INTERNATIONAL PROGRAMME ON CHEMICAL SAFETY
ENVIRONMENTAL HEALTH CRITERIA 180
Principles and Methods for Assessing Direct Immunotoxicity
Associated with Exposure to Chemicals
This report contains the collective views of an international group of
experts and does not necessarily represent the decisions or the stated
policy of the United Nations Environment Programme, the International
Labour Organisation, or the World Health Organization.
First draft prepared at the National Institute of Health Sciences,
Tokyo, Japan, and the Institute of Terrestrial Ecology, Monk's Wood,
United Kingdom
Published under the joint sponsorship of the United Nations
Environment Programme, the International Labour Organisation, and the
World Health Organization
World Health Organization
Geneva, 1996
The International Programme on Chemical Safety (IPCS) is a joint
venture of the United Nations Environment Programme, the International
Labour Organisation, and the World Health Organization. The main
objective of the IPCS is to carry out and disseminate evaluations of
the effects of chemicals on human health and the quality of the
environment. Supporting activities include the development of
epidemiological, experimental laboratory, and risk-assessment methods
that could produce internationally comparable results, and the
development of manpower in the field of toxicology. Other activities
carried out by the IPCS include the development of know-how for coping
with chemical accidents, coordination of laboratory testing and
epidemiological studies, and promotion of research on the mechanisms
of the biological action of chemicals.
WHO Library Cataloguing in Publication Data
Principles and methods for assessing direct immunotoxicity
associated with exposure to chamicals
(Environmental health criteria ; 180)
1.Immunotoxins 2.Immune system 3.Risk assessment I.Series
ISBN 92 4 157180 2 (NLM Classification: QW 630.5.13)
ISSN 0250-863X
The World Health Organization welcomes requests for permission to
reproduce or translate its publications, in part or in full.
Applications and enquiries should be addressed to the Office of
Publications, World Health Organization, Geneva, Switzerland, which
will be glad to provide the latest information on any changes made to
the text, plans for new editions, and reprints and translations
already available.
(c) World Health Organization 1996
Publications of the World Health Organization enjoy copyright
protection in accordance with the provisions of Protocol 2 of the
Universal Copyright Convention. All rights reserved. The designations
employed and the presentation of the material in this publication do
not imply the expression of any opinion whatsoever on the part of the
Secretariat of the World Health Organization concerning the legal
status of any country, territory, city or area or of its authorities,
or concerning the delimitation of its frontiers or boundaries. The
mention of specific companies or of certain manufacturers' products
does not imply that they are endorsed or recommended by the World
Health Organization in preference to others of a similar nature that
are not mentioned. Errors and omissions excepted, the names of
proprietary products are distinguished by initial capital letters.
CONTENTS
NOTE TO READERS OF THE CRITERIA MONOGRAPHS
PREAMBLE
WHO TASK GROUP MEETING ON PRINCIPLES AND METHODS FOR ASSESSING DIRECT
IMMUNOTOXICITY ASSOCIATED WITH EXPOSURE TO CHEMICALS
PRINCIPLES AND METHODS
ABBREVIATIONS
SUMMARY AND RECOMMENDATIONS
1. INTRODUCTION TO IMMUNOTOXICOLOGY
1.1. Historical overview
1.2. The immune system; functions, system regulation, and
modifying factors; histophysiology of lymphoid organs
1.2.1. Function of the immune system
1.2.1.1 Encounter and recognition
1.2.1.2 Specificity
1.2.1.3 Choice of effector reaction; diversity
of the answer
1.2.1.4 Immunoregulation
1.2.1.5 Modifying factors outside the immune
system
1.2.1.6 Immunological memory
1.2.2. Histophysiology of lymphoid organs
1.2.2.1 Overview: structure of the immune system
1.2.2.2 Bone marrow
1.2.2.3 Thymus
1.2.2.4 Lymph nodes
1.2.2.5 Spleen
1.2.2.6 Mucosa-associated lymphoid tissue
1.2.2.7 Skin immune system or skin-associated
lymphoid tissue
1.3. Pathophysiology
1.3.1. Susceptibility to toxic action
1.3.2. Regeneration
1.3.3. Changes in lymphoid organs
2. HEALTH IMPACT OF SELECTED IMMUNOTOXIC AGENTS
2.1. Description of consequences on human health
2.1.1. Consequences of immunosuppression
2.1.1.1 Cancer
2.1.1.2 Infectious diseases
2.1.2. Consequences of immunostimulation
2.2. Direct immunotoxicity in laboratory animals
2.2.1. Azathioprine and cyclosporin A
2.2.1.1 Azathioprine
2.2.1.2 Cyclosporin A
2.2.2. Halogenated hydrocarbons
2.2.2.1
2,3,7,8-Tetrachlorodibenzo- para-dioxin
2.2.2.2 Polychlorinated biphenyls
2.2.2.3 Hexachlorobenzene
2.2.3. Pesticides
2.2.3.1 Organochlorine pesticides
2.2.3.2 Organophosphate compounds
2.2.3.3 Pyrethroids
2.2.3.4 Carbamates
2.2.3.5 Dinocap
2.2.4. Polycyclic aromatic hydrocarbons
2.2.5. Solvents
2.2.5.1 Benzene
2.2.5.2 Other solvents
2.2.6. Metals
2.2.6.1 Cadmium
2.2.6.2 Lead
2.2.6.3 Mercury
2.2.6.4 Organotins
2.2.6.5 Gallium arsenide
2.2.6.6 Beryllium
2.2.7. Air pollutants
2.2.8. Mycotoxins
2.2.9. Particles
2.2.9.1 Asbestos
2.2.9.2 Silica
2.2.10. Substances of abuse
2.2.11. Ultraviolet B radiation
2.2.12. Food additives
2.3. Immunotoxicity of environmental chemicals in wildlife and
domesticated species
2.3.1. Fish and other marine species
2.3.1.1 Fish
2.3.1.2 Marine mammals
2.3.2. Cattle and swine
2.3.3. Chickens
2.4. Immunotoxicity of environmental chemicals in humans
2.4.1. Case reports
2.4.2. Air pollutants
2.4.3. Pesticides
2.4.4. Halogenated aromatic hydrocarbons
2.4.5. Metals
2.4.6. Solvents
2.4.7. Ultraviolet radiation
2.4.8. Others
3. STRATEGIES FOR TESTING THE IMMUNOTOXICITY OF CHEMICALS IN ANIMALS
3.1. General testing of the toxicity of chemicals
3.2. Organization of tests in tiers
3.2.1. US National Toxicology Program panel
3.2.2. Dutch National Institute of Public Health and
Environmental Protection panel
3.2.3. US Environmental Protection Agency, Office of
Pesticides panel
3.2.4. US Food and Drug Administration, Center for Food
Safety and Applied Nutrition panel
3.3. Considerations in evaluating systemic and local
immunotoxicity
3.3.1. Species selection
3.3.2. Systemic immunosuppression
3.3.3. Local suppression
4. METHODS OF IMMUNOTOXICOLOGY IN EXPERIMENTAL ANIMALS
4.1. Nonfunctional tests
4.1.1. Organ weights
4.1.2. Pathology
4.1.3. Basal immunoglobulin level
4.1.4. Bone marrow
4.1.5. Enumeration of leukocytes in bronchoalveolar lavage
fluid, peritoneal cavity, and skin
4.1.6. Flow cytometric analysis
4.2. Functional tests
4.2.1. Macrophage activity
4.2.2. Natural killer activity
4.2.3. Antigen-specific antibody responses
4.2.4. Antibody responses to sheep red blood cells
4.2.4.1 Spleen immunoglobulin M and
immunoglobulin G plaque-forming cell
assay to the T-dependent antigen, sheep
red blood cells
4.2.4.2 Enzyme-linked immunosorbent assay of
anti-sheep red blood cell antibodies of
classes M, G, and A in rats
4.2.5. Responsiveness to B-cell mitogens
4.2.6. Responsiveness to T-cell mitogens
4.2.7. Mixed lymphocyte reaction
4.2.8. Cytotoxic T lymphocyte assay
4.2.9. Delayed-type hypersensitivity responses
4.2.10. Host resistance models
4.2.10.1 Listeria monocytogenes
4.2.10.2 Streptococcus infectivity models
4.2.10.3 Viral infection model with mouse and rat
cytomegalovirus
4.2.10.4 Influenza virus model
4.2.10.5 Parasitic infection model with
Trichinella spiralis
4.2.10.6 Plasmodium model
4.2.10.7 B16F10 Melanoma model
4.2.10.8 PYB6 Carcinoma model
4.2.10.9 MADB106 Adenocarcinoma model
4.2.11. Autoimmune models
4.3. Assessment of immunotoxicity in non-rodent species
4.3.1. Non-human primates
4.3.2. Dogs
4.3.3. Non-mammalian species
4.3.3.1 Fish
4.3.3.2 Chickens
4.4. Approaches to assessing immunosuppression in vitro
4.5. Future directions
4.5.1. Molecular approaches in immunotoxicology
4.5.2. Transgenic mice
4.5.3. Severe combined immunodeficient mice
4.6. Biomarkers in epidemiological studies and monitoring
4.7. Quality assurance for immunotoxicology studies
4.8. Validation
5. ESSENTIALS OF IMMUNOTOXICITY ASSESSMENT IN HUMANS
5.1. Introduction: Immunocompetence and immunosuppression
5.2. Considerations in assessing human immune status related to
immunotoxicity
5.3. Confounding variables
5.4. Considerations in the design of epidemiological studies
5.5. Proposed testing regimen
5.6. Assays for assessing immune status
5.6.1. Total blood count and differential
5.6.2. Tests of the antibody-mediated immune system
5.6.2.1 Immunoglobulin concentration
5.6.2.2 Specific antibodies
5.6.3. Tests for inflammation and autoantibodies
5.6.3.1 C-Reactive protein
5.6.3.2 Antinuclear antibody
5.6.3.3 Rheumatoid factor
5.6.3.4 Thyroglobulin antibody
5.6.4. Tests for cellular immunity
5.6.4.1 Flow cytometry
5.6.4.2 Delayed-type hypersensitivity
5.6.4.3 Proliferation of mononuclear cells in
vitro
5.6.5. Tests for nonspecific immunity
5.6.5.1 Natural killer cells
5.6.5.2 Polymorphonuclear granulocytes
5.6.5.3 Complement
5.6.6. Clinical chemistry
5.6.7. Additional confirmatory tests
6. RISK ASSESSMENT
6.1. Introduction
6.2. Complements to extrapolating experimental data
6.2.1. In-vitro approaches
6.2.2. Parallellograms
6.2.3. Severe combined immunodeficient mice
6.3. Host resistance and clinical disease
7. SOME TERMS USED IN IMMUNOTOXICOLOGY
REFERENCES
RESUME
RESUMEN
NOTE TO READERS OF THE CRITERIA MONOGRAPHS
Every effort has been made to present information in the Criteria
monographs as accurately as possible without unduly delaying their
publication. In the interest of all users of the Environmental Health
Criteria monographs, readers are requested to communicate any errors
that may have occurred to the Director of the International Programme
on Chemical Safety, World Health Organization, Geneva, Switzerland, in
order that they may be included in corrigenda.
* * *
A detailed data profile and a legal file can be obtained from the
International Register of Potentially Toxic Chemicals, Case postale
356, 1219 Châtelaine, Geneva, Switzerland (Telephone No. 979 9111).
* * *
Funding and support for the preparation and finalization of this
monograph were provided by the United States Environmental Protection
Agency under Cooperative Agreement with the World Health Organization
No. CR 821767-01-0, by the German Federal Ministry for the
Environment, Nature Conservation and Nuclear Safety, and by the
Netherlands National Institute for Public Health and Environmental
Protection.
Environmental Health Criteria
PREAMBLE
Objectives
The WHO Environmental Health Criteria Programme was initiated in
1973, with the following objectives:
(i) to assess information on the relationship between exposure to
environmental pollutants and human health and to provide
guidelines for setting exposure limits;
(ii) to identify new or potential pollutants;
(iii) to identify gaps in knowledge concerning the health effects of
pollutants;
(iv) to promote the harmonization of toxicological and
epidemiological methods in order to have internationally
comparable results.
The first Environmental Health Criteria (EHC) monograph, on
mercury, was published in 1976; numerous assessments of chemicals and
of physical effects have since been produced. Many EHC monographs have
been devoted to toxicological methods, e.g. for genetic, neurotoxic,
teratogenic, and nephrotoxic effects. Other publications have been
concerned with e.g. epidemiological guidelines, evaluation of short-
term tests for carcinogens, biomarkers, and effects on the elderly.
Since the time of its inauguration, the EHC Programme has widened
its scope, and the importance of environmental effects has been
increasingly emphasized in the total evaluation of chemicals, in
addition to their health effects.
The original impetus for the Programme came from resolutions of
the World Health Assembly and the recommendations of the 1972 United
Nations Conference on the Human Environment. Subsequently, the work
became an integral part of the International Programme on Chemical
Safety (IPCS), a cooperative programme of UNEP, ILO, and WHO. In this
manner, with the strong support of the new partners, the importance of
occupational health and environmental effects was fully recognized.
The EHC monographs have become widely established, used, and
recognized throughout the world.
The recommendations of the 1992 United Nations Conference on
Environment and Development and the subsequent establishment of the
Intergovernmental Forum on Chemical Safety, with priorities for action
in the six programme areas of Chapter 19, Agenda 21, lend further
weight to the need for EHC assessments of the risks of chemicals.
The Criteria monographs are intended to provide critical reviews
of the effect on human health and the environment of chemicals,
combinations of chemicals, and physical and biological agents. They
include reviews of studies that are of direct relevance for the
evaluation and do not describe every study that has been carried out.
Data obtained worldwide are used, and results are quoted from original
studies, not from abstracts or reviews. Both published and unpublished
reports are considered, and the authors are responsible for assessing
all of the articles cited; however, preference is always given to
published data, and unpublished data are used only when relevant
published data are absent or when the unpublished data are pivotal to
the risk assessment. A detailed policy statement is available that
describes the procedures used for citing unpublished proprietary data,
so that this information can be used in the evaluation without
compromising its confidential nature (WHO, 1990).
In the evaluation of human health risks, sound data on humans,
whenever available, are preferred to data on experimental animals.
Studies of animals and in-vitro systems provide support and are used
mainly to supply evidence missing from human studies. It is mandatory
that research on human subjects be conducted in full accord with
ethical principles, including the provisions of the Helsinki
Declaration.
The EHC monographs are intended to assist national and
international authorities in making risk assessments and subsequent
risk management decisions. They represent a thorough evaluation of
risks and are not in any sense recommendations for regulation or
setting standards. The latter are the exclusive purview of national
and regional governments.
Content
The layout of EHC monographs for chemicals is outlined below.
* Summary: a review of the salient facts and the risk evaluation of
the chemical
* Identity: physical and chemical properties, analytical methods
* Sources of exposure
* Environmental transport, distribution, and transformation
* Environmental levels and human exposure
* Kinetics and metabolism in laboratory animals and humans
* Effects on laboratory mammals and in-vitro test systems
* Effects on humans
* Effects on other organisms in the laboratory and the field
* Evaluation of human health risks and effects on the environment
* Conclusions and recommendations for protection of human health
and the environment
* Further research
* Previous evaluations by international bodies, e.g. the
International Agency for Research on Cancer, the Joint FAO/WHO
Expert Committee on Food Additives, and the Joint FAO/WHO Meeting
on Pesticide Residues
Selection of chemicals
Since the inception of the EHC Programme, the IPCS has organized
meetings of scientists to establish lists of chemicals that are of
priority for subsequent evaluation. Such meetings have been held in
Ispra, Italy (1980); Oxford, United Kingdom (1984); Berlin, Germany
(1987); and North Carolina, United States of America (1995). The
selection of chemicals is based on the following criteria: the
existence of scientific evidence that the substance presents a hazard
to human health and/or the environment; the existence of evidence that
the possible use, persistence, accumulation, or degradation of the
substance involves significant human or environmental exposure; the
existence of evidence that the populations at risk (both human and
other species) and the risks for the environment are of a significant
size and nature; there is international concern, i.e. the substance is
of major interest to several countries; adequate data are available on
the hazards.
If it is proposed to write an EHC monograph on a chemical that is
not on the list of priorities, the IPCS Secretariat first consults
with the cooperating organizations and the participating institutions.
Procedures
The order of procedures that result in the publication of an EHC
monograph is shown in the following flow chart. A designated staff
member of IPCS, responsible for the scientific quality of the
document, serves as Responsible Officer (RO). The IPCS Editor is
responsible for the layout and language. The first draft, prepared by
consultants or, more usually, staff at an IPCS participating
institution is based initially on data provided from the International
Register of Potentially Toxic Chemicals and reference data bases such
as Medline and Toxline.
The draft document, when received by the RO, may require an
initial review by a small panel of experts to determine its scientific
quality and objectivity. Once the RO finds the first draft acceptable,
it is distributed in its unedited form to over 150 EHC contact points
throughout the world for comment on its completeness and accuracy and,
where necessary, to provide additional material. The contact points,
usually designated by governments, may be participating institutions,
IPCS focal points, or individual scientists known for their particular
expertise. Generally, about four months are allowed before the
comments are considered by the RO and author(s). A second draft
incorporating the comments received and approved by the Director,
IPCS, is then distributed to Task Group members, who carry out a peer
review at least six weeks before their meeting.
The Task Group members serve as individual scientists, not as
representatives of any organization, government, or industry. Their
function is to evaluate the accuracy, significance, and relevance of
the information in the document and to assess the risks to health and
the environment from exposure to the chemical. A summary and
recommendations for further research and improved safety are also
drawn up. The composition of the Task Group is dictated by the range
of expertise required for the subject of the meeting and by the need
for a balanced geographical distribution.
The three cooperating organizations of the IPCS recognize the
important role played by nongovernmental organizations, so that
representatives from relevant national and international associations
may be invited to join the Task Group as observers. While observers
may provide valuable contributions to the process, they can speak only
at the invitation of the Chairperson. Observers do not participate in
the final evaluation of the chemical, which is the sole responsibility
of the Task Group members. The Task Group may meet in camera when it
considers that to be appropriate.
All individuals who participate in the preparation of an EHC
monograph as authors, consultants, or advisers must, in addition to
serving in their personal capacity as scientists, inform the RO if at
any time a conflict of interest, whether actual or potential, could be
perceived in their work. They are required to sign a statement to that
effect. This procedure ensures the transparency and probity of the
process.
When the Task Group has completed its review and the RO is
satisfied as to the scientific correctness and completeness of the
document, it is edited for language, the references are checked, and
camera-ready copy is prepared. After approval by the Director, IPCS,
the monograph is submitted to the WHO Office of Publications for
printing. At this time, a copy of the final draft is also sent to the
Chairperson and Rapporteur of the Task Group to check for any errors.
It is accepted that the following criteria should initiate the
updating of an EHC monograph: new data are available that would
substantially change the evaluation; there is public concern about
health or environmental effects of the agent because of greater
exposure; an appreciable time has elapsed since the last evaluation.
All participating institutions are informed, through the EHC
progress report, of the authors and institutions proposed for the
drafting of the documents. A comprehensive file of all comments
received on drafts of each EHC monograph is maintained and is
available on request. The chairpersons of task groups are briefed
before each meeting on their role and responsibility in ensuring that
these rules are followed.
WHO TASK GROUP MEETING ON PRINCIPLES AND METHODS FOR ASSESSING DIRECT
IMMUNOTOXICITY ASSOCIATED WITH EXPOSURE TO CHEMICALS
Members
Dr A. Emmendörffer, Department of Immunobiology, Fraunhofer Institute
of Toxicology & Aerosol Research, Hanover, Germany
Dr H.S. Koren, Health Effects Research Laboratory, US Environmental
Protection Agency, Chapel Hill, NC, USA (Vice-Chairman)
Dr R.W. Luebke, Immunotoxicology Branch, Health Effects Research
Laboratory, US Environmental Protection Agency, Research Triangle
Park, NC, USA (Joint Rapporteur)
Dr M. Luster, National Institute of Environmental Health Sciences,
Research Triangle Park, NC, USA
Dr C. Madsen, Institute of Toxicology, National Food Agency of
Denmark, Ministry of Health, Soborg, Denmark
Dr P. Ross, Dalhousie University, Halifax, Nova Scotia, Canada
(c/o Seal Rehabilitation and Research Centre, Pieterburen,
Netherlands)
Dr H.J. Schuurman, Preclinical Research/Immunology, Sandoz Pharma Ltd,
Basel, Switzerland
Dr H. Van Loveren, Laboratory for Pathology, National Institute of
Public Health and Environmental Protection, Bilthoven, Netherlands
(Joint Rapporteur)
Dr J.G. Vos, National Institute of Public Health and Environmental
Protection, Bilthoven, Netherlands (Chairman)
Dr K.L. White, Jr, Immunotoxicology Group, Medical College of
Virginia, Virginia Commonwealth University, Richmond, VA, USA
Observers
IUTOX
Dr P. Montuschi, Department of Pharmacology, Catholic University of
the Sacred Heart, Rome, Italy
ECETOC
Dr R.W.R. Crevel, Environmental Safety Laboratory, Unilever Research
and Engineering, Sharnbrook, Bedfordshire, United Kingdom
Secretariat
Dr J.H. Dean, Sanofi Winthrop, Inc., Sanofi Research Division,
Collegeville, PA, USA
Mr V. Quarg, Federal Ministry for Environment, Nature Conservation &
Nuclear Safety, Bonn, Germany
Dr E. Smith, International Programme on Chemical Safety, World Health
Organization, Geneva, Switzerland
ENVIRONMENTAL HEALTH CRITERIA PRINCIPLES AND METHODS FOR ASSESSING
DIRECT IMMUNOTOXICITY ASSOCIATED WITH EXPOSURE TO CHEMICALS
A WHO Task Group on Principles and Methods for Assessing Direct
Immunotoxicity Associated with Exposure to Chemicals met at the World
Health Organization, Geneva, from 10 to 14 October 1994. Dr E. Smith,
IPCS, welcomed the participants on behalf of Dr M. Mercier, Director
IPCS, and the cooperating organizations. The Task Group reviewed and
revised the draft monograph and prepared the final text.
The first draft of the monograph was prepared by a group of
authors (listed below) under the coordination of Dr J.G. Vos and
Dr H. Van Loveren of the Dutch National Institute for Public Health
and Environmental Protection (RIVM), an IPCS Collaborating Centre for
Immunotoxicology and Allergic Hypersensitization. The second draft,
incorporating comments received after international circulation to
national experts of the first draft to IPCS contact points for
Environmental Health Criteria monographs, was prepared by Dr J.G.
Vos and Dr H. Van Loveren of the Netherlands and Dr Kimber White, USA.
Dr E. Smith of the IPCS Unit for the Assessment of Risk and
Methods was responsible for the scientific content of the monograph
and Mrs E. Heseltine for the editing.
The efforts of all who helped in the preparation and finalization
of the monograph are gratefully acknowledged.
The contributing authors were:
Dr J.H. Dean, Collegeville, PA, USA
Professor J. Descotes, Lyon, France
Dr F. Kuper, Zeist, Netherlands
Dr M. Luster, Research Triangle Park, NC, USA
Dr P.S. Ross, Bilthoven, Netherlands
Dr H.J. Schuurman, Basel, Switzerland
Dr M.J. Selgrade, Research Triangle Park, NC, USA
Dr R.L. de Swart, Bilthoven, Netherlands
Dr H. Van Loveren, Bilthoven, Netherlands
Professor J.G. Vos, Bilthoven, Netherlands
Dr P.W. Wester, Bilthoven, Netherlands
Professor A.G. Zapata, Madrid, Spain
ABBREVIATIONS
ACTH adrenocorticotrophic hormone
Ah aromatic hydrocarbon
AIDS acquired immunodeficiency syndrome
B bursa-dependent
CALLA common acute lymphoblastic leukaemia antigen
CD cluster of differentiation
CEC Commission of the European Communities
CH50 haemolytic complement
CML cell-mediated lympholysis
DMBA 7,12-dimethylbenz[ a]anthracene
DNCB dinitrochlorobenzene
ELISA enzyme-linked immunosorbent assay
EPO erythrocyte lineage differentiation factor
FACS fluorescence activated cell sorter
GALT gut-associated lymphoid tissue
G-CSF granulocyte colony-stimulating factor
GM-CSF granulocyte-macrophage colony-stimulating factor
GVH graft-versus-host
HCB hexaclorobenzene
HEV high endothelial venule
HIV human immunodeficiency virus
HPCA human progenitor cell antigen
HSA heat-stable antigen
ICAM intercellular adhesion molecule
IFN interferon
Ig immunoglobulin
IL interleukin
IPCS International Programme on Chemical Safety
LFA lymphocyte function-related antigen
LIF leukaemia inhibitory factor
LOAEL lowest-observed-adverse-effect level
LOEL lowest-observed-effect level
M microfold
MALT mucosa-associated lymphoid tissue
MARE monoclonal anti-rat immunoglobulin E
MARK monoclonal antibody anti-kappa
M-CSF macrophage colony-stimulating factor
MED minimal erythemal dose
MHC major histocompatibility complex
NCAM neural cell adhesion molecule
NK natural killer
NOAEL no-observed-adverse-effect level
NOEL no-observed-effect level
NTP National Toxicology Program
PAH polycyclic aromatic hydrocarbon
PCB polychlorinated biphenyl
PG prostaglandin
QCA quiescent cell antigen
RIVM Dutch National Institute of Public Health and
Environmental Protection
S9 9000 x g supernatant
SCF stem-cell factor
SCID severe combined immunodeficiency
SIS skin immune system
STM Salmonella typhimurium mitogen
TBTO tri- n-butyltin oxide
Tc cytotoxic T cell
TCDD 2,3,7,8-tetrachlorodibenzo- para-dioxin
TCR T-cell receptor
Tdth delayed-type hypersensitivity T cell
TGF transforming growth factor
Th T helper-inducer cell
THAM T-cell activation molecule
THI 2-acetyl-4(5)-tetrahydroxybutylimidazole
O,O,S-TMP O,O,S-trimethylphosphorothiate
TNF tumour necrosis factor
UVB ultraviolet B
UVR ultraviolet radiation
VCAM vascular cell adhesion molecule
VLA very late antigen
SUMMARY
1. The immune system has evolved to counter challenges to the
integrity of self from either microorganisms or cells that have
escaped the organism's control mechanisms. Recognition that
xenobiotics can impair the function of the immune system has led to
progress in immunotoxicology over the last two decades. Experimental
approaches (mainly in rodent species) have been developed and
validated in multilaboratory studies. In this monograph, the function
and histophysiology of the immune system are reviewed, and the
information necessary to understand and interpret the pathological
changes caused by immunotoxic insults is provided. Emphasis is laid on
the immune systems of humans and rodent species, but reference is made
to other species, including fish, that have been the object of
immunotoxicological studies. The pathophysiology of the immune system,
including the variable susceptibility of its components, alterations
to the lymphoid organs, and the reversibility of changes are important
for understanding the impact of immunotoxicity.
