4.3.2.3  Whole-body exposure systems	172
            4.3.2.4  Nose- and head-only exposure systems ..  .. 173
      4.3.3  Chamber Design	 174
            4.3.3.1  Chamber shape  	175
            4.3.3.2  Chamber size	178
            4.3.3.3  Chamber materials and construction	179
      4.3.4  Airflow Systems	            	182
      4.3.5  Sampling and Monitoring of Chamber Conditions ... 188
      4.3.6  Maintenance of Sanitary  Conditions  	202
      4.3.7  Specialized Equipment	 203
 4.4   GENERATION OF TEST ATMOSPHERES	 204

      4.4.1  Introduction	            	204
      4.4.2  Generation of Gas and Vapor Atmospheres   	206
      4.4.3  Generation of Aerosol Atmospheres    	208
            4.4.3.1  Introduction	208
            4.4.3.2  Generation of aerosols from solids	210
            4.4.3.3  Generation of aerosols from liquids,
                    solutions,  and liquid suspensions	    212
 4.5  CONCLUSIONS AND RESEARCH
     RECOMMENDATIONS  	   223
     4.5.1  Conclusions	      223
     4.5.2  Research Recommendations	          224
4.6  LITERATURE CITED	     225
                              vi

-------
                             FIGURES

 2.1   Representation of the susceptibility of the human
      embryo to teratogenesis, beginning with ferti-
      lization and continuing throughout intrauterine
      development  	      31

 3.1   Hypothetical mechanisms for cellular and humoral
      immune reactions	   77

 3.2   Direct fluorescent antibody test and "sandwich" test
      for staining antibody-producing cells	   92

 3.3   The four types of allergic reaction	    104

 4.1   Concentration-time relationships in a chamber
      operated for a long period of time	170

 4.2   Time-concentration curve for exposure to constant
      concentration using air-lock mechanism	170

 4.3   Schematic diagram of the Rochester exposure
      chamber	176

 4.4   Schematic diagram of the New York University
      exposure chamber	  177

 4.5   Small-scale exposure chamber based on a bell
      jar design	  179

 4.6   Whole body dynamic exposure chamber of
      1.3-m side	    182

 4.7   Cage arrangement and chamber design of Moss and
      Brown showing airflow pattern   	    183

 4.8   Schematic diagram of airflow system   	    184

 4.9   Sketch of chamber indicating the eight corner and
      one reference sampling positions	190

4.10   Limits of particle size measuring equipment	     193
                                 ¥11

-------
4.11  Hexhlet two-stage aerosol sampler 	       . .          ^

4.12.  Conicycle aerosol sampler	    •  195

 4.13   Performance curves for the hexhlet and conicycle
       samplers	     	      	           	  196

 4.14  Schematic diagram showing operation of cascade
       impactor   	      	         	197

 4.15  Schematic of Anderson Mark HI cascade impactor	198

 4.16    Calibration of an Anderson Mark III impactor	199

 4.17  Working principle of the inertial spectrometer	200

 4.18  Electrostatic precipitator using radioactive  tritium
       (H3) as an ion source	201

 4.19  Improved counter-current vaporizing apparatus,  with
       thermostating jacket       	     	207

 4.20  J-tube vaporization assembly	209

 4.21   The Wright dust feed mechanism	     ... 212

 4.22   Schematic view of Ettinger's modification of
       TimbrelFs fibrous aerosol generator  	213

 4.23   Diagram of the DeVilbiss nebulizer	      215

 4.24  Schematic view of ultrasonic nebulizer                         218

 4.25  Schematic drawing of a spinning disk generator
      used to produce monodisperse aerosols of both
      soluble and insoluble forms from solutions of
      suspensions   	                                   220

 4.26  Schematic view of Sinclair-LaMer monodisperse
      aerosol generator     ....                                      T-I
                                 Vlll

-------
                             TABLES
 2.1   Literature Survey of Experimental Parameters in
      Testing of Chemicals for Teratogenicity	11
 2.2   Thalidomide Action in Various Species	      17
 2.3   Aminopterin Action in Various Species	      22
 2.4   Methotrexate Action  in Various Species	   23
 2.5   Preimplantation Developmental Stages in Man and
      Laboratory Animals	   27
 2.6   Critical Periods of Organogenesis in Animals	     29
 2.7   Behavioral Testing Procedures	      34
 2.8   Behaviors Tested and Methods Used in Teratologic
      Evaluation	   35
 2.9   Species Susceptibility to Drugs   	     38
2.10   Thaliomide Teratogenesis in Primates	      39
2.11   Comparative Teratogenicity of Thalidomide   	      40
 3.1   Tests for Immunotoxicity	      73
 3.2   A Classification of Toxicity-Influencing Factors	      74
 3.3   Products of Activated Lymphocytes (PALS) In  Vitro	     76
 3.4   PFC Response in  Experimental Animals Following Exposure
      to Environmental  Pollutants	      91
 3.5   Definition of the Four Types of Allergic Reaction    	    103
 3.6   Respiratory Allergens	     107
 3.7   Testing Approaches for Evaluating the Immunobiologic
      Effects of Food Additives, Drugs, and
      Environmental Chemicals   	     114
 4.1   Some Physiological Indices of Man and Animals	  151
 4.2   Parameters Measured in Lavage Fluid	    160
 4.3   Organs Suggested for Microscopic Examination	    161
 4.4   Respiratory Function Tests to Evaluate Breathing
      Pattern, Lung Volumes, Pulmonary Pressures,
      and Lung Mechanics	    163
 4.5   Inhalation Exposure Apparatus: Basic Requirements	  168
                                 IX

-------
4.6  Measurement Principles Used in Continuous Monitoring
     Instruments for Gases and Vapors	
4.7  Particulate Sampling and Characterization Apparatus          192
4.8  Output Characteristics of Some Compressed Air
     Nebulizers	      216
 4.9  Output Characteristics of Some Ultrasonic Nebulizers          219
4.10   Operating Conditions at Which Model I Spinning Disk
      Was Used   	222

-------
                    ACKNOWLEDGMENTS

   The authors would like to thank Tim Ensminger, manager of the Infor-
mation Research and Analysis Section (IR&A),  Information Resources
Organization, Oak Ridge National Laboratory (ORNL), for his  support
during the preparation of this document. The  advice and support  of Drs.
Daljit  Sawhney, William Farland, and Diane Beal,  U.S. Environmental
Protection  Agency (EPA)  project officers,  and the  assistance of Helen
Warren,  Jan Pruett, and other  members of the  Toxicology  Information
Response  Center  are  gratefully  acknowledged.  Most  helpful  in  the
preparation  of  the  immunotoxicity  section  was  the  bibliography  The
Effects of Environmental Chemicals on the Immune System: A Selected
Bibliography   With  Abstracts,   1969-1980,  by   S.   G.  Winslow
(NLM/TIRC-81/2).
   Special thanks are extended to Drs. W. E.  Dalbey, N. Gengozian, and
H. R.  Witschi of the ORNL Biology Division; to  Drs. Harold Grice and
Clifford  Chappel of FDC Consultants,  Inc.;  to  Dr. Carl Wust of the
University of Tennessee at Knoxville; and to Drs. David Anderson, Larry
Chitlik, Ernest  Falke, Elaine Francis, Stan Gross,  Carole Kimmel, James
Murphy, Daljit  Sawhney, and John Whalan of the  U.S. EPA for their
technical  review and helpful  suggestions.  The authors are also  greatly
indebted to Judy Crutcher, Pat Hartman, Sherry Hawthorne, Frances Lit-
tleton, and Donna Stokes of the IR&A Publications Office for their assis-
tance in document preparation and to Carolyn  Seaborn and  Lois Thurston
of the Chemical Effects Information  Group for their  assistance in  the col-
lection and organization  of reference materials.
                                  XI

-------
                            ABSTRACT

   This document is the second of a two-part literature analysis of param-
eters  associated  with  the various toxicity testing  methods  (test animal
selection,  pathology,  etc.).  Acute, subchronic,  chronic, and  carcinogenic
testing  methods  are covered  in  ORNL/EIS-151. Testing  methods  for
developmental  toxicity, immunotoxicology, and inhalation toxicology and
research needs associated with these areas are  covered  in  this volume,
ORNL-6094. These reports were prepared for the purpose of  assisting and
supporting the U.S.  Environmental  Protection  Agency  in  its efforts to
develop guidelines for more efficient and economical testing procedures.
                                   Xlll

-------
                   1. EXECUTIVE SUMMARY

1.1 TERATOGENICITY

    Proper attention to the design of a teratogenicity experiment is essen-
tial. Parameters such as  the  number  and level of doses,  the number of
species, the number of  test animals per  dose group,  and the time during
pregnancy for the administration of the test agent are important considera-
tions.
    At least three dose levels should be tested; the low dose should permit
normal embryonic  development, the intermediate  dose(s) should produce
malformed offspring if the test chemical is teratogenic, and the  high dose
should be maternally toxic or produce embryotoxicity or fetotoxicity. At
least two species of test animals with  a  sufficient number  of animals per
dose group to permit meaningful statistical evaluation of the results should
be  used. Current guidelines recommend the use of at least 20  pregnant
animals of a  rodent species and 12 pregnant rabbits. Generally, the test
agent  is administered during  the  period of organogenesis  because this is
considered  to be the  developmental stage most sensitive to the action of
teratogenic chemicals.  The period of prenatal development that slightly
overlaps organogenesis but extends primarily  into the fetal period is known
as histogenesis.  Although teratogenic  agents that come into contact with
the developing fetus during this time can  cause minor structural deviations,
the abnormalities that are more likely to occur during the fetal period are
those involving growth or functional aspects of development.
    The term  behavioral teratogenesis  is used to  describe the branch of
teratology devoted  to studying behavioral  modifications that  result from
prenatally  administered  agents.  In addition to the histogenesis period,
administration of chemicals during organogenesis or pregestationally has
resulted in behavioral alterations in offspring.
    Although many  different animal species have been used as test animals
in teratogenicity testing experiments, no  ideal species has been  identified.
The rat and  mouse are the most  commonly used  rodent species, and the
rabbit is the most commonly  used nonrodent species. The  nonhuman pri-
mate has  received  considerable attention as a test animal because of its
anatomical and  physiological  similarity to man; however, the cost associ-
ated with the use of nonhuman  primates will likely restrict their use to
special situations (e.g., testing  a drug designed for use during pregnancy).
    In recent years alternative approaches to the standard methods of test-
ing agents for  teratogenicity  have been investigated, primarily due to
economic and practical considerations. These include an in vivo method
where determinations of  teratogenic potential are made on the basis of

-------
litter size  and weight and  several in vitro methods, such  as whole mam-
malian embryo culture,  a  Hydra attenuata system, and  a frog  embryo
assay.
    Possible areas for future research include further investigation  of post-
natal behavioral changes as  they relate to pregestational and  gestational
chemical exposure, further  research  directed toward standardization of the
methodology  for  screening chemicals  for  behavioral  teratogenicity, and
determination of  an  acceptable battery of tests for in vitro screening of
compounds for teratogenicity.

 1.2 IMMUNOTOXICOLOGY

     Investigations into the toxic  effects of industrial and  environmental
 chemicals have revealed that some of these  chemicals have distinct, and
 usually very  specific,   effects  on  the  immune  system. This   complex
 organization  of organs  and cells  plays a critical role  in the protection of
 man against  infection and neoplastic diseases; thus dysfunction of the sys-
 tem can have serious health  effects. These  issues are the basis for  the evo-
 lution  of  the field of  immunotoxicology  (the  study  of  immunologic
 alterations caused  by industrial  and environmental chemicals), and they
 have stimulated  interest  in test methods originally designed  for use in
 immunopharmacology  (to  assess  the  therapeutic  potential  of  immu-
 nosuppressant or immunopotentiating  drugs). Several toxicology groups
 have suggested the  use  of immunotoxicology assays as an adjunct to toxi-
 cology studies.
     The protective responses of the  body against foreign entities (antigens)
 are carried out  by the immune system,  which is composed of cells and
 organs of the lymphoreticular  system. Immune responses may be specific
 or nonspecific.
     The  specific  immune  response  is characterized by the highly specific
 recognition of (and  response to) antigen by lymphocytes and by the induc-
 tion of immunological  memory. Specific immunity includes  cell-mediated
 immunity  (an expression  of the activities of T-Iymphocytes),   humoral
 immunity (an expression of the activities of B-lymphocytes), and immuno-
 logical tolerance.
     When a  chemical  compound is  suspected  of altering  the   immune
response, both the cellular  and humoral responses must be tested  Tests of
cell-mediated immunity usually  measure T-cell activities, either in vivo or
in  vitro. Basic in  vivo procedures for evaluating the cell-mediated immune
response include tests of delayed  hypersensitivity, allograft reject'      H
graft vs host reactions. These are well-established, valuable procedures but
are generally cumbersome and time consuming, requiring large nu   b
animals. Newer tests have been developed, some utilizing isotone  th ^ °
more efficient, quantitative,  and  sensitive.  Some of these     h  ^ *K
radiometric ear test and  the  footpad assay, have been recomme &*A f
inclusion in toxicity screening protocols.

-------
   In  vitro tests of cell-mediated  immunity  are versatile and convenient
methods for the large-scale screening of chemicals. Tests recommended for
screening protocols include lymphocyte transformation by mitogens, mixed
lymphocyte cultures, and assays  for  the  migration inhibiton factor. For
these  assays, the  test substance  is  generally  administered  to the test
animal,  appropriate cells  are  removed, and  immunocompetence of these
cells is measured in vitro. In some assays, however, the test chemical can
be added to  the test system in vitro, thus bypassing the animal treatment
step.
   Tests of humoral immunity measure, in some cases,  both T- and B-cell
functions (T-cell dependent humoral response) and, in  others, only B-cell
functions (T-cell independent humoral response), depending on the antigen
used. Humoral immunity is generally  evaluated  on  the basis of circulating
antibody or local  production of antibody, the identification and quantita-
tion  of the various  immunoglobulins,  and the colony-forming and blasto-
genic capabilities of B-cells. Several of these  tests have  been suggested for
use in  toxicity  screening  protocols. These include assays for circulating
antibody, the classic hemagglutination and hemolysin titrations, and more
modern  techniques, such as single radial immunodiffusion (SRID)  assay
and  the enzyme-linked immunosorbent  assay (ELISA). Tests designed to
detect the local production  of antibody include the plaque-forming  assay
(localized   hemolysis   in   gel),   immunofluorescence   techniques,  and
erythrocyte-antibody-complement (EAC) rosette assay.
   The  nonspecific immune response does not  involve  the recognition of
antigen or the mounting  of an immune response but, instead, involves the
generalized activities of phagocytic cells, lysozymes, the interferon system,
the complement system, and the kinin system. Tests for macrophage func-
tion, hormone and complement activities, host resistance to infection, resis-
tance to tumor challenge, and endotoxin  hypersensitivity are typical pro-
cedures used to assess this system.
   Another facet of immunology test  procedures, also important in assess-
ing the toxic properties of industrial/environmental  chemicals, are tests for
allergencity,  which,  in contrast to immunotoxicity procedures that measure
immunocompetence, are designed to measure the sensitizing potential of a
chemical. Chemical allergens found in industrial environments, thought to
be  highly protein-reactive  chemicals,  usually  affect  a  small subset  of
workers who are hypersusceptible to the low-dose exposure legally permit-
ted in chemical plants. Sensitization can  occur via dermal or respiratory
routes,  and, in either case, the resulting allergic reaction can  have serious
ramifications for the worker involved.  Predictive tests in  animals, usually
guinea pigs, have been designed to assess the  allergenic  potential of chemi-
cals. Skin  procedures include the Draize test, Freund's complete adjuvant
test, the guinea pig maximization test, the "split adjuvant"  test,  and the
open cutaneous test. Respiratory sensitization in animals, although difficult
to attain, has been demonstrated in limited studies.

-------
   If assays  to evaluate the immunotoxic effects and allergic potential of
chemicals are to become incorporated in routine toxicity test protocols, the
selection of the proper combination of tests that would be most efficient,
economical, and informative becomes an important consideration. Experts
in the field of immunotoxicology have  made their independent recommen-
dations for the types of immunology procedures that should be included in
routine toxicity studies and have designed tier-testing schemes to demon-
strate how this can be accomplished.

1.3 INHALATION TOXICOLOGY

 1.3.1 General Test Design

    Toxicity tests designed to evaluate effects from inhalation  exposure
 involve many of the same  variables as tests utilizing other exposure routes.
 The selection of test species should be based on the anatomical and physio-
 logical similarity of the respiratory tracts of humans and the test species,
 as well as metabolic similarities and economic/practicality considerations.
 Larger animals  (e.g., horses,  monkeys) are  generally more similar to
 humans,  but  practical restraints  favor the use of smaller animals (e.g.,
 rats).
     Test  durations for inhalation  studies include acute (one exposure with
  14  days  of observation), subchronic (repeated  exposures for  14, 28, or 90
 days), and chronic (repeated exposures for 1 to 3  years). Due to the large
 costs involved in a  chronic  study, these should be undertaken only when
 preliminary studies  indicate the need for an  evaluation of  lifetime effects.
 Exposure to the  test agent can be intermittent (6 to 8 h per day, 5 days
 per week), based on occupational exposure, or continuous (22 to 24 h per
 day,  7 days per week)  to simulate environmental exposure. Intermittent
 exposure  allows recovery  from the test agent  (which can affect toxicity)
 but is generally simpler to  maintain and operate.
    Inhalation studies with repeated exposures should include at least three
 dose levels  and suitable controls. The determination  of the actual dose
 delivered to the test animal is a major difficulty in inhalation studies. The
common  method is to calculate the dose based on the concentration (C) in
the  test chamber multiplied by the length of exposure (T). However  this
indicates  only the maximum possible dose because not all of this quantity
is inhaled by the animal or, once inhaled, reaches the lungs. Dosimetric
formulas  have been  developed that consider the  volume intake of the lun&s
and  the percentage of the agent retained in the lungs. Unfortunately, these
variables  are difficult  to determine under actual use  conditions  and  th
CXT estimation of dose is still widely used.
   The evaluation  of toxic  effects resulting  from  inhalation  exposure
includes many of  the procedures used  in other  toxicity studies as   11
procedures specific for inhalation-caused  effects. Clinical observations of
the animals  should  be made  daily, before,  during,  and after
These include signs of irritation, behavioral changes, and  changes n

-------
functions. Biochemical and  hematological  tests should be performed on
blood samples (taken before and after  exposure) to aid in evaluating sys-
temic effects. Biochemical tests for evaluating lung damage are currently
being modified from similar tests used to evaluate hepatotoxicity. Patho-
logical  techniques  are the  most  useful evaluations for assessing  toxic
effects from  inhalation exposure. A thorough gross examination of all test
animals is necessary with special attention given to the respiratory system.
Microscopic  evaluations should be  performed on  animals  from the high
dose and control groups as well as  any  lesion-bearing animals found in the
gross examination.  The organs to be examined should cover any potential
targets  for systemic  toxicity. Respiratory tissues should be  examined  in
detail, using special  techniques  such as histochemistry, when indicated.
Respiratory function tests are techniques for evaluating inhalation expo-
sure effects that require special equipment and training.  They can quantify
changes at low-dose levels and can produce dose-response data.  However,
they are not  conclusive by themselves and should be correlated with other
toxicity evaluations.

1.3.2 Exposure Chamber Design

    The design of the inhalation exposure  chamber  is an important  con-
sideration in an  inhalation  study. It can be operated  in  a  static mode,
where a predetermined dose is introduced into a closed chamber, or in a
dynamic mode, where  a  continuous flow of air and test agent is vented
through the chamber. Static exposures  are  limited by problems associated
with the finite volume of air used without replacement and are generally
most useful for short-term  exposures or  when the test agent is available
only in limited quantities. Dynamic exposures are used for the majority of
inhalation tests and require accurate calculations of airflow and test agent
feed rates to achieve the  desired doses. The chambers are most frequently
designed to house the whole animal within the exposure area, particularly
for long-term studies. Nose- and head-only systems can  be used when der-
mal or  accidental oral exposure  must be avoided.  These chamber designs
require special equipment to restrain the animals and seals to prevent the
loss of test agent.
    The shape of most large, dynamic, whole-body exposure chambers is
based on a cubical or hexagonal design. Pyramidal or conical additions to
the top and bottom of the chamber increase the distribution efficiency of
air within the  chamber,  making the design  closer in operational charac-
teristics to the  most efficient shape, the sphere.  The  most widely  used
chambers of this  design  are the Rochester (hexagonal)  and New York
University  (cubical)  designs.  For  carcinogenesis  studies, more  isolated
chambers or  exposure rooms are often used because of the long duration of
the tests and the need to limit exposure to  the test compound. Small-scale
exposure chambers designed for pilot or acute studies occur in more varia-
tions including  bell  jars, Lucite cylinders,  and  scaled-down versions  of

-------
 larger units. Flexibility is important  in these designs since  the  unit will
 usually be used for many types of chemical exposures.
    Chamber  size  is governed  primarily  by air distribution and  animal
 volume  considerations.  The  maximum animal  volume  should rarely  be
 more than 5% of the total chamber volume, and chambers of 1 to 8 cubic
 meters usually provide  the best air distribution.
    The chambers are usually constructed of smooth, nonabsorbent materi-
 als that are resistant  to a wide range of chemicals. Stainless steel walls
 and glass  or Plexiglass® observation ports and doors are the  most common
 materials. Rubber or  plastic seals are necessary to prevent air loss and
 should be resistant to  the test chemical. Animal cages are generally made
 of stainless steel mesh without  any solid surfaces. Their arrangement
 should  facilitate exposure  with rotation  of  the individual  positions for
 long-term studies.
    Airflow  systems for the chamber should  provide  units  to condition,
 filter, meter, and control the flow of the air, test agent,  and  their mixture.
 The rate of airflow is  determined by the animal loading and the test agent
 feed rates. Various meters and  valve arrangements are used  to provide the
 necessary control, with valves located downstream from  filter units  to
 prevent clogging. An air pump is usually located in the exhaust line to pro-
 vide the  negative operational  pressure (0.1 to 0.5 in.  of water) for the
 chamber. Air enters  the chamber through the top pyramid  and  usually
 exits through an arm  of a Y-joint at the bottom pyramid. The air needs to
 be conditioned  to control temperature and humidity within  an acceptable
 range  (75 to 82°F and <55% humidity).  Automatic control systems for
 this are available. Filter banks should be  placed in the main supply line
 and in  the chamber and main exhaust lines. Usually several  types of filters
 are included in each bank to ensure satisfactory removal of contaminants.
     During operation  of the chamber, samples must be taken to monitor
 the exposure  conditions, including temperature, humidity,  and test agent
 concentrations (also  particle  sizes   for  aerosol atmospheres).  Sampling
 probes  should monitor all areas of the chamber distribution with added
 emphasis on conditions at the breathing zone of the animals.  Sampling and
 monitoring techniques  for vapor and gas atmospheres usually utilize filters
 with standard analytic techniques. Aerosol atmospheres  are  more difficult
 to sample, and a  whole array  of samplers  have been developed for this
 including elutriators, cascade impactors, centrifugal  force samplers  precip-
 itators,  and  optical monitors.   These usually   include  some  me'ans  for
 separating samples into size classes to determine particle  size  numbe    d
 concentration.
    To maintain  sanitary conditions within  the  chamber, techniques and
 equipment have been developed  to remove animal wastes Us  11
 manent  spray  ring  is  used  to  wash  down the chamber walls  with^
 appropriate cleaning solution. Cleaning procedures must  be followed trf^
prevent  cross  contamination  of different chambers  and  that    t
animal technicians.                                            °      e

-------
   Occasional  use of specialized chambers or equipment may be  neces-
sary.  Examples of such equipment include chambers that operate under
reduced or increased pressures and apparatus to evaluate effects of animal
activity during exposure to test agents.

1.3.3 Generation of Test Atmospheres

   Inhalation testing requires that the test agent  be put into an atmos-
pheric form  for  exposure.  Atmospheres of gas or vapor  test agents are
relatively simple  to generate, compared to substances that must be in an
aerosol form.
   Gases and  vapors can  be classified  as irritants  or asphyxiants. One
common method  for generating gas and vapor  atmospheres is  to store the
chemical  in  high pressure  cylinders.  For  batch introduction,  air  can be
mixed into the cylinder, while for dynamic systems  the gas can be released
into the airstream. Porous diffusion plugs or containers can also be used to
release gases. If  the substance is not volatile or chemically stable enough
to store  in  cylinders, it can be vaporized through controlled heating or
counter-current techniques.
   The generation of aerosols must include techniques for many types of
aerosols including dusts, fumes, smoke, mists, and fogs  using solid or fluid
feed materials. Aerosols occur in monodisperse forms where the  particles
are all within  a  narrow  size range or  polydisperse forms that include a
wide  range  of particle sizes. Monodisperse aerosols are more useful for
studying particle  size effects since they simplify the experimental condi-
tions, but polydisperse aerosols are more typical of actual exposure condi-
tions. The division  of the parent material into particles or droplets of a
small enough size to remain airborne intensifies the chemical and physical
activities of the test agent because of the increased surface area and total
space occupied.
   The generation  of aerosols from  solid test substances has been  per-
formed using loose  and compacted forms of the particles. Loose  particles
must  be kept  in  suspension by  vibrating,  rotating, or  stirring techniques
usually combined with an aspirator to distribute the particles. Such tech-
niques are  used  more often  for  handling fibrous dusts  or  to  produce
polydisperse  aerosols. The  most common technique for  generating dust
aerosols is to use a compacted plug of powder and then  scrape  off particles
to redistribute. The Wright dust feed generator is a typical apparatus of
this design and uses a horizontal, rotating scraper blade and a packed plug
of powder in a brass cup. By controlling the humidity of the  distributing
air stream, plus  the packing density  and particle sizes of the plug, there
should be no problem with redistributing the particles.
   The generation of aerosols from liquids, solutions,  or  suspensions can
be accomplished  with various techniques. Aspirators, which use a pressure
drop across an orifice and  a stream of pressurized air to create  droplets,
have  been widely used for  generating liquid  aerosols. These  are usually
combined with an impaction surface to limit the particle sizes produced. A

-------
                                    8

typical example of such an air pressure-driven nebulizer is the Vaponeph-
rin nebulizer,  which uses  a glass  sphere directly  opposite  from  an air-
stream jet passing over a  capillary tube to create  monodisperse aerosols.
Ultrasonic vibrations from a  piezoelectric crystal have also been  used  to
fractionate  liquids  in  a nebulizer  apparatus.  The most  common liquid
aerosol-generating technique is the spinning disk or top. These use centri-
fugal force to separate particles from a rotating flat surface  upon which
the test  liquid is  applied.  Such apparatus are capable of producing highly
monodisperse aerosols. Another technique for creating monodisperse aero-
sols is based on the controlled condensation of heated vapors on  suitable
nuclei.  Most  of  the  above  methods can  also  produce dry particles by
adding dry or  heated air to the droplets after they are generated.

-------
                      2. TERATOGENICITY

                            Robert H. Ross

2.1 INTRODUCTION

   The evaluation  of agents for teratogenic effects and  their impact on
human health is an area that only recently has become a  significant regu-
latory  concern. As  Wilson (1979) states, "The  notion that chemical and
physical agents in the environment need to be tested for their potential to
cause teratogenic effects when there is likelihood of exposure during preg-
nancy  is  a relatively new concept."  It was not until  the  thalidomide
tragedy in the early  1960s that the world's attention was focused  on  the
fact that the human embryo is not isolated in an impervious maternal body
where it  is shielded from all but genetic harm. Wilson points out that  the
year 1966 probably deserves  special recognition  because of the significant
efforts initiated at  that time to minimize the risks to  the unborn popula-
tion from radiation (by the International Committee for Radiation Protec-
tion Recommendations)  and  to minimize such  risks from drugs  (by  the
Food and Drug Administration Guidelines).
   This  chapter  will be  primarily  limited  to a discussion of chemically
induced  congenital malformations  and  functional impairment. The first
section will discuss some basic test parameters,  the  second  section will
examine  the influence of the time of administration with  regard to induc-
ing teratogenicity, the third will explore the science of behavioral  terato-
genicity,  the fourth will discuss the choice of species for teratological test-
ing, and the last section will address some of  the recent approaches in
teratogenicity  testing. The terms embryolethality or fetolethality will refer
to the  death of the embryo or fetus, teratogenicity will refer only to mal-
formed offspring (both morphological and behavioral), and embryotoxicity
or fetotoxicity  will  refer to teratogenic,  growth  retardation, abortifacient,
or intrauterine death responses of offspring to a chemical.

2.2 GENERAL EXPERIMENTAL CONSIDERATIONS

   Some  basic parameters of testing chemicals  for  teratogenicity include
dosage (number, levels, and duration), positive controls, number of species,
number of test animals  per  dose group, administration  route, and fetal
examination.
   In reviewing teratogenicity studies with thest  parameters in mind, it is
important to distinguish  between screening studi   -u d those aimed at  the
further delineation  of a teratogenic effect or thos^; lesigned to investigate
the toxicology  of a clinically used  dose. The protocols used in the latter
two instances will often be different from those  for the  screening study.

-------
                                 10

Most of the comments in the following subsections will be directed to the
screening  study. Table 2.1 presents a brief insight into the protocols  used
by different researchers when conducting screening studies. The selection
of studies in this table was determined by  searching the recent issues  of
(1) Teratology, (2)  Teratogenesis, Carcinogenesis,  and Mutagenesis, and
(3) Toxicology and Applied Pharmacology.

2.2.1 Dosage—Number and Level
    The Environmental Protection Agency (EPA) (USEPA 1982a, 1982b),
the  Organization for Economic Cooperation  and  Development (OECD)
(OECD 1982), and the Food  and Drug Administration  (FDA) (USFDA
 1982) each recommends the use of at least three doses. An examination of
 the data  in Table 2.1 (from 20 relatively recent research efforts) indicates
 that most researchers are apparently testing at least three concentration/
 dose levels.
    Ideally, the high dose level should produce either  maternal toxicity (but
 not  lethality),  embryotoxicity,  or  fetotoxicity  [Wilson  1975b; World
 Health Organization (WHO) 1967]. The  low dose  should permit normal
 embryonic development, and the intermediate dose(s) should produce mal-
 formed offspring if the test chemical  is teratogenic (Collins  and Collins
 1976). Extrapolation between the low dose and the intermediate dose or
 doses would provide an indication of the effect/no effect range of the test
 chemical.

 2.2.2 Dosage—Duration
     Schardein  (1976) states that, in general, acute  dosing of  a chemical
 results in a greater  teratogenic insult than  does prolonged dosing. One
 example  is that of the anticancer and  antibiotic drug, actinomycin  D, in
 rats. Malformations were induced in 27.9% of the surviving offspring with
 a single dose of 200 Mg/kg given on day 9  of gestation, but when the drug
 was given as ten daily injections of 25 ng/kg on each of days 0 to 9 of ges-
 tation,  only 8.7% of the surviving offspring was malformed (Wilson 1966).
 Comparable results  were seen  with the number  of  dead  and resorbed
fetuses—32.5% were  either  dead or resorbed  with  the  single  200-^g/kg
dose but  only  9.7% were dead or resorbed  when 25 /tg/kg was given on
each of days 0 to 9  of  gestation. Further experimentation  demonstrated
that the dam was sensitized by prolonged treatment with actinomycin D at
high doses so  that for the dam  or  her  offspring even moderately terato-
genic doses become lethal. The stimulation of drug-metabolizing enzymes
in the liver microsomes by prolonged dosing was suggested as the possible
mechanism of this phenomenon, which has been observed by Koppanyi and
A very in  experiments with more than   100 drugs  (Koppanyi  and Avery
1966, as reported by Schardein  1976).
   Similar results were demonstrated with the  administration of the benz-
hydrylpiperazine  antihistamine  drug chlorcyclizine   (King  et  al  1965^

-------
                              Table 2.1. Literature Survey of Experimental Parameters in Testing of Chemicals for Teratogenicity
     Chemical
Species—number
    per group
Number of doses/concentrations,
      duration, and route'
                                                                                                   Extent of
                                                                                               fetal examination
Positive
control
Reference
Concanavalin A
Rabbit—7 to 11     1  dose level—160 fig intracoelomically
Indole-3-acetic acid
Octoxynol-9
Chlordiazepoxide
(Cdz) and
amitriptyline (Amt)

Tridemorph
Tetrachlorobenzene
isomers

ONO-802"
Rat—26 to 40;      4 dose levels—5, 50, 200, or 500
mouse—25 to 38    mg/kg by gavage on gestation days 7

Rat—25            2 dose levels—0.5 or 5 mg/kg/day intra-
                    vaginally on days 6 to 15 of gestation

Hamster—11  to     1 dose level—28.5 mg/kg Cdz HCL or
   18                70.3 mg/kg Amt HCL or combination
                    intraperitoneally on day 8 of gestation

Rats—21 to 27;     3 dose levels—20.6, 60.2, or 189.2 mg/kg
mice—22 to 27     for rats and 27.5, 81.7, or 245.1 mg/kg
                    for mice on days 6 to 15 of gestation

Rat—10            3 dose levels—50, 100, or 200 mg/kg by
                    gavage on days 6 to 15 of gestation

Rabbit—19 to 23   3 dose levels—0.0125, 0.0625, or 0.25
                    mg/kg intravaginally on gestation days
                    6 to 18
                                      External (all live fetuses), skeletal            No      DeSesso 1979
                                      (all live fetuses), soft tissue (internal
                                      organs of all live fetuses examined
                                      grossly and selected fetuses prepared
                                      for visualization)

                                      External (all fetuses), skeletal (all            No      John et al. 1979
                                      fetuses), soft tissue (1/3 of fetuses)

                                      External, (all fetuses), skeletal (2/3 of        No      Saad et al. 1984
                                      litter), soft tissue (1/3 of litter)
                                      External, skeletal, and soft tissue (all
                                      fetuses)
  No      Beyer et al. 1984
                                       External (all fetuses), skeletal and soft        No      Merkle et al.  1984
                                       tissue (proportions not specifically stated)
                                       External (all fetuses) skeletal (2/3 of         No      Kacew et al. 1984
                                       litter), soft tissue (1/3 of litter)

                                       Skeletal (1/2 of fetuses), soft tissue           No      Petrere et al. 1984
                                       (1/2 of fetuses)

-------
                                                                  Table 2.1 (continued)
Chemical
Species — number
per group
Number of doses/concentrations,
duration, and route*
Extent of
fetal examination
Positive
control
Reference
Cyproheptadine         Rat—7 to 61C       2 dose levels in one series (25 or 50
chlorhydrate                               mg/kg) and 3 in another (15, 25, or 35
                                          mg/kg); all given by oral intubation on
                                          gestation days 6 to 15
Hexamethylmelamine   Rat—15 to 30;
                       Rabbit—15
Trimethyl phosphite     Rat—25
                                          Rat: 3 dose levels—10, 20, or 40
                                          mg/kg by gavage on gestation days
                                          6 to 15 or for 4-day periods during
                                          organogenesis; rabbit: 3 dose levels—20,
                                          40, or 60 mg/kg by gavage on gestation
                                          days 6 to 18

                                          3 dose levels—16, 49, or 164 mg/kg by
                                          gavage on gestation days 6 to  15

Triamcinolone          Rat—3             3 dose levels—0.125, 0.25, or 0.5 mg/kg
                                          intramuscularly on gestation days 9 to
                                          11,  12 to 14, or 15 to 17

Methyl chloride         Rat—25;           Rat: 3 concentrations - 100, 500, or  1500
                       mouse—33         ppm by inhalation  on gestation days 7 to
                                          19 (6  h/day); mouse: 3 concentrations—
                                          100, 500, or 1500 ppm by inhalation on
                                          gestation days 6 to 17 (6 h/day)
External (all fetuses); fetuses with
abnormal appearance studied for soft
tissue and skeletal defects, other
fetuses randomly selected for soft
tissue or skeletal examination

Rat—external (all fetuses), skeletal
(~2/3 of fetuses), soft tissue (—1/3
of fetuses); rabbit—all fetuses exam-
ined for external skeletal and soft
tissue malformations


External (all fetuses), skeletal (2/3 of
fetuses), soft tissue (1/3 of fetuses)

External (all fetuses), skeletal (1/2 of
fetuses), soft tissue (1/2 of fetuses)
                                                                                   Skeletal (1/2 of fetuses), soft tissue (1/2
                                                                                   for head and fetal trunk examination)
                                                                                                                             No
                                                                                                                             No
                                                                                                                             Yes
                                                                                                                             No
                                                                                                                             No
                                                                                                                                     Rodriguez-Gonzalez
                                                                                                                                     et al. 1983
Thompson et al.
1984
Mehlman et al.
 1984

Rowland and
Hendrickx 1983
                                                    Wolkowski-Tyl
                                                    et al. 1983

-------
                                                                   Table 2.1. (continued)
Chemical
Acetonitrile
Species - number
per group
Hamster — 6 to 12
Number of doses concentrations,
duration, and route*
4 concentrations/dose levels — 1800,
Extent of
fetal examination
External (all living fetuses), ratio of
Positive
control
No
Reference
Willhite 1983
Delalutin
Cocaine hydro-
chloride
Valproic acid
                      Mouse—11 to 14d
Rat—8 to 12";
mouse—8d
Mouse—5 to 21
 2-Nitro-p-phenylene   Mouse—25 to 39
 diamine
3800, 5000, or 8000 ppm by inhala-
tion on gestation day 8; 100, 200, 300,
at 400 mg/kg by intubation or intra-
peritoneally on gestation day 8

3 dose levels—42, 416, or 833 mg/kg sc
on gestation days 6 to 15

Rat: 3 dose levels—50, 60, or  75 mg/kg
mg/kg intraperitoneally on gestation days
8 to 12; Mouse:  1  dose level—60 mg/kg
intraperitoneally on gestation days
7 to 16

5 dose levels—1.0, 1.2, 1.4,  1.65, or 2.7
mmole/kg intraperitoneally on days 8 to 10
of gestation; 2 dose levels - 1.0 or 4.4
mmole/kg on days 11 to 13 of gestation
                    7 dose levels—32, 64, 128, 160, 192,
                    224, or 256 mg/kg sc on gestation
                    days 6 to 15
                                                                                       fetuses examined for skeletal and soft
                                                                                       tissue defects not given
                                                                External (all fetuses), skeletal (—1/2 of     Yes     Seegmiller et aL
                                                                fetuses), soft tissue (-1/2 of fetuses)                1983

                                                                External (all fetuses), 3:1 ratio for fetal      No      Fantel and
                                                                examination of soft tissue to skeletal                 Macphail 1982
                                                                All live fetuses examined for external,        No     Kaoetal. 1981
                                                                skeletal, and soft tissue abnormalities
                                            External (all live fetuses), skeletal (all      No      Marks et al. 1981
                                            live fetuses), soft tissue (at least 1/3
                                            of each litter, as well as all stunted
                                            fetuses and  those with external
                                            malformations)

-------
                                                                 Table 2.1. (continued)
Chemical
Phencyclidine
Species - number
per group
Mouse — 18 to 45
Number of doses concentrations,
duration, and route*
4 dose levels— 60, 80, 100, or 120 mg/kg
Extent of
fetal examination
External (all live fetuses), skeletal (all
/ , i -* 1 /i
Positive
control
No
Reference
Marks et al. 1980
                                        by gavage on gestation days 6 to 15
o-Toluenediamine   Rat—22 to 25;       Rat: 4 dose levels—10, 30, 100, or 300
                    rabbit—14 to 16     mg/kg orally on gestation days 6 to 15;
                                        rabbit: 4 dose levels—3,  10, 30, or 100
                                        mg/kg orally on gestation days 6 to 18

Acrylonitrile        Hamster—NGC      4 dose levels—0.09, 0.19, 0.47, or 1.23
                                        mmole/kg intraperitoneally on day 8 of
                                        gestation

L-Azetidine-2-      Hamster—5 to 16"   4 dose levels—100 mg/kg on gestation
carboxylic acid                          days 7 to 12; 200  mg/kg on gestation days
                                        7 to 9, 7 to 12, or 10 to  12; 300 mg/kg on
                                        gestation days 9, 10, or  11; 600 mg/kg on
                                        gestation days 7, 8, 9, 10, 11, or  12; all
                                        doses  administered intraperitoneally
live fetuses), soft tissue (at least 1/3
of each litter, as well as all stunted
fetuses and those with external
malformations)

Skeletal (1/2 of each litter), soft tissue
(1/2 of each litter)
Skeletal (all live fetuses)
External (all live fetuses), skeletal (1/3
of each litter), soft tissue (10 fetuses
from remaining specimens in each
group)
                                                                                                                             Yes     Becci et al.  1983
                                          No      Willhite et al. 1981
                                          No     Joneja 1981
"Although negative controls not specified, each experiment used them.
bSynthetic E, postaglandin.
'All but the high dose group had at least 20 animals.
dNumber of litters examined.
°NG = not  given.

-------
                                  15

Chlorcyclizine  hydrochloride administered to pregnant rats in  25-mg/kg
doses over a 4-day period (days 12 to 15  of gestation) produced 16% mal-
formed young;  the same dosage given over a 15-day period  (days 1 to 15)
resulted  in  malformations in  only 2% of the offspring. The  results  were
even more striking when a 50-mg/kg dose was administered. Treatment on
days 10 to 15 of gestation  resulted in 82% malformed individuals, but only
0.1% malformed offspring  was produced when a dosage of 50 mg/kg was
given over days 1 to  15  (implantation was not inhibited).  A subsequent
paper (King et  al. 1972) showed  that the explanation behind this lack of
teratogenic activity from prolonged treatment (defined in this instance as  a
15-day period)  of  chlorcyclizine was  the drug's self-stimulation of its own
metabolism.
    Wilson (1975b)  and the WHO  (1967)  also report that the repetitive
administration  of certain chemicals during pregnancy can alter or mask  a
teratogenic action because of the ability of these chemicals to  change their
own metabolism. Wilson states three  ways that some chemicals can change
the  rate  of their own metabolism: (1) the induction of  catabolizing
enzymes (microsomes) in the liver or other tissues, which increases  meta-
bolism  (also mentioned above);  (2)  the inhibition  of naturally present
enzymes which degrade chemicals; and (3) the induction of  impaired func-
tion or overt pathology in important homeostatic organs such as the liver
or kidneys. He  further states that it is assumed these changes can occur
within three or four days of the beginning of repeated treatments.
    What then  is the optimum  dosing period for testing chemicals for tera-
togenicity?  The foregoing  paragraphs suggest that dosing for only one day
or  for  short periods of three to four days  during  organogenesis  would
enable  detection of chemical  teratogens whose teratogenicity might go
undetected  if dosing duration was increased. Wilson  (1975b)  recommends
both short-term and repeated  tests.  His  recommended procedure consists
of dividing the  period of organogenesis into  shorter dosage periods of three
to four days but also  dosing some animals throughout organogenesis. He
states  that  fewer  animals would be needed  in each dose  group than  if
dosage were given throughout organogenesis because the range of likely
developmental defects, and hence overall variability,  would  be reduced by
the shorter treatment span. As  mentioned, some  animals  would still be
dosed  throughout organogenesis  to  detect  any teratogenic  effects  that
would result from the  cumulative action of the chemical. Collins and  Col-
lins (1976) list  three reasons why the single-dose technique—which,  they
report, is recognized as the most successful in producing abnormalities—is
not used in place of  the repetitive  dosing method  in screening studies.
These are: (1)  the necessity of treating different animals for each day  of
organogenesis  increases the cost of the experiment, (2) without  prior
knowledge  of  the compound's  effects,  there  is  no  way  of predicting
whether a  specific  organ or  the entire  developmental sequence will be
affected,  and  (3)  compounds  that are likely to  be tested  for teratogenic

-------
                                  16

action are usually encountered by humans as repetitive doses, and with sin-
gle dosing cumulative effects cannot be measured.
   Current guidelines  (OECD  1981,  USEPA 1982a,  1982b,  USFDA
1982) recommend dosing during the  period  of organogenesis.  The critical
exposure  periods for  induction  of behavioral  changes  in  the postnatal
animal are at present not well defined, and it is conceivable that  existing
guidelines will be amended as more information becomes available.

2.2.3 Positive Controls
   Although positive controls  are  not recommended in current guidelines
for teratogenicity testing (OECD  1981,  USEPA  1982a, 1982b, USFDA
1982), they have been used to demonstrate that the test animal will pro-
duce malformed offspring after exposure  to  an  established teratogenic
chemical as  well as to validate the soundness of a specific laboratory's or
researcher's  methodology. This use  of positive controls  in  teratology has
not been completely abandoned (Table 2.1).
    The  data generated from positive controls must be evaluated with some
caution  because an animal species can  be very sensitive to one teratogen
but  only mildly or not at all sensitive to another. Table 2.2 shows the tera-
togenic  effects  from thalidomide in  several animals, including man. Man,
 monkeys,  and rabbits are shown to be susceptible  to the teratogenic action
 of thalidomide, but several strains of rats, mice, and hamsters are not (at
 least by conventional testing protocols). The action of the human teratogen
 aminopterin (Table  2.3)  also  shows apparent species variability. Terato-
 genic effects have  been seen in man and mice but not  in rats  and monkeys
 (Table 2.3). The data in these tables clearly demonstrate that an animal
 may be susceptible to the teratogenic action of one chemical but not neces-
 sarily to that of another chemical. Therefore,  if a positive  control is esta-
 blished  in the animal species being  tested  and the test  chemical  does not
 produce teratogenic  effects, it cannot be  assumed that the test chemical
would not induce teratogenic effects  in another animal species or even
another strain of the same species. Conversely, if an established teratogen
such as  thalidomide  fails to induce teratogenicity  in an animal being used
as a positive control,  it cannot be assumed that the test chemical will not
produce teratogenic effects in the same species.

2.2.4 Number of Species

   As Tables 2.2, 2.3, and 2.4 indicate, the choice of species or strain can
determine whether or not a chemical is identified as a teratogen. Species
variability to potential teratogenic agents would thus dictate  that at least
two  species should be used. Current EPA guidelines recommend the use of
at least  two species,  with the rat  and rabbit being  the  preferred  species
when only two species are used (USEPA 1982a, 1982b). The  FDA guide-
lines (USFDA 1982)  also state that these are the preferred species.

-------
                                                 Table 2.2. Thalidomide Action in Various Species
Animal
Species
 strain
 Dose, route
and duration
                                                                                      Effects
         Reference
Man
Mouse    Schofield

          A, C3H, Swiss
          CF,, ICR, C57,
          CBA, SJL

 Rat
              Dose not specified, oral adminis-
              tration during first trimester

              Dose and duration not known; oral
              administration
              400 mg/kg orally throughout
              pregnancy

              A-31 to 625 mg/kg/da/; C3H,
              Swiss - 62 mg/kg/day; admin-
              istered orally on days 6 to 7 of
              gestation and continued for 4 to
              6 days
              200 mg/kg total given orally for
              gestation days 2 to 12

              2 g/kg total ipb on gestation day
               12, days 11 and 12, or days 12
              and 13
                          Missing limbs; auricle or pinna of ear decreased
                          in size; lesser deficiencies
                          Reports of several cases—hypoplasia and aplasia
                          of the extremity or individual bones, absence
                          and malformation of limbs; malformations of
                          other organs such as stenoses and malrota-
                          tion of the gastrointestinal tract and dysplasias
                          of the external ears and eyes often associated

                          No malformed fetuses in albino mice
                          Skeletal malformation; open eye; enlarged skull;
                          curvature of back; kinky tail; phocomelia;
                          micromelia
                          No malformed offspring but increased resorptions
                          in CF, strain

                          16 litters - two had abnormalities, complete
                          resorption in one, more than 50% resorption
                          in three, and normal offspring in ten; abnor-
                          malities consisted of stunting of extremities
                          and tails, absence of dorsal and  lumbar
                          vertebrae and ribs,  absence of digital bones,
                          and curved bones of the fore and rear extremities
Smithells 1962


Pfeiffer and Kosenow 1962,
Lenz and Knapp 1962
Somers 1963

DiPaolo et al. 1964





Fratta et al. 1965


Murphy 1962

-------
                                                   Table 2.2 (continued)
Animal Species
strain
Wistar
Wistar
Dose, duration,
and route
200 or 400 mg/kg total given
orally throughout pregnancy
45 mg/kg ivgon day 10, 11, or 12
of gestation
Effects
Fetal resorptions but no abnormalities
Skeletal deformities of thoracic ribs and of
spinal column in 33 and 56% of fetuses on
Reference
Somers 1963
Parkhie and Webb 1983
Sprague-Dawley    10 (ip), 20 (orally), 50 mg
                   (orally or ip) per day at
                   various dosing schedules be-
                   tween days 7 and 17 of gestation
Charles River      100 mg/kg 6 to 12 total given orally
                   on gestation days 6 to 12
Long Evans and    150 mg/kg total given orally on
Dunning Fischer    gestation days 3 to 12

Holtzmann         25-500 mg/day orally or ip at
                   various dosing schedules
                   during organogenesis
day 11 and 12, respectively; deformities of
eyeball as a result of administration on
day 10 or 12
520 young examined—36 grossly malformed (6.9%),
40 showed malformations after clearing of fetuses,
76 total malformations (14.6%); malformations
included malrotation of hind limbs, hamartoma of
the palate with accessory incisors, lack
of a tail in one instance, and a subcutaneous
cartilaginous-type mass of tissue from the
middorsal region to the tail

Resorptions  but no malformations


No apparent injury
More resorptions than controls; only one fetus
showed serious limb malformation: sig-
nificant teratogenic effect shown by missing
sternebrae and delayed ossification of
the sternum; malformations different
from humans
King and Kendrick 1962
                                 oo
Delahunt et al.  1966

Fratta et al. 1965
                                                   Moore 1965

-------
                                                             Table 2.2 (continued)
Animal
Species
 strain
Dose, duration,
   and route
                                                                                      Effects
                                                      Reference
Hamster   Not given


           Not given


           Inbred and random
Monkey   Rhesus and
           stump-tailed macaque
           Bonnet
           Baboon, bonnet,
           and cynomolgus
           Marmoset
                Up to 8000 mg/kg throughout
                pregnancy
                150 mg total given orally for
                gestation  days 3 to 12

                350 mg/kg/day orally throughout
                pregnancy
                 5 mg/kg/day orally on
                 gestation days 26 to 28
                 or 24 to 30 or 10 mg/kg on days
                 25 to 27, 24 to 30,  or 23 to 29

                 10 mg/kg orally on day 24, 25,
                 26, 27, 28, or 29 of gestation;
                 30 mg/kg orally on day 25 or 28

                 17.5-300 mg/kg total given
                 orally at various dosing
                 schedules on gestation days
                 18 to 43 (both single  and
                 multiple doses)

                 25 mg/kg/day orally on
                 gestation days 38 to 52
                         Not teratogenic
                                                                      Somers 1963
                                                                      Fratta et al. 1965
                                                                                                                Homburger et al.  1965
Not teratogenic


Inbred lines showed 6.2% incidence of grossly
malformed fetuses—acrania or split cranium,
abnormal positioning  of legs, kinking of
tail, and cleft palate;  random bred lines
showed no significant teratogenicity

Severe deformities in  most species—
missing digits, missing radii and  ulnae,
shortened humeri, kinking and in some
cases shortening of tail
Temporal and mandibular bones malformed     Hendrickx and Newman 1972
                                                                     Vondruska et al. 1971
                          Malformed fetuses - deformed limbs, spina
                          bifida, kinked tail, etc
                          Reduction in size of spinal cord and
                          dosal root ganglia; pattern of deformities
                          said to be similar to man
                                                                                                                Hendrickx 1970
                                            McBride and Vardy 1983

-------
                                                             Table 2.2 (continued)
 Animal
Species
 strain
Dose, duration,
   and route
                                                                                       Effects
                                                                                                         Reference
Rabbit     New Zealand
           white and
           Dutch-Belted
           New Zealand white
           New Zealand white
          Japanese white
                Not given
                150 mg/kg/day orally
                on days 8 to 16 of
                pregnancy

                150 mg/kg/day orally on
                gestation days 8 to
                16 or 2 to 9
                                 50, 150, or 300 mg/kg/day
                                 oralJy on gestation days
                                 6 to 18
                          Low incidence of deformed appendages in
                          New Zealand whites; increased incidence
                          of limb defects in Dutch-Belted—
                          agenesis or hypogenesis of the ulna,
                          radius, tibia, and/or fibula; absence
                          of some bones in limbs
                          Limb deformities in almost every litter
                         Sixty-seven percent of kits deformed
                         when administered on days 8 to 16—most
                         predominant effects seen in limbs
                         (arthrogryposis, micromelia, absence of
                         digits), but developmental failure of
                         visceral organs (kidneys, adrenals) and
                         in bony structures also observed; only
                         17% abnormalities in the one litter
                         exposed on days 2 to 9
                         External skeletal and internal anomalies
                         significantly (.P<0.01) different from
                         controls at 150 and 300  mg/kg day;
                         external anomalies included contracture
                         of forearm and club foot
                                                                                                Delahunt 1965
                                                                                                                 Somers  1963
                                                                                                                 Drobeck et al.  1965
                                                                                               Matsubara et al. 1983

-------
                                                              Table 2.2 (continued)
 .  .   .          Species                   Dose, duration,                              _.,                                  _  ,
Animal               .                           ,                                      Effects                              Reference
                  strain                       and route

           Japanese white—       50, 150, or 300 mg/kg/day         External, skeletal, and internal                 Matsubara et al. 1983
           NIBS                  orally on gestation days             anomalies significantly different at all
                                  6 to 18                            dose levels (P<0.01); external anomalies
                                                                     of highest frequency were holoprosen-
                                                                     cephaly, anencephaly, hypoplasia of
                                                                     ala nasi, and club foot

   "625 mg/kg dose resulted in 55% abortion rate and thus a minimum number of malformations resulted.
   blntraperitoneally
   'Intravenously

-------
                                                  Table 2.3.  Aminopterin Action in Various Species
 Animal     Spedes/
              strain
                Dose, route,
               and duration
                         Effects
                                                              Reference
 Man
                         29 mg orally over 10 days at ~6.5
                         to 8 weeks of gestation
                         20 mg orally over several weeks during
                         second trimester of pregnancy

                         12 mg orally over 12 days during early
                         pregnancy
 Mouse    ICR-DUB     25 mg ip'on day 11 or 12 of gestation
 Rat
           Wistar and
           others
           Wistar

 Monkey   Rhesus


           Rhesus and
           cynomolgus
0.05, 0.075, 0.1, or 0.2 mg/kg single ip
injections on various gestation days

0.1 mg/kg  ip on gestation day 11 or 12

0.12 mg/kg orally on days 24, 25, and
26 of gestation

0.1-0.2 kg/day on gestation days  21 to 33;
1.0 kg/day on gestation days 38 to 39
                                             Excessively large head; nasal bridge broad and flattened;
                                             eyes widely separated; malformed ears; mandibular
                                             hyoplasia; posterior cleft palate; absence of parietal
                                             bones in skull
                                             Multiple skull anomalies; left talipes equinovarus
Numerous head abnormalities—soft skull, no ossification of
parietal bones, eyes wide apart, broad nasal bridge,
posterior cleft palate, and others; large hands, other
anomalies also described

Short or small limbs; hemimelia; severe ectrodactly
(absence of one or more fingers or toes)

Injections on days 7 and  10 effective—embryolethality


No malformed offspring—fetuses resorbed with high doses

No injury to offspring


Two aborted and  one normal fetus, but no malformed
offspring
Rabbit     Not given     2 mg/kg by injection to six-day-old blastocyst   No injury
Emerson 1962




Meltzer 1956


Warkany et al.  1959




Kochhar 1975


Baranov 1965


Murphy 1962

Tanimura 1972


Wilson 1969


Hay 1964
"Intraperitoneally.

-------
                                                 Table 2.4. Methotrexate Action in Various Species
 .  .   .      Species/
Animal           .  '
              strain
           Dose, route,
           and duration
                          Effects
      Reference
Man
Mouse     ICR


Rat

Monkey   Rhesus



           Rhesus
 Rabbit    New Zealand
           white
           New Zealand
           white
5 mg/day orally through second
month of pregnancy

2.5 mg/day orally for five days ~
between 8th and 10th week of preg-
nancy
0.3 to 50.0 mg/kg ip'on day 10 of
gestation

0.2 mg/kg on day 9 of gestation

2.5 to 4 mg/kg various regimes on
gestation days 17 to 45

3 mg/kg ivb on  gestation days 29 to
32

9.6 mg/kg iv on day 10 of gestation

19.2 mg/kg iv during days  11 to 14
of gestation
Major abnormalities of skull of infant


Major anomalies—absence of frontal bone, absent lambdoid
and coronal sutures, multiple anomalous ribs, unusual facies,
and absence of all digits on right foot and all but one on left
foot

25 and 50 mg/kg produced congenital defects, primarily cleft
palate and reduction of digits

103 embryos exposed—64% resorptions and 30% malformations

Most fetuses normal, three aborted and one showed moderate
gut rotation; only 13 embryos exposed, so trivial teratogenicity
might reflect small sample size
Mildly embryolethal, nonteratogenic, transitory growth retar-
dation

50% fetal mortality and a 25% malformation rate in surviving
fetuses
Multiple anomalies—facial clefts, cleft palates, defects of fore
and hind limbs, etc.
Powell and Ekert 1971


Milunsky et al. 1968




Skalko and Gold 1974


Wilson 1971b

Wilson 1971b


Wilson et al.  1979


Jordan et al.  1970

Jordan 1973
 "Intraperitoneally.
 blntravenously.

-------
                                  24

   An economical alternative would be to test the two species separately
instead  of concurrently and test the second  species only when the first
species has given negative  or ambiguous results.  In the interest of human
safety, the test chemical would have to be considered potentially terato-
genic to  humans if  one  animal  species  produces  malformed offspring.
However, the more animal species that demonstrate a chemical to be tera-
togenic, the greater the possibility that the chemical will  be  teratogenic in
humans, and by testing two  species data would be available  for species
comparisons. Such information as  the difference or  similarity in  dose  or
anatomical  defects would become available  and would  be  beneficial  in
other teratogenic studies with these same species. In addition, the use of
two species would give a better understanding of the effect/no effect level
of the test chemical.

    The use  of only two species (a rodent and rabbit) appears  to be justi-
fied in  view of the fact, stated by the WHO in  1967, that all substances
shown  to be  teratogenic in man  have  also  demonstrated  teratogenic
activity in the mouse,  rat, and rabbit. The WHO cautions, however, that
negative  results  obtained  by testing  chemicals in these species provide  no
absolute  assurance that the chemical will not induce teratogenic effects in
man. This  situation  would particularly  be of concern when the chemical
tested is  a drug intended for use during  human pregnancy or a chemical to
which heavy exposure is  likely for  pregnant women. In these instances,
when the rabbit or  rodent does not demonstrate  teratogenicity,  a third
species closer to man physiologically probably should also be used as a test
animal.  This is similar to the testing scheme proposed by Wilson (1975),
who recommends the use of the rat, mouse, hamster, or rabbit as screening
species to establish the embryotoxic dose range and not as an end point for
teratogenic  studies  unless appreciable embryotoxicity is demonstrated at
appropriate  multiples of the anticipated human dose.

2.2.5 Number of Test Animals Per Dose  Group
    The number  of pregnant animals per  dose group is an important con-
sideration.  As Weil  (1970) points out, for  a statistical analysis of the
results of an experiment  designed  to assess teratogenicity, the number of
independent sampling units, N, is the number of dams or litters. Using the
number of pups as N will produce invalid conclusions.
    The number of animals used to  test chemicals for teratogenicity is vari-
able. The results of some  of the studies  cited in  Table 2.1 might be ques-
tioned because too few animals in  some dose groups were used to permit
reliable statistical analysis. The WHO (1967),  although not specifying
animal  numbers, states that the number  of  rodents  used must be large
enough  to satisfy statistical requirements.  For species more closely related
to man, the WHO recommends that the number  of animals  be  as large as
practicable,  in order to obtain  reproducible results. Current guidelines
recommend 20 pregnant rats,  20 pregnant mice, 20 pregnant hamsters

-------
                                 25

and  12 pregnant rabbits  per  dose group (OECD  1981, USEPA 1982a,
1982b, USFDA 1982). The FDA  guidelines add that these are minimum
numbers of animals at or near  term.

2.2.6 Administration Route

   The choice of administration route is usually dictated by the present or
expected human exposure to the test chemical (Collins and Collins 1976).
In  most  instances, the  chemical should  be administered  orally,  but
administration  by other  methods  such as dermal  or  inhalation  may  be
necessary  in order to simulate  human  exposure. The EPA (USEPA 198la,
1982b), the FDA (USFDA 1982), and the OECD (OECD 1981) guide-
lines recommend oral administration  by gavage unless the chemical and
physical characteristics of the  test chemical or the use pattern would sug-
gest another route.

2.2.7 Fetal Examination

   Current  guidelines (OECD 1981,  USEPA  1982a,  1982b, USFDA
1982)  are in agreement concerning the extent of fetal examination.  All
fetuses should be examined externally, and one-third to one-half of all rat,
mouse,  and  hamster fetuses should be examined for skeletal  defects with
the remainder examined for soft tissue malformations. All rabbit fetuses
should be examined for both  skeletal  and soft tissue malformations. The
data in Table 2.1 indicate that, while many researchers are following this
protocol, some are not (e.g., Mehlman et al. 1984).

2.3 STRUCTURE-ACnvrTY  RELATIONSHIPS

   As Schardein (1983)  states "One of the  most potentially promising
measures  of teratologic risk is  the  structure-activity relation of chemicals."
Schuler et  al.  (1984)  in  subjecting  15 glycol  ethers  to   an  in vivo
reproductive toxicity assay found  that the 5 ethers with terminal methyl
groups and 2 with  terminal ethyl  groups  produced few viable CD-I mice
litters  whereas  3 butyl  ethers, 3 glycol  ethers  with  terminal  hydroxy
groups, and 2 ethyl ethers did not produce such profound fetotoxicity.
   The evidence generated by  Willhite et al. (1984) in their study of the
structure-activity relationships  of retinoids suggested  that the changes in
teratogenic activity associated  with structural modification of vitamin A at
carbon  15 were primarily dependent upon the presence of or  biotransfor-
mation  to a free carboxyl or a moiety with an equivalent pKa at carbon
15. The molecular size of the substitutent or the sterochemical position
about carbon 13 was considered of secondary importance for the induction
of terata.
   A series of papers (Keeler  1970, Brown 1978, Brown and Keeler  1978)
have discussed  the  teratogenic  structural requirements of naturally occur-
ring  steroids.   Keeler  (1970)  studied  jervine,   11-deoxojervine,  and
3-glucosyl-ll-deoxojervin, three compounds teratogenic to sheep  (causing

-------
                                  26

clyclopia) which structurally  differ only in substitution at the 3 and 11
positions.  Keeler tentatively concluded  that the nature  of the substitution
of these two positions has little influence  on  teratogenicity. Testing with
more than  18 additional  compounds  including rubijervine, veratramine,
and  testosterone in sheep did not produce  cyclopia, although veratramine
did cause occasional  bowing of limbs,  joint flexure, and lack of muscular
control. None of these compounds possessed the fused oxide and piperidine
rings present  in  the  active compounds (i.e., jervine, 11-deoxojervine, and
3-glucosyl-l 1-deoxojervine).
   Brown (1978) investigated  the teratogenic activity in the hamster of
several derivatives of jervine and 11-deoxojervine. His findings showed that
highly teratogenic compounds present  a negatively charged center accessi-
ble to the steroid a face. Similarly,  Brown and Keeler (1978) stated that
"conventional steroids with a  secondary basic nitrogen bonded in  a position
analogous to hormone binding sites (accessible to the a face of the steroid)
would be predicted to be substantially teratogenic."
    Freese et  al. (1979) studied the correlation between the growth inhibi-
tory effects, partition coefficients, and teratogenic effects of a number of
lipophilic acids and found many of the acids that  were potent inhibitors of
mammalian cell replication to be also teratogenic.
    Schumacher (1975) reported that  very strict  structural  requirements
govern the  teratogenic  effects of  thalidomide  (in particular  an  intact
phthalimide or  phthalimidine  moiety) and,  although  investigations  with
thalidomide have revealed a  number of interesting and rather unique pro-
perties of this chemical, no conclusive  evidence connecting any special pro-
perty of thalidomide with its  teratogenicity has been presented.

2.4 TERATOGENESIS AND TIME OF ADMINISTRATION

    A  chemical  should be  considered an unlikely teratogen  only after
proper teratological experimental procedures have been performed. One of
the most critical aspects of testing chemicals for teratogenicity is determin-
ing the most susceptible time during gestation  for chemical administration.

2.4.1 Preimplantation

    In  both rat and  man, there is  an interval of  approximately six  days
between  the  fertilization of the oocyte  and  the  implantation  in the
endometrium  known  as the preimplantation or  predifferentiation  stage
(Leonard 1983; also  see Table  2.5). In this largely undifferentiated  state,
the ovum is generally considered to not be  at  risk  from teratogenic agents.
With respect  to drugs Leonard (1983) states, "It  appears that  drugs are
either  toxic to the entire embryo, thereby  resulting in  its  death, or affect
only a relatively small number of cells, that can be repaired without caus-
ing any obvious  physical damage." However,  consideration must be  given
to the possibility of  chemicals  with  long half-lives  whose effects may be
manifested days after exposure. An  example  is  the progestin cyproterone

-------
                                   27
      Table 2.5 Preimplantation Developmental Stages in Man and Laboratory Animals
                        Figures as Postconceptional Days
Stage
Fertilzed eggs in the
fallopian tubes
2 cells in oviduct
4 cells in oviduct
8 to 12 cells in oviduct
Blastocele begins to
develop - oviduct
Free blastocyst in uterus
Implanting blastocyst
Implantation in progress
Post implantation
Delivery

Human

1
2
2 1/2
3

4
5
6
6 to 13
19 to 260
260 to
280
Monkey &
baboon

1
2
3
4

5 to 6
7 to 8
9
9 to 15
18 to 160
164 to
168
Rabbit

1/4
1/2
1
2

3
4
7
7to8
8 to 29
30 to
32
Rat

1
2
3
3 1/4

4
5
6
6 to 8
9 to 21
22

Mouse

1/2
1
2 1/4
21/2

3
4
4 1/2
4 1/2 to 6
7 1/2 to 19
20

    The following days are those of high teratogenic risk:  human, 19 to 80; monkey and
    baboon, 18 to 50; rabbit, 8 to 18; rat, 9 to 17; mouse, 7  1/2 to 15.

    Source: Adapted from Leone 1977, Fig. 1, p. 19.
acetate whose long biological half-life (60 h) was responsible for abnormal
morphological development  when  administered in high  doses on  day 2 of
gestation in the  mouse (Eibs et al.  1982).
    With respect to behavioral teratogens, limited  evidence  suggests that
exposure  during the  predifferentiation  period may  result  in behavioral
abnormalities.  Werboff  and  Havlena  (1962)  found  that tranquilizers
administered to the gravid rat induced behavioral changes in the offspring
independent of  the trimester of gestation during which these  drugs were
given.  In  this  study, the  tranquilizers  reserpine, chlorpromazine,  and
meprobamate (0.1, 6.0,  and 60.0 mg/kg/day subcutaneously, respectively)
were administered  on days 5 to 8, 1 to 14, or 17 to 20 of gestation with
each daily  dose divided into three equal  injections. Because day 8 (Leo-
nard 1982) or day 9 (see Table 2.5) in  the rat is considered the first day
of the  differentation phase of development, administration on days  5 to 8
would  be during the predifferentiation phase.