2. Immunosuppression and immunostimulation both have clinical
consequences. Immunodeficiency states and severe immunosuppression,
such as can occur during transplantion and cytostatic therapy, have
both been associated with increased incidences of infectious diseases
(particularly opportunistic ones) and cancer. Exposure to immunotoxic
chemicals in the environment, however, may be expected to result in
more subtle forms of immunosuppression which may be difficult to
detect, leading to increased incidences of infections such as
influenza and the common cold. Studies of experimental animals and
humans have shown that many environmental chemicals suppress the
immune response. Immunotoxic xenobiotics are not restricted to a
particular chemical class. Compounds that adversely affect the immune
system are found among drugs, pesticides, solvents, halogenated and
aromatic hydrocarbons, and metals; ultraviolet radiation can also be
immunotoxic. Therapeutic administration of immunostimulating agents
can have adverse effects, and a few environmental chemicals that have
immunostimulating properties (beryllium, silica, hexachlorobenzene)
can have clinical consequences.
3. The complexity of the immune system results in multiple potential
target sites and pathological sequelae. The initial strategies devised
by immunotoxicologists working in toxicology and safety assessment
were to select and apply a tiered panel of assays to identify
immunosuppressive and immunostimulatory agents in laboratory animals.
Although the configuration of these testing panels may vary depending
on which agency or laboratory is conducting the test and on the animal
species employed, they all include measurement of one or more of the
following: altered lymphoid organ weights and histology; changes in
the cellularity of lymphoid tissue, peripheral blood leukocytes,
and/or bone marrow; impairment of cell function at the effector or
regulatory level; and altered susceptibility to challenge with
infectious agents or tumour cells.
The original test guideline No. 407 of the Organisation for
Economic Co-operation and Development, published in 1981, was not
designed to detect potential immunotoxicity, and modifications have
been proposed to make the guideline more useful for identifying
immunotoxicants. Tiered testing systems have been designed for more
extensive investigation of potential immunotoxicity, by the US
National Toxicology Program, the Dutch National Institute of Public
Health and Environmental Protection, the US Environmental Protection
Agency Office of Pesticides, and the US Food and Drug Administration
Center for Food Safety and Applied Nutrition.
Studies have been conducted in mice, and to a lesser extent in
rats, to investigate the specificity, precision (reproducibility),
sensitivity, accuracy, and relevance for the assessment of risk to
human health of a variety of measures of immune status. International,
interlaboratory validations of methods have been carried out within
the International Collaborative Immunotoxicity Study of IPCS and the
European Union, the Bundesinstitut für Gesundheitlichen
Verbraucherschutz, und Veterinärmedizin, and in studies of cyclosporin
A in Fischer 344 rats.
4. The tests used in the tiered testing schemes are described in
Section 3, which indicates the rationale for their selection and the
complexities involved in their performance. Although these protocols
were designed for studies of rats and mice, some have been applied
successfully for studying immunotoxicity in other animal species,
including non-human primates, marine mammals, dogs, birds, and fish.
A variety of factors must be considered in evaluating the
potential of an environmental agent or drug to influence the immune
system of experimental animals adversely. These include selection of
the appropriate animal models and exposure variables, inclusion of
general toxicological parameters, an understanding of the biological
relevance of the end-points being measured, use of validated measures,
and quality assurance. The experimental conditions should take into
account the potential route and level of human exposure and any
available information on toxicodynamics and toxicokinetics. The doses
and sample sizes should be selected so as to generate clear dose-
response curves, in addition to no-observed-adverse-effect or
no-observed-effect levels. The strategies are continually refined to
allow better prediction of conditions that may lead to disease. In
addition, techniques should be developed that would help to identify
mechanisms of action; these might include methods in vitro,
examination of local immune responses (such as in the skin, lung, and
intestines), and use of the techniques of molecular biology and
genetically modified animals.
5. The detection of immune changes after exposure to potentially
immunotoxic compounds is more complicated in humans than in
experimental animals. The testing possibilities are limited, levels of
exposure to the agent (i.e. dose) are difficult to establish, and the
immune status of populations is extremely heterogeneous. Age, race,
gender, pregnancy, acute stress and the ability to cope with stress,
coexistent disease and infections, nutritional status, tobacco smoke,
and some medications contribute to this heterogeneity.
An important factor in assessing the usefulness of a particular
study for risk assessment is epidemiological study design. The
commonest design used in immunotoxicity is the cross-sectional study,
in which exposure status and disease status are measured at one time
or over a short period. The immune function of 'exposed' subjects is
then compared with that of a comparable group of 'unexposed'
individuals. There are possible pitfalls in this study design.
Because many of the immune changes seen in humans after exposure
to a chemical may be sporadic and subtle, recently exposed populations
must be studied and sensitive tests be used for assessing the immune
system. Conclusions about immunotoxic effects should be based on
changes not in a single parameter but in the immune profile of an
individual or population.
Most of the tests for specific immunity (cell-mediated and
humoral), nonspecific immunity and inflammation were developed to
detect immune alterations in patients with immunodeficiency disease
and are not always adequate to detect subtle alterations induced by
environmental chemicals. IPCS, the Centers for Disease Control, and
the US National Academy of Sciences have each described procedures for
evaluating changes in the human immune system resulting from exposure
to immunotoxicants, but the tests described require evaluation for
this purpose.
6. Risk assessment is a process in which relevant data on the
biological effects, dose-response relationships, and exposure for a
particular agent are analysed in an attempt to establish qualitative
and quantitative estimates of adverse outcomes. Typically, risk
assessment comprises four major steps: hazard identification, dose-
response assessment, exposure assessment, and risk characterization.
Up until now, immunotoxicology has focused mainly on hazard
identification, and to some extent on dose-response assessment, and
very few studies have included exposure assessment or risk
characterization.
As in other areas of toxicology, uncertainties exist which may
affect the interpretation of data on immunotoxicity with regard to
human health risk. The two most problematic issues -- extrapolating
effects from individual cells to a whole organ or beyond and
extrapolating data from experimental animals to humans -- are common
to most non-cancer end-points. The first issue is due to uncertainties
associated with establishing a quantitative relationship between
changes in individual immune function and altered resistance to
infections and neoplastic disease. The second issue is due to
uncertainties associated with assessing risk to human health on the
basis of studies in laboratory animals.
The ultimate purpose of risk assessment is to protect human
health and the environment. Suitable model systems must therefore be
chosen. The toxicokinetics of the test material and the nature and
magnitude of the immune response generated in the model should be
comparable to that of humans.
Conventionally, empirical uncertainty factors are used in risk
assessment to derive an acceptable exposure limit from experimental
results. This approach does not take into account the functional
reserve or redundancy of the immune system. A more recent development
in risk assessment is use of in-vitro models as an adjunct to studies
of experimental animals. The advantages of this approach are that it
improves the accuracy of extrapolation of data from animals to man and
minimizes the use of animals; it also bridges the gap between those
data, particularly when human experimentation is limited for ethical
considerations. Chapter 6 cites two examples in which in-vitro data
make it possible to reduce the uncertainties in risk assessment
associated with exposure to ozone and ultraviolet radiation. The
difficulty in establishing quantitative relationships between
immunosuppression and clinical disease has limited the use of
immunotoxicological data in risk assessment.
RECOMMENDATIONS
Recommendations for the protection of human health
1. Chemicals should be screened to determine if they are potentially
immunotoxic to humans. If immunotoxicity is detected, the chemicals
should be investigated further as part of the risk assessment process.
2. Chemicals for which little or no information is available on
toxicity should be screened for potential immunotoxicity following a
protocol based on, for example, the revised OECD guideline No. 407.
When some information is available on the test material (e.g.
physicochemical properties, toxicokinetics, structure-activity
relationships), a flexible approach to testing is recommended which
permits a rational selection of test procedures.
3. The immunotoxic risk of mixtures of environmental pollutants, in,
for example, fish, to certain human consumer groups (e.g. fishermen)
should be assessed.
Recommendations for protection of the environment
1. Chemicals should be screened to determine if they are potentially
immunotoxic to wildlife species. If immunotoxicity is detected, the
chemicals should be investigated further as part of the risk
assessment process.
2. The immunotoxic risk of environmental pollution to the health of
the ecosystem should be assessed in laboratory, semi-field, and field
studies of the wildlife occupying high trophic levels or those species
judged to be sensitive.
Recommendations for further research
1. The panels of tests suggested for evaluating xenobiotic-induced
immunotoxicity in humans should be investigated to determine their
ability to detect subtle alterations in immune status.
2. The relationships between alterations in immune function and human
health should be established for use in immunotoxic risk assessment.
Epidemiological studies should be carried out that include assessment
of exposure, in order to establish dose-response relationships.
3. The relationship between immunotoxicity and the development of
neoplasia should be investigated.
4. Baseline immunological data should be established for the general
population and for subpopulations such as ethnic minorities, children,
the aged, and pregnant and lactating women in order to assess their
immune status.
5. Immunotoxicological assessment should be conducted for
subpopulations potentially susceptible to the effects of immunotoxic
compounds, including those at the extremes of age and those with
deficient nutritional status.
6. Biomarkers of exposure, effect, and susceptibility should be
identified, developed, and validated for use in epidemiological
studies of immunotoxicity in both humans and wildlife.
7. The quantitative relationship between immune function and host
resistance in animal models, including the nature, magnitude, and
significance of functional reserve and redundancy, should be explored
for risk assessment.
8. Since chemicals and biological agents enter the body via the
respiratory and alimentary tracts and the skin, more research should
be carried out on local immunity.
9. Preliminary observations in laboratory animals that suggest that
primary immunization does not compromise testing for subacute toxicity
should be substantiated by further research, so that functional
testing can be incorporated into toxicology testing.
10. Methods and reagents should be developed in order to characterize
the immune system of wildlife species and to assess their immune
status for immunotoxicological studies.
11. The mechanisms of the immunotoxic action of xenobiotics in humans
should be elucidated by a combination of studies in laboratory animals
in vivo and experiments with human and animal tissues and cell lines
in vitro.
12. In view of the sensitivity of the developing immune system to
immunotoxic injury, more emphasis should be placed on studies
involving perinatal exposure to a chemical or mixture of chemicals.
13. Studies should be conducted to establish whether exposure to
xenobiotics that are not themselves sensitizing adds to the risk of
allergic disease in general.
14. Autoimmune models in laboratory animals should be used to assess
whether xenobiotics can modulate autoimmune disease in humans.
15. The effects on immune function of confounding factors in humans
and animals, including age, race, sex, gender, nutritional status,
acute stress, and underlying disease, should be evaluated further in
order to determine their effects in tests for the immunotoxicity of
environmental chemicals.
16. Methods for assessing cytokines and their production in different
body compartments, including plasma, bronchoalveolar lavage fluid, and
nasal lavage fluid, and by cells isolated from various anatomical
sites should be validated for humans and laboratory animals, and their
applicability for assessing the risk of chemicals should be
established.
17. Data from clinical trials should be made more widely available;
and patients undergoing therapy with immunomodulatory drugs should be
monitored clinically and immunologically in a systematic way.
18. The toxicokinetics of immunotoxic chemicals should be further
investigated, particularly with regard to whether their concentrations
in human biological fluids indicate levels of environmental exposure.
19. The interactions between the immune system, the nervous system,
and the endocrine system should be further investigated, with
particular emphasis on how xenobiotics adversely affect them.
20. The significance of ultraviolet radiation-induced
immunosuppression for public health and the health of ecosystems
should be evaluated.
1. INTRODUCTION TO IMMUNOTOXICOLOGY
1.1 Historical overview
It is well established that each individual has an intrinsic
capacity to defend itself against pathogens in the environment, with a
defence known as the immune system. By general definition, the immune
system serves the body by neutralizating, inactivating, or eliminating
potentially pathogenic invaders such as microorganisms (bacteria and
viruses); it also guards against uncontrolled growth of cells into
neoplasms, or tumours. The major features of the structure and
function of the immune system have been elucidated over the last three
decades; in parallel, awareness grew of toxicological manifestations
after exposure to xenobiotic chemicals. Recognition of the interplay
between toxicology and immunology is relatively recent: A
comprehensive review, published in 1977 (Vos, 1977), was the first
survey of a large series of xenobiotics that affect immune reactivity
in laboratory animals and hence may influence the health of exposed
individuals. Most research groups focusing on toxicity to the immune
system started their activities during the last decade. Textbooks of
immunotoxicology date only from the early 1980s (Gibson et al., 1983;
Dean et al., 1985; Descotes, 1986), while one on clinical
immunotoxicology is more recent (Newcombe et al., 1992).
Immunotoxicology is the study of the interactions of chemicals
and drugs with the immune system. A major focus of immunotoxicology is
the detection and evaluation of undesired effects of substances by
means of tests on rodents. The prime concern is to assess the
importance of these interactions in regard to human health. Toxic
responses may occur when the immune system is the target of chemical
insults, resulting in altered immune function; this in turn can result
in decreased resistance to infection, certain forms of neoplasia, or
immune dysregulation or stimulation which exacerbates allergy or
autoimmunity. Alternatively, toxicity may arise when the immune system
responds to the antigenic specificity of the chemical as part of a
specific immune response (i.e. allergy or autoimmunity). Certain drugs
induce autoimmunity (Kammüller et al., 1989; Kammüller & Bloksma,
1994). The differentiation between direct toxicity and toxicity due to
an immune response to a compound is to a certain extent artificial.
Some compounds can exert a direct toxic action on the immune system as
well as altering the immune response. Heavy metals like lead and
mercury, for instance, manifest immunosuppressive activity,
hypersensitivity, and autoimmunity (Lawrence et al., 1987).
This monograph is concerned mainly with one aspect of
immunotoxicology: the direct or indirect effect of xenobiotic
compounds (or their biotransformation products) on the immune system.
This effect is usually immunosuppression, or the induction of a state
of deficiency or unresponsiveness. Allergy and autoimmunity will be
dealt with in a future Environmental Health Criteria monograph.
Toxicological research over the past decade has indicated that
the immune system is a potential 'target organ' for toxic damage. This
finding was the basis for a number of large scientific conferences on
immunotoxicology and sparked the active interest of national and
international organizations in this field. One of the milestones in
the development of the discipline was the international seminar on
'The Immunological System as a Target for Toxic Damage', held in
Luxembourg in 1984 and organized by the International Programme on
Chemical Safety (IPCS), and the Commission of the European Communities
(CEC). At the seminar, immunotoxicology was defined as 'the discipline
concerned with the study of the events that can lead to undesired
effects as a result of interaction of xenobiotics with the immune
system. These undesired events may result as a consequence of: (1) a
direct and/or indirect effect of the xenobiotic (and/or its
biotransformation product) on the immune system; or, (2) an
immunologically-based host response to the compound and/or its
metabolite(s) or host antigens modified by the compound or its
metabolites' (Berlin et al., 1987). Recommendations were made
concerning the significance to public health of immunotoxicology,
immunotoxicity testing, research and development in immunotoxicology,
the development of databases, and training and education. A subsequent
workshop on 'Immunotoxicity of Metals and Immunotoxicology', organized
by IPCS and the CEC, in collaboration with the International Union for
Pure and Applied Chemistry and German governmental agencies, was held
in Hanover, Germany, in 1989 (Dayan et al., 1990). A meeting on risk
assessment in immunotoxicology was organized by the United States
National Institute for Environmental Health Sciences in 1990. A
meeting on human immunotoxicology tests was organized by the Agency
for Toxic Substances and Disease Registry and the Centers for Disease
Control, in Atlanta, Georgia, United States of America, in 1992. In
1994, two meetings were held: one in Oxford, United Kingdom, organized
by IPCS, on risk assessment in human imunotoxicity, and one in
Washington DC, United States, organized by the International Life
Sciences Institute, on methods in immunotoxicology.
In parallel to these meetings, activities were started within
IPCS for the development and validation of methods for assessing
toxicity to the immune system. In this regard, a hallmark event was
the meeting in 1986 of a technical review and working group, in
London, United Kingdom (IPCS, 1986).
A number of tiered approaches to immunotoxicity testing have been
proposed, in rats (Vos, 1980; Van Loveren & Vos, 1989) and subsequently
in mice (Luster et al., 1988). These approaches have been evaluated
for their capacity to identify chemicals as immunosuppressive. Of a
group of 18 pesticides evaluated in rats, six were identified as
inducing immunotoxicity at doses similar to those that cause other
toxic effects, and five were immunotoxic at lower doses (Vos & Krajnc,
1983; Vos et al., 1983a). Effects were seen on different parameters
with different compounds and included lymphocytopenia, reduced thymic
and spleen weights, and increased levels of serum immunoglobulin (Ig)
G. One of the compounds identified was hexachlorobenzene (Vos, 1986),
which is further described in Section 2. In mice, the tiered approach
was used to assess the immunotoxicity of 51 chemicals, selected on the
basis of factors including structure-activity relationships with
previously identified immunotoxic substances, and use (Luster et al.,
1992). Of the spectrum of assays applied, the strongest associations
with immunotoxic potential were observed with the splenic IgM antibody
plaque-forming cell response and cell surface marker analysis; somewhat
weaker associations were found for natural killer (NK) cell activity,
cytotoxic T-lymphocyte cytolytic activity, lymphocyte proliferation
in vitro after mitogen stimulation, and thymus:body weight ratio.
The tiered approach in immunotoxicity testing is further described in
Section 3.
Multi-laboratory studies have been initiated to validate the
screening of immunotoxic compounds, including the IPCS-European Union
International Collaborative Immunotoxicity Study, a study in Fischer
344 rats, and the international study of the Bundesinstitut für
Gesundheitlichen Verbraucherschutz, und Veterinärmedizin, which were
designed to determine interlaboratory reproducibility. The
experimental animal used in these studies is the rat, and some
functional tests are included. Test methods are also being developed
and validated within the National Toxicology Program (NTP) in the
United States. This programme includes studies of carcinogenicity in
rats and mice, but because the immune system of mice is better
characterized than that of rats, the NTP chose the mouse as the
experimental animal for immunotoxicity assessment. The immunotoxicity
database of the NTP has been evaluated to determine the predictability
(sensitivity and specificity) of the assays. In the Netherlands, a
Committee for Immunotoxicology of the Dutch Health Council reviewed
methods that could be used to assess the immunotoxic properties of a
compound and for deriving information about risks to humans on the
basis of the results of laboratory experiments. The Committee also
examined the relationship between the immunotoxic properties of a
substance and its mutagenic and carcinogenic properties (Dutch Health
Council, Committee for Immunotoxicology, 1991).
The immune system was reviewed by the United States National
Research Council in order to identify the kinds of basic research that
might reveal markers of environmental exposure and disease. Major
emphasis was placed on biological markers of three types: those
originating from the immune system, those related to exposure to
immunosuppressive toxicants, and those of effects of environmental
pollutants. Markers of susceptibility to environmental materials were
also considered to be important, especially if they are of a genetic
nature and can be used to identify individuals susceptible to
autoimmune diseases. The National Research Council subcommittees on
pulmonary toxicology and on immunotoxicology, have published their
reports (US National Research Council, 1989, 1992).
Interest in immunotoxicology within the scientific community is
reflected by the existence of a special section on immunotoxicology
within the Society of Toxicology. An immunotoxicology discussion group
initiated in the United States has an international composition. The
European Union has a programme on science and technology for
environmental protection that includes immunotoxicology as an
important aspect.
There is growing concern in society about the effects of
xenobiotics, such as environmental pollutants, on public health; the
immune system is one of the targets of such effects. Some chemicals
present in the environment that have been reported to influence the
immune system are listed in Table 1 (IPCS, 1986). Immunotoxicity can
result in e.g. reduced resistance towards infection or generation of
tumours that escape immune surveillance. A number of substances
affect immunological parameters; these include halogenated
hydrocarbons such as polychlorinated biphenyls, polybrominated
biphenyls, polychlorinated dibenzo- para-dioxins, and polychlorinated
dibenzofurans (Elo et al., 1985; Lu & Wu, 1985; Bekesi et al., 1987;
Kimbrough, 1987; Hoffman 1992); pesticides and precursors (Fiore et
al., 1986; Deo et al., 1987; Nigam et al., 1993); organic solvents
(Capurro, 1980; Denkhaus et al., 1986); asbestos (Lew et al., 1986);
silica (Uber & McReynolds, 1982); and metals like lead (Ewers et al.,
1982; Reigart & Graber, 1976). Oxidant air pollutants, like sulfur
dioxide, nitrogen dioxide, and ozone, and particles in airborne dust
may affect immune function (Koren et al., 1989; Van Loveren et al.,
1994).
Immunotoxicity in humans is further discussed in Section 2. Few
epidemiological data have been published that indicate suppression or
altered resistance to infection and tumours. In general, the
usefulness of the epidemiological studies that have been published is
limited by the following: exposure is usually uncontrolled, mainly
occurring during accidents; the magnitude and pattern of exposure are
not known, and the exposure is often too low to alter the immune
system measurably; exposure is often not to one xenobiotic but to a
mixture; it is almost impossible to control for confounding
parameters, such as age, sex, genetic background, health status, and
nutritional status; and it is not always possible to define and
analyse appropriate control groups (US National Research Council,
1992). Environmental pollution and its effect on health status are
currently subjects of concern in eastern European countries and have
generated much interest in the worldwide environmental health science
community. Recent epidemiological studies have compared the possible
relationship between exposure to air pollutants and health effects in
the former German Democratic Republic and Federal Republic of Germany,
Table 1. Examples of compounds that are immunotoxic for humans or
rodents
Chemical Immune toxicity
-------------------
Rodent Human
2,3,7,8-Tetrachlorodibenzo-para-dioxin + +
Polychlorinated biphenyls + +
Polybrominated biphenyls + +
Hexachlorobenzene + Unknown
Lead + Unknown
Cadmium + Unknown
Methyl mercury compounds + Unknown
7,12-Dimethylbenz[a]anthracene + Unknown
Benzo[a]pyrene + Unknown
Di-n-octyltindichloride + Unknown
Di-n-butyltindichloride + Unknown
Benzidine + +
Nitrogen dioxide and ozone + +
Benzene, toluene, and xylene + +
Asbestos + +
N-Nitrosodimethylamine + Unknown
Diethylstilboestrol + +
Vanadium + +
From IPCS (1986)
and some of these studies included immunological data or end-points.
For instance, von Mutius et al. (1992) found a higher prevalence of
asthma among schoolchildren in western than eastern Germany, and
Behrendt et al. (1993) observed, surprisingly, that total serum IgE
levels were higher in schoolchildren in eastern than in western
Germany. Several factors were found to influence total IgE: history of
parasitic disease, number of persons per dwelling, and passive
smoking. Sex and passive smoking were the only variables that had a
significant effect in western German children. Air pollutants and
parasitic infections were suggested to be the major contributing
factors to increased IgE production in children in eastern Germany.
Remarkable differences in air quality were seen between eastern and
western Germany, and Behrendt et al. (1995) distinguished two types of
air pollution: type I, composed of sulfur dioxide particles and dust,
occurring predominantly in eastern Europe, is associated with
respiratory infections and other chronic inflammatory airway
reactions; type II, occurring both indoors and outdoors in the
environment in industrialized western countries, is composed mainly of
nitric oxide, nitrogen dioxide, ozone, volatile organic compounds, and
fine particles. The latter type of air pollution is associated with
allergic diseases and allergic sensitization, indicating that air
pollutants interfere with parameters of allergy at the level of
sensitization, elicitation of symptoms, and exacerbation of disease.
Until further epidemiological studies are conducted in humans,
assessment of immunotoxicity in rodents, with subsequent extrapolation
to the human situation, is still a good indicator of toxicity and can
serve as a basis for subsequent decisions and regulations by
authorities to reduce or prevent the risk of human exposure. This
aspect is discussed further in Section 5.
Humans are exposed to environmental contaminants mainly via food,
water, and air. Open water (e.g. rivers, lakes, and coastal areas) and
sediments often act as sinks for environmental pollution. This global
problem can be deduced from disease manifestations in fish that live
in coastal areas, especially those species that live in close contact
with contaminated silt. High levels of contaminants and the diseases
associated with them are not only of economic importance (i.e. to
fisheries) but also affect people who consume the fish, as seen in
studies showing increased levels of contaminants in people eating fish
from the contaminated Baltic Sea (Svensson et al., 1991) and in Inuit
and Indian populations in Canada who consume large quantities of fish
and marine mammals (DeWailly et al., 1992). There is now some evidence
that wildlife aquatic species have decreased resistance and enhanced
incidences of infection and tumours that may be linked to
environmental pollution (Vos et al., 1989; Wester et al., 1994).
Because the immune system of fish has not been characterized in such
detail as that of mammals, immunotoxicological studies have not been
extensively included in ecotoxicology, although a number of reports of
direct toxic actions of xenobiotics on fish species have been
published in this developing field (Wester & Canton, 1987; Payne &
Fancey, 1989; Anderson, 1990; Wester et al., 1990; Khangarot &
Tripathi, 1991; Secombes et al., 1992; Anderson & Brubacher, 1993;
Faisal & Hugget, 1993).
1.2 The immune system: functions, system regulation, and modifying
factors; histophysiology of lymphoid organs
1.2.1 Function of the immune system
In order to interpret pathological alterations of the immune
system in terms of altered function, the physiology of the system must
be understood. Since knowledge of the structure and function of the
immune system is growing rapidly, a review of this subject, focusing
on histophysiology, is presented. This section is not meant to serve
as a textbook on immunology but to provide sufficient information for
an understanding of pathological changes due to immunotoxic action.
For general textbooks on immunology, reference may be made to Sell
(1987), Klein (1990), Brostoff et al. (1991), Roitt (1991), Paul
(1993), and Roitt et al. (1993). The section covers mainly humans and
rodents, but reference is made to other species that are relevant in
immunotoxicity assessment, e.g. fish in ecotoxicology. It should be
noted that species differences can be large, despite fundamental
similarities between the immune systems of animals. It is therefore
difficult to conduct immunotoxicological studies in immunologically
less well characterized animal species, although comparative studies
that are under way may lessen the problems. Zapata and Cooper (1990)
have written a comprehensive textbook on phylogenetic aspects of
immunology. Phylogenetic data, from primitive fish to mammals, are
presented in Table 2 (Cooper, 1982; Klein, 1986; Du Pasquier, 1989;
Zapata & Cooper, 1990; Sima & Vetvicka, 1992). Relevant phylogenetic
aspects of the immune system are described below.
In mammals, the immune system and its reactions consist of a
finely tuned, complex interplay between various cell types and soluble
mediators secreted by those cells (Figure 1), some of which are listed
in Section 7.
Immune responses can be classified roughly as innate (natural and
nonspecific) and acquired (adaptive) responses, in which the reaction
is directed to a specific determinant (antigenic determinant or
epitope). The nonspecific response involves effector cells such as
macrophages (Vetvicka & Fornusek, 1992), NK cells (Herberman &
Ortaldo, 1981), granulocytes (Ross, 1992), and mediator systems
including the complement system (Tomlinson, 1993). Specificity is
based on recognition by specific receptors on lymphocytes or by
antibodies: The classical reaction to bacterial infection, resulting
in antibacterial antibody formation and antibody-mediated destruction
of the pathogen, is only one part of the intrinsic capacity of the
system. Further attributes of the system (Nossal, 1987) are summarized
below.
1.2.1.1 Encounter and recognition
The initiation of an immune response requires adequate
recognition of the pathogen. This recognition often occurs immediately
after entry, e.g. during or after passage through the epithelial
barrier of the body (skin or mucus-secreting epithelia in the
respiratory and gastrointestinal tract). The first defence includes
nonspecific inactivation, e.g. by nonspecific killer cells,
neutrophilic granulocytes, and cells of the mononuclear phagocyte
system (formerly called the reticuloendothelial system). It also
includes antigen processing and presentation to cells such as
lymphocytes of the T-helper-inducer type (Th) which can generate a
specific response.
Table 2. Evolution of immunologically important traits among vertebrates
A. Major histocompatibility complex (MHC) and transplantation
Species Graft MLR CML MHC control Serologically
rejection and/or of immune detectable
GVH response MHC antigens
Tunicata (sea squirts) + ? ? ? ?
Agnatha
Hagfish (Hyperotreti) + ? - ? ?
Lamprey (Hyperoartii) + ? - ? ?
Chondrichthyes (cartilaginous fish)
Shark, ray + ? + ? +
Osteichthyes (bony fish)
Sturgeon (Chondrostei) + ? ? ? ?
Bony fish (Teleostei) + ? + ? +
Lungfish (Dipnoi) + ? ? ? ?
Amphibia (amphibians)
Salamanders (Urodela) + ? ? ? +
Frogs, toads (Anura) + + + + +
Reptilia (reptiles)
Turtles (Chelonia) + + ? ? ?
Lizards, snakes (Squamata) + + + ? ?
Crocodiles, alligators (Crocodilia) + ? ? ? ?
Aves (birds) + + + + +
Mammalia (mammals) + + + + +
Table 2 (cont'd)
B. Complement and immunoglobulins (Ig)
Species Complement Immunoglobulins
IgM IgG-like IgA IgD IgE
Tunicata (sea squirts) ? - - - - -
Agnatha
Hagfish (Hyperotreti) ? ? - - - -
Lamprey (Hyperoartii) ? ? - - - -
Chondrichthyes (cartilaginous fish)
Shark, ray + + + - - -
Osteichthyes (bony fish)
Sturgeon (Chondrostei) + + + - - -
Bony fish (Teleostei) + + + - - -
Lungfish (Dipnoi) + + + - - -
Amphibia (amphibians)
Salamanders (Urodela) + + ? - - -
Frogs, toads (Anura) + + + + - -
Reptilia (reptiles)
Turtles (Chelonia) + + + - - -
Lizards, snakes (Squamata) + + + - - -
Crocodiles, alligators (Crocodilia) + + + ? - -
Aves (birds) + + ? + - -
Mammalia (mammals) + + + + + +
Table 2 (cont'd)
C. Leukocytes
Species Lymphocytes Plasma Macrophages
cells
Small T B
Tunicata (sea squirts) + - - - +
Agnatha
Hagfish (Hyperotreti) + - - + +
Lamprey (Hyperoartii) + - - + +
Chondrichthyes (cartilaginous fish)
Shark, ray + - ? + +
Osteichthyes (bony fish)
Sturgeon (Chondrostei) + - ? + +
Bony fish (Teleostei) + + + + +
Lungfish (Dipnoi) + ? ? + +
Amphibia (amphibians)
Salamanders (Urodela) + + + + +
Frogs, toads (Anura) + + + + +
Reptilia (reptiles)
Turtles (Chelonia) + + + + +
Lizards, snakes (Squamata) + + + + +
Crocodiles, alligators (Crocodilia) + + + + +
Aves (birds) + + + + +
Mammalia (mammals) + + + + +
Table 2 (cont'd)
D. Lymphoid organs
Species Bone Thymus Spleen Lymph glands
marrow or nodes
Tunicata (sea squirts) - - - -
Agnatha
Hagfish (Hyperotreti) - - - GALT
Lamprey (Hyperoartii) - - - GALT
Chondrichthyes (cartilaginous fish)
Shark, ray - + + GALT
Osteichthyes (bony fish)
Sturgeon (Chondrostei) - + + GALT
Bony fish (Teleostei) - + + GALT
Lungfish (Dipnoi) - + + GALT
Amphibia (amphibians)
Salamanders (Urodela) + + + -
Frogs, toads (Anura) + + + ?
Reptilia (reptiles)
Turtles (Chelonia) + + + ?
Lizards, snakes (Squamata) + + + ?
Crocodiles, alligators (Crocodilia) + + + ?
Aves (birds) + + + +
Mammalia (mammals) + + + +
+, positive; ±, to be confirmed; -, negative; ?, not investigated.
MLR, mixed leukocyte response; GVH, graft-versus-host; CML, cell-mediated lympholysis;
GALT, gut-associated lymphoid tissue
The importance of an epithelial barrier for primitive defence
mechanisms is clear in fish, in which a specific mucosal immune system
with local production of antibodies associated with mucus secretion
has been recognized. The gills are the main entry for both antigens
and pathogens living in water, and both migrating macrophages and gill
epithelial cells are involved in the process.
1.2.1.2 Specificity
The immune system can distinguish one particular determinant in
an immense spectrum of determinants. The discrimination between 'self'
and 'non-self' (i.e. the avoidance of autoreactivity) is an example of
this specificity.
Antigens can be polypeptides, carbohydrates, or lipids (lipopoly-
saccharides and lectins). A polypeptide antigen epitope is made up of
about 10 amino acids. Lymphocytes are central to antigen specificity,
as they express receptors for a single, distinct antigenic determinant
on their surface. On B lymphocytes, this antigen receptor is
essentially an antibody (immunoglobulin) molecule (Hasemann & Capra,
1989). The antigen-binding fragment of the surface receptor and the
antibodies produced after differentiation of B cells into plasma cells
have virtually identical structures -- a quaternary structure
comprising the dual heavy and light chains of the immunoglobulin
molecule (the so-called variable part of these protein chains). B-Cell
surface immunoglobulin and the immunoglobulin product of the plasma
cell progeny may differ in the constant part of the heavy chain. Like
the T-cell receptor (TCR), the B-cell receptor has a hetero-oligomeric
structure. After the antigen is bound to the surface immunoglobulin,
signal transduction occurs, in which one alpha-ß chain is linked to
the surface immunoglobulin. Biological responses involving tyrosine
phosphorylation and calcium mobilization are then induced, including
activation, tolerance, and differentiation, depending on the
differentiation stage of the B cell (Pleiman et al., 1994). On virgin
B cells, the surface receptor is an IgM or IgD molecule (with µ or
delta heavy chains, respectively). After so-called immunoglobulin
class switching (Vercelli & Geha, 1992; Harriman et al., 1993), IgG
(gamma chain), IgA (alpha chain), and IgE (epsilon chain) molecules
can be synthesized. Associated with the immunoglobulin molecule on the
cell surface is a dimeric transmembrane molecule, the Igalpha and
Igßchain, which functions in signal transduction and intracellular
activation of kinases of the src family (Pleiman et al., 1994). This
alpha-ß dimeric molecule is now used in identifying B cells.
For T cells, the antigen receptor is a heterodimeric molecule
(either the alpha-ß or the gamma-delta heterodimer), which has a
constant and a variable part, like those of immunoglobulin molecules
(Hedrick, 1989). Transmembrane signalling (tyrosine phosphorylation)
after antigen contact occurs when this heterodimer is linked on the
cell surface to the T3 (or CD3) molecule, which consists of at least
five invariant chains (Clevers et al., 1988; Chan et al., 1992).
The structural differences between the TCR for antigens and the
B-cell receptor (i.e. antibody) arise from differences in the gene
segments that encode the receptors. It is thus not surprising that
T cells recognize other determinants on the antigenic compound than
those recognized by B cells. For large antigens such as proteins,
distinct T-cell and B-cell epitopes can be identified, as illustrated
by the fact that the alpha-ß heterodimeric receptor on T cells
recognizes the antigenic determinant in the context of polymorphic
determinants of the major histocompatibility complex (MHC) antigens.
The antigenic determinant is either a small peptide, produced during
processing of antigen in the antigen-presenting cell and located in a
'groove' formed by the quaternary structure of the MHC molecule
(Adorini, 1990; Rothbard & Gefter, 1991; Germain & Margulies, 1993),
or a larger molecule associated with the MHC molecule outside the
groove. The latter is found for so-called 'superantigens', like
Staphylococcus enterotoxin A (Herman et al., 1991). Thus, with
either type of antigen, Th cells and delayed-type hypersensitivity
T cells (see below) recognize the antigenic determinant only when it
is presented together with the individual's own (self) determinant of
class II MHC. This phenomenon is called MHC class II restriction. In
contrast, T cells of the suppressor (Ts) and cytotoxic (Tc)
populations are MHC class I restricted. The processing of antigen by
antigen-presenting cells and subsequent complexing with MHC molecules
occur intracellularly, but with different pathways for MHC class
I-associated and MHC class II-associated complexing. In addition, MHC
class II-associated complexing may occur with proteins present at very
high concentrations in the extracellular environment (Neefjes &
Momburg, 1993; Engelhard, 1994). MHC restriction does not necessarily
apply to T-cell subsets within the pool expressing the alpha-ß
heterodimeric TCR (Haas et al., 1993). B Lymphocytes do not function
in an MHC restricted manner but recognize nominal antigen with their
surface immunoglobulin receptor. Therefore, recognition of antigens by
B cells and by most gamma-delta T cells, does not require antigen
presentation on cells carrying their own MHC class I or class II
determinants. The total repertoire of antigen recognition
specificities is about 107 for antibodies and somewhat less for the
TCR (about 106) (Roitt et al., 1993)
Phylogenetically, the capacity to reject an allograft is acquired
very early (e.g. sponges), and this has been interpreted as evidence
for an MHC complex. Information on the molecular features of MHC
antigens (Hughes & Nei, 1993) is available, however, only for the toad
Xenopus (Flajnik et al., 1991; Sato et al., 1993) and for fish
species (Hashimoto et al., 1990; Kasahara et al., 1992; Ono et al.,
1992; Hordvik et al., 1993). All vertebrate species produce specific
antibodies and typical cell-mediated immunity indicative of specific
responses, demonstrating that both T and B cells exist, as in birds
and mammals. Few data are available, however, because there are
virtually no reagents for lymphocyte identification and
classification. In chickens, monoclonal antibodies recognize alpha-ß
and gamma-delta TCR and a third TCR with a configuration ß-ß'; various
T-cell subsets have also been identified in this species with
appropriate antibodies (CD3, CD4, CD8; see below). MHC restriction of
antigen recognition by T cells has been demonstrated in Xenopus. In
fish, as in mammals, antigens are processed and presented by accessory
antigen-presenting cells, such as monocytes, to specific lymphocytes
in a seemingly alloantigen (presumably MHC or MHC-like) restricted
fashion (Vallejo et al., 1990; Stet & Egberts, 1991; Vallejo et al.,
1992). 'B-like' cells expressing immunoglobulins occur in all of the
fish and amphibian species that have been studied so far; however,
there is no IgD molecule in lower vertebrates. Presumably, all B-like
cells express on their surface an immunoglobulin molecule of high
relative molecular mass, similar to IgM in mammals.
All non-mammalian vertebrates produce immunoglobulins of high
relative molecular mass (Fellah et al., 1992; Wilson & Warr, 1992;
Wilson et al., 1992; Litman et al., 1993; Marchalonis et al., 1993).
There is probably also an IgG-like immunoglobulin of low relative
molecular mass that is functionally but not structurally equivalent to
mammalian IgG, but it has been observed in only a few species of bony
fish. There is also evidence for the presence of an IgA-like
immunoglobulin in some lower vertebrates. Non-mammalian vertebrates
have no IgD or IgE. The total repertoire of antigen recognition
specificities for antibodies is smaller in non-mammalian vertebrates
than in mammals.
1.2.1.3 Choice of effector reaction; diversity of the answer
After activation of Th cells, an immune response develops in
order to eliminate the antigen; in practical terms, the response
results in inactivation of the pathogen. The response is humoral
(antibody-mediated) and/or cellular (cell-mediated).
In the humoral response, Th cells together with antigen activate
specified B cells to become antibody-producing plasma cells. The
antibodies produced mediate the subsequent inactivation of the foreign
substance in a number of ways. When present in the form of immune
complexes, IgG and IgM either activate the complement system and
induce complement-mediated cytotoxicity (Hansch, 1992; Tomlinson,
1993) or activate secondary effects, which include: (i)
vasodilatation, increased vascular permeability, and attraction of
granulocytes, with subsequent release of lysosomal proteolytic enzymes
(i.e. components of acute inflammation); (ii) IgG antibody-mediated
cellular cytotoxicity, in which the antibody forms the antigen-
specific bridge between the killer cell (macrophage, binding of the Fc
fragment of IgG) and the target; (iii) IgG- and IgM-induced
opsonization and ingestion by phagocytic cells (granulocytes,
macrophages), with involvement of receptors for immunoglobulin Fc and
the complement split product C3d; (iv) binding of IgE to IgE receptors
on the surface of mast cells and basophilic granulocytes, which
induces degranulation with release of mediators after antigen binding.
Complement is phylogenetically very old, as genes that encode
complement components and complement proteins have been identified in
hagfish (Hanley et al., 1992; Ishiguro et al., 1992), and C3-like
activity exists in invertebrates and in all vertebrates. Complement C3
has been purified from all classes of vertebrates, including fish
(e.g. hagfish); in primitive fish like lampreys, it shows 30% homology
with human C3, whereas that in rodents is about 80% homologous with
that in humans (Lambris, 1993). Complement activation by immune
complexes occurs in most primitive vertebrates, but the secondary
effects emerged later in phylogeny. Antibody-mediated cellular
cytotoxicity involving NK cells has been documented in some bony fish.
Little is known about inflammatory reactions in lower vertebrates,
except for some histopathological data obtained in fish.
In the cellular response, Th cells activate precursors of Tc
cells, which subsequently kill the target after antigen-specific
recognition. Furthermore, precursors of lymphokine-producing cells
(for example, delayed-type hypersensitivity T cells) can be activated,
and the lymphokines thus secreted subsequently activate macrophages to
kill the target. Studies of Th clones in vitro have revealed the
existence of different types (Mosmann & Coffman, 1989). Th1 cells
synthesize interleukin (IL)-2, IL-3, tumour necrosis factors alpha and
ß, and interferon (IFN)gamma(cytokines are discussed below), provide
help to B cells (especially in IgG2a synthesis), activate macrophages,
and initiate delayed-type hypersensitivity reactions. Th2 cells secrete
IL-3, IL-4, IL-5, IL-6, IL-10, IL-13, and tumour necrosis factor alpha,
providing extensive help to B cells (for IgM, IgG1, IgA, and IgE
synthesis), and activate eosinophilic granulocytes. Th2 cells do not
play a role in the initiation of delayed-type hypersensitivity. Various
cytokines are involved in the generation of these Th populations, with
dominant effects of IL-4 (Th2) and IL-12 (Th1). Cytokines produced by
each type of Th cell subpopulation appear to downregulate the activity
of the others. The choice of effector reaction is determined in part
by cooperation between the various populations of Th cells: IFN gamma
from Th1 cells can downregulate Th2 cells, while IL-10 from Th2 cells
can downregulate Th1 cells (Mossman & Coffman, 1989).
Finally, it should be noted that the immune system exerts other
types of responses that do not involve activation of Th cells. These
include T cell-independent activation of B cells, usually when the
antigens consist of repeating polysaccharide units (present on
bacteria such as Escherichia coli and Pneumococcus). They also
include direct activation of T killer cells which bear antigen-
specific receptors comprising the gamma-delta heterodimer (Bell, 1989;
De Weger et al., 1989; Raulet, 1989; Haas et al., 1993).
1.2.1.4 Immunoregulation
The effector reaction is induced by a finely tuned interplay
between cells and soluble mediators (Figure 2) (Van Deuren et al.,
1992). On the one hand, cell-cell contact is required; for instance,
between the antigen-presenting cell with (processed) antigen on its
surface and the Th cell. On the other hand, mediators (cytokines or
interleukins) influence cell function at a distance, after binding to
specific receptors on the target cell and subsequent signal
transduction, resulting in cell activation (Foxwell et al., 1992; Taga
& Kishimoto, 1992). New interleukins and their surface and soluble
receptors are being identified and characterized rapidly: at least 15
different interleukins are known at present, in addition to various
growth factors. As a reflection of their function, these factors are
designated as chemokines (attracting cells) or cytokines (influencing
cell function, such as stimulation). Factors synthesized by
lymphocytes are called lymphokines; those synthesized by monocytes or
macrophages are called monokines. Among the consequences of cytokine-
cell interactions are chemotaxis (Miller & Krangel, 1992), production
of subsequent mediators in the cascade, and down-regulation of
cellular function (Kimchi, 1992). The role of mediators in cell
adhesion to vascular endothelium is described below.
Examples of mediators that influence immune reactions positively
are IL-1 (Dinarello, 1992), which is secreted by antigen-presenting
cells and stimulates Th cells; IL-2, which is secreted by Th cells and
stimulates a variety of T cells to amplify the response; and a number
of B-cell stimulatory cytokines (Figure 2). Surface receptors for
these interleukins are present on certain immune cells (and also in
neuroendocrine tissue, mentioned below) or are expressed during
activation. For instance, IL-2 receptors and HLA-DR molecules are
expressed on T cells as part of the activation process, and their
expression is used to assess the state of cell activation (Shabtai et
al., 1991).
In down-regulation, the immune response manifests active
processes. A number of mediators, such as prostaglandin E2, IL-10, and
transforming growth factor ß (Brigham, 1989; Fontana et al., 1992;
Larrick & Wright, 1992), can suppress subsets of T and B lymphocytes.
After stimulation of Th cells in the initiation of the response,
precursors of Ts cells are activated which subsequently inhibit Th
cells from further amplifying the response, in direct cell-cell
contact or by secretion of soluble inhibitors. The existence of Ts
cells has been disputed, as part of the function of these cells is
cytotoxicity, which is exerted by the closely related MHC class I-
restricted Tc population (Bloom et al., 1992). Immunoregulatory
circuits have also been documented at the antibody level, where a
first antibody generates a second one directed to itself. The
relevance of this antibody-anti-antibody network, the so-called 'anti-
idiotype' network, remains a subject of speculation.
The immune system is in a continuous state of homoeostatic
equilibrium. The introduction of an antigen (pathogen) disturbs that
balance by activating antigen-specific cell clones of T and B
lymphocyte origin. The system not only allows the proliferation and
amplification of relevant clones to cope with the antigen, but it also
searches for (and reaches) a state of newly defined homeostasis.
Little is known about the phylogenetic development of mechanisms
for regulating immune responses in vertebrates. Helper, cytotoxic, and
even suppressor lymphoid functions have been reported in ectothermic
vertebrates, but the existence of T-cell subpopulations has not been
demonstrated. T and B cells cooperate in teleost fish and in
amphibians, and there is indirect evidence that they do so in
reptiles; however, MHC restriction of these cell interactions has been
demonstrated only in Xenopus. The existence of clusters consisting
of lymphocytes, macrophages, and plasma cells has been documented in
various lower vertebrate species, which indicates the importance of
cell-cell interactions in evoking immune responses. Cytokine activity
resembling that of IL-1 and interferon has been identified in various
fish species. A T-cell growth factor was first characterized in
Xenopus. An idiotype-anti-idiotype network has been proposed to
explain the regulation of antibody production in some bony fish
(Zapata & Cooper, 1990).
1.2.1.5 Modifying factors outside the immune system
Communication with other homeostatic mechanisms in the body is an
important aspect of immunoregulation; mediators of the response have
effects not only on the internal regulatory network but also on
systems outside the immune system. Communication with the clotting and
kallikrein systems by complement components is one example;
communication with the central nervous system is another (Ader et al.,
1990). For instance, IL-1 generated by antigen-presenting cells
affects the temperature regulatory centre (induction of fever) and the
sleep regulatory centre (induction of slow-wave sleep) in the
hypothalamus (Dinarello, 1992).
One example of interaction between the immune system and the
central nervous system is the profound influence of stress on immune
reactivity (Khansari et al., 1990). Both stressful events and the way
in which individuals cope with stress are involved (Bohus et al.,
1991), as documented in studies mainly in rats but also in other
species, including fish (Faisal et al., 1989). In humans, conditions
of acute psychological stress include bereavement (Bartrop et al.,
1977), marital disruption (Kiecolt-Glaser et al., 1987), and
examination periods (Kiecolt-Glaser et al., 1986), and these can be
associated with a decrease in immune status (Kiecolt-Glaser & Glaser,
1986) which can result in increased risk for infection, including
common respiratory tract infections, e.g. influenza (Boyce et al.,
1977; Clover et al., 1989), infectious mononucleosis (Kasl et al.,
1979), and herpes virus reactivation (Glaser et al., 1985).