2.4.2 Organogenesis
    The  period  of embryological  differentiation (organogenesis) is con-
sidered to be the most sensitive developmental stage to  the action of tera-
togenic chemicals (Wilson 1975, Schardein 1976).  Schardein (1976)  lists

-------
                                  28

what he considers to be the critical periods for a number of animal species
(see Table 2.6); the period of greatest  susceptibility in man is similar to
that of the baboon and the rhesus monkey. As indicated in Table 2.1, dos-
ing periods  used  by  researchers vary.  Current regulatory  guidelines
(OECD   1982,  USEPA  1982a,  1982b,  USFDA   1982)   recommend
appropriate dosing periods for the mouse,  rat, hamster, and  rabbit.  These
guidelines are consistent with respect  to the mouse and rat (gestation days
6 to 15).  The USFDA guidelines are slightly different with respect  to the
hamster and rabbit; the  OECD and EPA recommend  dosing during  gesta-
tion days 6 to 14 for the hamster and 6 to 18 for the rabbit, whereas the
FDA recommends days 4 to 14 for the hamster and 7 to 18 for the rabbit.
    Although  Table  2.6 indicates  several days categorized as  in the critical
period, the embryologic differentiation of the  various organs  proceeds at
varying rates; thus,  the time of teratogenic insult is important with  regard
to  which  tissues are affected. This is illustrated in the research of King et
al. (1972), Kajii et  al. (1973),  Sadler  and Kochhar  (1975),  Inouye and
Murakami (1977), and  Eibs et al. (1982). Inouye and Murakami studied
 the teratogenicity in  mice of the hair dye constituent 2,5-diaminotoluene
 and  found that  subcutaneous or  intraperitoneal injections  of 50  mg/kg
 body weight  caused a low incidence of exencephaly (cranial malformation)
 and prosoposchisis (fissure of the face, e.g., hairlip) and a high incidence
 of skeletal malformations in  those animals treated on day 8 of gestation.
 No such malformed fetuses, however, were found in those animals treated
 on days 10 to 14 of pregnancy,  and  only  a very low incidence of vertebral
 and rib anomalies was  observed in the fetuses treated on gestation day 7
or 9.
    In studying the teratogenic  effects  of a single oral injection (intuba-
tion) of chlorambucil (14.2 or 20 mg/kg) on the 10th, llth, 12th, or 13th
day,  Sadler and Kochhar (1975) observed limb defects from treatments on
the llth  or 12th days of gestation and tail defects from treatment on all
days. Treatment on the 13th day resulted  in digital defects but without the
long-bone defects observed with day 11 and  12 treatments; defects were
observed  only at  the highest dose level.  Sadler  and  Kochhar also found
that  in vitro  limb  bud  response to chlorambucil was similar  whether the
limb  buds were taken from pregnant mice after chlorambucil treatment  or
the chlorambucil was applied to  limb bud  cultures after removal from non-
treated pregnant mice.
    Eibs et al. (1982) administered cyproterone acetate (30  mg/kg  subcu-
taneously) to mice on one of days 1  to  12 of gestation. The most sensitive
gestation  days for abnormalities  were days 5 and 6 for the urinary tract,
days 8 and 9  for the respiratory tract, and  days  10 and 11 for cleft palate.
    The significance of the critical period is also shown by the human tera-
togen  thalidomide. One  study identified  113  pregnant  women  who took
thalidomide between August 1959 and  December 1961, but only 7 women
took the  drug during the critical period (reported here to be between 34
and 50 days after the last menstrual period) (Kajii et al.  1973).  Of these

-------
                          29
  Table 2.6. Critical Periods of Organogenesis in Animals
Species
Hamster, golden
Mouse
Rat
Rabbit
Ferret
Cat
Dog
Guinea pig
Pig
Sheep
Monkey, rhesus
Monkey, baboon
Armadillo
Human
Cow
Horse
Mean duration
of gestation
(days)
16
19
21
31
43
63
63
68
114
150
168
175
225
278
284
336
Gestation
Days"
4 to 14
7 to 16
9 to 17
8 to 21
8 to 28
5 to 58; 5 to 15
most favorable
1 to 48; 8 to 20
estimated
1 1 to 20
12 to 34
14 to 36
20 to 45; 22 to 30
most susceptible
22 to 47
1 to 30
20 to 55
8 to 25
?
aP*»ri/t/1 f\f r*mhr\jn\ri{rir*ti] nro5mnoenp.sic fir nftrind nf
known susceptibility.
Source: Adapted from  Schardein 1976, Table 2-5, p.  17.
Data collected from several sources. Reprinted with per-
mission of the publisher.

-------
                                  30

seven women, three delivered malformed babies, and four delivered babies
without apparent malformations. Possible explanations for the four normal
babies include inaccurate pregnancy diagnosis time, so that thalidomide
possibly was not taken during the critical period,  or  actual resistance to
thalidomide.  Two additional  women  took the drug before the critical
period, on  day 19 and day 23, respectively; both women aborted (induced
and spontaneous abortions,  respectively). The remaining  104 women took
thalidomide after the critical  period with all offspring apparently normal,
except for one reported case of anal stenosis.
    One other example of the  importance of the critical period is shown in
rats administered |8-aminopropionitrile, a chemical that induces cleft palate
(King et  al.  1972). In this instance,  there were  a significant  number of
offspring  with cleft palate only when day 15  of the gestation  period was
included in the dosing schedule. All of the viable  fetuses had cleft  palate
when 875  mg/kg was given on day 15, or over days  13 to 15 or 14 to 15,
but when a total dose of 4320 mg/kg  was administered over days 12 to 14,
only 8% of the young were malformed.

 2.4.3 Histogenesis and Fetal Period

    The period  of  prenatal  development that slightly  overlaps organo-
 genesis but extends primarily into the fetal period is known as histogenesis
 (Fig. 2.1). Teratogenic agents that come into contact with the developing
 fetus during this time  of tissue  formation  and  development can cause
 minor structural deviations, but the abnormalities that are more likely to
 occur during  the fetal  period are those involving  growth or functional
 aspects of development (Wilson  1975b). An example is the administration
 of vitamin A during the early- and  mid-fetal period  (Hutchings et al.
 1973, Hutchings and Gaston  1974). Pregnant rats administered a terato-
 genic dose of vitamin A  during  the early  fetal period (days 14 and  15 of
gestation) produced offspring that showed a behavioral  deficit indicating
nonreinforcement (characterized by a decreased ability to inhibit respond-
ing to an auditory signal). In  addition to behavioral  effects, a  generalized
retardation in growth was induced, as evidenced  by delayed onset  of fur
growth and  eye-opening and  by reduced body weight as well  as reduced
brain size  with  obvious microcephaly in one animal. Pregnant rats treated
with vitamin A during the mid-fetal period (days 17 and 18 of gestation)
produced offspring with no retardation in growth or in brain size, but  with
a  possible motor deficit affecting coordination that resulted in slower rates
of response to auditory stimuli  than control animals.

2.5 BEHAVIORAL TERATOGENESIS

    As described  by Hutchings  (1983), behavioral  teratology [Vorhees
(1983) uses the synonym psychoteratology] is an integration of teratology
with experimental psychology.  The primary concern of behavioral teratol-
ogy is the study of  neurobehavioral changes that occur when germ cells,

-------
                                  31
                                               ORNL-DWG 79-13670

          TERATOGENIC  SUSCEPTIBILITY  IS  GREATEST
                  DURING EARLY ORGANOGENESIS
                                       I-FUNCTIONAL MATURATION-
              EMBRYONIC PERIOD         FETAL  PERIOD
                    ENTIRE DEVELOPMENTAL SPAN

   Figure 2.1. Representation of the susceptibility of the human embryo to
teratogenesis, beginning with fertilization and  continuing throughout intra-
uterine  development. (Source: Adapted from  Wilson  1975b,  Figure  4, p.
662).
embryos,  fetuses,  and  immature  postnatal organisms  are  exposed to a
variety  of environmental disturbances  and events. Regarding the state of
the  art  of  behavioral  teratology, Hutchings (1983)  writes,  "I  should
emphasize, however, the current state of behavioral teratology as a fled-
gling discipline;  its  subject  matter   remains  largely unexplored  and
uncharted, its methodology has been uneven and occasionally wanting, and
its accomplishments strike a tenuous balance between false starts and pal-
pable inroads made by a few pioneering researchers. But despite its  imma-
turity and tentative beginnings, there should be no doubt that it is emerg-
ing as a new scientific speciality."
   Although the  period of organogenesis is generally considered the time
the embryo  is  at greatest  risk to morphological alteration, the  exposure
period is not as well defined with respect to the induction of behavioral
changes. As discussed in  Section  2.4.3, the  histogenesis period can be a
sensitive period for  behavioral teratogens. However, as Coyle et al.  (1976)
reported, behavioral anomalies can also be induced in  the offspring as a
result of exposure during organogenesis.  They cite  Hoffeld and Webster
(1965)  and  Murai (1966),  who  found  that when  chlorpromazine  was
administered  during early pregnancy maze learning was impaired,  but no
effects were  observable when it was applied late in pregnancy. In addition,
Werboff and  Havlena  (1962)  demonstrated that behavioral changes
induced by the administration of tranquilizing drugs  are independent of
the trimester  of  gestation.  Compounding the situation even  further  are
studies by Gauron  and Rowley (1969)  and  Friedler (1978) which show

-------
                                  32

that the behavior of offspring can be affected even by pregestational expo-
sure to certain chemicals.
    An  investigation  of the  holdings  of the  Environmental  Teratology
Information  Center  (ETIC)  of the  Oak Ridge  National  Laboratory
indicates  that  of the over  25,000 citations indexed  concerning various
aspects  of teratogenicity testing,  2645 are related to behavioral teratology.
By species this breaks down as follows: rat (1078), mouse (297), hamster
(11), rabbit (69), monkey (38), and human (1152).

2.5.1 Historical Perspective

    Werboff and  his coworkers are credited with establishing the branch of
teratology known as behavioral  teratology.  They published several papers
(e.g., Werboff et al. 1961a,  1961b;  Werboff and  Havlena 1962; Werboff
and Kesner 1963) detailing the postnatal effects of prenatally administered
psychotropic  drugs.  As indicated  by  Leonard  (1983)  and Hutchings
(1983), it was not until the early 1970s that the work of Werboff and col-
leagues was extended. The realization that prenatal exposure can cause
behavioral  alterations has resulted in  both Japan and  Britain mandating
that  behavioral   evaluations  be  conducted  on  all  new  drugs  during
premarket reproductive testing (Vorhees and Butcher 1982, as reported in
Vorhees  1983). Although similar regulations have not been established in
the United States, Vorhees  (1983) cites several events which suggest that
U.S. regulatory authorities are giving  serious consideration to  establishing
guidelines for behavioral teratogenicity testing.

2.5.2 Behavioral Testing Methodologies

    Because no single test gives  a comprehensive behavioral  assessment, a
multitesting procedure is usually employed. Jensh (1983) in  his review  on
behavioral  testing procedures cites  several authors (Spyker  and  Avery
1976, Kimmel  1977,  Rodier  1978,  Silverman  1978,  Buelke-Sam and
Kimmel 1979, Lasagna  1979, Vorhees et al. 1979) who have suggested cri-
teria for an effective multitest procedure. These criteria are listed by Jensh
(1983):
  1.  The tests should be simple.

  2.  The tests should be comprehensive, examining a variety  of behavioral
     functions (global).

  3.  The  tests  should be sensitive to slight alterations  in the organism's
     functions,  as well as to alterations produced by a number  of different
     agents.

  4.  The behavioral  schedule should be simple, to minimize the training
     time of personnel.

  5.  The tests  should take a minimum amount of time.

-------
                                  33
 6.  The tests  should be  reliable,  reproducible,  and  valid in  the  same
    laboratory as well as in the laboratories of other investigators.

 7.  The tests should have a proven history of positive results.

 8.  The tests should be economical in terms of financial expense and utili-
    zation of animals.

 9.  The tests should be within the framework of existing guidelines.

10.  The protocol  should  move in  graded steps from elementary to  more
    sophisticated tests.

11.  Data should be quantifiable.

12.  The tests should be predictive of effects in humans.

   Table 2.7 lists what Jensh (1983) considers the more commonly used
tests either singly  or  as part of a behavioral  test  battery.  Earlier,  Buelke-
Sam and Kimmel  (1979) reported  on a survey they conducted concerning
behavioral methods in teratologic evaluation used by those researchers who
responded to their survey  (Table 2.8).  As both Tables 2.6 and 2.7 indicate,
a variety of  behavioral test methods  are being used to assess  behavioral
alteration.  This variety of  test methods as  well  as  the varying exposure
regimens  and  postnatal environmental conductions  used  by  researchers
emphasizes  the  need for  standardization  of  screening  methods  for
behavioral teratology (Buelke-Sam  and Kimmel 1979). Jensh (1983) states
that "standardization of test procedures is of paramount importance" and
adds that "the  investigator needs  to  have access to a standardized pro-
cedure for each of the chosen  tests, including experimental protocols, con-
struction designs,  techniques, past results, and interpretative implications
and limitations." Standardization procedures  are in fact under way (Kim-
mel et al. 1982, 1983; Kimmel 1984), and a workshop to  present the data
from a  six-laboratory  study was  held  on September  3-6,  1985, in Cincin-
nati, Ohio.

2.6 SPECIES COMPARISONS
   Proper selection  of a test  species is important. The large  number of
animals  necessary for toxicological research usually  dictates that  the  test
species  be easily available and economical. The following subsections  will
consider some  general aspects of species selection; will briefly  discuss the
advantages and disadvantages  of the rat, mouse, hamster, rabbit, monkey,
and other less  commonly  used species as models for  extrapolation to man;
and    will   compare   the   response   of   three    known   human
teratogens—aminopterin,    methotrexate,   and   thalidomide—with   the
response in test animals.

-------
                                    34
                  Table 2.7. Behavioral Testing Procedures'
                          I. Physical Development
Pinna detachment                            Testes descent
Incisor eruption                              Vaginal opening
Eye opening                                 Development of fur
Ear unfolding and opening
                           II.  Behavioral Analyses
A. Reflex development tests'*
   Crossed extensor                         Tail pinch
   Flexor dominance                        Tremors
   Roofing reflex                            Surface righting
   Limb withdrawal                         Air righting
   Forelimb  placing                         Auditory  startle
   Hindlimb placing                        Negative  geotaxis
   Forelimb and hindlimb grasp              Visual placing
   Suckling ability                          Vibrissae placing
   Tail hang                               Cliff avoidance
   Vibrissae stroking                        Pain reflex reaction
   Ear twitch                               Olfactory reflex
 B. Motor  and coordination tests0
    Bar holding
    Forelimb hanging
    Visual orientation
    Inclined plane
    Rotarods
    Parallel bars
 C. Spontaneous or nonforced behavior tests'1
    Spontaneous motor behavior
    Head elevation
    Head pointing
    Hindlimb elevation
    Rearing
    Pivoting
    Gait
    Crawl
    Walk
    Backward walking
    Grooming
    Circling behavior
    Sleeping and somnolence
D. Forced behavior tests'
    1. Mazes
       T-maze
       Y-maze
       Hebb-Williams  maze
       Lashley  III maze
       Olton maze
Ascending wire mesh
Ascending a vertical rod
Clinging to and descending
   a vertical rope
Swimming
 Hopping
 Homecage emergence
 Resistance to handling
 Feeding
 Undisturbed behavior
 Spontaneous and general activity
 Nose poke
 Head dip
 Step-down test
 Thirst test
 Maternal  behavior
Open-field
Activity wheel


2.  Operant conditioning tests
   Bar-(lever-) press
   Avoidance testing
   Discrimination reversal
     learning
"Data compiled from Jensh 1983.
bSome of these include complex motor activity.
'Primarily include forced, complex activities.
dMay include some motor evaluations.
'Primarily include learning skills.

-------
                             35
         Table 2.8.  Behaviors Tested and Methods Used
                   in Teratologic Evaluation"
Observation (23 )b
18C
3
2
1

Reflex
17
15
11
10
6
6
6
5
4
3
3
2
1

Motor
18
10
8
7
5
4
3
2
2

Sexual
12
general activity and condition
modified Irwin method
photography /video taping
residential unit

development (34)
surface righting
auditor startle
pupillary contraction
air righting
cliff avoidance
touch-escape
visual placing
Fox method
nonspecified, animal
negative geotaxis
Smart & Dobbing method
nonspecified, human
stabilimeter

coordination (35)
rotating rod/drum
swimming
hanging/grasping
gait
inclined plane
traverse rods
pivoting
ambulation times
nonspecified, human

behavior (15)
mating success
Activity (39)
23 open field
10 activity wheels
6 photoactomers
3 behavioral sequencing
2 hole board
2 inductance monitors
2 nonspecified human
2 stabilimeter
1 automated plus-maze
1 exploratory
1 home cage
1 motron
1 residential unit
7 some monitoring of activity
for ^ 24-h period

Aggression (6)
3 shock-induced
1 muricide
1 nonspecified, human
1 tail-rattling

Classical conditioning (2)
1 nonspecified, human
1 tone-shock pairing

Avoidance (26)
12 passive
12 2-way active
7 1-way active
2 nonspecified CAR
2 Sidman
2 Y-maze
5  male performance
2  female receptivity signs

-------
                               36
                      Table 2.8 (continued)
Operant behavior (17)
   13  simple schedules
      9   FR
      7   DRL
      4   CRF
      4   VI
      3   FI
      2   punishment
    9  discrimination procedures
    6  acquisition/extinction
    4  multiple  schedules
    3  progressive/titrating schedules
    2  various nonspecified types
    1  discrete trial
 Maze/instrumental behavior (30)
   15 water mazes
    9 discrimination procedures
    4 T-maze
    3 neonatal home-cage seeking maze
    3 nonspecified types
    3 straight alley
    3 Y-maze
    2 Hebb-Williams
    2 Lashley III
    7 reversal learning
    2 delayed responding
Sensory function (29)
   4  thresholds
      3 auditory
      2 taste/olfactory
      1 shock
    3  photophobia
   22  assessed as part of
      reflex development
    9  assessed as part of a
      discrimination paradigm

Pharmacologic/environmental
challenge (12)
    3 amphetamine
    2 barbiturate sleep time
    1  alcohol preference
    1  audiogenic seizure
    1  drug discrimination
    1  LD50
    1  metrazol seizure
    1  morphine analgesia
    1  neonatal CNS lesion
    1 neonatal 6-OH-dopamine
    1 nonspecified
    1 shock stress
 Social/group behavior (4)
    2 competitive situation
    1 group homing response (4)
    1 nonspecified
    1 residential unit
 Other (7)
    4 spontaneous alteration
    3 consummary patterns
    1 psychological evaluation,
      children
"Tabulated  summary of  methods used  by those researchers  who
 responded to the questionnaire.
bNumber of researchers who  reported using one or more methods
 to evaluate behavior falling into each major category.
°Number of researchers who reported using the specific method or
 monitoring the specific behavior.

 Source:  Buelke-Sam  and   Kimmel  (1979),  Table   2,  p.  24.
 Reprinted with permission of the publisher.

-------
                                  37

2.6.1 General Aspects

   No ideal species has been identified with regard to extrapolating exper-
imental results  from  teratogenic tests  in  animals  to man  (NAS  1977;
Schardein 1976, 1983; Palmer 1978). That is, there is no one species that
fits all  the criteria for an  ideal test animal as listed by Wilson  (1975):
(1) absorbs,  metabolizes,  and  eliminates  test substances as  does man,
(2) transmits  test substances  and their metabolites  across the placenta as
does man, (3) has embryos and fetuses with developmental and metabolic
patterns similar to those of man, (4) is easily bred and has large litters
and a short gestation period, (5) is inexpensively maintained under labora-
tory conditions,  and (6) does not bite,  scratch, kick, howl, or squeal. The
problem of  choosing the most  appropriate species is further complicated
when the test  chemical has  no record of human administration from which
to draw a comparison. Thus, with little scientific data available, the choice
of species is often based on availability, economic feasibility, and ease of
management (Brown  1963, as reported in Palmer 1978).
   Kalter  (1968)  writes that  there are three different types of inter-
specific and intraspecific susceptibility: (1) an agent teratogenic in some
species  or groups of animals  may be completely  or almost  completely
without teratogenic effect in  others, (2) a  teratogen may produce similar
defects  in various species, stocks, or strains of animals  but  with  varying
susceptibility (frequency) between one  strain and another or between one
species  and another,  and (3)  a  teratogen may induce one or more abnor-
malities in some species or groups but have entirely different or only some-
what different effects in others. Kalter indicates that these are not mutu-
ally  exclusive categories and certain  situations may contain features  of
more than  one   category.  These  differences  are  illustrated in  Tables
2.9-2.11 and also  in the following sections.
   The rat followed  by the mouse and the rabbit are the most commonly
used laboratory animals for the teratogenic  screening of chemical agents
(World Health  Organization  1967, Palmer 1978).  Further substantiation
of this  fact comes by  examination  of  the  holdings of ETIC, whose files
contain over 25,000 papers indexed on various aspects of teratogenic test-
ing.  Of these, approximately  8350 cite the rat as a test  animal, 4640 cite
the mouse, 600, the hamster,  1700, the rabbit, and 460, the nonhuman pri-
mate. In addition, nearly 9550 papers discuss teratogens and  teratogenicity
in relation to humans.

2.6.2 Rat
   Advantages of using  the  rat as a  test animal  for teratogenic testing
include (1) short  duration of  pregnancy (21 to 22 days), (2) high fertility
rate, (3) large litters and a relatively good resistance to the toxic effects of
most drugs, (4) a fairly good developmental stability, and (5) a low spon-
taneous rate  of  major  malformations (1  per 1000  fetuses,  0.001%)
(Tuchmann-Duplessis 1972, Palmer 1978).  The rat is also easily  handled

-------
                                 38
               Table 2.9.  Species Susceptibility to Drugs
Drug
Azathioprine
Cortisone
Demecolcine
Ethyndiol
Meclizine
Methylprednisolone
Perphenazine
Dosage
(mg/kg)
and route
10POa
5 or 10 IMb
1.5 IPcorSCd
1 PO or SC
125 PO
1 IM or SC
25 PO
Malformed species
(percent incidence)
Mouse
0
79
23
30
0
92
100
Rat
0
0
0
0
92
0
98
Rabbit
57
50
3
0
0
0
0
Others


0 (monkey)




aPO—per os (oral).
bIM—intramuscular.
°IP—intraperitoneal.
dSC—subcutaneous.
Source: Adapted from Schardein,  1976,  Table 2-8, p. 21.  Data collected
from several sources. Reprinted with permission of the publisher.
and economically maintained, and test materials can be administered by a
wide variety of routes (Palmer 1978). The main limitation of the rat is its
poor teratogenic susceptibility  to drugs such as cortisone (Thompson and
Schweisthal   1969,   Hansson  and   Angervall   1966),   azathioprine
(Tuchmann-Duplessis and  Mercier-Parot  1966), and thalidomide (Somers
1963). Thus, definite conclusions  based only on tests with  rats could be
misleading (Tuchmann-Duplessis 1972).

2.6.3 Mouse
    The advantages listed for the use of the rat in testing (Section  2.6.2)
also apply to the mouse, with  the  addition that the mouse  is even  more
economical, has a shorter gestation period (18 to  19 days), has a wide
variety of defined inbred strains for special studies, and is more susceptible
to some  teratogens (Tuchmann-Duplessis 1972, Palmer  1978).  Disadvan-
tages include: (1) small  size  of the fetus and consequent difficulty in
examination of malformations; (2) arrangement of  the  malformations in
clusters,  which creates difficulty  in assessment; (3) higher spontaneous
malformation  rate  than the  rat  (0.5%  in  the Swiss  albino  colony of
Tuchmann-Duplessis);  and (4)  resorption rates  much  higher than in the
rat, necessitating a large control group (Tuchmann-Duplessis  1972, Palmer

-------
                                   39
               Table 2.10. Thalidomide Teratogenesis in Primates
Species
Man
Cynomolgous
monkey
Baboon
Rhesus monkey
Bushbaby
Japanese
monkey
Stump-tailed
monkey
Marmoset
Bonnet monkey
Teratogenic
dose
(mg/kg,
oral route)
0.5 to 1.0
10
5
12 to 19
20
20
5 to 10
45
5 to 30
Gestation
days
treated
20 to 36
22 to 32
18 to 44
24 to 26,
27, or 30
16 to 30
24 to 26
24 to 30
25 to 35
24 to 29 or
41 to 44
Defect
Limbs (80%), ear (20%)
Limbs (67%), teratomas (33%)
Limbs and tail (40%)
Limbs (100%)

Limbs (100%), tail (17%),
central nervous system (17%)
Limbs (100%), tail (20%)
Limbs, ear, and jaw (100%)
Limbs (45%), visceral (52%)
  Source: Adapted from Schardein,  1976, Table 2-11, p. 25. Data collected from
  several sources. Reprinted with permission of the publisher.
 1978). The mouse is  also known to respond  differently than  the  rat to
 some drugs such as cortisone.
   With regard  to the human  teratogen  thalidomide  (see  also Section
 2.6.8.3),  mice embryos are relatively insensitive, as is evident in  a study by
 Somers (1963) in  which oral  administration of doses up to 400 mg/kg
 throughout pregnancy did not reduce the number of newborn mice or their
 ability to survive to weaning. Even when the doses were increased to 4000
 mg/kg only a small increase in resorption sites was observed.

 2.6.4 Rabbit

   The rabbit was the first laboratory  animal shown to be susceptible to
 the human teratogen thalidomide (Somers  1962)  and has  since been con-
 sidered by many biologists to be one of the most  favorable animals for
 teratogenic studies (Tuchmann-Duplessis  1972). The  susceptibility  of the
 rabbit to thalidomide was  also reported by Somers  (1963), who produced
limb defects in the offspring by administering 150 mg/kg to pregnant rab-
bits on days  8 to 16 of gestation, and by Shepard (1976), who  states that
limb defects in New Zealand rabbits were observed after administration of
 250 mg/kg on days 8 to 10 of pregnancy.

-------
                                  40
              Table 2.11. Comparative Teratogenicity of Thalidomide
Organ system
affected
Limbs
Ears
Eyes
Cardiovascular
Urinary
Genital
Gastrointestinal
Pulmonary
Prenatal and/or
postnatal death
Human
X
X
X
X
X
X
X
X
X
Nonhuman
primate
X
X
X
X
X
X
X

X
Rabbit
X


X
X
X
X

X
Rat Mouse
X (low%) X (low%)


X




X X
      Source: Hendrickx et al.  1983, Table II, p. 153. Reprinted with permission of
      the publisher.
    Advantages of using the rabbit include  the  larger size  of  the  rabbit
 fetus compared to the rat or mouse (Palmer 1978); the ability of the rab-
 bit to show a variety of spontaneous  malformations, which suggests that
 this species would be more susceptible  than some other species to a variety
 of teratogenic actions (Palmer 1968, as reported in Palmer 1978); and, for
 comparison purposes, the rabbit's physiological difference  from the rat and
 the mouse (Tuchmann-Duplessis 1972). The  necessity of having large con-
 trol groups, a higher spontaneous  malformation rate  (1.7%  for a  5-year
 period) (Tuchmann-Duplessis 1972), and the rabbit's dependence on gut
 flora for  nutrition, which  prevents  the  testing of antibiotics and  makes
 incorporation into the diet  a hazardous and inaccurate  means of adminis-
 tering test compounds (Palmer 1978), are some disadvantages of using the
 rabbit as  a model for testing compounds for teratogenicity. More practical
 disadvantages are problems associated with handling and intubation.

 2.6.5 Hamster
   Palmer (1978) reports that the hamster is mainly a  competitor for the
 mouse in  animal studies;  it is  almost  as economical to  maintain and  is
 reputed to be more stable genetically. The hamster has several advantages
 that make it  a potentially valuable animal for developmental and terato-
 genic studies: unique  reproductive features such as large litter size and
 short gestation period (16  days)  (Perm  1967). Palmer (1978), however,
lists  several disadvantages of using the hamster: (1) aggressive behavior
generally necessitates individual caging; (2) achieving precisely timed mat-
ing,  although  not difficult,  is not  as  convenient as  with  the mouse;

-------
                                  41

(3) intravenous injection is more difficult than in the mouse [Perm (1967)
does demonstrate a  method requiring anesthesia whereby the  injection is
made utilizing the lingual vein];  and (4) as with the rabbit, the hamster's
gut flora  renders it susceptible to change in the diet  and to  antibiotics.
Also, the hamster  has a poor historical data base. Tuchmann-Duplessis
(1972) concludes that from  available data the hamster  does  not  have  a
definite advantage as a test animal in the teratogenic screening of drugs.

2.6.6 Nonhuman Primates

    Much has been  written concerning the use of nonhuman primates as
animal models in the teratological testing of chemicals (Wilson  et al. 1968;
Wilson  1969, 1971a,  1971b, 1978; Hendrickx 1972; Poswillo et al. 1972;
Tanimura 1972; Siddall 1978). The rhesus monkey  (Macaco  mulatto) is
the  most popular  nonhuman  primate test  species  according to  Wilson
(1978), but teratological studies have been performed using other nonhu-
man  primates such as the cotton-eared marmoset  (Poswillo et  al. 1972;
Siddall  1978) and the  baboon (Hendrickx 1972).
    One  of the  primary advantages of using nonhuman primates in the
teratological screening of chemicals is their similarity in  many respects to
man. Wilson (1978) reports that: (1) in some instances the metabolism of
drugs in  man approximates that of other higher primates; (2) the repro-
ductive physiology  of  Old  World monkeys (rhesus monkeys and baboons)
closely resembles that of man, particularly the menstrual cycle, spermato-
genesis, and parturition; (3) the  placental structure of man and nonhuman
primates  appears to be quite similar, although the similarities of function
during the  critical  periods of  gestation have yet to be  fully  established;
(4)  the  anatomical  and  temporal  aspects  of  embryonic development
between the higher nonhuman primates and man  are  strikingly  similar;
and (5) similarity in response to thalidomide is well established [(the com-
parative response to  other agents, however, has  been variable  (Section
2.6.8)].
    Similarly, Leone (1977)  states in an excellent review of the develop-
mental  stages in mammals used for teratogenetic  tests  that  "Significant
differences are detectable between rodents and primates in the  early stages
of development, which especially affect the structure and functions of the
embryonic membranes and consequently influence the teratogenic process."
In his opinion, the morphological and comparative aspects of development
deserve as much attention as  do other parameters  (i.e., metabolic  path-
ways, dose threshold, etc.).
    In a discussion favoring greater diversity in test species used for terato-
genicity testing, Wilson (1975a)  states that rodents and rabbit embryos are
dependent during the  initial days of organogenesis on the inverted yolk-sac
placenta,  which is structurally and functionally different from the chorioal-
lantoic placenta of other mammals (e.g., primates)  at corresponding  times
in development.

-------
                                   42

   Wilson (1978)  cites scarcity; difficulties in handling, especially the
unpredictability  of behavior and general susceptibility to infections; and
low fecundity as disadvantages to using nonhuman  primates for teratologi-
cal testing of drugs. Two other disadvantages of nonhuman  primates are
the expense  (high  cost usually necessitates  a test group  size that is not
adequate) and the long gestation period.
   It should be mentioned  that  prosimians, and to a certain extent New
World monkeys such  as marmosets, have not been used extensively for
teratological evaluations  of chemicals.  The greater bush baby (Galago
crassicaudata),  a prosimian, was found by  Hendrickx (1972)  to  give a
negative response to thalidomide and  has since  received little  attention.
The  marmoset  has  shown  similar responses  as man  to  thalidomide
(Poswillo et al.  1972, Siddall 1978), but  the fact that marmosets  do not
menstruate causes  difficulty in following the reproductive cycle and possi-
bly accounts for their sparse use (Wilson 1978). Because of the advantages
afforded  by  their  smaller size, however,   Wilson (1978)  recommends
further testing of the suitability of New World monkeys and prosimians  as
animal models in teratology.

2.6.7 Other Species

    Other animals such as dogs, cats, and  swine have been used for terato-
logical testing but not to the extent of the previously mentioned species.

2.6.7.1 Dog

    Palmer (1978)  writes that the primary advantage of the dog is its con-
current use as a nonrodent  species for other  toxicity tests that may provide
further information of the  test material such as pharmacokinetics.  Earl et
al. (1973) tested the beagle with the known human teratogens thalidomide
and aminopterin as well as with methyl  mercuric chloride, hydroxyzine,
and hydroxyurea. They found that the dog  (1) responds to known  human
teratogens, but not in a classical manner, (2) does not appear to be as sen-
sitive an  indicator of  the above-mentioned compounds as  some primates,
and (3)  offers little advantage over other  laboratory animals. In addition,
the frequency of estrus (twice a year), a high cost of maintenance, and a
poor historical data base are also disadvantages.

2.6.7.2 Cat

   The cat has  been used infrequently in  teratology studies for such rea-
sons as the difficulty in exact determination  of the time of ovulation and
the time required  to establish a stable colony (Palmer  1968).  In addition,
the cat has a poor historical data base from which to make  comparisons.
Khera (1975) did  show  that thalidomide (10 to  480 mg/kg  per day)
induced a wide variety of fetal cardiovascular defects including ventricular
septal  defect, malpositioned great vessels, and narrowed left ventricular
chamber with hypertrophied walls.

-------
                                  43

   Other research by  Khera  and coworkers (Khera  and Iverson  1978,
Iverson et al. 1980), however, demonstrated that the cat is not as suscepti-
ble to  the  teratogenic effects  of some chemicals as  are  other  animal
species. Khera and Iverson (1978) administered ethylenethiourea (a degra-
dation  product  of the ethylenebisdithiocarbamate group  of fungicides)
orally in single daily doses of 0, 5, 10, 30, 60, or 120 mg/kg to 14, 9, 7, 8,
7, and  10 female  cats,  respectively, beginning on day 16 of gestation and
continuing  through pregnancy unless  maternal toxicity was  exhibited.
Although ethylenethiourea proved to be toxic to pregnant cats (several cats
died or were killed in a moribund condition in the 30-, 60- and 120-mg/kg
dose groups) and  had caused malformations  in  some of the dead fetuses
taken from  those  cats killed in the moribund state, no  clear evidence for
teratogenicity of ethylenethiourea in live fetuses was seen. This result is in
contrast to  the ethylenethiourea-induced teratogenesis  observed  in rats
(Khera 1973; Khera and  Tryphonas 1977, as reported by Khera and Iver-
son 1978). In a  follow-up study, Iverson et al. (1980) suggest that the rea-
son for the  absence of teratogenic effects in the  cat was the ability of the
cat to extensively metabolize ethylenethiourea to its S-methyl derivative;
examination of rat urine showed no evidence of the S-methyl derivative.

2.6.7.3 Pig
   Tuchmann-Duplessis (1972) and Palmer (1978) list  several advantages
of using the pig: (1)  it is easily  available and highly prolific compared to
other nonrodent species,  (2) its embryology and genetics are well known,
(3) pregnancy can be obtained year round, and (4) incorporation of the
test material into  the diet is  easily and accurately accomplished because of
the pig's voracious eating habits. These  same  authors cite disadvantages of
using the pig as the large amounts of the test material required due to the
size  of the  pig  and the excessive floor  space required.  Earl et al. (1973)
observed malformed offspring in  miniature swine that had been exposed to
thalidomide, hydroxyurea, and aminopterin  and conclude  that, although
the reaction to thalidomide was not the classical response, miniature  swine
offer certain advantages  such as low  incidence of spontaneous malforma-
tions. These researchers recognize that more data are necessary before the
role  of the pig in teratogenic studies can be determined.

2.6.8 Results  of Testing  Some Human Teratogenic  Chemicals  in Animal
Models

   Three known human  teratogens, aminopterin, methotrexate,  and thali-
domide, will be  used  to compare the teratogenic response between  humans
and  several  animal species;  the  data  are summarized  in Tables 2.2,  2.3,
and 2.4.

-------
                                  44
2.6.8.1 Aminopterin
   In addition to man, the folic acid antagonist aminopterin (Table 2.3)
induced  malformations  in  the  mouse  but  not  in  the  rat,  where
embryolethality  and resorptions were observed,  or  in the rhesus (Macaco
mulatto) and  cynomolgus (Macaco irus) monkeys.  The one reference con-
cerning aminopterin toxicity in the rabbit reports no injury to a six-day-old
blastocyst,  but the assumption  from these data that no injury would result
to the fetus probably would not be reliable.

2.6.8.2 Methotrexate

   Methotrexate, also a folic  acid  antagonist and a methyl derivative of
aminopterin,  is  teratogenic  to humans,  mice,  rats,  and  rabbits  but
apparently not to monkeys, although abortions were induced (Table 2.4).
The data in Table  2.4 indicate that (1) teratogenic defects in the man,
mouse,  rat, and rabbit are somewhat similar, especially with regard to rib
and  limb abnormalities; (2) the rat shows a teratogenic response to metho-
trexate but not to  aminopterin (Table 2.3); and (3) the monkey seems to
be resistant to the teratogenic action of both methotrexate and aminopte-
rin (Table 2.3). Skalko and Gold (1974) point out, however, that no tera-
togenic effects  are produced  in mice  at  doses that  are teratogenic in
humans, rats, and  rabbits and abortifacient in rhesus monkeys  (0.3 to 10
mg/kg), thus suggesting  that  the mouse embryo is more resistant to the
embryotoxic effects of methotrexate than any other mammalian  species yet
studied. The  minimum dose  required  to  produce teratogenicity in the
mouse was 25 mg/kg and is the same as that necessary for aminopterin to
be teratogenic (Table 2.3). In rabbits, methotrexate-induced abnormalities
increase as the  dose increases, a fact that is related to metabolism by the
maternal system (Jordan  1974). Jordan discovered  that when methotrexate
is present  in  low concentrations, it is rapidly converted by the maternal
enzyme aldehyde oxidase  to  a relatively  inactive  metabolite,  7-hydroxy-
methotrexate; at higher levels of methotrexate,  the enzyme becomes sat-
urated and, consequently, more unmetabolized drug reaches the embryo.

2.6.8.3 Thalidomide

   Table  2.2 indicates  that  human, monkey,  and rabbit embryos are
highly  susceptible   to thalidomide  but that  mouse,  rat,  and  hamster
embryos are not nearly as sensitive and show apparent strain differences.
In the study by Parkhie and Webb  (1983)  thalidomide teratogenicity was
demonstrated  in Wistar rats  by the administration of single intravenous
doses. This is in contrast to the data of Somers (1963) (see Table 2.2),
who, however, administered the compound  orally. Parkhie and Webb state
that  the intravenous administration of thalidomide may  circumvent the
factors involved in its loss of teratogenicity when given orally.
   Although both the monkey and the rabbit embryos  exhibit malforma-
tions similar to those observed in the human fetus, especially with regard

-------
                                 45

to limb defects, Delahunt (1965) concludes that the thalidomide-induced
abnormalities  in man are much more similar to those in the monkey than
they are to those in the rabbit. Although  Drobeck et al.  (1965)  do not
disagree with  Delahunt, they do believe that the results observed in the
rabbit  with thalidomide generally  confirm what  has been reported in
humans. They continue by stating that micromelia and phocomelia in the
human cases have been so popularized as to overshadow the wide range of
effects that has actually been observed including developmental failure of
visceral organs such as the kidneys and  adrenals as well as developmental
failure  in  bony structures. According  to Drobeck  and coworkers, these
effects are well characterized in the rabbit.
    Detailed examination  of  the data  on   rats, hamsters,  and mice in
Table 2.2  illustrates the apparent  strain differences in  susceptibility to
thalidomide. Fratta et  al. (1965) found no teratogenicity in mice strains
CFb  ICR,   C57,  CBA,  and   SJL;   DiPaolo  et   al.  (1964)  induced
malformations in  strain A, C3H, and  Swiss. Similar results are seen in
hamsters, where inbred lines are susceptible and random  bred  lines are
not, and in rats,  where the Sprague-Dawley and Holtzmann (descended
from Sprague-Dawley) strains and one  unidentified strain  produced  mal-
formed  young, whereas the Wistar (by oral  administration), Charles River,
Long Evans, and Dunning Fischer strains did not.
    Although  many explanations have been offered  for species  or strain
susceptibility  to the teratogenicity of thalidomide, the issue has  not  been
resolved. Gordon et al. (1981) postulated that thalidomide is metabolized
to a toxic  electrophilic intermediate that is not produced in vitro by liver
microsomes from  Sprague-Dawley rats,  which, they state are not a sensi-
tive species. This intermediate was produced by hepatic preparations  from
rabbits, monkeys,  and humans, species regarded by Gordon et al. (1981)
as sensitive. However, Table 2.2 shows that some strains of rats,  including
Sprague-Dawley, have demonstrated the  teratogenicity of thalidomide.
    Vaisman et al. (1983) have postulated that the mechanism of the tera-
togenic  action of thalidomide is the induction of a deficit of ascorbic acid
in species such as nonhuman primates and  man that do not have the abil-
ity to synthesize this vitamin.  In their tests  with the guinea  pig, which also
cannot synthesize  ascorbic acid, thalidomide was shown to induce a deficit
of ascorbic acid in both adults and  fetuses. Their premise is based on the
fact that  a deficiency of ascorbic acid leads to marked inhibition of col-
lagen synthesis (Matusis  1975, as reported  in Vaisman et  al. 1983) which
may damage  the  anlagen of the limbs.  However, teratogenicity has been
demonstrated  in  rabbits (including limb  malformations)  and,  although
perhaps not as convincingly, in rats, mice,  and hamsters (see Table 2.2),
which can synthesize ascorbic acid.
   Thus,  as stated by  Fabro (1981),  "We  must accept  the facts that a
clear understanding of the mechanism  by  which  thalidomide exerts  its
teratogenic effects escapes us  after almost two decades of intense scrutiny
and that the teratogen thalidomide still remains an embryologic enigma."

-------
                                  46
2.7 RECENT APPROACHES IN TERATOLOGY TESTING
   In recent years alternative approaches to the standard methods of test-
ing agents  for  teratogenicity  have been investigated,  primarily due to
economic and  practical considerations.  Most,  although not  all, of the
recently  proposed methods have been in vitro test  procedures. An investi-
gation of the current holdings of the ETIC concerning mammalian studies
reveals 889 indexed reference citations pertaining to various aspects of in
vitro teratogenicity. Analysis of these with respect to species indicates that
322  are  associated with rats, 409 with mice,  14 with hamsters, 48 with
rabbits,  1  with  monkeys, and  105 with humans. The total is small com-
pared to the approximately 25,000 citations recorded by ETIC for in vivo
testing, but, as many of the citations for in vitro testing are within the last
5 to 8 years, the current interest  in alternatives to whole mammal testing
is evident.

2.7.1 In  Vivo Methods

   An alternative to current embryo toxicity screening techniques was pro-
posed by Martian  and Jelinek (1979)  using a  caudal morphogenetic sys-
tem.  In  this proposed  teratology  screen, which the authors describe as a
compromise approach,  the  test substance is administered to  rats on days
10 and 11 of pregnancy, and the  effects on the  activity of the caudal mor-
phogenetic system  are  determined by measurements of  those parts  of the
trunk arising  through activity of that system of 13-day-old embryos. The
authors  concluded  that "the caudal morphogenetic system of rat embryos
reflected some  of  the  changes induced  in the  maternal-embryonic com-
plex."
    Chernoff and  Kavlock (1982) have proposed a teratology screen that
differs from previously  employed in vivo testing  methods in that it does not
require the labor-intensive examination of fetuses for soft-tissue and skele-
tal anomalies. Rather, the dams (CD-I mice were  used in the 1982  study)
are allowed to give birth, and determinations of teratogenic  potential are
made on the basis of litter size and weight on postpartum days 1  and 3 as
compared  with  concurrent  controls. Also,  because the  intent is priority-
setting and not testing,  only one dose level (at or near the maximally toxic
dose)  is  used, in comparison with at least three levels  in  conventional
teratology  test methods.
   In  the 1982 study,  Chernoff  and  Kavlock  tested 28  compounds of
known teratogenic  potential—15  that  were teratogenic by  standard test
criteria  did exhibit  some form   of developmental toxicity;  2 chemicals
known to produce only  reduced weight caused reduced weight (endrin) or
reduced  litter  size (sodium  selenite) in the proposed assay;  2 compounds
which  in standard  tests  caused  only  an  increased incidence of  super-
numerary ribs produced no effect in  the proposed  screen; and of 9 com-
pounds for which  no effects had  been  reported  in conventional test pro-
cedures,  6 produced no effects in the screen and  3 produced positive effects

-------
                                  47

(2 compounds caused reduction in litter weight and  1 caused reduced litter
size).
   Chernoff and Kavlock (1982) state that their teratology screen can be
used to divide chemicals into three categories: (1) those that induce peri-
natal death should be tested  as rapidly as possible, (2) those that induce
only perinatal weight changes would  be given a lower testing priority, and
(3)  those  producing no  effect  would be given the  lowest  priority. In
October  1984, the EPA released guidelines for the use of this preliminary
developmental toxicity  screen. Personal communication with Dr. Kavlock
has indicated that approximately 10  to 12 laboratories are currently  using
this  proposed method.  Results  of using this  teratology screen in four of
these laboratories have been reported by Schuler  et  al. (1984). Fifteen
glycol ethers were tested  using Charles River CD-I mice (no two labora-
tories tested the same chemical). Schular et al. provided results from con-
ventional testing for five of these. Of the five, three were known terato-
gens, and a correlation  with results of the proposed screen was seen in that
these three compounds  were designated either in a high priority  or middle
to high priority  group.  Two of the five compounds showed no reproductive
toxicity  in  conventional tests;  one was  assigned to  the middle priority
group and  the other to the low priority group in the in vivo screen.

2.7.2 In Vitro Methods
   Over the last few years several in vitro methods for  screening chemicals
for teratogenicity have been  proposed. Interest  in  testing for teratogenic
chemicals  in vitro  was evident as early  as  1975, when the  International
Conference on Tests of Teratogenicity In Vitro was held  (Kimmel  et al.
1982b). In August  1981 the consensus workshop on  In  Vitro Teratogenesis
Testing was convened in Arkadelphia, Arkansas, and the proceedings were
published   in   the   Journal  of  Teratogenesis,  Carcinogenesis,   and
Mutagenesis (Vol.  2, 1982).  Much of the information  presented below is
taken from those proceedings. As defined at the workshop and used in this
section, the term in vitro  refers to any system that uses test subjects other
than the intact pregnant mammal.

2.7.2.1 Whole mammalian embryo culture
   Sadler  et al. (1982) tested the potential teratogenicity of several com-
pounds using a whole  mouse embryo procedure. Their approach was to
administer the test compound at the  4- and 5-somite stages, which accord-
ing to them showed survival and a high percentage of malformations. They
found that early stages (primitive   streak to  0-1  somites) were  often
severely  affected and in some cases failed to survive. Sadler and coworkers
showed that glucose, ketone bodies, vitamin A, and hyperthermia produced
malformations (neural tube and/or cardiac effects)  in vivo. Although stat-
ing that the whole-embryo culture system may be useful as a screening

-------
                                   48

technique  for  potentially  teratogenic  substances,  they  recognize  that
unanswered questions, such as the appropriate exposure time, point to the
need  for  further  investigation  before  standardized  protocols  can  be
developed.
   The reader is referred to a paper by Kochhar (1975)  for a description
of early attempts to culture whole mammalian embryos.

2.7.2.2 Embryonic limb bud organ culture

   As described by Kochhar (1982), this method involves the removal of
limb buds from mouse embryos of known ages  (usually of the 12th day of
gestation). Limb buds are placed on top of ultrathin Millipore or Nucleo-
pore Filters and incubated in a nutrient medium in the presence or absence
of  test  chemicals.  End points  that  may be monitored  include  cell
proliferation, differential growth, morphogenetic cell death, size and shape
of limb  parts, chondrogenesis, and collagen or proteoglycan  biosynthesis.
Kochhar stated that  validation was still in the  early stages; to achieve full
potential  the   system  should  be  combined   with  an  efficient  drug-
metabolizing preparation.

2.7.2.3 Avian embryonic cells

    Wilk et al.  (1980) describe two in vitro systems using chick embryos,
one  using neural crest cells obtained from very early chick embryos and
the other, limb bud mesenchyme cells obtained at a later stage of develop-
ment.  Tests were performed both with and without S-9 added for meta-
bolic activation. The S-9 fraction was toxic to cells when added directly to
the medium, and thus the S-9 plus a NADPH-generating system and test
compound was combined in autoclaved dialysis bags and  added to the cul-
tures.  A combination of both the neural crest  and limb  mesenchyme sys-
tem correctly predicted the teratogenicity  of the known  teratogens tested
with the exception of thalidomide. Isoniazid and glutethimide, which are
not teratogenic in vivo, were also inactive in the Wilk et al. system.  Wilk
et al. (1980) state that a modification of their system by  incorporating the
use of postmitochondrial fractions from placental and  fetal tissues may
improve  its sensitivity and enable the detection of the teratogenic activity
of thalidomide.

2.7.2.4 Ascites tumor cell assay

   An in vitro teratogenicity assay  that identifies teratogens by their abil-
ity to inhibit the attachment of ascitic mouse ovarian tumor cells to plastic
surfaces  coated with concanavalin A has been  proposed by Braun et al.
(1982). This system  correctly identified 60 known teratogens, including
thalidomide, but failed to predict  the teratogenicity  of 14  known terato-
gens, including X  irradiation and methotrexate. With respect to predicting

-------
                                  49

nonteratogenicity, 21 nonteratogens were  correctly identified  but  7 com-
pounds  were false  positives. Recognizing  the limitations of any single in
vitro system, the authors state that "attachment inhibition in concert with
other, complementary, in vitro assay  systems can become a useful  method
for all  assessment  of the teratogenic potential of environmental agents."
[According  to  the  National Toxicology Program's annual plan for 1984
(NTP 1984), a complementary in vitro test measuring the chemical inhibi-
tion of growth  potential  of human  embryonic  mesenchymal cells  is
currently being conducted by two separate  laboratories. ]

2.7.2.5 Drosophila embryo cell assay

    Bournias-Vardiabasis  and  Teplitz (1982) report the development of an
in vitro assay that  detects teratogens by adding them to primary cultures
of embryonic Drosophila cells and analyzing the degree of change in cell
differentiation  and  tissue formation.  In validation tests of over 100 com-
monly used industrial chemicals, drugs, and food additives, only two false
negatives (ethyl alcohol and cyclophosphamide) and two false positives (p-
aminobenzoic acid  and  dimethylamino antipyrine) were observed. When
the serum of animals fed the  test agent is  added to the differentiating cul-
ture, metabolic products  of ingested  compounds can be tested for  terato-
genicity. The authors note that further testing, validation,  and incorpora-
tion of a metabolic  activation  system are needed.

2.7.2.6  Poxvirus morphogenesis
    Because the poxvirus has  a rapid, simple morphogenetic pathway that
is dependent upon  cell proliferation, the effect on the growth of poxvirons
in cell culture  can  be used as an indicator of teratogenic potential (Keller
and Smith 1982). In this assay, vaccinia WR-infected BSC 40 monolayers
are exposed to the  test agent for 24  h; the number of functional virons is
then  determined by plaque  assay. In tests with  42 known  teratogens,
Keller  and  Smith  observed  that  33  inhibited  the  number  of  virons
produced and 3 stimulated the production virons and were thus considered
teratogenic by  this  assay. Of  nine nonteratogens tested,  all but one (ascor-
bic acid) were  negative  in the poxvirus assay.  Thus six teratogens were
false negatives  and  one was a  false positive.

2.7.2.7  Neuroblastoma cells
    Mummery  et al. (1984) investigated the ability of 39 teratogenic  and
18  nonteratogenic  compounds to interfere  with  normal growth and  dif-
ferentiation of murine neuroblastoma cells (clone NIE-115).  This system
correctly predicted  the teratogenicity of 35 of the 39 and the nonteratogen-
icity of 14 of the  18. The  authors state that substances  such as thali-
domide that require metabolic activation for teratogenicity are unlikely to
be detected by this system.

-------
                                  50
2.7.2.8 Hydra attenuate system
   Johnson et al. (1982) describe a teratogenic screen that uses the fresh-
water Hydra attenuata. The first of a two-part experiment involves expos-
ing adult hydra to a test substance over a broad range of concentrations
and determining the  minimal toxic concentration to within one-tenth log.
In the second experiment, cells of dissociated hydra  are manipulated into a
configuration that permits  them  to  achieve  the  developmental  events
characteristic of any embryo and  undergo total whole-body  regeneration;
these cells are exposed during a 4-day ontogenesis  period to  the  test sub-
stance in the same manner as the  adult, and the minimally developmental
toxic concentration is determined to one-tenth  log.  Then, the ratio of the
adult (A) to the developmentally (D) toxic concentration is calculated. A
small A/D ratio indicates that the agent affects adults and developmental
events at generally the same dosage and, provided an adult toxic dose level
is not attained, the embryo will not be at risk. A large A/D ratio  indicates
the agent is  a primary developmental  hazard  with the  conceptus  as  the
most vulnerable target, and  at risk even at sub-adult toxic exposure levels.
This  system is also described by Johnson and Gabel (1982) and  by John-
son (1984).
    The  Hydra attenuata system  has  been  validated by comparing  the
A/D ratios for chemicals tested using this system with the A/D ratios for
chemicals tested using conventional developmental toxicity test procedures.
Data supplied by Dr.  Johnson of comparison of the  A/D ratios of approxi-
mately 40  test substances  indicate a good correlation in most instances.
The A/D ratios of mammals and hydra seldom varied by as much as a
factor of 2. One  exception  was 6-aminonicotinamide with A/D  ratios of
over  4  for  mammals  and  26 for hydra. Personal communication  with
Dr.  Johnson indicates that the use of the bioactivation in the  hydra system
is currently being  investigated under the aegis of the March of Dunes.

2.7.2.9 Planarian assay
    Best  and Morita (1982) describe the results of experiments using fresh-
water planarians as indicator  organisms in a teratogenicity  assay. Either
surgical   fragments  of planarians  that  undergo  regeneration or  intact
planarians can be used as test organisms.  When fragments are used,  dis-
tortion of the regeneration is the measure of teratogenicity, and with intact
planarians the development of morphologic or other observable abnormali-
ties is the indicator. Citing the results of testing several chemicals for tera-
togenicity, Best and Morita state that "both regeneration  of surgical frag-
ments and aberrant  remodeling  of whole  planarians model important
features  of embryogenesis and are potentially  useful for  assaying terato-
gens." The authors suggest that the planarian system might also be useful
in screening for chemicals that  cause behavioral  teratogenicity.

-------
                                   51

2.7.2.10 Frog embryo assay

   In  this method described by  Dumont et al. (1982) and  Dumont and
Epler  (1984), early frog embryos are exposed for 96 h to concentration
ranges of suspect agents. The following end points may be observed: LC50,
EC50 (concentration required to  produce 50% terata  among survivors),
growth and developmental  stages  attained,  anatomical abnormalities,
motility   (behavior),  and  pigmentation.   Personal  communication  with
Dr.  Dumont indicates that 7 of  41 compounds (17%) tested were false
negatives  and 5 of 41 (12%) were false positives.

2.8 CONCLUSIONS  AND  RECOMMENDATIONS  FOR  FURTHER
    RESEARCH

2.8.1 Conclusions

2.8.1.1 General experimental considerations

  1.  At least three dose levels of a chemical should be tested. The low dose
     should  allow normal development, the intermediate dose should show
     teratogenicity if the chemical tested is a teratogen, and the high dose
     should  produce either maternal toxicity, embryotoxicity, or fetotoxi-
     city.

  2.  Positive controls cannot provide certainty that the test chemical is not
     teratogenic if tests indicate no teratogenicity.
  3.  For each chemical at least two species should be tested for teratogeni-
     city.
  4.  The number of pregnant test animals per group must be sufficient to
     permit  reliable statistical  analysis.   A  minimum  of  20  pregnant
     animals of a rodent species and 12 pregnant rabbits has been recom-
     mended by several regulatory groups.
  5.  The administration route should be oral intubation unless the physical
     and chemical properties of the chemical or human use conditions dic-
     tate otherwise.
  6.  All fetuses should be examined for external malformations. Approxi-
     mately  one-third to one-half  of the fetuses of rats, mice,  and hamsters
     should be examined for skeletal defects and the remainder for soft tis-
     sue  malformation.  All rabbit fetuses should  be examined for both
     skeletal and soft tissue anomalies.

2.8.1.2 Structure-activity relationships
   Although selected  teratogenicity studies seem to illustrate certain
structure-activity  relationships,  generalizations regarding the  prediction  of
human risks from chemical structure are  not yet feasible. This conclusion

-------
                                  52

is  supported  by the statement of Zimmerman (1975) that "Prediction  of
teratogenesis by chemical structure will have to  await a future time" and
of that by  Schumacher (1975) that "Structure-activity  studies,  as  per-
formed until now, will not, we feel, provide the needed information and are
therefore  of  no predictive  value with regard to  teratogenicity in animals,
much less in humans."

2.8.1.3 Teratogenesis and time of administration

   When  a  chemical  is  tested  for  teratogenicity,   observations  for
behavioral as well as physical defects should be  performed. Although it is
generally  recognized  that  organogenesis is the most  sensitive prenatal
developmental period to the malforming action  of teratogenic chemicals,
the possibility of a test chemical inducing  behavioral  teratogenic effects
when administered in each of the three trimesters (and even pregestation-
ally, see Section 2.5) must be considered.

2.8.1.4 Behavioral teratogenesis

    After  a slow start, interest in behavioral  teratogenicity increased during
the early  1970s,  and today it is a widely recognized and respected science.
Although  some  countries  (the United  States  excluded) have  proposed
guidelines for the testing of agents for behavioral teratogenicity,  standardi-
zation of  test methodologies is still needed.

2.8.1.5 Species comparisons
    There is no ideal animal model  for use in teratological research. The
monkey (primarily the higher primates) has  received considerable attention
in recent years,  especially since the thalidomide syndrome appears to  be
mimicked to  a high degree  in  the  monkey.  This  animal offers  the
advantage of having a chorioallantoic placenta as does man, in contrast to
the inverted yolk-sac placenta  of rodents  and  rabbits,  and having other
anatomical  and  physiological similarities to man. Yet, studies using the
folic acid antagonists methotrexate and aminopterin, known human terato-
gens, have been unable to  demonstrate any significant teratogenicity in the
monkey, although abortions have been induced.  However, if a  compound
whose embryotoxicity is being  tested  was  shown to  have abortifacient
action but   not  necessarily  teratogenic potential  (inducing  malformed
offspring), the  use of the compound would probably  be restricted for
women of child-bearing age. Thus,  the decision on potential use of this
compound by  pregnant  women  would probably be the same whether it
induced malformations or  abortions.  Using  the  criteria of embryotoxicity
(embryolethality, malformations, or growth  retardation),  Wilson  (197la)
has shown that the effects induced by several agents in  man and in the
rhesus monkey (chemical and viral) are comparable. Wilson acknowledges
that before the rhesus monkey can be considered more reliable than other
laboratory animals for anticipating embryotoxicity in  man,  more  detailed

-------
                                   53

comparative  information on  drug  metabolism  and distribution between
laboratory animals must be obtained. However, even if the rhesus monkey
is shown to be more reliable  than other  laboratory animals for predicting
embryotoxicity in man, the expense will likely prohibit serious considera-
tion as a routine test animal for embryotoxicity.
   Rats,  mice, and rabbits  have been  and will  probably continue to  be
used more frequently than the monkey as laboratory test animals primarily
due to  availability,  ease  of handling, and  economy. Rats and mice have
been the  most commonly used, but use  of the rabbit  has increased since
the early 1960s because the rabbit was the first  laboratory animal  shown
to be susceptible to thalidomide teratogenicity.
   All  chemicals do not require the same degree of teratogenicity testing.
Compounds for which human exposure,  especially  to  pregnant women, is
considered a probability should receive more intensive  teratological  testing
than those chemicals for  which human exposure is  a remote possibility. In
the same vein, Wilson (1978) believes that despite  the limited availability
of higher  primates, testing  with macaques and baboons is essential for
those drugs needed for therapeutic purposes during  human pregnancy and,
perhaps, for the environmental chemicals to which women may be inadver-
tently heavily exposed prior to the  diagnosis  of pregnancy. Wilson  contin-
ues by  stating that potential teratogenic  effects to the  central nervous sys-
tem  can be adequately tested only in those higher  primates  that have, at
least roughly, a range of mental and nervous activities comparable to man.

2.8.1.6  Recent approaches in teratology testing
   Investigations  of practical alternatives to  standard   in  vivo  testing
methods for teratogenicity screening are needed because of the cost and
labor-intensive efforts. Although a few recent approaches have used in vivo
procedures, most  have used  systems other than the  whole  mammal.  At
present, although validations of some of these  in  vitro  methods are still
being conducted, no single test can be considered as a screen; a battery of
tests is more appropriate. For further insights  into in vitro  screening of
agents for teratogenicity the reader is referred to the overview  of the 1981
proceedings of the consensus workshop  on In Vitro Teratogenesis Testing
(Kimmel et al. 1982b).

2.8.2 Recommendations for Further Research
   The following are identified as possible research areas for the  terato-
genicity testing of chemicals:
•  An  extensive analysis, using existing  data, of the teratogenic action of  a
   select group of chemicals representing different chemical classes. Such
   an   analysis  should enhance our  understanding of species  specificity,
   mechanisms of action, and structure-activity relationships.

-------
                                  54
•  Further investigation of postnatal behavioral changes as they relate to
   pregestational and gestational chemical exposure.
•  Further research directed  toward standardization of the methodology
   for screening chemicals for behavioral teratogenicity.
•  Further research on variations on present in vivo test procedures, espe-
   cially with regard to economy and simplicity.
•  Continued validation of existing in vitro methods for screening com-
   pounds for teratogenicity as well as further research on the identifica-
   tion of additional in vitro test systems.
•  In  conjunction with the  above, determination of an acceptable battery
   of tests for in vitro screening of compounds for teratogenicity.


2.9 LITERATURE CITED

Baranov VS.  1965. Injurious effect of aminopterin at different  stages in
the embryonic development in rats. Embryology 16:448-450.

Becci  PJ, Reagan  EL, Knickerbocker MJ, Barbee  SJ, Wedig JH.  1983.
Teratogenesis study of o-toluenediamine in rats and rabbits. Toxicol Appl
Pharmacol 71:323-329.

Best JB, Morita M. 1982. Planarians as a model system for in vitro terato-
genesis studies. Teratog Carcinog Mutagen 2:277-291.

Beyer  BK, Guram MS, Geber WF. 1984. Incidence and potentiation of
external and internal fetal anomalies resulting  from chlorodiazepoxide  and
amitriptyline alone and in combination. Teratology 30:39-45.

Bournias-Vardiabasis N, Tepliz RL. 1982. Use of Drosophila embryo  cell
cultures as  an  in  vitro teratogen  assay.  Teratog  Carcinog  Mutagen
2:333-341.

Braun  AG, Buckner CA, Emerson DJ, Nichinson BB.  1982. Quantitative
correspondence between the in vivo and in vitro  activity of  teratogenic
agents. Proc Nat  Acad Sci USA 79:2056-2060.

Brown AM. 1963. Matching the animal with the experiment. In: Animals
for research.  Lana-Petter,  ed.  London:  Academic,  pp.  261-685.   (As
reported in Palmer  1978).

Brown D. 1978.  Structure-activity relation  of  steroidal amine  teratogens.
In: Effects of poisonous plants on livestock. Keeler RF, VanKampen KR,
James  LF, eds. New York: Academic, pp. 409-418.

-------
                                 55

Brown D, Keeler RF.  1978.  Structure-activity relation of steroid terato-
gens. I. Jervine ring system. J Agric Food Chem 26(3):561-569.

Brummett  ES,  Johnson  EM.  1979.  Morphological  alterations  in  the
developing fetal rat limb due to maternal injection of chlorambucil. Tera-
tology 20(2):279-288.

Buelke-Sam J, Kimmel CA. 1979. Development and standardization of
screening methods for behavioral teratology. Teratology 20:17-30.

Chernoff N, Kavlock RJ. 1982. An in vivo teratology screen utilizing preg-
nant mice. J Toxicol Environ Health 10:541-550.

Collins, TFX, Collins  EV.  1976.   Current  methodology  in  teratology
research. In: Advances in  modern toxicology. Mehlman MA, Shapiro  RE,
Blumenthal H, eds.  New York: John Wiley, pp. 155-174.

Coyle, I, Wayner MJ, Singer G. 1976. Behavioral teratogenesis: a critical
evaluation. Pharmacol Biochem Behav 4:191-200.

Delahunt CS.  1965. Teratogenic effects of thalidomide in the rabbit, mon-
key, and man. In:  Proceedings of second teratology  workshop, Berkeley,
Calif, pp. 51-81.

Delahunt, CS, Lassen LJ, Rieser  N. 1966. Some comparative teratogenic
studies with thalidomide. Proc Eur Soc Study Drug Toxic. 7:229-240.

DeSesso JM.  1979.  Lectin teratogenesis: defects produced by Concanavalin
A in fetal rabbits. Teratology 19:15-26.

DiPaolo,  JA, Gatzek H, Pickren  J.  1964.  Malformations  induced  in the
mouse by thalidomide. Anat Rec 149:149-156.

Drobeck  HP,  Coulston  F, Cornelius D. 1965. Effects of thalidomide on
fetal development in rabbits and  on establishment of pregnancy in mon-
keys. Toxicol Appl Pharmacol 7:165-178.

Dumont JN, Schultz TW, Newman  SM.  1982. A frog embryo terato-
genesis assay:  Xenopus (FETAX)—a model for teratogen screening. Tera-
tology 25:37A.

Dumont JN,  Schulz TW, Elper RG.  1983. The response of the FETAX
model to mammalian teratogens. Teratology 27:39A.

Dumont JN,  Elper  RG. 1984. Validation studies on the  FETAX terato-
genesis assay (frog embryos). Teratology 29(2):27A.

-------
                                  56

Dwornik JJ, Moore KL.  1965. Skeletal malformations in the Holzman rat
embryo following the administration of thaliomide. J Embryol Exp Mor-
phol 13(2):181-193.

Earl FL,  Miller E, Van Loon EJ. 1973. Teratogenic research in  beagle
dogs and miniature swine. In:  The  laboratory  animal in  drug  testing.
Spiegel A, ed. Stuttgart: Gustav Fischer Verlag, pp. 233-247.

Eibs HG, Spielmann H, Hagele M.  1982. Teratogenic effects of  cypro-
terone  acetate and medroprogesterone treatment during the pre- and post-
implantation period of mouse embryos. I. Teratology 25:27-36.

Eibs HG, Spielmann, Jacob-Muller U,  Klose J. 1982. Teratogenic  effects
of cyproterone acetate  and medroxyprogesterone treatment during the pre-
and postimplantation period of mouse embryos.  II.  Cyproterone acetate
and medroxyprogesterone  acetate treatment  before implantation  in vivo
and in vitro. Teratology 25:221-299.

Emerson  DJ. 1962. Congenital malformation due to attempted abortion
with aminopterin. Am J Obstet Gynecol ll(3):356-357.

Fabro  S.  1981.  Biochemical basis  of thalidomide teratogenicity.  In:
Biochemical  basis of chemical teratogenesis. Juchau MR,  ed. New  York:
Elsevier, pp.  159-178.

Fantel AG, Macphail  BJ.  1982. The  teratogenicity of cocaine. Teratology
6:17-19.

Perm VH. 1967. The  use  of the golden hamster  in experimental teratol-
ogy. Lab Anim Care 17(5):452-462.

Perm  VH,   Hanlon  DP.  1974.  Toxicity  of copper  salts in hamster
embryonic development. Toxicol Reproduction 11:97-101.

Fratta ID, Sigg  EB, Maiorana K.  1965. Teratogenic effects of thalidomide
in rabbits, rats, hamsters, and mice. Toxicol Appl Pharmacol 7:268-286.

Freese E, Levin  BC, Pearce R, et al. 1979. Correlation between the growth
inhibitory effects, partition  coefficients and teratogenic effects of lipophilic
acids. Teratology 20:413-440.

Friedler  G.  1978. Pregestational  administration  of morphine sulfate  to
female mice:  long-term effects on the development of subsequent progeny.
J Pharmacol Exp Ther  205( 1 ):33-39.

Gauron EF, Rowley VN. 1969.  Effects on offspring behavior of mothers'
early chronic drug experience. Psychopharmacologia (Berl.) 16:5-15.

-------
                                 57

Gordon  GB,  Spielberg  SP,  Blake  DA,  Balasubramanian  V.   1981.
Thalidomide teratogenesis:  evidence for a toxic  arene  oxide metabolite.
Proc Natl Acad Sci USA 78(4):2545-2548.

Hansson CG,  Angervall L.  1966. The parathyroids  in  corticosteroid-
treated pregnant rats and their offspring. Acta Endocrinol 53:547-552.

Hay MF. 1964. Effect of maternal treatment with various agents on early
embryonic development in the rabbit. In: Proceedings of 5th International
Congress  of Animal Reproduction and Artificial  Insemination,  Trento,
Italy, pp. 309-312.

Hendrickx AG. 1970. Teratogenicity of thalidomide in the baboon (Papio
cynocephalus), bonnet monkey (Macaco radiatd)  and cynomolgus monkey
(Macaca irus). Proc 3rd Int Congr Primat 2:230-237.

Hendrickx AG. 1972. A comparison of temporal factors in the embryologi-
cal development  of man, old world  monkeys, and galagos, and craniofacial
malformations induced  by thalidomide and triamcindone. Medical Prima-
tology (Switzerland), Proc  3rd  Conf Exp Med Surg Primates, Part III,
pp. 259-269.

Hendrickx AG,  Binkerd  PE,  Rowland  JM. 1983. Developmental  toxicity
and nonhuman primates. In:  Issues and reviews  in teratology.  Kalter H,
ed. New York: Plenum Press,  pp. 181-214.

Hendrickx AG, Newman L. 1972. Temporal and mandibular bone malfor-
mations induced by thalidomide in the bonnet monkey (M. radiata). Tera-
tology 5(2):257.

Hoffeld DR, Webster RL. 1965. Effect of injection of tranquilizing drugs
during pregnancy on the offspring. Nature 295:  1070-1072. (As reported
in Coyle et al.  1976).

Homburger F,  Chaube  S,  Eppenberger M,  Bogdonoff PD,  Nixon CW.
1965. Susceptibility  of certain  inbred strains of hamsters to teratogenic
effects of thalidomide. Toxicol Appl Pharmacol 7:686-693.

Hutchings  DE.  1983.  Behavoral  teratology:  a  new frontier  in  neuro-
behavioral research. In: Handbook of experimental pharmacology. Vol. 65.
Born GVR, Farah A, Herken  H, Welch AD, eds, pp. 207-235.

Hutchings DE,  Gibbon  J, Kaufman MA.  1973.  Maternal Vitamin  A
excess during the early  fetal period: effects on learning and development in
the offspring. Dev Psychobiol  6(5):445-457.

-------
                                 58
Hutchings DE, Gaston  J. 1974. The effects of Vitamin A excess admin-
istered during the  mid-fetal period on learning  and  development in rat
offspring. Dev Psychobiol 7(3):225-233.

Inouye M, Murakami  U.  1977. Teratogenicity of  2,5-diaminotoluene,  a
hair-dye constituent, in mice. Food Cosmet Toxicol 15:447-451.

Iverson F, Khera KS, Hierlihy SL. 1980. In vivo and  In vitro metabolism
of ethylenethiourea in the  rat  and  the cat.  Toxicol Appl  Pharmacol
52:16-21.

Jensh RP. 1983. Behavioral testing procedures: a review. In: Teratogenesis
and  reproductive toxicology.  Johnson EM  and Kochhar  DM,  eds.  New
York: Springer-Verlag,  pp. 171-206.

John JA,  Blog DD, Murray FJ, Schwetz BA, Gehring PJ. 1979. Terato-
genic effects of the plant hormone indole-3-acetic acid in  mice and rats.
Teratology 19(3):321-326.

Johnson  EM.  1984.  A  prioritization and biological decision tree for
developmental toxicity safety evaluations.  J Am Coll Tox 3(2):141-147.

Johnson EM, Gabel BEG. 1982. Application of the  Hydra assay for rapid
detection of developmental hazards. J Am Coll Tox 1(3):57-71.

Johnson  EM, Gorman RM,  Gabel  BE,  George  ME. 1982. The Hydra
attenuata system for detection  of teratogenic  hazards. Teratog Carcinog
Mutagen 2:263-276.

Joneja MG. 1981.  Teratogenic  effects of proline  analogue L-azetidine-
2-carboxylic  acid in hamster fetuses. Teratology 23:365-372.

Jordan RL.  1973. Response of the rabbit  embryo to  methotrexate. Teratol-
ogy 7(3): 19 A.

Jordan RL.  1974. Studies on  the relationship between methotrexate meta-
bolism and teratogenicity. Excerpta Med:89.

Jordan RL, Terapane JF, Schumacher JH. 1970.  Studies on the teratogen-
icity of methotrexate in rabbits. Teratology 3:198.

Kacew S, Ruddick JA, Parulekar M,  Valli VE,  Chu I,  Villeneuve DC.
1984. A teratological evaluation  and analysis of fetal tissue levels following
administration  of  tetrachlorobenzene  isomers   to  the  rat.  Teratology
29:21-27.

-------
                                 59
Kajii TM,  Kida M,  Takahashi K. 1973. The effect of thalidomide intake
during 113 human pregnancies. Toxicology 8:163-166.

Kalter H.  1968. Teratology of the central nervous system.  Chicago: The
Univesity of Chicago Press, 483 pp.

Kao J, Brown NA, Schmid B, Goulding EH, Fabro S. 1981.  Teratogeni-
city of valproic acid: in vivo and in vitro investigations. Teratog Carcinog
Mutagen 1:367-382.

Keeler  RF.  1970.  Teratogenic  compounds  in  Veratrum caifornicum
(Durand). IX. Structure-activity relation. Teratology 3:169-174.

Keller  SJ, Smith MK.  1982. Animal virus screens for potential teratogens.
I. Poxvirus morphogenesis. Teratog Carcinog Mutagen 2:361-374.

Khera  KS.  1973.  Ethylenethiourea: teratogenicity study in  rats and rab-
bits. Teratology 7:243-252. (As reported in Khera and  Iverson 1978).

Khera  KS.  1975.  Fetal cardiovascular and other defects induced by tha-
lidomide in cats. Teratology  11:65-72.

Khera  KS,  Tryphonas L. 1977.  Ethylenethiourea-induced hydrocephalus:
pre- and post-natal pathogenesis in offspring  from rats given a single oral
dose during pregnancy. Toxicol Appl Pharmacol 42:85-97. (As reported in
Khera  and Iverson 1978).

Khera  KS, Iverson F. 1978. Toxicity of ethylenethiourea in  pregnant cats.
Teratology 18(3):311-314.

Kimmel  CA. 1984. Behavioral teratology in  experimental animals.  Paper
presented at the European  Teratology  Society Meeting, The Netherlands,
September 1984.

Kimmel CA, ed. 1977. Final report of the committee on postnatal evalua-
tion of animals subjected to insult during development.  Research Triangle
Park, North Carolina: National Institute of Health. (As reported in Jensh
1983).

Kimmel  CA, Buelke-Sam  J,  Adams  J, Reiter  LW, Sobotka  TJ,  Tilson
HA.  1982a. Design considerations in  the  evaluation of standardized
methods  in a behavioral teratology protocol. Teratology 25:54A.

Kimmel  CA, Buelke-Sam  J,  Adams J, Reiter  LW, Sobotka  TJ,  Tilson
HA.  1983. An intralaboratory comparison  of selected  methods  in  a
behavioral teratology protocol.  Teratology 27:57A.

-------
                                  60
Kimmel GL, Smith K,  Kochhar DM,  Pratt RM.  1982b.  Proceedings of
the consensus workshop on in vitro teratogenesis testing. Teratog Carcinog
Mutagen 2:(34) i-ii.

King, CTG, Horigan  E, Wilk AL. 1972. Fetal  outcome from prolonged
versus  acute drug administration in the pregnant  rat. Drugs Fetal Dev,
pp. 61-75.

King CTG, Kendrick  FJ. 1962.  Teratogenic effects of thalidomide in the
Spague Dawley rat. Lancet 2:1116

King CTG, Weaver SA,  Narrod SA. 1965. Antihistamines and teratogeni-
city in the rat. J Pharmacol Exp Ther 147:391-398.

Kochhar DM. 1975. The use of in vitro procedures in teratology. Teratol-
ogy 11:273-288.

Kochhar DM. 1982. Embryonic limb bud organ culture in assessment of
teratogenicity  of environmental  agents.  Teratog   Carcinog  Mutagen
2:303-312.

Koppanyi T, Avery MA. 1966. Species differences  and the clinical trial of
new drugs: a review.  Clin  Pharmacol  Ther 7:250-270.  (As reported in
Schardein  1976).

Kuroki  T,  Drevon   C.  Montesano   R.   1977.  Microsome-mediated
mutagenesis in V79 Chinese hamster cells by various nitrosamines. Cancer
Res 37(4): 1044-1050.

Lasagna L. 1979. Toxicological barriers to providing better drugs. Arch
Toxicol 43:27-33. (As  reported in Jensh 1983).

Leonard BE.  1982. Behavioural  teratology: post-natal  consequences of
drug exposure in utero. Arch Toxicol Suppl 5, 48-58.

Leonard BE.  1983. Behavioral teratology  and toxicology. Psychophar-
macology (Amsterdam): l(l):248-299.

Leone  VG. 1977.  Comparative aspects  of developmental stages in mam-
mals used for teratogenetic tests. In: Methods in  prenatal  toxicology. Neu-
bert D, Merker H-J, Kwasigroch TE, eds. Massachusetts: PSG, pp. 14-24.

Lenz W, Knapp K. 1962. Thalidomide embryopathy. Arch Environ Health
5:110-105.

-------
                                  61
Marhan  O, Jelinek R.  1979.  Efficiency of embryotoxicity  testing pro-
cedures. I. A compromise approach. Toxicol Lett 4:385-388.

Marks TA, Worthy WC, Staples RE. 1980. Teratogenic potential of phen-
cyclidine in the mouse. Teratology 21:241-246.

Marks TA, Gupta BN, Ledoux TA, Staples RE. 1981. Teratogenic evalua-
tion   of  2-nitro-p-phenylenediamine,  4-nitro-o-phenylenediamine,  and
2,5-toluenediamine sulfate in the mouse. Teratology 24:253-265.

Matsubara Y, Goto M, Mikami T, Suzuki Y, Chiba T.  1983. Teratogenic
effects of thalidomide in the  rabbit: difference in  susceptibility between
two breeds. Cong Anom 23:223-229.

Matusis  II. 1975. Vitamins and  antivitamins.  Moscow. (As reported in
Vaisman et al. 1983).

McBride WG, Vardy PH. 1983. Pathogenesis of thalidomide teratogenesis
in the  marmoset (Callithrix  jacchus):  evidence  suggesting a  possible
trophic  influence of  cholinergic nerves  in limb  morphogenesis. Develop
Growth Differ 25(4):361-373.

Mehlman MA, Craig PH, Gallo  MA. 1984.  Teratological evaluation of
trimethyl phosphite  in the rat. Toxicol Appl Pharmacol 72:119-123.

Meltzer  HJ.  1956.  Congenital  anomalies due to  attempted abortion with
4-aminoteroglutamic acid. J Am Med Assoc 161(13):1253.

Merkle J, Schulz V, Gelbke HP.   1984. An embryotoxicity study of the
fungicide tridemorph  and its commercial formulation Calixin®. Teratology
29:259-269.

Milunsky AJ,  Graef  W, Gaynor MF  Jr.  1968.  Methotrexate-induced
congenital malformations. J Pediatr 72(6):790-795.

Mummery  CL, Van  Den Brink CE,  Van Der Saag PT, De Laat SW.
1984. A short-term  screening test for teratogens using differentiating neu-
roblastoma cells in vitro. Teratology 29:271-279.

Murai N.  1966.  Effect  of maternal medication during pregnancy  upon
behavioral development of offspring. Tohokv J  Exp Med 89:265-272. (As
reported in Coyle et al. 1976).

Murphy  ML. 1962. Teratogenic effects in rats of growth inhibiting chemi-
cals including studies of thalidomide. Clin Proc Child Hosp 18:307-322.

-------
                                62

NAS. 1977. National Academy of Sciences. Principles and procedures for
evaluating the toxicity of household substances. Washington, D.C.: Print-
ing and Publishing Office. 130 pp.

NTP. 1984. National Toxicology  Program. Fiscal year 1984 annual plan.
U.S. Department of Health and Human Services, p. 174.

OECD. 1981. Office  of Economic Cooperation and Development. OECD
guidelines for testing of chemicals. Section 4, No. 414.

Palmer AK.  1968. Spontaneous malformations of the New Zealand white
rabbit: the background  to  safety evaluation tests. Lab  Anim  2:195-206.
(As reported in Palmer 1978).

Palmer AK.  1978. The design of subprimate animal studies. In: Handbook
of teratology, Vol. 4. Research procedures and data analysis. Wilson JG,
Fraser FC, eds. New York: Plenum, pp. 215-253.

Parkhie M, Webb M.  1983. Embryotoxicity and  teratogenicity  of thali-
domide in rats. Teratology 27(3):327-332.

Petrere JA, Humphrey RR, Sakowsky R, Fitzgerald JE, de la Iglesia FA.
1984.  Teragology  study with the synthetic prostaglandin ONO-802 given
intravaginally to rabbits. Terato Carcio Mutagen 4:225-231.

Pfeiffer RA, Kasenow W.  1962. Thalidomide  and congenital  abnormali-
ties. Lancet 1:45-46.

Poswillo DC, Hamilton  WJ, Sopher D. 1972.  The marmoset as an animal
model for teratological research. Nature (London) 239(5366):460-412.

Powell HR,  Ekert  H.  1971.  Methotrexate-induced  congenital malfor-
mations. Med J Aust 2:1076-1077.

Randall CL,  Taylor WJ. 1979.  Prenatal ethanol exposure in mice: terato-
genic effects. Teratology 19(3):305-312.

Rodier PM. 1978.  Behavioral teratology. In: Handbook of teratology,  Vol.
4.  Research  procedures  and data analysis. Wilson JG, Fraser FC,  eds.
New York: Plenum, pp. 397-428. (As reported in Jensh 1983).

Rodriguez-Gonzalez MD, Lima-Perez MT, Sanabria-Negri JG. 1983.  The
effect  of  cypropeptadine  chlorhydrate  on rat  embryonic development.
Terato Carcino Mutagen 3:439-446.

-------
                                  63

Rowland  JM,  Hendrickx AG.  1983.  Teratogenicity of  triamcinolone
acetonide in rats.  Teratology 27:13-18.

Saad  DJC,  Kirsch RM, Kaplan, LL, Rodwell  DE.  1984.  Teratology of
intravaginally administered  contraceptive jelly containing  octoxynol-9 in
rats. Teratology 30:25-30.

Sadler TW,  Horton WE, Warner  CW. 1982. Whole embryo culture: a
screening technique for teratogens? Teratog Carcinog Mutagen 2:243-253.

Sadler TW, Kochhar DM. 1975. Teratogenic effects of chlorambucil on in
vivo and in vitro organogenesis in mice. Teratology 2:71-78.

Schardein JL. 1976. Drugs as teratogens. Cleveland, Ohio: CRC, 291 pp.

Schardein JL. 1983. Teratogenic risk assessment. In: Issues and reviews in
teratology. Kalter H, ed.  New York: Plenum, pp. 181-214.

Schuler RL,  Hardin BD, Niemeier RW et al. 1984. Results of testing  fif-
teen glycol  ethers in  a short-term in vivo reproductive  toxicity  assay.
Environ Health Perspect  57:141-146.

Schumacher  HJ.  1975. Chemical structure and teratogenic  properties.  In:
Methods for detection of environmental  agents  that produce congenital
defects. Shepard  TH,  Miller JR,  Marois M, eds.  New York: American
Elseiver, Inc., pp. 65-77.

Seegmiller RE, Nelson GW, Johnson CK.  1983.  Evaluation of the  terato-
genic potential of delalutin (17a-hydroxyprogesterone caproate)  in mice.
Teratology 28:201-208.

Shepard TH. 1976. Catalog of teratogenic agents. Johns Hopkins Univer-
sity. 291 pp.

Siddall RA. 1978. The use of marmosets (Callithrix jacchus) in teratolog-
ical and toxicological research. Primates Med 10:215-224.

Silverman P. 1978.  Animal behavior in  the laboratory. New York: Pica,
pp. xi-409. (As reported in Jensh 1983).

Skalko RG, Gold MP.  1974. Teratogenicity of methotrexate in mice. Tera-
tology 9:159-164.

Smithells RW, Bond MB. 1962.  Thalidomide and  malformations.  Lancet
1:1270-1273.

-------
                                 64
Somers GF. 1963. The foetal toxicity of thalidomide. Proc Eur Soc Study
Drug Toxic 1:49-58.

Somers  GF.  1962.  Thalidomide  and congenital  abnormalities.  Lancet
1:912-913.

Spyker JM, Avery  DL.  1976.  Behavioral toxicology and the developing
organism. In:  Proceedings,  workshop on  behavioral  toxicology. DHEW
Publication No. (NIH)76-1189.  Washington, B.C.: U.S. Dept. Health,
Education,  and  Welfare, pp.  95-109. (As reported in Jensh 1983).

Sullivan  FM.  1970.  Mechanisms of action of teratogenic drugs. Proc  R
Soc Med 63(12):42-43.

Tanimura  T.  1972.  Effects of macaque  embryos  of drugs  reported  or
suspected to be teratogenic  to humans: discussion paper.  Acta Endocrinol
(Copenhagen) Suppl. 16:293-308.

Thompson  DJ,  Dyke IL, Molello JA. 1984. Reproduction and teratology
studies on hexamethylmelamine in the rat and rabbit. Toxicol Appl Phar-
macol 72:245-254.

Thompson  JF,  Schweisthal MR.  1969. Study of closure of the embryonic
rat palate  in  vitro  with  the effects of certain chemicals.  J Dent  Res
48(4):568-572.

Tuchmann-Duplessis H.  1972.  Teratogenic  drug screening, present  pro-
cedures and requirements. Teratology 5:271-286.

Tuchmann-Duplessis H, Mercier-Parot L.  1966.  Effect of two chemically
related antimetabolites  on  the  embryo.  Bull Schweiz Akad  Med  Wiss
22:153-165.

USEPA.  1982a. U.S.  Environmental Protection Agency. Health effects
test guidelines.  Organ/tissue teratogenicity study. Washington, DC: Office
of Toxic Substances.  EPA 560/6-82-001.

USEPA.  1982b. U.S.  Environmental Protection Agency. Subdivision  F:
Pesticide assessment guidelines.  Washington,  D.C.:  Office of Pesticide
Programs, pp. 126-130.

USEPA.  1984. U.S.  Environmental  Protection  Agency.  Health  Effects
Test Guidelines.  Specific organ/tissue  toxicity—primary developmental
toxicity screen.  Washington,  D.C.: Office of Toxic Substances, 10 pp.

-------
                                  65

USFDA. 1982. U.S. Food and Drug Administration. Toxicological princi-
ples for the safety assessment of direct food additives and color additives
used in food. Washington,  D.C.:  U.S. Food  and  Drug Administration,
pp. 108-118.

Vaisman BL, Popov  VP, Ignat'eva TV.  1983. Reduction  in  the  tissue
ascorbic acid level  in guinea pigs by  thalidomide.  Byull Eksp Biol Med
96(7):27-29.

Vondruska JF, Fancher OE, Calandra  JC. 1971. An investigation into the
teratogenic potential of captan, folpet, and difolation in nonhuman pri-
mates. Toxicol Appl Pharmacol 18:619-624.

Vorhees  CV. 1983.  Behavioral  teratogenicity  testing  as   a  method  of
screening for hazards to human health: a methodological proposal. Neuro-
behav Toxicol Teratol 5:469-474.

Vorhees CV, Brunner RL, McDaniel CR, Butcher RE. 1978. The relation-
ship of gestational age to vitamin  A induced postnatal dysfunction. Tera-
tology 17:271-276.

Vorhees CV, Butcher  RE.  1982.  Behavioral teratogenicity. In: Develop-
mental toxicology.  Snell K, ed.  London: Croom-Helm,  pp.  247-298. (As
reported in Vorhees 1983).

Vorhees CV, Butcher RE, Brunner RL, Sobotka TJ. 1979. A developmen-
tal test battery for  neurobehavioral toxicity in rats:  a preliminary analysis
using  monosodium  glutamate,  calcium carrageenan  and  hydroxyurea.
Toxicol Appl Pharmacol 50:267-282. (As reported in Jensh 1983).

Warkany  J,  Beaudry PH, Hornstein  S. 1959.  Attempted  abortion with
aminopterin  (4-aminopteroylglutamic  acid).  AMA  Am   J  Dis  Child
97:274-281.

Weil CS.  1970. Selection of the valid number of sampling units and  a con-
sideration of their combination in  toxicological  studies involving reproduc-
tion, teratogenesis or carcinogenesis. Fd Cosmet Toxicol 8:117-182.

Werboff J, Gottlieb JS, Dembicki EL, Havlena J.  1961b. Postnatal effect
of  antidepressant  drugs  administered during  gestation.  Exptl  Neurol
3(6):542-555.

Werboff J, Gottlieb JS, Havlena J, Word TJ.  196la. Behavioral effects of
prenatal drug administration in the white rat. Pediatrics 27:318-324.

-------
                                  66

Werboff J, Havlena J.  1962. Postnatal behavioral effects of tranquilizers
administered to the gravid rat. Exptl Neurol 6:263-269.

Werboff J, Kesner R. 1963. Learning deficits of offspring after adminis-
tration of tranquillizing drugs to the mothers.  Nature 197:106-107.

Wilk AL,  Greenberg JH,  Horigan EA, Pratt RM,  Martin GR.  1980.
Detection of  teratogenic compounds using differentiating embryonic cells
in culture. In Vitro 16(4):269-276.

Willhite CC, Perm VH, Smith RP. 1981. Teratogenic effects of alphatic
nitriles. Teratology 23:317-323.

Willhite CC. 1983. Developmental toxicology of acetonitrile in the  Syrian
golden hamster. Teratology 27:313-325.

Willhite CC, Dawson MI, Williams KJ. 1984. Structure-activity relation-
ships of retinoids in developmental toxicology. Toxicol Appl  Pharmacol
74:397-410.

Wilson JG. 1966. Effects of acute and chronic treatment with actinomycin
D on pregnancy and the fetus in the rat. Harper Hosp Bull 24:109-118.

Wilson JG. 1969. Teratological and reproductive studies on nonhuman pri-
mates.  In: Methods for teratological studies in experimental animals and
man, II. Tokyo, Japan, pp. 6-31.

Wilson JG. 197la. Use of  primates in teratological research and testing.
In: Malformations congenitales des mammiferes, Paris, pp. 273-292.

Wilson JG. 1971b.  Use of rhesus monkeys in teratological studies. Fed
Proc 30(1): 104-109.

Wilson JG. 1975a. Critique of current methods for teratogenicity testing.
In:  Methods of environmental agents that produce congenital defects.
Shepard TH, Miller JR,  Marois  M, eds.  New York: American Elseiver,
pp. 29-48.

Wilson JG. 1975b. Reproduction  and teratogenesis: current  methods and
suggested improvements. J Assoc Off Anal Chem 58(4):657-667.

Wilson JG.  1977. Current  status of teratology  general  principles and
mechanisms derived from  animal  studies. In:  Handbook of teratology, Vol.
I. General principles in etiology.  Wilson JG, Fraser FC, eds. New York:
Plenum, pp. 47-74.

-------
                                  67
Wilson JG.  1978. Feasibilty and design of subhuman primate studies.  In:
Handbook of teratology, Vol. 4. Research procedures and data  analysis.
Wilson JG, Fraser FC, eds. New York: Plenum, pp. 255-273.

Wilson  JG.  1979.   The  evolution  of teratological  testing.  Teratology
29(2):205-212.

Wilson  JG.  Fradkin  R,  Hardman  AA.  1968.  Progress  report  on
teratological testing of drugs in rhesus monkeys. Teratology 1:223.

Wilson JG, Scott WJ, Ritter EJ, Fradkin R. 1979. Comparative distribu-
tion and embryotoxicity of methotrexate in pregnant rats and rhesus mon-
keys. Teratology 19:71-79.

WHO.  1967.  World Health  Organization. Principles for the testing of
drugs for teratogenicity. 20 pp.

Wolkowski-Tyl R,  Phelps  M, Davis JK. 1983.  Structural  teratogenicity
evaluation of methyl chloride in rats and mice  after inhalation  exposure.
Teratology 27:181-195.

Zimmerman, EF. 1975. Chemical structure and teratogenic mechanism of
action. In: Methods  for detection  of  environmental  agents that produce
congenital defects.  Shepard TH, Miller  JR, Marois  M,  eds.  New York:
American Elseiver, pp. 79-91.

-------
                   3. IMMUNOTOXICOLOGY

                         Mary Lou Daugherty

3.1 INTRODUCTION

   The science  of  immunology,  once limited  to  studies of host-defense
mechanisms  against  bacterial,  fungal,  viral,  protozoal,  and  helminth
diseases, has expanded into a  complex but well-studied  discipline that
includes allergy, tumor immunity, host-graft reactions, tolerance, and the
immune mechanisms  involved in the pathogenesis  of various occupational
disorders of the lung and skin.
   These  protective  responses of the body against  foreign  entities (or
antigens)  are  carried  out  by  the  immune  system,  a  complex  and
widespread organization of organs  and cells. Immune responses are of two
types: nonspecific (or innate) and specific (or acquired) (De Bruin 1976).
   Nonspecific immunity  is characterized by mechanisms for the disposal
of foreign  and  potentially  harmful  macromolecules,  microorganisms, or
metazoa which do not involve the recognition of antigen or the mounting
of an immune response. Nonspecific immunity is generalized and includes
the activities of phagocytic  cells, certain microbial inhibitors (lysozymes),
the interferon system, the complement system, and the kinin system.
   Specific immunity is characterized by (1) the  highly specific recogni-
tion  of and  response  to antigen by lymphocytes and (2) the induction of
immunological memory in long-lived  lymphocytes (memory cells),  which
are then capable of producing an enhanced immune response upon second
contact with an  antigen (Luster et al. 1982b). Specific immunity includes:
cell-mediated immunity, humoral immunity, and immunological tolerance.
The  major organs of the specific immune system in man are the thymus
and  bursa-equivalent, which are the  primary  lymphoid tissues,  and the
lymph nodes, the spleen, the bone marrow, the gut-associated lymphoid tis-
sue (GALT) (e.g. Peyer's  patches), and the bronchus-associated lymphoid
tissue (BALT),  which are the peripheral lymphoid tissues (Luster et al.
1982b). The major cells are the circulating lymphocytes. These cells  play a
central role in both  cellular and humoral immunity and and can be divided
into  two distinct populations, T-cells and B-cells, with different functions
and  properties  (Vos  1977).  Also important are the macrophages, both
fixed and free, which are involved  in and seem to be necessary for the ini-
tiation and the effector phases of humoral and cell-mediated immune reac-
tions  (Faith et  al.  1980).  Unlike the lymphocytes, however, the  macro-
phages do not clonally divide and are not antigen specific. Other cell types,
such as monocytes, granulocytes, mast cells, and reticuloendothelial cells,
facilitate in various  ways the expression  of the  immune response even
                                  69

-------
                                  70

though  they  are not equipped  ab initio to recognize and  interact with
antigen or to participate in the  induction of the immune response (Miller
1975).
   Because the immune system plays a critical role  in the protection of
man against  infectious  and neoplastic disease,  dysfunction of the system
can have  serious health consequences (Gardner et al. 1979);  it therefore
becomes important to be able to identify immunotoxic agents that may be
released into the workplace or environment.
   Although the  concept  of immunotoxicology is relatively new,  it has
been known for some time that  certain drugs can cause suppression of the
immune system.
   In  recent years, chemicals other than those  used clinically for immune
suppression have been recognized as agents capable of inducing alterations
in the  immune response. These chemicals are of environmental concern
and  include polychlorinated  biphenyls,  2,3,7,8-tetrachlorodibenzo-/>-dioxin,
hexachlorobenzene,  polybrominated  biphenyl,  chlorinated  dibenzofuran,
NO2,  SO2, benzo[a]pyrene,  tobacco  smoke,  heavy  metals,  organotins,
butylated  hydroxytoluene,  ortho-phenyl  phenol,  and  pesticides  such as
DDT, Dieldrin, organochlorines, and organophosphates (Vos 1977, Moore
1979,  Silkworth and Loose  1981, Luster et  al. 1982b). Information from a
variety of human studies (reviewed by Holt and Keast 1977) indicates that
increased prevalence of infections, and perhaps neoplastic diseases, result-
ing from  chronic exposure to air contaminants is associated with  impaired
immunological control  mechanisms. These  human data are supported by
the  results of animal  studies that demonstrate  the  impairment of host
resistance by  atmospheric  contaminants such  as pesticides, metals,  and
cigarette  smoke [reviewed  by Koller (1979a, 1979b), Treagan (1975), Vos
(1977), Moore and Faith (1976), Holt and Keast (1977), and Faith et al.
(1980)].  In addition, Loose et  al. (1977, 1978,  1979) demonstrated  that
polychlorinated biphenyl  and hexachlorobenzene were profoundly  immu-
nosuppressive  at clinically subtoxic levels and suggested that the evaluation
of immune parameters may prove to be a sensitive  indicator of general
toxicity.
   Procedures for  testing  the immunotoxicity of chemicals are  generally
not adequately represented in current routine toxicity testing schemes, and,
as a result, only a limited number of chemicals have  been shown to have
immunosuppressive properties in such tests. Vos (1977) suggested  that
these chemicals may be representative of a much  larger group  of  immuno-
toxic chemicals and advocated the inclusion of techniques for the evalua-
tion  of the function and morphology of the immune system in routine test
protocols.  In  order to accomplish this,  it  is  necessary to  incorporate
knowledge from the  expansive,  growing immunology discipline  into the
field of toxicology.
   The development of tests to  evaluate the immune response during toxi-
cological  testing is  gaining interest. Older in vivo test  methods, which can
be expensive and  cumbersome,  are undergoing revaluation,  and in vitro

-------
                                  71

correlates that  are  cheaper, easier  to handle, and more reproducible are
being developed.
   The primary purpose of this section is to review those tests of immune
function, immunocompetence in  particular, that appear to be suitable for
inclusion in routine  toxicity testing batteries. In addition to tests for immu-
nocompetence, tests for the induction of  clinical allergy (those diseases in
which  hypersensitivity  plays a part, such as  contact hypersensitivity, hay
fever, and asthma)  will be mentioned briefly.  Recommendations for tier
testing schemes for immunotoxicity will be presented.
   An in-depth review of the organ systems and the complex cellular and
humoral interactions that are necessary  for,  and ultimately result in, an
immune response is not within the  scope  of this document. A glossary of
immunological terms, prepared for the nonimmunologist, is provided at the
end of the summary. For a more comprehensive survey of basic immunol-
ogy, the references listed in Section  3.8 are recommended.
3.2 TESTS TO DETECT ALTERATIONS IN THE IMMUNE RESPONSE
   Chemicals capable of inducing alterations in the immune response may
alter both  specific  and nonspecific responses,  and the response may be
enhanced  or suppressed,  depending on  a  variety of  circumstances that
include  chemical and  dosage level and  the species and  age of the  test
animal (Luster and Faith 1979). For example, inorganic metals have been
shown to affect primarily humoral immune responses (Koller  1979a) while
halogenated aromatic  hydrocarbons affect  cell-mediated  immunity; low-
level exposure to certain chemicals may lead to immunoenhancement while
higher  levels are immunosuppressive (Koller et al.  1977,  Holt and Keast
1977);  hexachlorobenzene is reported to produce opposite  responses in the
rat and the mouse (Moore 1979); and the developing immune system may
be more sensitive to the effects of environmental chemicals than the adult
immune system (Vos and Moore 1974, Moore and Faith 1976).
   It has  generally been thought that no agent selectively suppresses the
antibody-forming mechanism  but  that  any agent  that  suppresses  cell
proliferation  or interferes  with protein  or nucleic  acid  formation will,
under  proper circumstances,  suppress  the antigen-antibody mechanism
(Loomis 1974). However,  there  is evidence to suggest  that this  is not
necessarily the case, but that the various cell types that participate in the
immune response can be specifically depressed (Nicolin  et al. 1980).
   When  a  chemical  compound  is  suspected  of altering  the immune
response, both the  cellular and humoral systems  must be tested (Koller
1979a). Tests of cell-mediated immunity  generally  measure  the  specific
activities of the T-lymphocyte populations and may be performed using in
vivo  or  in vitro techniques. Tests of humoral immunity generally measure
cellular synthesis  of  antibody  and  circulating  antibody,  primarily  B-
lymphocyte properties.  There are also tests  to measure macrophage proper-
ties and other indirect parameters of the immune system. Thus, tests are

-------
                                   72

available to help determine which  of the  following steps in the immune
response are altered by a particular chemical (Vos 1977):

 1. antigen processing (phagocytosis),
 2. antigen recognition by lymphocytes,
 3. adaptive  phase   (differentiation  and  proliferation  of  T-  and  B-
    lymphocytes),  or

 4. effector phase.

    Representative  tests, primarily the original techniques or the modifica-
tions  of those techniques that  seem  suitable for toxicity screening,  are
listed in Table 3.1 and are reviewed in the following sections.
    As in any discussion  of toxicity testing procedures, the importance of
proper  test conditions should  be recognized.  Factors  that may affect  the
outcome of the tests are  listed in Table  3.2 and  are discussed in detail by
Doull (1975).

3.2.1 Assessment of Cell-Mediated Immunity

    The main categories  of  cell-mediated immunity include: classical cell-
mediated protective immunity, which is mainly effective against protozoal,
viral, fungal, and  some bacterial infections;  delayed  hypersensitivity skin
reactions to  extracts or whole suspensions of organisms,  also identified as
tuberculin hypersensitivity; chemical contact sensitivity; allograft rejection;
immunological surveillance to  tumors; and certain organ-specific autoaller-
gic diseases, such as thyroiditis, encephalomyelitis (following rabies vacci-
nation), adrenalitis, and orchitis  (Turk 1975, Faith et al. 1980).
    T-cells are  required  for the induction of the cell-mediated immune
response. When  stimulated  by antigen,  T-cells form  a pool  of sensitized,
antigen-specific lymphocytes  which  then  can function  as  memory cells
(which  produce the secondary  immune response), effector cells, killer cells,
helper cells, or suppressor cells (Faith et al.  1980). T-cells do not produce
antibody (Roller 1979a).
    Effector  cells,  in  delayed-type  hypersensitivity,   produce biologically
active products  called  lymphokines.  Lymphokines are soluble  nonantibody
products of  lymphocyte  activation  by antigen  and   mitogens which  are
thought to act  as molecular mediators  of  cellular immune  responses
(Merely et al.  1978). Lymphokines  exhibit the following  characteristics:
(1) produce  increased vascular  permeability  following intradermal  injec-
tion; (2) increase  tritiated thymidine uptake  by lymphocytes in culture;
and (3) inhibit macrophage migration in  vitro. Lymphokines retain biologi-
cal activity after removal of the antigen that stimulated  them. They may
also be  involved in  both expression and regulation of lymphoid cell activity,
and they may represent a pathway  for the expression of T-cell function.
Some lymphokines, such as the migration inhibition factor, the chemotac-
tic  factor,  and  the macrophage-activating factor, influence macrophage

-------
                                           73
                           Table 3.1 Tests for Immunotoxicity
Assessment of cell-mediated immunity
    In vivo tests

         Tests for delayed hypersensitivity
             Cutaneous hypersensitivity to tuberculin
             Radiometric ear test
             Footpad assay
             Allograft rejection
         Graft-vs-host reaction

         In vitro tests
             Transformation of lymphocytes by mitogens
             Mixed lymphocyte culture
             Macrophage migration inhibition factor (MIF)
             Macrophage aggregation
             Leukocyte adherence inhibition (LAI)

Assessment of humoral immunity

    Assays for local production of antibody
         Plaque-forming assays
         Immunofluorescence
    Measurement of circulation antibody and immunoglobulins
         Precipitation
         Hemagglutination and hemolysis
         Passive hemagglutination
         Enzyme-linked immunosorbent assay (ELISA)
    Other tests of B-cell function
         EAC rosette technique
         Transformation of lymphocytes by B-cell mitogens

Assessment of indirect parameters  of immunity
    Macrophage functions
         Ingestion
         Intracellular killing
    Host resistance to infection

Tests to detect the sensitizing potential of chemicals

    Skin sensitization
         Draize test
         Freund's complete adjuvant test
         Guinea pig maximization test
         Split adjuvant technique
         Buhler test
         Open epicutaneous test
    Respiratory sensitization

-------
                                   74
        Table 3.2 A Classification of Toxicity-Influencing Factors


 1.  Factors Related to the Toxic Agent
    Chemical composition (pH, choice of anion, etc.)
    Physical characteristics (particle size, method of formulation, etc.)
    Presence of impurities or contaminants
    Stability and storage characteristics of toxic agents
    Solubility of the toxic agent in biologic fluids
    Choice of the vehicle
    Presence  of  excipients:  adjuvants,  emulsifiers, surfactants,  binding
         agents,  coating agents, coloring agents, flavoring  agents, preser-
         vatives,  antioxidants,  and  other  intentional and  nonintentional
         additives

 2. Factors Related to the Exposure Situation
    Dose, concentration, and volume of administration
    Route, rate, and site of administration
    Duration and frequency of exposure
    Time of administration (time of day, season of the year, etc.)
 3. Inherent Factors Related to the Subject

    Species and strain (taxanomic classification)
    Genetic status (littermate, siblings, multigeneration effects, etc.)
    Immunologic status
    Nutritional status (diet factors, state of hydration, etc.)
    Hormonal status (pregnancy, etc.)
    Age, sex, body weight, and maturity
    Central nervous system status  (activity, crowding, handling, presence
         of other  species, etc.)
    Presence of disease or specific organ pathology
 4.  Environmental Factors Related to the Subject
    Temperature and humidity
    Barometric pressure (hyper- and hypobaric effects)
    Ambient atmospheric composition
    Light and other forms of radiation
    Housing and  caging effects
    Social factors
    Chemical factors

Source: Adapted from Doull (1975).

-------
                                    75

functions (Faith et al. 1980). The properties of lymphokines or products of
activated lymphocytes  (PALs) have been summarized by  Bloom et al
(1974) in Table 3.3.
   Killer cells are cytotoxic and participate in the rejection of tumor  cells
and transplants.
   Helper cells  interact with B  cells in  antibody response  against many
antigens, while suppressor T-cells function to suppress the functions of B-
cells or effector T-cells. [There is also evidence for the existence of B-cells
that modulate T-cells (Turk 1975).]
   The induction and manifestation of the cell-mediated immune response,
illustrated in Figure  3.1, proceeds as follows (Park and Good 1974, Miller
1975, Luster et al. 1982b):

  1. Antigen interacts with T-cells, usually outside the lymphoid tissues (at
    the site of antigen accumulation or within a graft).
  2. Lymph nodes draining the site of sensitization  become  enlarged, and
    on  day two  of the  response the  small T-lymphocytes in the thymus-
    dependent areas differentiate  to  large pyroninophilic  cells  and divide
    to  produce  a progeny  of effector cells  (small lymphocytes), which
    enter the circulation after 3 to 4 days. The onset of delayed hypersen-
    sitivity coincides roughly with the appearance of these cells in the cir-
    culation.

  3. The small sensitized  lymphocytes (effector cells) now in the  circula-
    tion come in contact  with the antigen and are  stimulated  to produce
    lymphokines  or  effector  molecules,  which  recruit nonspecific  cells,
    such as monocytes,  to the site of the reaction.
  4. Those monocytes that  accumulate are  then transformed  to macro-
    phages, which through the release of  lysozymal  enzymes produce  vari-
    ous effects   leading  to  the  ultimate   expression  of  the  cellular
    response—a  delayed  hypersensitivity lesion, a graft rejection,   or  a
    microbiocidal effect.

   In general, in vivo  assays of  cell-mediated immunity  mainly measure
the effector mechanism, while in vitro assays  may  measure cellular com-
ponents  (Luster  et al.  1982a). In vivo and in vitro assays that are com-
monly employed  to  measure T-lymphocyte  functions will  be described.
These will be discussed with respect to preferred test  species,  advantages
and  disadvantages, usefulness  in  detecting the  effects of environmental
chemicals,  and, when possible, applicability of the test to particular chemi-
cal classes.

3.2.1.1 In vivo tests
   Basic in vivo  procedures for  evaluating  the  cell-mediated immune
response include tests of  delayed hypersensitivity, allograft  rejection, and
graft-vs-host reactions.  These are well-established and valuable procedures,

-------
                                           76
                  Table 3.3 Products of Activated Lymphocytes (PALs) In Vitro
A.  Affecting Macrophages
    Migration Inhibitory Factor (MIF)
    Macrophage Aggregation Factor
    Macrophage Spreading Factor

    Macrophage Activating Factor (MAP)
    Macrophage Chemotactic Factor

    Cytophilic Antibodies
B.  Affecting Lymphocytes
    Blastogenic or Mitogenic Factors (BF, MF)

    Thymus Replacing Factor (TRF) or
       Helper Factor (HF)
    Suppressor Factor

    Transfer Factor (TF)

C.  Affecting Neutrophils
    Leucocyte Inhibitory Factor (LIF)
    Leucocyte Chemotactic Factor

D.  Affecting Eosinophils
    Eosinophil Chemotactic Factor (ECF)

    Eosinophil Promoting  Factor
E.  Affecting Other Cell Types
    Lymphotoxin (LT)
    Proliferation Inhibitory Factor (PIF)
    Cloning Inhibitory Factor (C1IF)
    Osteoclast Activating Factor (OAF)

    Colony Stimulating Factor (CSF)
     Interferon (IF)
Inhibits migration of normal macrophages
Agglutinates macrophages in suspension
Increases adherence and surface area of
  macrophages
Increases glucose-C-1 oxidation and phagocytosis
Causes macrophages to migrate along gradient
  through micropore filter
Confer some specific reactivity to antigens

Induce blast cell transformation and thymidine
  uptake by normal lymphocytes
Permits differentiation of B cells to
  Ab secreting cells
Inhibits activation of, or antibody production
  by, B cells
Converts nonreactive cells to cells capable
  of reacting to specific antigens

Inhibits migration of polymorphs
Causes polymorphs to migrate through
  micropore filter

Together with Ab-Ag complex, causes migration
  through micropore filter
Increases migration from agarose drops

Cytotoxic, especially for L-cells
Inhibits proliferation without killing cells

Causes release of 45Ca from embryonic bone
  cultures
Stimulates differentiation of bone marrow
  stem cells into granulocytic and/or
  monocytic cells
Renders cells resistant to virus infection
Source: Adapted from Bloom et al. 1974.

-------
                                                                          ORNL-DWG 85-10739
                                          /^^•W"-^
                                            D
                                            i
                                  stimulauon
                                               -lymphocytary factors -
          lymphoblast — »8
                   sensibihzed
                   lymphocyte
stimulation T-iymphocytes
celldivision
i ce
1 immur
©i
0
rthymus 	 » $ Q 	
T- lymphocytes ^~
r> ® °
iwrnphrsi^ U ,»v ^
1 activation
llular versus (A) '
le reaction |
1
I
1
1
] 1 01
contact 	 	 	 >^\ ^ /^\
. y 1 1 HJMUIU ^A ^ 	 • 	 	 ^ *- - — — -. 	 , , * 	 , ; .
.stem' cells 0 O stimijlot'°n V V c* — x X_y
, ' _,'f\ (A) RNA (P) coated phagocyte
                                    lyrnphOld  1
                                                                               (P)
                                                                                         (A)
                   8-iymphocytes
1'  membrane-bound,
x~ __m_odi^ied_gntige_n_
                                        humoral
                                      immune reaction
                                  activation   |
                                B-lymphocytes    T- lymphocyte 7
                                             I      i        —s,   synthesis o<
                                            & 	»	» (• ^
                                                          V^_y immunoglobulm
                                                        plasma cell
                                                                                         (A)
                                                 (developed from B-lymphocyte)
                                                                             Qntlboclies
   Figure  3.1.  Hypothetical  mechanisms  for  cellular  and  humoral  immune  reactions.
(Source: De Bruin  1976. Used with permission of the publisher.)

-------
                                  78

but they are generally cumbersome and time consuming and require large
numbers of animals. Newer tests have been introduced, some utilizing iso-
topes, that are quantitative and  more sensitive.  Some of these have  been
recommended for inclusion in toxicity screening protocols.
   Tests of delayed-type faypersensitivity. Cell-mediated immunity is often
measured  by  delayed  hypersensitivity  reactions  (Koller  1979a).  The
evaluation  of delayed hypersensitivity is generally based on dermal reac-
tions  that  lead to erythema, induration, and increase in skin  thickness.
More recent tests have been introduced that are based on the radioactive
labeling of  bone  marrow precursors  of monocytes  that subsequently
migrate into sites of inflammation (Lefford 1974) or on the extravasation
of  human  serum albumin into the area  of  hypersensitivity  response
(Paranjpe  and Boone  1972). All of these tests are important to the assess-
ment of  immunocompetence.  Examples  are described in the following
paragraphs.

Skin Reactions (Useful in Immunotoxicological Studies)

    An  important  test for measuring cell-mediated immunity is the delayed
cutaneous  hypersensitivity reaction to tuberculin as described by Flax and
Waksman  (1962). In the  original  experiment rats  were  injected  in  one
hind footpad with a suspension of heat-killed  tubercle bacilli and then  were
challenged with a purified protein derivative of tuberculin (FDD), which
was injected intradermally  into the  flank. The animals were observed rou-
tinely at 4 to 6, 24, 48, and 72 h, and the diameter  and degree of indura-
tion of the skin reactions  were determined. The reaction was maximal at
24 to 48 h.
    In addition to the rat, the guinea pig is often used in this system, and
in the past was considered to be the species of choice.
    The usefulness of this skin test in the evaluation of immunotoxicologic
properties  of chemicals, as well as  species  specificity of the test,  was
demonstrated by  Vos et al. (1973) and by Vos et al. (1979b).  The  pro-
cedure  used was similar to that of Flax and Waksman (1962); for a higher
degree  of  quantitation, the thickness  of the  reaction was measured as an
end point, in addition to diameter and degree of induration. In the study
by Vos et al.  (1973) the  skin test  reaction was evaluated in guinea  pigs
and rats  following  administration  of  2,3,7,8-tetrachlorodibenzo-/?-dioxin
(TCDD). Guinea  pigs were dosed for 8 weeks with 0.008, 0.04, 0.2, or 1.0
jig/kg (body weight). Skin  reactions were significantly reduced at the 0.2-
and 0.04-jtg/kg levels.  All guinea  pigs treated with  1.0  Mg/kg died or
displayed overt signs  of toxicity. Thymus atrophy and lymphopenia  were
observed. The rats were dosed for 6 weeks with higher doses of 0, 0.2, 1.0,
and 5.0 ng/kg, and, although animals at the 5-/*g/kg level had lower body,
thymus, and adrenal weights, the skin reactions were not altered.
    In the  second  study, Vos et  al. (1979b) included the skin test in a bat-
tery of immunological assays to  test the effects of  hexachlorobenzene in

-------
                                   79

the rat. The cell-mediated immune response was not altered by the chemi-
cal, but the antibody response was stimulated.
   The cutaneous hypersensitivity reaction to tuberculin is not a very sen-
sitive test for measuring cell-mediated immunity (Vos 1977), and measure-
ments of  the  reaction  are extremely  subjective and  vary  considerably
between observers (Carter and Bazin 1980). However, the skin test offers
several  test sites per animal, is quick and easy to perform (Crowle 1975),
and  can yield reproducible results with  little variation if careful measure-
ments are made. Lefford (1974) introduced a  new test using radiolabeling
techniques, which is more  quantitative  and less subjective  than the more
traditional skin test. This procedure  is described in the following section.

Radiometric Ear Test

   The Lefford  (1974) technique for measuring delayed-type hypersensi-
tivity (DTH) is based on the radioactive labeling of bone marrow precur-
sors  of monocytes,  which subsequently migrate  to  sites of inflammation
and  accumulate  in  DTH reactions. Rats  were sensitized intravenously or
subcutaneously with viable bacillus  Calmette-Guerin (BCG), then  injected
subcutaneously  with thymidine  methyl-3H. Twenty-four hours later  the
animals were  injected  in  the  left ear with  tuberculin purified  protein
derivative (PPD). The right ear, injected with diluent, served  as a control.
Twenty-four hours after the injection of antigen the central portion of each
ear was removed, and radioactivity  of the tissue was measured in  a liquid
scintillation spectrometer. Counts of the test ear were then  compared with
those of the control ear, and it was possible to calculate the left ear/right
ear ratio  as the parameter of delayed hypersensitivity.
   Mice  and guinea pigs may also be tested by  this method, and  various
protein  antigens, other  than PPD,  can  be used;  however, PPD is  recom-
mended for rats and  guinea  pigs  while other protein antigens, including
bovine  serum  albumin,  are  generally  used in mice  (Luster  and Faith
1979).
   Moore and  Faith (1976)  introduced a modification  of  the  Lefford
method  in which rats  were sensitized to oxozalone by skin-painting the
ears. Tritiated  thymidine was injected intraperitoneally 10 days later,  and
24 h after thymidine injection a challenging dose of oxozalone in  olive oil
was  painted on one ear. Standard procedures were then  used  to determine
the  ratio of  radioactivity in  the treated  and untreated ears. Moore  and
Faith (1976),  by using this  method, were able to detect  suppression of
cell-mediated immunity  in Fischer rat pups exposed to TCDD on gestation
day  18  and on postnatal days 0, 7, and 14.
   In a later study, Fraker (1980) used a similar technique to  test the
effects  of polybrominated  biphenyls (PBB)  in  mice.  The animals  (fed
PBB) were sensitized percutaneously and  challenged on one ear lobe with
dinitrofluorobenzene. After intravenous  injection of 125I-deoxyuridine, the
ratio of  radioisotope  incorporated  in  the challenged versus  the  unchal-
lenged  ear  lobe  served  as  a measure  of  T-cell-dependent  delayed

-------
                                  80
hypersensitivity.  No  effect  was  seen  on this response. This  was  of
particular interest because, in the same study, the IgG and IgM plaque-
forming cells were impaired  by PBB, indicating that both helper T- and
B-cells were affected. T-cells involved in delayed-type hypersensitiviy and
those involved  in  humoral  immunity  are of  the  same  phenotype  and
probably follow  the same developmental pathway;  thus  the  differential
effects of PBB were thought to be  due to differences in  susceptibility to
the chemical of the various subclasses of T-cells (Fraker 1980).
    There are certain disadvantages  to this method:  first, each ear can be
tested  only once, so each animal can receive  a maximum of two  tests;
furthermore, if the left  ear/right ear ratio is calculated the animal can be
used only once. Advantages of the technique  are objectivity, reproducibil-
ity, and sensitivity, and  these far outweigh the disadvantages. It appears to
be  particularly useful as a general screen for  chemical effects because the
participation of many cell types is required for a normal reaction.

Footpad Assay

    The footpad assay is based on the  observation  that 125I-human serum
albumin (125I-HSA) will extravasate into the  edematous area produced by
a  delayed  hypersensitivity response  (Paranjpe  and Boone 1972). In the
procedure described by Kauffmann et al.  (1982),  Sanders et al. (1982),
and Munson et al. (1982), recommended  for  testing the effects of chemi-
cals on cell-mediated immunity, mice are immunized in the left hind foot-
pad with sheep red blood  cells (SRBC). Four  days later the mice are chal-
lenged in the same footpad  with SRBC and 17 hours later are injected
intravenously  with 125I-HSA. The  animals are killed and both  feet are
removed and measured for radioactivity in a  gamma  scintillation counter.
Control animals are not sensitized but are challenged in order to determine
nonspecific swelling. The stimulation index (S.I.) is calculated:

     _  Left footpad -  _   Left footpad unsensitized
        Right  footpad     Right footpad unsensitized

    Minor aspects  of the  technique,  such as the  number  and route of
immunizing injections,  the timing of the challenge,  and timing of the iso-
tope  injection,  may  vary between  laboratories.  Also, footpad  swelling
instead of extravasation  of  isotope  may  be  the end  point of the  assay
(Muller et al.  1977). Mice are generally used  in this assay. However, it
has been noted that the strain of mouse may  influence detectability of the
response (Crowle 1975).
    Muller et  al.  (1977) used  the footpad assay  to demonstrate the
suppressive effects of lead on DTH  in mice.  Alterations of both primary
and secondary  responses to SRBC were correlated with  the concentrations
of lead in the  blood. In  another study, DTH,  assayed by footpad injection,
was also shown to be suppressed in  male mice by trichloromethane and in
males  and females by bromodichloromethane  (but not by  bromoform and

-------
                                  81

dibromochloromethane)  (Schuller  et  al.  1978).  Smith  et  al.  (1978)
reported a similar inhibition of the footpad response with poly chlorinated
biphenyls (Aroclor 1254).  Sanders et al. (1982), using the footpad assay,
observed a suppression of the stimulation index by  subchronic administra-
tion of trichloroethylene in female (but not male) mice.
    According to Dean  et  al. (1979a),  the footpad assay is probably the
best single  parameter for immunologic competence because  it measures
both  the  recognition and effector  phases of the  cell-mediated immune
response and requires the cooperation of the macrophage/monocyte series.
    The isotope system  is  sensitive  and  quantitative,  and the reaction
appears to be of the pure delayed type  without the  humoral component of
the immediate anaphylactic  or  arthus type, which  often accompanies the
delayed hypersensitivity reaction (Paranjpe and Boone 1972).
    The assay can be used to determine the effects of chemicals on the
ability  of  the host  to  be immunized  or  to  respond  to challenge  after
immunization (Dean et al.  1979a).
    Allograft Rejection.  Skin grafting is a useful technique that has been
applied to the study of various biological problems. The assay is dependent
upon  the  ability of  an  experimental host  animal to reject a tissue trans-
planted  from  a  donor  that  differs  from  the host  at an  important
histocompatibility  locus. Allograft  rejection time is used as the  end  point
for detecting  alterations in the cell-mediated immune response  of experi-
mental  animals.
    Standard techniques for skin grafting in rabbits, guinea pigs, and mice
were described by Billingham and Medawar (1951). These classic methods
can be  laborious and time consuming and  are not  practical for situations
that require large numbers of grafts.
    A rapid method  of grafting skin on tails of mice  was described by
Baily and Usama (1962). This widely used procedure consists of slicing an
approximately 5- by 2-mm section  of skin  from the  tail and transferring it
to a raw area of the host tail from  which the reciprocal graft has just been
removed. Grafts are placed on the host  so that the direction of hair growth
is opposite to that of the  host in order that  "takes" can be distinguished
from  possible host replacement. The grafts adhere well and do not require
extraneous pressure  or  dressing. Also,  they are  protected by,  and can be
observed through, a length of glass tubing fitted over the tail.
    This technique can also be successfully performed on rats. Using  allo-
graft  rejection as one of the parameters, Vos and Moore (1974) detected
impairment of cell-mediated immunity in the offspring of rats treated, dur-
ing gestation and postnatally, with  TCDD.  In a more recent study, Vos et
al.  (1979b) incorporated allograft rejection into a battery of function  tests
used to ascertain the effects of hexachlorobenzene on the immune system
of the rat. The chemical had no effect on the in vivo cell-mediated immune
response but did enhance humoral immunity.
    The tail skin-grafting technique is simple and rapid, and  a  particular
advantage is that the graft is in plain view at all times; however, the  assay

-------
                                  82

is not applicable  to studies requiring the use of large grafts or host areas
other than skin, and it is not particularly quantitative.
   Graft-vs-host reaction. In the graft-vs-host  reaction (a classic assay of
cell-mediated immunity), immunocompetent cells are injected into hybrids
of parental strains of animals that differ from each other at the major his-
tocompatibility locus. The donor cells do not possess any antigens foreign
to the host and thus  are tolerated by the hybrids.  However, the immuno-
competent donor  cells do react against the foreign antigen in the host and,
therefore, induce  a host reaction that can be measured by several methods,
including the spleen weight assay and the renal enlargement assay. These
methods have been reviewed by Ford (1978). In immunotoxicology studies
the test compound is given to the donor animals.
   A particularly useful technique was described by Ford et al. (1970) in
which  donor spleen or thoracic  duct lymphocytes were injected into the
feet  of  young F[ hybrid  rats. The recipients were  killed seven days later,
and  the weights  of  their popliteal lymph nodes were  plotted against the
dose of cells injected,  the degree  of enlargement of the draining lymph
nodes being a measure of the graft-vs-host activity of the donor cells.
   Differences in the effects of various chemicals on the immune response
have been demonstrated with  the graft-vs-host popliteal lymph node assay.
For example, TCDD suppressed the  response in rats and mice (Vos et  al.
1973, Vos and Moore  1974), while  Aroclor  1016 and  hexachlorobenzene
enhanced the response.
   The popliteal lymph node  assay is a sensitive and quantitative measure
of cell-mediated  immunity.  Its in vitro analog, the mixed lymphocyte cul-
ture reaction, is discussed in a later section.

3.2.1.2  In vitro tests

   In vitro tests  of cell-mediated immunity were initially developed for the
purpose of analyzing the mechanisms of the response (Bloom et al. 1974),
but it  soon became  evident that these  methods for testing  cell-mediated
immunity also offered  possibilities, at  the  clinical  level,  to objectively
assess  the status or  level of  delayed hypersensitivity without  risk to the
patient.  These assays are also useful in the detection of alterations in the
immune response of  experimental animals following exposure to potential
immunotoxicants. Those tests that seem most suitable for this purpose are
described.
   Transformation of lymphocytes  by mitogens.  The ability of  cultured
lymphocytes to undergo transformation  and to incorporate labeled thymi-
dine (3H-TdR) into DNA following nonspecific stimulation  with mitogens
is the  basis of one of the most  valuable tools of  immunotoxicology. The
mitogenic response of lymphocytes is a parameter of the adaptive phase of
the immune response (Vos et  al. 1979c). Some mitogens selectively stimu-
late  the cellular  immune (T-cell) components while others stimulate the
humoral  immune (B-cell)  components.  Phytohemagglutinin  (PHA)  and
concanavalin A (Con A) are the  T-cell mitogens of choice  (Luster  and

-------
                                  83

Faith 1979), while lipopolysaccharides seem to stimulate primarily B-cells
(Smith  1972). On the other hand, pokeweed mitogen stimulates both B-
and T-cells and can be used to assay peripheral blood lymphocytes that do
not respond to other  mitogens (Luster and Faith  1979). The response of
lymphoid cells to Con A and PHA is discussed in greater detail by Stobo
et al. (1972).
   Thorpe and Knight (1974) describe a microplate culture technique for
lymphocyte  transformation and define optimal conditions  for performing
the assay. Essentially, various  concentrations of mouse mesenteric lymph
node cells were cultured in microplate wells in the presence of mitogens.
The plates were incubated at 37°C in 5% CO2. Twenty-four hours prior to
harvesting, 3H-TdR was added to each culture. Cultures were usually har-
vested  after 72 h and then processed  for scintillation  counting to measure
3H-TdR incorporation. An  alternative test that  measures mitogen- (and
alloantigen-) induced  lymphocyte proliferation in a one-way  mixed leuko-
cyte culture was  described by  Dean et al. (1979a), who advocate its use
for defining changes in host immunocompetence.
   Examples of the use of mitogen stimulation in immunotoxicology stu-
dies  are numerous. Mitogen stimulation  is  usually performed on  mouse
lymphocytes, because  Fischer rats, which are commonly used in toxicology
studies, respond  poorly  to the usual mitogens (Tada et  al.  1974, as
reported in Luster and Faith 1979). The response to T-cell mitogens corre-
lates well with cell-mediated  immunity and can be applied to  cells that
have been exposed to  chemicals in vivo or in vitro. Vos and Moore (1974)
tested the mitogen response of lymphoid cells from mice and  rats that had
been exposed to  TCDD at different  ages.  In rats whose  mothers were
exposed during gestation and  postnatally, the PHA response was reduced
on a cell-for-cell  basis, but both the  PHA and Con A responses recover-
able from the whole  thymus were depressed. In mice treated with TCDD
at 1  month of age, reduced responsiveness of spleen cells to PHA was seen
only at a clearly toxic dose level; in 4-month-old mice PHA responsiveness
was  not depressed. The thymus was  the main organ affected by TCDD,
and  other  parameters of cell-mediated immunity were  depressed  in the
very young animals.  The  data suggest that TCDD-suppression  of cell-
mediated  immunity may be age related.  In a later study, Sharma  et al.
(1978)  similarly reported that PHA  and pokeweed mitogen responses in
mice and rabbits  were suppressed with toxic levels of TCDD and, further-
more,  that  low  levels of the  chemical caused enhancement of immune
reponsiveness similar to that produced by antigenic substances.
   A related chemical, 2,3,7,8-tetrachlorodibenzofuran (TCDF), also pro-
duced a moderate suppression of the lymphoproliferative response (to PHA
and lipopolysaccharides, but not to Con A)  of splenic lymphocytes  taken
from guinea pigs treated orally with the chemical (Luster et al. 1979).
   Other industrial chemicals that have been shown to alter the  mitogenic
response of  lymphocytes include  polybrominated biphenyl  (Firemaster
FF-1),  which decreased the responsiveness  of spleen cells from  rats and

-------
                                 84

mice  (Moore et  al.  1978);  dioxane,  which  suppressed  mouse  T-cell
responses while augmenting B-cell responses (Thurman et al. 1978); and
toluene  di-isocyanate, which stimulated the response  of both human and
mouse lymphocytes (Thurman et al. 1978).
   Chronic exposure of mice to fresh cigarette smoke resulted in a marked
depression of the PHA response of lymphocytes  taken from the regional
lymph nodes  of the respiratory  system and from the blood (Thomas  et al.
1974a).
   The  sensitivity of the mitogen response to the effects  of  heavy metals
has been  tested  by several  investigators. In  one study, Gaworski and
Sharma (1978) observed  depressed responses to PHA  and pokeweed  mito-
gen in splenic lymphocytes taken from mice  fed high, but subtoxic,  levels
of lead, cadmium, or mercury. Similar  effects were observed in rat lym-
phocyte responses to PHA and Con A following chronic prenatal and post-
natal  exposure to low  levels  of lead (Faith  et al. 1979). However, in a
third  study (Roller et  al. 1979), lead and cadmium  did  not significantly
affect the response of mouse lymphocyte proliferation  by Con A. Koller et
al. (1979) suggested that the conflicting results of the two mouse studies
could be due to variation in  amount and length  of exposure to the com-
pounds, strain  differences, or differences in the  response of lymphocyte
subpopulations to PHA and Con A.
   Lymphocyte transformation  is a  semiquantitative and reproducible
technique, and the microplate method is  practical and labor saving (Vos
1977; Luster and Faith  1979).  The procedure has been  adapted for  use
with human cells in clinical laboratories to determine the immune status of
patients (Thurman et  al.  1978).  Luster and  Faith (1979) point out, how-
ever, that it does not measure effector function nor does it represent a nor-
mal response.
   Mixed Lymphocyte Culture.  The mixed lymphocyte  culture  (MLC)
reaction is an in vitro  analog of the allograft rejection response (Thurman
et al. 1978, Luster and  Faith  1979). Similar to the  mitogen stimulation
assay, lymphocytes in  culture are normally stimulated by allogeneic lym-
phocytes to  undergo blastogenesis, and  lymphoproliferation is then meas-
ured by the amount of tritiated  thymidine incorporated by the test cells.
The  response, probably a measure of both  B-  and T-cell proliferation, is
stimulated by  cell surface  antigens  and  requires macrophages  for  the
response to be initiated (Thurman et al.  1978).
   The  MLC assay has not been used extensively in immunotoxicology
studies.  However,  the system was employed by Thurman  et  al.  (1978) as
one of several parameters examined in  the evaluation  of the immunotoxi-
cologic effects of toluene  diisocyanate, dioxane, and vinyl chloride in mice.
   Their procedure was  as follows: spleens were removed from untreated
CBA/J  and  C57B1/6J mice and cell  suspensions were prepared; blasto-
genesis  of these cells was blocked with mitomycin. For both control  and
test  groups, responding  (unblocked) thymus, spleen,  or  lymph node  cell
suspensions were prepared from  CBA/J mice that had been treated with

-------
                                  85

the different chemicals, and those cells were cultured in microtiter plates
for 72 or 96  h.  Control  cultures received  0.1  mL of syngeneic, blocked
CBA/J lymphocytes,  and the test  cultures  received allogeneic, blocked
C57B1/6J lymphocytes. The cultures were pulse-labeled and  prepared  for
scintillation  counting  to  measure  lymphoproliferation  of  the  unblocked
CBA/J cells.
   The proliferation  response  of the  spleen, thymus, and  lymphoid cells
was generally inhibited by concentrations of 0.001  to 0.05% toluene diiso-
cyanate but apparently was stimulated by 0.1% toluene diisocyanate. The
response was  lowered slightly  by 10% dioxane and was not affected  by
vinyl chloride  (500 ppm for 7 h).
   The assay is fairly simple to perform; spleen, thymus, and lymph node
cells can be tested; and treatment with chemicals can be in vitro as well as
in vivo,  which qualifies  the test for use  with  human cells.   However,
Oehler et al. (1977) found that it is sometimes difficult  to demonstrate an
MLC response in spleens from rats because of  the presence of suppressor
macrophages.  Therefore,  it  has  been  suggested that blood  lymphocytes
from  most  species  can be substituted for  spleen  lymphocytes,  following
Hypaque-Ficol density gradient centrifugation (Luster and Faith 1979).
   Macrophage Migration Inhibition  Factor (MIF). Bloom  and Bennett
(1966)  observed  that sensitized peritoneal  lymphocytes, upon interaction
with  specific  antigen  in  vitro,  elaborated into  the medium a  soluble
material that could  inhibit  migration of normal  exudate  cells.  In  their
experiment, peritoneal macrophages, from guinea pigs sensitized to tuber-
culin  with Freund's adjuvant, were drawn into capillary tubes and centri-
fuged. The capillary tubes were cut  at the cell-medium interface and were
placed in special  chambers containing  various supernatants from  sensitized
peritoneal lymphocyte cultures. Macrophages, exposed to the supernatants
from  cultures of lymphocytes  growing in medium  alone, migrated out of
the tube onto a specially placed coverslip. Macrophages  that were exposed
to the supernatant from cultures of lymphocytes growing in purified pro-
tein derivative (PPD) did  not migrate.  The soluble material elaborated into
the supernatant by the PPD-stimulated lymphocytes has been identified as
the lymphokine MIF. This type of cellular interaction can be detected in
experiments typified by the following (David and David 1971):

  1. In  the  simplest  migration assay,  sensitized peritoneal exudate cells
    (macrophages  and  lymphocytes)  are   assayed  in  a  capillary  tube
    against the appropriate antigen.

  2. In another assay, lymphocytes from lymph nodes or other organs  are
    mixed with normal peritoneal exudate cells  in order to detect sensitive
    cells in  the various organs.
  3. In the "indirect" assay, lymphocytes are incubated  with antigen, and
    the cell-free  supernatants  are assayed for MIF activity using normal
    peritoneal exudate cells.

-------
                                  86

   The in vitro migration system has been used primarily in guinea pigs,
but also in  mice, rats,  hamsters,  and monkeys (David and  David 1971);
and it has been adapted to the assay of human cells (Thor 1971).
   The principle of this reaction has been applied to the measurement of
the effector stage of  cell-mediated immunity in immunotoxicological stu-
dies (Luster and Faith 1979).  For example, Luster et al. (1979) used  an in
vitro MIF assay to demonstrate a modest, but dose-related suppression of
cell-mediated immunity in TCDF-exposed lymphocytes. These results were
in agreement with the mitogen  stimulation assay, another in vitro test of
cell-mediated immunity.
   The migration assay is semiquantitative and not extremely expensive;
however, it  does require attention to technical  detail  (Luster and  Faith
1979, Morley et al. 1978). Assay (1), above, is thought to be more sensi-
tive than other in vitro methods of assessing delayed hypersensitivity pro-
vided high concentrations of antigens are used (Carter and Bazin 1980).
   Macrophage Aggregation.  Macrophage aggregation, another  in  vitro
correlate of DTH, may become more widely used than MIF tests because
of its relative simplicity (Crowle 1975).
   During investigations of the  interaction  of antigen and  sensitive lym-
phoid  populations,  aggregation of sensitive peritoneal exudate cells was
observed in  tubes in the presence of antigen. Lolekha et al. (1970) studied
this  reaction using  BCG and egg albumin as antigens. Guinea pigs were
immunized via footpad injections, and the test was performed 4 to 8 weeks
following BCG immunization and 10 to 14 days after egg  albumin. The
animals were injected intraperitoneally with  light mineral oil and 5 days
later  were  exsanguinated.  Peritoneal  exudate  cells  were  removed by
conventional methods  and  were washed,  counted, and  resuspended in
sterile tubes containing PPD or egg albumin. After 24 h at  37°C and 5%
CC>2 the tubes  were shaken to resuspend the cells for observation of aggre-
gates. The reaction was antigen-specific, and it was also found that super-
natant  fluids from lymphoid cells cultured from antigen-sensitive animals
caused  aggregation of nonsensitive peritoneal exudate cells in the presence
of antigen.  The investigators  concluded  that  exposure  to antigen leads to
the release of a macrophage aggregation factor (MAP) from sensitive lym-
phoid cells in culture.
   The soluble material MAP, which  causes aggregation  of peritoneal
exudate cells may be the same  as MIF  (Lolekha et al. 1970). The  MAP
assay is simpler and less time consuming than the MIF assay (Lolekha et
al.  1970, Crowle 1975) and  lends itself  to semiquantitation by  routine
titration.
   No  examples were found for the  use of this test in immunotoxicology
studies.
   Leukocyte Adherence Inhibition (LAI).  This assay, originally described
by Halliday and Miller (1972),  measures the release of a diffusable factor
from sensitized T-lymphocytes following contact with  antigen. The factor,
which inhibits  the adherence of leukocytes to glass or plastic surfaces, has

-------
                                   87

been used to assess cell-mediated immunity. A modified microtest, adapted
from Halliday and Miller by Holt et al. (1974), was used in the following
study to detect immunological effects of cigarette smoke.
   In this  study (Holt et al. 1976) mice were exposed to high- or low-tar
cigarette smoke for 36  or 31  weeks,  respectively, then  inoculated with
BCG. Purified protein derivative of BCG was dispensed into microtest tis-
sue culture plates, and  splenic lymphocytes  from the BCG-treated mice
were added to the plates. After a 2-h incubation period the  plates were
washed and  stained, and  adherent  cells were  counted.  The  numbers of
adherent cells in  the  wells  containing  antigen were compared with the
numbers in wells  containing  no antigen.  Exposure to high-  and  low-tar
cigarette smoke  resulted in suppression of the LAI, indicating suppression
of cell-mediated immunity.
   The reaction of LAI  is similar to that of macrophage migration inhibi-
tion, but the procedure  is much simpler to perform (Halliday and  Miller
1972).