The hypothalamus-pituitary-adrenal axis is an important pathway
in the communication between the central nervous system and the immune
system, resulting in synthesis of glucocorticosteroid hormone by the
adrenal gland induced by adrenocorticotrophic hormone from the
pituitary gland (Buckingham et al., 1992). Other mechanisms are those
mediated by the direct action of neuropeptides, such as opioid
peptides (van den Bergh et al., 1991), on immune cells; these are
either stimulatory or down-regulatory, depending in part on
experimental design and conditions. In addition, almost all lymphoid
tissues are innervated (Bulloch, 1987; Felten et al., 1987; Kendall &
Al-Shawaf, 1991), although the role of this neuroregulatory pathway is
largely unknown (Freier, 1990). The immune system and the
neuroendocrine system have a number of biologically active mediators
in common, including cytokines and neuromediators (Fabry et al.,
1994), and are strongly interrelated (Weigent & Blalock, 1987; Sibinga
& Goldstein, 1988; Heijnen & Kavelaars, 1991; Heijnen et al., 1991;
Knight et al., 1992). Drugs that act on the central nervous system,
like some tranquillizers, antidepressants, benzodiazepines,
antiepileptics, anaesthetics, and levodopa (an antiparkinson drug),
may cause immunosuppression (Descotes, 1986). Thus, the principles of
immunotoxicology may find application in neurotoxicology and endocrine
toxicology (Snyder, 1989). Some examples of factors that modify the
immune system, based on studies of the thymus of mice during pregnancy
and of birds after hatching, are given below. Exogenous conditions may
have pivotal influences on the structure and function of the immune
system. In ectothermic vertebrates, these include seasonal changes
like temperature and photoperiod and are governed by corticosteroid
and sex hormones. In spring, during the mating period, there is low
immune reactivity, with thymic involution; in the postmating period
and the first part of summer, there is maximal development of lymphoid
tissue and immune reactivity; at the end of summer, activity declines,
and the lymphoid organs undergo pronounced involution, which persists
throughout the autumn and winter (Zapata et al., 1992). The lymphoid
system of ectothermic vertebrates therefore cannot be described in
morpho-functional terms without taking into account the season in
which studies were conducted. Seasonal variation in
immunoresponsiveness is also seen in laboratory animals (Ratajczak et
al., 1993).
1.2.1.6 Immunological memory
A special feature of the immune response is the generation of
memory after initial contact with an antigen (Gray, 1993). The first
response includes activation and amplification of antigen-specific T
or B cells to exert effector reactions; it ends with the return of
antigen-recognizing cells to the normal resting state (small
lymphocytes). The second contact with the antigen results in
recruitment of more antigen-specific cells to give a stronger signal
and more efficient elimination of the antigen or pathogen. The
reaction also occurs faster, and there is a stronger binding of
antibody to antigen. The gradual increase in the binding capacity of
antibodies during the immune response is known as 'affinity
maturation'. The memory pool of lymphocytes mediates a faster response
than the virgin (unprimed) pool.
The memory effect is most evident in the humoral arm of the
immune response. The first response gives rise only to IgM class
antibody, but antibodies of other immunoglobulin classes (especially
IgG in the internal system) are generated after subsequent contact and
immunoglobulin class switch. The antibodies thus formed may show
increased affinity as a result of somatic mutation (Kocks & Rajewski,
1989).
The cellular basis of immunological memory is largely unresolved.
It appears to be located within the T-cell population, because its
presence within the B-cell population is apparently of short duration
and is associated with the presence and stimulatory activity of the
antigen in germinal centres of secondary lymphoid tissue. The
generation of memory is the basis for vaccination, performed to
prevent contracting infectious disease by bringing about contact with
the pathogen in an attenuated or inactivated, nonpathogenic form.
It is not known whether immunological memory exists in primitive
vertebrates. The classical differentiation into primary and secondary
responses does not exist, as it is mainly immunoglobulin of high
relative molecular mass that is present, and the repertoire of
antigen-recognizing specificities is smaller than that in birds and
mammals. Some fish species may have immunological memory.
1.2.2 Histophysiology of lymphoid organs
1.2.2.1 Overview: structure of the immune system
Components of the immune system are present throughout the body
(Figure 3). The lymphocyte compartment is lodged in lymphoid organs,
from which cells can move to sites of infection or inflammation.
Phagocytic cells of the monocyte-macrophage lineage occur in lymphoid
organs and also at extranodal sites, such as Kupffer cells in the
liver, alveolar macrophages in the lung, mesangial macrophages in the
kidney, and glial cells in the brain (Figure 4). Polymorphonuclear
leukocytes are present mainly in blood and bone marrow as mature and
progenitor cells. These cells accumulate at sites of inflammation.
The lymphoid organs can be classified roughly into two types:
primary or central (antigen-independent) and secondary or peripheral
(antigen-dependent). This classification is based on the antigen-
dependence of cell proliferation and differentiation. It does not hold
for lower vertebrates, including fish. The bone marrow in higher
vertebrates is a primary organ, in which are found the pluripotent
haematopoietic stem cells which differentiate into progenitors of
myeloid cells (which in turn differentiate into granulocytes,
monocytes, erythrocytes, and platelets) and lymphoid progenitors
(Figure 1). The differentiation is not antigen-dependent or antigen-
driven, but this does not exclude a role for antigens in the process.
Factors secreted in the periphery during antigen-specific stimulation
can promote haematopoiesis in the bone marrow (Figure 2) (Fletcher &
Williams, 1992; Kincade, 1992; Saito 1992; Williams & Quesenberry,
1992) or inhibit haematopoietic activity (Wright & Pragnell, 1992).
The bone marrow also functions as a secondary lymphoid organ, because
terminal antigen-induced lymphoid cell differentiation can occur in
its microenvironment. For instance, the bone-marrow cells include both
the memory lymphocyte pool and the major plasma cell population, which
contributes to the intravascular pool of immunoglobulins. Normally,
plasma cell differentiation follows antigen presentation at peripheral
sites, and stimulated cells subsequently migrate to the bone marrow
for final differentiation. The secondary or peripheral lymphoid organs
in the body include the lymph nodes, spleen, and lymphoid tissue along
secretory surfaces like the gastrointestinal and respiratory tracts.
A second classification is based on the location of lymphoid
organs, divided into internal organs (some lymph nodes and the spleen,
in addition to thymus and bone marrow) and external organs (lymphoid
tissue along secretory surfaces and the lymph nodes that drain the
mucosa-associated lymphoid tissue (MALT)) (Figure 3). Lymphoid organs
at these two locations behave somewhat independently in host defence,
for instance in immunoglobulin synthesis. The main function of the
external or secretory immune system is to produce (secretory) IgA
antibody, whereas the internal immune system (mainly bone marrow)
produces IgG or IgM antibody; the major site of IgE synthesis in the
body is also along secretory surfaces. The extent of the secretory
immune system should not be underestimated: About half of the body's
lymphocytes are located in the secretory immune system, and its
capacity for immunoglobulin synthesis is about 1.5 times that of the
internal system. In all vertebrates, intraepithelial lymphocytes and
nonencapsulated lymphoid infiltrates occur in the intestine. True
lymphoid organs, such as the tonsils, Peyer's patches, and the
appendix, have been reported only in higher vertebrates. In birds, the
caecal tonsil, a blind appendix of the posterior intestine, becomes an
important peripheral lymphoid organ after involution of the bursa of
Fabricius.
Another organ that contributes to the immune system is the skin.
It does not contain organized lymphoid tissue, but immune components
in skin are interconnected with other immune organs, leading to the
concept of the skin immune system, or skin-associated lymphoid tissue
(Stingl & Steiner, 1989; Bos, 1990; Nickoloff, 1993; see section
1.2.2.7).
Immune cells and cellular products are transported between
lymphoid organs by blood and lymph vessels (Duijvestijn & Hamann,
1989). For example, Langerhans cells on their way from the skin to a
lymph node are present in lymph as 'veiled macrophages'. The blood
circulation contains only a minor part of the body's total pool of
lymphocytes (estimated at about 1%) and only a selected population,
i.e. the recirculating lymphocyte pool (Figure 5). Therefore,
assessment of only the blood lymphoid compartment does not give a
complete inventory of the body's immune system, as it ignores the
activities of the non-recirculating cells. In general terms, the blood
lymphoid compartment does not include cells that are in a state of
activation, proliferation, or differentiation; such cells are
typically tissue-bound. This rule is not absolute. For instance, in
the case of a highly activated B-cell system with hyperplasia of
germinal centres in lymph nodes, activated B cells may occur in the
circulation. These cells are normally not part of the recirculating
pool. With regard to the non-lymphoid cells of the immune system,
macrophages are tissue-bound (histiocytes), and monocytes are their
counterpart in blood. Dendritic cells, which are present in very low
proportions in blood, are the counterparts of Langerhans cells in
skin, veiled macrophages in lymph, and interdigitating dendritic cells
in lymphoid tissue. The follicular dendritic cells, i.e. the antigen-
presenting cells in follicles in lymphoid tissue (see below), do not
exist in the circulation. Within the myeloid series, the mast cell is
typically tissue-bound (Figure 6).
The endothelium lining the blood vasculature has a major role in
the passage of cells from blood to tissue parenchyma. Adhesion
molecules on circulating cells and endothelium are involved in this
process (Patarroyo, 1991; Gahmberg et al., 1992; Gumbiner & Yamada,
1992). These glycoproteins belong to three families of molecules, the
immunoglobulin supergene family, the integrins, and the selectins
(Springer, 1990; Lasky, 1992; Bevilacqua, 1993). They may be present
on endothelium in the resting state but often require alteration to
become biologically active in inflammatory processes (e.g. under the
influence of mediators like IFN gamma). In lymphoid tissue, they are
called addressins, to indicate their role in the selective homing of
passenger lymphocytes into internal or external lymphoid tissue.
The structure and histophysiology of the bone marrow, thymus,
lymph node, spleen, and MALT are presented in more detail in the
following sections. The different microenvironments of these organs
are summarized in Table 3. A detailed description of the histological
and pathological aspects of rat lymphoid tissue is given by Jones et
al. (1990).
Not only conventional histological features but also the
expression of cell surface markers (immunological phenotype) are
emphasized. Monoclonal antibodies to marker substances
(differentiation antigens) on cell populations are used widely in the
identification of leukocyte populations, especially in cell
suspensions (flow cytometry) and tissue sections (immuno-
histochemistry). The wide range of monoclonal antibodies that currently
exists have been grouped according to the 'cluster of differentiation'
(CD) nomenclature, in which they are classified according to their
reactivity to the same cell marker molecule (but not the same epitope
on those molecules) (Clark & Lanier, 1989; Knapp et al., 1989;
Schlossman et al., 1994, 1995). The CD nomenclature has been adapted
for species other than man (Holmes & Morse, 1988; Jefferies, 1988;
Schuurman et al., 1992a). Some monoclonal antibodies that can be used
in the identification of leukocytes and stromal cells in tissue sections
and cell suspension are described in section 4.1.2.
Table 3. Microenvironments in lymphoid tissue
Microenvironment Cells present Function
Bone marrow Haematopoietic cells organized as Differentiation of stem cells into cells
islands within fatty tissue, mature of the erythroid, myeloid/monocytoid,
leukocytes, plasma cells platelet and lymphoid lineage; antibody
synthesis, memory cells
Thymus
Cortex Reticular epithelium, immature Generation of T-cell competence,
T cells T-cell receptor rearrangement, positive
selection (MHC restriction), negative
selection (autoreactive cells), phenotypic
changes
Medulla Reticular epithelium, dendritic cells, Final generation of T-cell competence
T lymphocytes (negative selection), thymic hormone
synthesis, antigen presentation
Lymph node and spleen
Paracortex (lymph node), Interdigitating cells, Th and Ts cells Lymphocyte entry through high periarteriolar
lymphocyte sheath lymphoendothelial venules (lymph node) or
central arteriole (spleen), antigen
presentation to Th cells, T-cell proliferation,
differentiation, and regulation (Ts cells)
Primary follicles, follicular mantle Dendritic cells (subtype of follicular Storage of virgin and memory B cells,
of secondary follicles dendritic cells), dendritic macrophages, recirculating B cells (surface IgM+IgD+)
B cells, small number of T cells
Table 3 (cont'd)
Microenvironment Cells present Function
Germinal centre Follicular dendritic cells, dendritic T Cell-dependent B-lymphocyte
macrophages (starry-sky macrophages), differentiation, antigen presentation in the
B cells (centrocytes, centroblasts), form of immune complexes (with/without
Th cells complement C3)
Medulla (lymph node), Plasma cells, T effector cells, Termination of antigen-specific reaction:
red pulp (spleen) reticular cells, polymorphonuclear antibody synthesis and immune
granulocytes complex-mediated clearance, Tdth and
Tc cell response
Marginal zone (spleen) Marginal zone macrophages, T Cell-independent B-lymphocyte
marginal metallophilic cells, proliferation and differentiation, e.g. to
marginal zone B cells bacterial polysaccharides, B-cell memory
(surface IgM+IgD- cells)
Mucosa-associated lymphoid tissue
Epithelium covering lymphoid M (microfold) cells Transport (uptake) of exogenous
tissue (e.g. Peyer's patches) substances
Follicles and interfollicular See 'lymph node and spleen' For antibody synthesis: precursors of IgA
areas plasma cells
Mucosal epithelium Epithelial cells, Tc cells, natural First line of defence, synthesis of secretory
killer cells, gamma-delta T cells component, transport of IgA (IgM) to
lumen
Lamina propria Plasma cells, macrophages Synthesis of IgA antibody, phagocytosis
and killing
MHC, major histocompatibility complex; Th, T helper; Ts, T suppressor; Ig, immunoglobulin; Tdth, delayed-type hypersensitivity T cells;
Tc, T cytotoxic
Typing of cells in the T lymphocyte lineage is a good
illustration of immunological phenotyping. Subsets with other
functions usually cannot be identified by conventional cytology but
can be recognized by immunological phenotyping. For example, T cells
with a helper-inducer function are labelled by antibodies to CD4
antigens, and cells with a cytotoxic function are labelled by
antibodies to CD8. The reverse does not hold, however; for example,
not all CD4+ cells are helper-inducer cells, because T cells in the
separate subset effecting delayed-type hypersensitivity and some
macrophage populations are also CD4+*CD8+ T cells are in either
the cytotoxic or the suppressor subset. Thus, there is no unequivocal
relationship between CD4 or CD8 expression and cell function, because
these cell-surface molecules are expressed in relation to the way in
which antigen is recognized: T cells recognize antigen in the context
of MHC molecules -- MHC class II molecules for CD4+ cells and MHC
class I antigens for CD8+ cells (Engleman et al., 1981; Meuer et
al., 1983). This phenomenon is further discussed in sections 1.2.2.3
and 1.2.2.4.
Monoclonal antibodies have been developed to most subtypes of
leukocytes, including NK cells. The specificity of antibody 3.2.3,
anti-NKR-P1, for identifying these cells was described by Chambers et
al. (1989). Immunological phenotyping of these cells in rat spleen and
lung is illustrated in Figure 14d.
NK cells and their activities are an example of the differences
between cell function, cytology, and immunological phenotype. NK cells
are characterized by their killer function against selected targets
in vitro. Cells with cytological features that include cytoplasmic
granules, called large-granular lymphocytes, presumably have a natural
killer function, as do cells with an immunophenotype defined by
certain antibodies. The cells identified in these ways are by no means
identical and belong to different, overlapping subpopulations.
1.2.2.2 Bone marrow
The microenvironment of the bone marrow is found in all of the
hollow bones of the body and occupies the medullary space of the
skeleton (Figure 7). Blood enters the marrow from nutrient arteries at
the place where they bifurcate to form the central artery of the
medullary canal. Branches of the central artery terminate in
capillaries within the medullary space or penetrate the endosteum,
where arterial blood from the nutrient artery mixes with blood from
muscular arteries in the periosteal capillaries. The blood returns via
the vascular sinuses to the central sinus and vein. Blood cells in
various stages of development (Figure 1) intermingle with their
microenvironment, which is composed of reticular cells (adventitial
and fibroblastic), adipocytes, endothelial cells, and extracellular
matrix. These, together with passenger accessory cells (macrophages,
monocytes, and lymphocytes, especially T cells), support the
haematopoietic process (Lichtman, 1981; Weiss & Sakai, 1984; Chanarin,
1985; Weiss & Geduldig, 1991; Mayani et al., 1992).
In many mammals during ontogenesis, haematopoiesis is first
located in the yolk sac and fetal liver, until the bone marrow
develops and becomes functional. In rodents, haematopoiesis also
occurs in splenic red pulp later in life. This effect is less
pronounced in other species, but haematopoiesis can occur at other
sites in the body, e.g. the thymus. In lower vertebrates, a bone
marrow containing haematopoietic stem cells occurs firstly in
amphibians. It contains only granulopoietic and lymphoid cells;
erythropoiesis and thrombopoiesis usually occur in the blood sinusoids
of the splenic red pulp. In fish and more primitive vertebrates,
lymphoid tissue is found at many ectopic locations, such as intestine,
brain, heart, gonads, pro- and mesonephrons, and liver and is
functionally and structurally similar to the bone marrow of higher
vertebrates. In birds, the thymus is the main site of erythropoiesis
during certain periods of life (Kendall & Ward, 1974; Kendall &
Frazier, 1979).
Haematopoiesis: The pluripotent haematopoietic stem cells in
the bone marrow proliferate and differentiate to progenitors of
myeloid, erythroid, and lymphoid cells. Some further differentiation
occurs in the marrow, but the final maturation mainly occurs elsewhere
(Figure 1). The common lymphoid progenitor cell differentiates into a
T- and a B-lymphoid progenitor. Progenitor T cells then move to the
thymus for further maturation. B-Lymphoid progenitors mature via the
pre-B cell stage into virgin B cells, which then leave the
microenvironment of the bone marrow and lodge in the periphery. In
birds, this function in B-cell maturation is performed by the bursa of
Fabricius, an epithelial organ located adjacent to the termination of
the gastrointestinal tract (B stands for bursa-dependent) (Cooper,
1982; Du Pasquier, 1989; Zapata & Cooper, 1990). Although the bursa is
considered essential for the production of B lymphocytes, chickens
bursectomized early in embryonic life have B cells in their peripheral
lymphoid organs and can produce antibodies, even if their repertoire
is very limited. Thus, the bursa seems to be involved in the
generation of immunoglobulin diversity rather than just in B
lymphatopoiesis.
The development of stem cells into more mature cells requires the
presence of a microenvironment, as described above, and is regulated
by haematopoietic cytokines (also called haematopoietins) (Fletcher &
Williams, 1992; Kincade, 1992; Quesniaux, 1992; Saito, 1992; Williams
& Quesenberry, 1992; Wright & Pragnell, 1992; see also Figure 2).
Unmyelinated nerve fibres may terminate in the haematopoietic spaces
(Lichtman, 1981), and neuropeptides may play a role in the regulation
of haematopoiesis. A microgeographical distribution of the various
cell lineages in microenvironments has been suggested, such as the
preferential associations between fibroblasts and granulocytes,
megakaryocytes, and sinus endothelial and adventitial cells and
between macrophages and erythroid cells.
The classical bioassay for haematopoietic stem cell activity is
measurement of colony-forming units in the spleen of irradiated mice
after injection of bone-marrow cells. More differentiated progenitors
can be cultured in vitro to provide information on the specific
lineage of the colony-forming cell, to evaluate the progenitor
activity of macrophages, granulocytes, megakaryocytes, eosinophilic
granulocytes, and mast cells, as well as that of several cell
lineages, such as macrophages and granulocytes. Progenitors of
erythrocytes are assayed as early progenitors or as erythroid burst-
forming units. Studies of the mechanisms involved have revealed a
number of factors that promote the differentiation of progenitor
cells, like granulocyte-macrophage colony-stimulating factor,
macrophage colony-stimulating factor, granulocyte colony-stimulating
factor, and c-kit ligand. These mediators are produced by various cell
types, mainly after activation, and promote haematopoiesis in vivo
in conditions of e.g. infection and inflammation. Other soluble
mediators produced by T cells and by other cells involved in the
antigen-driven immune response also promote haematopoiesis; these
include IL-1, IL-3, IL-6, IL-11, interferon and tumour necrosis
factor. Haematopoiesis in the bone marrow is thus not independent of
antigenic stimulation and exposure but is not typically antigen-
driven. For instance, in the case of acute infection, the bone marrow
produces a large proportion of polymorphonuclear granulocytes within a
short time, manifested in blood as leukocytosis and a shift in the
differential count to band-type immature granulocytes. This shift may
be related to migration of T lymphocytes into the bone marrow
microenvironment, where they activate haematopoiesis. Antigen-induced
lymphoid cell differentiation also occurs in the bone marrow; for
instance, bone marrow cells include the memory lymphocyte pool and the
major plasma cell population, which contribute to the intravascular
pool of immunoglobulins.
Development and aging: The bone marrow is the primary site of
haematopoiesis throughout life, generating over 95% of the
haematopoietic activity in adult mammals (Mayani et al., 1992).
Essentially all of the medullary space in the bones is occupied by
haematopoietic tissue in mice and rats; in contrast, in adult humans,
dogs, and rabbits, haematopoietic tissue is restricted to the proximal
epiphyses of the long bones, the central skeleton, and the skull; most
bone marrow is gradually replaced by fat cells. The normal mean
proportions of bone marrow cells and changes with age in rats were
reported by Valli et al. (1990).
1.2.2.3 Thymus
The thymus is a two-lobed organ located in the mediastinum,
anterior to the major vessels of the heart (Kendall, 1991; Von
Gaudecker, 1991; Schuurman et al., 1993), although it is located in
the neck region in the guinea-pig. Its anatomical location complicates
complete thymectomy in vivo. The two independent lobes, attached to
each other only by connective tissue, consist of smaller lobules,
which have basically the same architecture, with a subcapsular and
outer cortical area, a cortex, and a medulla (Figure 8). With
conventional histological stains, the cortex is strongly stained,
owing to a dense population of small lymphocytes, and the outer cortex
and medulla show a paler colouring.
Blood vessels enter the lobules at the cortico-medullary junction
and extend radially into the cortex. Nerves course along the blood
vasculature. Fenestrated capillaries are very infrequent in the
cortex. The thymus is unique among the lymphoid organs in that its
microenvironment consists of a reticular epithelium (in birds, the
bursa of Fabricius also has an epithelial framework). Macrophages
derived from the bone marrow are found in the cortex and medulla as a
transient population. Dendritic cells in the medulla, resembling
interdigitating dendritic cells in lymph nodes, have a major function
in antigen presentation and strongly express MHC class II antigen.
The thymus appears for the first time in vertebrate phylogeny in
cartilaginous fish (sharks and rays). More primitive vertebrates lack
a thymus, although they can manifest cellular immune responses. Except
in bony fish, the thymus is histologically similar in all vertebrates,
although it is derived from distinct pharyngeal pouches in different
species. The classical cortex-medulla demarcation is not present in
fish thymus, and most thymocytes seem to occupy the central part of
the organ. In all species, epithelial cells organize a supporting
network in both cortex and medulla. Hassall's corpuscles, which are
epithelial aggregates with centrally located cell debris, occur in the
medulla of the human thymus but are scarce or absent in the rodent
thymus and in the thymus of ectothermic vertebrates. In the latter,
epithelial cysts are frequent. The thymus of lower vertebrates, in
contrast to that of mammals, contains numerous myoid cells; in avian
thymus, significant erythropoiesis has been documented (Kendall &
Ward, 1974; Kendall & Frazier, 1979), which may also occur in mammals
(Kendall, 1980; Kendall & Singh, 1980).
T-Cell maturation: selection in the thymus: T Cells reside in
the thymus during their maturation from progenitor cells to
immunocompetent T cells. This gland has a privileged function in
promoting the maturation process (Brekelmans & Van Ewijk, 1990;
Shortman et al., 1990; Van Ewijk, 1991; Boyd et al., 1992). In
congenitally athymic, 'nude' animals (Schuurman et al., 1992b,c) and
in thymic aplasia in children with complete Di George's syndrome, the
absence of a functionally active T-cell system is causally related to
the aplasia of the organ. The process of T-cell maturation includes a
number of steps located in different microenvironments (Schuurman et
al., 1993): The least mature cells, which enter the lobules from the
bloodstream at the cortico-medullary junction, first move to the outer
subcapsular cortex, where they appear as large lymphoblasts. They then
pass through the cortex, where they become small lymphocytes with a
scanty cytoplasm. Finally, they move to the medulla, where they appear
as medium-sized lymphocytes. These translocational stages in
development are monitored on the basis of the immunological phenotype.
For the CD4-CD8 phenotype, the cells change from CD-CD8- (so-called
double-negative) at a very immature stage, then change to a CD4-CD8+
stage and into a CD4+CD8+ (double-positive) phenotype, which is
found on almost all lymphocytes in the cortex. In the medulla, T cells
have the phenotype of mature cells, with distinct CD4+CD8- (about
two-thirds) and CD4-CD8+ (about one-third) populations.
This phenotypic change is accompanied by a crucial aspect of
intrathymic T-cell maturation: genesis of the TCR, which consists of
the alpha-ß heterodimer. Initially, the DNA genomic organization that
encodes these chains is in germline configuration, with a variety of
gene segments encoding the variable part of the receptor molecule.
Before transcription and translation into TCR becomes possible,
combinations have to be made of the gene segments that encode the
variable and constant parts of the TCR. This process of gene
rearrangement requires the thymic microenvironment, and only after it
is completed can the cell synthesize the receptor. The receptor is
then expressed on the cell membrane together with the CD3 molecule,
which may act as the transmembrane signal transducing molecule after
TCR stimulation; the CD3 molecule is already present in the cytoplasm
of the cell even before the TCR has been synthesized. T Cells at this
stage of maturation can be recognized by cytoplasmic staining with CD3
reagents.