3.2.2 Assessment of Humoral Immunity

   Humoral immunity is defined as that specific immunity which is medi-
ated by antibodies (immunoglobulins) present in  the  plasma,  lymph, and
tissue  fluids of  the body  and which may also become  attached to cells
(Herbert and Wilkinson 1971). The cells ultimately responsible for the
secretion of antibody are the  B- (bursa-derived; bone marrow-derived)
lymphocytes.  When  stimulated by antigen, B-cells,  specific  for that
antigen, either proliferate and develop into plasma cells (antibody-secreting
cells) or serve as memory cells.
   The introduction of antigen is usually followed by a lag phase when no
antibody can be detected in the circulation (Park and  Good 1974). During
this  period (the  first 24 to  48 h  after  antigen)  the  following  occur:
antigen-processing by macrophages, active synthesis of mRNA and rRNA,
transformation of  the lymphocytes, and proliferation of these  cells, which
differentiate  into  antibody producers. Morphologic changes occur  in the
red pulp or at the  edges  of the white  pulp of the spleen or in the medullary
cords or zones surrounding the germinal centers in the cortical  areas of the
lymph nodes.
   The lag phase  is followed by a logarithmic phase during which the rate
of production of  19S IgM  antibody, the first detected, follows a loga-
rithmic curve. After IgM production, 7S IgG antibody appears and enters
the log phase. In a typical antibody  response the plateau of antibody pro-
duction is  reached about  8  days after introduction of  antigen, and the
decay  period which then follows is  short. The primary response is weak,
sluggish, and short lived.
   In contrast to  the primary response, the secondary response  (antibody
production   after  the  second  immunization) is characterized  by  faster
development  and higher levels of antibody titers and a  predominance of
IgG early in the response.

-------
                                 88

   It has been determined (Robertson 1981)  that immunoglobulin  M is
synthesized by the B-cells at early stages of maturation and that IgG, IgA,
or IgE are synthesized at later stages.
   For most antigens, cooperation between both T-  and B-cells is neces-
sary for the induction of antibody-formation (this is the T-cell-dependent
humoral response) (Claman et al. 1966, Miller and  Mitchell  1968). How-
ever, certain antigens, such as type III pneumococcal polysaccharide and
E.  coli  lipopolysaccharide, can  stimulate a  good  immune response  in
athymic animals (e.g., nude mice)  (this is the  T-cell-independent humoral
response) (Moller and Michael  1971, Andersson and  Blomgren 1971).
Tests of humoral immunity include those  which measure circulating anti-
body  or  local production  of antibody  to  both thymus-dependent  and
thymus-independent  antigen; those  which identify  and quantitate the vari-
ous immunoglobulins; and those which assess the colony-forming and blas-
togenic  capabilities   of  B-cells.  Although numerous useful tests  (and
modifications  of these tests)  have been developed  for  the purpose  of
evaluating the humoral immune  response, the  following  discussion will be
limited to those which, because of convenience, speed,  sensitivity,  etc.,
appear to be likely candidates for inclusion in toxicity-screening protocols.

3.2.2.1 Assays for local production of antibody
    Many methods have  been developed for the enumeration  of individual
cells producing antibody. Dresser (1978) defined  three  approaches to the
study of these cells with  respect to  localization  of antibody at various sites:
(1) detection of antibody as it is secreted from the cell;  (2) detection of
intracellular antibody; and  (3) detection of antibody on the cell surface.
Tests designed to  identify antibody at these sites include  plaque-forming
assays (localized hemolysis in gel),  immunofluorescence  techniques, and
the "erythrocyte-antibody-complement" (EAC)  rosette assay, respectively.
    Plaque-forming assays  (localized hemolysis in gel). Jerne and  Nordin
(1963) first  described this method  for  enumerating antibody-producing
cells, and it has become  one of the most widely used tests in experimental
immunology.
    According to the original  assay, cells are obtained from the spleen or
other  lymphoid organs of  an animal that has been immunized  against
SRBC. These cells are mixed with SRBC and  molten agar, the mixture is
poured into a petri  dish and  allowed to set, and the dish  is incubated at
37°C.  After incubation the release by each cell of hemolysins (antibodies
capable of lysing red blood cells in the presence of complement) is revealed
with the  addition  of complement to the  dish.  A  clear zone  of  hemolysis
appears around each antibody-forming cell.
    The  technical  aspects of  the  plaque-forming cell  assay have   been
described in detail by Dresser  (1978).
    The PFC technique of Jerne and Nordin  (1963) detected  only  those
cells producing antibody,  primarily IgM,  which could fix complement and
lyse cells directly.  These  cells are referred to as  direct  plaque-forming

-------
                                  89
cells.  Other classes of immune-globulins (IgG, IgA)  are either  secreted in
insufficient amounts or are incapable of fixing complement, unless they are
complexed with antiglobulin antibody. Cells that produce these antibodies
are referred  to as indirect plaque-forming cells  and are revealed  when
developing serum is added to  the plates. The  PFC technique can  be used
to measure both primary and secondary responses.
   Modifications of the original technique have been introduced that are
more  convenient,  sensitive,  and economical than  the original  procedure.
Methods have  been described that utilize  agarose instead of agar as the
support medium (Plotz et al. 1968, Yamada and Yamada 1969, Duke and
Harshman 1971,  Urso  and  Gengozian  1973) or  that  use  no  support
medium at all (Cunningham  and Szenberg  1968, Kennedy and Axelrad
1971).  Some of these procedures employ microscope slides instead of petri
dishes (Plotz et al.  1968, Urso and Gengozian  1973). Three of these assays
were  compared for sensitivity by Peterman  and  Wust (1975), who also
defined optimal conditions for the  assay by testing  spleen cells from rats
immunized with sheep or mouse erythrocytes in combination with human
or guinea pig complement. They found that the most sensitive  system was
one using SRBC, no agar or agarose,  and either human or guinea pig com-
plement.
   Other modifications of particular importance  to immunotoxicology
have  been introduced by: (1)  Mishell and  Dutton (1967), who performed
the plaque-forming cell  assay following in vitro immunization of normal
spleen cells; (2) Archer et al. (1978),  who used the Mishell-Dutton modifi-
cation to develop  an  assay for the in vitro  immunosuppresive effects  of
water soluble and insoluble chemicals; and (3) Tucker et al. (1982), who
adapted the Mishell-Dutton assay  for testing chemicals that require meta-
bolic  activation.
   The plaque-forming  cell assay can be performed with a  variety  of
antigens such  as polysaccharides, synthetic polymers, proteins, and  hap-
tens,  which can be bound covalently to indicator erythrocytes  (Peterman
and  Wust 1975).  Thus, both T-cell dependent  and T-cell independent
antigens can  be tested  with this  procedure. Protein antigens are  particu-
larly  recommended when  Fischer  rats or guinea pigs are used  because
these  species  do not respond well to SRBC (Luster and Faith 1979).
   The plaque-forming cell assay and its modifications have been used fre-
quently in immunotoxicology  studies. It has been  shown to be a sensitive
indicator of alterations in the immune response. Table 3.4  lists examples of
the plaque-forming cell  response  in experimental  animals following  expo-
sure to environmental  pollutants.
   It  is noteworthy that certain of these chemicals  were  reported to pro-
duce  suppression of the  plaque-forming cell response at dose levels which
produce  no  signs of  overt  clinical illness  [i.e.,  hexachlorobenzene,
polychlorinated biphenyl (Loose et al. 1977,  1978, 1979), polybrominated
biphenyl (Fraker 1980), cadmium chloride (Roller et  al.  1975)]. In addi-
tion, the plaque-forming cell response was  depressed by nickel  chloride at

-------
                                  90

levels falling below the threshold limit value for humans (Graham et al.
1978).
   The plaque-forming cell response was  suppressed  by all the chemicals
listed in Table 3.4 except for nickel oxide (Graham  et al.  1978) and the
platinum  ethylene diamines (Berenbaum  1971).  The lack of  effect of
nickel oxide  could be attributed to insolubility of the compound; lack of
effect of the  platinum ethylene diamines could be attributed to their chem-
ical nature because other platinum diamines were highly toxic to mice at
the doses given.
   The advantages of the plaque-forming  cell assay include the following:
(1) it is sensitive to the immunosuppressive effects of industrial chemicals;
(2) it is versatile  and convenient and can be performed with a variety of
antigens; (3) it lends itself well to the analysis of multiple samples;  and
(4) the rate of antibody secretion can be determined mathematically by
measuring the direct plaque size (Luster and Faith 1979).
   Immunofluorescence.  Antigens and antibodies can be made fluorescent
by chemically binding them to a fluorescent dye, most often isocyanate or
isothiocyanate  of  fluorescein (Gell and Coombs 1975). This fluorochrome
labeling procedure can provide rapid, accurate  localization of the site of
antigen-antibody  interaction when one  of the reactants  forms part of a
cell, tissue, or other biological structure.
    Fluorochromes are  substances that will  absorb  radiation  (ultraviolet
light) and become excited. The excited molecules emit observable radiation
which ceases almost immediately after the exciting radiation is withdrawn.
    There are  several approaches to  fluorescent labeling. The basic  pro-
cedures for these  have been  described in detail by Johnson and Holborow
(1973). These microscopic methods are especially applicable to localization
of 7-globulin and specific antibody.  7-Globulin in  cells  can be demon-
strated with a fluorescent conjugate  prepared from antiglobulin  serum
(direct  method) while specific antibody can  be shown with the "sandwich"
(or  indirect) technique,  where  the  antibody-containing cells  are  first
treated with the antigen and then free valencies of the antigen are used to
combine a fluorescent specific antiserum (Figure 3.2).
    The application of immunofluorescence in the detection of cellular anti-
bodies  using the  "sandwich"  technique  involves these basic procedures
(Johnson and Holborow  1973):
  1.  The preparation  of the substrate  material (cells or tissue) for micros-
     copic examination. (Care must be taken to avoid  fixatives that destroy
     reactivity of  immunoglobulins as antibodies. Cryostat or liquid nitro-
     gen methods  may be used  for tissue sections; dried  smears may be
     used for cell suspensions.)

  2.  The preparation and immunological  characterization of the fluores-
     cent reagents.

  3.  Staining with the fluorochrome dye. (Care must  be taken  to  keep the
     slides moist and to remove nonspecifically bound reactants.)

-------
                                   91
             Table 3.4 PFC Response in Experimental Animals Following
                      Exposure to Environmental Pollutants
Chemical
Hexachlorobenzene
Polychlorinated biphenyl
Polybrominated biphenyl mixture
Trichloroethylene
Cigarette smoke
Benzo[a]pyrene
Carbon and sulfur dioxide
Sulfur Dioxide
Carbon
Carbon dust
Cadmium chloride
Chromium chloride
Nickel chloride
Nickel sulfate
Nickel oxide
Platinum diamine dichloride
Platinum diamine tetrachloride
Platinum ethylene diamines
Methyl mercury
Lead
Species
Mouse
Mouse
Mouse
Mouse, rat
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse"
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Mouse
Rat
Mouse
"Some doses produced immunosuppression without
bExposed in utero.
"Doses below threshold limit value.
PFC
Response
Suppressed
Suppressed
Suppressed
No effect
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed
Suppressed
No effect
Suppressed
Suppressed
Suppressed
No effect
Suppressed
Suppressed
No effect
Suppressed
Suppressed
Suppressed
Suppressed
signs of overt
Reference
Loose et al. 1977'
Loose et al. 1978"
Loose et al. 1979"
Moore et al. 1978
Fraker 1980"
Sanders et al. 1982
Thomas et al. 1973
Nulsen et al. 1974
Thomas et al. 1974b
Urso and Gengozian 1980
Zarkower 1972
Zarkower 1972
Zarkower 1972
Zarkower and Merges 1972
Koller et al. 1975"
Graham et al. 1978
Graham et al. 1978
Graham et al. 1975
Graham et al. 1978C
Graham et al. 1978
Graham et al. 1978
Berenbaum 1971
Berenbaum 1971
Berenbaum 1971
Koller et al. 1977
Ohi et al. 1976
Luster et al. 1978
Koller and Kovacic 1974
clinical illness.
 4. Microscopy under conditions suitable for observing fluorescence.  (To
    achieve maximal fluorescence,  a careful choice of filters is  necessary,
    based on  the excitation and emission characteristics of the  fluoro-
    chrome in question.)
   Fluorescent techniques have been used successfully in immunotoxicol-
ogy studies, and their versatility is illustrated by the following examples.
   Verschuuren et  al.  (1970)  used a fluorescent-antibody  technique to
demonstrate  a decrease  in  immunologically active cells  in  the popliteal
lymph nodes of guinea  pigs which were  fed triphenyltin  acetate (TPTA)
for 90 days  and  then injected with tetanus toxoid. The popliteal lymph

-------
                                  92
                                                     ORNL-DWG 85-10740
Direct fluorescent antibody test
(for nuclear antigen and antibody)
Nucleated cells
on slide
Treated with
fluorescent
nuclear antibody
Washed



M^MMMMt


JH AiA

Key
A Fluorescent nuclear antibody
| Fluorescent globulin
V Antibody in ceil to egg albumin
O Egg albumin
1 Fluorescent antibody to egg albumin
Sandwich test
(for specific onhbody- producing cells)
Single cell
producing antibody
( V ) t° egg albumin
Treated with
specific antigen
O (egg albumin)
Washed
Treated with
fluorescent
antibody^'
to egg albumin
Washed


rAsassf^

i — iftft — i

r±Vt&^

r=>§Ki 	 1

   Figure  3.2. Direct fluorescent antibody test  and "sandwich test"  for
staining antibody-producing cells. (Source: Gell and Coombs  1975.  Used
with permission of Blackwell Scientific Publications, Ltd.)
nodes  were subsequently removed and quick-frozen. For detection  of  all
gamma-globulin-containing cells, the cryostat sections were treated with
fluoresceinated  rabbit  anti-guinea pig  gamma globulin. For the specific
detection of tetanus antitoxin-producing cells, serial  sections were treated
first with  tetanus  toxoid, then with  rabbit tetanus  antitoxin serum, and
finally with fluoresceinated goat anti-rabbit serum.
    In  a later study, Vos and de Roij (1972) used a similar technique to
determine  the effect  of PCB on gamma globulin synthesis in  popliteal
lymph nodes  of guinea pigs  stimulated with  tetanus toxoid. Photographs
were taken of the fluorescein-labeled  tissue  sections using a fluorescence
microscope,  and  the   photographs  were ranged in order  of increasing
number  of 7-globulin-containing  cells to make statistical evaluation  possi-
ble. The investigators  made the  observation that this technique is a  sensi-
tive parameter for the detection of immunosuppressive activity of PCB in
tetanus toxoid-stimulated animals.

-------
                                   93

   Bozelka et al. (1977) used a direct immunofluorescence method to elu-
cidate the  mechanism of cadmium-induced splenomegaly in mice. Follow-
ing  SRBC  injection,  increased  numbers  of  cells  labeled  with  goat
antimouse  IgM and IgG were observed in both the white and red pulp of
the enlarged  spleens of cadmium-intoxicated animals. This  led the authors
to conclude  that the B lymphocyte was responsible for the hyperplastic
changes induced by cadmium.
   Fluorescent techniques require special equipment, which may be expen-
sive. Also, unwanted fluorescein staining may occur with  both the direct
and  indirect  methods. However, the procedure is rapid  and specific and
can be performed on tissue sections so that antibody-producing cells can be
observed in situ. In toxicity testing the procedure would be  especially valu-
able  in  answering specific  questions regarding  antibody  synthesis once
immunotoxicologic effects have been established.

3.2.2.2 Measurement of circulating antibody and immunoglobulins
   The production and secretion of antibody is a highly specific response
of plasma  cells that have been exposed to antigen. Analysis of serum anti-
body provides a direct and  quantitative measurement of the production of
specific antibody by these cells.  Most procedures  require the interaction, in
vitro, of antigen and antibody; these procedures can  be applied to measure-
ment of the  primary and secondary responses to both  thymus-dependent
and   thymus-independent  antigens.   These  procedures,  such  as  routine
hemagglutination and hemolysin titrations, are simple to perform and  are
widely used.  More sophisticated methods, such as radial immunodiffusion
and  the enzyme-linked  immunosorbent assay, have been  applied to  the
identification and  quantitation of the individual immunoglobulins present
in the antisera.
   Several of these procedures,  sensitive to alterations in humoral immune
functions, will be described in the following sections.
   Precipitation.  When  soluble antigen is  mixed with antiserum specific
for that antigen  a precipitate is formed in a few minutes that can be
observed and measured (Gell and Coombs 1975). The interfacial or ring
technique,  a  qualitative test  for precipitating antibodies, was developed by
Ascoli in  1902 (as cited in Gell and Coombs 1975): A drop or two of
antigen is layered above a similar volume of undiluted antiserum, and in a
positive reaction a ring of precipitate forms at the  interface. This type of
test is more sensitive for the detection of antigen than for antibody.
   Precipitation techniques  have been standardized  and  quantitated.  A
sensitive and  reproducible method for the detection of antibody, which can
be completed in less than 2 h, was described by Raney and  McLennan
(1979).  To  test  their  procedure, samples of rabbit antisera produced
against a  nucleoside-protein conjugate  were  added to  tubes  containing
increasing  amounts  of  this  antigen in   a  buffered  solution  of  2%
polyethylene  glycol. The tubes were  incubated at 37°C for 30 or 60 min
and stored at 4°C for 15 minutes to  1 week with daily resuspension of the

-------
                                  94

precipitates. The precipitate was washed and dried, and the protein content
was determined at various time points.  These results were compared with
those  of duplicate samples  that were assayed by the standard procedure
that  does not incorporate the  use of polyethylene glycol. In  comparison
with the standard procedure, the method of Raney and McLennan requires
less time for complete precipitation (2 h vs 2 to 10 days); it is also more
sensitive and more reliable.
   Another precipitation  technique, gel diffusion, has become  widely used
for the  qualitative detection of antibodies to soluble antigens, and, though
slightly slower than the ring test, it  is more versatile. In single diffusion
one reagent is incorporated in the supporting  gel while the other,  in liquid
form, diffuses into it.
   A quantitative modification of this procedure, single radial diffusion
(SRID). was developed by Mancini et al. (1965). The technique, useful for
quantitative determination of different  immunoglobulins in serum, is per-
formed as follows (Ouchterlony and Nillson 1973): a melted  3% agar solu-
tion  is  mixed with an equal volume  of antiserum that has  been  properly
diluted. The agar mixture  is then poured as  a  1-mm thick layer onto a
glass  plate.  Holes 2 mm wide are punched in the gel and filled with accu-
rately measured volumes of the  antigen solutions to be tested along  with
references of known concentration of antigen. The plate is incubated in a
humid  atmosphere  until the resulting precipitate becomes stationary. The
area of the  halo of precipitate is determined. Quantitation is possible since
the  area of  the  halo is directly proportional to  the  concentration of
antigen. The "reversed" SRID technique allows quantitative determination
of specific  antibody when the antigen constitutes the internal reactant and
the antiserum to be tested constitutes the external reactant.
   The SRID technique is  not very sensitive but can be used satisfactorily
when high antibody titers are present,  as in the secondary response to an
antigen (Vos 1977).
   Immunodiffusion  methods  have been  employed  in  immunotoxicology
studies  for the assessment of antibody formation and for the measurement
of immunoglobulins.  Vos et al. (1973)  measured serum tetanus-antitoxin
levels  in immunized guinea pigs using a RID  procedure  in which the
tetanus  toxoid was incorporated into  the agar  and the  antiserum was
allowed to diffuse out from the well. A slight suppression of the humoral
immune response  by  sublethal doses of  TCDD was detected with this
method.
   In a similar study, Luster et al. (1979), using a RID assay, quantified
serum IgG in the serum of guinea pigs that had been exposed to TCDF. A
minimal depression in serum  IgG  was observed, which  correlated  with
antibody production to bovine gamma globulin.
   The RID assay was used by Loose et al. (1979) to measure serum IgG,
IgA,  and  IgM immunoglobulin  concentrations in  mice that had  been
treated with polychlorinated biphenyls or hexachlorobenzene, then immun-
ized with SRBC. There was a reduction in serum IgA, IgG, and IgM that

-------
                                  95

was concomitant with a decrease in the spleen plaque formation response.
In the response  to a secondary challenge  only  IgA  was depressed.  The
chemicals were,  however, administered at subtoxic levels.
   Precipitation methods are simple  and convenient, especially  since com-
mercial plates have been introduced.  There  seems to be a good  correlation
between  antibody and Ig levels measured by SRID and antibody produc-
tion measured by agglutination or plaque formation.
   Hemagglutination   and  hemolysis.  The  direct  agglutination  test  is
another classical serological reaction which involves, simply, the clumping
of a cell suspension by specific antibody (Gell and Coombs 1975). The test
is performed by mixing dilutions of antiserum with the cell suspension and
observing  agglutination of  the cells.   Antibody combines  rapidly with
antigen,  but no agglutination occurs until the cells come in contact with
each other. Early agglutination tests were performed in tubes. A more effi-
cient  micromethod of the  test was described by Stavitsky (1954) for use
with tannic acid and  protein-treated cells (see below); the test is also quite
useful with untreated red blood cells. A modification of the method, one of
several procedures recommended by White  et al.  (in press) for assessing
the effects of toxicants or immunocompetence, will be outlined below. By
adding complement to the system,  the  hemagglutinin  reaction can be con-
verted to a hemolytic reaction (Vos 1977). There is generally good correla-
tion between the hemagglutination and  hemolysis titrations.
   In the method described by White et  al.  (unpublished  observations),
mice  are exposed to the test chemical for the desired length of time and on
day  0 are immunized  intraperitoneally with 1 x  109 SRBC.  On day  7,
blood is  collected in an anticoagulant (sodium citrate), and  following cen-
trifugation  an  aliquot of the plasma is removed and heat  inactivated  at
56°C for 30 minutes. Serial 1:1 dilutions are made in phosphate-buffered
saline in  conical microtiter plates  to a final volume  of 0.1  mL per well.
Eight  samples  can be  diluted at  once using a  semiautomatic technique.
One-tenth milliliter of a 0.5% SRBC suspension is added to each well, and
the plates are covered and placed in  a  37°C humidified incubator. After 2
h the plates are observed for agglutination using  a magnifying mirror. The
end point is the serum dilution at which no  agglutination occurs. (The titer
is expressed as  the reciprocal of the highest dilution  that gives a positive
reaction.) The secondary antibody  response can also be tested in this way,
and estimation  of the proportions of the IgM and  IgG antibodies can  be
achieved by degradation of IgM with 2-mercaptoethanol.
   The sheep erythrocyte hemagglutination and/or hemolysin  techniques
have been used  frequently to assess the humoral immune response follow-
ing  exposure   of  experimental  animals   to  chemicals,   such   as tri-
chloroethylene  in mice (Sanders et al.  1982); PCB  in monkeys (Thomas
and Hinsdill 1978),  carbon dust in  mice (Zarkower and Morges  1972),
and cigarette smoke in mice (Esber et al. 1973; Nulsen et al. 1974, Holt et
al. 1976).

-------
                                   96
   Sanders  et al.  (1982)  detected slight  suppression  in the  antibody
response to  SRBC following exposure of mice  to  subtoxic levels  of tri-
chloroethylene. Also,  Thomas and Hinsdill (1978) observed suppression of
the SRBC,  but not  the  tetanus typhoid, antibody  response  of  monkeys
after exposure to PCB. They speculated that the SRBC response is more
sensitive to damage than the tetanus toxoid response.
   The hemagglutination and hemolysin assays are thus sensitive, easy to
perform, economical,  and quantitative. They  can  be used to titrate primary
and secondary antibody responses and to measure the relative amounts of
IgG and IgM  produced in response to an antigen.
   Passive hemagglutination. The chief difference in the precipitation reac-
tion  and  the  hemagglutination reaction  is  that in  the former  case the
antigen molecules are free  in solution and in the latter they are fixed on
the surface  (membrane) of the cell (Coombs and Gell 1975). Agglutina-
tion  is far  more sensitive  than  precipitation  in  detecting  the  antigen-
antibody reaction (Herbert  1973, Coombs and Gell 1975). Boyden  (1951)
thus developed a technique  for the adsorption of various  proteins  on the
surface of tannic acid-treated  (tanned cell technique)  SRBC  and demon-
strated the  agglutination of such  cells by highly diluted specific antisera.
The  term  "passive"  agglutination is  used  because  the red  cells  act as
passive carriers  of  the  protein antigens. Stavitsky (1954) adapted the
technique to  the titration of antitoxin and other protein sera using the
microtiter assay previously discussed.
   Technical  details for the preparation of the antigen-coated cells  for use
in agglutination procedures  have been outlined by Herbert (1973). These
will be briefly summarized here:
   Choice of Cells    Although it would seem logical to use cells of the
same species for agglutination as that from which the antiserum is derived,
most  investigators use SRBC or human O  cells. Both of  these are fairly
easy to obtain and  have been satisfactory  in the  tanned cell technique,
human cells being the more sensitive of the two. The cells should be  fresh.
   Preservation of Cells   Unless  the fresh cells can be  used right away
they must be preserved (or fixed). This can be done before or after  coating
but usually takes place prior  to  coating  because fixation may  alter the
coating antigen. The most  commonly used fixative is formalin. Preserved
cells are, however, considerably less sensitive than fresh ones.
   Adsorption of Antigen  by  Red Cells    Coating  of the red cells with
antigen may  be  accomplished in  three  ways:  by  direct adsorption of
antigen by the cells,  by  the tanned cell technique, and by chemical bond-
ing of antigen to the cell.
   Direct Adsorption  Direct coating of fresh cells can be done by simply
incubating the antigen and  cells at 37°C for approximately 2 h. Such cells
are used in both hemagglutination and,  with the addition  of complement,
hemolytic tests. These cells are, however, very  insensitive to antibody in
comparison  to tanned cells. Preserved (fixed) cells can also be coated this

-------
                                  97

way but with a longer incubation period, with constant mixing, and at the
pH that is appropriate for the particular antigen.
    Certain haptens can  also become directly attached to red blood cells,
and hemagglutination tests with such cells appear to be quite sensitive.
    Tanned Cell Technique  It is generally assumed that tannic acid acts
on  red cells in such  a way to cause them to take up protein antigens, but
Herbert (1973) points out that the chief function may well be to increase
the instability of the  cells, making a normally nonagglutinating reaction
result in agglutination. Cells that have been treated  with tannic acid will
show nonspecific agglutination settling patterns, but the tendency to agglu-
tinate  may be balanced by  adding normal  serum  as a  stabilizer. The
coated cells are thus  highly sensitive to agglutination, in the presence of a
very small amount of antibody.
    The tanning procedure itself is easy  to perform  and involves simply
incubating the fresh  or  fixed cells with tannic  acid  in buffer for a brief
period of time. The cells can  then be stored for  subsequent coating with
antigen, which can be accomplished by incubating the cells and antigen at
room temperature for a  very short time.  During  the tanning and  coating
procedures the cells  must be  carefully washed at each step, the pH must
be  carefully monitored,  and optimal concentrations of the reactants must
be  used. The cells thus prepared are ready for use in  conventional titration
procedures.
    Attachment of Antigen to Red Cells by Use of an Intermediate Com-
pound  Bis-diazotized benzidine  is a suitable chemical for use in this pro-
cedure. Coupling the antigen to the cells is accomplished by simply incu-
bating cells (usually formalinized), bis-diazotized benzidine, and  antigen
together for a short  time  and then washing the sensitized cells. Other
chemicals recommended  for this  include chromic  chloride, difluorodinitro-
benzene, and carbodiimide.
    Passive hemagglutination  techniques have been employed to test the
humoral response to the  following:  (1) E. coli lipopolysaccharide following
exposure of rats to lead acetate (Luster et  al.  1978) and of mice  to carbon
(Zarkower 1972; Zarkower and Morges 1972); (2) human gamma globulin
following  exposure  of rats to cadmium  chloride (Jones et  al.  1971);
(3) diphtheria toxoid  following exposure of guinea pigs to DDT (Gabliks
et al.  1973); (4) Brucella abortus antigen following exposure of mice to
methyl mercury (Spyker 1975); and (5) tetanus toxoid following exposure
of monkeys to PCB (Thomas and Hinsdill 1978).
    Luster et al. (1978) tested the humoral immune response to both sheep
erythrocytes   (SRBC),  a   thymus-dependent   antigen,   and  E.   coli
lipopolysaccharide, a  thymus-independent  antigen, in  mice exposed to lead
acetate. The SRBC response was suppressed by lead while the lipopolysac-
charide response was  not,  indicating an alteration of the T-, rather than
the B-, lymphocyte population.
    The reciprocal  effect  was demonstrated in  the study  reported  by
Spyker (1975). Mice were exposed to methyl mercury in utero  and were

-------
                                  98

challenged  14  months later with Brucella abortus  antigen, a  thymus-
independent antigen, and SRBC. In this case B-cell function was impaired
while T-cell function was unaffected. (No impairment was observed in the
response to either antigen when mice were challenged at 4 months of age,
indicating that  impaired immune function may be  a delayed effect of pre-
natal exposure to methyl mercury.)
   In summary, agglutination methods are objective, quantitative, sensi-
tive, and versatile. Passive hemagglutination techniques are a bit  compli-
cated, but  they are valuable tools for  the  study  of  responses to soluble
antigens. Micromethods simplify the procedures and allow the use of small
quantities of reagents, thereby lowering the cost of the tests.
   Enzyme-linked  immunosorbent assay. Another test that,  similar to the
radial immunodiffusion technique, has the capability  for the  measurement
of both  antigen and antibody in sera is the  enzyme-linked immunosorbent
assay (ELISA), which was  introduced by Engvall and Perlmann in  1971.
The technique was adapted  by Vos et al. (1979a, 1979b,  1979c)  for use  in
studies on the humoral immune response of rats. The  following "sandwich"
ELISA  was used by them for the measurement of rat IgG and IgM:
    Polystyrene tubes were coated  (by rotation for 18 h) with antimmuno-
globulin solution; test and reference serum were  then  diluted and added  to
the appropriate tube and the tubes were incubated under rotation  for 1  h.
After washing,  1 mL of a conjugate of  horseradish peroxidase and antiim-
munoglobulin (IgG  or IgM) was added. The tubes  were then  incubated
under rotation  for  2  h.  Excess conjugate  was removed, and  1 mL  of
specific hydrogen peroxide substrate was added. Extinction  was measured
at 490 nm following incubation for one hour at  room temperature. Vos et
al.  (1979a) reported that the measurements of IgM and  IgG  correlated
well with results obtained  by SRID.  ELISA is  less  time-consuming than
the SRID technique, and the macro-ELISA used here required less serum.
    Vos  et al. (1979a)  also utilized ELISA for determining antibody titers
to E. coll  lipopolysaccharide and tetanus toxoid.  The method performed
and described by them was that of Ruitenberg et al. (1975):
    Polystyrene microplates were coated with the antigens for 2 h at 37°C
(LPS did not attach well and special procedures were required here). The
microplate trays were incubated  with 100 pL of  test or control serum dilu-
tions for  1  h at  37°C.   After  washing,  100 juL of the horseradish
peroxidase-antiimmunoglobulin  conjugate was  added, and the microplates
were incubated and  rotated for 2 h. The plates were washed  and substrate
was added. After incubation for  1  h at room temperature the reaction pro-
duct was evaluated  visually and expressed as  2log of the highest dilution
giving a positive reaction.  The ELISA results were compared with those
obtained with passive hemagglutination, and ELISA was found to be more
sensitive except when secondary  response  titers  to  tetanus toxoid were
compared.
    Vos et  al. (1979b,  1979c) included  the ELISA with routine studies to
assess immune function  in  rats  exposed prenatally and postnatally, or  as

-------
                                  99

adults, to hexachlorobenzene. The procedure appeared  to provide a satis-
factory assessment of the humoral immune response in the rats.
   According to Vos et al.  (1979a) and Roller (1982)  the procedure is a
reliable,  sensitive,  specific,  and economic method for detecting IgM and
IgG antibodies and antigens. It has the potential for large-scale screening
because mechanization of all steps of the assay is feasible.  In  the labora-
tory of Roller (1982), ELISA  has been accepted as the test procedure of
choice for assessing humoral immunity in animals exposed to chemicals.

3.2.2.3 Other tests of B-ceU  function

   B-cell function can be evaluated by procedures that measure parame-
ters  other than direct antibody formation. Two of these, the EAC rosette
technique and transformation by B-cell mitogens, are discussed.
   EAC  rosette technique.  There is  a population of lymphoid cells that is
capable  of  binding "erythrocyte-antibody-complement"  (EAC)  to form
rosettes.  This is possible because B-cell membranes  possess a receptor  for
the third component of complement (C3).
    Erythrocytes are prepared for this test by first reacting them with rab-
bit antierythrocyte  antibody  (forming  the complex EA) (Bianco  et  al.
1970). The  EA complex can then be treated either with serum from mice
genetically  deficient in  C3 or with purified  human  complement com-
ponents.  The complex thus formed  by antibodies, erythrocytes, and com-
plement is the EAC complex.
    Roller and Brauner (1977)  applied  the  EAC rosette  assay  to  the
enumeration of B-cells from mice that had been exposed to lead acetate or
cadmium chloride. In their experiment,  spleen lymphoid cells were incu-
bated with EAC for 5 min at  37°C; the  cells were then centrifuged for 5
min and incubated on ice for 1 to 2 h. After gentle resuspension, the cells
were stained with  crystal violet and examined  microscopically  in a hema-
cytometer. Cells with three or more bound erythrocytes were counted as
EAC rosettes. Fewer rosettes were formed  by  the  B-cells  from mice
exposed to the chemicals than by B-cells  from control  mice. Cell  viability
tests indicated that this effect was not due to toxicity of the lead and cad-
mium.
   The rosette test is a simple assay for the enumeration of B-cells and is
one  of the tests routinely  employed by  Luster and Faith (1979) for  the
detection of alterations in the humoral immune response.
   Transformation of  lymphocytes  by   B-ceU mitogens. The  technical
aspects  of the mitogen-stimulated  lymphocyte  transformation  assay have
been discussed in  Section 3.2.1.2. The procedure is the same  for both T-
and  B-cell transformation  with the  exception of the mitogens used. Bac-
terial lipopolysaccharides (LPS) are generally  used to stimulate the mito-
gen response of B-cells, while concanavalin A and phytohemagglutinin  are
primarily T-cell mitogens.  Pokeweed mitogen, which  stimulates  both T-
and B-cells,  is also used frequently.

-------
                                  100

   The B-cell  response  has been  shown  to  be sensitive to the  immu-
nosuppressive effects of metals  (Gaworski and Sharma 1978, Koller et al.
1979), TCDD (Sharma et al. 1978), and TCDF (Luster et al. 1979), but
the response has been either enhanced  or unaffected by hexachlorobenzene
(Vos et al.  1979b, I979c).

3.2.3 Assessment of Indirect Parameters of Immunity
   A number of indirect factors must be  considered when evaluating the
immune  response. Indirect parameters  commonly assessed include  macro-
phage functions, hormones, complement, host resistance to infection, resis-
tance to  tumor challenge, and endotoxin hypersensitivity (Luster and Faith
1979). Of  these, macrophage functions and host resistance to infection are
often assayed in  immunotoxicology studies. Some  procedures for evaluat-
ing these parameters are discussed in the following sections.

3.2.3.1 Macrophage functions
    Polymorphonuclear  and mononuclear  phagocytes  are the two major
classes of  cells that are  characterized by the  ability to ingest and  destroy
invading microorganisms (van  Furth et al. 1978).  This process of phago-
cytosis is completed in the following  stages: opsonization of the particles
(viable and nonviable) by serum factors, attachment of the opsonized par-
ticles to  the cell  surface, engulfment of such particles, intracellular killing
of microorganisms, and digestion  of  microorganisms and other ingested
matter. Of these stages, ingestion and intracellular  killing are most fre-
quently tested.
    Experiments for the  study  of macrophage functions can be performed
in vivo or in  vitro. In vitro methods  are  preferred by some  investigators
(Vos  1977, van Furth et al. 1978, Luster and Faith 1979). The advantages
of these  methods are that they allow the use of homogeneous cell  popula-
tions  so  that the  functions of individual cell types (e.g., monocytes, granu-
locytes, and macroc-hages) can  be studied and they eliminate the effects of
serum and other factors on the  macrophage function.
    The main disadvantage is that phagocytosis  and intracellular killing do
not  appear  to be extremely  sensitive  to the  chemicals administered
systemically in studies on the immune response. For  example, in a few of
the experiments  reviewed,  even though other  immune  parameters were
altered  by exposure  to  trichloroethylene (Sanders  et  al.  1982)  or
hexachlorobenzene  (Vos et  al.  1979b,  1979c) macrophage functions were
not affected.  Methods  for  obtaining  macrophages for  study have been
reviewed in detail by Stuart et al. (1978) and van Furth et al. (1978).
    Ingestion. Peritoneal  macrophages  of mice are  frequently used in the
study  of phagocytosis.  In the  procedure  described by  van Furth et  al.
(1978) phagocytes  and an equal number of bacteria are mixed together in
a tube containing serum. The mixture  is incubated  at 37°C under continu-
ous rotation. At various  time points (0, 30, 60,  90, and 120 min) samples

-------
                                   101

are taken from the suspension  and centrifuged, and supernatants  of  the
samples are plated on agar for  subsequent bacterial  colony counts;  or  the
cell suspensions are examined microscopically after 15 minutes incubation
for morphological assessment of phagocytosis.  If colony counts are made,
phagocytosis is measured by the decrease in the number of extracellular
organisms in the supernatants. If morphological assessment is made, cells
are dried on a slide and stained, and cells that have ingested bacteria  are
counted.
   Sanders  et al.  (1982) describe an alternative method for assessing
phagocytic  activity in  which peritoneal  exudate cells  are collected  and
counted  and  allowed  to adhere to a  24-well Costar  culture dish.  The
adhered cells  are washed, and chromated SRBC (opsonized with IgG) is
added. The  plates are incubated and removed at 10 to 45 min. The cells
are washed  and scraped from the  dishes and  placed  in tubes for gamma
counting.
   Intracellular Killing. Intracellular  killing by  peritoneal macrophages
from the mouse can be assessed by the following method (van  Furth et al.
1978). The  cells can be removed and  incubated with bacteria in vitro for
30 min, or  the mouse can be injected  intraperitoneally with the bacteria.
Macrophages  can ingest bacteria in vivo  in 3 min. After the initial period
of phagocytosis, the macrophages are washed, suspended  in medium, and
counted. The suspension is divided into  aliquots, which are incubated at
37°C (one  at 4°C).  After 15,  30,  or  60 min of incubation, aliquots  are
removed, the  cells  are disrupted by freezing and thawing, and the  mix is
plated on agar.  The number of viable  bacteria is  determined from  the
number of colonies.
   The assays just described are performed with  peritoneal macrophages
from mice.  However, similar tests  can  be performed with alveolar macro-
phages. In addition, macrophages can be obtained  from other species such
as the rat and guinea pig. Clearance  and intracellular  killing by alveolar
macrophages  appear  to  be sensitive indicators of injury by inhaled gases
and particles  such  as nitrous oxide (Acton and Myrvik 1972, Vassallo et
al. 1973); cigarette smoke (Rylander 1971), ozone (Goldstein et al.  1971),
and carbon  black (Rylander 1969).

3.2.3.2 Host resistance to infection
   The assessment of  host resistance  to infection following exposure to
chemicals represents a  practical  approach to defining immunosuppression.
For  testing  host  resistance Listeria   monocytogenes  and Streptococcus
pneumoniae are  recommended  because  resistance   to  these   organisms
appears to  be dependent primarily upon normal T-cell and B-cell func-
tions, respectively (Vos  1977, Luster and Faith 1979). All  phases of the
cell-mediated  immune response, including macrophage killing in the effec-
tor stage, participate in the resistance to Listeria (Vos 1977).
   Tripathy and Mackaness (1969) first described the Listeria procedure
and  applied it  to  testing the  effects  of several  drugs  on  the immune

-------
                                 102

response.  The assay can be performed on mice and is based on the princi-
ple that effective suppression  of the immune response by drugs or  other
chemicals  (which can be administered at the onset of infection or during
the course of the infection) would result in continued  multiplication  of
Listeria,  particularly in the liver and spleen. The animals are  injected
intravenously with the live bacteria, and resistance can be measured in two
ways: (1)  mortality  rates of treated animals are compared with those of
the controls or (2) bacterial growth in the spleen is enumerated at days 4
and  6  after  inoculation,  when  cell-mediated  immunity interrupts  the
growth of the organism in vivo. The spleen is homogenized, and the homo-
genate is  plated  for the  subsequent  enumeration  of colonies  formed  by
viable bacteria.
   The advantages of the technique as listed by  Tripathy and  Mackaness
(1969) are as follows:

 1. the immunosuppressive effect is quickly revealed

 2. the method provides an accurate and quantitative assessment of  the
    effect of chemicals.
   When  Streptococcus pneumoniae  is the infectious agent, resistance to
infection is simply assessed by comparing  mortality rates  of  controls  to
those of treated animals (Vos 1977).
   Numerous alternatives to the use of these organisms include the use of
viruses and other bacteria and parasitic expulsion. Increased mortality due
to impaired host  resistance to infectious organisms has been demonstrated
in various laboratory animals treated with chemicals such as PCBs (Friend
and Trainer  1970, Thomas and Hinsdill  1978, Loose et al. 1979), metals
(Exon  et  al.  1975,  Gainer 1977, Roller 1979b,  Exon et al.  1979), and
TCDD (Thigpin et al. 1975).

3.3 ALLERGIC REACTIONS TO ENVIRONMENTAL CHEMICALS
   Coombs and Gell classified allergic reactions in 1963. A detailed dis-
cussion of these reactions can also be found in Coombs and Gell (1975).
The four  classifications are briefly described in  Table 3.5 and are illus-
trated  in  Figure  3.3. Coombs and Gell  stress that the circumstances in
which these four  basic types of reactions may  be studied in an uncompli-
cated form are limited and that the pattern seen in any one human disease
is  often complex,  perhaps involving several of the responses listed in  Table
3.5.
   The Type  I reaction in man includes both general  and local  anaphy-
laxis. In generalized anaphylaxis  if the patient does not die rapidly, local
symptoms develop such as bronchial asthma, hay fever, pulmonary edema,
and  urticaria. An example of local  anaphylaxis  is the appearance of a
wheal at the  site  of antigen application, as in the prick or scratch diagnos-
tic skin tests.
   The Type II  reaction is manifest  as: complement-dependent antibody
cytotoxicity (transfusion reactions, hemolytic disease of  the newborn, or

-------
                                  103


       Table 3.5. Definition of the Four Types of Allergic Reaction

I (Anaphylactic, reagin-dependent)

Initiated by  allergen or antigen reactng  with  tissue cells (basophils  and
mast  cells) passively  sensitized  (allergized) by  antibody produced else-
where,  leading to  the  release  of  pharmacologically active substances
(vasoactive amines). An  example of local anaphylaxis is the appearance of
a wheal at the site of antigen application, as in the prick or scratch diag-
nostic skin tests

II (Cytotoxic or cell stimulating)

Initiated by antibody reacting with either  (1) an  antigenic component of a
cell or tissue element or (2) an antigen or hapten which has  become inti-
mately associated with these; damage may  then  occur in the presence of
complement or of certain kinds of  mononuclear cells. Stimulation of secre-
tory organs may  also occur (e.g., the  thyroid).  The  Type II reaction  is
manifest as  transfusion  reactions, hemolytic disease  of the  newborn, or
lesions produced in tissues by action of antibody and complement

in (Damage by antigen-antibody complexes)
Initiated when antigen reacts in  the tissue spaces with potentially precipi-
tating antibody, forming microprecipitates in and around the small vessels,
causing damage to cells secondarily, or being precipitated in  and  interfer-
ing with the function of membranes, or when  antigen in excess reacts in
the blood stream with potentially  precipitating  antibody, forming soluble
circulating complexes which are deposited in the blood-vessel walls or in
the basement membrane and cause local inflammation or massive comple-
ment activation. Antigen  excess complexes mediate  allergic damage of
three general  kinds: (1) the Arthus reaction (situations in which antigen is
in excess locally in  the body); (2)  serum  sickness (antigen-antibody union
takes place  in the  bloodstream  and  complexes  are  deposited  at sites
throughout  the body);  (3)  "massive  complement  activation" (especially
associated with endotoxin shock)

IV (Delayed, tuberculin-type, cell-mediated)

Initiated essentially by  the reaction of  actively allergized  lymphocytes,
probably of the T (thymus-derived) population responding specifically to
allergen by the release of lymphokines, and/or the development of cytotox-
icity without the participation of free antibody.  Locally, it is manifested by
the  infiltration  of cells  at  the  site  where   the  antigen  is  injected


Source: Adapted from Coombs and Gell (1975).

-------
                                    104
                                                        ORNL-DWG 85-10741
Key.

• A *    Antigens
	1—   Liberation of histamine and other
,        pharmacologically active substances

         Antibody

     --^Sites of involvement of complement or
         non-allergized  lymphocytes

     -    Specific antigen-combining  receptors on
         membrane of specifically allergized lymphocytes
     Figure  3.3. The four  types  of  allergic  reaction. (Source. Coombs  and
  Gell 1975.)

-------
                                  105

lesions produced in tissues by  action of antibody  and complement);  cell
dependent antibody cytotoxicity; antibody reacting  with hapten or antigen
combined with or adsorbed on tissue cells to cause cytotoxicity (drug sensi-
tivity,  e.g., purpura, hemolytic anemia,  and  granulocytopenia);  antibody
reacting   with  tissue  cells   to cause  stimulation  (thyroid  stimulator,
"enhancement" of malignant and other grafts).
    The  Type  III  reactions  are characterized by  antigen-antibody com-
plexes formed in moderate antigen excess so  that they do not precipitate
but remain soluble, thus becoming toxic to the surrounding tissue (precipi-
tated complexes are  normally nonirritating  and  are easily cleared  by
phagocytes).  Antigen excess complexes mediate allergic damage of three
general kinds: (1)  the  Arthus reaction  (situations  in  which  antigen is in
excess locally  in the body);  (2)  serum  sickness (antigen-antibody  union
takes place  in  the bloodstream,  and complexes  are deposited at  sites
throughout the body); and (3) "massive complement activation" (especially
associated with endotoxin shock).
    Type IV reactions are referred to as delayed hypersensitivity, cellular
hypersensitivity,  cell-mediated allergy, and  others.  Nomenclature is con-
stantly under discussion. An  essential  characteristic  of delayed reactivity is
that the state can be  transferred by cells  although  not  by serum. It is
currently thought that  the cell-mediated response is a property  of T-cells
and has  no need for B-cells. The development  of  the delayed lesion was
described in Section 3.2.1.
    Chemical  allergens  found in industrial environments  usually affect a
small subset of workers who are hypersusceptible to the low-dose exposure
legally permitted in chemical plants (Adkinson  1977). For example, from
1  to  10% of industrial workers  exposed to legally  permissible  levels of
organic isocyanates will  develop hypersensitivity reactions following low-
dose exposure to the chemicals  (Pepys 1976).  Similarly, a small group of
highly  susceptible  workers  becomes  immunologically  sensitive  when
exposed to numerous other industrial chemicals, and in almost all cases the
principal allergen has been identified  as a chemical that is highly protein-
reactive (Adkinson  1977).
    Landsteiner and Jacobs (1935,  1936) were the first to demonstrate the
antigenicity of chemicals. They  induced reproducible immunological effects
in guinea pigs with simple compounds, halogenated dinitrobenzene deriva-
tives. They discovered,  however, that  if the halogen  was  replaced with a
hydrogen,  hydroxyl,  methane,  or an amino group   the compound was
immunologically  inactive.  The ability  to  sensitize seemed  to  be related to
the lability of the Cl ion when  treated with alkali.  The investigators con-
cluded that this property facilitated the binding of a  simple nonantigenic
chemical with an endogenous protein carrier, creating an antigenic com-
plex.
   This  concept  led  to  the  "hapten  theory,"  named  for  the chemical or
"hapten" or "fastening agent" involved (Loomis  1974). A hapten is, there-
fore, a compound of low  molecular weight  that combines with a protein

-------
                                  106

carrier to form a product having antigenic properties that can elicit antibo-
dies  in the animal. Although many chemicals do not appear to react with
protein  in vitro, these  complexes are nonetheless  formed in the  intact
animal, probably through metabolic alteration of the chemical.
   It is  now known that simple chemicals of molecular weight  less than
1000 cannot elicit an immune response unless they  are capable of covalent
linkage to carrier molecules (Adkinson 1977).

3.3.1 Allergic Response to Inhalants
   The  lung  is constantly exposed  to  biologically active materials in
inhaled  gases and particulate  matter. The immune response to inhaled
physical and  chemical  agents and the  injurious  consequences of  this
response have  been reviewed in  detail  by Burrell (1977)  and De  Bruin
(1976).
   Burrell (1977)  recognized the  respiratory route as an excellent  means
of immunization,  particularly  in the stimulation of the  IgA  and  IgE
classes.  He also recognized that  these immune responses are not always
beneficial to the host.
   Inhalants  thought to  evoke  allergic  responses include physical  and
chemical agents and antigens  of  plant and animal origin.  Some of these
are listed in Table  3.6.
   Illnesses  due  to  inhalation of organic dusts  include  farmer's lung,
bagassosis, byssinosis, ptilosis, tabacosis,  malt worker's lung, and suberosis
(De  Bruin  1976).  The  organic dusts set up  toxic asthma  and allergic
alveloitis, responses thought to be of the Arthus type. Illness due to inhala-
tion  of  inorganic particles include inorganic pneumoconiosis and chronic
pneumoitis. (The latter  disorder is a consequence of a delayed hypersensi-
tive  reaction induced by  exposure  to beryllium.)

3.3.2 Allergic Response to Dermal Sensitizers

   Many chemicals have been  demonstrated to evoke allergic reactions of
the  skin. The  most common  response  is  allergic contact dermatitis, a
delayed-type hypersensitivity reaction. Among  these allergenic  chemicals
are:  2,4-dinitrochlorobenzene, p-nitrodimethylaniline, p-phenylenediamine,
hexavalent chromium, nickel, beryllium, and cobalt.
   For the induction of contact sensitization the eliciting contact allergen
must be applied to the skin, penetrate the epidermis and become  combined
with protein, with which a complex antigen is formed (De Bruin 1976).
   Contact  sensitivity   is  apparently  a  complex  hypersensitivity to
numerous conjugates (formed in situ) of a chemical with various analogous
proteins.

3.3.3 Tests to Detect the Sensitizing Potential of Chemicals

   Marzulli  and Maibach  (1977) stated that three types  of  tests  are
needed to evaluate  skin sensitization potential:

-------
                                   107
                     Table 3.6. Respiratory Allergens
Physical agents
     Asbestos
     Coal dust
     Silica
Chemical agents
     Beryllium
     Detergent enzymes
     Isocyanates
     Miscellaneous chemical incitants
Antigens of plant origin
     Aspergilli
     Cotton and other textile dusts
     Miscellaneous hypersensitivity pneumonitis incitants
     Thermophilic actinomycetes
Antigens of animal origin
Source: Adapted from Burrell (1977).
 1. the predictive test to identify allergenic substances
 2. the diagnostic test to identify a substance that may be actually pro-
    ducing skin reactions, and
 3. the use test to provide information regarding the safety of ingredients
    in combinations before they enter the marketplace.
   Some of the predictive tests of skin allergy will  be described in the fol-
lowing sections. These  are useful for testing chemicals to which  occupa-
tional exposure may occur via the skin. Also,  a new test for the detection
of allergy induced via respiratory exposure will be described.
3.3.3.1 Skin sensitization
   According to the OECD (1981), skin  sensitization  tests require an ini-
tial exposure to a test substance, followed by a challenge exposure which is

-------
                                  108

administered no less than a week later. Sensitization is evaluated from the
reaction to the challenge exposure, usually 24 and 48 h after the pertinent
exposure,  based on  the proportion of  each  group that became sensitized
and the severity of the reaction in  each animal. The recommended species
for animal testing is the guinea  pig,  with the number and sex determined
by the method  used.  For  the induction  exposure, the concentration used
should produce  a  skin reaction;  for the challenge exposure the concentra-
tion should be nonirritating.
   Skin predictive test procedures in guinea pigs have been identified by
Klecak (1977)  for ascertaining the  allergenic  potential  for  chemicals.
These include the Draize test, Freund's complete adjuvant test, the guinea
pig maximization test, the "split adjuvant"  test,  the Buhler test, and the
open  cutaneous test.
    The Draize Test   In this test,  guinea pigs are sensitized intradermally
on the flank every other day for 18 days, and skin reactions are read 24 h
after  each injection. The challenging dose is administered intradermally on
the opposite flank on day  35. Controls are treated only with the challeng-
ing dose.  On days 36 to 37 the  test  areas are shaved, and the intensity of
erythema and occurrence  and  size of  edema   of  the  test  reaction are
recorded.  The test and control reactions are compared. The test is difficult
to quantitate but it  is easy  to perform,  and   material and operational
expenditures are minimal.
   Freund's Complete Adjuvant (FCA) Test -  In  the  FCA  test, guinea
pigs are  sensitized every other day for 10 days intradermally in the flank
with  a mixture of test material and FCA (killed  mycobacteria). Skin reac-
tions  are read 24 h after each injection. The challenging doses of appropri-
ate concentrations are applied topically to  the  opposite flank. The chal-
lenging dose is not  mixed with Freund's  adjuvant.   Skin reactions are
evaluated as for the Draize test.
    The technique is simple and inexpensive,  but Klecak (1977) believes
that it is more useful for identifying sensitizers than as a predictive test.
    The Guinea Pig Maximization  Test   For  this  test,  guinea pigs are
immunized  on the shoulder both intradermally with  the  test agent in FCA
and topically (test agent in petrolatum) 7 days later. The challenging dose
is applied topically  to the flank on day  21, and  the reactions are read on
days  23  and 24. This test combines  the factors that favor sensitization in
the guinea  pig  and is considered  to be the most sensitive procedure for
detecting allergenic potential  in  these animals. There is  also a high degree
of correlation between the results of  this test and the human maximization
test (Klecak 1977).
    The  Split  Adjuvant Technique   This  test  makes use of  the finding
that  intradermal  challenge of FCA  beneath the  site of  topical application
of allergen  potentiates sensitization. Sensitizing  doses are administered to
the back  (to which dry ice has been applied) of  the guinea pig on days 0,
2, 4,  and 7. On day 4 FCA is injected intradermally twice  into the sensiti-
zation site.  The challenging dose is applied  to the back on day  20, and the

-------
                                   109
reaction  is  read on  days  22,  23,  and  24.  Klecak  (1977) believes  this
method is particularly  effective with  weak  sensitizers, but  it  is time-
consuming.
    The Buhler Test   In this test, the chemical is applied to the flank and
held there with an occlusive patch for 6 h on days 0, 7, and 14. On day 28
the challenging dose is applied to the same site for 24  h. The skin reaction
is read on days 29, 30, and 31,  and the results are compared with those of
negative and/or vehicle controls.  The  test  is fairly  expensive, but  it
corresponds to the practical needs of  a predictive test method  (Klecak
1977).
    The Open Epicutaneous Test   In the open epicutaneous test, guinea
pigs are immunized on the  flank three or five times weekly for  four weeks,
always using the same site, with undiluted compound and/or  its  progres-
sive dilutions. The reactions are read once a week or 24 h after each appli-
cation. On days  21  and 35 the minimal irritant (previously determined)
and some lower  concentrations  are applied to the contralateral flank. A
concentration is considered to be allergenic when at least one animal  in a
group shows a positive reaction with nonirritant concentrations. The test is
suitable for testing simple chemicals as well as finished products, and the
topical  application methods  used  are  similar  to  common use  (Klecak
1977).
    Technical  details and original references for the skin sensitization tests
can be found in Klecak (1977).

3.3.3.2 Respiratory sensitization
    Although  experimental induction of respiratory  hypersensitivity  has
been readily accomplished  with  natural  antigens (such as animal danders
and pollen) and  with purified proteins,  the induction of such a  response
with hapten-protein complexes has been more difficult (Karol et al. 1978).
    Karol  et al.  (1978),  however, were successful in producing respiratory
hypersensitivity  in guinea pigs with repeated exposure to  aerosols of
hapten-ovalbumin conjugates. The  animals were  exposed  to  aerosols of
p-azobenzenearsonate  ovalbumin and  p-tolylureido ovalbumin in a Plexig-
las  inhalation chamber.  Respiratory hypersensitivity was evaluated on the
basis of changes in respiratory rate, a sensitive indicator of the response.
The hypersensitivity  response was  hapten-specific,  as were most of the
antibodies produced  during the course of the study. The guinea pig model
was used in a later study in which respiratory hypersensitivity was induced
with hexyl  isocyanate-ovalbumin  aerosol  (Karol  et  al. 1979).  Hapten-
specific antibodies were  detected as well as hapten-specific IgE antibodies.
Karol  et  al.  (1978)  recommend  their  method for  screening industrial
chemicals for their potential induction of respiratory hypersensitivity. In
more recent studies using the guinea pig  model, Karol  (1983) and  Karol et
al.  (1981) demonstrated that the immunologic response to  toluene diiso-
cyanate following inhalation exposure is dose dependent and that respira-
tory tract hypersensitivity can be induced by dermal contact.

-------
                                  110
3.4 TIER TESTING
   Tests suitable for assessing the various cellular and humoral functions
of the immune system have been discussed in the previous sections. If the
evaluation of immunocompetence is to become a permanent fixture in rou-
tine toxicity test protocols, then the selection of the proper combination of
tests for use  in these protocols becomes an important consideration. First,
it seems obvious that both humoral and cellular functions should be tested.
Second,  it is important  that tests be  chosen that  will efficiently  and
economically provide maximum information.
   Investigators  in  the  field  of immunotoxicology  have  made their
independent recommendations for the types of immunology procedures to
be included in routine toxicity studies or  in tier testing protocols. Sugges-
tions  of Moore  and  Faith (1976), Luster and Faith (1979), Speirs  and
Speirs (1979),  Dean et  al. (1979b),  White et al. (unpublished  observa-
tions), and Luster et al. (1982b) are presented in the following sections.

3.4.1 Moore and Faith (1976)
   Moore and  Faith, interested in the developing immune system, recom-
mended the rat and mouse as the species  to be tested, initially at weaning.
The following methods were selected because they were generally  accepted
by immunologists and were within the capabilities of many laboratories.

 A. Tests for humoral (or B-cell) immune function

      1.  Hemagglutination techniques using
          a.  lipopolysaccharide (thymus independent)

          b.  SRBC (thymus dependent)
         In a minimal experimental design where only one antigen can be
         tested,  one should select the  SRBC test because both  T- and B-
         cells can be evaluated with this procedure.

      2.  Mitogen stimulation of thymus and spleen cells with lipopolysac-
         charide (a B-cell stimulator)

 B. Tests of cellular (or  T-cell) immune function

      1.  Radiometric ear test

      2.  Mitogen   stimulation  of  thymus  and   spleen  cells  with phy-
         tohemagglutinin and Concanovalin A, T-cell stimulators.

Moore and Faith further recommended that:

  1. Chemicals that depress immune function in adults should be evaluated
    in the fetus/neonate.

  2. Routine teratology studies should include an evaluation of the thymus
     and spleen  (organ weights and histologic examination). If  effects are
    observed on maturation of the tissue, then one should consider subse-
    quent experiments that specifically assess the immune response.

-------
                                 Ill
3.4.2 Luster and Faith (1979)
   These  investigators outlined the immune tests used in their laboratory
at the National Institute of Environmental  Health Sciences.  The initial
assessment of the effects of a particular chemical on the immune system is
generally  made in a routine toxicity study. At necropsy  Luster and Faith
perform the following procedures:
 A. Examination of lymphoid organs  (thymus, spleen, lymph  nodes, and
    bursa of Fabricius) and perhaps the adrenals
      1. Gross examination
      2. Histopathology
      3. Lymphoid organ to body weight ratio
      4. Quantitation of viability for the determination of cytotoxicity
 B. Hematology
      1. Differential
      2. Leukocyte count
 C. Immunology
      1. Serum immunoglobulin concentrations—electrophoresis
      2. Lymphocyte transformation (PHA and LPS)
      3. T-cell rosettes (not practical in mice or rats)
      4. Surface antigens on lymphocytes  (theta antigen in mice)
 D. Clinical signs, infections, tumors
Further assessment of the immune system can be  accomplished  with the
following:
 A. Tests of cell-mediated immunity
      1. In vitro transformation of lymphocytes by concanavalin A
      2. In vitro mixed lymphocyte cultures
      3. Radiometric ear test
      4. Assay for migration inhibition factor
 B. Tests of humoral-mediated immunity
      1. In vitro  transformation of  lymphocytes  by  dextran sulfate  or
        pokeweed mitogen
      2. B-cell surface antigens
          a. EAC rosettes
          b. Fluorescent labeling of B-cell immunoglobulins

-------
                                 112
     3.  Quantitation of serum  immunoglobulins by  radial immunodiffu-
        sion
     4.  Antibody response to  T-dependent  and T-independent antigen
        using the plaque assay
C.  Tests of indirect immunity
     1.  Macrophage
         a.  In vivo uptake of aggregated 125I-BSA or colloidal carbon
         b.  In vitro or in vivo determination of phagocytosis and intra-
             cellular killing of bacteria
         c.  Cytotoxicity and proliferation
     2.  Hormone concentrations
         a.  Cortisol/cortisone
         b.  Estrogens or androgens
     3.  Complement (measurement of C-dependent hemolysis)
     4.  Host resistance to infection
     5.  Resistance to tumor challenge
     6.  Endotoxin hypersensitivity
   The  tests were selected by Luster and Faith for their reliability and for
their ability to detect subtle differences in immune status.
3.4.3 Speirs and Speirs (1979)
   The  Immunotoxicology Program at the National Center for Toxicologi-
cal Research has devised  a model for  in vivo assessment of the  effects of
toxic agents on immunocompetence. This procedure is summarized as fol-
lows:
 1.  Immunize mice with an antigenic mix consisting of:
               Diphtheria toxoid
               Pertussis vaccine
               Polyvalent pneumococcal polysaccharide  antigens
               Tetanus toxoid
 2.  Determine levels of antibody  and other plasma  proteins using radio-
    immunoassay procedures.
 3.  Assess functional capabilities of macrophages, lymphocytes, and other
    blood cells using cytochemical analysis.
 4.  Develop an immune profile for each  toxicant tested. Compare with
    profles of known immunosuppressants and other toxic agents.

-------
                                 113

 5. Compare the immune profiles  of mouse and man with respect to a
    particular toxicant.

   In  their model Speirs and Speirs (1979)  use licensed  vaccines  for
immunization,  and commonly used  pharmaceutical agents  as  prototype
immunosuppressants so that  animal  and human data can be  more readily
compared.  Risk assessment is based on comparison of  immune profiles of
animals treated with known immunotoxicants with those of animals treated
with a potential toxicant.

3.4.4 Dean et al. (1979b)

   This group at Litton Bionetics has been involved for the last  few years
in developing and refining assays of humoral and cell-mediated immuno-
competence. They have proposed a tier approach  for screening for immu-
nological effects  (tier I) and to help determine  which mechanisms  are
responsible for  the effects observed  (tier II). The assays were selected on
the basis of relevance to the human  experience, cost, reproducibility, ease
of performance,  and application to  routine toxicology studies.  The tests
included in tiers I and II are  listed in Table 3.7.
   After 90 days of a chronic or subchronic toxicity study,  15 animals  are
removed for immunology studies.  Testing can then proceed in either a
hierarchical or  nonhierarchical fashion. Hierarchical testing assesses chem-
icals in a stepwise procedure following administration of a dose of the test
agent with minimal or no overt toxicity. The testing begins with the sim-
plest  or least  expensive  assays  and continues through the  more time-
consuming and expensive in  vitro tests. It  is assumed that those  chemicals
with deleterious effects would be  detected with the initial assays and would
require further testing only when  information  regarding mechanism is
required.
   In nonhierarchical testing the assays (tier I) would  all be  performed at
once. This  procedure  is useful when  the number of compounds to be tested
is small or  when the data  are needed quickly.
   If no immunological effect is observed in tier  I, no further studies  are
performed. If a positive effect (suppression or enhancement) is observed,
the tier II assays are performed.

3.4.5 White et al. (Unpublished Observations)
   This group  at the Medical College of Virginia has developed  a pro-
cedure to investigate the effects of a given  chemical on  the immune system
against a background of standard toxicological procedures.
   An acute study to determine  the  LD50 is performed first.  Next,  a sub-
chronic 14-day  range-finding  study  is performed, usually  at  1/10  and
1/100 the LD50. The parameters  measured in the  14-day subchronic study
are:

-------
                                        114
         Table 3.7. Testing Approaches for Evaluating the Immunobiologic Effects of
                   Food Additives, Drugs, and Environmental Chemicals
                              Tier I: Screening
                                       Tier II: Mechanism
Clinical laboratory
Pathology
Cell-mediated
  immunity
Humoral immunity
Susceptibility
  Immunepotentiation
  Tumor challenge
Hematology
Blood chemistry
Urine

Lymphoid organs
 •  Relative weight
 •  Cell viability
 «  Histology

Delayed hypersensitivity (DHS)
 •  T-dependent antigen (keyhole
   limpet hemocyanin, bovine
   gammaglobulin, tetanus toxoid)
   toxoid study using the radio-
   metric assay
Lymphocyte proliferation
 •  T- and B-cell mitogens
 •  Mixed leukocyte culture
Ig  levels
Specific antibody tier
 •  T-dependent antigen
Jerne plaque assay
 •  T-dependent antigen

Tumor challenge - TO,^
Specific blood or tissue levels
  of compound
Hormone levels

Cell surface markers
 • T-cells
 - B-cells
 • Null cells
 • Fc-cells

DHS
 • T-independent antigen
Lymphocyte proliferation antigen
Helper cell function
Macrophage function
Mishell-Dutton assay
Local product of antibody
Specific antibody tier
 • T-independent antigen (SIII)
Cytotoxicity

Bacterial challenge
Virus challenge
Source: Adapted from Dean et al. (1979b).
 A.  Standard toxicology
       1.  Body weights6
       2.  Necropsy—gross pathology*5

       3.  Organ weights'" (brain,  liver, spleen, lungs, thymus, kidneys, and
           testes)

       4.  Hernatologyb (leukocyte counts, hematocrit, hemoglobin)
       5.  Clinical chemistries3  (glutamic pyruvic transaminase, lactic dehy-
           drogenase, blood urea nitrogen)
       6.  Blood coagulation15
  B.  Immunotoxicology

       1.  Cell-mediated  immunity0—delayed-type  hypersensitivity  response
           to sheep erythrocytes (footpad assay)

-------
                                  115
     2.  Humoral immunity*—spleen antibody-forming  cell response to
         sheep erythrocytes (IgM plaque-forming cells)
Three  sets of mice  are  used; the superscripts a, b, and  c  designate the
group on which the test is performed.
   The 90-day subchronic study is then performed to demonstrate the tar-
get organ. The following variables are measured  during and at the end of
the 90-day study:
 A. Standard toxicology
      1.  Body weights, twice weekly
      2.  Fluid consumption
      3.  Necropsy3—gross pathology
      4.  Organ  weights8—brain, liver, spleen, lungs, thymus, kidneys, and
         testes
      5.  Urinalysis—pH, glucose, protein, bilirubin, blood, and urobilino-
         gen
      6.  Hematology"—hematocrit,  erythrocytes, leucocytes, differential,
         platelets, and hemoglobin
      7.  Coagulation3—prothrombin time,  activated partial thromboplas-
         tin time, and fibrinogen
      8.  Clinical  chemistries'"—sodium, calcium, potassium, serum  glu-
         tamic pyruvic transaminase, serum glutamic oxaloacetic transam-
         inase,  lactic  dehydrogenase, total protein, albumin, creatinine,
         and blood urea nitrogen
 B. Hepatic microsomal mixed functional oxidase parameters
      1.  Liver weight
      2.  Microsomal protein
      3.  Glutathione levels
      4.  P-450 content
      5.  Cytochrome b5
      6.  Aminopyrine N-demethylase
      7.  Aniline hydroxylase
 C. Cell-mediated immunity
      1.  Delayed-type hypersensitivity response to sheep erythrocytesb
      2.  Spleen cell  response to Concanavalin Ab
 D. Humoral immunity
      1.  Spleen  antibody-forming   cell  response to  sheep  erythrocytesb
         (footpad assay)

-------
                                 116

     2. Spleen cell response to lipopolysaccharide
     3. Basal levels of IgM, IgG, and IgA
 E.  Functional activity of the reticuloendothelial system
     1. Vascular clearance rate of sheep erythrocytesd
     2. Organ uptake of sheep erythrocytesd
     3. Chemotaxis, adherence, and phagocytosis  of  peritoneal exudate
        cells6
 F.  Bone  Marrow3
     1. DNA synthesis
     2. Granulocyte—monocyte stem cells
Five sets  of  mice are used; superscripts a,  b, c, d, and  e designate the
group on which a test is performed.
    White et  al. demonstrated that these assays,  used to evaluate immune
status, are quantifiable and reproducible and that such assays can be easily
incorporated  into a routine toxicology study.
3.4.6 Luster et al. (1982b)
    In  November  1979,  the National  Institute of  Environmental Health
Sciences held a Consensus Meeting to develop a list of relevant immunolo-
gic  parameters suitable  for evaluating chemically induced immunotoxici-
city. The  group of more than  35 immunologists and toxicologists of varied
experience and expertise  prioritized the following scheme of immunologic
parameters and methods considered necessary to measure  immunotoxicity.
These tests are to be performed in the order listed.
Chemical  for immunological evaluation —>•  Chronic or Subchronic Study
                                I            I
                          Immunology Screening Panel
 1.  Pathotoxicology-hematology,  liver,  chemistry,  serum proteins, lym-
    phoid organ weights,  and histology.
 2.  Host  resistance-tumor challenge and infectious agent challenge.
 3.  Radiometric  delayed hypersensitivity.
 4.  Lymphoproliferative responses-PHA, Con A, LPS, and MLC.
 5.  Humoral immunity-Ig levels, specific antibody titer,  plaque-forming
    cell assay.
 6.  Macrophage  function assays.
 7.  Bone  marrow progenitor cell assays.

-------
                                  117
           Data Evaluated
             I           I
    No immuno-      Positive immune-   —>•    Study mechanism of
    logical effect        logical effect            immunological effect
             I           I
           No further study

3.5. SUMMARY

   Reports  of  immunotoxic effects of environmental chemicals, some  at
clinically subtoxic  levels, have stimulated interest in tests of immune func-
tion  and in the inclusion of  such  tests in routine toxicity  test protocols.
These  procedures   provide  information related to  direct  injury to the
immune  system and,  equally important, are capable of  detecting subtle
alterations in the  immune  response. This system  could be an  extremely
sensitive indicator  of the general toxicity of certain chemicals.
   The preceding  review  was undertaken to examine some of the test
methods capable of detecting chemically induced alterations in the specific,
as well as the nonspecific, immune response.
   Tests of cell-mediated,  humoral, and indirect immune functions were
described, and  examples of the use of these tests in  the assessment of the
potential immunotoxic effects of chemical were  presented.  A few tests that
are useful for  detecting the sensitizing  effects of chemicals  were also
included.
   There was no evidence to indicate that any particular test or battery of
tests would be  more suitable for evaluating a particular class of chemicals
than others.  However, there are indications that certain  chemical classes
may be selectively toxic to  the humoral immune system, while others may
affect   cellular immunity.  For  example,  lead,  cadmium,  mercury,
polychlorinated biphenyls, and DDT impair primarily the humoral immune
response,  while  organochlorines   impair   primarily  the  cell-mediated
response. These observations, however, should not preclude the assessment,
in toxicity testing,  of both humoral  and cell-mediated immunity.
   Information related to  the costs of these tests was not found in the
literature. Generally,  immunology procedures are regarded as being  fairly
expensive; but the  introduction of in vitro techniques may enhance the pos-
sibility of large-scale testing at more reasonable  costs.
   Recommendations  for the types of  immunology tests that  should  be
included  in  routine   toxicity  studies  or  in   tier-testing schemes   were
presented.

-------
                                 118
3.6 GLOSSARY

ADJUVANT. A  substance  administered with or before an antigen that
    nonspecifically  enhances  or  alters  the  cell-mediated or  humoral
    immune  response  to that  antigen  (Herbert and  Wilkinson  1971,
    Sharma 1981b).

ALLERGEN.  An  antigenic substance capable  of  eliciting an  allergic
    response or an allergic  state (Herbert and Wilkinson  1971,  Sharma
    1981b).

ALLERGENIC. Causing allergy (Sharma 198 Ib).

ALLERGY.  (1)  Hypersensitivity; (2) heightened reactivity to  antigen,
    according to  Gell  and Coombs,  which is synonomous with immunity
    (Herbert  and Wilkinson  1971);  (3)  abnormal  or  harmful  reaction
    which includes delayed, immediate, or contact hypersensitivity and
    may  also include a number of reactions that have  no  immunological
    basis (Platts-Mills  1982).

ALLOANTIGEN.  Cell-surface  antigens  in  different strains of the same
    species (Wust, unpublished observations).

ALLOANTISERA. Mouse  or  rat antisera  prepared by immunizing one
    inbred strain with  cells from another strain (Wust, unpublished obser-
    vations).

ALLOGRAFT. A  syngeneic graft,  or  one  that  has  been  exchanged
    between two  genetically dissimilar members of the same species, such
    as of two different inbred strains (Herbert and Wilkinson 1971).

ANAPHYLAXIS.  An immediate hypersensitivity  or  Type  I  reaction
    resulting from the administration of  antigen to a primed subject. The
    reaction is caused  by the release of vasoactive substances, such as his-
    tamine  and leukotrienes, when antigen combines  with antibody on cell
    surfaces. The reaction may be generalized or localized. The  systemic
    reaction can  be violent  and can cause death in  minutes (Herbert and
    Wilkinson 1971; Wust, unpublished observations).

ANTIBODIES. Immunoglobulins  that are produced in  the body by cells
    of the lymphoid series, particularly plasma cells, in  response to stimu-
    lation by  antigen that are  capable  of specific  combination  with
    antigen (Herbert and Wilkinson  1971, Sharma 1981b).

ANTIBODY TITER.  In serological  reactions, a measure of units of anti-
    body in an  antiserum  per  unit volume of the  undiluted serum.  To

-------
                                  119

    determine the titer, the serum is serially diluted, antigen is added, and
    the end point is determined (Herbert and Wilkinson 1971).

ANTIGEN. A substance that  elicits a specific immune response,  either
    cell-mediated  or  humoral, when  introduced into the tissues  of an
    animal. An antigen may also induce specific immunological tolerance
    (Herbert and Wilkinson 1971). Can  be a protein, nucleic acid,  car-
    bohydrate, or lipid (Wust, unpublished observations).

ANTIGEN PROCESSING. Macrophage-antigen interaction in which the
    macrophage  localizes the  antigen  and  brings  it  closer to  the
    antibody-forming cell. This interaction  may also  result in  enhanced
    immunogenicity of the antigen (Koller 1981).

ANTIGEN RECOGNITION SITE. The site on the surface of a lympho-
    cyte which may  be identical to an antibody-combining site. Can react
    specifically with  antigen and thus initiate humoral or  cell-mediated
    immune resposes (Herbert and Wilkinson 1971).

ANTIGEN-ANTIBODY  REACTION.  The  interaction  between  an
    antigen and  the  antigen-binding receptor on the  surface of a small
    lymphocyte  which stimulates  the  differentiation and  proliferation of
    the  cell,  leading to  the  production  of a progeny of  effector cells
    (which execute the immune response) and of memory cells (which are
    responsible for the enhanced  or secondary response following a  subse-
    quent  encounter with the same antigen) (Miller  1975). The immune
    system responds  to an  unlimited  variety of antigens,  some of them
    normally not present in nature (Polak 1977).

    There  is now evidence for the existence, on  the cell membrane of
    small  lymphocytes,  of  antigen  recognition  units that  can  bind
    antigenic determinants and of Ig molecules with specificity similar or
    identical to that of secreted antibody (Miller 1975). There is also evi-
    dence  to show that the Ig determinants  on the cell membrane are the
    antigen-binding  receptors. Apparently   both  T-  and  B-cells  have
    antigen-binding receptors on  their surface, but they are of such low
    density on nonimmune cells that they cannot  be detected by routine
    antigen-binding techniques.

ANTISERA. Serum,  from an animal, that contains antibodies  specific to
    a given antigen [e.g. anti-ovalbumin (Herbert and Wilkinson 1971)].

ARTHUS   REACTION.  An  inflammatory  reaction  that  follows the
    administration of antigen to an animal that already possesses precipi-
    tating  antibody to that antigen. The reaction is caused by the forma-
    tion, in the presence of complement, of antigen-antibody  complexes
    that adhere to the endothelium of vessels and become surrounded by

-------
                                 120

    fibrin,  platelets and neutrophils. This is followed by plugging of the
    vessels with thrombi and  exudation  of neutrophil-rich fluid into the
    surrounding tissues.  The  reaction  is  manifested  as an  ulcer  that
    appears several  hours  after intradermal injection of antigen into a
    primed animal (Herbert and Wilkinson 1971).

B-LYMPHOCYTE.  B-cells arise from stem cells in the bursa of Fabricius
    in birds and in the bursal equivalent in mammals (thought to be gut-
    associated  lymphoid  tissue or bone marrow) (Faith et  al. 1980). B-
    cells mature independently of thymic influence (Koller  1979a). They
    circulate and rest  temporarily in the thymus-independent areas of the
    spleen (follicles  and  peripheral regions of the white pulp) and lymph
    nodes  (follicles and medulla)  (Faith et al. 1980), where they consti-
    tute 10 to 20% of  the nucleated free cells. They are not present in the
    thymus (Wust, unpublished observations).
    B-cells are stimulated by antigen to become antigen-specific cells that
    proliferate and differentiate into plasma cells, which secrete antibody
    or become memory B-cells.

BACILLE CALMETTE-GUERIN (BCG). An attenuated, living bovine
    strain  of Mycobacterium tuberculosis used to vaccinate against tuber-
    culosis and leprosy (Herbert and Wilkinson 1971).

BLASTOGENESIS. Stimulation of cell (lymphocyte) blast formation by
    an antigenic substance (Sharma 1981b).

BLASTOGENIC.  Capable of  producing  cell blasts.  In  immunology,
    antigenic substances capable of producing cell blasts (Sharma  1981b).

BONE MARROW. The principal source of the lymphoid  stem cell in
    postnatal  life  and the only general  hemopoetic  tissue in  the healthy
    adult.  The vascular compartment of the bone marrow is  occupied by
    the vessels that  distribute nutrients  throughout the marrow cavity and
    the hematopoetic  compartments   (red   marrow),  which   contain
    erythrocytes, granulocytes, lymphocytes, megakaryocytes, monocytes,
    macrophages,  plasma cells, mast cells, and their precursor stem cells.
    In adult animals much of the red marrow is replaced by fatty tissue
    and becomes yellow marrow. The bone marrow does not contain  signi-
    ficant numbers of fully  competent  thymus-dependent lymphocytes.
    However,  theta-positive  incompletely developed  T-cells  and  more
    mature T-cells are demonstrable. The marrow also contains significant
    numbers  of  B-lymphocytes  (fully  differentiated  immunoglobulin-
    producing  lymphocytes) and plasma cells (Bloom and Fawcett  1969,
    Herbert and Wilkinson 1971, Park and Good 1974).

BURSA-EQUIVALENT. A  poorly identified tissue  in  mammals,  con-
    sidered to be equivalent to the bursa of Fabricius in birds, which gives

-------
                                 121

    rise to B-cells (Luster et al. 1982b). The bursa is a saclike lymphoep-
    ithelial  structure arising as a dorsal diverticulum from the cloaca of
    young birds that contains  many lymphoid follicles in which lympho-
    poiesis proceeds until the structure involutes about the time of sexual
    maturity (Herbert and Wilkinson 1971). The lymphocytes are  B-cells
    that  migrate  from  the  bursa  to  colonize  the  secondary lymphoid
    organs (Wust, unpublished observations). The bursa is associated with
    humoral immunity,  demonstrated by  the  fact that bursectomized
    chickens fail to  make antibodies to a variety  of  antigens and lack
    plasma  cells  and germinal centers  in their lymphoid  tissues (Herbert
    and Wilkinson  1971).

CELL-MEDIATED IMMUNITY (CMI). The type of specific immunity
    that  is  effected  by  specifically sensitized  lymphocytes  and  can  be
    transferred with  those cells. (Vos 1977, Faith et al. 1980). The main
    categories of cell-mediated immunity  include: classical cell-mediated
    protective immunity,  which is mainly effective against  protozoal, viral,
    fungal,  and  some  bacterial  infections; delayed hypersensitivity skin
    reactions to extracts or whole suspensions of organisms, also identified
    as tuberculin hypersensitivity;  chemical contact sensitivity;  allograft
    rejection;  immunological surveillance  to  tumors;  and certain  organ-
    specific autoallergic  diseases,  such as thyroiditis, encephalomyelitis
    (following  rabies vaccination),  adrenalitis,  and orchitis (Turk  1975,
    Faith et al. 1980).

CELL SURFACE MARKERS. Antigens or alloantigens on the surface of
    lymphocytes  (IgD and IgM on  B-cells; theta and TL antigens on  T-
    cells;  and  Ly  alloantigens on  both T-  and B- cells) (Luster  et  al.
    1982b).

CF. Chemotactic factor.

CHEMOTACTIC  FACTOR.  A  product of  sensitized  T-lymphocytes
    exposed to antigens that direct  the migration of leucocytes to sites of
    inflammation (Allison 1982).

CLINICAL ALLERGY.  A disease or diseases  in which  hypersensitivity
    plays  a part, such as contact hypersensitivity, hay fever, and asthma.

CLONAL THEORY. The theory of F. M.  Burnet (1959), who believes
    that if a cell  comes in contact with an  antigen, and if any part of that
    antigen happens  to fit the antigen-binding site of the cell, the cell is
    stimulated to divide and  increase its production of that particular glo-
    bulin. Because  each region of a given antigen may more or less fit a
    number of different antigen binding sites, a considerable array of dif-
    ferent antibodies is likely to  be produced, directed toward the  several
    antigenic determinants of the antigen molecule.

-------
                                 122

COMPLEMENT.  An  enzymatic  system  of  11   serum  proteins,  9
    components,  which  act  in  a  specific  sequence,  or cascade.  The
    complement system is activated by many antigen-antibody reactions
    and  is essential  for  antibody-mediated immune  hemolysis  and  bac-
    teriolysis.  Complement participates in other biological reactions  such
    as  phagocytosis,  opsonization, chemotaxis,  and immune cytolysis
    (Park and Good 1974, Herbert and Wilkinson  1971).

    The complement proteins have  been named according to the numeri-
    cal order  of their role in the immune reaction, except for  C4, which
    was named before the sequence was described (Roller  1981).

CONCANAVALIN  A.  A mitogen derived  from Canavalis ensiformis,
    selective to immature T-cells (Sharma 1981e).

CONTACT DERMATITIS.  Delayed  hypersensitivity reaction of the skin
    resulting from exposure to chemical allergens, including environmen-
    tal and industrial chemicals, drugs and cosmetics, and the catechols of
    poison oak and poison ivy (Ligo and James 1974, as reported in Gigli
     1982).

CYTOTOXIC. Capable of producing cell lysis (Sharma  198le).

CYTOTOXIC FACTOR. A  lymphokine, released by a  sensitized lympho-
    cyte upon contact with antigen, that is capable of producing cell lysis
    (Sharma  1981c).

DELAYED HYPERSENSITIVITY.  The state mediated by lymphocytes,
    which is evident only when lesions appear about  24 h after contact of
    the  primed subject with antigen. This reaction is a  manifestation of
    cell-mediated immunity and can be transferred to another animal with
    cells (Herbert and Wilkinson 1971).

DIRECT PFC.  Plaque-forming  cells  producing antibody, primarily  IgM,
    that can  fix complement and  lyse cells directly (Jerne and Nordin
     1963).

EFFECTOR T-CELLS.  Cells that, in delayed-type hypersensitivity, pro-
    duce biologically active products called lymphokines, following contact
    with specific antigen (Luster et al. 1982b).

FREUND'S COMPLETE ADJUVANT (FCA).  A water-in-oil emulsion
    adjuvant  in  which  killed, dried  mycobacteria (usually  M.  tubercu-
    losis) are suspended in the oil phase (Herbert and Wilkinson 1971).

GRAFT-VS-HOST REACTION.  The reaction  of a  graft  containing
    immunologically competent  cells  against the  tissues  of a  genetically
    different host (Herbert and Wilkinson  1971).

-------
                                 123

HAPTEN. A compound of low molecular weight, not antigenic itself, that
    can be  combined with a protein  carrier to form a product having
    antigenic properties and can thus  elicit antibody  formation (Sharma
    1981c).

HELPER T-CELLS. T-lymphocytes that, through the release of soluble
    factors,  activate B-cells and  amplify  their responses  against many
    antigens (Sharma 1981b, Roller 1981).

HEMAGGLUTINATION.  The clumping of a cell suspension  by  specific
    antibody (Gell and Coombs  1975).

HEMOLYSIS.  The bursting of  red  blood cells.  In immunology,  this
    occurs when complement is  added  to the hemagglutinin reaction (Vos
    1977).

HISTOCOMPATIBILITY ANTIGEN. An  isoantigen carried on the sur-
    face of  nucleated cells of many tissues which may induce an immune
    response that causes the rejection of a graft from one individual to
    another of  the  same species whose tissues do  not carry that  antigen
    (Herbert and Wilkinson 1971).

HISTOCOMPATIBILITY LOCUS. The locus on a chromosome at which
    the genes that  determine the histocompatibility antigens are  located
    (Herbert and Wilkinson 1971).

HUMORAL IMMUNE RESPONSE.  The type of specific immunity that
    operates through  antibody-producing  cells and  is  transferable  by
    serum (Vos 1977,  Faith et  al. 1980). The antibodies (immunoglobu-
    lins) are present in the plasma, lymph, and tissue fluids of the body
    and may become attached to cells (Herbert and Winkinson  1971).

    Upon  antigenic stimulation,  a   small  number  of B-lymphocytes,
    specific  for that antigen, proliferate and differentiate into plasma cells
    that secrete antibody in a primary response or serve as memory cells
    for the secondary response (Faith et al.  1980).

HYPERSENSITIVITY. State of  abnormal susceptibility  or  sensitivity,
    usually in reference to allergy (Sharma 1981b).

IMMEDIATE HYPERSENSITIVITY. Antibody-mediated  immune reac-
    tion that results from the release  of histamine and other vasoactive
    substances that can be  transferred from one individual  to  another by
    serum alone. The antibody is usually of  the IgE type, which fixes par-
    ticularly to  mast cells. The reaction  usually  appears  within a  few
    seconds  to  30 minutes after contact with  antigen  (Herbert  and
    Wilkinson 1971).

-------
                                 124
IMMUNE RESPONSE.  The  manifestation  of  the interaction of  the
    organs and  cells of the immune system, which  is expressed through
    protective mechanisms  such as  rejection of allografts or delayed
    hypersensitivity  (cellular  immunity)  or  production  of  antibodies
    (humoral immunity).  The responses of organisms to  antigens  are of
    two types: innate (or nonspecific) immunity and acquired (or specific)
    immunity (De Bruin 1976).

IMMUNE   SUPPRESSION.  Suppression  of  the  immune  response
    (Sharma 1981c).

IMMUNE SYSTEM. The organization of organs and cells which interact
    at various levels to effect the responses of the body to foreign entities
    (or antigens) (Miller 1975).

IMMUNIZATION. Administration of  antigen  to  produce an immune
    response to that antigen. Administration of either antigen to produce
    active immunity or antibody to produce passive immunity to protect
    against  the harmful  effects  of antigenic  substances  or  organisms
    (Herbert and Wilkinson 1971).

IMMUNOASSAY.  Analytical  techniques that  utilize antigen-antibody
    type reactions (Sharma 198 Ib).

IMMUNOCOMPETENCE. The capacity of the body to  cope with infec-
    tion and malignant growths. Immunocompetence has been character-
    ized  as  (1) the establishment of barriers to maintain integrity  and
    protect the body from the environment; (2) the  capacity to recognize,
    neutralize, isolate, and reject foreign agents; (3)  diverse reactions such
    as  inflammation, granulomatous reactions, cell-mediated and humoral
    responses,  and  healing and repair  processes; and (4)  a  process of
    "adoptive immunization" in which cells react in  a  greatly  augmented
    or  more effective  manner  to sequential  exposures  to a particular
    pathogen (Speirs and Speirs 1979).

IMMUNOGLOBULINS. Proteins, each of which is made up of two light
    and two heavy  chains  that are usually linked  together by disulfide
    bonds.   The various  immunoglobulins differ  from  each  other in
    primary structure  even  though  some amino   acid  sequences   are
    remarkably constant throughout phylogeny. All  antibodies  are immu-
    noglobulins  although  all immunoglobulins may  not be  antibodies.
    Antibodies  are produced in the body by  plasma cells in response to
    stimulation  by antigen, and they are capable of specific combination
    with  antigen. Five  major classes of antibody proteins, or immunoglo-
    bulins,  have been identified in man: IgG, IgA,  IgM, IgD, and  IgE
    (Herbert and Wilkinson 1971).

-------
                                  125

    The heavy and light chains of the immunoglobulins are symmetrically
    arranged, each consisting of a variable and a constant region. Accord-
    ing to Goldstein et al. (1974):
    In the constant region of a  given  immunoglobulin class,  the polypep-
    tide chain is  the  same  (or  nearly the  same)  in  its  amino acid
    sequence. This similarity is  common  among various species and  is
    found even in primitive organisms. In  contrast, the variable region  is
    composed of a variety  of different  amino acid sequences, while the
    total number  of amino acid residues remains the same.  The antigen
    binding sites  are apparently located in  the  variable portions of the
    light and heavy chain. Generally,  lymphoid cells  differentiate in such
    a way that a  given cell produces only a  single immunoglobulin while
    all  lymphoid cells together produce thousands of different immunoglo-
    bulins.

IMMUNOLOGICAL  TOLERANCE.  An  acquired  state  of  specific
    unresponsiveness to an antigen (Moreno 1982).

IMMUNOSUPPRESSION. Inhibition of  the  stimulation of an  immune
    response by an antigen by  a physical, chemical, or biological agent
    (Herbert and Wilkinson 1971).

INDIRECT PFC. Plaque-forming cells that  produce  antibody that either
    is  insufficient  for or is incapable of fixing  complement unless it  is
    complexed with antiglobulin  antibody.  Revealed in  the PFC assay
    when developing serum is added to the plate (Dresser 1978).

INTERFERON. A protein (lymphokine) released by cells in response to
    virus infection. When taken up by other cells, interferon inhibits the
    replication of  viruses within them. Interferon is also released after the
    injection of bacterial endotoxin (Park and Good 1974,  Herbert  and
    Wilkinson 1971, Sharma 198la).

IgA. The major immunoglobulin component of secretions in the respiratory
    and gastrointestinal tracts and the second most abundant immunoglo-
    bulin class (10% of total) (Park  and Good 1974, Rowe 1975), but
    present in relatively  low concentrations in the serum. The molecule  is
    synthesized locally  in plasma cells  of the submucosa;  the  secretory
    component is  synthesized in the  epithelial cells  (Roller  1981).  The
    IgA molecules of external secretions are  primarily 11S, while those of
    the serum are  mainly 7S. Although IgA does not fix complement,  it
    may activate  C3 by alternate pathways. IgA  is active  against viral
    and bacterial  antigens at the sites  of secretion,  but the role of the
    antibodies in protection is not  clear. It has been suggested  that IgA
    antibodies may activate the alternate  complement pathway and that
    they may be active  in promoting phagocytosis and intracellular killing
    of organisms by peritoneal macrophages.

-------
                                  126
IgD. Prominent surface-bound immunoglobulin in the newborn; constitutes
    only 0 to 0.2% of immunoglobulins of normal serum  (Roller 1981).
    Half-life about 3 days. Antibody  activity  against penicillin, insulin,
    milk proteins, diphtheria toxoid, and nuclear and thyroid antigens has
    been demonstrated in IgD molecules (Park and Good  1974).

IgE. The  immunoglubulin that  is associated  with  the  reaginic antibodies
    active in Type I (see Section 3.4)  allergic  reactions  (Bennich and
    Johansson  1971, Ishizaka and Ishizaka 1971).  IgE is produced by the
    plasma cells of the regional lymph nodes and mucous membranes of
    the respiratory  and gastrointestinal tracts (Roller  1981). The concen-
    tration of IgE in the serum is very small. It binds to tissue mast cells
    and blood  basophils, and exposure of these cells to a specific allergen
    (the antigen to which the IgE is directed)  leads to the release of his-
    tamine and other mediators. No protective role has been demonstrated
    for IgE  antibodies. Rowe (1975) postulates that IgE  antibodies may
    assist  in protection  against certain parasites  either directly through
    pharmacological mediators or  by the induction of an increase of vas-
    cular  permeability allowing freer  access of antibodies and cells from
    the blood.

 IgG.  The  most  abundant  (85% of  total)  immunoglobulin in man, syn-
     thesized primarily during the  secondary immune  response.  Has half-
     life  of about 23 days (Park and Good 1974, Rowe 1975, Faith et al.
     1980). It diffuses  readily into extravascular tissue, where it can react
    with bacteria,  parasites, and  viruses  (Park and  Good 1974),  and is
     responsible for long-lasting protection  to these agents  (Roller  1981).
     IgG is selectively  transported  across the placenta, and at birth  is the
     predominant immunoglobulin  in  the  circulation (Park  and  Good
     1974).

 IgM. Often  the  first immunoglobulin to  be detected following antigenic
    challenge and the most effective specific first line of defense  (Rowe
     1975, Faith  et al. 1980). The IgM antibodies constitute 5 to 10% of
    the total immunoglobulins in the blood (Park and Good 1974). One of
    the  characteristic properties of  IgM  is its  high molecular  weight
    (900,000). It has a serum half-life  5 to  6 days.

 RILLER LYMPHOCYTES. Special cytotoxic T-lymphocytes that partici-
    pate in the delayed hypersensitivity reactions (Sharma  198la).

 RININS.  Small  polylpeptides  that  are  released  from  a plasma  alpha-
    globulin, kininogen,  by proteolytic enzymes such  as kallikrein.  These
    enzymes are  released when  polymorophonuclear cells  phagocytize
    immune complexes. Rinins  are active as vasodilators  and facilitate the
    increase of  vascular permeability. They  also cause  contraction  or

-------
                                  127
    relaxation  of  smooth  muscle (Park  and Good  1974,  Herbert and
    Wilkinson 1971, Sharma 198la).

LIPOPOLYSACCHARIDES. Compounds in which lipid  is  linked  to
    polysaccharide. A common component of cell membranes.  In  immu-
    nology usually refers to components of O-antigen (endotoxin) complex
    of gram-negative bacilli such as E. coli,  Salmonella, and Bordetalla
    pertussis.  E.  coli lipopolysaccharide is a  thymus-independent antigen
    (Luster et al. 1978, Herbert and Wilkinson 1971).

LYMPH  NODE.  A kidney-shaped accumulation  of reticuloendothelial
    cells and  lymphoid tissue, connected  to  other  lymph nodes by lym-
    phatic vessels. These vessels enter the  node at various places but leave
    it only at the hilus (a slight indention  on the  side of the node) on their
    way to the thoracic or right lymphatic ducts. The lymph nodes act as
    filters  through which  foreign materials  must  pass.  These materials
    thus must  come  in contact with macrophages  and lymphocytes;  the
    lymph nodes are important centers for phagocytosis and  for the initia-
    tion and development of the cellular and humoral immune responses.

    The lymph node is divided into the outer  cortical and inner medullary
    parts.  The cortex consists primarily of dense lymphatic tissue which
    continues  into the medulla as medullary  cords.  In the mid-cortex  are
    thymus-dependent areas, which appear selectively depleted  of cells in
    neonatally  thymectomized animals.  The cortex  contains  lymphatic
    nodules about  1  mm  in  diameter, which develop and disappear in
    response to various  stimuli. These nodules  contain dendritic macro-
    phages which are capable of fixing antigens at their surface.  In  the
    embryo and  the immediately postnatal  stage  they  lack the central
    "germinal"  areas which are  thought   to  develop  in  response  to
    antigenic stimulation. In old age the germinal areas may  disappear.
    The medulla consists of the same cellular constituents as the  cortex,
    but the medullary cords rarely contain nodules (Bloom and Fawcett
    1969, Park and Good 1974, Herbert and Wilkinson 1971).

LYMPHOCYTES.  During  embryological development, lymphocyte  pro-
    genitor cells migrate in the prenatal yolk sac through the liver to the
    bone marrow, with lymphocyte maturation occurring  outside the bone
    marrow (Faith et al.  1980). Some cells enter the thymus, where they
    proliferate  and mature under the influence of thymic epithelial cells.
    A  small  number of lymphocytes (T-cells) are released  from  the
    thymus, circulate through the body, and rest temporarily in thymus-
    dependent areas of the  spleen and lymph nodes. Lymphocytes play a
    central role in both cellular and humoral  immunity, but these lympho-
    cytes consist of two distinct populations,  T-cells and B-cells, with  dif-
    ferent functions and properties (Vos 1977).

-------
                                 128
LYMPHOCYTE  TRANSFORMATION.  The  change  in  lymphocyte
    morphology when the  cells  are cultured in the  presence  of phy-
    tohaemagglutinin (or other mitogen)  or of an antigen to which they
    are primed. The changes include: increased size, increased cytoplasm,
    and  increased visibility  of the nucleoli. After about 72 h the altered
    cells resemble blast cells (Herbert and Wilkinson 1971).

LYMPHOID ELEMENTS. Cells of the circulating blood that originate in
    the  lymphatic  tissue  (e.g.,  lymphoctes  and probably  monocytes)
    (Bloom and Fawcett 1969).

LYMPHOKINES. Soluble nonantibody products  of lymphocyte activation
    by antigen and mitogens that are thought to act  as molecular media-
    tors of cellular immune responses (Morely et al.  1978). Lymphokines
    exhibit the following characteristics: (1)  produce  increased  vascular
    permeability  following  intradermal injection; (2) increase  tritiated
    thymidine uptake by lymphocytes in culture; and  (3) inhibit macro-
    phage migration in  vitro. Lymphokines retain biological activity after
    removal of  the  antigen  that  stimulated them.  They may also be
    involved in both expression and regulation of lymphoid cell activity,
    and they may represent a pathway for the expression of T-cell func-
    tion.  Some  lymphokines,  such  as the  migration inhibition  factor
    (MIF), the chemotactic factor  (CF), and the macrophage-activating
    factor (MAP), influence macrophage functions (Faith et al. 1980).

LYMPHOPROLIFERATIVE  RESPONSE.   Lymphocyte   proliferation
    induced by mitogens; the  basis for in vitro assays of T- and B-cell
    function (Luster et al. 1982a).

LYSOZYMES. Enzymes found  in  many body  fluids  which lyse certain
    bacteria,  primarily  gram-positive cocci, and enhance the bactericidal
    action of antibody and complement for gram-negative bacteria (Park
    and Good  1974, Herbert and Wilkinson 1971).

MACROPHAGE. Nucleated cells that  have been classified as part of the
    mononuclear phagocyte system, which replaces the concept of the reti-
    culoendothelial system (Vos 1977).  Macrophages, both fixed  and free,
    are  involved in and seem  to  be necessary for the initiation and the
    effector phases of humoral and cell-mediated immune reactions (Faith
    et al.  1980). The mechanisms of these activities are not clear. Macro-
    phages do not clonally divide, as do  lymphocytes, and they are not
    antigen specific (Pierce and  Kapps  1976, as reported in Faith et al.
     1980). They do, however, process the antigen, either breaking it down
    internally  or  retaining  some material on the  cell  surface.  It  is this
    surface-bound  antigen and perhaps a  soluble factor from the macro-
    phage which interact with  lymphocytes in immune  induction. Interac-
    tions of macrophages and sensitized lymphocytes in the effector phase

-------
                                 129

    of the cell-mediated immune response are probably mediated by lym-
    phokines (Faith et al. 1980).

MACROPHAGE AGGREGATION FACTOR  (MAP). Soluble material
    released by sensitized lymphocytes in culture (lymphokine), following
    exposure to antigen, that causes the aggregation of peritoneal exudate
    cells (Crowle 1975).  MAF may be the same as MIF (Lolekha et al.
    1970). Basis for in vitro test for delayed hypersensitivity.

MACROPHAGE MIGRATION INHIBITION FACTOR  (MIF). Solu-
    ble material (lymphokine) elaborated by T-cells,  and perhaps other
    cell types, following  interaction  with  specific antigen in vitro.  MIF
    inhibits  the  in vitro  migration of normal exudate cells. (Bloom  and
    Bennett 1966, Sharma 198la). Can be used to test for delayed hyper-
    sensitivity in vitro.

MAF. Macrophage aggregation factor.

MEMORY  CELLS. Long-lived small lymphocytes (B-cells),  which are
    differentiated  from  virgin  B-cells  that have  been  stimulated  by
    antigen.  These  cells  are concentrated  in the thoracic  duct  and are
    responsible for secondary antibody responses (Koller 1981).

MIF. Macrophage Migration Inhibition Factor.

MITOGENS. Molecules  of certain lectins derived from plants or similar
    macromolecules that stimulate lymphocyte  transformation  (blast cell
    formation) in vitro. This is analogous to the in vivo formation of blast
    cells that precedes clonal division of B or T  lymphocytes when there is
    an antigen challenge  (Sharma 1981d).

MITOMYCIN C. Antibiotic obtained from Streptomyces. Has antitumor
    activity  (Hawley 1981).

MYELOID  ELEMENTS. Cells of the circulating blood that originate in
    the myeloid  tissue (e.g.,  erythrocytes and granular  leukocytes) (Bloom
    and Fawcett 1969).

NATURAL  KILLER  CELLS. A  population of lymphocytes  having
    natural  cell-mediated cytotoxicity against  various  tumor cell  lines
    (Luster  et al. 1982b).

NONSPECIFIC IMMUNE  RESPONSE.  The  means of disposal of
    foreign  and  potentially  harmful  macromolecules,  microorganisms, or
    metazoa that, does not involve the recognition of antigen or the induc-
    tion of an immune response. Nonspecific immunity is generalized and

-------
                                 130

    includes the activities of phagocytic cells, microbial inhibitors (lyso-
    zymes), the interferon system, the complement system,  and the kinin
    system (Herbert and Wilkinson 1971, Vos 1981).

NUDE  MOUSE.  Born without  a  thymus  and unable  to  make  T-
    lymphocytes. Responds poorly to  thymus-dependent  antigens,  but
    mounts a  good antibody  response  to  thymus-independent antigens.
    Can serve  as a laboratory model for a congenital immunodeficiency
    syndrome and  may be  useful in testing B-cell responses (Hayward
    1982).

NULL  CELLS. Lymphocytes that have characteristics of  neither B- or
    T-cells. Useful in cellular cytoxicity assay (Luster et al. 1982b).

OPSONIN. (1) An antibody that, combined with antigen,  facilitates the
    phagocytosis of that antigen by  a macrophage or  polymorphonuclear
    leucocyte.  (2)  Heat labile substances  (nonantibody) found in blood
    plasma that facilitate phagocytosis.  Probably  activated components of
    complement, C3 in particular (Herbert  and Wilkinson 1971).

PERIARTERIOLAR  LYMPHOID SHEATH.  T-lymphocyte-dependent
    lymphoid masses  arranged  around  arteries in the white pulp of the
    spleen (Kociba 1981).

PEYER'S  PATCHES. Gut-associated lymphoid tissue, histologically simi-
    lar to secondary  (peripheral) lymphoid tissue (Kociba  1981). T-cells
    of Peyer's  patches are reported to have specific suppressor activity fol-
    lowing oral antigen administration (Ngan and Kind 1978, as reported
    in Doe 1982).

PFC.  Plaque-forming cells.

PHAGOCYTES. Cells, both fixed and  free, that ingest and often digest
    large particles such as effete blood cells, bacteria,  protozoa, and dead
    tissue  cells. These include  macrophages, Kupffer cells,  and  neutro-
    phils. They are also active in ingesting foreign materials such as finely
    divided carbon (Park and Good 1974, Herbert and Wilkinson 1971).

PHYTOHEMAGGLUTININS.  Lectins extracted from   the  beans of
    Phaseolus   vulgaris or  P.  communis  that  stimulate  lymphocyte
    transformation  and causes  agglutination of  certain red blood cells
    (Herbert and Wilkinson 1971).

PLAQUE-FORMING CELLS (PFC). Cells obtained  from the spleen or
    other lymphoid organs of an animal, immunized  against SRBC or  a
    protein antigen,  which are  mixed  with  sheep  red  blood cells  (or
    antigen-coated  SRBC if protein  antigen was  used) and molten agar,

-------
                                 131

    poured into a petri dish, and allowed  to set. The dish is then incu-
    bated at 37°C. After incubation the release by each cell of hemolysins
    (antibodies capable of lysing red  blood cells in the presence of com-
    plement) is revealed with the addition  of complement to the  dish. A
    clear  zone of hemolysis appears  around  each antibody-forming cell.
    The PFC  technique can be used to measure both primary and secon-
    dary response (Jerne and Nordin 1963).

PLASMA  CELLS.  Antibody-forming  cells,   differentiated  from  B-
    lymphocytes, that have a short half-life and secrete specific antibody
    (Sharma 1981b).

PPD. Purified  protein derivative.

PRECIPITATION. The formation of a visible complex on mixing soluble
    antigen with antiserum specific for that antigen (Gell  and Coombs
    1975).

PRIMARY IMMUNE RESPONSE.  The response of the body to the ini-
    tial encounter  with antigen. The  response  is  weak,  sluggish, and
    short-lived (Herbert and Wilkinson 1971).

PRIMARY LYMPHOID  ORGANS. The thymus  and  bursa-equivalent
    (Luster et al.  1982b). The  characteristics  that  distinguish primary
    organs from  other lymphoreticular tissues are as follows:  (1) the lack
    of cellular responses to antigen, such as those leading to  antibody for-
    mation; most cell-mediated  immune responses; and the  induction of
    immunological  memory (the  medullary cells of the thymus, however,
    can respond  to mitogens and participate  in graft-vs-host reactions);
    (2) the  intensive  production  of  lymphoid cells  in  these  organs,
    independent of antigenic stimulation; and (3) a reticular framework of
    epithelial  cells that is not  present  in other lymphoreticular  tissues
    (White 1975).

PRIMED.  (1) Of a whole animal: activation of appropriate cells  in lym-
    phoid  tissue  by exposure to antigen  so  that  further contact of a
    primed host  with  the same  antigen usually results in  a rapid and
    vigorous  secondary immune  response.  (2) Of  cells: activation  of a
    given cell with antigen so that the cell can either  produce more cells
    that have  been  so activated, or synthesize immunoglobulin, or mediate
    the reactions of cell-mediated  immunity  (Herbert and  Wilkinson
    1971).

PURIFIED PROTEIN DERIVATIVE (PPD). A soluble protein fraction,
    precipitated  from  the medium in  which  M.  tuberculosis has  been
    grown (Herbert and Wilkinson 1971).

-------
                                132
RED  PULP  (SPLENIC).  Consists  of large  numbers  of blood-filled
    sinusoids in which phagocytosis of effete red cells takes place. Many
    phagocytes and plasma cells  are  found in  the cords of Billroth, a
    modified lymphatic tissue that merges into  the white pulp.  In many
    mammals the red pulp of the  spleen contains various sized groups of
    myelocytes, erythroblasts, megakaryocytes,  and hemocytoblasts.  The
    red pulp is also heavily populated  with all the cells of the circulating
    blood (Bloom and Fawcett 1969, Herbert and  Wilkinson 1971, Park
    and Good 1974).

SECONDARY  IMMUNE  RESPONSE. Antibody production  after  the
    second  immunization.  This  response  is  characterized by faster
    development, higher levels, and greater persistence of antibody titers
    and a predominance of IgG (Roller 1981).

SECONDARY LYMPHOID ORGANS. The  spleen,  lymph nodes,  cir-
    cuiting lymphocytes, GALT, and BALT (Luster et al.  1982b).

SPECIFIC  IMMUNE  RESPONSE.  The  result  of  the  recognition of
    antigen in such a way that antibody or  primed lymphocytes  can react
    specifically  with  that  antigen.   Specific  immunity  includes  cell-
    mediated immunity, humoral immunity,  and immunologic  tolerance
    (Herbert and  Wilkinson 1971).   A major  characteristic of specific
    immunity is the establishment of immunological memory (Vos 1981).

SPLEEN. Situated in the bloodstream, unlike the lymphatic tissue, which
    is scattered in the lymph stream.  The spleen, as does the lymph node,
    has a collagenous  framework within which is suspended a reticular
    framework. There  is also a  capsule which  is  thickened at  the hilus
    where the veins enter the organ  and the arteries leave it. Continua-
    tions of the capsule, the trabeculae, penetrate the organ and  form part
    of  its framework.  The reticular  framework and  the cells  form  the
    splenic tissue, which is composed of typical lymphatic tissue,  the white
    pulp, and an atypical lymphatic tissue, the red pulp.

    The spleen functions to (1)  provide a  site  of differentiation for the
    lymphocytes and hematopoetic stem cells,  (2)  trap  the blood-borne
    foreign and altered  endogenous particles, (3) provide reservoir space
    for the circulating  blood, and (4) form antibodies, particularly IgM,
    after stimulation by antigen in   the circulating  blood  (Bloom  and
    Fawcett 1969, Herbert and Wilkinson 1971,  Park and Good 1974).

SRBC.  Sheep red blood  cells. Thymus-dependent antigen, commonly used
    in testing the humoral immune response  (Vos 1981).

SUPPRESSOR T-CELLS. Lymphocytes that suppress  antibody synthesis
    by  inhibiting B-cells (Koller  1981).

-------
                                 133
T-LYMPHOCYTES. T-cells are stimulated by antigen to form a pool of
    sensitized,  antigen-specific  lymphocytes which then  can  function as
    memory  cells, effector  cells, killer cells,  helper cells, or suppressor
    cells (Faith et al.  1980).  T-cells do  not  produce  antibody  (Roller
    1979a).

THYMUS-DEPENDENT ANTIGEN. Antigen, such as SRBC or tetanus
    toxoid, that  elicit an  immune response only when  the thymus or
    thymus cells are present (Vos 1981).

THYMUS-INDEPENDENT ANTIGEN. Antigens,  such as  lipopolysac-
    charide, that can elicit an immune response by B-cells in  the absence
    of the thymus (Moreno  1982).

THYMUS.  In mammals the thymus is considered to  be the primary
    lymphoid  organ. In most mammals the thymus consists of  two lobes
    located in the anterior part of  the  thorax,  ventral  to  the  trachea
    (Herbert  and Wilkinson, 1971).  Histologically, it consists mainly of
    lymphocytes  distributed throughout  distinct  cortical  and medullary
    areas  on  a  network  of  reticular   cells. These   lymphocytes are
    continually being produced  in the gland,  but following the neonatal
    period only about 5%  seem to leave. Hassal's corpuscles, which are
    groups of flattened,  concentrically arranged cells, are present in the
    medulla.
    During the neonatal period the thymus supplies small lymphocytes,
    those responsible for cell-mediated immunity,  to the  blood, lymph and
    thymus-dependent areas of the peripheral lymphoid organs.
    The thymus also plays  a part in  humoral immunity. Neonatally thy-
    mectonized animals can mount only  a poor antibody response to cer-
    tain antigens such as SRBC and tetanus toxoid.
    Investigations of thymic functions have established that the  thymus is
    essential for the development of adaptive immunity. It appears to pro-
    vide the microenvironment  for the differentiation and proliferation of
    primitive  stem cells migrating from the bone marrow; to elaborate a
    humoral  factor  that  influences  immunocompetency  of lymphocytes
    elsewhere; and to foster cell-mediated immunity  (Bloom and Fawcett
    1969, Herbert and Wilkinson 1971, Park and Good 1974).

WHITE PULP  (SPLENIC).  Contains a  stroma composed  of reticular
    filters  closely joined to primitive reticular  cells and fixed macro-
    phages. As in  the lymph nodes, the meshes of  the framework are filled
    with free lymphocytes of various sizes which form both  diffuse and
    nodular lymphatic tissue.  The amounts  of dense  and nodular lym-
    phatic tissue vary in  response to various stimuli. The lymphatic nodule

-------
                                 134

    contains a corona or peripheral zone  and a central germinal center
    that may be  involved in the  immune response (Bloom and Fawcett
    1969, Herbert and Wilkinson 1971, Park and Good 1974).

3.7 LITERATURE CITED

Acton J, Myrvik  QN. 1972. Nitrogen dioxide effects on alveolar macro-
phages.  Arch Environ Health 24:48-52.

Adkinson  NF, Jr. 1977.  Environmental influences on the immune system.
Environ Health Perspect 20:97-103.

Allison  AC. 1982. Macrophages and  neutrophils. In:  Clinical  aspects of
immunology, 4th  ed.  Lachmann PJ, Peters DK,  eds.  Boston: Blackwell
Scientific  Publications, pp. 101-129.

Andersson  B,  Blomgren  H.  1971.  Evidence  for  thymus-independent
humoral antibody production in mice against  polyvinylphenolidone and E.
coli Lipopolysaccharide.  Cell Immunol 2:411-424.

Archer  DL, Smith  BG,  Bukovic-Wess  JA.   1978. Use of an  in  vitro
antibody-producing  system for recognizing potentially immunosuppressive
compounds. Int Arch Allergy Appl Immunol 56:90-93.

Ascoli  M.  1902.  Ueber  den  Mechanismus der  Albuminurie  durch
Eireiweiss. Munch med  Wschr  49:398. (As reported in Gell and Coombs
1975).

Bailey DW, Usama B.  1962. A rapid method of grafting skin on tails of
mice. Transplant  Bull 7:424-425.

Bennich H, Johansson SG. 1971. Structure and function of human immu-
noglobulin E. Advan Immunol 31:1.

Berenbaum MC.  1971.  Immunosuppression by platinum diamines.  Br J
Cancer  25:208-211.

Bianco  C, Patrick R, Nussenzweig V. 1970. A population  of lymphocytes
bearing a membrane receptor for antigen-antibody-complement complexes.
I. Separation and  characterization. J Exp Med 132.

Billingham RE, Medawar PB. 1951. The technique of free skin grafting in
mammals. J Exp Biol 28:385-405.

Bloom  BR, Bennett B.  1966. Mechanism of reaction in vitro associated
with delayed-type hypersensitivity. Science  153:80-82.

-------
                                135
Bloom BR, Stoner G, Fischetti V, Nowakowski M, Muschel R, Rubinstein
A. 1974. Products of activated lymphocytes  (PALs) and the virus plaque
assay.  In: Progress in immunology II, Vol. 3.  Brent  L, Holborow J, eds.
Amsterdam: North-Holland, pp. 133-144.

Bloom W, Fawcett DW. 1969. A textbook of  histology, 9th ed. Philadel-
phia: W.B. Saunders  Co., 858 pp.

Boyden SV.  1951. The adsorption of proteins on erythrocytes treated with
tannic acid  and subsequent hemagglutination  by antiprotein  sera. J Exp
Med 93:107-120.

Bozelka  B,  Burkholder  PM,  Chang  LW.  1977.   Cadmium  induced
immunosuppression    and  splenomegaly   in   mice.   Am   J   Pathol
86(2):21a-22a.

Burnet  M.  1959. The  clonal selection  theory of  acquired immunity.
Nashville, TN: Vanderbilt Press, 209 pp.

Burrell R. 1977. Immunological reactions to inhaled physical and chemical
agents. In: Handbook of physiology. Section 9:  Reactions to environmental
agents. Lee  D, Falk  H,  Murphey S, Geiger S,  eds. Bethesda, MD: Ameri-
can Physiological Society, pp. 285-298.

Carter PB,  Bazin H. 1980. Immunology.  In: The laboratory  rat, Vol. II,
Research applications. Baker  HJ, Lindsey JR,  Weisbroth SH, eds. Ameri-
can College  of Laboratory Animal Medicine Series.

Claman HN, Chaperon  EA, Triplett RF. 1966. Thymus-marrow cell com-
binations. Synergism in antibody  production.  Proc  Soc  Exp Biol  Med
122:1167.

Coombs  RRA, Gell  PGH.  1963. The classification  of  allergic reactions
underlying disease. In: Clinical aspects of immunology, Gell PGH, Coombs
RRA, eds. Oxford: Blackwell, pp. 317-359.

Coombs RRA, Gell PGH. 1975. Classification  of allergic reactions respon-
sible for clinical hypersensitivity and disease. In: Clinical aspects of immu-
nology, 3rd ed.  Gell PGH, Coombs  RRA,  Lachman P,  eds. Oxford:
Blackwell, pp. 761-781.

Crowle AJ.  1975. Delayed hypersensitivity  in the mouse.  Adv Immunol
20:197-264.

Cunningham AJ, Szenberg A. 1968. Further improvements in the plaque
technique  for   detecting  single  antibody-forming   cells.  Immunology
14:599-601.

-------
                                 136
David JR, David R.  1971. Assay for inhibition  of macrophage migration.
In:  In  vitro methods  in cell-mediated immunity.  Bloom  BR, Glade PR,
eds. New York: Academic Press, pp. 249-261.

Dean JH, Padarathsingh ML, Jerrells TR. 1979a.  Application of immuno-
competence  assays for defining  immunosuppression. Ann NY  Acad  Sci
320:579-590.

Dean JH, Padarathsingh ML, Jerrells TR. 1979b. Assessment of immuno-
logical effects induced by chemicals, drugs, or food  additives. I. Tier test-
ing and screening approach. Drug Chem Toxicol 2(1/2): 5-17.

De Bruin A. 1976. Biochemical toxicology of environmental agents. New
York: Elsevier Publishing Company, 1544 pp.

Doe WF. 1982. Immunological aspects  of the gut.  In: Clinical  aspects of
immunology, 4th  ed.  Lachmann PJ, Peters  DK, eds. Boston:  Blackwell
Scientific Publications, pp. 985-1010.

Doull J. 1975. Factors influencing toxicology. In: Toxicology: the basic  sci-
ence of poisons. Casarett LJ, Doull J, eds. New York: MacMillan Publish-
ing Co., Inc., pp.  133-147.

Dresser DW.  1978.  Assays  for immunoglobulin-secreting cells.  In: Hand-
book of experimental immunology, Vol. 2,  3rd  ed. Weir DM, ed.  Oxford:
Blackwell Scientific Publications, pp. 28.1-28.25.

Duke L, Harshman  S. 1971. In vitro induction of  anti-DNP antibody by
immunogenic RNA from rabbit liver. Immunochemistry 8:431-445.

Engvall  E,  Perlmann P.   1971.   Enzyme-linked  immunosorbent  assay
(ELISA).  Quantitative assay of  immunoglobulin  G. Immunochemistry
8:871-874.

Esber  H,  Menninger F,  Bogden A.  1973.  Immunological  deficiency
associated with cigarette smoke inhalation by  mice: primary and secondary
hemagglutinin response. Arch Environ Health  27:99-104.

Exon JH, Koller  LD, Kerkvliet  NI.   1979. Lead-cadmium interaction:
effects  on  viral-induced mortality  and tissue  residues in  mice.  Arch
Environ Health 34:469-475.

Exon JH, Patton NM, Koller LD.  1975. Hexamitiasis in cadmium-exposed
mice. Arch Environ Health 31:463-464.

-------
                                137
Faith RE,  Luster  MI, Kimmel CA.  1979. Effect of combined pre- and
postnatal  lead  exposure on  cell-mediated immune  function. Clin  Exp
Immunol 35:413-420.

Faith RE,  Luster  MI, Vos JG.  1980. Effects on immunocompetence by
chemicals of environmental concern. In: Reviews in biochemical toxicology
2. Hodgson E, Bend JR, Philpot RM, eds. New York: Elsevier/North Hol-
land, pp. 173-211.

Flax MH, Waksman BH. 1962. Delayed cutaneous reactions in the rat. J
Immunol 89:496-504.

Ford WL. 1978. Measurement of graft-versus-host activity. In: Handbook
of experimental immunology, Vol. 2, 3rd  ed.  Weir  DM, ed.  Oxford:
Blackwell Scientific Publications, pp. 30.1-30.12.

Ford WL, Burr W, Simonsen  M. 1970. A lymph node weight assay for the
graft-versus-host   activity  of   rat  lymphoid   cells.   Transplantation
10:258-266.

Fraker PJ. 1980. The antibody-mediated and delayed type hypersensitivity
response of mice exposed to polybrominated biphenyls. Toxicol Appl Phar-
macol 53(l):l-7.

Friend M, Trainer DO.  1970. Polychlorinated biphenyl:  interaction with
duck hepatitis virus. Science 170:1314-1316.

Gabliks J, Askari EM, Yolen  N.  1973. DDT and immunological responses.
I. Serum antibodies and  anaphylactic  shock  in guinea pigs. Arch Environ
Health 26(6):305-308.

Gainer  JH.  1977.  Effects of heavy metals and of deficiency of zinc on
mortality rates in  mice  infected with encephalomyocarditis virus.  Am J
Vet Res 38:869-872.

Gardner DE, Hu EPC, Graham JA, Chernoff N, and Hueter FG.  1979.
Review  of some toxicology models to assess health  effects of environmental
pollutants. In: In vitro toxicity testing of environmental agents, current and
future  possibilities. Part B:  Development of risk assessment  guidelines.
Kolber AR, Wong TK, Grant LD, DeWoskin RS, Hughes TJ, eds. New
York: Plenum Press, pp. 443-460.

Gaworski CL, Sharma  RP. 1978.  The  effects of  heavy metals  on
3H-thymidine uptake in lymphocytes. Toxicol Appl Pharmacol 46:305-313.

-------
                                138
Gell PGH, Coombs RRA. 1975. Basic immunological methods. In: Clini-
cal  aspects of immunology, 3rd ed. Gell PGH, Coombs RRA, Lachmann
PJ,  eds. Oxford: Blackwell Scientific Publications, pp. 3-53.

Gigli I. 1982. Immunological aspects of skin diseases. In: Clinical aspects
of immunology, 4th ed.  Lachmann PJ, Peters DK, eds. Boston: Blackwell
Scientific Publications, pp.790-821.
Goldstein  AL, Aronow,  Kalman SM. 1974. Drug allergy. In: Principles of
drug action:  the basis of  pharmacology,  2nd ed. New York: John Wiley
and Sons,  pp. 489-515.

Goldstein  L,  Tyler WS, Hoeprich PD, Eagle C. 1971. Ozone and antibac-
terial  defense  mechanisms  of  the  murine  lung.   Arch  Intern  Med
127:1099-1102.

Graham JA, Gardner DE, Miller FJ, Daniels MJ, and Coffin DL.  1975.
Effect of  nickel chloride  on primary antibody production in the spleen.
Environ Health Perspect 12:109-113.

Graham JA, Miller FJ, Daniels  MJ, Payne EA, and Gardner DE.  1978.
Influence  of  cadmium, nickel and chromium on primary immunity in mice.
Environ Res  16(l-3):77-87.

Halliday  WJ, Miller S.  1972. Leukocyte adherence inhibition: a simple
test for cell-mediated immunity and serum blocking  factors. Int J Cancer
9:477-483.

Hawley GG. 1981.  The condensed  chemical dictionary,  10th  ed. New
York: Van Nostrand Reinhold Co., p.  697.

Hayward  A.  1982. Immunodeficiency. In: Clinical aspects of immunology.
Lachmann PJ, Peters DK, eds.  Boston:  Blackwell Scientific  Publications,
pp.  1658-1712.

Herbert WJ. 1973. Passive hemagglutination with  special reference to the
tanned cell technique. In: Handbook of experimental immunology, Vol. 1,
2nd  ed.  Weir   DM,  ed.   Oxford:  Blackwell  Scientific   Publications,
pp.  20.1-20.20.

Herbert WJ, Wilkinson PC. 1971. A dictionary of immunology. Oxford:
Blackwell  Scientific Publications,  195 pp.

Holt PG,  Chalmer J,  Roberts LM, Papadimitriou JM, Thomas WR,  Keast
D.  1976.  Low-tar  and high-tar cigarettes: comparison of effects in mice.
Arch Environ Health  31:258-265.

-------
                                 139

Holt PG, Keast D. 1977. Environmentally induced changes in immunologi-
cal function: acute and chronic effects of inhalation of tobacco smoke and
other atmospheric contaminants in  man  and experimental animals. Bac-
teriol Rev 41(1):205-216.

Holt PG,  Roberts LM, Keast  D.  1974. The MLA test: a rapid micropro-
cedure for the demonstration of delayed type hypersensitivity in vitro. J Int
Res Comm 2:1416.

Ishizaka K, Ishizaka T.  1971. Mechanisms of reaginic hypersensitivity: a
review. Clinical Allergy 1:9.

Jerne  NK, Nordin  AA.  1963.  Plaque  formation in  agar  by single
antibody-producing cells.  Science 140:405.

Johnson GD, Holborow EJ. 1973. Immunofluorescence. In: Handbook of
experimental  immunology,  Vol.  1,  2nd  ed.  Weir DM, ed.  Oxford:
Blackwell Scientific Publications, pp.  18.1-18.20.

Jones RH, Williams RL, Jones AM. 1971. Effects of heavy metal on the
immune response. Proc Soc Exp Biol Med 137:1231-1236.

Karol  MH.   1983.  Concentration-dependent  immunologic response  to
toluene diisocyanate (TDI) following inhalation  exposure. Toxicol Appl
Pharmacol 68:229-241.

Karol  MH, Hauth BA, Alarie Y.  1979.  Pulmonary  hypersensitivity  to
hexyl  isocyanate-ovalbumin   aerosol  in  guinea  pigs.   Toxicol  Appl
Pharmacol 51(1):73-80.

Karol  M  H, Hauth  BA, Riley EJ,  Magreni CM.  1981. Dermal contact
with toluene diisocyanate (TDI) produces respiratory tract hypersensitivity
in guinea pigs. Toxicol Appl Pharmacol 58:221-230.

Karol MH, loset  HH, Riley EJ, Alarie YC.  1978. Hapten-specific respira-
tory hypersensitivity in guinea pigs. Am Ind Hyg Assoc J 39(7):546-556.

Kauffmann BM,  White KL, Sanders VM, et al. 1982.  Humoral and cell-
mediated  immune status in mice  exposed to chloral  hydrate.  Environ
Health Perspect 44:147-151.

Kennedy J, Axelrad M.  1971. An improved assay for hemolytic plaque-
forming cells. Immunology 20:253-257.

-------
                                140
Klecak G.  1977. Identification of contact allergens:  predictive  tests in
animals.  In: Advances in  modern toxicology, Vol. 4:  Dermatotoxicology
and  pharmacology.  Marzulli FN,  Maibach HI,  eds. Washington,  DC:
Hemisphere Publishing Corporation, pp. 305-339.

Kociba RJ. 1981. Morphologic considerations in the detection of immune
suppression in routine toxicity studies.  In: Immunologic  considerations in
toxicology,  Vol.  II.  Sharma  RP,  ed. Boca  Raton, FL:  CRC Press,
pp. 123-131.

Roller LD. 1979a. Effects of environmental contaminants on the immune
system. Adv Vet Sci Comp Med 23:267-295.

Koller LD. 1979b. Some immunological  effects  of lead, cadmium,  and
methylmercury. Drug Chem Toxicol 2(l/2):99-110.

Koller LD. 1981. Humoral immunity. In:  Immunologic considerations in
toxicology. Sharma RP, ed. Boca Raton, FL: CRC Press,  pp. 19-30.

Koller L D.  1982.  In  vitro assessment of humoral  immunity following
exposure to heavy metals.  Environ Health Perspect 43:37-39.

Koller LD, Brauner JA.  1977. Decreased B-lymphocyte response  after
exposure to lead and cadmium. Toxicol Appl Pharmacol 42:621-624.

Koller LD, Exon JH, Brauner JA.  1977.  Methylmercury: decreased anti-
body formation in mice. Proc Soc Exp Biol Med 155:602-604.

Koller LD, Exon JH, Roan JG.  1975.  Antibody suppression  by cadmium.
Arch Environ  Health 30(12):598-601.

Koller LD,  Kovacic S.  1974.  Decreased  antibody  formation  in  mice
exposed to  lead.  Nature 250:148-150.

Koller LD, Roan JG, Kerkvliet N.  1979. Mitogen stimulation  of lympho-
cytes in CBA mice exposed to lead and  cadmium. Environ Res 19:177-188.

Landsteiner K, Jacobs J. 1935. Studies  on the sensitization of animals  with
simple chemical compounds. I. J Exp Med 61:643-656.

Landsteiner K, Jacobs J.  1936.  Studies on sensitization of  animals  with
simple chemical compounds. II. J Exp Med 64:625-639.

Lefford MJ. 1974. The measurement of tuberculin hypersensitivity in  rats.
Int Arch Allergy 47:570-585.

-------
                                 141
Ligo RN, James RB.  1974. Occupational skin disease:  the scope of the
problem. Cutis 13:527. (As reported in Gigli 1982).
Lolekha S, Dray S, Gotoff SP. 1970. Macrophage aggregation In vitro: A
correlate of delayed hypersensitivity. J Immunol 104:296-304.

Loomis  TA. 1974. Abnormal response to chemicals. In: essentials of toxi-
city, 2nd ed. Philadelphia: Lea and Febiger, pp. 105-117.

Loose LD, Pittman KA,  Benitz KF,  and Silkworth JB.  1977. Polychlori-
nated biphenyl  and hexachlorobenzene induced humoral immunosuppres-
sion. J Reticuloendothel Soc 22(3):255-272.

Loose LD, Pittman KA, Benitz KF, Silkworth JB, Mueller W, Coulston F.
1978. Environmental  chemical-induced  immune dysfunction.  Ecotoxicol
Environ Saf 2(2):173-198.

Loose LD, Silkworth JB, Mudzinski SP, Pittman KA, Benitz KF, Mueller
W. 1979. Modification of the immune response by  organochlorine xenobi-
otics. Drug Chem Toxicol 2(1-2): 111-132.

Luster M I, Dean JH, Boorman GA.  1982a. Cell-mediated immunity and
its application in toxicology. Environ Health Perspect 43:31-36.

Luster MI,  Dean JH, Moore JA.  1982b. Evaluation of immune functions
toxicology. In: Principles and methods of toxicology. Hayes AW, ed. New
York: Raven Press, pp. 561-586.

Luster  MI, Faith  RE.  1979. Assessment of immunological  alterations
caused  by  halogenated  aromatic  hydrocarbons.  Ann NY   Acad  Sci
320:572-578.

Luster MI,  Faith RE, Kimmel CA. 1978. Depression of humoral immun-
ity in rats following chronic developmental lead exposure. J Environ Pathol
Toxicol  1:379-402.

Luster MI,  Faith RE,  Lawson LD.  1979. Effects of 2,3,7,8-tetrachlorodi-
benzofuran  (TCDF) on the  immune system in guinea pigs. Drug  Chem
Toxicol  2(l-2):49-60.

Mancini G,  Carbonara AO, Heremans JE. 1965. Immunochemical quanti-
tation of antigens by single radial immunodiffusion.  Immunochem 2:235.

Marzulli FN, Maibach HI. 1977. Contact  allergy: predictive testing in
humans. In: Advances in modern toxicology, Vol. 4, Dermatology  and
pharmacology. Marzulli FN, Maibach HI, eds. Washington, DC: Hemi-
sphere Publishing Corporation, pp. 353-372.

-------
                                 142
Miller JFAP. 1975. Induction of the immune (allergic) response. In: Clini-
cal aspects of immunology, 3rd ed., Gell PGH, Coombs RRA, Lachmann
PJ, eds. Oxford: Blackwell Scientific Publications, pp. 447-469.

Miller  JFAP, Mitchell GF. 1968.  Cell to cell interaction in  the immune
response.  I. Hemolysin-forming cells in  neonatally thymectomized mice
reconstituted with  thymus  of thoracic duct  lymphocytes.  J  Exp Med
128:801-820.

Mishell RI, Dutton RW.  1967. Immunization of dissociated mouse spleen
cell cultures from normal mice. J Exp Med 126:423-442.

Moller  G,  Michael  G.  1971. Frequency  of antigen-sensitive cells  to
thymus-independent antigens. Cell  Immunol 2:309-316.

Moore  JA.  1979.  The  immunotoxicology  phenomenon (polychlorinated
biphenyls, halogenated hydrocarbons). Drug Chem Toxicol 2(1/2): 1-4.

Moore JA, Faith RE. 1976. Immunologic response and  factors affecting its
assessment. Environ Health Perspect 18:125-131.

Moore  JA, Luster MI,  Gupta BN, McConnell  EE.  1978. lexicological
and immunological effects  of a commercial  polybrominated biphenyl
mixture (Firemaster FF-1). Toxicol Appl Pharmacol 45(l):295-296.

Moreno C. 1982. Tolerance.  In: Clinical  aspects of immunology, 4th ed.
Lachmann PJ, Peters  DK, eds. Boston: Blackwell Scientific  Publications,
pp. 199-242.

Morley J, Wolstencroft  RA,  Dumonde DC. 1978. Measurement of lym-
phokines. In: Handbook of experimental immunology, Vol. 2,  3rd ed. Weir
DM, ed. Oxford: Blackwell Scientific Publications, pp. 27.1-27.28.

Munson AE, Sanders VM, Douglas KA, et al. 1982. In vivo assessment of
immunotoxicity. Environ Health Perspect 43:41-52.

Muller S, Gilbert KE, Krause C,  Gross U, Age-Stehr JK, and Diamant-
stein T.  1977.  Suppression of delayed-type hypersensitivity of mice by
lead. Experintia 33(5):667-668.

Ngan  J,  Kind  LS. 1978. Suppressor T cells for IgE  and  IgG in  Peyer's
patches of mice made tolerant by  the oral administration of  ovalbumin. J
Immunol 120:861. (As reported in  Doe 1982).

Nicolin A., Cosco A, Marelli O, Sarra F.  1980.  Toxicology  of immunity.
In: The principles and methods in modern toxicology.  Galli  CL, Murphy
SD, Paoletti R, eds. Amsterdam: Elsevier/North Holland, pp.  295-310.

-------
                                 143
Nulsen A,  Holt PG, Keast  . 1974. Cigarette smoking, air pollution  and
immunity: A model system. Infect. Immun. 10:1226-1229.

OECD.  1981.  Organisation for Econonic Co-operation  and Development.
Skin sensitization. In:  OECD guidelines for testing  of chemicals.  Paris,
France: OECD, Section 406,  pp. 1-9.

Oehler JR, Herberman RB,  Campbell DA, and Deu  JK.  1977. Inhibition
of rat mixed lymphocyte cultures by suppressor macrophages. Cell Immu-
nol 29:r38-m50.

Ohi G, Fukuda M,  Seto H,  Yagyu H. 1976. Effect of  methylmercury on
humoral immune  responses in mice under conditions simulated to practical
situations. Bull Environ Contam Toxicol 15:175-180.

Ouchterlony O, Nillson LA.  1973. Immunodiffusion  and immunoelectro-
phoresis.  In: Handbook of experimental immunology, Vol.  1, 2nd ed. Weir
DM, ed. Oxford: Blackwell Scientific Publications, pp. 19.1-19.39.

Paranjpe MS, Boone CW. 1972. Delayed hypersensitivity to Simian virus
40  tumor cells in BALB/c mice demonstrated by a radioisotopic footpad
assay. J Nat Cancer Inst 48:563-566.

Park BH, Good RA. 1974. Principles of modern immunobiology: basic and
clinical. Philadelphia: Lea  and Febiger, 618 pp.

Pepys J.  1976. The role of industrial agents in the etiology of asthma. Adv
Asthma Allergy (Fisons Labs.) 3:1.

Peterman GM, Wust CJ. 1975.  Comparison of three hemolytic plaque
assays using sheep and mouse erythrocytes as  antigens in rats. Proc  Soc
Exp Biol Med 150:289-294.

Pierce  CW,  Kapp  JA.   1976. The  role  of  macrophages in antibody
responses in vitro. In: Immunobiology of the macrophage. Nelson DS,  ed.
New York: Academic Press, p. 2. (As reported in  Faith et al. 1980).

Platts-Mills TAE. 1982. Type I or immediate hypersensitivity:  Hay fever
and asthma.  In:  Clinical  aspects of immunology.  Lachmann  PJ,  Peters
DK, eds.  Boston: Blackwell Scientific Publications, pp. 580-686.

Plotz P,  Talal N, Asofsky R. 1968.  Assignment of direct and facilitated
hemolytic plaques in mice to specific immunoglobulin classes. J Immunol
100:744-751.

-------
                                144

Polak L. 1977.  Immunological aspects of contact sensitivity. In: Advances
in modern toxicology, Vol. 4, Dermatotoxicology and pharmacology. Mar-
zulli FN, Maibach HI, eds. Washington: Hemisphere Publishing Corpora-
tion, pp. 225-288.

Raney JP, McLennan BD.  1979. A modified quantitative precipitin  test
which   is  rapid,  sensitive  and  reproducible.  J  Immunol  Methods
29(1):65-70.

Robertson  M.  1981.  Genes of lymphocytes. Diverse  means to antibody
diversity. Nature 290:625-627.