The TCR gene rearrangement is similar to the rearrangement of
genes that encode immunoglobulin heavy and light chains, which takes
place in the bone-marrow microenvironment. Once TCR has been expressed
at the surface, however, the cell undergoes a process unique to T
cells, namely, specific selection on the basis of recognition
specificity (Blackmann et al., 1990; Sprent et al., 1990; Von Boehmer,
1990). First, the cell is examined for its capacity to recognize an
antigen in the context of its own MHC (self-restriction); then it is
allowed to expand (positive selection). Second, the cell is examined
for its capacity to recognize a self-antigen (autoreactivity). If it
recognizes a self-antigen, it is blocked from further differentiation
(negative selection). In this way, the random pool of antigen
recognition specificities of T cells is adapted to the host's
situation; the total repertoire of the alpha-ß T-cell population
(estimated at 1012 different epitopes) changes into the potentially
available repertoire (estimated at recognition of about 106
epitopes). Current theories of negative selection state that this step
is not feasible for all putative autoantigens in the body. Rather, it
applies to a selection of potentially harmful specificities (in
particular MHC antigens). If a cell is not selected during positive or
negative selection, it dies, possibly by suicide or apoptosis
(McDonald & Lees, 1990). A hallmark of apoptosis is endonuclease-
induced chromosomal fragmentation into 200 base-pair fragments
(McConkey et al., 1990). Histologically, apoptosis is recognized by
the presence of condensed, sometimes fragmented nuclei, which can be
found in phagocytic macrophages ('tingible body' or 'starry-sky'
macrophages) (Kendall, 1991).
Function of the microenvironment: It is generally accepted that
the epithelial microenvironment of the thymic cortex plays a major
role in positive selection. This microenvironment expresses MHC class
I and class II products and shows close interactions with lymphocytes
morphologically (at the electron microscopic level). This close
interaction is reflected in the complete inclusion of lymphocytes
inside the epithelial cytoplasm ('thymic nurse cells') (Van Ewijk,
1988). Negative selection has been ascribed to either the epithelial
compartment or the medullary dendritic cells. The different processes
occurring in early (cortical) and late (medullary) maturation are
associated with differences in the microenvironment. Epithelial cells
in the cortex and medulla differ in antigen expression,
ultrastructural characteristics, and their capacity to synthesize
thymic hormones such as thymulin, thymic humoral factor, thymosin, and
thymopoietin. These hormones have a major function late in intrathymic
T-cell maturation, and the major site of thymic hormone synthesis is
the medullary epithelium (Dabrowski & Dabrowski-Bernstein, 1990).
The cortex can be considered a primary lymphoid organ because it
is an antigen-free microenvironment with a blood-thymus barrier. In
contrast, antigens can move relatively freely into the medulla and
encounter antigen-presenting dendritic cells as well as antigen-
reactive T cells. Thus the medulla has the properties of a secondary
lymphoid organ (Van Ewijk, 1988; Kendall, 1991).
Ontogeny, growth, and involution: The thymus in rodents reaches
full development at around day 17 of gestation, that is about five
days before birth. In humans, a fully developed thymus is first seen
at the 16th to 17th week of gestation (Von Gaudecker, 1986), which is
relatively earlier in the gestation period than in rodents, since the
human immune system is more mature at birth than that of rodents.
Nevertheless, a thymus that appears histologically to be fully
developed may be functionally somewhat immature. For instance, fetal
thymus from humans (McCune et al., 1988; Namikawa et al., 1990) or
rats (De Heer et al., 1993) can be transplanted into mice with the
severe combined immunodeficiency (scid) mutation and grow for longer
than tissue obtained postnatally (see also section 4.5.3). After
birth, when the individual first comes into contact with exogenous
antigens, the thymus is called upon to provide large numbers of T
cells to the periphery, and the organ grows in a relatively short time
-- in humans within three to four weeks, from about 15 to 50 g. In
rats, the organ reaches it largest relative size about one week after
birth; the absolute weight is greatest at about six to eight weeks of
age. After adulthood is reached, the thymus starts to involute
(Steinmann, 1986; Kuper et al., 1990a; Schuurman et al., 1991a; Kuper
et al., 1992a), a process that may be related to changes in the
hormonal status of the individual; circulating thymic hormone is
reduced to very low levels in adults. The underlying mechanisms are
not fully understood. The consequences of age-associated involution
are obvious: emigration of lymphocytes from the thymus decreases
dramatically, from, for instance, 1.6 × 106/day in one-month-old
mice to 4 × 104/day in one-year-old mice (Stutman, 1986; Shortman et
al., 1990). Apparently, the persistent generation of a new antigen-
recognition repertoire in the T-cell population of adults is not
needed. Instead, the body can defend itself using the established
repertoire and extrathymic self-renewal of the T cells. Similar
processes may occur after artificial involution of the thymus caused
by toxic compounds or acute stress (including acute disease); this
aspect is further discussed below.
It should be emphasized that the basic architecture of the thymus
is not a fixed histological entity; its features depend on the age and
the stress hormone status of the individual. A 'normal' architecture
can be expected only between the late gestational period and young
adulthood, before the start of age-associated involution. This
phenomenon has important implications for the selection of rodents
according to age for studies of immune function and in the
interpretation of studies of immunotoxicity.
In mice, severe but reversible changes occur in the thymus during
pregnancy (Clarke, 1984; Clarke & Kendall, 1989; see also Figure 9).
The weight of the thymus shows an initial small rise in early
pregnancy, from about 35 to 40 mg, and dramatically decreases to 15 mg
or less at day 17 of pregnancy (Clarke, 1984). This involution is
associated with severe lymphodepletion of the cortex. Cell death is
seen by the presence of apoptotic figures and phagocytosis in
macrophages and epithelial cells. Remarkably, the large lymphoblasts
in the outer cortex remain relatively unattached, and the same applies
for thymocytes in thymic nurse cells. In birds, changes with breeding
sessions have been found (Kendall & Ward, 1974; Ward & Kendall, 1975).
In a study of a wild population of adult red-billed queleas, cyclical
enlargement and regression of the thymus were documented; at the time
of mating and laying, most birds, irrespective of sex, showed an
involuted thymus; on subsequent egg incubation the thymus size
increased, with a decline in the latter half of the rearing period.
1.2.2.4 Lymph nodes
A finely branched lymph vessel system (lymphatics) is involved in
the return of interstitial fluid in tissue to the blood circulation,
with lymph nodes spaced at regular intervals (Dunn, 1954; Tilney,
1971). The major sites of lymph nodes (or groups of lymph nodes) are
shown in Figure 3; a scheme of the areas drained by distinct lymph
nodes or lymph node groups is given in Figure 10, a schematic
presentation of individual lymph node architecture is presented in
Figure 11, and the histology of a lymph node is shown in Figure 12.
The main functions of lymph nodes are to filter pathogens from the
afferent lymph and then to initialize immune reactions. The afferent
lymphatics penetrate the lymph node capsule and connect with the
subcapsular sinus, which in turn connects with the cortical and
medullary sinuses. On the basis of the lymph flow through the node,
basic units can be recognized, each of which is supplied by its own
afferent lymph vessel and which comprise part of the paracortex
(Bélisle & Sainte-Marie, 1981; Sainte-Marie et al., 1990). Afferent
and efferent blood vessels are connected to the organ at the hilus,
where the lymph leaves the node via the efferent lymphatic(s). The
efferent lymphatics drain into other lymph nodes or directly into the
thoracic duct, which enters the bloodstream, in the rat via the left
subclavian vein.
Lymph vessels and lymph nodes occur only in mammals. In some
birds and monotremes (primitive mammals), primitive lymph nodes
directly interposed in the lymph circulation have been described.
Ectothermic vertebrates do not have lymph nodes. Some small lymphoid
organs such as lymph glands and jugular bodies have been described in
some species of frogs but not in others. These organs presumably
filter antigens from both blood and lymph and have been claimed to be
phylogenetic precursors of mammalian lymph nodes. Likewise, small
lymphoid aggregates associated with the cardinal veins occur in some
reptiles.
Lymph nodes are surrounded by a connective tissue capsule. The
nodes comprise various compartments or microenvironments: (i) the
outer cortex, with follicles and interfollicular areas; (ii) the inner
cortex or paracortex; and (iii) the medulla, with medullary cords and
medullary sinuses. These compartments are easily differentiated into
sections after conventional histological staining. In the cortex,
interfollicular areas and the paracortex (T-lymphocyte area) are
differentiated from follicles by the presence of blood vessels lined
by high endothelium (high endothelial postcapillary venules, discussed
below). Follicles (B lymphocyte area) are rounded structures, located
mainly immediately underneath the capsule; they present as
accumulations of small lymphocytes (primary follicle) or a pale-
stained centre with large lymphoid cells (centrocytes, centroblasts)
and tingible-body macro-phages surrounded by a mantle with small
lymphocytes (secondary follicles). The interfollicular areas are
continuous with the paracortex, and the latter is continuous with the
medullary cords.
The arterial blood supply enters the node at the medulla and ends
in the paracortex as arteriolar capillaries, with branches in the
follicles. The capillaries feed venules that are lined with high
endothelial cells. The high endothelial venules run from the
paracortex into the medullary cords and then leave the node via the
vein in the hilus. Lymphocytes migrate through the high endothelial
venules after adhering to the endothelium by specific receptor-ligand
interactions (Picker & Butcher, 1992). The adhesion molecules on
lymphocytes and endothelium that are involved in this binding and
subsequent passage thought the endothelial layer are the addressins,
reflecting the difference in receptor-ligand interactions that exists
between lymph nodes of the internal and external lymphoid systems.
Subsequently, lymphocytes can specifically reach the various nodes,
using the same route (the blood). After the lymphocytes have migrated
into the parenchyma, they move into their microenvironment or
compartment.
Antigen encounter and immune reactivity: The main route of
access for antigens and pathogens is the afferent lymph flow; antigens
can also come into contact with the lymph node tissues via the blood.
Antigens in the lymph, either free or processed by veiled macrophages,
enter the node through the subcapsular sinus, which is rich in
macrophages that can phagocytose free antigen. From there, antigens
move to the paracortex, where they are presented to CD4+ Th cells by
the antigen-presenting cells for initiation of the immune response.
The main antigen-presenting cells in the lymph node are the
interdigitating dendritic cells, the tissue equivalents of veiled
macrophages, which can arise from Langerhans cells in the skin (for
e.g. lymph nodes draining the skin). The antigen-presenting cells
express MHC class II antigens in high density, enabling the alpha-ß
TCR of the Th cells to recognize the antigenic determinant complexed
with the polymorphic ('self') MHC class II molecule. The CD4+
molecule on the Th cell surface binds to a non-polymorphic determinant
of MHC class II molecules and strengthens the binding between Th and
antigen-presenting cells (Janeway, 1992). The cellular interaction
triggers the synthesis of cytokines like IL-1 and IL-2. This process
is down-regulated by Ts cells in a way that is not yet completely
understood.
Follicles are involved in antigen-driven B-cell activation,
somatic mutation, positive and negative selection, and memory and
plasma cell development (Szakal et al., 1989; Kroese et al., 1990; Liu
et al., 1992) and are known as primary follicles in the resting state.
They contain small, IgM+IgD+ virgin B cells in a framework of
follicular dendritic cells. During stimulation by antigens, the
follicles change into secondary follicles consisting of a germinal
centre surrounded by a mantle. Antigen may be transported into the
follicle by immune (T) cells, but this route is not yet fully
established. Antigen is presented to B cells in immune complexes, with
complement split products like C3b, which are trapped in cytoplasmic
extensions of the follicular dendritic cells. The interaction between
complement and complement receptors on these cells has a pivotal role
in the adherence of antigen to them. Fc receptors also play a role,
but only in rodents. The complement split products in the trapped
immune complexes have an accessory function in antigen presentation.
Local CD4+ Th cells assist in B-cell activation. Antigen-driven
B-cell activation and proliferation in the germinal centre are
accompanied by an isotype switch of the immunoglobulin class
synthesized by the B cell. In addition, the affinity of the antibody
increases as a result of somatic mutation (Kocks & Rajewski, 1989). A
kind of selection mechanism has been proposed in this antigen-driven
process, in which cells that produce antibody of higher affinity are
selected preferentially, and cells that produce antibody of lower
affinity are not selected and subsequently die, perhaps by apoptosis
(programmed cell death) (Liu et al., 1989). This selection resembles
that of developing T lymphocytes in the thymus; it differs from T-cell
selection by the absence of negative selection and the occurrence of
somatic mutation. Antigen may remain in the follicular compartment for
quite some time, thereby causing persistent activation of B cells,
related to the state of immunological memory within the B-cell
population. After the antigen disappears, the immunological memory in
the B cells is short-lived and disappears. B-Cell activation in
germinal centres leads to activated cells with a specific morphology,
the so-called centrocytes and centroblasts. Finally, B-cell activation
leads to the formation of plasma cells, both in the periphery of
germinal centres but more predominantly in the medullary cords of the
lymph nodes.
The main site of effector immune reactions is the medulla. The
medullary cords house macrophages, granulocytes, activated effector T
cells, and plasma cells. The effector T cells include CD4+ delayed-
type hypersensitivity T cells (mediator-producing cells) and CD8+ Tc
cells. The Tc cells bear an alpha-ß TCR that recognizes antigen in the
context of the polymorphic determinant of MHC class I molecules. The
CD8 molecule has an accessory function in this process, since it binds
a non-polymorphic class I determinant (Janeway, 1992). Antigen
recognition by Tc cells is thus different from that by Th cells. The
reaction products of the effector cells, such as lymphokines, and
effector cells like plasma cell precursors leave the lymph nodes via
the efferent lymph or blood circulation to go to other sites of the
body.
Development and aging: Lymph node morphology is dynamic: its
appearance throughout life is directly related to the type and amount
of antigenic stimulation. After antigenic contact, the organ increases
in size within a relatively short time, with high proliferative
activity of lymphocytes and germinal centre formation, depending on
the type of reaction and the choice of immunological reaction. In the
case of B-lymphocyte reactions, hyperplasia of follicles is seen (e.g.
after bacterial infection); in the case of T-cell reactions, the
interfollicular areas or paracortex become enlarged (e.g. in viral
infection). After the reaction is terminated or is transferred to the
next draining lymph node, it regains its normal small size. Germinal
centres with an interfollicular microenvironment can develop
extranodally, especially at sites of chronic inflammation.
The dynamics of the lymph node can be illustrated by several
examples. After immunization of the footpad with antigen mixed with an
adjuvant, such as Freund's complete adjuvant, containing killed
mycobacteria, the draining popliteal lymph node becomes enlarged (in
rats, from about 5 mg to more than 100 mg), and granulomatous
reactions (epitheloid-cell granuloma) can be seen histologically. The
swelling of lymph nodes is used to assess reactivity towards chemicals
and in the evaluation of immunomodulatory drugs (Gleichmann et al.,
1989). In the popliteal lymph node assay, a test substance is injected
subcutaneously into one footpad and the contralateral side is left
untreated or injected with vehicle only. The effect of the substance
is subsequently estimated from the difference in weight between the
popliteal lymph nodes. Further evidence for immunostimulatory activity
in vivo is obtained by histological appearance, often manifested as
follicular hyperplasia.
Lymph nodes develop relatively late in fetal life: At birth, the
anlage of unstimulated lymph nodes is present, containing few lymph
cells. The lymph nodes develop quickly after exposure to many new
(exogenous) antigens. In adults, they may become relatively quiescent
and small, with virgin T and B cells and primary follicles. The lymph
nodes in aged rats are capable of the same degree of activity as those
in young individuals upon antigenic stimuli. The state and type of
activation in the various nodes of adult and aged animals differ under
normal housing conditions and are a reflection of the absence or
existence of continuous local stimulation with antigens or disease
processes, like inflammation and the presence of a tumour in the
drained area (Ward, 1990). The central nodes of the mandibular and
superficial cervical group, which are continuously exposed to
(aero)antigens via the oronasopharynx, may contain a considerable
number of plasma cells and precursors in the medullary cords and
relatively well-developed germinal centres. Sinal histiocytosis (that
is, considerable numbers of macrophages in the sinuses) and
accumulations of pigmented macrophages are often present in the
mesenteric lymph nodes, which are continuously exposed to antigens via
the digestive tract. An extensive review of lymph node development and
aging is given by Losco & Harleman (1992).
1.2.2.5 Spleen
The spleen consists of two main compartments: the red and white
pulp (Van Rooijen et al., 1989; Dijkstra & Sminia, 1990; Laman et al.,
1992; Van den Eertwegh et al., 1992). A schematic drawing of the
spleen is presented in Figure 13 and histological views in Figure 14.
The red pulp consists of blood-filled sinusoids and Bilroth's cords
containing macrophages, lymphocytes, plasma cells, and NK cells.
Macrophages perform major functions in clearing blood cells (for
instance, old red blood cells) and in phagocytosis, especially of non-
opsonized particles. This high-volume filter function is made possible
by two factors: the direct contact, unobstructed by blood-vessel
walls, between phagocytic cells and blood-borne particles; and the
large blood supply, estimated at about 5% of the total blood volume
per minute. There are no lymphatics in the spleen.
The phagocytic function is especially important in the case of
intravascular pathogenic microorganisms, before antibody formation and
subsequent opsonization occur (early bacterial septicaemia). The
mononuclear phagocyte system of the liver (Kupffer cells) plays a
major role in the removal of opsonized particles. Together with the
hepatic phagocytic system, splenic macrophages synthesize complement
components, although this is done mainly by hepatocytes. In rats and
mice, the red pulp contains nests of (extramedullary) haematopoiesis,
characterized histologically by megakaryocytes and normoblasts. In the
case of systemic septicaemia, when pathogenic microorganisms have
reached the blood either directly or after inadequate filtering
through lymph nodes, the red pulp increases and contains large
proportions of (immature) granulocytes. Differentiation between
septicaemia and extramedullary haematopoiesis is not always easy; the
decreased or absent white pulp in septicaemia can be helpful in making
this differentiation.
Phylogenetically, cartilaginous fish are the first species that
have a spleen, which consists of lymphoid follicles and a red pulp
that generally houses developing erythroid cells (Zapata & Cooper,
1990). The lympho-haematopoietic masses of the intestine that are seen
in some lower vertebrates (e.g. lampreys) are not primitive spleens
but rather primitive lymphohaematopoietic organs equivalent to
mammalian bone marrow. In most bony fishes, the white pulp is poorly
developed, probably reflecting the existence of other peripheral
lymphoid organs, e.g. the kidney, which participate actively in the
immune response. After antigenic stimulation, the amount of splenic
lymphoid tissue increases considerably in all lower vertebrates,
although germinal centres do not occur. At the cellular level,
however, antigen-presenting cells and cells retaining immune complexes
on their surface have been described in the spleen of some bony
fishes, anurous species, and reptiles.
White pulp: The spleen contains about one-quarter of the body's
total lymphocyte population; during lymphocyte recirculation, more
cells pass through the spleen than through all the lymph nodes.
Lymphocytes in the spleen reside in the white pulp, which consists of
a central arteriole surrounded by the periarteriolar lymphocyte
sheath, a T-lymphocyte area. The outer sheath contains B lymphocytes
and, after antigenic stimulation, plasma cells. Adjacent follicles
contain B cells. Around the periarteriolar lymphocyte sheath and
follicles is a corona containing B cells, called the marginal zone;
this region is easily distinguished, especially in rats. The
periarteriolar lymphocyte sheath has a microenvironment and a
passenger leukocyte content similar to that of the lymph node
paracortex. Some sources claim that the spleen is a rich source of Ts
cell activity, exceeding that of lymph nodes. The follicles are not
essentially different in structure and function from those of lymph
nodes. The spleen performs a major function in humoral immunity by
synthesizing IgM class antibodies, especially to blood-borne antigens.
The microenvironment of the marginal zone is unique to the
spleen. Histologically, the B cells at this site are of medium size;
on histological staining, they are larger and paler than B cells in
primary follicles and in the follicular mantle of secondary follicles
(Figure 14). In addition, they do not show the morphology of the
centrocytes or centroblasts found in germinal centres, and the
phenotypic expression (surface IgM+IgD-) indicates that the B
cells in the marginal zone are a separate population. Special
macrophage types are present, which are known as marginal zone
macrophages and marginal metallophilic macrophages. The latter are
located at the periphery of the white pulp, along the inner border of
the marginal sinus, and can be stained by silver impregnation.
The physiological function of the marginal zone has been
characterized recently. First, the site retains B-lymphocyte memory;
second, it mediates humoral responses that do not directly involve
T cells. These T-independent responses are elicited by polysaccharide
antigens of encapsulated bacteria, which are present in repeating
units on the microorganism and are presented to the B cells by
marginal zone macrophages. The response may not be completely T cell-
independent in all cases, as T cell-derived factors enhance the
response to some of these antigens. The antibodies generated are
mainly of the IgM class, as T-cell help is required for an isotype
switch.
In conclusion, the main immunological function of the spleen is
to defend the body's vascular compartment by generating T cell-
independent IgM antibody responses to bacterial polysaccharides and by
exerting an enormous phagocytic power. This function is lost after
splenectomy, when reduced nonspecific phagocytosis of non-opsonized
particles, lowered serum IgM levels, and increased susceptibility to
infections by encapsulated bacteria have been described.
1.2.2.6 Mucosa-associated lymphoid tissue
The secretory epithelial surfaces of the body form a major route
of entry for potentially pathogenic substances. These surfaces include
the epithelia of the gastrointestinal, upper and lower respiratory,
and urogenital tracts (Miller & Nicklin, 1987; Sminia et al., 1989).
The host response at these locations ranges from nonspecific
constituents, such as a physical or mechanical component (epithelial
barrier, motility of the gastrointestinal tract, and the mucociliary
escalator in the respiratory tract), to a chemical component (low
gastric pH, mucus, lysosomal and digestive enzymes), and antigen-
specific components of the immune system.
Nonspecific killer cells are found in significant numbers in the
lungs and along the epithelium of the gastrointestinal tract, where
lymphocyte-like cells have been found to kill pathogens, presumably
without prior sensitization (Hanglow et al., 1990). In mice, these
cells have been characterized as T cells with a gamma-delta TCR,
which, in contrast to Tc cells that express alpha-ß TCR, kill targets
in an MHC-nonrestricted manner (Raulet, 1989). The cells have antigen
specificity that is encoded at the DNA level by variable gene
segments, but the repertoire appears to be smaller than that which
encodes TCR alpha or ß chains. These gamma-delta T killer cells are
not generated under the strict influence of the thymus, as are their
alpha-ß T-cell counterparts (Bell, 1989; Haas et al., 1993). Apart
from their killer activity, these cells may serve as inducing elements
for the response mediated by alpha-ß TCR-expressing T subsets. In this
initiating activity, gamma-delta TCR molecules shed from the
lymphocyte surface may act as antigen-specific factors (De Weger et
al., 1989).
Lymphoid tissue: Lymphoid tissue occurs just underneath the
secretory epithelium, in the duodenum and jejunum as Peyer's patches
(Figure 15), in the appendix of the large intestine, along the bronchi
(Sminia et al., 1989), and in the oro- and nasopharyngeal regions
(Kuper et al., 1992b). These mucosal lymphoid tissues share structural
and functional characteristics and are strongly interrelated. The
common designation 'mucosa-associated lymphoid tissue' (MALT) is
therefore used to refer to bronchus-associated, gut-associated, and
nasal-associated (nasal cavity and nasopharynx) lymphoid tissue.
Nasal-associated lymphoid tissue has been identified in horses,
monkeys, and rats (Figure 16) (Kuper et al., 1990b). In humans and
domestic animals, the larger lymphoid nodules in the pharyngeal region
are called tonsils. Together with the intermediate lymphoid tissue,
the tonsils form Waldeyer's tonsillar ring.
The organization of MALT is similar to that of lymph nodes, with
B cell-containing follicles and T cell-containing interfollicular
areas. Afferent lymph vessels are lacking, because pathogens can enter
the tissue through the covering epithelial layer. The epithelial cells
at this location (the 'M' or microfold cells) are often thinner than
those at other secretory sites, in order to enable efficient passage
of antigens. Stimulated gut-associated lymphoid tissue and human
tonsils often have prominent follicles with germinal centres. In
contrast, germinal centres are scarce in stimulated bronchus-
associated and nasal-associated lymphoid tissue in rodents, due to the
fact that immunological reactions occur mainly in the draining
cervical lymph node. Th medulla-like areas seen in lymph nodes are
absent in MALT. Lymphocytes and NK cells are found in the lymphoid
tissue, in interstitial tissue in the lung, and in the lamina propria
along the gastrointestinal tract.
The homing specificity of lymphocytes into lymphoid tissue by
migration through high endothelial venules is described above. The
specificity of the homing phenomenon to MALT has the advantage that
the same circulation pathway (i.e. the blood) is used by the secretory
and internal immune system (with the intrinsic possibility of mutual
contact). In addition, the antigen message received at one secretory
site is followed by effects at all secretory surfaces. Thus, after
antigen presentation in the gastrointestinal tract, effector cells
(e.g. IgA antibody-synthesizing plasma cells, see below) are found at
the site of stimulation and at other secretory sites (e.g. the
respiratory tract). Thus, the major function of MALT is to initiate
immune responses, which are then passed on to draining lymph nodes,
such as mesenteric lymph nodes in the gastrointestinal tract.