Rowe DS. 1975. Immunoglobulins. In: Clinical aspects  of immunology, 3rd
ed. Gell PGH, Coombs  RRA,  Lachmann  PJ,  eds.  Oxford:  Blackwell
Scientific Publications, pp. 285-306.

Ruitenberg EJ,  Steerenberg PA,  Brosi BJM. 1975. Micro-system for the
application of ELISA (enzyme-linked immunosorbent assay) in the serodi-
agnosis of Trichinella spiralis infections. Medikon Nederland 4:30-31.

Rylander R. 1969. Alterations of lung defense mechanisms  against air-
borne bacteria.  Arch Environ Health  18:551-555.

Rylander R. 1971. Lung clearance  of  particles and  bacteria. Effects of
cigarette smoke exposure. Arch Environ Health 23:321-326.

Sanders  VM, Tucker AN, White  KL, et  al.  1982.  Humoral and cell-
mediated immune  status in mice exposed to  trichloroethylene in the drink-
ing water. Toxicol  Appl Pharmacol 62:358-368.

Schuller GB, Kaufmann BM, Borzelleca  JF, Sanders VM,  Munson  AE.
1978. Effect of four haloalkanes  on  humoral and cell-mediated immunity
in mice. Toxicol Appl Pharmacol 45(1):329.

Sharma RP.  198la.  Cellular immunity.  In: Immunologic considerations in
toxicology, Vol. 1. Sharma  RP, ed. Boca  Raton, FL: CRC Press,  pp.
31-36.

Sharma  RP. 198 Ib. Glossary. In: Immunologic considerations in toxicol-
ogy, Vol. 2. Sharma RP, ed. Boca Raton, FL: CRC Press, pp. 157-160.

Sharma  RP. 1981c.  Immunity and the  immune system. In:  Immunologic
considerations in toxicology, Vol. 1. Sharma RP,  ed. Boca Raton, FL:
CRC Press, pp. 9-18.

-------
                                 145

Sharma RP. 198Id. Splenic lymphocyte transformation in culture as a tool
for immunotoxicologic evaluation of chemicals. In: Immunologic considera-
tions in toxicology, Vol. 2. Sharma RP, ed. Boca Raton, FL: CRC Press,
pp. 133-145.

Sharma  RP. 1981e. Tissue  or  cell injury  in  immune mechanisms.  In:
Immunologic considerations  in toxicology, Vol. 1. Sharma RP, ed.  Boca
Raton, FL: CRC Press, pp. 37-44.

Sharma RP, Kociba RJ, Gehring PJ. 1978. Immunotoxicologic effects of
2,3,7,8-tetrachlorodibenzo-p-dioxin  in laboratory  animals.  Toxicol  Appl
Pharmacol45(l):333.

Silkworth   JB,   Loose   LD.   1981.   Assessment  of   environmental
contaminant-induced lymphocyte dysfunction.  Environ  Health  Perspect
39:105-128.

Smith  RT.  1972.  Specific recognition reactions  at  the  cellular level  in
mouse  lymphoreticular cell subpopulations. Transplant Rev 11:178-216.

Smith  SH, Sanders  VM, Barrett BA, Borzelleca JF, Munson AE. 1978.
Immunotoxicological  evaluation  on  mice  exposed  to  polychlorinated
biphenyls. Toxicol Appl Pharmacol 45(1):330.

Speirs  RS, Speirs EE.  1979.  In vivo models for assessing effects of drugs
and toxicants on immunocompetence. Drug Chem Tox 2(1-2): 19-33.

Spyker JM. 1975. Assessing the impact of low level chemicals on develop-
ment:  behavioral and  latent effects. Fed  Proc Fed  Am  Soc  Exp  Biol
34:1835-1844.

Stavitsky AB.  1954. Micromethods for the study of proteins and antibo-
dies. I.  Procedure   and  general applications  of hemagglutination  and
hemagglutination inhibition reactions with tannic  acid and protein-treated
red blood cells. J Immunol 72:360-367.

Stobo  JD,  Rosenthal AS, Paul  WE. 1972.  Functional  heterogeneity  of
murine lymphoid cells. I.  Responsiveness to and  surface binding of con-
canavalin A and phytohaemagglutinin. J Immunol  108:1-17.

Stuart  AE,  Habeshaw  JA, Davidson AE. 1978. Phagocytes in vitro.  In:
Handbook  of experimental immunology, Vol. 2,  3rd ed.  Weir DM,  ed.
Oxford: Blackwell Scientific Publications, pp. 31.1-31.30.

Tada NK,  Itakura K, Aizawa M. 1974. Genetic control of the  antibody
response in inbred rats. J Immunogenet 1:265-273. (As reported in Luster
and Faith 1979).

-------
                                146
Thigpin JE, Faith RE,  McConnell  EE, and Moore JA.  1975. Increased
susceptibility  to   bacterial  infection  as  a  sequela   of  exposure  to
2,3,7,8-tetrachlorodibenzo-p-dioxin. Infect Immun 12:1319-1324.

Thomas PT, Hinsdill RD.  1978. Effect of polychlorinated biphenyls on the
immune responses of Rhesus monkeys and mice. Toxicol Appl Pharmacol
44(l):41-52.

Thomas WR, Holt PG,  Keast D.  1973. Effect of cigarette smoking on pri-
mary  and  secondary  immune  response  in  mice.   Nature  (London)
243:240-241.

Thomas  WR,  Holt PG,  Keast  D. 1974a.  Alterations in  the primary
immune response of mice by exposure  to fresh cigarette smoke. Int Arch
Allergy Appl Immunol 46:481-486.

Thomas WR, Holt PG,  Keast D.  1974b. Mice chronically exposed to fresh
cigarette smoke. Arch Environ Health 27:372-375.

Thor DE. 1971. The capillary tube  migration inhibition technique applied
to human peripheral lymphocytes using the guinea pig peritoneal exudate
as the indicator  cell population. In: In vitro methods  in cell-mediated
immunity.   Bloom  B,  Glade  P,  eds.  New  York:   Academic  Press,
pp. 273-280.

Thorpe PE, Knight SC.  1974. Microplate culture of mouse lymphoid cells.
I. Quantitation  of responses to  allogeneic  lymphocytes, endotoxins and
phytomitogens. J  Immunol Methods  5:387-404.

Thurman GR, Goldstein AL, Kilian DJ. 1978. The effects of organic com-
pounds used in the manufacture  of plastics on  the responsivity  of murine
and human lymphocytes. Toxicol Appl Pharmacol 44(3):617-641.

Treagan  L. 1975. Metals and the immune response: a review. Res  Com-
mun Chem Pathol 12(2): 189-220.

Tripathy SP, Mackaness GB. 1969. The effect of cytotoxic agents on the
primary immune response  to Listeria monocytogenes. J Exp Med 130:1-16.

Tucker AN, Sanders VM, Hallett P, Kauffmann B, Munson AE. 1982. In
vitro  immunotoxicological assays for  detection of compounds requiring
metabolic activation. Environ Health Perspect 43:123-127.

Turk JL.  1975. The mechanism and mediators of cellular hypersensitivity.
In: Clinical aspects of  immunology, 3rd ed., Gell  PGH, Coombs  RRA,
Lachmann   PJ,    eds.    Oxford:   Blackwell   Scientific   Publications,
pp.  1019-1042.

-------
                                 147
Urso P, Gengozian  N. 1973.  T cell deficiency in mouse allogeneic radia-
tion chimeras. Immunol 111:712-719.

Urso P, Gengozian  N. 1980.  Depressed humoral immunity  and increased
tumor incidence in mice following in utero exposure to benzo[a]pyrene. J
Toxicol Environ Health 6:569-576.

van Furth R, van Zwet TL,  Leijh  PCJ. 1978. In  vitro determination of
phagocytosis and  intracellular killing by polymorpho-nuclear and mononu-
clear phagocytes. In: Handbook of experimental immunology, Vol.  2, 3rd
ed.  Weir  DM,  ed.   Oxford:   Blackwell  Scientific   Publications,
pp. 32.1-32.19.

Vassallo CL, Domm BM,  Poe RH,  Duncombe ML, Gee JBL. 1973. NO2
Gas and NO2 effects  on alveolar macrophage phagocytosis and metabol-
ism. Arch Environ Health  26:270-274.

Verschuuren HG, Ruitenberg EJ, Peetoom F, Hellman PW, van Esch GJ.
1970. Influence of  triphenyltin acetate  on lymphatic  tissue and immune
responses in guinea pigs. Toxicol Appl Pharmacol 16:400-410.

Vos JG. 1977. Immune suppression as related to toxicology.  Crit Rev Tox-
icol 5(1):67-101.

Vos JG.  1981. Screening and function tests to detect immune suppression
in toxicity  studies. In: Immunologic considerations in  toxicology, Vol. II.
Sharma RP, ed. Boca  Raton, FL: CRC Press, pp. 109-122.

Vos JG, Buys J, Beekhof P, Hagenaars AM.  1979a. Quantification of total
IgM and IgG  and specific IgM  and IgG to  a thymus-independent (LPS)
and a thymus dependent  (tetanus toxoid) antigen in the rat by Enzyme-
linked Immunosorbent Assay (ELISA). Ann NY Acad Sci 320:518-534.

Vos JG, De Roij T. 1972.  Immunosuppressive activity of a polychlorinated
biphenyl preparation on the humoral immune response in guinea pigs. Tox-
icol Appl Pharmacol 21:549-555.

Vos JG, Moore JA. 1974. Suppression  of cellular immunity in rats and
mice by maternal treatment  with 2,3,7,8-tetrachlorodibenzo-p-dioxin.  Int
Arch Allergy Appl Immunol 47:777-794.

Vos  JG,  Moore JA,  Zinkl  JG.  1973. Effect  of 2,3,7,8-tetrachlorodi-
benzo-p-dioxin  on the immune  system of  laboratory  animals. Environ
Health Perspect 5:149-162.

-------
                                148

Vos  JG, Van  Logten MJ, Kreeftenberg  JG, and  Kruizinga  W.  1979b.
Hexachlorobenzene-induced stimulation of the humoral immune response
in rats. Ann NY Acad Sci 320:535-550.

Vos JG, Van Logten MJ, Kreeftenberg JG, Steerenberg PA, Kruizinga W.
1979c. Effect of hexachlorobenzene on the immune system of  rats follow-
ing  combined  pre-  and  postnatal  exposure.  Drug  Chem  Toxicol
2(l-2):61-76.

White RG. 1975. Immunological  functions of lymphoreticular tissues. In:
Clinical  aspects  of immunology, 3rd  ed.  Gell PGH,  Coombs  RRA,
Lachmann PJ, eds. Oxford: Blackwell Scientific Publications, pp. 411-445.

White KL, Jr.,  Sanders  VM, Kauffmann BM et al.  Immunotoxicological
investigations in the mouse I. General approach and methods. (In press).

Yamada   H.   Yamada   A.    1969.    Antibody   formation   against
2,4-dinitrophenyl-hapten  at the cellular level. J Immunol 103:357-363.

Zarkower  A.  1972. Alterations  in antibody response induced by chronic
inhalation of SO2 and carbon. Arch Environ Health 25:45-50.

Zarkower  A,  Morges M. 1972.  Alteration in antibody responses induced
by carbon inhalation: a model system. Infect Immun 5:915-920.
 3.8 GENERAL REFERENCES

 Lachmann PJ, Peters DK, eds. 1982. Clinical aspects of immunology, 4th
 ed., Vols. I and II. Boston: Blackwell Scientific Publications, 1751 pp.

 Weir  DM,  ed.  1978. Handbook of experimental  immunology, 3rd ed.
 Oxford: Blackwell Scientific Publications.

-------
                4. INHALATION TOXICOLOGY

                           Michael G. Ryon

4.1. INTRODUCTION

   Inhalation toxicology  is the branch  of toxicology that deals with expo-
sure by the inhalation or respiratory route and  with the resulting effects
(Campbell  1976).  Inhalation toxicology exists as a special  subset of toxi-
cology because  of  the unique exposure situation  of the lungs and the spe-
cial technique and  equipment needs of the inhalation toxicity test.
   The lung is the organ with the greatest  direct  contact between the
environmental air  (and the contaminants therein)  and  its vital, functional
elements as a direct  unavoidable consequence of living (Casarett 1975).
This contact of the lung with the  air is even  more significant because of
the tremendous surface area and number of structures  involved. There are
several hundred million alevoli in  the  human lung, with an approximate
surface  area  of 100 m2  when distended (Hatch  and  Gross 1964, Cam-
bridge 1973). These are  serviced by 2000 km of lung  capillaries that are
separated by only  1  fj,m  from the gas in  the alevoli  (Hatch and Gross
1964, USDHEW 1977). Due to this large surface area/blood contact, even
the rate of intake of highly soluble compounds can be ten times greater for
the inhalation route than for the oral route. Added to these structural con-
siderations  is the tremendous amount of air (10 m3) processed each day by
the lungs (Hatch and Gross 1964). Thus, the  contact with a potential lung
toxicant can be extremely hazardous. The concern about the impact of air
pollution exposure  on man (Stokinger  1953) and the tremendous death
rate from lung cancer [in 1967 in the United  States over 50,000  deaths or
one every 10 minutes (Saffiotti 1969)]  underscore  the importance of inha-
lation toxicity studies.
   Inhalation toxicity also receives special emphasis because of the unique
equipment  and  techniques required for its evaluation. Because  the lungs
not only are responsible for maintaining the oxygen supply and the excre-
tion of carbon  dioxide, but also have a role  in  several body defense sys-
tems, inhalation toxicity studies must  deal with the  systemic as  well as
specific lung damage (Casarett 1975). Thus, there is no way that only one
standard and  accepted protocol can be used for all inhalation  studies
(MacFarland 1975). However, base guidelines can be  and  have  been pro-
posed to ensure that  minimum requirements  of  toxicity testing  are being
met; these  guidelines  should still allow the  flexibility and freedom to sub-
stitute more stringent or  comprehensive procedures (Gross 198la, USEPA
1982). Inhalation  studies  are  faced with unique problems  with  regard to
                                 149

-------
                                  150

determining dosage parameters and designing exposure  conditions. A spe-
cial  technology  has been  developed to duplicate  man's environment for
study in the laboratory, including exposure chamber designs, generating
and  monitoring  equipment for the various types of test atmospheres,  and
procedures for the evaluation of effects  (Drew  and Laskin  1973).  These
special  requirements  increase the average cost  of a toxicity study using
inhalation  exposure to two or three times that of a study  done by  oral
exposure (WHO 1978).
   It is for these reasons  that inhalation toxicology studies  have received
special  emphasis.  This report will discuss further  the special needs  of the
inhalation  study and  in particular will cover basic equipment considera-
tions and unique additions to test designs. Current guidelines for inhalation
testing  are not emphasized  in  this report;  for  a  discussion of these, see
Gross (198la) and USEPA (1982).

4.2  GENERAL TEST DESIGN

4.2.1 Introduction

     Toxicity tests designed to evaluate the effects  of a chemical after inha-
lation  exposure involve many of the  same variables as tests using other
exposure routes. All toxicity tests require the selection  of test species, test
 duration, exposure regimen, dosage levels, and procedures to evaluate toxic
 effects. These  variables  will  be discussed briefly  in this section  with
 emphasis on those aspects restricted to inhalation exposure.

4.2.2 Test Species
     In selecting the most suitable species for inhalation tests, the anatomi-
cal  and physiological similarity between  human  and test animal respiratory
systems  must   be  considered  along   with   metabolic   similarity  and
economic/practicality restrictions. The anatomical and  physiological differ-
ences include variations in the  structure of nasal cavities, trachea, and
 lower respiratory tree, in the site of particle deposition, and in the clear-
 ance and defense mechanisms (Giovacchini 1972). Table 4.1 lists some of
 the  physiological variations between man and various test species. Because
 no  single  animal species  is  anatomically similar to man  in all respects,
 researchers have  used mice, monkeys, rats, rabbits, hamsters, guinea pigs,
cats, and dogs for inhalation studies (Drew and Laskin  1973, WHO  1978).
     McLaughlin, Tyler, and Canada studied the respiratory system of vari-
ous  animals, examining such subgross anatomical variables  as the lobular-
ity,  pleura and  interlobular septa, arterial supply  to the pleura, and termi-
nal  bronchioles.  They classified species according to  their similarity to
humans and concluded that the  horse is anatomically  the most similar to
 man (McLaughlin et al. 196la,  1961b). The rat, rabbit, guinea pig, dog,
 cat, and monkey were considered to have basically the same type of lung
 structure  and  degree  of  relative similarity to man (McLaughlin  et  al.
 1966).  Among  these species,  Roe (1968) found  that  the  smaller  species

-------
                         Table 4.1. Some Physiological Indices of Man and Animals*
Physiological index
Body surface (m2)
Relation body surface to body
weight (m2/kg)
Basal metabolism (kJ/kg)
Frequency of respiration (min)
Size of alveoli (nm)
Surface of lungs (m2)
Relation of lung surface to
body weight (m2/kg)
Inhaled air (mL)
Lung ventilation (mL/min)
Relation of lung ventilation
to body weight (mL/min/g)
Consumption of oxygen (mL/kg/h)
Elimination of CO2 (mL/kg/h)
Coefficient of respiration
Pulse frequency for 1 min
Man
1.8
0.0257
105
14-18
150
50
0.7
616
8732
0.13
203.1
168.8
0.82
70-72
Dog
0.528
0.044
222
10-30
100
100
8.3
40-60
NG
NG
3600
NG
NG
90-120
Cat
0.2
0.066
NGb
20-30
100
7.2
2.8
NG
1000
0.30
9420
NG
NG
120-180
Rabbit
0.18
0.072
188
50-100
NG
5.21
2.5
NG
600
0.29
522.7
NG
0.83
150-240
Guinea
Pig
0.040
0.12
360
80-135
NG
1.47
3.2
1.75
155
0.33
2180
NG
NG
206-280
Rat
0.030
0.15
615
110-135
50
0.56
3.3
0.865
73
0.05
2199
2650
0.82
300-500
Mouse
0.006
0.3
711
140-210
30
0.12
5.4
0.154
25
1.24
3910
4240
0.85-1.33
520-780
aAdapted from Sanockij (1970a) as reported in
bNG = not given.
WHO (1978).

-------
                                  152

were  markedly  different from  man, especially in the  factors that affect
particle deposition. The  mouse, hamster, and rabbit lack the extensive tra-
cheal or bronchial mucus-secreting glands that humans possess, while the
rat possesses such glands only in the trachea. Although the cat does have
these glands, its pattern of secretion is so unlike man  as to make its use
generally unsuitable. The guinea pig, which also has these glands, is lim-
ited  in  its applicability  because of  a predisposition  to  asthmatic-type
attacks (Roe 1968). Roe felt that the dog  and monkey  were closer to man
in gross anatomy and response but that their size reduced their applicabil-
ity. Additionally, the monkey  is useful only when  laboratory-bred speci-
mens are available. In contrast to some other researchers, Roe considered
the mouse to be the preferred  species  for inhalation carcinogenesis studies
due to its high susceptibility, which enhances the chances for tumor forma-
tion.
    In an evaluation of dust retention in small animals, Palm et al. (1956)
found that the  retention pattern  for the monkey was more similar to man
than  that  of the guinea pig. In general, the smaller the animal the  lower
the rate of lung deposition for particles above 1  jim in diameter with less
alveolar deposition. Also, the overall retention is greater in small animals,
most likely due to greater  impingement  in the upper  respiratory  tract.
However,  Palm et al. (1956) did find  that the most favorable particle size
(1  /im) for alveolar  deposition was similar  for man  and small animals.
Another anatomical variable that influences particle deposition is the hor-
izontal layout of respiratory tracts of quadrupeds, which is totally different
from the  vertical tract of  man  and  other primates  (Palm et al.  1956,
WHO  1978). Carney (1979) found  that, for evaluations of immunological
sensitivity to particles, the  rat  was totally  unsuitable due to  sensitivity
problems and that  the  guinea pig was a more  suitable model. Brain and
Mensah (1983)  have reviewed many  aspects of anatomical differences  of
species that can affect inhalation  testing.  They discussed  variations  in
deposition, clearance,  and biological response and concluded the following:
(1)  the fraction of  aerosol actually  deposited  in  the respiratory tract
appears independent of body size; (2)  different species  breathing the same
aerosols do not  necessarily receive the  same lung doses, and thus exposure
concentration is  not an  adequate description of lung dose; (3) factors that
affect the lung dose  include changes  in ventilation, collection efficiency,
lung  anatomy, and clearance mechanisms; and  (4) interspecies variations
in  substance  metabolism and innate  biological responsiveness  make  it
unlikely that lung  damage  will  be the same  even if  the  lung dose  is
equivalent.
    The available literature clearly indicates that there is no species ana-
tomically identical to  man in all  respiratory  aspects. The most similar test
species are generally the larger animals (horse, dog, and monkey), which
are impractical for the majority  of inhalation  chambers and  tests. It is
therefore not surprising to find guidelines that recommend the rat (OECD
1979, Gross 198la)  or other rodent   species, especially for acute tests

-------
                                 153

(Drew and Laskin 1973, NAS 1977).  Other guidelines suggest that meta-
bolic similarity should also be stressed, because systemic effects are impor-
tant (Troy  1974),  that  a multiple  species  system is better (Hammond
1970, Drew and Laskin 1973), or that a rodent and nonrodent are needed
for chronic  studies (NAS 1977). A recent  anatomical and physiological
evaluation  of the ferret  indicates that  its  airways  are  more similar to
humans than  the dog's and  other favorable aspects of its lung  anatomy
make it  a more useful  and  less expensive choice for a nonrodent  test
species (Vinegar et al. 1981). As with other types of toxicity tests, no sin-
gle species can be found to satisfy all the needs of inhalation tests, and the
specific species chosen should be the one judged to be best suited for that
particular chemical and exposure condition (Troy 1974).

4.2.3 Duration and Mode of Exposure

    Similar to tests using other routes of exposure, there are three principal
test durations for inhalation  studies  (acute, subchronic, and chronic). The
initial evaluation of toxicity  is often obtained through an  acute test. This
test is defined as a single uninterrupted exposure of high concentration,
usually with a duration  of less than 8 h  (Drew  and Laskin 1973, NAS
1977, WHO 1978). An acute test can provide various information, includ-
ing approximate comparative toxicity of the chemical, the range of doses
needed for  studies  of longer durations,  and the  nature of possible toxic
effects (Drew and Laskin 1973, Troy 1974, NAS 1977). This information
is  provided  primarily  by determination  of an  LC50  value,  which is the
atmospheric concentration that  will  kill 50% of the test animals within a
predetermined period  of time  (Drew and Laskin  1973, Carney 1979).
Often the LC50 is a statistically derived value (OECD 1979)  and is given
in units  of  test substance weight per volume of air (mg/L, mg/m3, or
ppm) (Troy  1974,  OECD 1979). Guidelines have suggested  an  exposure
period of 1  to 4 h with  a  14-day observation period (NAS  1977, Gross
1981a). The length  of the observation period  is critical, because a short
period might miss delayed mortality and  a long period could  allow repair
of any tissue damage (Carney 1979). EPA guidelines state that the obser-
vation period should be flexible; a length beyond 14 days could be justified
by  the toxic  reactions, rate  of onset, and  length  of recovery  (USEPA
1982). Evaluations  of acute toxic effects  can include  clinical  observations
(NAS  1977), airway  irritancy  responses  (Giovacchini 1972, Cambridge
1973, Wells 1979),  gross necropsy, and  microscopic examination of the
lungs or  other target organs (NAS  1977,  OECD 1979).  Occasionally, no
acute effects can be obtained even with the maximum concentration that
can  be generated  (Carney  1979).  In this  case testing  may be halted
(OECD 1979) unless potential acute human exposure is expected (Carney
1979).
    To initially evaluate toxicity resulting from repeated exposure  to a sub-
stance, a  subchronic test  is  usually performed. The duration of a sub-
chronic inhalation test is no more  than  10%  <   the lifespan of the  test

-------
                                 154
animal, with typical lengths of 14, 28, or 90 days (OECD 1979). An expo-
sure time of 6 h/day is  often used for these tests (Wells  1979, OECD
1979).  Information   on  the  nature of  chronic  toxic  effects,  detailed
exposure-effect relationships, and range of doses for further chronic studies
represents  the  principal  results of a  subchronic test (Drew and Laskin
1973). The degree of information obtained depends on the type of evalua-
tions in the test design.  For maximum information from the  subchronic
inhalation  test, evaluations could include clinical observations, biochemical
tests, hematology tests, gross and microscopic pathology, and  pulmonary
function tests (Giovacchini 1972, Carney  1979, OECD 1979, Page et al.
1980).
    A  chronic inhalation  test is  necessary to fully evaluate  the effects of
lifetime exposure to a substance. To simulate lifetime exposure in humans,
the duration of  the  chronic test  should be greater than  10%  of the test
animal's lifespan, which  for most animals would be 1  to 5 years (NAS
 1977,  OECD 1979).  Additionally, the exposure period should be patterned
either after human occupational exposure  (6 to  8 h/day, 5 days/week) or
possible environmental exposure  (22 to 24 h/day, 7 days/week) depending
on the intended  use of the product (Drew and Laskin 1973, WHO 1978).
Procedures to  evaluate the chronic effects could include clinical observa-
tions,  biochemical tests, hematology  tests,  gross necropsy,  microscopic
pathology, and pulmonary function tests  (NAS 1977,  OECD 1979). To
use these evaluations for the duration  of a chronic test, large numbers of
test animals (e.g., a  minimum of 400 rodents) and exposure chambers are
required (NAS  1977). Therefore the  chronic test should be  undertaken
only after  extensive  preliminary testing  has indicated the need for an
evaluation of lifetime exposure effects.
    As  mentioned in  discussion of chronic  tests, exposure to the test sub-
stance can be scheduled on a continuous or an intermittent basis.  Although
this allows the substance to be evaluated in  the potential mode of human
exposure, there are certain factors of the two methods that must be con-
sidered.
    With intermittent exposure (6 to  8 h/day,  5  days/week),  the dose is
pulsed  with substantial periods of time, both daily and over the weekend,
for recovery from the test substance. These periods of recovery can  affect
the toxicity of the chemical (Haun 1972,  Drew and Laskin 1973, WHO
 1978).  Wright (1957) studied the effects of continuous  versus intermittent
exposure on dust retention in  rats  in  a  3-month subchronic  study and
found that rats exposed for 2 h  at a high concentration retained less dust
(20 to 30%)  than rats  exposed  for  20 h  at a low concentration,  even
though both groups  had the  same daily  concentration  X  time (CXT)
dose. Sidorenko  and  Pinigin (1976) examined the two  exposure modes in
short-term studies with rats and mice and  found that, generally, intermit-
tent exposure is  less  harmful (in terms of mortality  and  time  to onset of
effects) than continuous exposure. As the intervals between exposures were
increased  in duration, the toxic  effect of a substance in  mice declined

-------
                                  155

exponentially or  proportionally, depending  on the specific substance.  In
rats, the interrupted exposure had a smaller effect on the central nervous
system and  blood than did  continuous exposure.  However, Sidorenko and
Pinigin did find some contradictory evidence in the rate of onset of effects
with increasing time intervals between exposures. The varying response to
discontinuous exposure is related to the quantity of test substance that can
cross  lung membranes  and/or be absorbed by the aveolar  blood supply
during a given  time and is  perhaps related  to the processes of adaptation
and accumulation in the organism (Sidorenko and Pinigin 1976). Gardner
et al. (1979) studied the variation between  the two exposure modes using
NO2 and concluded that the concentration  factor is more important than
the length of exposure. When a constant  CXT dose was employed, a short
duration at a high concentration  produced  a greater effect than  a long
duration at a low concentration.
   Intermittent exposure has  some  practical  advantages that makes it
easier and cheaper  to  use than continuous  exposure.  The chamber design
can be much simpler, as there is no need to  feed or water the animals dur-
ing exposure (Drew and Laskin 1973, WHO 1978).  Similarly,  the  atmo-
sphere  generating  equipment  and  monitoring  devices  can be  simpler
because  they operate  only  for 6 to  8  h/day (Drew and Laskin  1973,
OECD 1979). The shorter period actually spent in the chamber also allows
more animals to  be exposed per cage (Drew and  Laskin  1973). In large
studies, these advantages can be substantial.
   The advisability of initially evaluating a substance using actual expo-
sure conditions has been questioned. The National Academy of Sciences
(NAS 1977) has suggested that  controlled studies can  often give more
informative data, especially  if the actual use conditions are unique, and
standard exposures  also make comparisons of relative toxicity easier. Such
tests basically  evaluate the toxicity  of  the test  substance (which is  an
inherent  property of the compound) and do  not fully address the hazard of
the compound (which is a combination of toxicity and exposure conditions)
(Gross 198la).  Additionally, even  though intermittent exposure may simu-
late the  time span  of  potential human exposure,  actual concentrations in
the environment  vary, with peaks higher than the average concentration
value. Thus, intermittent exposures may  not be realistic or representative
of the actual use  conditions (Drew and Laskin 1973, WHO 1978).

4.2.4 Dosage
   The  design of  inhalation exposure  studies considers two factors  of
dosage:  first, the appropriate number of levels (including controls) and,
second, how to  accurately determine the dosage.
   The number of concentrations varies with the duration and goals of the
test. Most guidelines recommend  a  minimum of  three dose levels of the
test substance for all studies (WHO 1978, OECD 1979, Page et al. 1980).
Acute tests  require  that the dose levels be designed to produce a range of

-------
                                  156

responses to permit construction of a mortality-concentration curve for cal-
culation of an LC50 value. Subchronic and chronic tests should use a dose
range with the high dose producing toxic effects, but not necessarily mor-
tality, and the low dose producing minimal or no effects and being roughly
equivalent  to  the  No-Observable-Effect  Level (NOEL)  (WHO  1978,
Gross  198la).  At least one intermediate dose is necessary if information
on  the dose-response relationship is desired (Page  et al.  1980,  Gross
1981a).
    Incorporation of control groups is more  extensive in chronic  and sub-
chronic tests. There  should be  at  least  one control group  that undergoes
the same procedures, except for actual  exposure to the test  substance,  as
the treatment groups. In addition,  a control  group is necessary to evaluate
the administration vehicle, if one is employed. The use of a vehicle control
group is also recommended for acute studies  (OECD 1979).
    The principal uncertainty associated with dosage in inhalation studies
is the  determination  of the actual effective  dose. For example, there may
be  variations in the  dose to the lung, dose  to the blood, and dose to the
target organ. The  factors that determine the dose delivered to  the lung
include, of course, length of exposure  and atmospheric concentration  of the
test agent  in the chamber (Roe 1968, MacFarland 1976).  However, these
factors describe, at best,  only the amount of test agent  available to the test
animals. The critical value is the dose  that  reaches the sites of  action  in
the lung (Hatch and Gross 1964, MacFarland 1975, Clark  1977).  Many
factors can affect this parameter, particularly test species variables such  as
minute volume of  ventilation;  frequency of respiration; bronchial dimen-
sions,  pulmonary  and  bronchial  arterial bloodflow; rate of  elimination
including natural clearance mechanisms; metabolic conversion; activity lev-
els of  the animals; and percentage of retention (DuBois and  Rogers 1968,
Drew and  Laskin 1973, NAS 1977, USDHEW 1977). Physical and  chem-
ical properties of the test agent also determine the effective dose, especially
the solubility, diffusion coefficient, and, for aerosols, particle size (DuBois
and Rogers  1968,  NAS 1977). The  degree that  these factors affect  the
dose can also vary. It has been shown,  at least theoretically, that the pul-
monary factors may be less important in  determining dose for tests  of
chronic duration  than  those  of shorter exposures  (DuBois and  Rogers
1968).  Further,  defense mechanisms  can  concentrate,  via phagocytic
macrophages, a relatively diffuse and even  exposure to create "hot  spots"
of  high test agent concentration (USDHEW 1977). This effect may  be
especially important if thresholds for effects apply to the test agent.
    In  addition to  these  variables, there is  often interference from other
accidental  routes of  exposure. Clearance of  the test agent from the lung
via the mucociliary  escalator  can result  in unintended  gastrointestinal
exposure (Clark 1977). Similarly, animals can ingest the test agent during
grooming behavior  or experience dermal exposure  if the test is performed
in a whole-body exposure chamber (Drew  and Laskin  1973, Kawai and
Nozaki 1978).

-------
                                   157

   Because one of the primary objectives of the inhalation test is to eluci-
date upon  the dose-response  relationship, the variables and interferences
discussed above have influenced  how the inhalation toxicologist calculates
dose and relates it to effects. The product of the atmospheric concentration
(C)  and  the  duration  of  exposure  (T)  is  often  used  to  determine
dose-response relationships. This calculation does not produce a dose per
se, as a dose by definition is a mass  quantity  (MacFarland 1976). How-
ever, it is still used to estimate the dose on the theory that the CXT dose
is proportional to the inhaled dose, which is proportional to the absorbed
or effective dose at  the  site of action  (MacFarland 1976, Carney 1979).
The  principles behind this theory  were revealed by Haber (1924)  in his
studies of chemical weapons and  is often known as Haber's rule. This rela-
tionship holds true at most exposures, but at the high and low extremes
the observed values  deviate from the  predicted values. Because the CXT
dose  is proportional  to  the  real  dose, it has  been used  to  calculate
L(CXT)50 data in acute tests  and to develop dose-response data.
   The main problem with the use of the CXT dose and  Haber's rule to
calculate dose-response relationships is that they represent  a simplification
of the real  situation by ignoring interspecies and individual variations in
minute volume and retention  percentage (Weston and Karel 1946, Carney
1979). These factors are  especially important  because  animal  data are
used as a basis  for extrapolation to human exposure conditions. The large
variation in the animal data could  produce  human exposure levels that are
still within hazardous concentrations (MacFarland 1976).
   Recognizing  this  potential   problem,   Weston  and   Karel  (1946)
developed a  dose formula  incorporating the minute volume (MV) and the
percentage of inhaled agent that is retained (a). This Dosimetric Formula,
dose = aCTrMV, can be considered  a true dose because it results in  a
value with dimensions of  a mass quantity. However,  the Dosimetric For-
mula has proved to be difficult  to use in  practical situations. Values for
the minute volume of animals at rest  for different species  are available in
the literature, but because it is the values under the conditions of exposure
that are important, studies should  allow for measurement of the MV dur-
ing actual  exposure  (MacFarland 1976). The retention percentage is even
more difficult to quantify as the values vary depending on the chemical
(particularly its solubility and particle size),  duration  of exposure,  test
species, and  exposure chamber conditions  (e.g., humidity). Despite these
drawbacks, the Dosimetric Formula represents a better approach than does
the CXT dose, especially for evaluating comparative  toxicities or differ-
ences between routes of exposure (MacFarland 1976).
   The difficulties associated  with the  CXT dose and the Dosimetric For-
mula have prompted researchers  to experiment with alternative methods of
evaluating  inhalation  doses, particularly in regard to the retained  dose.
Radioactive isotopes have  been used to measure the actual amount of the
test substance retained in the lung (Albert  et al.  1967, Roe 1968). Griffis
et al. (1981b) reported a method using acid-insoluble sodium to determine

-------
                                 158

amounts of glass fiber retained in lungs after exposure to aerosols. Magne-
tometric techniques for measuring lung burdens of ferrimagnetic dust have
also  been reported (Oberdoerster and Freedman  1982). Circulating blood
levels or rates of excretion have been used to estimate the body burden
and  are in  some cases better estimates  of dose than the CXT product
(MacFarland  1976,  Clark 1977). Such techniques still require  additional
development and validation before they can be incorporated into a routine
inhalation protocol.

4.2.5 Observations for Effects
   In  inhalation studies, many  methods have been used  to observe and
determine the toxic effects of the test agent. To detect systemic effects,
most of the tests and observations are the same as those used in other toxi-
city  studies (Poynter 1977).  However,  special procedures are used  in
observing toxic effects on the portal  of entry, the lung. The observation of
systemic effects will be only briefly discussed, with the majority of the dis-
cussion focusing on  observations unique to inhalation studies. No attempt
is made to  suggest  a rigid or standard protocol because the flexibility to
choose between a battery  of basic,  simple tests and  an array  of special
tests for inhalation effects can often be the most efficient  and informative
procedure  (MacFarland  1975).  With the  vast array of special tests that
are  being developed and the many  potential  effects that  could be moni-
tored,  the difficult task is to decide  which  aspects of inhalation  toxicity to
examine, as a truly comprehensive evaluation for all agents is impractical
(Campbell  1976, Clark 1977).

4.2.5.1 Clinical observations
    During  and following exposure, observation of the test animals for clin-
ical  signs can often indicate systemic or lung damage.  All animals should
be observed at least  daily and preferably twice  a day (NAS 1977, OECD
1979,  Gross  198la).  Observations should  include any signs of irritation;
changes in  skin, fur, eyes, or  mucous membranes; tremors; convulsions;
salivation;  bleeding; coughing; frothing at  the mouth;  presence and condi-
tion of urine and/or feces; lethargy; and changes in respiratory rate or pat-
tern (NAS  1977,  OECD 1979,  Page et al. 1980). The observer should be
particularly alert  for moribund animals,  in  order  to prevent loss from
autolysis or cannibalism.  In long-term studies,  the  animals can also be
weighed once  a week and the food consumption calculated.

4.2.5.2 Biochemical and hematological tests

    Hematological and biochemical  tests, including enzyme assays, should
be performed on all  animals before  and  after exposure (or at any interim
sacrifice during exposure).  Frequently performed hematology tests include
hematocrit, hemoglobin concentration,  erythrocyte  count, total and  dif-
ferential leucocyte counts, prothrombin time,  partial thromboplastin time,

-------
                                  159

platelet count,  and  a blood  smear  examination  (OECD  1979,  Gross
198la). Many biochemical  tests have  been used  in inhalation studies,
including  alkaline  phosphatase,  serum glutamic-pyruvic  transaminase,
serum  glutamic-oxaloacetic transaminase, blood urea nitrogen, total  pro-
tein, albumin, globulin,  cholesterol,  fasting  glucose,  calcium,  sodium,
chloride, and potassium (OECD 1979). Such tests can give information on
systemic effects  or  suggest  additional  special  tests.  The use of special
biochemical evaluations of lung tissue has also been  suggested, particularly
for intermediate  metabolism (Tierney  1974),  and  pulmonary surfactants
(King  1974). The majority of biochemical tests indicating lung damage are
currently in the  developmental stage with  many methods being modified
from  established  hepatotoxicity  evaluations (Witschi  1975). Although
these studies have not yet identified a biochemical  parameter that clearly
indicates toxic effects  in  the  lung, an  encouraging  beginning has been
made (Witschi 1975). Biochemical analysis of pulmonary  lavage fluid has
also shown promise  as an evaluative tool of inhalation toxicity. Henderson
and coworkers (Henderson et al.  1978, 1979b, 1980) have used the  lavage
techniques developed by Mauderly (1977) as a short-term  in vivo screen  of
acute  toxicity in the rat and hamster. By  measuring levels of  various
enzymes and  components in the  fluid (Table 4.2) after exposure to toxic
agents, they determined that lavage was a more sensitive  and earlier indi-
cator of initial lung  injury than was analysis of lung tissue. With continued
development, it may be a  useful technique for evaluation of chronic pathol-
ogy. Other in vivo and in vitro assays may find some applications as alter-
natives to traditional inhalation toxicity  studies  (Brown and Poole  1983).
These  alternatives include in vitro studies using immunocompetent cells  of
the lungs, cytoxicity and cell morphology of cell cultures,  and specific cel-
lular biochemical activities as end points (Brown and Poole 1983).

4.2.5.3 Pathological  observations

   Evaluation of effects  by pathology  studies  is  the  main method for
determining inhalation toxicity. As in other types  of toxicity tests,  it  is
essential to  have pathological  observations that include both a thorough
gross examination of all  animals  and a  detailed microscopic examination
of, at  least,  the high-dose,  control,  and lesion-bearing  animals  and the
suspected target organs  of animals  at the  intermediate levels  (Gross
1981a).
   The gross examination or necropsy is often more involved for inhala-
tion studies than for other routes of  exposure (Dungworth  et al.  1976),
with examination of additional respiratory tissues  (nasopharynx, larynx,
tracheobronchial  tree,  and  parenchyma).  The  gross necropsy typically
includes an  examination of the external  surfaces of the body; all orifices;
and the cranial, abdominal, and thoracic cavities, including the organs and
tissues contained within them (OECD 1979). Improved evaluations of the
morphology of the nasal  turbinates and mucosa have been recommended
by Jersey and Kociba (1979), who suggest that grading the extent and

-------
                                  160

             Table 4.2. Parameters Measured in Lavage Fluid

                 Parameter                   Indication

         Cytoplasmic enzymes:         Membrane damage
           lactate dehydrogenase;
           glucose-6P-dehydrogenase
           organ specific enzymes;
           isoenzymes

         Lysosomal enzymes:          Phagocytic activity
           acid phosphatase;
           beta-glucuronidase

         Membrane enzyme:          Broken membranes
           alkaline phosphatase

         Mucin component:            Increased mucin secretion
           sialic acid

         Cell count:                   Inflammatory response
           total; differential

         Source: Adapted from Henderson et al. (1980).
severity  of  the  lesions can give  dose-response relationships. During  the
necropsy, examination of the structures within the lung depends  on  the
extent of gross damage; when edema  or exudative lesions are absent,  the
airways and parenchyma are best examined after fixation (Dungworth et
al. 1976).
   After the necropsy, samples of all organs and tissues are preserved by a
fixative solution. The lung requires  special  attention to ensure  correct
preservation. The preferred technique is to perfuse the lung via the  air-
ways, because this maintains the  correct dimensions  and configuration of
the tissues while permiting rapid  fixation  (Dungworth et al. 1976, NAS
1977).  The  fixative  (10%  neutral  buffered formalin  or  Karnovsky's
formaldehyde/glutaraldehyde formula) should be perfused at a pressure of
25 to 30 cm of water for 1 to 4 h (Roe 1968, Dungworth et al. 1976).
   The microscopic examination should include a wide array of body  tis-
sues  and organs in addition to respiratory structures in order to assess both
the systemic effects  and the specific  effects  from inhalation. The exact
organs to be sampled vary with the chemical and its potential effects; some
suggested lists  are given in Table 4.3.  The  assessment  of  specific lung

-------
                                    161
            Table 4.3. Organs Suggested for Microscopic Examination*
         NASb
         WHOC
        OECD"
Bone
Bone marrow
Brain
Eye
Gonads
Heart
Intestine, large
Intestine, small
Kidneys
Liver
Lung
Lymph node, mesenteric
Lymph node, peribronchial
Muscle, skeletal
Nasal cavity
Nerve, peripheral
Pancreas
Pituitary
Prostrate
Salivary gland
Skin
Spleen
Stomach
Thyroids
Trachea
Urinary bladder
Uterus
Accessory genital organs
Adrenals
Aorta
Bone marrow, sternum
Brain
Caecum
Colon
Duodenum
Gall bladder (if present)
Gonads
Heart
Ileum
Jejunum
Kidneys
Liver
Lung
Lymph nodes, axillary
Lymph nodes, mesenteric
Mammary glands
Muscle, thigh
Nerve, peripheral
Oesophagus
Pancreas
Pituitary
Rectum
Salivary gland
Spinal cord (at three levels)
Spleen
Stomach
Thymus
Thyroid
Urinary bladder
Accessory genital organs
Adrenals
Aorta
Brain
Caecum
Colon
Duodenum
Gall bladder (if present)
Gonads
Heart
Ileum
Jejunum
Kidneys
Liver
Lung
Lymph nodes, axillary
Lymph nodes, mesenteric
Mammary glands
Nerve, peripheral
Oesophagus
Pancreas
Pituitary
Rectum
Salivary gland
Spleen
Stomach
Thymus
Thyroid
Urinary bladder
"Gross abnormalities or lesion-bearing tissues detected in gross necropsy should also be
examined.
bNAS = National Academy of Sciences (1977).
CWHO = World Health Organization (1978, Chapter 5).
dOECD  = Organization for Economic Cooperation and Development (1979).

-------
                                  162

effects requires sampling that  is wide in distribution but specific in ana-
tomic location because the size, diversity, and variable responses of the
respiratory tissue are so great (Dungworth et al. 1976). Generally, samples
to be taken include: (1) a longitudinal section of the distal trachea and the
bifurcation into bronchi; (2) vertical sections of one lung and the cranial,
middle, and caudal  lobes of  the  other  lung; (3)  the  proximal  trachea;
(4) the lobar bronchus; (5) the pharynx; (6) the larynx; and (7) the prox-
imal and  distal regions of the  nasal sinuses and turbinates (Dungworth et
al.  1976). These  samples should be initially evaluated by light microscopy
and/or scanning electron  microscopy.  If  additional  detail  is  desired,
transmission electron microscopy can be used (Dungworth et al. 1976).
    Occasionally, to fully elucidate  the toxic effects, special pathology tech-
niques may  be employed, including autoradiography, histochemistry, and
morphometry studies. Autoradiography is useful in determining cytokinet-
ics of damaged pulmonary cells (Dungworth et al.  1976) and the correla-
tion between deposition  of particles and  effect (Lisco 1959). Histochemis-
try is useful in  localizing biochemical changes to specific cells  or tissue
types (Dungworth et al. 1976), whereas morphometic analysis is useful for
accurately measuring the degree of damage and for statistically confirming
subtle effects (Dungworth et al. 1976; WHO 1978). Although these pro-
cedures  are  not routine, they  may  serve  a useful function in toxicity
evaluations.

4.2.5.4 Respiratory function tests
    The  respiratory function tests are observation methods for  assessing
toxic effects that are uniquely used in inhalation studies. These tests allow
the inhalation toxicologist  to  observe  the  effects  of  the test  agent on
breathing pattern,  lung volumes, pulmonary pressures,  flow volumes, and
lung mechanics  (Table  4.4). Other tests can be  performed to determine
changes in gas exchange and gas levels in the blood (Likens and Mauderly
1979).  Generally,  the respiratory  function tests can be applied to intact
animals,  can be performed rapidly  and repeatedly without affecting the
animal, and will produce dose-response relationships, often with low varia-
bility  (Swann et al.  1965,  Alarie 1966, Hiett 1974,  MacFarland  1976,
WHO 1978,  Wong and Alarie 1982, Juhos  et al.  1985). One key to these
methods is the understanding of baseline or  mean  values to allow interpre-
tation of responses as toxic or  nontoxic  (Kennedy  and Trochimowicz
1982). Because respiratory function can vary greatly between test animals,
it is vital to have a series of baseline data for each animal prior to dosing.
Concurrent evaluations of control animals can demonstrate any changes in
respiratory function over time  due to  aging  and  repeated  testing. The
respiratory function tests are useful in acute studies for delineating effects
and in chronic tests for following the development of effects (MacFarland
1976).  They can also provide information  on the mechanisms  of action
(Amdur  1958,  MacFarland  1976), although such data  are not  always

-------
           Table 4.4. Respiratory Function Tests to Evaluate Breathing Patterns, Lung Volumes, Pulmonary Pressures, and Lung Mechanics
           Test
                                       Definition
                                       Method of measurement"
                                                       Reference
Frequency (Respiratory rate)
Number of respirations per
minute
Tidal volume (VT)
Minute volume (MV)
Vital capacity (VC)
Inspiratory capacity (1C)
Expiratory reserve volume
(ERV)
Volume of air inspired during
a quiet breath
Volume of air expired per
minute
Change in volume between
maximal inspiration and maxi-
mal expiration

Maximum volume that can be
inhaled after a quiet expiration

Volume that can be exhaled
after a quiet inspiration
Visual count; volume pneumograph
(detects pressure changes in thorax);
impedance pneumograph (uses electrical
circuit to measure impedance change
resulting from thorax movement); tem-
perature change across nostrils; non-
rebreathing valve

Respirometer; plethysmograph used  with
a volume transducer; non-rebreathing
valve (calculated value); pneumotachc-
graph

Non-rebreathing valve; respirometer (cal-
culated  value); plethysmograph with vol-
ume transducer (calculated value); pneu-
motachograph

Plethysmograph (used with applied pres-
sure); air injected into lung via syringe
Plethysmograph (used with applied pres-
sure); air injected into lung via syringe

Plethysmograph (used with applied pres-
sure); air injected into lung via syringe;
pneumotachograph
McCutcheon 1951, Alarie 1966, Barrow
et al.  1971, Mauderly and Tesarek 1975,
Crossland et al. 1977
McCutcheon 1951, Leong et al. 1964,
Thomas and Morgan 1969, Diamond and
Lipscomb 1970, Barrow et al. 1971,
Barer et al. 1976

Truog and Standaert 1978, Leong et al.
1964, Diamond and Lipscomb 1970,
Thomas and Morgan 1969
Comroe et al. 1954, Drorbaugh 1960
Comroe et al. 1954, Drorbaugh 1960
Comroe et al. 1954, Drorbaugh 1960

-------
                                                            Table 4.4 (continued)
            Test
                                        Definition
                                                                     Method of measurement"
                                                                                                                     Reference
Total lung capacity (TLC)
Residual volume (RV)
Functional residual capacity
(FRC)
Intrapleural pressure


Esophageal pressure



Transpulmonary pressure
Volume of gas contained in
lungs at maximum inspiration
Volume of gas remaining in
the lung after maximal expira-
tion
Air remaining in lungs at end
of quiet expiration
Pressure within the pleural
space

Pressure within the esophagus
(approximates intrapleural
pressure)

Pressure difference between
pleural pressure and tracheal
opening
O2 absorption technique (used with
plethysmograph) calculated by adding
residual volume and vital capacity values

Calculated by subtracting expiratory
residual volume from functional residual
capacity values; determining amount of
inhaled inert gas retained in lungs after
exhalation

Inert gas dilution (uses tracheal cannula,
O2 chamber  and percent N2 to calculate
value): FRC = ERV + RV

Intrapleural catheter
Fluid-filled or balloon-tipped esophageal
catheter
Differential pressure transducer con-
nected to pleural or esophageal catheter
and mask or breathing chamber
Cavagna et al. 1967
King 1966a, Caldwell and Fry 1969,
Koo et al. 1976, Koen et al. 1977,
Diamond and O'Donnell  1977

Amdur and Mead 1958
 Cavagna et al.  1967, Caldwell and Fry
 1969, Koo et al. 1976
Amdur and Mead 1958, Palecek 1969

-------
                                                            Table 4.4 (continued)
           Test
                                       Definition
                                       Method of measurement"
                                                                                                                    Reference
Compliance
Resistance
Forced vital capacity; forced
expiration volume; maximum
expiratory flow-volume
Distensibility of the lungs and
thorax (relation of transpul-
monary  pressure to lung vol-
lume changes, C = AV/AP)
                              Represents the frictional and
                              viscous forces that occur with
                              respiration (relation between
                              transpulmonary pressure and
                                R = AP/AV)
Dynamic forced respiratory
characteristics useful for
detecting changes in upper and
lower airway resistance
Dynamic: plethysmograph (for volume)
and pleural or eosphageal pressure (cal-
culated on analog computer, from oscillo-
scope displays, or by manual calcula-
tions); static:  pressure volume curves gen-
erated by plethysmograph and air
injected in lungs; forced oscillation of res-
piratory movements in a plethysmograph
using volume  changes measured by a
pressure transducer

Dynamic: plethysmograph (for volume),
pleural or esophageal pressure, and pneu-
motachograph (for flow) (calculated
using same techniques as compliance);
static: change in trans-pulmonary pres-
sure at end expiration (using a pneumo-
tachograph);  forced oscillation of respira-
tory movements in plethysmograph

Whole-body plethysmograph used with
negative and positive  pressure chamber;
whole-body respirator and pneumotacho-
graph
Davis and Morris 1953, Radford 1957,
Amdur and Mead 1958, Dennis et al.
1969, Nattie 1977, Diamond et al. 1973,
Hiett 1974, Decker et al.  1979
                                                                        Amdur and Mead 1958, Mead 1960,
                                                                        Murphey and Ulrich 1964, Frank and
                                                                        Speizer 1965, King 1966b, Palecek 1969,
                                                                        Giles et al. 1971, Diamond et al. 1973
Diamond and O'Donnell 1977, Moorman
et al., 1977
"Note: pneumotachograph can be used to measure all lung volumes except for RV. It is generally the most precise and easy to use means for measuring
flow and volume (by intergration).

Source: Adapted from Likens and Mauderly (1979).

-------
                                   166

necessary in toxicological studies.  One important advantage that  respira-
tory function tests have over other evaluations of toxic effects is that they
can quantify certain changes (e.g., reflex bronchoconstriction) at  concen-
trations  below  the  levels  that  produce  morphological damage (Amdur
1958,  WHO  1978).  Therefore,  they  may  allow measurement of  such
effects  at  levels of reversible rather  than irreversible changes. Another
advantage,  particularly  for  chronic  tests,  is  that  the  same  animal's
responses can be followed over  the entire exposure period, thus avoiding
the variability introduced by testing different groups at each point in time
(Gross  1981b).  Finally,  similar  data can often be  generated for  humans
and test animals, which facilitates comparisons (WHO 1978).
    The  respiratory  function  tests  do have  disadvantages. First,  they
require  special  equipment  and  training  for proper  use  (NAS  1977).
Second, many of the tests (e.g.,  vital capacity or residual volume)  used on
humans cannot  be used on test species without  some  modification or
anesthesia because,  as currently administered, the tests require the under-
standing  and cooperation of the  test subject (Roe  1968, MacFarland
1976). Third, some  tests can be  used only with anesthetized animals under
highly unnatural physiological conditions (Roe 1968). Finally, the informa-
tion gathered by these tests is rarely conclusive regarding the total effects
of  a  test  substance and  must  be  supplemented  by the other evaluation
methods discussed in  earlier sections, especially histopathology (Roe 1968,
MacFarland 1976).
    In addition to the  tests of the mechanics,  volumes, and pressures associ-
ated  with respiratory function,  tests  to  evaluate ventilation distribution,
diffusion, and mucociliary action have been  developed. A test that can be
used  to evaluate effects on the  ventilation distribution is the nitrogen
wash-out technique (MacFarland 1976). Generally, the technique  meas-
ures the decline in nitrogen concentration of expired air after switching to
pure oxygen inspiration. A rapid, exponential decline indicates normal ven-
tilation distribution, but  an uneven or less steep slope  indicates  an irregu-
lar ventilation  distribution. Effects on the  diffusion  capabilities can be
tested  by determining the rate at  which  carbon monoxide or  oxygen
transfers from the alveolus to the hemoglobin (MacFarland 1976, Likens
and Mauderly  1979). One way to  test this process is a  closed-circuit
rebreathing technique using carbon monoxide. The rate of decline of the
carbon monoxide indicates the diffusing capabilities. Diffusion capabilities
can also be determined  during a  single vital capacity inspiration and
expiration.  Measurement of blood  gas (oxygen, carbon  dioxide)  levels is
another technique suggested for evaluating respiratory function. However,
these tests can  be difficult to perform in small animals and often require
surgical approaches (Likens and Mauderly 1979). To  determine effects on
the mucociliary capacity of the lungs, a tracheal preparation can be taken
from an  exposed  animal, test particles (e.g., pollen grains) applied, and
their progress between two points observed. Another test uses radiolabeled

-------
                                  167

particles in a live test animal and monitors their progress with a scintilla-
tion detector (Troy 1974). This latter approach was used by Kenoyer et al.
(1982)  to determine the toxic effects  of  ozone on short-  and long-term
clearance rates. They found that ozone exposure delays the  initial clear-
ance (perhaps  by  affecting ciliary beat rate  or mucus changes), but the
long-term clearance  is  accelerated (perhaps due to greater activity of
deep-lung macrophages). These  tests would  be  most useful in  a study
designed to evaluate only effects  on the lung,  because they are too compli-
cated for use in a routine test of systemic toxicity.

4.3. EXPOSURE CHAMBER DESIGN

4.3.1 Introduction

   The design  of the exposure chamber should be one of the first factors
considered in  an inhalation test to ensure that the requirements  of the
exposure protocol are met. Specifically designed buildings and facilities are
preferred for such tests.  The building should have a high ceiling to accom-
modate chambers  and necessary air  ducts, have plenty of floor space to
allow access to at least  two and preferably all  sides of the chamber, pro-
vide good illumination without  allowing direct sunlight exposure  of the
chambers, and  have an  adequate air supply system with appropriate con-
trols for humidity, temperature, and flow (Roe  1968, Drew and Laskin
1973, WHO 1978).  The exposure chamber  units should  be designed to
meet  the majority of the  basic requirements  listed  in  Table 4.5. This
chapter will deal with the necessary equipment to meet such requirements,
including discussions of the various  types of exposure  systems, chamber
design,  air supply and exhaust  systems, monitoring of the chamber,  and
use of  specialized  equipment. Consideration of test  agent  delivery  and
atmosphere generation equipment are covered  in Section 4.4.

4.3.2 Types of Exposure  Systems

   This section will focus on the various operation modes possible for
inhalation chambers. The administration  of  the test agent  can  be in a
static or dynamic mode  of operation. Also the  animals can be exposed to
the test agent  via whole-body,  nose-only  or  head-only exposure  systems.
The advantages, disadvantages, and uses of each system will be discussed.

4.3.2.1 Static systems

   A static exposure  system consists of a closed chamber into  which a
predetermined  dose  of  the  test  agent is  introduced and  mixed  (Lodge
1968). The  test animals remain  in this closed system for  the duration of
the test without  any replacement  of the  test atmosphere. Obviously the
lack of replacement  air will limit the duration of  exposure, because  it
allows depletion of the oxygen concentration,  increased carbon dioxide lev-
els, increased  temperature  (from buildup of  animal  body heat),  and

-------
                                  168
       Table 4.5. Inhalation Exposure Apparatus: Basic Requirements
Exposure chamber
Adequate size for exposure of sufficient animals for statistical evaluation
Even distribution of test gases, aerosols, or particle clouds
Temperature control
Humidity control
Easily cleaned and decontaminated
No hazard to personnel
Identical chambers available for untreated or vehicle-only treated controls
Test material
Gases  — reliable control of concentration in chamber
Aerosols and particle clouds — reliable monodisperse system
as well as control of concentration; vehicles, if used, should
be without effect on respiratory system
Monitoring of exposure
Apparatus should  be  available for monitoring the concentration
of gases, particle cloud density, and distribution of particle
size at breathing zone, continuously during exposure
Safety
The absence of leaks  in the chambers or
connecting pipe systems should be ascertained
repeatedly, preferably continuously, by a suitably designed
monitoring system where hazardous agents are under test
The atmospheric pressure within the inhalation exposure system
should be slightly below the ambient atmospheric pressure
Suitable traps and filters should guard both ends of all pipeflow systems
Source: Adapted from Hinners et al. (1966), as reported in Roe (1968).

-------
                                  169
increased humidity (MacFarland 1976, Clark 1977, NAS 1977). An expo-
sure duration  of 0.5 to 1 h has been  suggested for smaller static systems
(MacFarland  1976).
   In addition to the problems associated with a finite volume, there is
also  a loss  of test  agent by deposition  on the chamber walls or by removal
for sampling  purposes (Lodge  1968,  Drew  and Laskin  1973).  Although
many of these problems can be eliminated or reduced by use of a large-
volume  chamber  (e.g.,  a minimum of 500  L/kg test animal weight  for
every hour of exposure, Clark  1977), nonintrusive sampling techniques, or
nonreactive wall materials, there is still one major disadvantage. Because
the total amount  of test  agent  must be introduced at the beginning of the
exposure, the  concentration declines with time from an initial peak higher
than the desired exposure to  a  final level below the desired exposure. Even
though  the average concentration value of the  entire  exposure  may be at
the desired level,  the variations introduce uncertainty and prevent the uni-
form exposure  concentration  that is  preferred  for inhalation  studies
(MacFarland  1976).
   Despite these  drawbacks, there are several situations  for  which the
static system  is the  best choice.  Static systems  are  useful for assessing
acute toxicity and in performing  preliminary pilot studies  (WHO  1978).
Also, if the test agent is available only in limited quantities, a static expo-
sure should be used (MacFarland 1976).  A static  system is likewise
preferred in the testing of biological aerosols as these are difficult to gen-
erate continuously (Drew and Laskin 1973).  The majority of inhalation
tests, however, will use the dynamic system.

4.3.2.2 Dynamic systems
   A dynamic exposure  system consists of a chamber that has a continu-
ous flow through  of air and test agent  at a constant rate and in a predeter-
mined ratio (Silver 1946). The constant airflow avoids many of the prob-
lems that limit the use of static systems and is usually generated by plac-
ing an  air pump in the exhaust  line of the system. Such  a  placement
creates  a  negative pressure in  the chamber, which  will aid in  preventing
loss  of  the test agent through  leakage (Silver 1946, MacFarland  1976).
Occasionally,  to ensure an even airflow, buffer tanks are used to eliminate
pump-induced flow fluctuations (Silver 1946). Usually the test animals  are
inserted in the chamber prior to exposure and then the gas flow is started
(MacFarland  1976). This creates  an exposure curve of the form shown in
Figure 4.1. If immediate exposure to  a constant  concentration is desired,
an airlock  device,  usually built into the chamber door, can be used (Silver
1946). The animals are placed  in  the airlock, the outer door is closed, and
the airlock is  opened to the chamber atmosphere. This produces the expo-
sure curve  shown in Figure 4.2.

-------
                                 170
                                                    ORNL-DWG 81-11341
 O
  QC
  LU
  O
  2
  O
  O
      0
                      '99
                                 TIME
    Figure 4.1. Concentration-time relationships in  a  chamber operated for
 a long period of time. ta = Start of flow through  chamber; t99 =  time at
 which equilibrium  concentration is reached; tb = time at which test agent
 is no longer added to airflow;  and tc = end of airflow through chamber.
 (Source: Adapted  from  MacFarland  1976. Used   with  permission  of
 Academic Press.)
                                                    ORNL-DWG 81-11340
         CC
         LJ
         O
         2
         O
         O
                        '0
                                TIME
   Figure 4.2.  Time-concentration curve for exposure to constant concen-
tration using air-lock mechanism. t0 = Animals introduced into chamber
and i\  =  animals removed from chamber. (Source: Adapted from MacFar-
land 1976. Used with permission of Academic Press.)

-------
                                 171
   In normal use situations, it takes  a certain amount of time for the
chamber concentration to reach the desired level (see Figure  4.1).  Silver
(1946) developed the following equation to calculate this time lag:
                                         ^w-|
                        C  =  -11 -
where
      C  =  the concentration (mg/L) at time t,
      w  =  mg agent introduced per minute,
      a  =  the volume of chamber (L),
      b  =  the volume of air passing through the chamber each minute,
      e  =  natural logarithm

Two assumptions, constant flow and complete mixing, which are generally
satisfied, are required for this equation (MacFarland 1976).  Because this
is  an exponential rate,  Silver assumed a value of 99% to be sufficiently
close to the desired equilibrium value  for testing; based on that value the
following  equation  can  be used  to  determine when  the  chamber has
reached equilibrium:

                           T99=4.6052-f-  .
                                       b

MacFarland (1981) points out that,  after the T99  concentration level is
reached, the aerosol distribution should be uniform within the chamber, as
inlet air would not differ by more than 1% from the chamber concentra-
tions in most chamber arrangements.  Other numerical values for the con-
stant (4.6052) are given by Silver (1946) for percentages lower than 99%.
Because the T99 value depends on  chamber size and flow rate,  one  could
reduce the time to equilibrium by increasing the flow (for both  test  agent
and total flow) above that required for the desired exposure (MacFarland
1976). Once the T99  equilibrium has  been reached, the flow rate can be
reduced. Another technique to avoid the delay would be to use dual rate
supply systems. With the latter method, one must be careful  that the con-
centration does not exceed the desired level (MacFarland 1976). The  T99
equilibrium  is a theoretical value,  and,  although  the chamber  should be
uniformly mixed, the actual chamber concentrations may be lower due to
losses of test agent from adsorption on walls or animal  fur and  uptake by
the  animals. Therefore, monitoring of the  actual  concentration in  the
chamber is always necessary to ensure  proper exposure levels (Lodge  1968,
Drew and Laskin 1973).
    The appropriate flow rate to  be used in  dynamic chambers must  fall
within a limited range if the chamber  is to avoid the  heat buildup and oxy-
gen loss problems that occur with static chambers. MacFarland (1976)
discusses the ranges  of flow  rates commonly used  and recommends  the
following equations:

-------
                                  172

        F  =  0.1 to 0.2V for gases,
        F  =  0.2 to 0.5V for volatile vapors,
        F  =  0.5 to 0.7V for light aerosols, and
        F  =  0.7 to 1.0V for heavy aerosols,
where

        F  =  flow rate in liters per minute
        V  =  chamber volume in liters.

For example, if the chamber size is 1000 L, the flow rate should be in the
range of 100 to 200 L/min for a gas, but 700 to 1000 L/min for a heavy
aerosol. If flow rates are calculated in this manner and test animals loaded
in the chamber at a density (animal volume per chamber volume) of 5% or
less, the use  of special air conditioning systems are not necessary. EPA
guidelines require a dynamic air flow of 12 to  15  air changes per hour to
ensure an oxygen content of 19% (USEPA 1982). The flexibility shown by
the dynamic  exposure system makes it the preferred choice  for chronic,
subchronic, or any repeated exposure studies where test agent availability
is not a limiting factor. Dynamic chambers are required for testing within
the EPA guidelines (USEPA  1982).

4.3.2.3 Whole-body exposure systems
    Whole-body exposure systems,  especially when operated dynamically,
are the most frequently used exposure systems in inhalation  testing. The
animals are placed in cages within the exposure chamber  and exposed to
the test  agent  atmosphere. Whole-body exposure requires no surgical or
anesthetic preparation prior  to  testing (MacFarland 1976). The animals
are free to move within their cages, and the use of exercise devices can
provide information on  the  effects in active  animals.  With whole-body
exposure systems, large  numbers of animals or several different species can
be  exposed simultaneously in one chamber (MacFarland 1976). Uniform
exposure conditions for  all animals are obtained even when large numbers
of animals are exposed. The  whole-body system does, however, have some
disadvantages. Deposition of the test agent (especially aerosols)  on animal
fur with the  potential for ingestion during grooming behavior is a major
problem (MacFarland   1976,  Nettesheim  and  Griesemer  1978).  The
animals should be housed in  individual cages to reduce such ingestion and
to avoid dosage interference resulting from  filtration through the fur of
huddling  animals  (MacFarland  1976, Clark  1977).  When whole-body
exposure systems are  operated in the usual dynamic mode,  large quantities
of the test agent are  required, which can be a disadvantage (MacFarland
1976), and, if the test agent is extremely hazardous,  special handling tech-
niques and equipment,  such  as glove  ports,  must be used to protect the
laboratory  technicians  (Nettesheim and  Griesemer  1978).  These  draw-
backs have stimulated the development of the alternative exposure systems
discussed in Section 4.3.2.4.

-------
                                  173

4.3.2.4 Nose- and head-only exposure systems

   In some inhalation studies, exposure of the whole-body is  not  accept-
able; for these situations nose-  or  head-only  exposure  systems have been
developed.  Basically, these systems  consist of a central chamber into which
the test agent is introduced and separate holding devices that  position the
animal  so  that only the  nose or  head is protruding into the  test atmo-
sphere.  The design of the central chamber may be circular, hexagonal, or
linear and  be constructed of Plexiglass® or stainless steel (Thomas and Lie
1963,  MacFarland  1976).  The  nose,  muzzle, or  neck of the  exposed
animals must  be  sealed to prevent loss or dilution of the test agent. The
holding  devices   include  cylindrical   Plexiglass®   containers   (e.g.,
plethysmograph) used mainly for rodents, sling arrangements for dogs, and
adjustable  chairs for primates (MacFarland  1976).  If exposure of many
animals is  desired, cylindrical or hexagonal chambers  can be stacked to
provide a larger central chamber for exposure (Drew and Laskin 1973). If
a  linear arrangement  is  used  to repeatedly  expose  large  numbers of
animals, the positions  of the animals  must be  rotated  between exposures
because the linear arrangement can affect test agent  concentrations. An
alternative  method involves using a molded face mask to deliver the test
atmosphere. This system also requires  some type of restraining device and
an airtight seal between mask and animal to prevent  air loss and eliminate
dead space. Latex  molding techniques have proved satisfactory for main-
taining  a tight seal and have been fitted to  dogs,  rodents,  and primates
(Dubin  and Morrison 1969, MacFarland 1976).
   The nose-  and head-only systems are most useful for studies where no
or little dermal or oral exposure can be tolerated (Drew and Laskin 1973,
MacFarland 1976). Both systems reduce the  total area exposed, with the
face mask  system  exposing  the least  surface  area. A  comparison of a
whole-body exposure  system and  two  nose-only  exposure  systems  (a
cylindrical  polycarbonate tube and  a stocklike restrainer) using radioactive
tracers  indicated  that  the  nose-only exposures  were  much more accurate
for delivery of inhalation doses  (Henry et al.  1983). Data show that only
13% of the total dose was deposited in the lungs of mice during a 130-min
whole-body exposure, with 60 to 80%  of the dose eventually reaching the
gastrointestinal tract. The nose-only  exposure systems delivered approxi-
mately threefold to fivefold more test  material  to the lungs and threefold
less material to  the gastrointestinal tract under the  same test conditions.
These  systems are also advantageous  when  the  test  agent is extremely
hazardous  (e.g., radioactive aerosols) or when  it is available only in limited
quanities (Drew and Laskin 1973,  WHO 1978). For extremely hazardous
substances, the entire exposure apparatus can  be set up within  an enclosed
inhalation  chamber with glove boxes or a rolling drawer (Moorman 1978).
This provides  a double measure of safety, enables animals to  be manipu-
lated,  and  allows  rapid  reestablishment  of  exposure  conditions. Other
advantages include the  potential for slower air flow through the chamber,
because body heat buildup is not a problem  inside the chamber,  and the

-------
                                  174
capability for rapid exposure of the animals to the test atmosphere (espe-
cially for aerosols)  due to the  small  chamber  size  (Wells 1979). One
unique application of these systems is a design that exposes animals either
dermally and/or by inhalation, depending on how the  animal holder tubes
are inserted into the chamber (Francovitch  et al. 1982). Most applications
of these systems are for acute testing, as such tests do not require  feeding
or watering  of the  animals during  exposure (Drew  and  Laskin  1973).
Long-term testing is  difficult due to the need to restrain the animals.
   One of the major disadvantages of these exposure  systems is the diffi-
culty of exposing large numbers of animals (MacFarland 1976). Often a
large crew of technicians  is needed if  many  animals are  to be exposed
(Nettesheim  and Griesemer 1978). Other disadvantages are caused by the
extra handling and restraints required. The physical stress  created or the
anesthetic required to avoid this stress can affect the outcome of the expo-
sure (Drew and Laskin 1973). One aspect of physical  stress is the  buildup
of body  heat  within  the  restraining tube. Perforations in  the tube or
other means  of  ventilation  can offset  this  disadvantage  (Kennedy  and
Trochimowicz  1982).  Nettesheim  and  Griesemer (1978)  found reduced
body weight  gain and increased  adrenal gland weight in animals exposed
by  these  systems.  However, Smith et al. (1980) reported on a nose-
exposure system using a polycarbonate tube for  chronic aerosol exposures
of fibers that did not affect stress as measured by body weight, tempera-
ture, blood counts,  and plasma  corticosterone  levels.  Thus,  some of the
nose- and head-only exposure  systems may  be  useful  for  long-term,
repeated exposure tests  but  are  best utilized for the  specialized exposure
conditions described.

4.3.3 Chamber Design
   As mentioned previously,  the majority of  inhalation  chambers are
designed for dynamic, whole-body exposures. The discussions in this sec-
tion will  pertain to  this type of chamber  unless otherwise specified. In
designing an  inhalation  chamber the principal objectives are performance
and  ease of operation (Fraser et al. 1959).  Performance refers to the abil-
ity of the  chamber to meet the  requirements of the  particular  inhalation
study being  undertaken. Ease of operation  includes ready  accessibility of
animals  for   sampling  and  handling,   and maintenance  and   sanitation
considerations. Due  to the wide variety of potential  test substances and
their different properties, many modifications  of  a  few  basic chamber
designs have  been developed. This section  will deal with the main design
considerations of shape, size, construction materials, and some operational
features of the basic  chambers. Modifications of these basic chambers will
be discussed, depending on their potential utility.

-------
                                   175

4.3.3.1 Chamber shape

   The shape of inhalation chambers is  determined by requirements for
uniformly distributing the test agent and minimizing the internal surface
area in order to reduce adsorption and deposition of the test agent (Silver
1946). Theoretically, a  large  volume with  a  small surface area  is  the
desired condition.  This is achieved best by a chamber  of spherical shape
(Fraser et al.  1959); however, such a chamber is difficult to build and fit
with doors or windows (MacFarland 1976). The practical solution is to use
a chamber with a  cubical shape, but this results in 25% more surface  area
than does the  sphere (Frazer et al.  1959, Hinners et  al.  1966, MacFarland
1976). An intermediate design is a hexagonally shaped chamber, but the
gain in efficiency  is usually  not worth the increased cost of construction
(MacFarland  1976).  By adding pyramid (or cone)  structures to the  top
and bottom with air entering at the top and exhausting  at the bottom,  and
by avoiding animal placement in the corners, the air  distribution efficiency
can be increased and a cubical chamber used for most applications (Fraser
et al. 1959, Roe 1968).
   Large-scale exposure chambers,  incorporating most of the  shape  con-
siderations discussed above, are best represented by the Rochester and the
New York University chambers. The Rochester chamber (Figure 4.3)  is  a
hexagonal-shaped  chamber with conical additions to the top and bottom
(Fraser et al.  1959). The New York University chamber (Figure 4.4) is
based on a cubical shape with pyramid additions (Drew and Laskin 1973)
and is considered the system of choice for repeated exposure studies (NAS
1977).
   Other large-scale designs have  been used for specialized exposures or
when unique equipment is available. Spherical chambers composed of clear
Lucite® plastic have been used for exposures to carcinogens at  low  flow
rates  and can be built at a lower cost  than steel chambers (Stuart et al.
1970). The Longley chamber is a modification of the standard rectangular
design and is  subdivided into two  or  four  separate sections  (MacEwen
1978). This  allows simultaneous exposure of control and three dose levels
of test animals. Also, the  design  is  flexible enough to permit  testing of
multiple dose levels for acute  studies.
   Carcinogenic studies often require specialized chamber designs because
of the long duration of the studies,  the large number of animals,  and need
to isolate the carcinogens. Laskin et al. (1970) designed a system of three
box chambers  that are connected by internal  sliding doors. One chamber is
used for  the inhalation exposure while the other two chambers house the
control and  experimental animals.  Pass boxes and  neoprene gloves built
into the system  allow the animals, food, and wastes  to be introduced  into,
moved within, or removed from the system. Thus, the animals never have
to be removed from the unit and can be totally isolated. Entire rooms have
also  been designed  as large-scale  exposure  units  for exposures  of many
animals.  Chambers with an 2.4 m  X  2.4 m floor area and a submarinelike
door hatch at each end  were designed  at Dow  Chemical Corporation for

-------
                                176
                                          ORNL-DWG 85-10742
          CONTAMINANT
              SUPPLY
SUPPLY
  AIR
           EXHAUST
            AIR'
   Figure 4.3.  Schematic diagram  of  the  Rochester exposure chamber.
(Source:  Adapted from Drew and Laskin 1973. Used with permission of
Academic Press.)

-------
                                 177
              EXHAUST  AIR
SUPPLY  AIR
ORNL-DWG 85-10743

 CONTAMINANT
     SUPPLY
                                    DRAIN
   Figure  4.4. Schematic diagram of the  New York University exposure
chamber. (Source: Adapted from  Drew and Laskin 1973. Used with per-
mission of  Academic Press.)

-------
                                 178

exposure of large  numbers of test animals  (MacEwen 1978).  The doors
were connected to a clean corridor and  a "dirty" corridor to control the
spread of the test  materials via the movement of animal racks  in and out
of the  chamber.  The  current  state of  large  chamber design is  "one
chamber-one room-one project," which achieves the performance of indivi-
dual experiments in an integrated framework (MacEwen 1978).
   Smaller-scale exposure chambers offer more variations in design than
the larger units because they are  used  for  single exposures or short test
durations. One  of the first and simplest  small-scale designs (Figure 4.5) is
a glass bell jar mounted and sealed to a wooden  or metal base with  ports
for air inlet and exhaust (Fraser et al. 1959). These can be operated in a
dynamic or static mode and are usually enclosed in a fume hood for added
safety. Similar to  this approach,  Laskin and Drew  (1970)  designed a
Lucite cylinder with injection molded domes at each end, supported on a
wooden frame.  The top dome can be removed for access and is sealed with
rubber O-rings. This system is relatively inexpensive and its size (36 cm in
diameter and 60 cm  in length)  makes it ideal for acute or pilot studies.
Another approach to small-scale units  is to scale down  a  larger design.
Fraser et al. (1959)  describe a dynamic, cubical unit with two stainless
steel walls, one Plexiglass® wall, a  removable Plexiglass®  door, and a  max-
imum size  of 25 ft3 (7.6 m3). Drew and Laskin (1973) described another
scaled-down unit,  called a California Hood, that utilizes the  New  York
University  chamber design mounted within a square exhaust chamber.
Besides these small-scale exposure chambers, a unique chamber  design of a
converted,  cylindrical  human respirator with a  horizontal  layout and a
once-through airflow from end  to end has  been  described for short-term
exposure of monkeys (Thiede et  al. 1974). An important factor that should
be incorporated into any of  the small chamber designs  is adaptability of
the components (Fraser et al. 1959). Because the purpose of these designs
is to perform acute, pilot, or other limited studies, component adaptability
will  allow  the  unit  to  be  used  repeatedly  for different purposes or test
chemicals and consequently will reduce cost.

4.3.3.2 Chamber size

   As discussed  in  the previous section, exposure  chambers  occur in a
wide  range of sizes. The most important  consideration for  size  is  the
volume  occupied by the animals in relation to the total  chamber volume.
The animal volume should never be greater than 5%  of the total volume to
prevent heat and carbon dioxide buildup and to avoid  interfering with air
circulation  (Silver 1946, Fraser  et  al. 1959). In this regard, it  is better to
use a chamber that is too large rather  than one that may be too  small
(MacFarland 1976). The size range of large-scale chambers that maintain
good air distribution  (especially for dusts  and  aerosols) is   1 to  8  m3
(MacFarland 1976). Chambers of  a size outside of this range can be used
if alternative methods for mixing and distributing the  test atmosphere are
used or if a static system is employed. For example, Schreck et al. (1981)

-------
                                   179
                   Mixing flask
 Vapour generator
      rotameter
   Vapour
  generator
  Door -
   Chamber exhaust outlet
                  Compressed air inlet
              ORNL-DWG 85-10744

Diluting air rotameter


          - 7 gallon battery jar
                      Wooden
                      clamping bar

                       Wing nut
                                                          Battery jar
                                                          support
             Wooden frame
   Figure 4.5. Small-scale  exposure chamber based on  a bell jar design.
(Source: Adapted from Leach 1963, as reported in Drew and Laskin 1973.
Used with permission of Academic Press.)
describe a large Rochester-type chamber (~12.6 m3) that uses an exhaust
manifold to aid in maintaining  an acceptable air  distribution. The large
system was designed to reduce maintenance time and allow large numbers
of test animals to be chronically exposed. However, it is  generally recom-
mended that the chamber size be within the l-to-8-ft3 range and that more
chambers be used if space is insufficient for the test purposes.

4.3.3.3 Chamber materials and construction

   Ideally, materials used in constructing the inhalation exposure chamber
should be determined by the nature of the test agent (Eraser et al.  1959).
However, most chambers will be used repeatedly for tests of different sub-
stances  and they  are  generally  constructed of materials  that are  highly
resistant to a wide range of substances. The interior of the chamber  should
be simple in design and have smooth finished surfaces with no sharp  edges.
   The walls are usually constructed of, or lined with, a smooth nonabsor-
bent material  (Fraser et al. 1959). Common wall materials include  alumi-
num, stainless steel, safety glass, and plastic. If less resistant materials are
used (or if the test agent reacts with one of the above materials), the walls
can be lined with  glass, enameled, or lacquered (Silver 1946). Galvanized
iron  riveted together has been suggested for the  pyramid funnels, depend-
ing on the test substance (Timbrell et al.  1970). Stainless steel is the most

-------
                                  180

commonly used metal, but it is susceptible to crevice corrosion after weld-
ing. If welding is used, Type 304 stainless steel is preferred due to its resis-
tant properties (Hinners et al. 1968). In large chambers,  it may be neces-
sary to reinforce the walls with 3/4-in. pipe to provide support (Hinners et
al. 1968).
   The  door and observation ports are made of plate glass or Plexiglass-
type plastics (Fraser et al. 1959). The door should permit easy access and
be as large  as practically possible.  If a  large glass door  is used, stainless
steel framing and cross members may be necessary for support.  The door
can be of a sliding design (either vertically or horizontally) or hinged with
full-length piano hinges  (Fraser et  al. 1959, Hinners et al. 1966). Toggle
clamps,  construction clamps, or magnetic devices have been used to main-
tain the tight seal for the long periods that is necessary for subchronic and
chronic  inhalation studies (Hinners et al.  1968, Timbrell et al.  1970). A
stainless steel angle  around the  interior  edge of the door will prevent the
test animals  from clawing  or chewing  the  door gaskets  (Hinners  et al.
1968). Seals for the door and removable ports  can be made  of  neoprene,
PVC plastic, or rubber but must be resistant to the test substance. Reversi-
ble neoprene gloves attached to the door have been used in some studies to
allow handling of the test  animals in  the chamber without opening the
door  (Timbrell  et  al.  1970).  At  least one observation  port  should be
included, either at the back of the  chamber or in the  front of the top
pyramid.
   In addition to observation  ports, access is  needed for sampling  and
monitoring of the test atmosphere.  These sampling ports  may be specially
designed inserts or merely holes cut into the observation port or door; they
should  be  sealed  with rubber or  PVC  stoppers.  Hinners et  al.  (1968)
recommend  specially designed ports located at the back of the chamber at
various  heights. These  are  1.3-cm diameter stainless steel full  couplings
with PVC plugs through which  steel sampling  probes or  other monitoring
equipment can  be inserted into the test atmosphere. Further discussion on
sampling equipment  is presented in  Section 4.3.5.
   An  important  part  of the  chamber design is the construction  and
arrangement of animal cages within the  chamber. In general the following
conditions for cages  should be followed whenever possible: (1) for exposure
to aerosols,  animals should be housed individually  in  separate  cages;
(2) larger  animals should be placed in the bottom tier  of cages; (3) no
pans or solid surfaces  should be used in cage designs; (4) a layer of open
space should separate each cage layer;  and (5) cage positions  should be
rotated  to maintain  equal exposures  for all animals (Fraser  et  al. 1959).
The individual  cages should be made of a mesh material, most frequently
stainless steel. Some recommended sizes and materials for the cages are:
46 X 46 X 46 cm of stainless steel mesh for dogs and monkeys, 23 X 23
X 38 cm of 1.3-cm  mesh hardware cloth for rabbits, and 7.5  X 7.5 X 15
cm of 0.6-cm hardware cloth for small rodents (Fraser et al.  1959). If the
test  animals are to be  continuously exposed,  larger cage sizes may be

-------
                                  181
necessary, such as those recommended by the U.S. Department of Health,
Education, and Welfare (1974). The small rodent cages can be combined
into a single tray of 20 or so components to  aid in handling (Fraser et al.
1959). If the chamber door is large enough,  the cages can be mounted on
a cart, which can then be rolled in  and out  of the chamber as needed. If
this is not possible, cage  supports should be added to the chamber walls.
These can be metal pins welded into the corners at six levels 25 cm apart
or metal tracts  mounted on the chamber sides (Hinners et al. 1966). Often
metal grates or mesh partitions are used to support the smaller rodent
cages.  An  interesting modification  of  these standard cage  arrangements
has been reported by Wells (1979) and Carney (1979). This consists of a
"ferris wheel" type arrangement (Figure 4.6) that allows the cages to be
rotated  automatically during exposure.  Such  a cage arrangement can help
stabilize and equalize exposure, especially for dust exposures, by eliminat-
ing shadow effects from  the cages.  It can  also  be  used to increase the
number of animals per chamber (Carney 1979). A variation of this ferris
wheel design was discussed  by Doe and  Tinston (1981);  air enters the
chamber at the top front, flows in a cyclical pattern  concurrent with the
rotation of the cage  wheel,  and  is  exhausted at the  top rear  of the
chamber.  Moss  and Brown  at Battelle  Pacific  Northwest Laboratory
designed a chamber modification with another alternative cage arrange-
ment  (Beethe et al. 1979). This design provides constant  concentrations
and size distributions of  aerosols within the chamber as well  as  reduced
handling time for the animals. The  cages are constructed in compartmen-
tal units that occupy one-half of the cross  section of the  chamber. The
units are arranged at six different heights and are designed to operate with
waste catch pans in place (Figure 4.7). Air enters the top of the chamber,
flows  down the sides, and is diverted into distributing eddies by the catch
pans.  However, verifying studies showed that lung deposition in rats varied
somewhat  for different locations in  the chamber. Griffis et al.  (1981b)
tested the  chamber  design and found  that  animals on tiers 2, 4, and 6
(Figure 4.7) had lung burdens  8 to 11% greater than animals on the other
tiers.  Also, animals on the lower tiers  had burdens 5 to 6% greater than
animals on the upper tiers.  Griffis et al. recommended that animals be
rotated  among the tiers for chronic tests.
   Other   design  considerations  include  heating  and  cooling   of  the
chamber,  which is usually accomplished through the air  supply (see Sec-
tion  4.4).  Light  can be  provided  by  indirect  sunlight or preferably  by
artificial illumination. If artificial illumination is used, light fixtures should
be placed outside the  chamber to  avoid the necessity of specially sealed
fixtures (Fraser et al. 1959). Equipment and techniques for maintaining
the sanitary conditions in the chamber are discussed in Section 4.3.6.

-------
                                 182
                                                     ORNL-DWG 85-10745
   Figure 4.6. Whole-body dynamic exposure chamber  of 1.3-m side. The
animals are accommodated in 16 mesh cages that are supported on the
ferris wheel. (Source: Adapted from Wells 1979. Used with permission of
Blackwell Scientific Publishers, Ltd.)
4.3.4 Airflow Systems

    The airflow  system of an inhalation exposure chamber should include
the following components:  a source of air, several banks of filters (for both
supply and exhaust lines), air conditioning  units,  a system of ducts, lines,
or other containers for distribution, and devices to control and measure the
flux and/or  amount of  air,  test  substance,  and final mixtures (Lodge
1968). There  are  many ways of incorporating these features into the  air-
flow system. Use of common air supply and exhaust lines  is recommended
to minimize costs  (Fraser et al.  1959). The system  should be designed with
flexibility of use in mind, since individual components of the system may
have to be modified or moved to allow testing  of varied  substances. A
schematic diagram of a "typical" airflow system for dynamic, whole-body
exposure is presented in Figure 4.8.
    Control of the airflow within the system relies on the proper position-
ing  of values, meters, and pumps  to determine  its rate, distribution,  and

-------
                                 183
                                          ORNL-DWG 85-10746
   Figure 4.7. Cage arrangement and chamber design of Moss and Brown
showing airflow pattern. (Source: Adapted from Beethe et al.  1979. Used
with permission of the publisher.)

-------
                                 184
                                                 ORNL-DWG  81-H342
                                           AIR  INTAKE
         ^        LJ                   LJ
    CONDITIONING  -	  COMPRESSOR	  FILTERS
                   n                   n
                                   COMMON  SUPPLY LINE
VENTURI
 FLOW
 METER
INJECTION PORT FOR
 TEST SUBSTANCE
                                   COMMON EXHAUST LINE
                                EXHAUST
                                I PUMP |


                                FILTERS
   ANIMAL
  EXPOSURE
  CHAMBER —
                             AIR
                          RELEASE
                           WASTE DRAIN
   Figure 4.8.  Schematic diagram  of airflow system. (Source:  Adapted
from Wright 1957 and Wehner et al. 1972.)
operational pressure. An integrated airflow control system was designed by
Koizumi and Ikeda (1981) that automatically controlled exposure concen-
trations, flow  rates, and  sampling for vapor  atmospheres. Their system
consisted of four  exposure  chambers with  vapor generators, a  tank  of
methane gas for calibration, a gas liquid chromatograph,  a computer with
a printer, and a flow controller.  Depending on the computer program, a
variety  of atmosphere types can be generated, including continuous  expo-
sure at  a given and  constant concentration; mixed vapors exposure;  expo-
sures with  commanded geometric means  and standard  deviations;  wave
patterns; and repeated short-term peaks. This variety allows the researcher
to duplicate real-life exposure situations, such as variable concentrations of
workplace atmospheres, with low variability (coefficients of  variation  of
3.7 to 6.1%) and with greater reproducibility.
    The  air pumped into the exposure chamber is generally taken from the
room atmosphere  (Hinners et al. 1968)  or from  a  central pressurized
laboratory  supply  line  (Fraser  et al.  1959,  Timbrell et al.  1970).  These

-------
                                  185

sources provide "pure air," which  can be operationally defined as air free
of anything known to interfere with the study (Lodge  1968). Of course, to
do this, the internal supplies  of air must process air from the environment
through filtration, compression, and conditioning units (Figure 4.8.).
   The rate that the air is delivered  through the  system is limited at the
lower end by the amount of air needed to overcome heat buildup generated
by the animals and on the upper end by the atmosphere generation or feed
rate  (Fraser et al. 1959). For most exposure situations, a rate for small
chambers of 2 to 10  ft3/min  (0.6 to 3  m3/min) or for  large chambers of 2
to 20 ft3/min  (0.6 to 6 m3/min) is recommended (Roe 1968). A venturi-
type flow meter, placed in the supply duct  and hooked to  a magnehelic
pressure  gauge, is used to monitor the rate  (Fraser et al. 1959). Various
types of  valves have  been used to control the rate, including sliding gates
mounted  on, or geared cones inserted into, orifice  plates in the supply and
exhaust lines (Fraser et al.  1959); a butterfly damper in the supply line
and  a globe valve in  the exhaust  line (Hinners et al. 1966); and flexible
pinch valves in the exhaust line (Fraser et al. 1959). The valves should be
placed downstream from filter units to prevent clogging (Drew and Laskin
1973). It is necessary to verify the accuracy of the construction  and the
calibration of valves and meters  after  actual  placement in  the  system
(NAS 1977). Alarms can also be  installed in the  airflow system to signal
when valves, pumps, or other apparatuses  have  malfunctioned and dis-
rupted the airflow system (Schreck et al.  1981).
   The   distribution   of  the  air   (and test  substance) in  the  exposure
chamber  must avoid a direct flow  from  inlet to outlet and areas  of stag-
nant air  (Fraser et al. 1959). Criteria for an acceptable chamber exposure
were  proposed  by  Carpenter  and   Beethe  (1978)  and  include:  (1)
nonrelational uniform flow from inlet to outlet, (2) no  aerosol recirculation
from  below to above  animal  levels, and  (3)  uniform aerosol size distribu-
tion  and  concentration  distribution. The  pyramidal top provides  an  area
for mixing of the atmosphere and works quite well when the air enters
tangentially to the short cylinder at the top of the  pyramid (Hinners et al.
1968). The  test  substance  is injected similarly  (Figure  4.8). However,
there is a tendency for the inlet air to  flow down the sides of the chamber
instead of uniformly.  A venturi-type mixing section attached to the top of
the pyramid has been suggested as a remedy (Fraser et al. 1959). The air
is  usually exhausted  from the bottom  center of the chamber through  one
arm  of a Y joint (Roe 1968). It is just as important to maintain  uniform
flow for air leaving the chamber as for air entering. A distribution plenum
inserted below the exposure  zone  of the chamber  has been  recommended
to provide this uniform exhaust flow  (Fraser et al. 1959, Wehner et  al.
1972). Another exhaust  alternative producing uniform exhaust flow uses
several exhaust ports spread over the bottom of  the  chamber  leading to
one central  exhaust  (Carpenter  and  Beethe  1978). Modifications of the
standard  chambers often include  airflow  alterations. For example, Holm-
berg and  associates modified  the Rochester chamber to handle dense diesel

-------
                                 186

aerosols by adding a dispersing cone and screens to produce  a more lam-
inar  airflow  (Holmberg and  Moneyhun  1980, Holmberg et al.  1981).
Another variation in airflow design involved a switch to horizontal,  nearly
laminar airflow  (Hemenway and  MacAskill  1982). This  design used a
standard rectangular tank with a unique inlet and outlet baffle/plenum to
create an evenly  distributed flow through the chamber. A prechamber mix-
ing tube allowed the test substance to be blended at the proper concentra-
tion  prior to injection  into the chamber. Fecal  pans set below the cages
prevented the flow  from being totally laminar,  but particle  size samples
indicated no variation in size distribution. The horizontal laminar flow was
also  used in a chamber design by Ferin (1978). This design  used four
modules, with one  module  for filters, the  second  for  injecting  the  test
atmosphere (using perforated tubes), the third containing the animal cages
and  a perforated plate (diffuser) to convert the  airflow from turbulent to
laminar, and the fourth containing  an exhaust manifold.
   Among the various designs mentioned, one study was found that com-
pared the efficiency of their airflow  systems. This study used a model to
calculate from test  data the amount of dead space and turbulence  in the
design  (Hemenway et  al. 1982).  The  model was derived from  the mass
balance  of   the  system  and   incorporated  a  dispersion  coefficient  as
representative of the amount of turbulence.  The results  showed that  a con-
ventional design had 84.5 to 89.5% dead space  and a  negative dispersion
coefficient (often the tracer profile appeared at the outlet  point before it
showed  at  the inlet point). The  horizontal flow design had dead space
areas of  34  to 37% and a dispersion coefficient of 0.14 m2/s. The  multi-
tiered chamber (Figure 4.7) showed  almost no dead space  (0 to  6%)  and
had  a dispersion  coefficient of 0.12 m2/s.  These tests  were done without
the presence of animal  cages.  Dead space increases time to concentration
equilibration  at  start-up  and  shutdown,  decreasing  the  utility  for
pharmacokinetic studies.
   The operational pressure of the exposure chamber is controlled by the
valves  and meters  placed in  the  airflow system. It is generally recom-
mended that the chamber be  operated at a negative pressure (relative to
the external atmosphere) of 0.25 to 1.3 cm of water (Hinners et al. 1966,
Roe  1968,  USEPA 1982). This pressure is maintained by exhausting air
through a pump in the exhaust line at a faster rate than the rate of incom-
ing air (Fraser et al. 1959).  A pressure of 15 to 28 cm of water at an air-
flow rate of 1000 to 2000 ft3/min (305 to 610 m3/min) in the exhaust line
will  provide  the  necessary suction for a negatively pressurized  chamber
(Roe 1968). As  a  result of the  negative  chamber pressure, precautions
must be built into the  valve control system to prevent asphyxiation  of the
test animals if the air supply in interrupted (Wright 1957).
   Conditioning of the  air in the exposure  chamber is  a primary function
of the  airflow system.  This includes maintaining proper temperature and
humidity levels and removing any  impurities by  filtration. Temperature of

-------
                                  187

air flowing through the exposure chamber is affected by temperature of air
from the external source, temperature  of air  inside the exposure building,
and heat released by the animals. Bernstein and Drew (1980) have shown
that, when intake air is at room temperature,  the majority (95%) of excess
heat is  radiated through  the  chamber walls. Cooling of the air stream
allowed more heat  to  be  removed.  With  low  animal  loading  of the
chamber (eight rats in a 380-L chamber)  40  to 80% of the  excess heat is
removed by the cooled airstream, but as more animals are used, only  17 to
30% of the animal heat is removed by the airflow. Thus, if high animal
loading is necessary and high flow rates cannot be  used, then the exposure
room temperature should  be lowered to provide the necessary cooling to
maintain proper chamber temperatures. By maintaining the exposure  room
and chambers in the same temperature range, heat transfer  between  these
two  areas, and radiant heat  exchange between  the  chamber walls and
nearby animals are avoided (Fraser et al.  1959). The  recommended tem-
perature range for exposure chambers  and rooms is based upon preferred
temperatures  of the  most frequently used  test species and is 75  to  82° F
(24 to 28°C)  (Fraser et  al. 1959,  Hinners et al. 1966).  The proper tem-
perature is even more important in inhalation studies because it can affect
respiration rate and pattern and thereby the extent of exposure. Fraser et
al. (1959) discussed the calculations necessary to select the proper cooling
unit and recommended a  1/2-ton refrigeration unit for chambers of 145
ft3/min (44 m3/min). If extremely  high flow rates (>50 ft3/min  or >15
m3/min) are  used, a 3/4-ton  unit is suggested. If the local climate does
not require such a large cooling capacity, then copper cooling coils can be
wrapped around the  incoming  air ducts to  provide  the  proper temperature
(Hinners et al. 1966). In  either case the thermostat should  be located in
the exhaust line.
   Similarly  to the detrimental  effects on test animals of high tempera-
ture, excessive humidity can affect inhalation exposure and,  depending on
the test substance, test  agent  concentration.  It is  necessary to know the
heat load and the moisture generation rate  of the animals  in order to
determine the proper controls for humidity. Fraser et al. (1959) discuss the
necessary  calculations to determine these  factors.  Frequently, the cooling
process will reduce the humidity to the desired levels; however, it may also
be necessary to cool the incoming air to below the desired temperature and
then heat it to remove excess humidity and provide the correct tempera-
ture level. The desired humidity levels are generally accepted as  equal to
or lower than 55% relative humidity (Fraser  et al.  1959). EPA guidelines
require a humidity of 40 to 60% (USEPA 1982), but exceptions can occur
if the test agent makes this difficult.
   Another element  of chamber atmospheres that can be a confounding
factor is the presence of ammonia. This ammonia  comes from the expired
air of the test animals (Barrow  and Steinhagen 1980) or arises from the
action of bacteria on urine  and feces  of test  animals. Elevated ammonia
concentrations can react with test chemicals (Barrow 1978)  and affect the

-------
                                 188

health  of test animals (Weissman et al. 1980). Several  studies evaluated
the potential impact of ammonia  in whole-body chambers  and nose-only
exposure systems. As  expected, ammonia concentrations  increased with
time. Several mechanisms could be used to acceptably control the levels,
including use of bacteriostatic cageboard (impregnated with neomycin); an
increase in  airflow rates; frequent changing of  cageboard; and frequent
removal of animal wastes (Barrow 1978, Weissman et al. 1980). A recom-
mended procedure would  be to  use  an  airflow  rate  equivalent  to 14
chamber volumes per hour, use bacteriostatic cageboard, change the cage-
board  daily, and clean the chambers and  cages  weekly; this  should keep
levels below 2.7 ppm for mice (Weissman et al.  1980).
   To  remove potentially interferring impurities and achieve the "pure air"
necessary  for inhalation studies,  it is  widely  recommended  that several
banks  of filters be employed in both supply and  exhaust stages (Fraser et
al.  1959, Hinners et al. 1968, Roe  1968).  The most common impurities
are water vapor and oil droplets (Wright 1957). However, potentially more
harmful substances can be found in most air supply sources, and filtration
should be  designed  to handle a  wide range of  pollutants. Lodge (1968)
recommends the use of several filters  of  increasing efficiency,  including
charcoal beds to remove organics and polar  gases, a furnace  for combus-
tion of trace organics, and a final filter to remove particulates from the
furnace.  Drew  and   Laskin (1973)  recommend  filtration  through an
absolute (HEPA) filter and charcoal beds. Most recommendations require
at least two different filters for incoming air, with the first filter to handle
large materials (Hinners et al. 1968, NAS 1977). The  air with traces of
test agent  leaving the  exposure chamber must  also be thoroughly filtered,
preferably  at both the  individual chamber  exhaust and  the centralized
exhaust line (Roe 1968, Drew and Laskin 1973). Timbrell et al. (1970)
recommend the use  of two primary filters followed by one absolute filter
for each chamber exhaust. Drew and Laskin (1973) recommend filtration
of chamber exhaust  through a roughing filter (to remove animal hair and
debris), an absolute filter, and a  charcoal filter. They also recommend  a
similar filter composition for the central exhaust. In special circumstances
it may  be  necessary to  use  scrubbers,  electrostatic precipitators, or
cyclones (Drew and  Laskin 1973). As an added precaution, Timbrell et al.
(1970) recommend high release stacks  to help dilute the released exhaust
air in the environment.

4.3.5 Sampling and Monitoring of Chamber Conditions

    During  operation of the  exposure chamber,  it  is necessary to  sample
the atmosphere inside the  chamber to verify that correct  exposure condi-
tions are  being maintained.  Environmental factors that should be  moni-
tored include temperature, humidity, pressure, and  certain gas concentra-
tions.  In addition, the test agent distribution  and  concentration and the
airflow rates must be monitored (USEPA 1982).

-------
                                  189
   The temperature of the chamber can be monitored by mounting a ther-
mometer inside the chamber or by installing, in the chamber wall,  thermo-
couples that are  connected  to dial  thermometers  outside the chamber
(Fraser et al.  1959, Timbrell et al. 1970). The latter method is preferred
when large chambers are used with  samples taken from various locations
in the air system and exposure chamber (Figure 4.9.). This sampling pat-
tern  can also  be used to determine humidity and to verify adequate air
mixing and distribution within  the chamber.  Humidity determinations can
be obtained by  remote probes connected to a hygrometer or by a  hair
hygrometer placed inside the chamber (Timbrell et al. 1970). Nelson and
Taylor (1980) have described an integrated, automatic system that moni-
tors  and controls  chamber temperature, humidity, and airflow rate.  The
system uses a commercially  built temperature-humidity indicator with  a
specially designed  airflow-temperature-humidity control module. The mon-
itors provide feedback control for the system and, once calibrated,  the sys-
tem  provides flow rates with variations of 0.2%, temperatures with varia-
tions of  0.1°C, and  humidities  with variations  of  0.5%. This  type  of
automated system requires less operator time and produces excellent  con-
trol.  Correct operational pressures can be monitored by magnehelic gauges
connected to the chamber,  exhaust line, and supply line (Drew and Laskin
1973). The concentrations  of certain gases  (e.g., oxygen, carbon  dioxide)
may need to be monitored occasionally; standard chemical analyses have
proved satisfactory for this monitoring (Fraser et  al. 1959). For all these
monitoring techniques, the  accuracy of the  procedure and equipment
should be  periodically  verified using reference standards  (Fraser et al.
1959).
   The main monitoring effort is normally  focused on test agent  concen-
tration and distribution, using different techniques for gases and aerosols.
In general, to determine concentrations  accurately, you  must sample the
actual concentrations in the chamber and  at the breathing  zone of the
animals (Figure 4.9.)  and not rely on theoretical concentrations calculated
from rate of airflow and test substance feed rate (Fraser et al.  1959). Even
if concentrations are  measured accurately,  there will be variation in the
dose calculations as a result of variations in the precision of the analytical
method.  Also, with most monitoring  techniques, there is a  delay  between
sampling and analysis  that can limit  its applicability (Fraser et al. 1959).
In systems that  use a blower-powered chamber exhaust,  a potential prob-
lem is dilution of the test atmosphere by withdrawing air for the sample. If
a blower is used to withdraw the sample, the test  agent concentration can
be affected; to avoid this dilution the discharge from the sampling blower
should be  routed  into the chamber  discharge in front of  the  exhaust
blower. This will ensure that the total amount of air withdrawn from the
chamber  remains  constant  (MacFarland   1981).  Automatic sampling
techniques can  be used,  especially  where  chemical or  photoelectronic
methods  are employed. These are usually cheaper in the long run,  produce
a permanent record of exposure,  and  if rapid responses are possible can be

-------
                                  190
                                           ORNL-OWG 85-10747
                                                      w reference
                                                      corner
   Figure 4.9. Sketch of chamber indicating the eight corner and one refer-
ence sampling positions. Letters indicate  positions  of thermocouples used
for measurement  of  temperature distribution within chamber. (Source:
Adapted from Fraser et al. 1959.)
linked to a feedback method for controlling chamber feed rates (Fraser et
al.  1959). Currently, the number of automated procedures is increasing
(Hinners 1978, Van Stee and Moorman 1978, Kennedy and Trochimowicz
1982),  but those in operation  should still be verified by calibration  or by
nonautomated procedures to  ensure representative sampling and instru-
mental accuracy (Lodge 1968, Tillery et al. 1976, WHO 1978).
    Sampling techniques are comparatively simple when the test agent is a
gas or vapor. One technique  is  to obtain  a definite  volume of the test
atmosphere at a known temperature and pressure in a container (e.g., eva-
cuated flask). An  alternative  method is to pass a known volume of test
atmosphere  through  a  suitable  collective  material  (e.g., impregnated
paper). The samples obtained  can then be analyzed by standard analytical
techniques (Drew  and Laskin 1973). Direct methods for measuring gase-
ous concentrations can also be  used and adapted for  automatic analysis
(Table 4.6) depending on the  test agent. Automatic techniques for sample
analysis should be  selected based on their specificity, sensitivity, reliability,
response time, and cost (Bryan  1970). Nader (1971) reviewed the instru-
mentation aspects of these continuous monitors.

-------
                                  191

    Table 4.6. Measurement Principles Used in Continuous Monitoring
                   Instruments for Gases and Vapors

     Absorption spectrometry

       Visible and ultraviolet    O3, NO2, SO2, CHO'
       Infrared                CO, hydrocarbons

     Chemiluminescence        O3

     Electrometric

       Conductometry          SO2, acid gases
       Coulometry             O3, SO2, and other electroreducible
                                 and oxidizable gases

     lonization (flame)          Organics

     Electron capture*          Peroxyacylnitrates, halogenated
                                 organics

     aSemiautomatic sequential analysis with gas chromatography.

     Source: Adapted from Bryan (1970).
   Sampling and monitoring of aerosol and dust exposure studies are more
difficult, because both mass concentration and particle size must be deter-
mined. Some of the varied sampling methods used for aerosols are listed in
Table  4.7  with the particle size  range for which  they are appropriate
shown in Figure 4.10. Sampling methods and equipment for mass  concen-
tration and particle size determinations  are essentially of three types. One
type of sampler roughly separates  particles into two groups representative
of the  upper and  lower  respiratory tract, another  type  is  designed  to
separate the  test particles  into several  classes  of  different  sizes  or aero-
dynamic behaviors;  and  some  samplers collect or  analyze  the  particles
without any physical separation based on size.
   The first  aerosol sampling devices were designed to mimic the natural
separation  of particles that occurs between the upper and lower respiratory
tract. Generally, particles of 7  to  10 pm  diam are considered to  be non-
respirable and  represent  the particle size found in  the  upper respiratory
tract (Hatch and Gross 1964, Roe 1968).  Gravity sedimentation was one
technique used to simulate  this separation pattern. With laminar flow of
air  through a  horizontal  channel, particles will settle out depending  on
their size and settling velocities (Hatch and Gross 1964). Wright (1954)
designed a collector (Hexhlet sampler) using this principle with an elutria-
tor  composed of two banks of horizontal aluminum  plates separated by a

-------
                                             Table 4.7. Participate Sampling and Characterization Apparatus
               Apparatus
               Quantification
                                                                                                                 Remarks
 Dust jars, screens, plates, boxes


 Cascade jet impactors

 Single jet impactors

 Fine-fiber impinger
 Greensburg-Smith impinger

 Glass, asbestos, metal cellulose, and
  other esters in porous membranes
 Paper tape high-volume


 Cyclones

 Light scattering devices

Direct microscopy

Precipitators
                Sedimentation
Count, weight, chemical analysis

                Impingement
Count, chemical weight, micrurgic

Count, chemical weight, micrurgic

Count, chemical, micrurgic
Count, chemical
                  Filtration
Weight, chemical,  visual, etc.

Chemical, color, light
  transmission
                Centifugation
Weight, chemical,  count
                   Optical
Particle count concentration and size fractiona-
tion by light scattering, tyndallometry
Count, visual examination, sizing
            Electrostatic thermal
Count, micrurgic, visual
Simple, inexpensive, readily available. Accepted method for
pollen. Long settling time for submicron particles. Not isomatric

Rapid aerosol fractionation. If to count, not good for high
concentrationa. Can calibrate for BAD
Efficient to 1 pm, then less efficient. Distinguish  solid from
liquid aerosols
Simple, compact. Solid-liquid distinctions
Soluble gases and particulates. Commercially available, portable

Wide range of aerosols and applications. Inexpensive, readily
available (e.g., Millipore, Gelman)
Smoke and aerosol density
Wide range of sizes, materials, utilities. Available, inexpensive

Commerially available. Expensive. Need careful calibration
and operation. Automatic, rapid, small sample requirements
On impaction, filter, or electrostatically collected samples

Electrocharge, thermal gradient principles. Submicron particles
Source: Adapted from Campbell (1976).

-------
                                                                                                                   ORNL-DWG 81-11343


—

ULTRA
ULn


CENTF
CRAMICROSCC
PE

ELECTRON MICROSCOPE
ULTRAFILTRATION
1IFUGE
X-RAY DIFFRACTION






—

M


CENTRIFl


CROSCOPE





GAS ELUTRIATION
JGE





SEDIMENTATION & GRAVITY SETTLING
'

PERMEABILITY

TURBIDIMETRY
LIGHT SCATTERING













SIEVING
VISIBLE TO EYE


MAC

MINE TOOLS





(MICROMETER, VERNIER CALIPER, etc.)




0.0001
          0.0005 0.001
                           0.005 0.01
                                           0.05   0.1         0.5    1          5
                                                        PARTICLE SIZE (microns)
                                                                                 10
                                                                                            50  100
                                                                                                           500  1000
5000  10,000
         -*- PARTICLE SIZE LIMITS UNDER AVERAGE CONDITIONS.

         •»- STATED METHOD IS OF DOUBTFUL UTILITY IN THESE SIZE RANGES.
     Figure  4.10.  Limits of  particle size  measuring  equipment.  [Source: Sheehy  et al.  1967,  as  reported in  Campbell  1976.
 Reprinted from Clinical Toxicology 9(6):863 (1976) by courtesy of Marcel  Dekker Inc.]

-------
                                  194
vertical partition into  120 inlet ducts for the first separation stage and a
soxhlet filter to collect the "respirable" particles in the second stage (Fig-
ure 4.11). This unit will  collect about 1 g of particles when  operated at
100 L/min for 8 h (Hatch and Gross 1964).
   Another two-stage collector, the conicycle,  uses  centrifugal  force to
both reject the coarse  upper  respiratory tract particles and collect the fine
respirable particles  (Figure  4.12).  In  Figure  4.12, large particles  are
rejected at A and small  particles enter B  with deposition on  surface C.
Very small particles may  be  partially deposited or exit through orifices O.
The  sampler operates  at  a rotational speed  of 8000 rpm and a sampling
rate  of 10 L/min (Wolff and Roach 1961  as cited in Hatch and Gross
1964). Performance curves  for  the conicycle  and hexhlet samplers  are
shown in Figure 4.13.  A  two-stage continuous flow cascade impactor that
produces a separation window between  4 and  1 nm from a polydisperse
aerosol  has also been developed (Gat  1980).  Other two-stage samplers
include the cyclone separator,  which uses  centrifugal  force  to  separate
large particles and a  filter to trap  small particles,  and the preimpinger,
which uses changes in  air direction and water filled-glass flasks to separate
and  collect  the different  particle sizes (Fraser et al.  1959, Hatch  and
Gross  1964).  Generally,  there  is little difference in collection efficiency
between these methods, except for the potential for disaggregation of large
particles or  aggregations  with the  cyclone  or  preimpinger (Hatch  and
Gross  1964).  These differences  can be important, because  the  particles
must be collected in the  form occurring in  the aerosol cloud to be valid
particle size determinations.
   Aerosol samplers have also been  designed to produce multiple stages of
particle separation  based on size or aerodynamic behavior.  The cascade
impactor (Figure 4.14) was the first sampler designed with this capability
(Hatch and Gross 1964).  It consists of a series of impingement slots (usu-
ally  5  to 10)  arranged in order of decreasing width and distance followed
by a high-efficiency  filter. It works on the principle that progressively finer
particles are deposited at  each stage of impingement  as a result of increas-
ing jet velocities and decreasing jet dimensions (Hatch and Gross 1964). A
commercial model and its particle size separation series are shown in Fig-
ures 4.15 and 4.16. The  separation series compares rather well  to lung
separation  patterns with,  stages 1 and 2 corresponding to nasal  chamber
deposition, stages 3 to 5 corresponding to  upper respiratory airways  and
pulmonary air spaces  deposition, and stages 6 to 8 corresponding to lung
deposition. Thus it can be used to identify different  sites of deposition for
different size particles. The  cascade impactor does have two main  draw-
backs. First,  it tends to break up aggregations of  particles that  would
deposit as  large particles, and,  second, only a limited amount of material
can  be safely collected on each surface before the  risk of dislodgement
becomes a factor (Hatch and Gross  1964).
   A variation of this sampler  is the Electronic Cascade Impactor (ECI).
It was designed by Tropp et  al. (1980) to provide the following desired

-------
                                    195
                                                        ORNL-DWG 85-10748
                                                         Inches
   Figure  4.11.  Hexhlet two-stage  aerosol  sampler. A  = First stage of
gravity elutriator with size  separation based  on the equation P = 1 — d2/50,
where P  =  percentage penetrating and d =  diameter of unit  density
sphere in microns; B = tapered critical orifice; C = ejector to maintain a
constant airflow rate;  and D  = soxhlet  filter. (Source: Adapted  from
Wright 1954. Used with permission of the publisher.)
                                                        ORNL-DWG 85-10749
                     1 cm
   Figure  4.12.  Conicycle  aerosol  sampler. A  =  Inlet  for atmosphere;
B  =  inner rotational space; C =  deposition surface (removable) for smaller
particles; O  = air outlet; Z = rotational  axis; and / = radial distance
(only  particles coarse enough to travel this distance in the time required for
the air to pass through the chamber will be deposited). (Source: Adapted
from  Wolff and Roach 1961.)

-------
                                  196
                                  ORNL-DWG 85-10750
                     o
                     CL
                        02468
                        EQUIVALENT PARTICLE  DIA, MICRONS
   Figure 4.13. Performance curves for the hexhlet and conicycle samplers.
1  =  Experimental pulmonary deposition curve  of  Brown et al.;  2 =
Findersen's  theoretical  curve for human  alveolar deposition; 3  =  hexhlet
calculated performance curve;  and 4  =  conicycle calculated performance
curve. (Source: Adapted from Wolff and Roach 1961.)
characteristics: (1) a response time of one second or less for the entire size
spectrum;  (2)  capability  to  be  calibrated absolutely;  (3) capability  to
measure very dense (without a dilution step) as well as very  dilute aero-
sols; (4) ability to classify (in time sequence, if desired) aerosol particles
for subsequent chemical or physical analyses; (5) a large volumetric sam-
pling rate to minimize errors introduced when sampling  a spatially hetero-
geneous  aerosol; (6) measurement of the aerodynamic particle  diameter
rather than  such  quantities  as  optical diameter  and electrical mobility
diameter; (7) low cost in comparison  with available instruments; (8) real-
time output of data that is easily interfaced to recording or data acquisi-
tion equipment. Generally, the ECI operates as does a standard cascade
impactor, except that  the aerosol particles are given a  unipolar  electrical
charge as they enter, each collection stage is isolated electrically from the
other stages, and each stage  is connected to an  electrometer detector. As
particles strike  the collection surfaces, they give rise to a  current  that is
used  to  measure  and  record the particles.  Because  each collection stage
captures only a certain  size range of particles,  both the number and size
distribution can be determined. Other automatic analysis techniques have
also been combined with cascade impactors including piezoelectric  sensors
and sensors that detect  decreases in beta  radiation resulting from particle
deposition  (Smith et al.  1978).

-------
                                 197
                                           ORNL-DWG 85-10751
                                     -• SMALLER PARTICLES
                                           1  \  PATH OF
                                              \  SMALL PARTICLE
   Figure 4.14. Schematic diagram showing operation of cascade impactor.
(Source: Adapted from Smith et al. 1978).

-------
                                 198
                                    ORNL-DWG 85-10752
                                      JET STAGE (9 TOTAL!
    Figure  4.15.  Schematic  of  Anderson  Mark  in  cascade impactor.
(Source: Adapted from Smith et al. 1978.)
   The collection performance of two cascade impactors and a new instru-
ment,  the  inertial  spectrometer, are compared  by De Zaiacomo et al.
(1983). The inertial spectrometer was designed by Prodi et  al. (1979) and
works  by injecting  the aerosol into a clean air stream  above a 90° bend.
The bend separates the particles by inertia, and they leave the original air
stream at distances as a  function of their aerodynamic  size (Figure 4.17).
The particles then  impact  on a  filter that is the outer edge of the bend.
This design allows the particles to be separated while they are airborne, in
a continuous distribution,  and avoids  the problems of bouncing  and re-
entrainment (De Zaiacomo et al. 1983). The two cascade impactors were a
multislot design and  a multiorifice design with both having eight stages
and a  backup filter. The comparison found that  for smaller size particles
the three designs performed equally well. For larger aerosols, the multislot
impactor overestimates the median  size and geometric  standard deviation
in comparison with the other two designs.
   Another multistage sampler design is the conifuge developed by Sawyer
and Walton (1950). It uses gravitational force (similar to  the elutriator)
with  an increase in force supplied by centrifugation. Air  is introduced at

-------
                                   199
                                                       ORNL-DWG 85-10753
   100
      .3   .4  .5 .6 .7 .8.91.0         2      3    4   5  6  7  8 9 10

                         PARTICLE DIAMETER, micrometers


   Figure 4.16. Calibration of an Anderson Mark III impactor. Collection
efficiency is  plotted vs particle  size  for stages  1  through 8.  (Source:
Adapted from Smith et al. 1978.)
the apex of a double-walled, hollow, rotating cone, and the particles are
deposited  on microscopic slides on the outer wall in  decreasing order of
size from  top to bottom (Roe 1968). This  design produces excellent size
separation without breaking up large or aggregated particles but is limited
by a slow sampling rate (Hatch and Gross 1964).
   Samplers have also been designed to collect particles without any dif-
ferential separation. Simple  filters and membrane filters are occasionally
operated without separators. Precipitator  devices  can  be fitted with some
type of separator but are usually operated without them. Optical detection
devices are designed to operate without the need for a separator.
   Precipitators have been designed that collect particles by thermal or
electrostatic methods. The thermal precipitator consists  of a narrow chan-
nel with a heated  wire or surface opposed by a cool collection plate or
slide. The particles are struck by molecules  with a higher velocity from the
heated  side and are  forced toward the cool deposition  surface (Tillery et

-------
                                 200
                                         ORNL-DWG 85-10754
    Figure  4.17.  Working principle of the inertial spectrometer. (Source:
Adapted from DeZaiacomo et al.  1983. Used with permission of the pub-
lisher.)
al.  1976). Electrostatic precipitators are usually operated on a point to
plane discharge design, with a corona discharge from a needle point charg-
ing particles for deposition on a conductive carbon-coated surface (Tillery
et al.  1976).  An electrostatic precipitator that uses a radioactive tritium
source to provide ions for charging particles is shown in Figure 4.18. Pre-
cipitators generally operate at low flow rates of 5 to  10 cm3/min and may
occasionally show some size separation.
   Optical  devices  may  be  used to analyze  particle  size, mass,  and
numbers. In general, they operate  by using photomultiplier tubes to  detect
absorbtion or refraction of projected  light by  particles (Drew and Laskin
1973). The photomultiplier produces a pulse proportional to the size of the
particle.  One  optical monitor  uses a  laser Doppler  velocimeter (LDV) to
measure  particulate velocity in an acoustic field (Mazumder and Kirsch
1977). A particle  entering the sensing field of the LDV  is excited  by an
acoustic  wave, and the particle oscillations are  monitored. By measuring
the phase lag between  the oscillation signal and  the acoustic wave  signal
(measured by  a microphone)  the aerodynamic  diameter  and number of
particles  can be determined. A prototype instrument using this technique,
the single particle  aerodynamic  relaxation  time  (SPART) analyzer, has

-------
                                  201
                                            ORNL-DWG 85-10755
                   0-RING
           DIRECTION
           Of FLOW
        _» DIRECTION
          OF FLOW
    MEMBRANE FILTER

ELECTRON
MICROSCOPE
GRID
                                     TOGROUNO
   Figure 4.18.  Electrostatic precipitator using radioactive tritium (H3) as
an ion source. (Source: Adapted from Tillery  et al.  1976. Used with per-
mission of the publisher.)
been developed (Mazumder et al.  1979).  It can analyze particulate sam-
ples in the size range of 0.2 to 10 MHI at a rate of 200 particles per second.
Another laser (He:Ne) device for  continuously monitoring the mass con-
centration of monodisperse aerosols was reported by Carlon et al. (1980).
Optical devices are most useful when nondestructive or noninvasive tech-
niques for monitoring particle size and number are necessary. However,
caution must  be  used with these devices, because  light scattering is  not
simply a function of size (Tillery et al. 1976); many variables can  affect
the results (Drew and Laskin  1973). The use of optical techniques to mon-
itor particulates in industrial process streams has been reviewed by Smith
et al. (1978). Potential problems with other nonoptical aerosol  samplers
have also been reviewed.  Heindryckx and  Dams (1981) investigated some
errors associated with cascade impactors used in field situations, and some
of their information  also applies to inhalation study  sampling. Lund et al.
(1979) reviewed practical problems arising from the relationship between
dust generators and  collectors. Formignani et al. (1982) discussed calibra-
tion apparatuses for filter media of aerosol samplers.
   Except for the optical techniques just discussed, the  above  sampling
techniques must be  followed  by  analysis of the sample. If the sample is
obtained on a filter media, various routine  gravimetric or  analytic tech-
niques (e.g.,  chemical,  UV absorption,  chromatography)  can be used to
determine mass concentration. The filter media to be used is often deter-
mined by the  analytical technique employed. Microscopic techniques (both
light and electron) can  be used to determine particle number and size  and
are most frequently  used with samplers  that deposit the particles on  a
removable  surface (e.g.,  an  electron microscopic grid in an  electrostatic
precipitator).  Computerized scanning  devices can   be used to automate

-------
                                 202

analysis of microscopic samples and simplify counting and sizing of parti-
cles (Smith et al.  1978). In general, the various sampling and monitoring
methods  produce comparable results.  Halbert et al. (1981) compared par-
ticle size distribution analysis using three techniques.  They found little
difference    in    analysis     results    between     an    electrostatic
precipitation-microscopic   technique,   a  cascade   impactor-gravimetric
technique, and a laser-acoustic relaxation-time technique.

4.3.6 Maintenance of Sanitary Conditions

   It is  important to maintain sanitary conditions  in the exposure room
and chambers to avoid  potential interference from disease or illness. This
maintenance includes both design features of  the chamber and  cleaning
techniques.
   To aid collection of test  animal wastes,  the chamber should have a
pitched floor, smooth interior  walls,  and a centrally  located drain (Drew
and  Laskin  1973). If the chamber  is used  for  continuous exposures, a
spray ring is usually installed to wash down wastes.  The preferred location
is at the top of the bottom pyramid below the  metal floor grate (Hinners
et al.  1966). The ring  is a  l/2-in.-diam stainless steel pipe with four to
eight wide-angle  nozzles  and  is controlled  by  an  external  hand valve
(Hinners et  al. 1968).  At the bottom of the chamber a 3-in.-diam gate
operated by a quick release is used to drain the  wash water. This is usually
connected to a sewer line,  unless the test substance is dangerous enough to
require collection  in containers (Drew and Laskin 1973). If the chamber is
used for intermittent or acute exposures that allow removal of the animals,
a spray ring can be installed at the lower edge of the top pyramid to wash
down the sides or an external hose can be  used  (Hinners et al. 1966, Drew
and  Laskin 1973). To avoid problems with the airflow systems (especially
clogging with animal hair),  the lines  should  be free from rough surfaces,
restrictions, or sharp bends (Hinners et al.  1968).
   During cleaning, it is important to prevent cross contamination between
chambers  and to protect  the health of the technicians.  Thus only one
chamber at a time should be opened and  cleaned (Timbrell et  al.  1970).
While cleaning, the operators should wear protective  clothing and respira-
tors  (Timbrell  et al. 1970). If a hose is used to wash down the chambers,
the water pressure should  be low  to avoid creating respirable mists of the
water-chemical  mixture  (Smith and  Spurling  1974). For chambers with
spray rings, the water temperature should be  no more than 102°F (39°C),
to prevent steaming within the chamber (Hinners et  al. 1968).  Chambers
with stainless steel components must be cleaned frequently to maintain the
useful life of the chamber. Cleaning solutions should  not remain in contact
very long. A cleaning solution  of acid detergent with  wetting, penetrating,
and low-foaming agents has been used successfully for cleaning deposits on
stainless steel (Hinners et al. 1968). After all  chambers have been cleaned,
the exposure room should  also be washed  down or vacuumed (not swept)

-------
                                  203
to ensure that no contaminants or waste materials remain (Timbrell et al.
1970).
   One  aspect  of  maintaining sanitary  conditions  is the  protection of
workers  associated  with the testing procedures. In addition to  the steps
mentioned above,  efforts  are needed to  ensure that workers  are  not
accidentally exposed to remnant or leaking test atmosphere. Such exposure
could occur through loading of the vapor  generation system during start-
up, leakage during  the test procedure, and release during shut-down  pro-
cedures (Klein and  Geary  1982). Automatic sampling of the chamber and
test  room  will  assist in determining  when  hazardous or safe conditions
exist. A special  problem  occurs  when particulate substances are being
tested, especially particulate carcinogens  (Drew 1978).  In these cases,
workers can be exposed to particulates on the cages, animals, and walls of
the test  chamber. In addition to  dermal contact, these particulates  may
reenter the atmosphere and be breathed by the workers (Drew 1978).  Use
of a service module or  glove and pass boxes will  reduce this potential for
exposure. Also,  use of and monitoring for an inert indicator substance
(such as  sulfurhexafluoride) in the aerosol airstream  can indicate when
small leaks occur. This approach is especially useful when rapid monitor-
ing techniques are  not  available for the test substance (Moorman 1978).
These procedures along with a disease control program will avoid interfer-
ence with the exposure studies.

4.3.7 Specialized Equipment

   In addition to the standard equipment described in earlier sections,  spe-
cialized equipment  is often used in inhalation studies. This includes whole
chambers designed  for unique exposure conditions  as  well  as equipment
designed to test nonroutine conditions of inhalation.
   Specialized  exposure   chambers  have  been  developed  by  research
laboratories of the armed services. These include chambers that operate at
pressures higher or lower than normal atmospheric pressure, and systems
for studying biological aerosols at low temperatures  (Drew  and Laskin
1973). Outer space  and underwater exploration programs have prompted
many of these special chamber designs.
   Another special  inhalation exposure chamber design for nose- or head-
only exposure was discussed in Section 4.3.2.4.  A part of this system, the
body plethysmograph,  is also used for  measuring  functional alterations in
respiration (see Section 4.2.5.4). The  body plethysmograph  consists of an
airtight metal or plastic (Lucite® or Plexiglass®) cylinder designed to  con-
tain  the animal with its nose and head inserted in the test atmosphere and
to allow  measurement of pressure changes resulting from breathing move-
ments. For rodents, the cylinder varies in  size from 9  to 23 cm in length
and  from 3 to 12 cm in width (Swann et all965, Alarie  1966). The front
is usually a combination of straight or curved rigid plastic plates and flexi-
ble sheets of sealing material (neoprene rubber, latex, etc.)  with a  cen-
trally located hole  for  the head or nose (Thomas and Lie  1963, Murphy

-------
                                 204

and Ulrich 1964). Occasionally, a sealed face mask is used  to deliver the
test agent and provide the necessary front seal. The cylinder is connected
to three  tubing  lines, one  a transducer  pressure line to monitor  pressure
changes,  another to conduct the pressure wave from the pump, and a third
to bleed  off  excessive pressure (Murphy and  Ulrich  1964,  Swann et al.
1965). Dorato et al. (1983) discuss a modification  of the body plethysmo-
graph  that uses an  intraesophageal catheter and an analog computer to
produce real-time  calculations  of pulmonary parameters in rats. This pro-
cedure did not require surgery  or prolonged anesthesia and was capable of
providing multiple evaluations  over extended time periods (40 weeks). A
body plethysmograph for use with dogs  has  been described by Boecker et
al. (1964) and  consists of  a large rectangular Plexiglass chamber with an
internal sling for restraining the animal and  a latex face mask for delivery
of the test agent.  Although  the test animals  are  usually kept in the
plethysmograph for only a  few  hours at a time, it is best to acclimate them
prior  to the study by placing them in the container for increasing periods
(Smith and Spurling 1974). This will avoid unwanted stress.
   Specialized  equipment  has  also been designed  to be inserted  into the
chamber for  use. One such type of equipment allows evaluation of inhala-
tion  effects on  animals that are exercising. For  rodents, this consists of a
revolving cylindrical wire cage (46 cm diam and 61 cm long) that rotates
at a speed of 6.5 rpm and can hold 10 to 15 animals (Fraser et al.  1959).
Another version designed by Illing et al. (1975), consists of a wheel 15 cm
in diameter and 3.8 cm wide  that is  open on one side  and closed by the
converging wires on the other. The cages are mounted on a  steel rod run
through  the center and separated by Teflon® sheets. A small motor rotates
the cages via the rod at speeds of 1 to 6 rpm.  Mice or hamsters can be
tested individually in this  design. To  test dogs, a rubber treadmill with an
effective area of 56 cm by 1.4 m was designed with operational speeds of
0.9 to 7.5 mph  (Fraser et  al. 1959). Such equipment is not  part of a rou-
tine study, but would be useful if the expected exposure of humans would
involve activity.


4.4. GENERATION OF TEST  ATMOSPHERES

4.4.1 Introduction

   Testing of chemicals by the inhalation exposure route requires  the gen-
eration of many types  of  test  atmospheres.  The production of an atmo-
sphere of a precisely known composition is one of  the most  difficult tasks
the toxicologist  faces  because  the ratio of  agent to air can be one to
several million  (Lodge  1968).  The wide variety of potential atmospheres
precludes a universal apparatus suitable for generating atmospheres for all
inhalation tests  (Clark  1977).  Each type of atmosphere must be treated
differently, and, even within each type, chemicals with different properties

-------
                                  205
can require unique generation techniques or equipment. The desired objec-
tive is to produce the required atmosphere  with  the  most efficient tech-
nique available (Carney 1979).
   The types of atmospheres generated  can be grouped into two broad
categories, gases or vapors, and aerosols. Gases and vapors were classified
according to  their mode  of physiological action rather than their structure
by Casarett  (1975). Asphyxiants are one class of  gases  and vapors and
include both  simple  and chemical forms. An asphyxiant is any substance
that has the capacity to deprive tissues of oxygen by any means other than
impairment of respiration  mechanics (Casarett 1975). Simple asphyxiants
are usually physiologically inert gases (e.g., nitrogen)  that prevent oxygen
from  reaching the respiratory surfaces,  while chemical asphyxiants (e.g.,
carbon  monoxide) prevent the body from utilizing the oxygen (Casarett
1975). Irritants are  the other major type of gases or  vapors  and produce
inflammation in tissues with which they are in contact (Casarett 1975).
   Aerosols  include  any  system  of liquid  droplets  or solid  particles
dispersed in  air with  small particle sizes (usually <10 p.m) and  conse-
quently low settling  velocities to possess considerable stability as an aerial
suspension (Hatch and Gross  1964). Even more than gases or  vapors, aero-
sols exhibit a  high  interdependence between  physical  characteristics and
toxicity (Casarett 1975). There  are many types of aerosols;  some  of the
more  common are discussed below.
  1. Dusts are fine particles (diameter range  of angstrom to 100 /tm) of
    solid material formed by  mechanical disintegration of matter and
    dispersion into air (Hatch and  Gross 1964). They are chemically and
    constitutionally the same  as their parent  material (Casarett 1975).
  2. Fumes are clouds of particles (diameter  <0.1  /mi) formed by combus-
    tion, sublimation,  or condensation and  are usually accompanied by a
    chemical change  (Casarett  1975).  Often this term is  restricted to
    descriptions of droplets from metal oxides of zinc, magnesium, iron,
    etc. (Hatch and Gross 1964).

  3. Smoke refers to a mixture of fine liquid droplets and solid  particles
    (<0.5 /um) with liquid  and vapor phases produced by combustion
    (usually  of organic material) (Hatch and Gross  1964, Casarett 1975).

  4. Mists and fogs also occur in liquid and  vapor phases with liquid dro-
    plets of  any size range formed by condensation of vapors on suitable
    aerial nuclei  or by  uptake of water by hydroscopic particles (Hatch
    and Gross 1964, Casarett 1975).
Each  of these aerosol types and the  gases and vapors require a unique gen-
eration methodology; the following sections will discuss the equipment used
to produce such atmospheres.

-------
                                  206
4.4.2 Generation of Gas and Vapor Atmospheres
   The generation of  test atmospheres for gases  and vapors is  relatively
simple,  involving  fewer  potential complications than  the generation of
aerosol atmospheres. Gases and vapors can be tested in a static or dynamic
mode. The only requirement is that the method of introducing the contam-
inant  must be compatible with the mode of operation (batch introduction
for static systems and continuous feed for dynamic systems) (Lodge  1968).
   A commonly  used batch method for generating an atmosphere is to
mix and store the contaminant in  a high-pressure cylinder (Cotabish et al.
1961, Tillery et al. 1976). The gas or volatile liquid is introduced into the
empty cylinder  by  syringe injections  or  manometer measurements  and
diluted with air to the desired concentration before release into  the expo-
sure chamber. This technique  can  be  used with stable contaminants and, if
the release rate is slow, can also be used with a dynamic system. Another
simple technique uses  high-pressure cylinders containing only the contam-
inant. The contaminant  is metered into a  flowing stream of pure  air by
valves and a calibrated rotometer, which delivers the proper concentration
to the exposure chamber (Amdur 1957, Saltzmann 1961). This  technique
is also  limited to stable  gases  or highly  volatile liquids  during dynamic
exposures. An expansion chamber or  reservoir is often used in conjunction
with such a system (Carson et al. 1962).
   When  the  gas or vapor  is not  highly volatile  or  chemically  stable
enough to store in high-pressure cylinders, other methods exist to introduce
contaminants into a flowing air  stream. One of the earliest used techniques
for liquids with high boiling points is to inject the liquid through a motor-
ized syringe onto a  heated block or into a heated  tube which vaporizes it
(Carpenter et al.  1949, Tillery et al.  1976). The vaporized contaminant is
then introduced into the air stream by a pump or aspiration needle.
   A problem with using such direct heating systems to vaporize liquids is
the potential for charring at the heat source or thermal decomposition of
the sample (NAS 1977).  To avoid this problem, two techniques have been
recently reported. Potts  and Steiner  (1980) redesigned a counter-current
system  to use a  perforated-plate distillation column (Figure 4.19).  The
liquid contaminant is introduced at the top of the column and flows down-
ward. Air is introduced  at  the  bottom of the column  and flows upward,
becoming saturated with the vapor as it passes  over the distillation  plates.
This  system provides  a  vapor source  of constant concentration, without
thermal decomposition or condensation resulting from cooling of a  heated,
saturated  vapor.  However,  the concentration  must be measured  in the
exposure chamber because no direct measurement of the amount of liquid
vaporized is possible. Also, since the  vapor pressure of low volatile  liquids
is not accurately known  at room temperatures,  the experimenter must use
trial and  error to determine the correct parameters necessary to  achieve
the desired concentrations. A  second new technique  to avoid thermal
decomposition is  designed to handle liquids that are  neither  highly or

-------
                                   207
                                                      ORNL-DWG 85-10756
       Ambient Air
          Inlet
   Figure 4.19. Improved  counter-current vaporizing apparatus, with ther-
mostating jacket.  (Source: Adapted from Potts and Steiner  1980. Used
with permission of the publisher.)

-------
                                  208

minimally volatile (Miller et al. 1980). A J-tube design vapori/er (Figure
4.20) is filled with 1/4-in. glass beads to increase the vaporization surface.
The  liquid is introduced at the bottom of the J through a capillary tube.
Heated air flows up through the system,  vaporizing the liquid as it passes
over the glass beads. By preheating the air, chances for charring and ther-
mal  decomposition are greatly reduced. This design also provides a greater
safety margin when potentially explosive liquids are vaporized.
   Another technique for vaporizing liquids  is based on two saturation
chambers and is similar in principle to the counter-current column of Potts
and  Steiner (1980). An inert carrier gas is passed through two gas washing
bottles containing test liquid, with the first bottle  at a higher temperature
than the second (Saltzmann 1961, Drew and Laskin 1973). The use of two
bottles ensures  saturation, but a filtering  step following the chambers may
be necessary to trap unwanted liquid droplets.
   Some unique methods for dispersing vapors and gases are based on dif-
fusion through  permeable materials. In one system vapor diffuses  at a con-
stant rate through a  porous asbestos plug (Lodge 1968). Certain plastics
(e.g., Teflon®)  have also been used (O'Keeffe and Ortman 1966, Drew and
Laskin  1973).  These  systems require constant flow and  temperature. The
rate of diffusion can be controlled by  varying wall or plug thickness and
area (Drew and Laskin 1973, Tillery et al. 1976).