The secretory IgA antibody response: The immune response in
MALT differs from that at other sites of the body in that it is
devoted to the generation of an IgA antibody response. Thus, MALT
contains precursors of IgA antibody plasma cells and populations of T
cells capable of promoting a B-cell immunoglobulin class switch into
IgA-producing B cells or plasma cells. B-Cell differentiation into
IgA-producing plasma cells after local antigen presentation is
accompanied by lymphocyte migration and specific homing. Precursors
move through draining lymph nodes into the blood and from there to the
secretory surface, where they lodge as IgA plasma cells in the lamina
propria. Specific homing mechanisms exist by which these cells are
able to select the secretory surface of their final mucosal
destination.
In contrast to IgA produced by the bone marrow and circulating in
blood, IgA synthesized by plasma cells of MALT consists of a dimeric
immunoglobulin subunit. The two monomers are linked by a polypeptide
called the J chain (about 15 kDa). These IgA antibodies have their
main effect outside the body itself, for example in salivary and
gastrointestinal secretions. The transport from the site of synthesis
across the epithelial barrier is specifically adapted for dimeric IgA,
and to a lesser extent for pentameric IgM. Epithelial cells express a
receptor for these immunoglobulins, called secretory component
(a polypeptide of about 70 kDa). After binding to this receptor, the
molecule is transported through the epithelium, possibly through its
cytoplasm, and excreted on the luminal surface. During this process,
secretory component attaches to the immunoglobulin molecule (coiled
around the Fc fragments); the composite molecule, comprising dimeric
IgA, the J chain, and secretory component, is called secretory IgA. In
rodents, a similar secretory component-mediated transport occurs in
the liver (Brown & Kloppel, 1989). Here, a secretory component on the
hepatocyte surface mediates the passage of dimeric IgA from the
sinusoids to the bile canaliculi. In this way, dimeric IgA entering
the liver by the portal vein efficiently recirculates to the bile and
from there into the gastrointestinal lumen. Secretory IgA is more
resistant to luminal conditions (especially proteolytic enzymes) than
dimeric IgA and is thus better able to function there.
IgA lacks the effector reactivity of IgM and IgG in complement
activation by the classical cascade, opsonization and phagocytosis, or
antibody-mediated cellular cytotoxicity. This lack appears to be
related to the absence of effector systems (complement, phagocytes) in
secretory fluid. The main function of IgA is to prevent the entry of
potentially pathogenic substances into the body, by a specific antigen
exclusion function in which the epithelium is coated with 'antiseptic
paint'.
Induction of immunological tolerance: A final feature of MALT
is its capacity to generate immunological tolerance. After antigenic
priming at secretory surfaces, subsequent systemic antigenic challenge
often results in nonresponsiveness (Strobel & Ferguson, 1984;
Challacombe, 1987; Holt & Sedgwick, 1987; Mowat, 1987). Suppressor
cells have been found in the spleen and suppressor factors in the
circulation after local immunization. The induction of tolerance
pertains primarily to dead microorganisms and inactivated proteins
which come into contact with the MALT. The mechanism of tolerance
induction and different responses to live and dead microorganisms is
not completely defined but is important in tolerance to food antigens
and the development of food allergies.
In summary, the host's defence in MALT is different from the
response of the internal immune system. The responses at this first
line of defence range from NK cell activity in the epithelium to
specific IgA antibody-mediated exclusion in the secretory fluid.
Immune responses to antigens entering the body at secretory sites are
initiated by lymphocytes in the epithelium, in lymphoid tissue
immediately underneath the epithelium, and in the lymph nodes that
drain the site (e.g. the mesenteric node). The liver may also
contribute to the response, as antigens passing directly into the
portal vein are efficiently removed and processed by the hepatic
mononuclear phagocyte system (Kupffer cells). Because MALT can
function independently of the internal immune system, blood analysis
alone may not provide complete information on MALT. Instead, analysis
of secretory fluids, such as saliva (for IgA antibody), bronchoalveolar
lavage fluid, and jejunal fluid, or direct investigation of the tissue
itself, are more appropriate approaches.
1.2.2.7 Skin immune system or skin-associated lymphoid tissue
As the skin is the largest organ of the body, its principal
physical function is to act as a barrier to water-soluble compounds,
to mechanical trauma, and to trauma caused by potentially pathogenic
microorganisms and the photons of sunlight. The physicochemical
characteristics of the outermost layer, the corneal or horny substance
of the epidermis, underlie the resistance to exogenous pathogenic
substances. The skin also has a host defence function that can be
designated as immunological. Some studies have suggested that the skin
might function as a primary organ (Fichtelius et al., 1970; Bos &
Kapsenberg, 1986), but most of the relevant immune reactions in the
skin appear to be antigen driven. The components of the skin immune
system, or skin-associated lymphoid tissue, are the following
(Streilein, 1983, 1990) (Figure 17): (i) Langerhans cells in the
epidermis, which are adapted for processing antigen and transporting
it to the draining lymph node, where they are called interdigitating
cells and present the antigen to lymphocytes; (ii) epider-motrophic
recirculating T lymphocyte subpopulations (homing T lymphocytes);
(iii) keratinocytes, which can synthesize cytokines after activation,
thereby influencing T-cell differentiation and haematopoiesis; they
can have an antigen-presenting function, especially after activation
resulting in MHC class II expression; (iv) Thy-1+ dendritic epidermal
cells, described in rodent skin epidermis: a special T cell that bears
the gamma-delta TCR and has an antigen-presenting function; and (v)
skin-draining lymph nodes comprising high endothelial venules through
which lymphocytes enter from the blood circulation.
Immune components exist not only in the epidermis but also in the
dermis. At this location, T cells and macrophages have preferential
distributions, especially in the papillary region. T Cells,
macrophages, mast cells, endothelial cells, and dendritic cells are
found in the connective tissue of the dermis, as in connective tissue
at other locations in the body. The reactivity of these cells in the
dermis may differ, however, from those at other locations. For
instance, skin mast cells (Van Loveren et al., 1990a) behave
differently from mast cells at other places. These immunological
components cannot always be recognized in conventional histological
preparations of skin. For instance, Langerhans cells and dendritic
epidermal cells require special immunohistological staining (Figure
17).
Various inflammatory and immune mediators are also present in
skin. These include antimicrobial peptides, complement components,
immunoglobulins, cytokines, fibrinolysins, eicosanoids, and
neuropeptides. They are partly derived from the blood circulation and
are partly of local origin (Bos, 1990). Streilein (1990) described the
function of the skin-associated lymphoid tissue as follows: induction
of primary immune responses to new cutaneous antigens, expression of
immunity to previously encountered antigens, and avoidance of
deleterious immune responses to non-threatening cutaneous antigens.
1.3 Pathophysiology
1.3.1 Susceptibility to toxic action
The dynamic nature of the immune system renders it especially
vulnerable to toxic influences. The major target sites of the immune
system for toxicity are presented in Figure 18. The reactions of
lymphoid cells are associated with gene amplification, transcription,
and translation, and compounds that affect these processes of cell
proliferation and differentiation are especially immunotoxic,
particularly to the rapidly dividing thymocytes and haematopoietic
cells of the bone marrow. Thus, the disappearance of lymphoid cells
from bone marrow, blood, and tissue, and thymus weight may be the
first and most obvious signs of toxicity, as was seen in application
of the tiered approach to assessing the immunotoxicity of pesticides,
mentioned previously (Vos & Krajnc, 1983; Vos et al., 1983a). Effects
on the constituents of the framework (stationary stroma), which
support and steer the activation, proliferation, and differentiation
of lymphoid cells, are observed less often (Krajnc-Franken et al.,
1990; Schuurman et al., 1991b). Such effects result mostly in
degeneration, ending in atrophy and fibrosis. Alternatively, framework
cells and passenger leukocytes may persist but are rendered unable to
function by the toxic insult, and the delicate interactions between
these cells may be affected. This result of toxicity is not always
observed by conventional histology; it may be visualized by changes in
immunolabelling for markers that have functional significance.
Otherwise, functional assessment in vivo or in vitro is required.
As shown in Figure 18, the effects on specific cells in lymphoid
tissues are generally reflected in altered histology of lymphoid
organs, but this is not always the case for effects on responses.
The skin, respiratory tract, and gastrointestinal tract together
form an enormous surface that is in close contact with the outside
world and is potentially exposed to a vast magnitude of microbial
agents and potential toxicants. The toxic effects on these components
of the immune system can differ in (histo)pathological manifestation
from those on internal lymphoid organs. In the human respiratory
tract, these effects include asthma, fibrosis, and pulmonary
infections. Examples of inhaled pollutants that may induce these
effects are oxidant gases and particulates such as silica, asbestos,
and coal dust. The cellular and biochemical profiles of
bronchoalveolar lavage constituents after exposure of experimental
animals and humans by inhalation (Koren et al., 1989) are valuable for
screening immune-mediated lung injury. The products of pulmonary
epithelial cells and alveolar macrophages appear to be key factors. A
number of studies have indicated that lung disease progresses with
postactivational release of cytokines, such as IL-1, tumour necrosis
factor, platelet-derived growth factor, and transforming growth
factors. Alveolar macrophages secrete not only cytokines but also a
variety of short-lived products that may contribute to altered
resistance to pulmonary infections and inflammation; these include
reactive oxygen species, such as superoxide, nitric oxide, and
hydrogen peroxide, and arachidonic acid metabolites. The overall
suppression of these humoral systems, in combination with effects on
e.g. NK cells, may predispose individuals to infectious agents or
tumour development or may alter the inflammatory and degenerative
response (Van Loveren et al., 1990b; Khan & Gupta, 1991; Denis, 1992;
Denis et al., 1993).
The skin is an important target in immunotoxicology, for instance
for chemical allergens (Kimber & Cumberbatch, 1992) and ultraviolet B
(UVB) radiation (Goettsch et al., 1993). The skin can respond to many
xenobiotics by a specific immune response (contact hypersensitivity)
or by a nonspecific inflammatory response (contact irritancy); both
responses are associated with the induction of pro-inflammatory
cytokines. The cells of the immune system are readily recruited from
the circulation to the skin in response to dermal stimulation by
xenobiotics. In addition, various resident immune cells can be
activated, for instance Langerhans cells during the induction of the
contact hypersensitivity response. Soluble mediators can be produced
locally, and antigen-antibody complexes can be formed at the site of
inflammation. Exogenous factors such as ultraviolet light (Applegate
et al., 1989) and 7,12-dimethylbenz[ a]anthracene (DMBA) (Halliday et
al., 1988) can cause the disappearance of Langerhans cells from the
skin (or the loss of their function), with consequent disturbance or
dysregulation of the skin's immune function. Keratinocytes, which
comprise the vast majority of cells in the epidermis, have an
important role in immune and inflammatory responses, serving as a
significant source of cytokines, which contribute either
quantitatively or qualitatively to the nature of the response of the
skin to exogenous stimulation.
Reactions to drugs illustrate the skin's susceptibility to toxic
influences. Cytostatic drugs used in cancer therapy often induce bone-
marrow depression as a major side-effect, resulting in an increased
risk for infections, so that blood leukocyte counts, which reflect
bone-marrow depression, must be monitored during administration.
Conversely, a number of cytotostatic drugs are immunosuppressive. A
well-known example is azathioprine (see section 2.2.1.1). A number of
new immunosuppressive drugs inhibit DNA synthesis, including
Mizoribine (Bredinin, an imidazole nucleoside), Brequinar sodium (a
quinoline carboxylic acid derivative), and RS61443 (morpholino-
ethylester of mycophenolic acid) (Thomson, 1992). Their specificity to
cells of the immune system may be related to the distinct pathways in
purine and pyrimidine metabolism that are preferentially used by
lymphocytes, such as guanine nucleotide synthesis promoted by inosine
monophosphate dehydrogenase in the case of Mizoribine and RS61443.
Another example of the particular sensitivity of the immune system to
toxic damage is its response to irradiation. Ionizing radiation is
commonly used in cancer therapy. Of the body's constituents, the
haematopoietic system is particularly sensitive to irradiation; when
pluripotent stem cells are affected, regenerative activity is lost.
Other systems destroyed by this treatment, like the intestinal
epithelium, have an intrinsic self-renewal capacity and do not need
replacement therapy. The lymphoid constituents of the immune system
differ in radiosensitivity. The dose of radiation that causes a
reduction in the cell population by a factor of 1/e (or 0.37), the
D0, has been estimated at about 0.9 Gy for bone-marrow lymphoid stem
cells, 0.4-0.9 Gy for pre-B cells, and 0.7 Gy for peripheral
lymphocytes (Anderson & Williams, 1977; Anderson et al., 1977). The
thymus is very sensitive to irradiation (Sharp & Watkins, 1981;
Anderson et al., 1986; Adkins et al., 1988). Cortical lymphocytes
manifest a D0 value of about 0.6 Gy. The intrathymic T-cell
precursor has a D0 value of about 1.4 Gy. A refractory population
(about 20-30%) is present within the medulla, but most cells are
radiosensitive (D0, about 0.7 Gy). The capacity of thymic stroma to
support precursor T-cell processing is radioresistant (Huiskamp et
al., 1988). In addition to cell depletion, peripheral lymphoid tissues
manifest decreased lymphocyte recirculation at a dose of only 0.5 Gy
(Anderson et al., 1977), suggesting that high endothelial venules, or
cell surface molecules involved in lymphocyte migration through these
venules, are radiosensitive. The sensitivity of lymphocytes to
ionizing radiation cannot be ascribed solely to their susceptibility
to death during proliferation. Cells in the resting state disappear
after irradiation, at a time that does not correspond to their
physiological half-life. Apparently, the death of lymphocytes occurs
at phases between cell division. This phenomenon, known as interphase
death, appears to be a distinct characteristic of lymphocytes and
sensitizes the immune system to radiation. Thus, only peripheral blood
lymphocytes need be assessed as dosimeters after accidental
irradiation (such as at Chernobyl in 1986).
Toxicity to the thymus: Of the lymphoid cells of the body, the
lymphocytes of the thymus (thymocytes) are especially susceptible to
the action of toxic compounds (Schuurman et al., 1992d). Table 4
presents data on the susceptibility of various thymic components to
toxic damage, illustrating the particular vulnerability of the
passenger lymphocyte population. The microenvironment appears to be
more resistant, mainly on the basis of histopathological assessment.
Thymocyte depletion, suggestive of toxicity towards this population,
may actually be an indirect effect, in that the microenvironment is
damaged and unable to support thymocyte growth. This situation may
exist in the thymus after exposure to 2,3,7,8-tetrachlorodibenzo-
para-dioxin (TCDD), which preferentially attacks the thymic
reticular epithelium, resulting in lymphocyte depletion histologically
(see section 2.2.2.1). Similarly, the reduction in the cellularity of
the medulla of the thymus after treatment with cyclosporin A is due to
the absence of the medullary lymphocyte population, but the basis of
the reduction is the disappearance of dendritic cells (as seen by
immunohistochemistry; see section 2.2.1.2).
The susceptibility of thymocytes to toxicity is also related to
their fragile composition, especially cortical thymocytes, and to the
delicate interactions between them and their microenvironment. For
instance, they are programmed to enter apoptosis when activated during
the physiological process of selection. A decrease in size or
involution of the organ may thus be the first manifestation of
toxicity. It should be noted that stress itself can induce thymic
involution; furthermore, thymic status is dependent on nutritional
status and age. The main function of the thymus is to generate the
T-cell repertoire during fetal and early postnatal life. Its
susceptibility to toxic compounds and the subsequent effects on the
cell-mediated immune system are most evident during this period of
life.
Table 4. Sensitivity of cell populations in the thymus to toxic chemicals
Cells Location Compound
Lymphocytes
Immature lymphoblasts Outer cortex Some organotin compounds,
2,3,7,8-tetrachlorodibenzo-para-dioxin
Small cells Cortex Glucocorticosteroids, cytostatic drugs
(e.g. azathioprine)
Intermediate-sized cells Medulla Ammonia caramel (THI)
Epithelial cells Cortex 2,3,7,8-Tetrachlorodibenzo-para-dioxin
Medulla Cyclosporin, FK-506
Dendritic cells Medulla Cyclosporin, FK-506
Macrophages Cortex Ammonia caramel (THI)
THI, 2-acetyl-4(5)-tetrahydroxybutylimidazole; caramel colour III
An ever-increasing number of toxic compounds has been shown to
affect the immune system. Induced and spontaneous immunopathology in
rodents have been described in reviews and textbooks (Dean et al.,
1982, 1985; Irons, 1985; Descotes, 1986; Lebish et al., 1986; Gopinath
et al., 1987; Luster et al., 1989; Vos & Luster, 1989; Jones et al.,
1990; Krajnc-Franken et al., 1990; Vos & Krajnc-Franken, 1990;
Schuurman et al., 1991b; Dean et al., 1994; Frith et al., in press).
The weight and histology of lymphoid organs are the main parameters in
evaluating toxicity. The examples given in section 2.2 illustrate the
various ways in which the immune system is injured.
1.3.2 Regeneration
The dynamic nature of the immune system provides it with a strong
regenerative capacity. In principle, the leukocyte population is
generated by a single pluripotent progenitor cell in the bone marrow.
In the thymus, single thymocyte precursors have an enormous capacity
for expansion (Ezine et al., 1984). Thus, after exposure to xenobiotic
compounds that destroy the immune system, regeneration can occur
within a relatively short time: The original architecture is restored
within three to four weeks after involution caused by radiation or
treatment with glucocorticosteroid or organotin compounds. The
restoration process can occur in waves, depending on the sensitivity
of the precursor T cells in the thymus and bone marrow (Huiskamp et
al., 1983; Penit & Ezine, 1989). Regeneration does not occur after
destruction of the white blood cell system, e.g. by sublethal
irradiation, when the stem cells in the bone marrow are affected. In
such cases, bone-marrow transplantation is required, and the anlage of
the lymphoid organs then supports generation of the newly built immune
system. After bone-marrow transplantation, polymorphonuclear
granulocytes are the first cells to appear in the circulation, within
two to three weeks; lymphocytes appear after three to four weeks;
however, establishment of a fully developed T-cell system takes about
six months (De Gast et al., 1985). These examples illustrate the
vulnerability of the bone marrow-derived component to toxic action;
regeneration requires substitution of this component and not the
relatively more resistant stromal component in lymphoid tissue.
1.3.3 Changes in lymphoid organs
The weight and gross morphology of lymphoid organs are the first
parameters studied in assessing toxicity, as the response to injury is
often expressed as changes in tissue or organ weight, size, colour,
and gross appearance. These observations are combined with leukocyte
counts, studies of differentiation, and the results of
histopathological evaluations of lymphoid organs and tissues.
Conventional histopathology allows evaluation of the effects of
xenobiotics on the main cell subsets on the basis of their distinct
morphology, tissue location, and density. In this way, the effects on
lymphocytes, lymphoblasts, and stromal cells can be evaluated
(Schuurman et al., 1994; Kuper et al., 1995). The disappearance of
lymphocytes from blood and tissues and the decrease in size and weight
are often first seen in the thymus, as this organ is very sensitive to
toxicity (see above). Evaluations of effects on distinct
subpopulations must be confirmed by immunohistochemical
characterization (Schuurman et al., 1992a). The sensitivity of
histopathology can be increased by quantitative microscopy and
morphometry and organ cell counts.
During microscopic examination, important aspects to be
considered are the cell density and the size of the various
compartments of the lymphoid organs, as well as qualitative changes
like germinal centre development; however, microscopic examination of
lymphoid organs reveals these highly dynamic and complex processes
only at a static stage. Changes in the number of cells are best
reported by descriptive terms like 'increases' or 'reductions' in
cellularity rather than by interpretive terms like 'involution',
'atrophy', or 'hyperplasia'. The pathology working group of the IPCS-
European Union international collaborative immunotoxicity study has
started to develop such a descriptive approach. A window of control
values and gross and microscopic appearance must be established for
recognition of deviations from normal; however, normal and control
values and histological features are influenced by various endogenous
and exogenous factors, including the age and hormonal (especially sex
hormone) status of the animal (Kammüller et al., 1989; Goonewardene &
Murasko, 1992; Kuper et al., 1992a; Losco, 1992; Losco & Harleman,
1992). The influence of sex hormonal status on the histology and
histophysiology of lymphoid organs is illustrated by the following
examples. Considerable changes are seen in the lymphoid organs of mice
during pregnancy (Clarke, 1984), in birds during hatching (see
Section 2), and in ectothermic vertebrates during different seasons of
the year (see section 1.2.1.5). Castration results in an increase in
thymic size in old animals (Kendall et al., 1990), as has been
observed in old rats with an involuted thymus after treatment with an
analogue of luteinizing hormone-releasing hormone (Greenstein et al.,
1987). Figure 19 shows the wet thymic weight of groups of male rats in
order to illustrate these phenomena: the mean values were about 500 mg
in young male rats at 2.5 months of age, about 125 mg in rats at 12-15
months of age, about 250 mg in aged rats after surgical castration,
and about 300 mg in aged rats after treatment with the hormone
(Kendall et al., 1990).
The neuroendocrine system and the sex steroid hormone balance may
also underlie the changes in lymphoid organs of ectothermic
vertebrates with seasonal changes like temperature and photoperiod
(mentioned above). The thymic status is also dependent on nutritional
status (Van Logten et al., 1981; Corman, 1985; Mittal et al., 1988;
Good & Lorenz, 1992), and stressful conditions influence the
appearance of lymphoid organs, especially the thymus. Finally, housing
conditions, diet, and microbiological status are important. For
instance, the presence of gamma-delta T cells in the rodent intestinal
epithelium is dependent on the presence or absence of microbiological
stimulation (that is, whether the animals are bred and maintained
under specified pathogen-free conditions) (Bandeira et al., 1990;
Dobber et al., 1992; Umesaki et al., 1993). The role of genetic
factors is reflected in strain differences in lymphoid organ
histology, e.g. in rats (Losco & Harleman, 1992; Schuurman et al.,
1992e), seen on examination of data on concurrent and historical
controls.
Genetic factors may also contribute to the individual variability
in response to a given compound. This variability can be relatively
high, even among random-bred and inbred rats, as illustrated in the
following example (Figure 20), derived from a study of the toxicity of
the immunosuppressive drug azathioprine in random-bred Wistar rats.
The data on haematological and histopathological parameters were
subjected to factor analysis in order to facilitate examination of
individual and group responses to the drug. Factor analysis involves
clustering of parameters into composite groups, or 'factors', by
arithmetic manipulation of the data, so that the parameters within a
cluster are related mathematically. It is up to the investigator to
decide whether the clustering is meaningful biologically. Figure 20
shows that not all of the 10 animals that received the highest dose of
azathioprine had the same response: three were considered to be high
responders, six low or medium responders, and one a non-responder. The
high individual variability in response illustrates the significance
of outliers in studies with a limited number of animals.
The relevance of changes observed in a study of exposure in vivo
must be based on knowledge of the nature of the response. Current
understanding of the relationship between the structure and function
of the lymphoid organs and their components often allows only a
provisional hypothesis to be made about the mechanisms of toxicity.
Moreover, different mechanisms of tissue injury can yield similar
histopathological features. Therefore, in-depth, specific histological
examination of tissue response and immune function are indispensable
for interpreting changes and for risk assessment. In interpreting
quantitative changes, like changes in cellularity, it should be noted
that particular components of the immune system may be decreased in
number or size (suppressed or involuted) or increased (stimulated or
expanded), but this does not necessarily reflect the overall effect on
the immune system or lymphoid organs.
The presence of tissue damage, protein complex deposits, and
inflammatory cell infiltrates may indicate the induction of
autoimmunity or the presence of allergy or hypersensitivity. The site
at which such responses are seen is often not a lymphoid organ, but
blood vessels, renal glomeruli, synovial membranes, thyroid, skin,
liver, or lung, which are well-known sites of autoimmunity and
hypersensitivity. Non-lymphoid organs should also be examined in order
to determine whether the effects on the lymphoid system are secondary,
for instance mediated by acute stress. This aspect is not considered
in detail in this monograph, which concerns mainly the direct toxic
action of xenobiotics on the immune system.
2. HEALTH IMPACT OF SELECTED IMMUNOTOXIC AGENTS
2.1 Description of consequences on human health
2.1.1 Consequences of immunosuppression
Since immunosuppressive drugs were introduced clinically to
prevent allograft rejection more than 25 years ago, the human
consequences of immunosuppression are well known. Infections and
cancers are the main consequences of immunosuppressive therapy, as
exemplified by a number of isolated case reports and epidemiological
studies (IARC, 1987; Descotes, 1990; IARC, 1990). The cancers are
often lymphomas and carcinomas, which are likely to be of viral
origin, especially in immunosuppressed patients. Recurrent respiratory
viral infections should also be considered as sentinel conditions for
immunotoxicity, both in individuals and in community-based epidemics,
including, but not confined to, opportunistic infections. Additional
evidence that immunosuppression can enhance the risk of cancer is the
increased incidence of an atypical form of Kaposi's sarcoma and of
lymphomas frequently located in the brains of patients with AIDS. It
is important to distinguish between profound immunodepression (mainly
seen clinically, e.g. after renal transplantation or cytotoxic therapy
for neoplasia) and the less severe suppression of immune function that
is more likely to be associated with exposure to an environmental
immunotoxic agent.