4.4.3 Generation of Aerosol Atmospheres

4.4.3.1 Introduction

   The generation of aerosol  atmospheres for inhalation testing involves
more variables than the generation of gas or vapor atmospheres. The divi-
sion of solids or liquids  into small particles or droplets and their dispersal
in air produces two major changes, increased surface area and total space
occupied, that  affect  the toxicity (Drew and  Laskin  1973). These changes
intensify the basic chemical and physical activity of the test agent, includ-
ing its rate of  oxidation, solubility, evaporation, and  deposition. Also, the
particle size that the parent material is divided into determines the depth
of penetration  and deposition site in  the  respiratory tract. Therefore, it is
necessary to determine the particle-size capability as well as the concentra-
tion  capability  of the  aerosol generation  equipment in order to select the
proper unit for each test. The equipment must produce  and maintain the
desired mass concentration and the desired mean effective particle size
with  an  acceptable dispersion  around  the mean (Lodge  1968).  Size fre-
quency information is  generally not useful for dosage evaluations, so most
discussions  refer to  mass  frequency  because  the  mass  quanity  (mass
median diameter) is generally a cube function of the easily measured par-
ticle diameter (Drew and Laskin 1973).
   Aerosols can be produced in two forms, monodisperse and polydisperse.
Polydisperse aerosols  include a  wide  range of particle diameters with the
mass median diameter having a large geometric standard deviation and are

-------
                                     209
                                                 ORNL-DWG 85-10757
                 1 IH HOHUSILICAH GLASS
                WITH Jb/?«, SHHtHICAL JOINTS
                                                        FILL WITH
                                                      1/4" GLASS BEADS
   Figure  4.20.  J-tube  vaporization  assembly.  (Source:  Adapted from
Miller et al.  1980. Used with permission of the publisher.)

-------
                                 210

considered the normal form for aerosols occurring in actual exposure con-
ditions (Drew and Laskin 1973, Casarett 1975). Monodisperse aerosols are
composed of particles with a narrow range of sizes  resulting  in  a mass
median diameter having such a small geometric standard deviation  that no
significant  error is introduced into  the experiment  (Wilson and  LaMer
1948, Mercer 1973).  Monodispersity of an aerosol  thus varies depending
on the experimental design or requirements and is an arbitary determina-
tion  of  the  experimenter (Mercer  1973).  Fuchs and  Sutugin  (1966)
defined it as those aerosols having a geometric standard deviation  of 1.22
or less when the particles have a log normal distribution; this definition is
reasonable for  most  applications (Mercer 1973).  The advantages of a
monodisperse aerosol relate to  simplification for  the study  of  factors
affecting toxicity and handling of data (Wilson and LaMer  1948). Mono-
disperse  aerosols are  also useful for  studying deposition and clearance pat-
terns and for calibrating the monitoring equipment (Mercer 1973).
   Many techniques  and types  of equipment for  generating aerosols,
whether  monodisperse or polydisperse, have been developed. The following
sections will  discuss the major designs for generating aerosols using solid
feed  materials   and   liquid,  solution,  or  liquid-solid  suspension  feed
materials.

4.4.3.2 Generation of  aerosols from solids

   The  earliest  generating  designs for solid  or  dust aerosols  involved
dispersion of loose noncompacted particles. Air was blown through a con-
taining device in which the loose dust particles were being kept in  suspen-
sion by vibrations, rotation of the container, or stirring. These devices pro-
duced  polydisperse aerosols  unless  the loose dusts had  been previously
sorted  and  sifted to  make a monodisperse  aerosol. Unfortunately,  these
techniques did not produce aerosol clouds with very consistent characteris-
tics, due in  part to  the  inability  to  predict the quantitative effects  of
changes  in  their methodology (Wright 1950, Roe 1968). Details of the
early (1885-1940) designs can be found in Wright (1950). Later dust aero-
sol generators of this type have been designed to  avoid many  of these
problems.
   Drew  and  Laskin  (1971) designed  a dust  generator to  use with
material  that is not  readily dispersed from the preferred prepacked plug
(e.g., polyester-fiberglass  dusts). For the fluidizing chamber, they used an
inverted  1-gallon  glass jar with a  3-in., four-blade fan mounted  in the
screw  cap.  Two 3/4-in.  access  holes  were provided  in  the top of the
chamber for  an  air inlet and an  aspirator baffle chamber  connection. The
dust (either polydisperse or monodisperse) is kept in suspension by the fan
and  removed via  an  aspirator.  Upon leaving  the  aspirator,  the particles
impact on the walls of the Lucite® baffle chamber (mounted directly above
the glass fluidizing chamber),  which  removes the larger particles.  The
remaining particles are  piped into  the  exposure  chamber. The chamber
concentrations depend  on the  initial  dust charge,  the  overall  airflow

-------
                                  211

through the chamber, and the aspirator airflow. In addition to providing a
generation method  for noncompactible  dusts,  this  device  also avoids
agglomeration  resulting from  electric  charges normally  associated  with
dusts.
   Another alternative  design  for generating aerosols from loose dusts is
the fluidized  bed generator (FBG).  The  fluidized bed  uses  air blown
through a porous bottom of a container to put dry particles in suspension.
The FBG can be combined with a vertical aspirator, an inverted funnel in
the bed (Tillery et al.  1976), or a horizontal eductor to remove  the dust
particles (Guichard 1976). To  produce fine particles, the bed can be con-
structed  from  100  to  200-Min-diam glass beads (Guichard 1976)  or
100-jtm diam nickel spheres (Willeke et al. 1974). The concentration of
the aerosol and the particle size distribution are dependent on the operat-
ing conditions of the  bed (Carpenter and Yerkes  1980). Bed loading and
superficial air velocity  influence the  mass median diameter  and the con-
centration (bed volume  also affects concentration) in such a  manner that
high bed loadings and volumes and low superficial velocities produce max-
imum concentrations while moderate  bed loadings and high velocities pro-
duce the smallest particle sizes. Thus, a trade-off must be made  between
size and concentration  in  selecting operating  conditions  (Carpenter and
Yerkes 1980). The FBG has been used to generate fibrous dusts (Spurney
et al. 1976, Marple et al. 1978) and shows good potential in this  area.
   The two  systems  just described are typical of  the  recent  loose  dust
dispensing devices. The  majority of these use an  aspirator to remove the
loose particles from various types of  fluidizing chambers.  In  general, this
type of aerosol generator is most useful for processing dusts that cannot be
successfully compacted and redistributed.
   The most  successful and widely  used  technique for generating  dust
aerosols utilizes compressed plugs of powder that are ground or abraded to
release the particles. The Wright dust feed mechanism (Figure 4.21) was
one of the first of these designs and  is still  widely  used today. Basically,
the design consists of a gear-train assembly powered  by a  synchronous
motor which drives a brass cup packed with dust down a spindle shaft in a
rotating motion and against a scraper blade  that continuously abrades  off
a uniform amount of dust (Wright  1950,  Fraser  et al. 1959).  The brass
cup is slightly larger than  the scraper  blade, which is a piston with a
spring  top having  a radial cut edge  turned  up to  form a scraper. Air is
blown into the  lower part of the brass cup; passes through an air channel
under the scraper  blade, where it  picks up the dust; goes down a hollow
shaft that supports the blade; and after leaving the shaft, strikes an impac-
tor plate that breaks up any aggregated particles (Wright 1950, Tillery et
al. 1976). If the dust plug is compacted with a  consistent density and if
90% of the particles have a diameter of 10 pm or less, there  should be no
problem in redistributing the dust  at  the proper size, provided the relative
humidity of the airstream is controlled (Wright 1950, Roe 1968, Drew and
Laskin 1973). Any fluctuations in dust concentration (e.g., due to buildup

-------
                                  212
 Scraper Blade
ORNL-OWG 85-10758



       Dust T'.oe


       Scraper Blade

       Scraper Head
 - Scraper Head
                                     Outlet Nozzle
                                                        mpactor Plate
   Figure 4.21. The Wright dust feed  mechanism. (Source: Adapted from
Leach et al.  1959.)
on the  scraper blade)  can be smoothed out  by adding a  large reservoir
after the impactor.
   Another dust generator using a horizontal feed of a compacted powder
plug has been designed to handle fibrous dusts (Timbrell et al. 1968). The
horizontal feed mechanism consists of a piston-cylinder arrangement with
a threaded  rod controlling the extent  that the powder plug is  introduced
into  the scraper  (Figure 4.22).  A  specially  designed rotor driven by an
electric motor shaves off a layer of dust inside the dispersing chamber. The
particles are carried away  by air entering through  a  side or bottom port
and  exiting through a  port in the  top.  As with the  Wright mechanism,
proper packing of the plug and control of relative humidity are needed to
produce the desired concentrations. The plug should be compacted all at
once  to prevent the formation of  layers.  The Timbrell  generator was
specifically  designed to process asbestos and  is capable of separating indi-
vidual fibers or  fiber bundles for distribution.  Compacted powder  plugs
have  also  been  abraded to  produce  dust  particles  by  using   air jets
(Dimmick 1959, Hounam 1971).

4.4.3.3 Generation of aerosols from liquids, solutions,  and liquid suspensions

   More types of aerosol generators exist to create  particles from liquids,
solutions,  and suspensions  of solids in liquids than from  dust or solids.

-------
                                   213

                                                      ORNL-DWG 85-10759
                                 -AEROSOL EXHAUST
                                            VIEWING PORT
 ROTOR
                                                 COMPACT  FIBROUS
                                                 PLUG
                         AIR
                       INLET
    Figure  4.22. Schematic view of Ettinger's  modification of Timbrell's
 fibrous aerosol generator. (Source: Adapted from Tillery et al. 1976. Used
 with permission of the publisher.)
Depending on the design, these can produce either  liquid droplet or solid
particle aerosols.
   One type of liquid aerosol  generator designed from  some of the same
principles used in dust dispensers is based on  the aspirator. Basically  the
aspirator functions by creating a pressure drop across an orifice or capil-
lary tube, which draws the liquid test agent up  the tube where it is sheared
into particles by the  air stream. Although it is  used  with dusts, the aspira-
tor can function more efficiently with liquids because the flow characteris-
tics are more favorable  and easily achieved.  Thus, many types of liquid
aerosol generators use the aspirator to atomize the liquid into particles.
   A typical atomizer based on the above principle  is the Laskin atomizer
(Fraser et al. 1959). Pressurized air  is pumped up  a central, hollow tube
and out four radially drilled holes at the top. Directly beneath these holes
is a collar with four  vertical capillary tubes lined up with the radial holes.
As air passes over the capillaries, liquid is drawn up and broken into dro-
plets.  The test  agent feed rate in this system  is dependent on the venturi

-------
                                  214

effect at the top of the capillaries and the air flow rate through the atom-
izer. Because  the most efficient atomization of liquids  occurs near the
maximum air  flow rate (Gage 1953), only a limited range of output varia-
tion is  possible by  changing  the  air flow  rate. A motor-driven  syringe
delivery system for the test liquid can be used to avoid this limiting factor
(Gage 1953).  Although this  system is  used primarily with liquids, solid
particles can be created by mixing in dry or hot air with the droplets to
evaporate the  liquid (Gage 1968),  leaving crystals (when  using a solution)
or solid particles (when using  a solid-liquid suspension).
    A special category of aspiration atomizers is often  referred to as nebu-
lizers. Nebulizers atomize liquids,  suspensions,  and solutions, usually with
compressed air to  form particles,  and  also include size-selective devices,
most frequently impaction surfaces.
    The Vaponephrin nebulizer is an all-glass design consisting of a nozzle
directing a  stream  of air across  the top of a capillary tube, with the bot-
tom of the tube resting in the test agent liquid.  Directly opposite from the
nozzle is  a glass sphere  that  alters the direction of the  air stream. This
change  in  direction impacts  larger particles  on the sphere  and  allows
smaller particles to  remain airborne (Fraser et al. 1959). The  whole com-
plex is  contained  in a glass  sphere to  control loss of liquid. The  critical
dimensions  of the design  are the diameter of the nozzle orifice, which lim-
its the  air  volume  and determines the air  pressure  needed; the distance
from the nozzle to  the capillary; and the distance from the capillary to the
impaction sphere.
    Another simple  nebulizer design (Figure 4.23)  is  the DeVilbiss nebu-
lizer. This  is  also  a glass design with  a  vertical aspirator  (Raabe  1976).
The impaction surface is  the  curved area of the exhaust tube, which  res-
trains about 99%  of the particles  produced (Mercer et al. 1968b). Other
impaction  surfaces include: a  series of  baffle holes that require the parti-
cles to  follow a tortuous path  (the  Dautrebande  nebulizer), the curved
walls of  the  generator  (the Lauterbach nebulizer),  a hollow  cylinder
around  the aspirator jet (the  Collison  nebulizer), and a sharp right angle
with a flat surface  or a rounded  curve baffle (Wright, Retec, and Lovelace
nebulizers)  (Whitby et al. 1965, Mercer et  al.  1968, Mercer 1973, Raabe
1976). Some  of the output characteristics of the common nebulizers are
given in Table 4.8  (Tillery et  al. 1976). In general, the differences in out-
put characteristics  are due to the  internal geometry of the chambers  and
the flow rates through them (Tillery et al. 1976). For specific applications,
these factors,  and the critical dimensions mentioned above, can be modi-
fied to produce the desired output concentrations and particle sizes.
    A unique variation of the  nebulizer design is fractionation of the liquid
stream  by ultrasonic  vibrations rather than by air pressure. Piezoelectric
crystals are typically  used to create the vibrations, which are transmitted
through  a coupling fluid  (Figure  4.24) to the  test liquid  (Mercer 1973,
Tillery et al.  1976).  The test  liquid reacts to the vibrations by forming a
fountain above the crystal with droplets breaking off  at the top. The size

-------
                          215
                                      ORNL-DWG 85-10760
     VENT
LIQUID INLET
     TUBE
              COMPRESSED  AIR
                       IN

  Figure 4.23. Diagram  of the DeVilbiss nebulizer. (Source: Adapted
from Mercer et al. 1968b. Used with permission of the publisher.)

-------
Table 4.8. Output Characteristics of Some Compressed Air Nebulizers
Nebulizer
Vaponefrin
DeVilbiss #40
DeVilbiss #40
DeVilbiss #40
Bennett Twin (2814)
Puritan (R6-051)
Lauterbach
Lauterbach
Lauterbach
Dautrebande D-30
Dautrebande D-30
Dautrebande D-30
Lovelace
Lovelace
Lovelace
Collison
Jet
pressure
(psi)
12
10
20
30
7.5
23
10
20
30
10
20
30
10
20
30
15
Mass
median
diameter
(^m)
5.6
4.1
3.2
2.8
6.8
6.5
3.8
2.4
2.4
1.7
1.4
1.3
NGd
5.4
NG
NG
Geometric
standard
deviation
1.80
-1.85
-1.85
-1.85
1.80
1.90
-2.05
-2.05
-2.05
-1.65
-1.65
-1.65
NG
1.90
NG
NG
Specific output
(liL solution/L
jet air)
-29
16.0*
13.8*
12.8*
23.8
26.6
3.9
5.7
6.0
1.42
2.3
2.4
15.3
30
35
8.7
Output (fiL
solution/rain")
117
155
229
270
119
266
30C
67C
91
21
49
65
14e
39e
58.3e
53
Reference
Mercer et al. 1965
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1965
Mercer et al. 1965
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1968b
Mercer et al. 1968b
May 1973

-------
                                              Table 4.8. (continued)

Nebulizer
Collison
Collison
Retec
Retec
Retec

Jet
pressure
(psi)
25
30
20
30
50
Nlass
median
diameter
(Mm)
1.9
NG
5.7
3.6
3.2

Geometric
standard
deviation
2.5
NG
1.8
2.0
2.2

Specific output
(nL solution/L
jet air)
6.7
5.8
35.2
35.9
31.9

Output (/*L
solution/mina)
55
55
208
284
376

Reference
May 1973
May 1973
Raabe 1972
Raabe 1972
Raabe 1972
"Dioctyl sebacate is the nebulized liquid for the Collison, while water containing salt or a fluorescent dye is the nebu-
lized liquid for the other nebulizers.
bZero auxiliary air flow.
cOrifice diameter, 0.032 in.
dNG = not given.
'Orifice diameter, 0.001 in.
Source: Adapted from Tillery et al. (1976).

-------
                                218
                                             ORNL-OWG 85-10761
           AEROSOL OUT
AIR  IN
             J
                        COUPLING  FLUID
                          TRANSDUCER
   Figure 4.24. Schematic view of ultrasonic nebulizer. (Source: Adapted
from Tillery et al. 1976. Used with permission of the publisher.)
of the droplets depends on the frequency of the acoustic field, the proper-
ties of the test chemical, and the rate at which they are removed from the
fountain (Raabe 1976). A stream of air is routed through the chamber to
disperse the accumulating aerosol. Some output characteristics  of various
ultrasonic  nebulizers are given in Table 4.9. In general, ultrasonic nebuliz-
ers produce more monodispersed aerosols than do typical compressed air
nebulizers (Mercer 1973). Kreyling and Perron (1983) describe an  ultra-
sonic nebulizer design that produces aerosols from low concentrated  parti-
cle solutions, can be optimized to produce  specific particle sizes by varying
air pressure and impaction plate angle, and can deliver a constant concen-
tration for  more than one hour. Another form of  ultrasonic nebulizer (the
Hartman Whistle) utilizes compressed air  shot into a resonant chamber to
produce sound waves of 25 kilocycles, which immediately disperse  any
introduced liquid into a fine aerosol (Fraser et al. 1959).
   The  most commonly  used technique  for generating  liquid aerosols,
especially  monodisperse  aerosols, is the  spinning  disk  generator first
developed  by Walton and Prewett (1949).  The generator consists of  a flat
horizontal  disk surface,  which is  rotated  at high speeds (up to  100,000
rpm) by a direct drive air or electric motor (Raabe 1976). The  test  liquid
is fed onto the center of the disk from a fixed-height reservoir or a syringe
pump and is spread out in a thin layer by centrifugal force (Figure  4.25).

-------
                         Table 4.9. Output Characteristics of Some Ultrasonic Nebulizers
Nebulizer
Mist O2 Gen,
50-mL cup'
Mist O2 Gen,
10-mL cup"
DeVilbiss,
gain setting 2b
DeVilbiss,
gain setting 3b
DeVilbiss,
gain setting 4b
Denton-Swartz
Denton-Swartz
Frequency
(MHz)
1.40
1.40

1.35

1.35

1.35

1.00
3.00
MMD
(/.m)
6.5
6.5

5.7

6.9

6.9

NGC
NG
CMD 0.34X Output
(
-------
                                 220
                                                ORNL-DWG 85-10762

                                      Liquid Feed
                                               Spinning Disk
                                             /('IOO.OOO rpm)
                                                      Primary
                                                      'Droplets
        Satellite
        Droplets
        Satellite
        Air Flow
   Figure 4.25.  Schematic drawing of a spinning  disk generator used to
produce monodisperse aerosols  of  both  soluble and insoluble forms from
solutions of suspensions. Air flow into the satellite  collector is adjusted so
that the inertia of the primary particles  allows them to enter the. main air
flow.  (Source:  Adapted from  Raabe  1976.  Used  with  permission  of
Academic Press.)

-------
                                  221

As the liquid  layer reaches and accumulates at the edge of the disk, dro-
plets  are  thrown off when the centrifugal  force exceeds  the  restraining
capillary action (Mercer 1973). The majority of the liquid mass forms par-
ticles  of the desired  size, but smaller satellite droplets  are also  formed,
usually four for every primary droplet (May 1949). A stream of air,  insuf-
ficient to  overcome the inertia of the primary droplets, is  used to control
the satellite droplets.
   A similar generator, the spinning top, also was designed  by Walton and
Prewett (1949) and modified by  May (1949, 1966).  It used  a flat disk
with a conical bottom cut into turbine blades. Pressurized  air  directed at
the blades provided support and rotation  for the disk.  Generally, spinning
tops  are  smaller and operate  at higher speeds (up to  240,000  rpm) than
the spinning disk (Mercer 1973, Tillery et  al. 1976).  Using the spinning
top, May was able  to generate monodisperse particles  with coefficients of
variation as low as 0.05 when the  disk surface was properly treated. The
May  spinning top generator  is  usually  mounted  within the exposure
chamber,  as droplets are thrown  off in a 360° area  around  the top. A
recent modification incorporates  an  encapsulating plenum  to control  the
aerosol generated, allowing it to be  used  outside the chamber  (Gussman
1981).
   The droplets  created  by spinning  disks  and  tops are  mostly mono-
disperse when the liquid feed rate  is slow with particle size varying with
the rotational  speed (Roe 1968). Table 4.10 shows the change in  particle
size with  the  rotational speed of a spinning disk that used an air-driven
tool grinder motor (Whitby et al. 1965). The range of particle  sizes  possi-
ble with  the spinning disks  is usually above 0.5 ^m  and can be up  to  50
iim (Raabe 1976). Dry particles can also be  created with the spinning disk
by evaporation of the liquid solvent. Hurford (1980)  describes a modifica-
tion of the May generator that produces a  monodisperse, solid aerosol with
a size range of 1 to 6 microns and a standard deviation of 1.1. The sizes
were  verified  by optical and scanning electron microscopy. The modifica-
tion to the May generator, by Gussman (1981), also provides the  smooth,
nonturbulent column  of rising air  needed to evaporate the liquid solvent
and produce particles of 5 pm  or less in diameter.
   The condensation of vapors on nuclei has also  been  used to create
monodisperse  aerosols.  Sinclair and LaMer (1949)  used   the  apparatus
shown in Figure 4.26 to form particles with a  narrow size range. The
liquid test  agent is first vaporized in  the boiler, then  nuclei (from  an elec-
tric spark or NaCl-coated wire source) are added, and the  mixture is car-
ried by a stream of air to the reheater. Because the reheater is  at a higher
temperature, complete vaporization of the liquid  is  assured. The  mixture
then enters the chimmey, where it is cooled to allow condensation  of the
vapor on the nuceli.  By carefully controlling the  nucleus  size  and  the
cooling process,  particles  of uniform  size are created (Mercer 1973,  Tilley
et al  1976) Modifications in the basic design were made by Rapaport and
Weinstock  (1955), Muir (1965),  and Liu  et al. (1966).  These modified

-------
               Table 4.10. Operating Conditions at Which Model I Spinning Disk Was Used*
Operating
conditions
Main droplet diameter (pm ± 10%)
Dye concentration by weight (g/g)
Solid particle diameter (jim)
Discharge air flow (cfm)
Orifice diameter (in.)
Satellite air flow (cfm)
Satellite funnel diameter (in.)
Liquid feed rate (mL/min ± 20%)
Feed needle size (number)
Particle generation rate
(particles/min ± 20%)
Projection of the hypodermic needle
below the satellite funnel (in.)
70,000
19
5 X 10-5to
3 X Iff3
0.6-2.4
100
6
15
2.5
2.5
22
350 X 106
1/16
Air
45,000
28
1 X 10-5to
3 X 10-3
0.6-3.5
100
6
15
3
3
22
130 X 106
3/32
motor speed (rev/n
30,000
41
5 X ia5to 5
3 X 1(T3
1.3-5.1
70
7.5
15
4
3
22
40 X 106
1/8
nin)
20,000
60
X 10-5to
3 X 10°
1.9-7.5
50
9
16
5
M
19
15 X 10'
3/16
12,000
100
5 X iasto
3 X Iff3
3.2-12
40-50
15
18
6
4
19
3 X 106
3/16
"Data obtained using methylene blue dye in alcohol. Disk diameter = 4.65 cm.
Source: Adapted from Whitby et al. (1965).

-------
                                223
                                FILTERED
                                   AIR
                                                  ORNL-DWG 85-10763
                  ELECTRIC
                  HEATER
;\   ,
 -INSULATION
             CONDENSATION
             NUCLEI GENERATOR
                                                  AEROSOL
                                                  MATERIAL
                                             VAPORIZER
             AEROSOL
               OUT
ELECTRIC
 HEATER
   Figure  4.26. Schematic view  of  Sinclair-LaMer monodisperse aerosol
generator.  (Source:  Adapted from Tillery et al.  1976. Used with permis-
sion of the publisher.)
designs are capable of producing monodisperse aerosols in the range of 0.3
to 1.4 /im (Mercer 1973, Tillery et al. 1976).

4.5. CONCLUSIONS AND RESEARCH RECOMMENDATIONS
4.5.1 Conclusions

   Toxicity tests designed to evaluate effects from inhalation exposure
involve many  of the same variables as tests using other exposure routes.
Exceptions to the standard protocols include:  (1) selection of test species
should consider anatomical and physiological similarity of respiratory tract
to that of humans;  (2) exposure modes can be intermittent (6 to 8 h/day
for 5 days/week) or continuous; (3) determination of actual. effective dose
is more complicated — generally the chamber concentration times length of
exposure  is used; and (4) observations for  effects include more extensive
pathological examination of respiratory structures and  use of respiratory
function tests.
   An important  aspect  of  inhalation testing is the  design of the test
chamber  of equipment for generating the exposure  atmosphere, and  of
equipment to monitor chamber atmospheric conditions and test material

-------
                                 224
concentrations. In  most  tests,  operation  of the  test  chamber  is  in a
dynamic mode, where  a continuous stream of test agent and air is vented
through the chamber. The test chamber is usually designed for whole-body
exposure, although nose-  and head-only exposure  systems  are also  used.
The shape of most large chambers is cubical or hexagonal with pyramidal
or conical additions at top and bottom to increase distribution efficiency of
air within the chamber. Chamber size is governed by air distribution and
animal volume considerations. Construction materials need to be smooth-
surfaced,  nonabsorbant materials that are resistant to a wide range of
chemicals.  Airfow systems for the chamber should contain units to condi-
tion, filter, meter, and control the flow of air and test agent to and from
the chamber. The chamber should be maintained at a negative operational
pressure and within narrow temperature and humidity ranges.
    Sampling and monitoring of the test atmosphere and chamber condi-
tions should be performed during operation. Sampling techniques for moni-
toring  gas and vapor  atmospheres usually rely on  filters and  automated
analytical procedures.  Aerosol atmospheres are more difficult  to sample,
particularly at the breathing zones, and a whole array of equipment has
been developed for this, including elutriators, cascade impactors, centrifu-
gal force samplers, and optical monitors. Some type of size  separation for
the particles is usually part of the procedure.
    Generation of the  test atmospheres also differs based on the type of
test material.  Gas and vapor generation is  usually achieved by release into
the airstream  from high  pressure cylinders or through vaporization  using
controlled heating or counter-current techniques. Aerosol atmospheres can
be generated in monodisperse or polydisperse forms. Generation  of aerosols
from solid materials can  be accomplished  from loose materials by use of
aspirators and a vibrating, rotating, or stirring technique to create a source
suspension. Compacted forms of test materials can  be put into an  aerosol
suspension by  abrading the  surface with a  scraper blade in the  path  of an
airstream.  Generating aerosols from  liquid test materials can be done in
more varied methods.  Aspirators with an  impaction surface can produce
particles of limited  size  ranges.  This technique  can be combined with a
fractionating device  to produce aerosols.  Applying the liquid to a spinning
disk or top is another technique frequently used to produce liquid aerosols.

4.5.2 Research Recommendations
    More research is needed on:

•   Applicability of test species,  particularly regarding metabolic similarity
    and particle deposition sites in the respiratory tract.

•   Determination of actual  effective doses  in inhalation testing.  This  could
    include actual dose delivered to  animal lungs, dose delivered to  blood
    of test animals, and dose reaching target organs.

-------
                                   225
•  The use of alternative testing methods for inhalation toxicity, perhaps
   more in vivo tests such as lavage fluid tests or in vitro tests that could
   be used to determine acute or target organ (lung) effects.
•  The use of respiratory  function  tests  to widen their  applicability,
   increase knowledge as to the significance of  their data, and provide
   data for comparisons between test animals and humans.

•  The various types of  atmosphere  generating  equipment,  particularly
   aerosol generators, including comparisons of efficiency, ability to main-
   tain desired particle  size and distribution ranges, and  avoidance of
   operating problems (e.g.,  particle clumping on size separators in aerosol
   generators).
•  The potential  for  extrapolation  of results from oral or  intravenous
   exposure  studies to  inhalation studies  for non-respiratory-tract effects
   or for  certain chemical classes. This might allow reduction  in the total
   number  of inhalation studies or in those studies aimed  at systemic
   effects of inhalation exposure.
•  The extrapolation of inhalation toxicity  data  gained  from  animal  stu-
   dies to human exposure.  Although this problem is not specific to inha-
   lation  studies,  it is one of the most important areas  in which existing
   data are insufficient to determine the true relationships between animal
   and human studies.
•  The relationship between toxicity evaluation of a chemical  and hazard
   evaluation of a chemical, particularly  as  applied to inhalation studies.
   The regulatory guidelines require toxicity data  applicable  to general
   situations, but the real life situation requires hazard (toxicity X expo-
   sure situation)  data  for  determining the impact of a chemical in a
   specific situation. How to satisfy both requirements without duplication
   of testing efforts needs more study.


4.6. LITERATURE CITED

Alarie  Y.  1966. Irritating properties  of airborne  materials  to the  upper
respiratory tract. Arch Environ Health 13:433-449.

Albert  RE, Lippmann M, Spiegelman J,  Luzzi A, Nelson N.  1967.  The
deposition and clearance of radioactive particles in the human lung. Arch
Environ Health 14:10-15.

Amdur MO. 1957- The influence of aerosols upon the respiratory response
of guinea  pigs to sulfur dioxide.  Am Ind Hyg Assoc Q 18:149-155.

Amdur MO. 1958. The respiratory response of guinea pigs to sulfuric  acid
mist. Arch Ind Health 18:407-414.

-------
                                226

Amdur MO,  Mead J.  1958.  Mechanics of respiration in  unanesthetized
guinea pigs. Am J Physiol 192:364-368.

Barer GR, Edwards CW, Jolly AI. ^1976. Changes in the carotid body and
the ventilatory response to  hypoxia  in chronically hypoxic rats. Clin Sci
Mol Med 50:3111313.

Barrow C.  1978.  Is ammonia a  problem? In: Proceedings, Workshop on
Inhalation  Chamber Toxicology, Drew  RT,  ed.  Brookhaven  National
Laboratory Associated Universities, Inc. BNL 51318. pp. 55-58.

Barrow CS, Steinhagen WH.  1980.  NH3 concentration in  the expired air
of the rat:  importance  to inhalation toxicology. Toxicol Appl Pharmacol
53:116-121.

Barrow RE,  Vorwald AJ,  Domier  E.  1971. The measurement of small
animal respiratory volumes by capacitance  respirometry.  Am Ind Hyg
Assoc J 32:593-598.

Beethe RL, Wolff  RK, Griffis  LC,  Hobbs CH, McClellan RO. 1979.
Evaluation of a recently designed multi-tiered exposure chamber. Lovelace
Biomedical and Environmental Research Institute, LF-67. 23 pp.

Bernstein DM, Drew RT. 1980. The major parameters affecting tempera-
ture inside inhalation chambers. Am  Ind Hyg Assoc J 41:420-426.

Boecker BB, Aguilar FL, Mercer TT. 1964. A canine inhalation exposure
apparatus  utilizing   a  whole  body  plethysmograph.  Health   Phys
10:1077-1089.

Brain JD, Mensah GA. 1983. Comparative  toxicology of  the respiratory
tract. Am Rev Respir Dis 128:S87-S90.

Brown RC, Poole A. 1983.  Inhalation toxicology. In: Animals and alterna-
tives in toxicity testing, Balls  M, Riddell RJ, Worden AN, eds. Academic
Press, New York, pp. 300-309.

Bryan RJ.  1970.  Generation  and monitoring of gases for  inhalation stu-
dies. In: Inhalation Carcinogenesis, Hanna Jr. M, Nettesheim P, Gilbert J,
eds. U.S. Atomic Energy Commission Symposium Series 18. pp.  193-205.

Caldwell  EJ,  Fry DL.  1969.  Pulmonary mechanics in the  rabbit.  J Appl
Physiol 27:280-285.

Cambridge GW.  1973. An  examination  and  evaluation  of  inhalation
toxicity studies. Aerosol Age. 18:32-35, 67-68.

-------
                                 227

Campbell KI. 1976. Inhalation toxicology. Clin Toxicol 9(6):849-921.

Carlon HR, Kimball  DV,  Wright RJ.  1980. Determination of aspirated
aerosol sampler efficiencies using laser reference techniques. Army  Arma-
ment  Research and Development  Command. Aberdeen Proving Ground,
MD. ARCSL-TR-80003; AD-E410266. 21  pp. (abstract only).

Carney  IF.  1979.  Biological  problems in  defining hygiene standards  for
particulates. Ann Occup Hyg 22:163-173.

Carpenter CP, Smyth HF, Pozzani UC. 1949.  The assay of acute vapor
toxicity, and the grading and interpretation of results on 96 chemical com-
pounds. J Ind Hyg Toxicol 31:343-346.

Carpenter RL, Beethe RL. 1978. Cones, cone angles, plenums, and mani-
folds. In: Proceedings, Workshop on inhalation chamber toxicology, Drew
RT, ed.  Brookhaven  National Laboratory Associated Universities,  Inc.
BNL 51318. pp. 21-30.

Carpenter RL, Yerkes K.  1980. Relationship between fluid bed aerosol
generator operation and the  aerosol  produced. Am  Ind Hyg  Assoc J
41:888-894.

Carson TR, Rosenholtz MS, Wilinski FT, Weeks MH. 1962. The response
of animals  inhaling nitrogen dioxide for single,  short-term exposures. Am
Ind Hyg Assoc J 23:457-462.

Casarett LJ. 1975. Toxicology of  the respiratory system. In:  Toxicology,
the basic science of poisons, Casarett LJ, Doull J, eds. MacMillan Publish-
ing Co., Inc., New York. pp. 201-224.

Cavagna GA, Stemmler EJ,  DuBois  AB. 1967.  Alveolar resistance  to
atelectasis.  J Appl Physiol 22:441-452.

Clark DB.  1977.  Long-term  inhalation toxicology studies. In: Current
approaches in  toxicology,  Ballantyne B, ed.  John Wright and  Sons Ltd.,
Bristol, pp. 105-115.

Comroe  JH Jr, Nisell OI, Nims  RG.  1954. A simple method for con-
current measurement of compliance and resistance to breathing in anesthe-
tized animals and man. J Appl Physiol 7:225-228.

Cotabish HN, McConnaughey PW,  Messer HC. 1961.  Making known
concentrations  for  instrument   calibration.  Am Ind   Hyg  Assoc  J
22:392-402.

-------
                                228
Crossland NJ, Horsfall GB, Oxenham ST, Shaw JS, Turnbull MJ. 1977.
A  simple device for measurement of  respiratory rate in the mouse. Br J
Pharmacol 61:490P-491P.

Davis J,  Morris B.  1953. Factors affecting the intrapleural pressure and
pulmonary ventilation in rats. Austral J Exp Biol 31:201-214.

Decker JR, Park JF, Craig DK.  1979. A method for measuring respiratory
volume parameters of  large animals during exposure to aerosols. Am Ind
Hyg Assoc J 40:567-577.

Dennis MW,  Douglas JS, Casby JU, Stolwijk JAJ, Bouhuys A. 1969. On-
line  analog computer for dynamic lung compliance and pulmonary resis-
tance. J Appl Physiol 26:248-252.

Denton MB, Schwartz DB. 1974. An improved ultrasonic nebulizer system
for the generation of  high  density aerosol dispersions. Rev Sci Instrum
45:81-83.

De Zaiacomo T, Tarroni G, Prodi V,  Melandri C,  Formignani M, Lom-
bardi CC,  Barilli L,  Olivieri P, Niccolai I.  1983. An  intercomparison
among aerodynamic  sizing instruments. J Aerosol Sci 14:314-317.

Diamond L, Lipscomb W. 1970. Design of a small animal pneumotacho-
graph. J Appl Physiol 29:720-722.

Diamond L, Adams GK, Bleidt  B, Williams B. 1973. Experimental study
of a potential anti-asthmatic agent: SCH 15280. J. Pharmacol Exp Therap
193:256-263.

Diamond L, O'Donnell M. 1977. Pulmonary mechanics in normal rats. J
Appl Physiol: Respirat Environ Exer Physiol 43:942-949.

Dimmick RL. 1959. Jet dispenser for compacted powders  in the one-to-
ten-micron range. Arch Ind Health 20:8-14.

Doe JE, Tinston DJ. 1981. Novel chambers for long term inhalation stu-
dies. In: Proceedings of the inhalation toxicology and technology sympo-
sium, Kalamazoo,  MI, October 23-24, 1980. Ann Arbor Science Publish-
ers, Inc., Ann Arbor, MI. pp. 77-88.

Dorato MA,  Carlson  KH,  Copple DL.  1983. Pulmonary mechanics in
conscious Fischer  344 rats: multiple  evaluations using nonsurgical  tech-
niques. Toxicol Appl Pharmacol 68:344-353.

-------
                                 229


 Drew  RT.  1978. Exposure  of rodents  to  participate  carcinogens,  first
 presentation. In:  Proceedings, workshop on inhalation chamber toxicology,
 Drew  RT, ed. Brookhaven  National Laboratory Associated  Universities,
 Inc. BNL51318. pp. 63-64.

 Drew  RT, Laskin S. 1971.  A  new dust-generating system for inhalation
 studies. Am Ind Hyg Assoc J 32:327-330.

 Drew  RT,  Laskin  S.  1973.  Environmental  inhalation  chambers.  In:
 Methods of animal experimentation. Vol. IV.  Environment and the special
 senses, Gay WI, ed. Academic Press, New York. pp.  1-41.

 Drorbaugh JE.  1960. Pulmonary  function in different  animals. J  Appl
 Physiol 15:1069-1072.

 Dubin  ED, Morrison G.  1969. A face mask and mouthpiece for respiratory
 studies in unanaesthetised beagle dogs. J Appl Physiol 27:104-105.

 DuBois AB, Rogers RM. 1968.  Respiratory factors determining the tissue
 concentrations of inhaled toxic substances. Respir Physiol  1968(5):34-52.

 Dungworth DL, Schwartz W, Tyler WS, Phalen RF. 1976. Morphological
 methods for evaluation of pulmonary toxicity  in animals. Ann Rev Phar-
 macol Toxicol 16:381-399.

 Ferin  J.  1978.  Can  laminar flow systems  be used?  In:  Proceedings,
 workshop  on inhalation  chamber  toxicology,  Drew  RT,  ed.  Brookhaven
 National Laboratory Associated  Universities, Inc. BNL 51318.  pp.  39-42.

 Formignani M, Melandri C, Tarroni  G, Prodi V, De Zaiacomo T. 1982.
 Calibration  apparatus  for  aerosol samplers. Comitato Nazionale  per
 1'Energia  Nucleare Bologna (Italy). CNEN-RT/PROT-(82)4.  32  pp.
 (abstract only).

 Francovitch RJ, Raitano L, Lee T, George WJ. 1982. Comparison of der-
 mal and inhalation toxicities of volatile agents employing a dual  function
 chamber. Fed Proc 41 abstract 7583.

 Frank  NR,  Speizer FE.  1965.  SO2 effects on  the  respiratory system in
dogs. Arch Environ Health 11:624-634.

Fraser   DA, Bales  RE,  Lippmann M,  Stokinger  HE.  1959.  Exposure
chambers for research in animal inhalation. Public Health Monograph No.
 157. 54pp.

-------
                                 230
Fuchs NA, Sutugin SG.  1966. Generation and use of monodisperse aero-
sols. In: Aerosol science, Davis CN, ed. Academic Press, London.

Gage JC. 1953. A controlled fluid-feed atomizer. J Sci Instrum 30:25.

Gage JC.  1968. Toxicity  of paraquat and diquat  aerosols generated by a
size-selective  cyclone:  effect of particle size distribution. Br J Ind Med
25:304-314.

Gardner DE, Miller FJ, Blommer EJ, Coffin DL.  1979. Influence of expo-
sure mode on the toxicity  of NO2. Environ Health Perspec 30:23-29.

Gat N.  1980. Development study of a two-stage continuous flow impactor.
Army Armament Research and  Development Command. Aberdeen Prov-
ing Ground,  MD.  TRW-33437-6001-RU-00;  ARCSL-CR-80009; AD-
E410241. 88 pp. (abstract only).

Giles RE, Finkel MP, Mazurowski J. 1971. Use of an online computer for
the evaluation of pulmonary resistance and dynamic  compliance in  the
anaesthetised dog. Arch Pharmacodyn 194:213-222.

Giovacchini  RP.  1972. Old and new issues in  the safety evaluation of
cosmetics and toiletries. Crit Rev Toxicol l(4):361-378.

Griffis LC, Henderson JH, Pickrell JA.  198la. A method for determining
glass in rat lung after  exposure to a glass fiber aerosol. Am Ind Hyg Assoc
J 42:566-569.

Griffis  LC,  Wolff RK, Beethe  RL, Hobbs RF,  McClellan  RO.  198Ib.
Evaluation of a multitiered inhalation exposure chamber. Fund Appl Toxi-
col 1:8-12.

Gross SB. 198la. Regulatory guidelines for inhalation toxicity testing. In:
Proceedings  of the inhalation toxicology and technology symposium, Leong
BKJ, ed. Ann Arbor Science Publishers, Inc. pp. 279-298.

Gross  KB.  1981b.  Pulmonary  function testing  of animals  chronically
exposed to diluted diesel exhaust for 267 days. Environ Int 5:331-338.

Guichard JC. 1976. Aerosol generation using fluidized  beds. In: Fine par-
ticles, Liu BYH, ed. Academic Press, New York. pp. 173-193.

Gussman RA. 1981. A further modification of the May spinning top aero-
sol generator. Am Ind Hyg Assoc J 42:208-212.

-------
                                 231
Haber R. 1924. "Funf vortrage aus den jahren 1920-1923." No. 3. Die
Chemie in  Kriege,  No.  5.  Zur  Geschichte  des  Gas  Kampfes.  Julius
Springer, Berlin.

Halbert MK, Mazumder MK,  Bond RL.  1981. Size-distribution analysis
of respirable particulates in  cosmetic aerosols: a methodological compari-
son. Food Cosmet Toxicol 19:85-88.

Hammond  PB.  1970.  The  use  of  animals in  toxicological  research
(prepared  for  the  National  Institute  of Environmental Health  Sciences
Task  Force  on  Research Planning in  Environmental Health  Sciences)
Natl.  Library  of Medicine,  Bethesda, MD.  (As reported  in  Drew and
Laskin 1973).

Hatch TF, Gross P.  1964. Pulmonary deposition and retention of inhaled
aerosols. Academic Press, New York. 189 pp.

Haun CC. 1972. Continuous animal exposure to methylene chloride. In:
Proceedings  of  the  2nd annual  conference on environmental toxicology,
Dayton, Ohio, 1971. U.S. Government Printing Office, pp. 309-326.

Heindryckx R, Dams R.  1981. Investigation of errors  associated  with the
use of cascade impactors. Sci Total Environ 19:167-178.

Hemenway DR, MacAskill  SM.  1982. Design, development  and  test
results of a  horizontal flow  inhalation toxicology facility. Am  Ind Hyg
Assoc J 43:874-879.

Hemenway DR, Carpenter  RL, Moss OR.  1982.  Inhalation  toxicology
chamber performance:  a quantitative  model.  Am  Ind Hyg  Assoc  J
43:120-127.

Henderson RF, Damon EG, Henderson  TR.  1978.  Early damage indica-
tors in the lung. I. Lactate dehydrogenase  activity in the airways. Toxicol
Appl Pharmacol 44:291-297.

Henderson RF, Rebar AH, DeNicola DB.  1979a. Early damage indicators
in the lung.  IV.  Biochemical and cytologic response of the lung to lavage
with metal salts. Toxicol Appl Pharmacol 51:129-135.

Henderson RF, Rebar AH, Pickrell JA,  Newton GJ. 1979b.  Early damage
indicators in the lung.  III.  Biochemical and cytological response  of the
lung to inhaled metal salts. Toxicol Appl Pharmacol 50:123-136.

-------
                                 232
Henderson  RF, Rebar AH, DeNicola DB, Henderson TR.  1980. The use
of pulmonary lavage fluid in screening for early indicators of lung injury:
In: Pulmonary toxicology of respirable particles, Saunders CL, Cross FT,
Dagle GE, Mahaffey JA, eds. U.S. Department of Energy, pp. 378-391.

Henry CJ,  Hwang  KK,  Kanagalingam KK, Kouri RE.  1983. Recent
developments in inhalation toxicology: evaluation  of selected short-term
endpoints following 'nose-only' exposure of rodents. In:  Safety evaluation
and regulation of chemicals, Homburger F, ed. Karger, NY.  pp. 233-242.

Hiett DM.  1974. Tests of ventilatory function for use in long-term animal
studies. Br J Ind Med 31:53-58.

Hinners RG. 1978. A system for automatically monitoring  chamber tem-
perature,  humidity,  and  pollutant  concentrations.   In:  Proceedings,
workshop on inhalation chamber toxicology, Drew  RT, ed. Brookhaven
National Laboratory Associated Universities, Inc. BNL 51318. pp. 89-92.

Hinners RG, Burkart JK, Contner GL. 1966.  Animal exposure chambers
in air pollution studies. Arch Environ Health 13:609-615.

Hinners RG, Burkart JK,  Punte CL. 1968. Animal inhalation  exposure
chambers. Arch Environ Health 16:194-206.

Holmberg RW, Moneyhun  JH. 1980. Inhalation exposure system for toxi-
cological studies of diesel oil  aerosols. Oak Ridge National Laboratory.
CONF-801156-2.  25 pp. (abstract only).

Holmberg RW, Moneyhun JH, Dalbey WE.  1981. An exposure system for
toxicological studies of  concentrated oil  aerosols.  In: Proceedings of the
inhalation toxicology and technology  symposium, Kalamazoo, MI, October
23-24,  1980.  Ann  Arbor Science  Publishers,  Inc.,  Ann  Arbor,  MI.
pp. 53-64.

Hounam RF. 1971. The konimiser—a dispenser for the continuous genera-
tion of dust clouds from milligram quantities of asbestos. Ann Occup  Hyg
14:329-335.

Hurford MJ.  1980.  Production  of  monodisperse  soild  aerosols  with  the
May spinning-top  generator. United Kingdom Atomic Energy Establish-
ment, Winifrith, England. AEEW-M-1816. 26 pp. (abstract only).

Illing JW,  Gardner DE, Coffin  DL.  1975. Motorized exercise wheel for
exposure of small  animals  to  environmental  pollutants. Lab Animal Sci
25:103-105.

-------
                                 233

Jersey GC, Kociba RJ. 1979.  The importance and results of proper mor-
phological evaluation of the nasal turbinates in inhalation toxicity studies.
Toxicol Appl Pharmacol 48:A1. (abstract only).

Juhos LT,  Sasmore DP, Falconer DG, Tyson CA.  1985.  Tests for pul-
monary function. In:  Organ function tests  in  toxicity evaluation, Tyson
CA, Sawhney DS, eds.  Noyes Publications, Park Ridge, NJ.  pp.  123-167.

Kawai K, Nozaki K.  1978.  Inhalation techniques for developmental toxi-
city testing. Teratology 18:13la.

Kennedy GL,  Trochimowicz HJ.  1982.  Inhalation toxicology. In: Princi-
ples and  methods of toxicology, Hayes AW, ed. Raven Press, New York.
pp. 185-207.

Kenoyer  JL,  Phalen RE, Davis  JR.  1981. Particle  clearance from  the
respiratory  tract as a test of toxicity: effect of ozone on  short and long
term clearance. Exper Lung Res 2:111-120.

King RJ. 1974. The surfactant system of the lung. Fed Proc Fed Am Soc
Exp Biol 33:2238-2247.

King TKC. 1966a.  Measurement of functional residual capacity in the rat.
J Appl Physiol 21:233-236.

King TKC. 1966b. Mechanical properties of the lungs in the rat. J Appl
Pharmacol 21:259-264.

Klein GW, Geary  DL.  1982.  Analytical monitoring for  inadvertent test
chemical release during the conduct of  inhalation toxicity  studies. Abstr
Pap Am Chem Soc 183:CHAS-23. (abstract only).

Koen PA, Moskowitz GD, Shaller TH. 1977. Instrumentation for measur-
ing functional residual capacity in small  animals. J Appl Physiol: Respirat
Environ Exercise Physiol 43:755-758.

Koizumi  A, Ikeda M.  1981. A servomechanism  for vapor concentration
control  in  experimental exposure  chambers.  Am  Ind  Hyg  Assoc  J
42:417-425.

Koo  KW, Leith DE, Sherter CB, Snider GL. 1976. Respiratory mechanics
in normal hamsters. J Appl  Physiol 40:936-942.

Kreyling WG,  Perron  GA. 1983. An aerosol  nebulizer  for  low concen-
trated particle suspensions. J Aerosol Sci 14:264-267.

-------
                                 234


Laskin S, Drew RT.  1970.  An inexpensive  portable inhalation chamber.
Am Ind Hyg Assoc J 31:645-646.

Laskin S, Kuschner  M, Drew RT. 1970. Studies  in pulmonary carcino-
genesis. In:  Inhalation  carcinogenesis, Hanna,  Jr. M, Nettesheim P, Gil-
bert  J, eds. U.S. Atomic  Energy  Commission  Symposium  Series  18.
pp. 321-351.

Leach  LJ.  1963. A  laboratory  test  chamber  for  studying  airborne
materials.  AEC  Progress   Report  UR-629.  University  of  Rochester,
Rochester, NY. pp. 1-12.

Leach  LJ, Spiegl  CJ, Wilson RH, Sylvester GE, Lauterbach KE. 1959. A
multiple chamber  exposure unit designed for chronic inhalation.

Leong  KJ,  Dowd  GF, MacFarland HN.  1964. A new technique for tidal
volume measurement in unanesthetized small animals. Can J Physiol Phar-
maco!42:189-194.

Likens  SA, Mauderly  JL.  1979.  Respiratory  measurements in  small
laboratory   mammals:  a  literature  review.   Lovelace  Biomedical  and
Environmental Research Institute Publication LF-68. 13  pp.

Lippmann M.  1961.  A compact cascade impactor for field survey sam-
pling. Am Ind  Hyg Assoc J 22:348-353.

Lisco H. 1959. Autoradiographic and histopathologic studies in radiation
carcinogenesis  of the lung. Lab Invest 8:162-168.

Liu BYH, Whitby KT, Yu HYS. 1966. A condensation aerosol generator
for producing  monodisperse aerosols in  the size range, 0.036um to 1.3um.
J Recherches Atmos 3:397.

Lodge  JP.  1968. Production  of controlled test  atmospheres. In: Air pollu-
tion Vol. II. Analysis, monitoring, and surveying, Stern  AC, ed. Academic
Press, New York.  pp. 465-483.

Lund IE, Mills B, Moyes AJ.  1979.  Review of practical problems arising
from the relationship between  dust generators and dust collectors. J Inst
Energy 52:32-43.  (abstract only).

MacEwen JD. 1978. Nonconventional systems. In: Proceedings, workshop
on inhalation  chamber  toxicology,  Drew RT, ed.  Brookhaven  National
Laboratory Associated Universities, Inc. BNL 51318. pp.  9-16.

-------
                                235

MacFarland HN. 1975. Inhalation toxicology.  J Assoc Off Anal Chem
58:689-691.

MacFarland  HN.  1976.  Respiratory   toxicology.  Essays in  Toxicol
7:121-154.

MacFarland HN.  1981. A problem and a non-problem in chamber inhala-
tion studies. In: Proceedings  of  the inhalation toxicology and technology
symposium,  Kalamazoo, MI,  October  23-24,  1980.  Ann Arbor Science
Publishers, Inc., Ann Arbor, MI.  pp. 11-18.

Marple VA, Liu BYH,  Rubow KL.  1978. A dust generator for laboratory
use. Am Ind Hyg Assoc J 39:26-32.

Mauderly  JL.  1977.   Bronchopulmonary  lavage of  small  laboratory
animals. Lab Anim Sc. 27:255-261.

Mauderly  JL,  Tesarek JL.  1975. Nonrebreathing valve for  respiratory
measurements in unsedated small mammals.  J Appl Physiol 38:369-371.

May KR.  1949. An improved  spinning top homogeneous spray apparatus.
J Appl Phys 20:932-938.

May KR.  1966. Spinning-top  homogeneous  aerosol generator with shock-
proof mounting. J  Sci Instrum  43:841-842.

May KR.  1973.  The   collison nebulizer: description, performance,  and
application. Aerosol Sci 4:235.

Mazumder  MK, Kirsch KJ. 1977. Single particle aerodynamic relaxation
time analyzer. Res Sci Instrum 48:622-624.

Mazumder  MK,  Ware RE,  Wilson  JD,  Renninger RG, Hiller  FC,
McLeod PC, Raible  RW, Testerman  RG. 1979. SPART analyzer: its
application  to aerodynamic size  distribution measurement. J Aerosol Sci
10:561-569.

McCutcheon FH. 1951. The mammalian  breathing mechanism.  J  Cell
Comp Physiol 37:447-476.

McLaughlin RF,  Tyler WS, Canada RO. 196la.  A study of the subgross
pulmonary  anatomy in various mammals. Am J Anat 108:149-164.

McLaughlin RF,  Tyler WS, Canada RO. 1961b. Subgross pulmonary ana-
tomy in various mammals and man. J Am Med Assoc 175:694-697.

-------
                                 236
McLaughlin RF, Tyler WS, Canada RO.  1966. Subgross pulmonary ana-
tomy of the rabbit, rat, and guinea pig with additional notes on the human
lung. Am Rev Respir Dis 94:380-387.

Mead  J.   1960.  Control  of  respiratory  frequency.  J  Appl  Physiol
15:325-336.

Mercer  XT.  1973.  Production  and characterization  of  aerosols.  Arch
Intern Med 131:39-50.

Mercer  TT,  Goddard  RF,  Flores  RL.  1965,  Output  characteristics  of
several commercial nebulizers. Ann Allergy 23:314-326.

Mercer  TT,  Goddard  RF, Flores RL.  1968a.  Output characteristics  of
three ultrasonic nebulizers. Ann Allergy 26:18-27.

Mercer  TT,  Tillery  M,  Chow HY. 1968b. Operating characteristics  of
some compressed air nebulizers. Am Ind Hyg Assoc J 29:66-78.

Miller RR, Letts RL, Potts WJ, McKenna MJ.  1980. Improved methodol-
ogy for generating  controlled  test  atmospheres.  Am  Ind  Hyg Assoc  J
41:844-846.

Moorman WJ, Hornung RW, Wagner WD.  1977. Ventilatory functions in
germ-free and conventional rats exposed to coal dusts.  Proc Soc Exp Biol
Med 155:424-428.

Moorman WJ.  1978.  Exposure  of rodents to  particulate  carcinogens,
second presentation. In: Proceedings, workshop on inhalation chamber tox-
icology,  Drew  RT,  ed.  Brookhaven  National  Laboratory  Associated
Universities, Inc. BNL  51318. pp. 64-74.

Muir  DCF.  1965.  The production of monodisperse aerosols by a LaMer
Sinclair generator. Ann Occup Hyg 8:223-228.

Murphy  SD, Ulrich CE. 1964. Multi-animal test  system  for  measuring
effects of irritant gases and vapors on respiratory function of guinea pigs.
Am Ind Hyg Assoc J 25:28-36.

Nader JS. 1971.  Direct-reading instruments for analysing  airborne gases
and  vapors.  In: Air  sampling instruments,  4th  ed.,  Lippmann M, ed.
American Conference of Governmental Industrial Hygiene, p. J-l.

NAS. 1977. National Academy of Sciences. Inhalation exposure. In: Prin-
ciples and procedures for evaluating the  toxicity of household substances.
Washington, DC. pp. 60-73.

-------
                                 237

Nattie EE. 1977. Breathing patterns in the awake potassium-depleted rat.
J Appl Physiol: Respirat Environ Exercise Physiol 43:1063-1074.

Nelson GO, Taylor RD. 1980. An automated flow-temperature-humidity
control system. Am Ind Hyg Assoc J 41:769-771.

Nettesheim P, Griesemer RA.  1978. Experimental models for studies of
respiratory tract  carcinogenesis.  In: Pathogenesis and  therapy  of lung
cancer: lung biology in health  and disease, Vol. 10, Lenfant C, Harris C,
eds. Marcel Dekker, Inc., New  York. pp. 75-188.

Oberdoerster  G, Freedman AP.  1982.  Lung retention of  iron-59 labeled
magnetite in rats  comparing magnetometric and  radioactive measurement
techniques. Am Rev Respir Dis 125:165.

O'Keeffe AE, Ortman GC. 1966. Primary standards for trace gas  analysis.
Anal Chem 38:760-763.

OECD. 1979.  Organization for Economic Cooperation and Development.
Final  report  of the  OECD short-term  and long-term toxicity  groups.
unpublished. 185 pp.

Page N,  Sawhney DS, Ryon MG, eds. 1980. Proceedings of the workshop
on subchronic toxicity testing.  Denver,  Colorado, May 20-24,  1979. Oak
Ridge National Laboratory Publication ORNL/EIS-189. 63 pp.

Palecek F. 1969. Measurement of ventilatory mechanics in  the rat. J Appl
Physiol 27:149-156.

Palm  PE, McNerney  JM, Hatch T. 1956. Respiratory  dust retention in
small  animals, a comparison with man. Arch Ind Health 13:355-365.

Potts  WJ, Steiner EC. 1980. An apparatus for generation  of vapors from
liquids of low volatility for use in inhalation studies. Am  Ind Hyg Assoc J
41:141-145.

Poynter  D.  1977. Inhalation  toxicology: the questions  which  must be
answered. In:  Current approaches in toxicology. John Wright and Sons
Ltd., Bristol, pp. 86-104.

Prodi  V, De  Zaiacomo  T,  Melandri C,  Tarroni  G,  Formignani  M,
Hochrainer D.  1979. An inertial spectrometer for aerosol particles. J Aero-
sol Sci 14:411-419.

-------
                                 238
Raabe OG.  1972. Operating characteristics of two compressed air nebuliz-
ers used in inhalation experiments.  Fission Product  Inhalation Program
Annual Report (1971-1972). LF-45.

Raabe OG. 1976.  Generation of aerosols of fine particles.  In: Fine parti-
cles, Liu BYH, ed. Academic Press, New York. pp. 57-110.

Radford EP, Jr. 1957. Recent  studies of  mechanical properties of mam-
malian  lungs.  In: Tissue  elasticity, Remington JW, ed. American Physio-
logical Society, Washington, DC. pp.  177-190.

Rapaport E, Weinstock SE.  1955. A generator for homogeneous aerosols.
Experientia  11: 363-364.

Roe FJC. 1968.  Inhalation tests. In: Modern trends in toxicology, Vol. I,
Boyland  E,  Goulding  R,  eds.  Appleton-Century-Crofts,  New  York.
pp. 39-74.

Saltzmann BE. 1961. Preparation and analysis of calibrated low concen-
trations of sixteen toxic gases. Anal Chem 33:1100-1112.

Saffiotti U.  1969. Experimental respiratory tract carcinogenesis.  Progr Exp
Tumor Res  11:302-333.

Sanockij IV. 1970. Complete toxicometry. In: Methods for determining the
toxicity and hazards of chemicals. Moscow. Medicina:50-54.  (In Russian).

Sawyer KF, Walton WH. 1950. The "conifuge", a size separating sam-
pling device for airborne particles. J Sci Instrum 27:272-275.

Schreck  RM,  Chan TL, Soderholm SC.  1981. Design,  operation and
characterization of large volume exposure chambers.   In: Proceedings of
the inhalation toxicology and  technology  symposium,  Kalamazoo, MI,
October 23-24, 1980. Ann Arbor Science Publishers, Inc., Ann Arbor, MI.
pp. 29-52.

Sheehy JP, Archinger WC, Simon RA. 1967. Handbook of air pollution.
U.S. Department of Health, Education, and Welfare, pp. 9-10.

Sidorenko GI, Pinigin MA.  1976.  Concentration—time relationship for
various  regimens of inhalation of organic compounds. Environ Health Per-
spect 13:17-21.

Silver SD. 1946.  Constant flow gassing chambers: principles influencing
design  and operation. J Lab Clin Med 31:1153-1161.

-------
                                239
Sinclair D, LaMer V. 1949.  Light scattering as a measure of particle size
in aerosols. Chem Rev 44:245-267.

Smith SW, Spurling NW. 1974. Device for exposing the respiratory tract
of the rat to medicinal aerosols. Lab Practice 23(12):717-721.

Smith WB, Cavanaugh PR, Wilson RR. 1978. Technical manual, a survey
of equipment and methods for  particulate sampling in industrial process
streams. U.S. Environmental Protection Agency Report No. 600/7-78-043.
217 pp.

Smith DW, Ortiz LW, Archuleta RF, Spalding JF, Ettinger HJ.  1980.
Method for  chronic nose-only exposures of laboratory  animals to inhaled
fibrous aerosols.  Los Alamos National Laboratory,  NM. LA-UR-81  963;
CONF-8010202-1. 34 pp. (abstract only).

Spurney KR, Boose  C, Hockrainer  D, Monig FJ.  1976. A  note on the
dispersing of fibrous powders. Ann Occup Hyg 19:85-87.

Stokinger HE. 1953.  Toxicologic perspective in planning air pollution stu-
dies. Am J Public Health 43:742-751.

Stuart  BO, Willard  DH, Howard EB. 1970.  Uranium  mine contaminants
in dogs and hamsters.  In: Inhalation carcinogenesis, Hanna  Jr  M, Net-
tesheim P,  Gilbert J, eds. U.S. Atomic  Energy Commission Symposium
Series,  No.  18. pp. 413-428.

Swann HE,  Brunol D, Balchum OJ.  1965.  Pulmonary resistance  measure-
ments of guinea pigs. Arch Environ Health  10:24-32.

Thiede FC, Hackney JD, Linn WS, Spier C, House W. 1974. Animal
atmospheric exposure chamber  system using  a modified tank respirator.
Am Ind Hyg Assoc J 35:370-375.

Thomas RG, Lie R. 1963. Procedures and equipment used in  inhalation
studies on small animals. AEC Research and  Development Report LF-11.
30pp.

Thomas RG, Morgan  BN.  1969. The effects of age,  weight, strain  and
anesthesia upon the breathing patterns of rats. Matrix 69:13-18.

Tierney DG. 1974. Intermediary metabolism of the lung. Fed  Proc  Fed
Am Soc Exp Biol 33:2232-2237.

-------
                                 240
Tillery MI,  Wood GO, Ettinger HJ.  1976.  Generation  and characteriza-
tion  of  aerosols and vapors for inhalation  experiments. Environ Health
Perspect 16:25-40.

Timbrell V,  Hyett AW, Skidmore JW. 1968. A simple dispenser for gen-
erating  dust clouds  from standard reference samples  of  asbestos.  Ann
OccupHyg  11:273-281.

Timbrell V, Skidmore JW,  Hyett AW, Wagner JC.  1970. Exposure
chambers for inhalation experiments with standard reference  samples of
asbestos  of  the international  union against  cancer (UICC). Aerosol Sci
1:215-233.

Tropp RJ, Kuhn  PJ, Brock JR. 1980. A new method for  measuring the
particle  size  distribution of aerosols. Rev Sci Instrum 51(4):516-520.

Troy  WR.  1974.  Testing  for  inhalation toxicity. J  Soc Cosmet Chem
25:283-296.

Truog WE,  Standaert  TA. 1978. Nonrebreathing valve  for spontaneously
breathing or mechanically ventilated  animals.  J  Appl  Physiol: Respirat
Environ Exercise Physiol 44:974-976.

USDHEW.  1974. U.S. Department of Health, Education, and Welfare.
Guide for the care and use of laboratory animals, 4th ed. National Insti-
tutes of Health.  DHEW Publication  Number (NIH)74-23, Washington,
DC.  56  pp.

USEPA. 1982.  U.S Environmental Protection Agency. Health  effects test
guidelines. Washington, DC. EPA 560/6-82-001.

Van  Stee EW,  Moorman MP. 1978.  Monitoring for temperature, humi-
dity, concentration. In:  Proceedings, workshop on inhalation chamber toxi-
cology, Drew RT, ed. Brookhaven National Laboratory Associated Univer-
sities, Inc. BNL 51318. pp. 81-88.

Vinegar  A,  Stinett EE, Kosch PC, Miller ML. 1981. Pulmonary physiol-
ogy of the  ferret  and  its potential  as  a  model for inhalation  toxicology.
Teratology 24:17A. (abstract only).

Walton  WH, Prewett  WC. 1949. The production of  sprays and mists of
uniform  drop size  by means of spinning disc type  sprayers.  Proc Phys Soc
London  Sec  B 62(6):342-350.

Wehner  AP, Craig DK, Stuart BO. 1972. An aerosol exposure system for
chronic  inhalation studies with  rodents.  Am Ind Hyg Assoc J 33:483-487.

-------
                                 241
Weissman SH, Beethe RL, Redman HC. 1980. Ammonia concentrations
in an animal inhalation exposure chamber. Lab Animal Sci 30:974-980.

Wells AB.  1979. Inhalation testing of toiletries in animals.  Int J Cosmet
Sci 1:135-146.

Weston RE, Karel  L.  1946. An application of the dosimetric method for
biologically  assaying  inhaled  substances.   J  Pharmacol  Exp   Ther
88:195-207.

Whitby KT, Lundgren DA, Peterson CM. 1965. Homogenous aerosol gen-
erators. Int J Air Water Pollut 9:263-277.

WHO 1978. World Health Organization. Inhalation exposure. In: Princi-
ples  and  methods  for  evaluating  the  toxicity  of  chemicals.  Geneva.
pp. 199-235.

Willeke K,  Lo CSK, Whitby KT. 1974. Dispersion characteristics of a
fluidized bed. Aerosol Sci 5:449-455.

Wilson  IB,  LaMer VK.  1948. The retention of aerosol  particles in the
human respiratory tract as  a function of particle radius. J Ind Hyg Toxicol
30(5):265-280.

Witschi H. 1975. Exploitable biochemical approaches for the evaluation of
toxic lung damage. Essays Toxicol 6:125-191.

Wolff HS, Roach SA. 1961. The conicycle selective sampling system. In:
Inhaled particles and vapors, Davis CN, ed. Oxford: Pergamon Press.

Wong K, Alarie Y. 1982. A method for repeated evaluation of pulmonary
performance in unanesthetized, unrestrained  guinea  pigs  and its applica-
tion  to detect  effects  of  sulfuric acid mist inhalation. Toxicol Appl Phar-
macol 63:72-90.

Wright BM. 1950. A new dust feed  mechanism. J  Sci Instrum 27:12-15.

Wright BM. 1954.  A size selecting sampler for  airborne dust.  Br J Ind
Med 11:284-288.

Wright BM. 1957. Experimental studies on the relative importance of con-
centration  and  duration  of  exposure to  dust  inhalation.  Br J  Med
14:219-228.

-------
                                          243
                                    TECHNICAL REPORT DATA
                             (Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-560/6-84-004
                                                  o
                                                             3. RECIPIENT'S ACCESSION NO.
4. TITLE ANOSUBTITLE                                    /
Scientific Rationale  for  the Selection of Toxicity
Testing Methods II. Teratology,  Immunotoxicology,
and  Inhalation Toxicology
                                                             5 REPORT DATE
                                                              September 1985
                                                             6. PERFORMING ORGANIZATION CODE
 M.  G.  Ryon, D. S. Sawhney,  M.  L. Daugherty,  and
 R.  H.  Ross
                                                             3. PERFORMING ORGANIZATION REPORT NO.
                                                              ORNL-6094
9, PERFORMING ORGANIZATION NAME AND ADDRESS
 Information Research  and Analysis
 Information Resources Organization
 Oak Ridge National  Laboratory
 Oak Ridge, TM 37831
                                                             11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
 Office of Toxic  Substances
 U.S.  Environmental  Protection Agency
 Washington, DC 20460
                                                             13. TYPE OP REPORT AND PERIOD COVERED
                                                              Final
                                                             14. SPONSORING AG6NCY CODE

16. ABSTRACT
      This document  is  the second of a two-part  lite
 associated with  the various toxicity testing  method
 pathology, etc.).   Acute, subchronic, chronic,  and
 are covered in ORNL/EIS-151.  Testing methods for d
 immuno toxicology, and  inhalation toxicology and res
 these areas are  covered in this volume, ORNL-6094.
 for the purpose  of  assisting and supporting the U.S
 Agency in its efforts  to develop guidelines for mor
 testing procedures.
                                                      rature analysis of  parameters
                                                        (test animal selection,
                                                      carcinogenic testing methods
                                                      evelopmental toxicity,
                                                      earch needs associated  with
                                                       These reports were prepared
                                                        Environmental Protection
                                                        efficient and economical
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                               b. IDENTIFIERS/OPEN ENDED TEH MS
                                                                             COSATI [ icld/Group
 Reviews
 Toxicity
 Tests
                                                Toxicity Testing
                                                Methodology
                                                                            06
 8. DISTRIBUTION STATEMENT

  Release unlimited
                                                 Unclassified
                                                                            262
                                               20 SECURITY CLASS (Thu page}
                                                 Unclassified
EPA Form 2220-1 (R*
                                   •irll.S. GOVERNMENT PRINTING OFFICE: 1985-631-056/20177

-------