2.1.1.1 Cancer
Evidence from three sources, namely cancer patients on
chemotherapy, organ transplant patients, and patients with autoimmune
disorders undergoing long-term immunosuppressive therapy, demonstrates
that immunosuppressed patients are at a higher risk than others of
developing malignancies (Boyle et al., 1984; Penn, 1988; IARC, 1990;
Barrett et al., 1993; Bouwes Bavinck et al., 1993; Penn, 1993a,b;
Descotes & Vial, 1994), although not all immunosuppressive drugs have
been shown to be carcinogenic, e.g. prednisone and methotrexate (IARC,
1987). Immunotoxic effects might result in tumour formation through
reduced immune surveillance, i.e. tumours might escape the guard of
the immune system. Reduced immune surveillance can thus be regarded as
tumour promotion.
The risk for second malignancies after prolonged cancer
chemotherapy has been shown in numerous case reports and
epidemiological studies (Henne & Schmähl, 1985; Boivin, 1990; Blatt et
al., 1992). Acute leukaemia is the most frequently reported second
cancer (Kyle, 1984); overall, iatrogenic leukaemias account for 10% of
all leukaemias, with an incidence 5-100 times higher than in the
general population. Non-Hodgkin's lymphoma develops in 0.5-4.5% of
patients within 10 years after cytotoxic therapy. The risk for solid
tumours, e.g. carcinomas of the lung, skin, breast, colon, and
pancreas, is also increased after cytotoxic therapy but with a
different trend: the increase in risk is more prolonged and slower
(Swerdlow et al., 1992).
Although most cytotoxic drugs are genotoxic, their
immunosuppressive effects may also account for the increased risk of
second cancers, as indicated by results obtained in organ transplant
recipients. The Cincinnati Transplant Tumour Registry collected data
on more than 3600 cases of cancer in transplant patients up to June
1988 (Penn, 1988). Large numbers of cancers were also included in the
Australian and New Zealand Combined Dialysis and Transplant Registry
(Sheil et al., 1991). Overall, 1-15% of organ transplant patients
developed cancer within the first five years after transplantation;
whatever the therapeutic regimen, the incidence of cancer was at least
three times that of the general population and increased
logarithmically with the length of follow-up to reach more than 50%
after 20 years in some series. Cancers of the skin and lips were
reported in 18% of patients after 10 years of immunosuppressive
therapy. Squamous-cell carcinoma was the most frequent skin cancer and
was about 250 times more frequent in transplant patients than in the
general population (Hartevelt et al., 1990). Lymphomas accounted for
14-18% of neoplasms in transplant patients, and high-grade non-
Hodgkin's lymphomas accounted for 95% of these lymphomas. The
incidence of Kaposi's sarcoma was 400-500 times more frequent in
transplant patients than in the general population.
A similar pattern of neoplasias was observed in patients on
immunosuppressive therapy for autoimmune diseases (Sela & Shoenfeld,
1988). Non-Hodgkin's lymphomas were 11 times more frequent than in the
general population, and other cancers, namely leukaemias, primary
liver cancers, and squamous-cell carcinomas, were also found to be
more frequent (Penn, 1988; Descotes & Vial, 1994). The respective
roles of immunosuppressive therapy and of the underlying disease
remain to be established, however.
Cancers have been reported to occur after immunosuppressive
therapy with both cytotoxic and noncytotoxic drugs. Even though the
respective roles of genotoxicity and immunosuppression are difficult
to ascertain, cancers have been described in patients on azathioprine
after organ transplantation (Wessel et al., 1988; Singh et al., 1989)
or on low-dose methotrexate for autoimmune disorders (Ellman et al.,
1991; Shiroky et al., 1991; Kingsmore et al., 1992; Kamel et al.,
1993); immunosuppressive drugs increased the risk of malignancies
(especially lymphomas) in the treated patients. Interestingly, all of
the noncytotoxic immunosuppressive drugs were reportedly associated
with a variety of malignancies presumably related to immuno-
suppression. Whatever the type of tumour, the time to tumour
development after treatment with cyclosporin A was shorter (26 months
on average; 14 months for lymphomas) than after conventional
immunosuppressive therapy (60 months on average) (Penn, 1988). The
murine monoclonal antibody OKT3 was also shown to increase the risk
for lymphoproliferative disorders (Swinnen et al., 1990): Lymphomas
developed 1-18 months after starting OKT3, and a correlation was found
between the dose and time to neoplasm development. Lymphoproliferative
disorders were also shown to occur following treatment with FK 506
(Reyes et al., 1991). There are insufficient data to conclude a direct
causal relationship between immunosuppression induced by environmental
chemicals and the development of cancer; however, there is
epidemiological evidence that exposure to various potentially
immunotoxic chemicals (e.g. pesticides, benzene) is associated with
increased risks for cancers that also occur in immunosuppressed
patients (e.g. non-Hodgkin's lymphoma and leukaemia).
No marked difference was found in the relative risks for
lymphoproliferative disorders associated with the various
immunosuppressive drugs currently used, suggesting that
immunosuppression is the causative factor, particularly when account
is taken of the different carcinogenic potentials of azathioprine and
cyclosporin A, both clinical reference immunosuppressive agents
(Ryffel, 1992). Reactivation of latent viruses, e.g. Epstein-Barr
virus, due to immunosuppression was suggested to be involved. Indeed,
most lymphoproliferative disorders induced with cyclosporin A or
methotrexate were B-cell malignant lymphomas associated with this
viral infection (Starzl et al., 1984; Kamel et al., 1993). Uninhibited
proliferation of Epstein-Barr virus in B lymphocytes caused by the
efficient immunosuppression of one or more kinds of controls by
T lymphocytes is the commonly accepted mechanism. Interestingly,
Epstein-Barr virus-associated lymphomas in patients on
immunosuppressive therapy are usually reversible upon cessation of
treatment (Starzl et al., 1984).
2.1.1.2 Infectious diseases
Whatever the primary cause of the immune deficiency, patients
develop more frequent, more severe, recurring, and often atypical
infections, depending on the type and severity of the deficiency. The
complications associated with severe immunosuppression include
bacterial, viral, fungal, and parasitic infections (Waldman, 1988;
Barr et al., 1989; Mandell, 1990; Tieben et al., 1994). The pathogens
most frequently encountered in immunodeficient patients include the
bacterial agents Staphylococcus aureus, Streptococci, Escherichia
coli, Pseudomonas aeruginosa, Listeria monocytogenes, Mycobacterium
tuberculosis, and atypical mycobacteria. Herpes virus,
cytomegalovirus, Epstein-Barr virus, and human papillomavirus are the
leading causes of viral infections in immunosuppressed patients.
Fungal opportunistic infections include those induced by Candida,
Aspergillus, and Cryptococcus species. The immunotoxicity induced
by environmental chemicals often results in subtle changes in the
immune system, which have been suggested to result in increased
incidences of common infections like influenza and the common cold
(see section 2.4).
The respiratory tract is a primary target for infectious
pathogens, especially in immunosuppressed patients. Pulmonary
infections and infections of the upper respiratory tract are the most
common (Frattini & Trulock, 1993). Cytomegalovirus infections, often
asymptomatic, are particularly frequent in renal transplant patients
(Rubin, 1990). In addition, Pneumocystis carinii can cause a
particular form of pneumonia in immunosuppressed patients. Even though
respiratory diseases usually predominate, gastrointestinal infectious
diseases may constitute the leading consequences of immunosuppression
(Bodey et al., 1986). Infections of the central nervous system and
isolated fever are also extremely frequent. Interestingly, the type of
infection that develops in immunosuppressed patients is largely
dependent on the type of immune defect, as illustrated in Table 5.
Infectious complications have been commonly described in patients
treated with various cytotoxic drugs for cancer and with
immunosuppressants, such as cyclosporin A, for the prevention of
allograft rejection or the treatment of autoimmune disorders (Kim,
1989; Descotes & Vial, 1994).
Table 5. Pathogens frequently associated with immune defects
Humoral Cellular Neutrophil Complement
immunity immunity functions system
Campylobacter Candida Aspergillus Neisseria
Echovirus Coccidioides Bacteroides Staphylococci
Gardia Cryptococci Escherichia coli Streptococci
Haemophilus Cytomegalovirus Klebsiella
Pneumococci Herpes virus Pseudomonas
Salmonella
Legionella
Mycobacteria
Staphylococci
Histoplasma
Toxoplasma
Pneumocystis
Human papillomavirus
2.1.2 Consequences of immunostimulation
The health consequences of immunostimulation are less well
established than those associated with immunosuppression. A number of
adverse effects have, however, been reported after treatment with
immunostimulating drugs, including influenza-like reactions,
facilitation or exacerbation of underlying diseases, and inhibition of
hepatic drug metabolism (Descotes, 1992).
Patients with influenza-like reactions present with mild to
moderate fever associated with chills, malaise, and hypotension. The
reaction usually develops within hours after taking an
immunostimulating drug, and the patient recovers uneventfully within a
few hours. Such reactions are relatively uncommon with most
immunomodulating agents but have been shown to limit treatment with
several recombinant cytokines, e.g. IL-1 and tumour necrosis factor.
Facilitation and/or exacerbation of underlying diseases have been
ascribed to most immunostimulating drugs, but the incidence of this
adverse event differs markedly from one drug to another. Exacerbation
of chronic infections, psoriasis, and Crohn's disease have been
reported. More interestingly, several autoimmune diseases are more
frequent in patients treated with various (recombinant) cytokines,
e.g. autoimmunity treated with IFN gamma (Jacob et al., 1987),
thyroiditis with IL-2 (Vial & Descotes, 1993), and lupus erythematosus
with IFN alpha (Vial & Descotes, 1993). Exacerbation or facilitation
of allergic reactions to unrelated allergens has also been reported:
starting an immunostimulating treatment has been associated with
exacerbation of underlying eczema, asthma, or rhinitis. Allergic
reactions to radiological contrast media have been shown to be more
frequent in IL-2-treated patients than controls (Vial & Descotes,
1992).
Oxidative drug metabolism by the hepatic cytochrome P450 system
has been shown to be inhibited by immunostimulating drugs (Descotes,
1985) and by administration of bacillus Calmette-Guérin (BCG) vaccine
or interferon (Vial & Descotes, 1993). Although the mechanism of this
inhibition is still unknown, activation of macrophages resulting in
the release of IL-1 and IL-6 has been suggested to be involved.
Likewise, vaccination has been shown to compromise drug metabolism to
a sufficient extent that normally therapeutic doses of theophylline
caused acute toxicity in humans (Renton et al., 1980). Stimulation of
the immune system has also been shown to alter drug metabolism in
humans (Renton, 1986). A similar effect of infection has been reported
in laboratory animals (Selgrade et al., 1984); infection with mouse
cytomegalovirus before exposure to the insecticide parathion reduced
the total P450 concentration and dramatically increased the toxicity
of parathion.
So far, only a few environmental chemicals have been shown to
exert immunostimulating properties, e.g. hexachlorobenzene and
selenium. There have been no reports of clinical reactions to such
chemicals that are similar to the adverse effects seen with
immunostimulating drugs.
2.2 Direct immunotoxicity in laboratory animals
The following are some illustrative examples of immunotoxic
chemicals.
2.2.1 Azathioprine and cyclosporin A
The immunosuppressive effects of azathioprine and cyclosporin A
are considered because they can shed light on the direct
immunotoxicity of environmental chemicals.
2.2.1.1 Azathioprine
Azathioprine is a thiopurine that is used as cytostatic drug in
the treatment of leukaemias and as an immunosuppressant in patients
who have received allogeneic organ transplants or who have autoimmune
diseases. When used as an immunosuppressant, its main side-effect is
bone-marrow depression, reflected in blood leukocytopenia; its
administration must therefore be monitored through blood leukocyte
counts. Another side-effect, especially after long-term
administration, is tumour formation (IARC, 1987).
In rats, azathioprine is cytotoxic for all cell lineages in the
bone marrow, and strong cellular depletion is observed histologically.
It decreases the cellularity in thymus, blood, and peripheral lymphoid
organs, but it is mainly in the thymus that the immature lymphocyte
population of the cortex is affected. This effect is a general feature
of most cytostatic drugs. A similar effect is seen in the thymus after
treatment with glucocorticosteroids, but the molecular mechanism
resulting in lymphocyte depletion is obviously different: interference
with DNA synthesis resulting in lymphocyte proliferation in contrast
to binding to glucocorticosteroid receptors and cell down-modulation.
Azathioprine affects a number of indicators of immune function, like
macrophage cytotoxicity (Spreafico et al., 1987), lymphocyte
proliferation in vitro after mitogen stimulation (Weissgarten et
al., 1989) and in the mixed leukocyte reaction (Mellert et al., 1989),
and cytotoxicity by NK cells (Pedersen & Beyer, 1986; Spreafico et
al., 1987; Versluis et al., 1989). Both stimulation and suppression of
these functions have been found in experimental animals, depending on
the dosage and the time of testing after exposure. These findings are
in accordance with those in azathioprine-treated patients, who showed
no change in primary antibody response, a decrease in secondary
antibody response, and some or no effect on lymphocyte proliferation
in vitro after mitogen stimulation. The time of testing after the
start of exposure to azathioprine was a crucial factor in the
detection of effects. Azathioprine was tested in the IPCS-European
Union international collaborative immunotoxicity study (see section
1.1) and showed a significant strain-dependent sensitivity.
2.2.1.2 Cyclosporin A
Cyclosporin A is one of the most powerful immunosuppressive drugs
(Kahan, 1989). It is a neutral lipophilic cyclic peptide consisting of
11 amino acids (relative molecular mass, 1203 Da) isolated from the
fungus Tolypocladium inflatum. Its main use is in bone-marrow
transplantation to prevent transplant rejection and graft-versus-host
reactions. It is also used in the therapy of various autoimmune
diseases.
A complication of cyclosporin A treatment is nephrotoxicity.
Another side-effect, especially after long-term administration, is
tumour formation (IARC, 1987). In its immunosuppressive action,
cyclosporin A does not affect resting lymphocytes but blocks the
events occurring after stimulation, particularly the synthesis of
lymphokines, including IL-1 and IL-2, and IL-2 receptors. The
synthesis of IL-1 by antigen-presenting cells and of IL-2 by Th cells
is inhibited, and the synthesis of IFN gamma and tumour necrosis
factor is blocked. These events occur inside the cell at the
transcriptional level. Cyclosporin A binds to an intracellular
receptor, cyclophilin, forming a complex with calcineurin; this
complex in turn interferes with the activation of genes, resulting in
inhibition of lymphokine gene transcription (Baumann et al., 1992;
Sigal & Dumont, 1992).
An interesting feature of cyclosporin A is its specific action on
the thymus and the induction of autoimmune phenomena. Rats treated
with total body irradiation and syngeneic or autologous bone-marrow
transplantation, followed by treatment with cyclosporin A at a dose of
about 10 mg/kg body weight per day subcutaneously for four weeks,
developed signs of acute graft-versus-host reactions, with lymphocytic
infiltration at multiple epithelial sites (Glazier et al., 1983). A
similar pseudo-graft-versus-host reaction has also been evoked in
mice. It is associated with thymic changes, because it can be
transferred in whole thymus or thymocytes (Sakaguchi & Sakaguchi,
1988). Histologically, the medullary area is diminished (Beschorner et
al., 1987a; Schuurman et al., 1990; see also Figure 21). The medullary
stroma shows a decrease in MHC class II expression, indicating a loss
of dendritic cells, which has been confirmed by electron microscopy
(De Waal et al., 1992a). As these cells normally contribute to the
negative selection process, their depletion (or reduced MHC class II
expression) may be related to an absence of negative selection. The
autoreactive T cells may even attack the medullary epithelium.
The effect of cyclosporin A on thymic functions, i.e. the
induction of 'leakiness', with export of T cells that have not been
negatively selected, has not yet been studied for other drugs, but may
not be specific to cyclosporin A. It represents a distinct mechanism
of autoimmunity induced by the action of toxic compounds on the immune
system, mediated via thymic selection. Although the medullary area is
reduced in young rats after treatment with cyclosporin A, this is not
the case in one-year-old rats, which presumably have a lesser output
of mature T cells because of thymic involution (Beschorner et al.,
1987b).
The effect of cyclosporin A in inducing syngeneic graft-versus-
host disease in rodents has an application in clinical medicine:
Patients treated for cancer with high-dose chemotherapy and/or total
body irradiation, followed by autologous bone-marrow transplantation,
develop a recurrence of the original tumour at a higher incidence than
patients who receive an allogeneic bone-marrow transplant. This
difference has been ascribed to the addition of a graft-versus-tumour
effect to the graft-versus-host reaction. Trials have now been
initiated to induce a graft-versus-host reaction with cyclosporin A
treatment after autologous bone-marrow transplantation, in order to
reduce tumour recurrence. The initial results are promising (Hess et
al., 1992; Yeager et al., 1993; Kennedy et al., 1994).
Interestingly, two other immunosuppressive drugs, FK-506 and
rapamycin, which also interfere with gene activation in T lymphocytes,
do not bind to cyclophilin but to another intracellular receptor, the
FK-binding protein. The effect of FK-506 on the thymus is similar to
that of cyclosporin A, i.e. a decrease in the medulla (Pugh-Humphreys
et al., 1990), and rapamycin causes severe acute involution with
disappearance of lymphocytes from the cortex (Zheng et al., 1991).
These findings indicate that the two compounds have different
molecular mechanisms of action on the thymus from those of cyclosporin
A, which have not yet been elucidated.
2.2.2 Halogenated hydrocarbons
2.2.2.1 2,3,7,8-Tetrachlorodibenzo-para-dioxin
The halogenated hydrocarbon most closely studied for its
immunotoxic effects is TCDD. It has a variety of toxic effects, with a
remarkable interspecies variation; however, it causes atrophy of the
thymus and immunotoxicity in all species investigated (Vos & Luster,
1989; Holsapple et al., 1991; Neubert, 1992; Kerkvliet & Burleson,
1994). Atrophy of the thymus is reflected histologically by lymphocyte
depletion of the cortex (Figure 22). Functionally, cell-mediated
immunity appears to be suppressed in a dose-dependent fashion, as
manifested in delayed-type hypersensitivity responses, rejection of
allogeneic skin transplants, graft-versus-host reactivity, and
lymphocyte proliferation in vitro after mitogen stimulation. This
immune suppression is age-related: more severe immunotoxic effects are
observed after perinatal administration than after administration in
adulthood (Vos & Moore, 1974; Thomas & Hinsdill, 1979). TCDD can also
impair antibody-mediated immunity after primary or secondary
immunization. A sensitive parameter of the immunotoxicity of TCDD and
TCDD congeners in mice is suppression of the T cell-dependent antibody
response to sheep red blood cells in mice (Vecchi et al., 1980; Davis
& Safe, 1988; Kerkvliet et al., 1990). No effects have been observed
on classical macrophage functions.
In mice, susceptibility to TCDD is genetically determined and is
segregated at the locus that encodes a cytosolic protein which
mediates aryl hydrocarbon hydroxylase activity (Poland & Knutson,
1982). This Ah (aromatic hydrocarbon) receptor has a high affinity for
TCDD and is strongly active in mouse and rat thymus (Gasiewicz &
Rucci, 1984), particularly in epithelial cells (Greenlee et al., 1985;
Cook et al., 1987). Ah receptor-dependent immunotoxicity has been
demonstrated in mice for thymic atrophy and the antibody response to
sheep red blood cells (Tucker et al., 1986; Kerkvliet & Burleson,
1994); however, the importance of Ah receptor-mediated events in
chronic, low-level TCDD immunotoxicity is controversial (Morris et
al., 1992).
Immunosuppression in adult mice manifests almost exclusively as
suppressed antibody responses and does not appear to be related to
thymic atrophy in experiments in thymectomized (Tucker et al., 1986)
and nude (Kerkvliet & Brauner, 1987) mice. Both T and B lymphocytes
involved in antibody responses can, however, be affected by TCDD. For
example, exposure to TCDD in vivo alters regulatory lymphocyte
function (Kerkvliet & Brauner, 1987) and antigen-specific T lymphocyte
activation (Lundberg et al., 1992). TCDD also inhibits T-independent
antigen responses (Vecchi et al., 1983) and T-dependent responses when
only B cells are treated (Dooley & Holsapple, 1988). Studies of the
effects of TCDD on enriched B-cell populations in vitro have shown
that it selectively inhibits late stages of the cell cycle and the
development of B cells into plasma cells after antigen-specific
activation (Luster et al., 1988). The molecular events responsible for
TCDD immunosuppression have not been examined in detail. While early
events in B-cell maturation, such as inositol phosphate accumulation,
are not affected (Luster et al. 1988), activation of protein kinase
(Kramer et al., 1987) and tyrosine kinase (Clark et al., 1991) have
been observed.
A consequence of TCDD-induced immunosuppression is impaired
resistance to infection by bacterial, viral, and protozoan
microorganisms (Vos et al., 1991). In various mouse strains with
different treatment schedules, TCDD suppressed resistance to models
of infectious diseases with Salmonella bern, S. typhimurium,
Streptococcus pneumoniae, herpes II, Plasmodium yoelli and influenza
viruses. Various effects have been reported on resistance to
L. monocytogenes. TCDD had no effect on the mortality of mice
infected with Herpes suis (pseudorabies), whereas the mortality of
mice infected with influenza virus was enhanced by a single oral dose
of TCDD as low as 10 ng/kg body weight (Burleson et al., in press).
Many studies have been performed to investigate the mechanisms of
TCDD-induced thymic atrophy, and a number have presented evidence that
the effect may occur through an action on epithelial cells:
1. The enhanced lymphoproliferation of thymocytes after coculture
with cultured mouse and human epithelial cells was reduced when
the epithelial cells were pretreated with TCDD (Greenlee et al.,
1985; Cook et al., 1987).
2. In mouse radiation chimaeras, TCDD-induced suppression of Tc
lymphocytes is determined by the host (epithelium) and not the
donor (bone marrow, subsequently thymocytes) (Nagarkatti et al.,
1984).
3. Histological and electron microscopy studies of TCDD-exposed rats
reveal formation of epithelial aggregates and a more
differentiated state of cortical epithelium, indicating that TCDD
acts on the thymic epithelium (De Waal et al., 1992b, 1993).
A direct action of TCDD on rat thymocytes has also been
documented in vitro as cell death due to apoptosis (McConkey et al.,
1988), but this effect requires higher concentrations than those that
affect epithelial cell function in vitro. In bone marrow, TCDD
affects myelopoiesis (Luster et al., 1985a) but may be more selective
for prothymocytes (Fine et al., 1989, 1990; Holladay et al., 1991;
Blaylock et al., 1992), thus indirectly affecting thymic function.
2.2.2.2 Polychlorinated biphenyls
Polychlorinated biphenyls (PCBs) are important environmental
chemicals shown in numerous studies to have immunotoxic properties.
PCB mixtures alter several morphological and functional aspects of the
immune system in rodents, guinea-pigs, rabbits, and chickens (Vos &
Luster, 1989). The first suggestion that PCBs might affect the immune
system came from observations on the weight and histology of lymphoid
organs. Oral exposure of chickens to PCBs resulted in small spleens
(Flick et al., 1965) and atrophy of lymphoid tissue (Vos & Koeman,
1970). Similar effects were noted in rabbits and guinea-pigs
(Figure 23). Dermal application of PCBs to rabbits caused lymphopenia,
atrophy of the thymic cortex, and a reduced number of germinal centres
in spleen and lymph nodes (Vos & Beems, 1971). Oral exposure of guinea-
pigs significantly reduced the number of circulating lymphocytes
and the relative thymus weight (Vos & Van Driel-Grootenhuis, 1972).
Functional tests have been focused on humoral immune responses.
Exposure of guinea-pigs, rabbits, mice, and rats to PCBs at different
regimens reduced antibody production to foreign antigens, including
tetanus toxoid, pseudorabies virus, sheep red blood cells, and keyhole
limpet haemocyanin (Vos & Van Driel-Grootenhuis, 1972; Koller &
Thigpen, 1973; Loose et al., 1977; Wierda et al., 1981; Exon, 1985;
Kunita et al., 1985). These data are in line with the observations of
Loose et al. (1977) and Thomas & Hinsdill (1978) that exposure to PCBs
lowered circulating immunoglobulin levels in mice. No reduction was
reported in antibody responses to bovine serum albumin (Talcott &
Koller, 1983).
The response to sheep red blood cells in the plaque-forming cell
assay has been used to establish dose-response relationships for
several potentially immunotoxic Aroclors in mice given a single
intraperitoneal injection of PCB mixtures (Davis & Safe, 1989). These
studies indicate that the higher chlorinated PCB mixtures are more
immunotoxic than the lower chlorinated Aroclors (Allen & Abrahamson,
1973; Loose et al., 1978; Tryphonas, in press). Data on the effects of
PCBs on total serum immunoglobulin levels have not been reported in
non-immunized animals.
While the suppressive effects of PCBs on humoral immunity are
well documented, the effects on cell-mediated immune parameters are
less clear. The delayed-type hypersensitivity reaction to tuberculin
was suppressed in guinea-pigs (Vos & Van Driel-Grootenhuis, 1972) but
not in rabbits treated with PCBs (Street & Sharma, 1975). Decreased
delayed-type hypersensitivity reactions were reported in mice by Smith
et al. (1978) but not by others (Talcott & Koller, 1983). Kerkvliet &
Baecher-Steppan (1988) reported that 3,4,5,3',4',5'-hexachlorobiphenyl
reduced Tc lymphocyte activity in the spleens of mice. In contrast,
the graft-versus-host reaction was increased following PCB treatment
(Carter & Clancy, 1980). Studies on the mitogen-induced responses of
splenic mononuclear leukocytes from PCB-treated mice in vitro
resulted in either enhanced or unaltered responses (Bonnyns &
Bastomsky, 1976; Wierda et al., 1981; Davis & Safe, 1989; Smialowicz
et al., 1989).
Functional impairment of the non-specific resistance of macro-
phages has been reported, including reduced phagocytic activity and
clearance of pathogenic bacteria by the spleens and livers of PCB-
exposed animals (Smith et al., 1978) and decreased NK cell activity
(Talcott et al., 1985; Smialowicz et al., 1989). Exposure of mice to
PCBs also enhanced their sensitivity to endotoxin shock (Loose et al.,
1978; Thomas & Hinsdill, 1978).
PCB treatment was shown to protect mice and rats against
Ehrlich's tumour (Keck, 1981) and Walker 256 tumours (Kerkvliet &
Kimeldorf, 1977), shown as reduced tumour growth and metastasis after
transplantation; in other studies, however, no influence of PCB on
tumour-cell implants was reported (Koller, 1977; Loose et al., 1981).
PCBs also affect the resistance of animals to infectious diseases.
Thus, ducklings exposed to low levels of PCBs were more susceptible to
challenge with duck hepatitis virus (Friend & Trainer, 1970), and mice
were more susceptible to challenge with Moloney leukaemia virus
(Koller, 1977), Plasmodium berghei (Loose et al., 1978),
S. typhimurium (Loose et al., 1978), L. monocytogenes (Thomas &
Hinsdill, 1978), and Herpes simplex and Ectromelia viruses
(Imanishi et al., 1980).
The immunotoxic effects of PCBs have also been investigated in a
number of studies with non-human primates. Decreased titres of anti-
sheep red blood cells have been observed in PCB-exposed rhesus (Thomas
& Hinsdill, 1978) and cynomolgus monkeys (Hori et al., 1982; Truelove
et al., 1982; Kunita et al., 1985). Immunotoxic effects were also
reported in adult female rhesus monkeys and their infants (exposed
in utero and through lactation) after low-level exposure (Tryphonas
et al., 1989, 1991a,b). In this study, five groups of female rhesus
monkeys were administered PCB (Aroclor 1254) at 0, 5.0, 20.0, 40.0, or
80.0 µg/kg body weight per day orally. Immunological effects were
reported after both 23 and 55 months of exposure and comprised
significantly decreased IgM and IgG responses to sheep red blood cells
at the lowest dose. Alterations in T-cell subsets were reported in the
group receiving the high dose in comparison with the controls, which
were characterized by an increase in Ts/Tc (CD8) cells and a reduction
in the relative numbers of Th/inducer cells (CD4) and in the CD4:CD8
ratio. No effects were seen on total lymphocytes or on B cells or on
total serum IgG, IgM, and IgA levels. A further study indicated that
Aroclor 1254 had no effect on B lymphocytes, since antibody responses
to T-independent pneumoccocal antigen were not significantly affected.
A trend for reduced incorporation of 3H-thymidine by mitogen-induced
lymphocyte proliferation was noted only for the T mitogens
phytohaemagglutinin and concanavalin A and not for the B pokeweed
mitogen. A significant augmentation of NK cell activity was noted at
the highest dose. Total serum complement activity (CH50) was also
increased. The serum levels of corticosteroids (hydrocortisone), which
were measured throughout the study, were not affected by treatment
(Loo et al., 1989), clearly indicating that the changes in several of
the immune parameters were direct effects of Aroclor 1254 on the
immune system.
2.2.2.3 Hexachlorobenzene
Hexachlorobenzene (HCB) is a highly persistent chemical which was
used in the past as a fungicide. Emissions to the environment now
occur owing to its use as a chemical intermediate and its presence as
a by-product in several chemical processes. It is an immunotoxic
compound (Vos, 1986), with different effects in rats and mice. In
rats, the main changes seen after subacute exposure are increased
weights of spleen and lymph nodes; the serum levels of IgM are also
increased. Histologically, the spleen shows hyperplasia of follicles
and the marginal zone (Figure 24); the lymph nodes have more follicles
with germinal centres and greater proportions of high endothelial
venules, indicating activation (Figures 25 and 26). High endothelial-
type venules are also induced in the lung (Figure 27), and macrophages
accumulate in lung alveoli (Kitchin et al., 1982; Vos, 1986; see also
Figure 28).
Functional assessment showed an increase in cell-mediated
immunity (delayed-type hypersensitivity) and an even greater increase
in antibody-mediated immunity (primary and secondary antibody response
to tetanus toxoid). Macrophage functions were unaltered. Stimulation
of immune reactivity occurs at dietary levels as low as 4 mg/kg after
combined pre- and postnatal exposure for six weeks, whereas the
conventional parameters of hepatotoxicity are not altered at this dose
(Vos et al., 1979a; Vos, 1986). The developing immune system of the
rat therefore seems to be particularly vulnerable to the immunotoxic
action of HCB. Reduced NK cell activity has also been found in the
lung after oral exposure to 150-450 mg/kg HCB in the diet (Van Loveren
et al., 1990c).
Studies on the mechanism of action of HCB indicate a role for
T cells: congenitally athymic rnu/rnu rats, which lack T cells, do
not manifest the hyperplasia of B lymphocytes in splenic follicles and
the marginal zone after administration of the compound; but
endothelial cell proliferation and macrophage accumulation in the lung
are apparently T cell-independent, as these effects were seen in
athymic animals (Vos et al., 1990b). In contrast to the
immunostimulatory effect in rats, HCB suppresses cell-mediated and
antibody-mediated immunity in mice, as well as their resistance to
protozoan infections ( Leishmania and Plasmodium berghei) and to
inoculated tumours (Loose et al., 1977, 1978, 1981). The
susceptibility of mice to HCB is also higher after pre-or perinatal
administration (Barnett et al., 1987). Recent studies indicate that
the immunostimulatory effect of HCB in rats may be related to
autoimmunity:
1. Exposure to HCB of Lewis rats, which develop autoimmune disease
after sensitization with complete Freund's adjuvant (adjuvant
arthritis) or with guinea-pig myelin (experimental allergic
encephalomyelitis), had clear effects (Van Loveren et al., 1990c):
Whereas the allergic encephalomyelitis response was severely enhanced,
the arthritic lesions were strongly suppressed.
2. Wistar rats treated with HCB produce antibodies to autoantigens;
thus, IgM, but not IgG, levels against single-stranded DNA, native
DNA, rat IgG (representing rheumatoid factor), and bromelain-treated
mouse erythrocytes (indicating that phosphatidylcholine is a major
autoantigen) were elevated. It has been suggested that HCB activates a
B-cell subset committed to the production of these autoantibodies and
associated with various systemic autoimmune diseases (Schielen et al.,
1993).
2.2.3 Pesticides
A large number of studies have focused on the immunotoxicity of
pesticides. Because of the chemical heterogeneity of these compounds
as a class, the reported effects vary widely (Barnett & Rodgers,
1994).
2.2.3.1 Organochlorine pesticides
The evidence for the immunotoxicity of organochlorine pesticides
as a class is inconclusive.
DDT: Wistar rats treated with 40 mg/kg body weight per day DDT
orally for 60 days showed increased anti-bovine serum albumin titres
(Lukic et al., 1973). Studies by Vos et al. (Vos & Krajnc, 1983, Vos
et al., 1983a), however, showed no changes in thymus or spleen
weights, leukocyte counts, or total serum IgG and IgM levels at doses
up to 800 mg/kg body weight per day.
Chlordane: Prenatal exposure of mice to chlordane was reported
to reduce contact and delayed hypersensitivity responses, suggesting
an effect on T-cell responses. Attempts to elucidate the mechanism
have, however, been unsuccessful. Johnson et al. (1986) observed
increased lymphocyte proliferation only at a dose of 8 mg/kg body
weight in B6C3F1 mice and concluded that chlordane has no significant
immunotoxicity in this model.
Chlordecone: Chlordecone reduced thymus and spleen weights by
40% in Fischer rats at at a dose of 10 mg/kg body weight per day but
had no significant effect at or below 5 mg/kg body weight per day.
T-Lymphocyte proliferation was unaffected at all doses (Smialowicz et
al., 1985a).
Lindane: Lindane had various effects on the anti-sheep red
blood cell response, depending on the immunization protocol. Specific
IgM levels were unchanged by parenteral immunization after four weeks'
treatment with 150 mg/kg in the diet, but specific IgG2b levels were
raised after intragastric immunization (André et al., 1983); however,
the duration of Giardia muris infection was significantly prolonged.
Five weeks' oral treatment with up to 12 mg/kg of diet decreased the
antibody titre to TY3 vaccine in rabbits in a dose-dependent manner
(Desi et al., 1978).
Toxaphene: Toxaphene given at 100 or 200 mg/kg of diet
decreased anti-bovine serum albumin antibody titres in Swiss mice
treated for eight weeks and in the offspring of dams given the same
diet. Macrophage activity was also reduced in these offspring, but
there were no changes in the delayed-type hypersensitivity response to
purified protein derivative (Allen et al., 1983).
Endosulfan: Endosulfan had no immunotoxic effect in Wistar rats
(Vos & Krajnc, 1983; Vos et al., 1983a).
2.2.3.2 Organophosphorus compounds
Single doses of the insecticides parathion, malathion, and
dichlorvos cause significant reductions in anti-sheep red blood cell
plaque-forming cell responses (Casale et al., 1983, 1984). The
relevance of these findings is questionable, however, as they occurred
only if cholinergic or parathion symptoms were also induced.
Administration of multiple doses of malathion resulted in conflicting
findings: C57Bl6 mice given four doses of 240 mg/kg body weight over
eight days had unchanged plaque-forming cell responses to sheep red
blood cells. In contrast, rabbits treated with 5-10 mg/kg body weight
per day over 5-6 weeks had reduced antibody titres after vaccination
with S. typhimurium. Parathion also failed to suppress anti-sheep
red blood cell plaque-forming cell formation when given as four doses
of 4 mg/kg body weight (Casale et al., 1983). Parathion-methyl given
to rabbits for four weeks did not affect immune responses (Desi et
al., 1978). In mice, however, both cellular and humoral responses were
reported to be suppressed by subacute administration of parathion
(Wiltrout et al., 1978).
The immunotoxicity of MPT-IP (the industrial compound for the
production of Wofatox EC50, containing 60% parathion-methyl) was
studied in mice given single oral doses of 8.9 mg/kg body weight or
repeated doses of 0.890 or 0.445 mg/kg body weight for four weeks.
Depending on the day of administration, the single dose increased the
IgM plaque-forming cell content of the spleen and the serum anti-sheep
red blood cell antibody titre. In the subacute system, the smaller
dose (0.445 mg/kg) increased the splenic plaque-forming cell content
and serum antibody titre (Institoris et al., 1992).
Dimethoate was administered by gavage to three generations of
Wistar rats at doses of 14.1, 9.39, and 7.04 mg/kg body weight
(equivalent to 1/50, 1/75, and 1/100 of the LD50), and parathion-
methyl was administered at doses of 0.436, 0.291, a,d 0.218 mg/kg body
weight. The highest dose of dimethoate significantly decreased the
plaque-forming capacity of spleen cells in the first generation and
increased thymic weight in the second and third generations. All three
doses of parathion-methyl decreased the number of red blood cells and
the haematocrit value, and the two highest doses decreased the
leukocyte count. The nucleated cell content of the bone marrow was
increased in the second and third generations, and decreased relative
thymic weight was seen at all three doses in the third generation
(Institoris et al., 1995). In a similar experiment, dichlorvos was
administered at doses of 1.85, 1.24, or 0.972 mg/kg body weight. A
significant decrease in leukocyte count, lowered spleen cellularity,
and decreased plaque-forming capacity were seen with the highest dose
in the second generation. In the third generation, there was a dose-
dependent decrease in femoral bone-marrow cellularity (Institoris et
al., in press). In vitro, 250 µmol/litre of paraoxon, a parathion
metabolite, suppressed mitogenic lymphocyte proliferation in spleen
cells from Sprague-Dawley rats (Pruett & Chambers, 1988).
Reduction of antibody titre against Ty3 vaccine was observed by
the end of six weeks' oral treatment of rabbits with 5-100 mg/kg body
weight of malathion or with 1.25 or 2.5 mg/kg body weight of
dichlorphos (Desi et al., 1978). In the same system, a dose-dependent
decrease was observed in the tuberculin skin reaction after
administration of 0.31, 0.62, or 1.25 mg/kg body weight of
dichlorphos. The cholinesterase activity of red blood cells was
decreased only by the two higher doses.
Convincing evidence for immunotoxicity has been obtained only for
O,O,S-trimethylphosphorothiate ( O,O,S-TMP), a contaminant of
various commercial organophosphorus formulations, such as malathion,
fenitrothion, and acephate. This compound was shown to suppress
humoral and cellular immunity in mice exposed to 10 mg/kg body weight
orally (Devens et al., 1985). Several organophosphorus derivatives can
alter some immune functions in vitro, including mitogen-induced
lymphocyte proliferation (Pruett & Chambers, 1988), T-lymphocyte
cytotoxicity, and production of hydrogen peroxide by macrophages
(Pruett, 1992), at concentrations that can theoretically be attained
in vivo.
Several mechanisms have been proposed to explain organo-
phosphorus-induced immunosuppression (Pruett, 1992). A direct
cholinergic mechanism is unlikely to be involved, as the addition of
various cholinergic agonists does not suppress immune responses in
vitro. In addition, O,O,S-trimethylphosphorodithioate, a structural
analogue of O,O,S-TMP, modulates cholinesterase activity but does
not alter immune competence. An indirect mechanism involving
stress caused by neurotoxicity has also been proposed. Finally, a
direct action on cells of the immune system, and particularly
macrophages, has been suggested to be involved. Mice treated with
O,O,S-TMP, which is not neurotoxic, generate a population of
macrophages, contraindicating lymphocyte proliferation. Antigen
processing and presentation by these highly activated (inflammatory)
macrophages are severely impaired; however, the changes in macrophage
function are not correlated with suppression of humoral or cellular
immunity. While there is no direct evidence that B and T lymphocytes
are the predominant targets of organophosphorus compounds, their
mechanisms of action on macrophages are largely unknown.
2.2.3.3 Pyrethroids
Dose-dependent decreases in the serum anti- S. typhimurium
antibody titre and in the tuberculin skin reaction were observed in
rabbits fed 25, 12.5, or 6.25 mg/kg body weight of technical-grade
cypermethrin (93.5%) for seven weeks (Desi et al., 1985). Single oral
doses (23.5, 20.7, or 18.7 mg/kg body weight) of supermethrin, the
active substance of the pyrethroid pesticide Neramethrin EC 50,
decreased the number of IgM plaque-forming cells in the spleens of
mice but had no effect on the delayed-type hypersensitivity reaction.
Repeated doses of 2.97, 1.49, and 0.743 mg/kg body weight caused only
slight changes in the leukocyte count and in the nucleated cell
content of femoral bone marrow (Siroki et al., 1994).
2.2.3.4 Carbamates
Carbaryl induced marked increases in serum IgG1 and IgG2, but not
IgA, IgG3, or IgM, levels of mice exposed to 150 mg/kg of diet for one
month (André et al., 1983). Rabbits given carbaryl at 4-150 mg/kg of
diet for four weeks had no changes in anti-sheep red blood cell
haemolysin or haemagglutinin titres or in the delayed-type
hypersensitivity response to tuberculin, whereas oral treatment with
carbofuran at 0.5-20 mg/kg of diet for four weeks induced a 60-75%
decrease in the delayed-type hypersensitivity response (Street &
Sharma, 1975). Aldicarb induced no changes in a large battery of
assays for immune function and host resistance in B6C3F1 mice exposed
to 0.1-1000 mg/litre of drinking-water daily for 34 days (Thomas et
al., 1987).
2.2.3.5 Dinocap
Dinocap is a dinitrophenol compound used as a fungicide. Female
C57Bl/6J mice were given doses of 12.5-50 mg/kg body weight per day by
gavage for 7 or 12 days. All mice given the highest dose died after
four days. Mice given 25 mg/kg for 12 days had decreased thymus
weights and cellularity and increased spleen weights but no changes in
body weight, leukocyte count, lymphoproliferative response to B- or
T-cell mitogens, mixed lymphocyte reaction, or NK cell activity of
spleen cells; lymphoproliferative responses to concanavalin A and
phytohaemagglutinin in thymocytes were reduced. In mice exposed for
seven days to 25 mg/kg body weight per day, the cytotoxic T lymphocyte
response to P815 mastocytoma cells was enhanced, and there was a
significant reduction in the IgM and IgC plaque-forming cell response
to sheep red blood cells. In vitro in murine thymocytes, a
concentration of 10 µg/ml dinocap for 72h suppressed the proliferative
response to concanavalin A and phytohaemagglutinin; exposure for as
little as 30 min suppressed the mitogen-stimulated response with no
direct cytotoxicity (Smialowicz et al., 1992a).
2.2.4 Polycyclic aromatic hydrocarbons
A major concern for human health is the carcinogenic potential of
most polycyclic aromatic hydrocarbons (PAHs). Interestingly, those
which are carcinogenic also have potent immunosuppressive properties,
whereas those which are not carcinogenic lack marked immunotoxic
effects (Ward et al., 1985; White, 1986). Suppression of humoral
immunity has been observed frequently after exposure to a number of
PAHs, including benzo[ a]pyrene, DMBA, and 3-methylcholanthrene (Ward
et al., 1985). Structure-activity studies by White et al. (1985), in
which the antibody-forming cell response was used to evaluate 10 PAHs
in B6C3F1 and DBA/2 mice, demonstrated a wide spectrum of activity:
compounds like benzo[ e]pyrene and perylene were not immunotoxic,
whereas dibenz[ a,h)anthracene and DMBA were potent immunosuppressors
of the plaque-forming cell response. Interestingly, the DBA/2 mice
were more susceptible to the immunosuppressive effects than the B6C3F1
mice.
PAHs also suppress cell-mediated immunity. T-Lymphocyte
cytotoxicity and mixed lymphocyte responsiveness were found to be
impaired by most PAHs. Differences between PAHs are seen, however, in
that benzo[ a]pyrene may be less suppressive of cell-mediated
immunity than DMBA, accounting for the greater host susceptibility to
L. monocytogenes and PYB6 sarcoma challenges in DMBA- than in
benzo[ a]pyrene-treated rodents (Ward et al., 1985). Thurmond et al.
(1987) evaluated immunosuppression in B6C3F1 ( Ah-responsive) and
DBA/2 ( Ah-nonresponsive) mice and in Ah-congenic C57Bl/6J
(responsive B6-AhbAhd and nonresponsive B6-AhdAhd) mice after
exposure to DMBA in a battery of immunological assays, including
evaluation of organ weights, plaque-forming cell response, mitogen
responses, and mixed lymphocyte responses. The authors concluded that
the immunosuppressive action of DMBA was independent of the Ah locus
and associated induction of cytochrome P1-450 metabolizing enzymes.
The mechanisms of PAH-mediated immunosuppression remain to be
elucidated. PAHs may exert their immunotoxic effects as the parent
compound or as metabolites. In vitro many of the metabolites of
benzo[ a]pyrene and DMBA are immunosuppressive, the diol metabolites
being the most potent (Kawabata & White, 1987; Ladics et al., 1991).
Several possible mechanisms of action have been proposed, including
altered interleukin levels (Lyte & Bick, 1986; Pallardy et al., 1989),
a direct effect on transmembrane signalling (Pallardy et al., 1992),
and alterations in intracellular calcium mobilization (Burchiel et
al., 1991; Davis & Burchiel, 1992).
Earlier studies suggested that Th cells or faulty antigen
recognition by T cells were possible mechanisms of DMBA-induced
immunosuppression (House et al., 1987, 1989). Myers et al. (1987) also
reported that benzo[ a]pyrene alters macrophage antigen presentation.
Studies by Ladics et al. (1992) demonstrated that the only splenic
cell type capable of metabolizing benzo[ a]pyrene was the macrophage
and that the predominant immunosuppressive metabolite formed was the
benzo[ a]pyrene-7,8 diol epoxide, which is also believed to be the
ultimate carcinogenic metabolite of benzo[ a]pyrene.
2.2.5 Solvents
2.2.5.1 Benzene
Exposure to benzene is associated with myelotoxicity, and a
strong correlation was noted between lymphocytopenia and abnormal
immunological parameters. The myelotoxicity may be due, in part, to
altered differentiation of marrow lymphoid cells, as suggested by the
finding that acute exposure of IgM+ cell-depleted marrow cultures to
hydroquinone, an oxidative metabolite of benzene, blocked the final
maturation stages of B-cell differentiation (King et al., 1987). In
addition, it was shown that the hydroquinone metabolite inhibits
lectin-stimulated lymphocyte agglutination and mitogenesis by reacting
with intracellular sulfhydryl groups (Pfeifer & Irons, 1981).
Immunosuppression associated with exposure to benzene was found
in rabbits to be an impaired antibody response together with an
increased susceptibility to tuberculosis and pneumonia. Similarly,
C57Bl/6 mice exposed to benzene had a lower antibody response and
reduced mitogen-induced lymphocyte proliferation (Wierda et al.,
1981). Chronic inhalation of concentrations as low as 30 ppm impaired
resistance to L. monocytogenes (Rosenthal & Snyder, 1985).
Similarly, increased susceptibility to PYB6 tumour cell challenge was
seen at concentrations that also impaired Tc lymphocyte function.
The mechanism of benzene-induced immunosuppression is unclear.
Cellular depletion may be the major effect, although B- and T-cell
dysfunction may also be involved. The antiproliferative effects of
benzene may be related to its ability to alter cytoskeletal
development through inhibition of microtubule assembly. Polyhydroxy
metabolites of benzene ( para-benzoquinone and hydroquinone) have
been shown to bind to sulfhydryl groups on the proteins necessary for
the integrity and polymerization of microtubules. This effect may
alter cell membrane fluidity and may explain the sublethal effect of
benzene on lymphocyte function.
2.2.5.2 Other solvents
Hexanediol (1.2 mg/kg per day for seven days) decreased thymus
and spleen weights, antibody production, and delayed-type
hypersensitivity in mice (Kannan et al., 1985). Humoral immunity was
suppressed to a greater extent in female than in male mice after a
four-month exposure to trichloroethylene in the drinking-water at
doses of 0.1, 1.0, 2.5, or 5.0 mg/ml; cell-mediated immunity and bone-
marrow stem-cell colonization were inhibited only in females (Sanders
et al., 1982). The immunotoxicity of glycol ethers and some of their
metabolites has been studied in rats by measuring the plaque-forming
cell response to trinitrophenyl lipopolysaccaride. The glycol ethers
2-methoxyethanol and 2-methoxyethylacetate were immunosuppressive, as
was the principal metabolite of the latter, 2-methoxyacetic acid. The
glycol ethers 2-(2-methoxyethoxy)ethanol, bis(2-methoxyethyl) ether,
2-ethoxyethanol and its principal metabolite 2-ethoxyacetic acid,
2-ethoxyethyl acetate, and 2-butoxyethanol were not immunosuppressive
(Smialowicz et al., 1991, 1992b, 1993)
Dichloroethylene did not induce immunotoxic changes in mice given
up to 2 mg/litre per day for 90 days (Shopp et al., 1985). Similarly
negative findings were obtained with trichloroethane (Sanders et al.,
1985).
2.2.6 Metals
Heavy metals have been shown to alter immune responsiveness in
laboratory animals (Koller, 1980). Alterations in B lymphocyte
function have been observed most frequently after exposure to lead and
cadmium, but T-cell and macrophage changes have also been described.
In addition, exposure to metals is correlated better with impaired
resistance to experimental infections than with changes in B-
or T-cell functions. Interestingly, immunostimulation has been
shown to occur at levels of exposure lower than those associated with
immunosuppression. Metals have also been shown to induce immuno-
potentiation, at lower doses than those that cause immunosuppression.
2.2.6.1 Cadmium
Conflicting results have been obtained with regard to the effect
of cadmium on humoral immunity in animals (Descotes et al., 1990).
Cell-mediated immunity, however, is consistently depressed after both
short- and long-term exposure, and phagocytosis and NK cell activity
are found to be depressed. Susceptibility to L. monocytogenes, Herpes
simplex 1 and 2, and influenza virus was increased in B6C3F1 mice
exposed for long periods (Thomas et al., 1985a).
2.2.6.2 Lead
Experimental studies suggest that lead has immunosuppressive
effects in rodents (Lawrence, 1985; Descotes et al., 1990; Koller,
1990). Early studies demonstrated that lead can suppress the humoral
immune response of mice exposed as adults (Koller & Kovacic, 1974) and
of rats exposed pre- and postnatally (Luster et al., 1978). In
contrast, no change in humoral immunity was found in mice exposed to
0.08-10 mmol/litre in drinking-water (Lawrence, 1981) or after a
10-week oral treatment with 13, 130, or 1300 mg/kg of diet (as lead
acetate) (Koller & Roan, 1980). Delayed-type hypersensitivity was
found to be depressed by lead acetate and lead chloride but not in
mice treated with lead oxide, nitrate, or carbonate. The most
consistent finding in experimental studies of the effects of lead on
host resistance, however, is increased susceptibility to infectious
agents (McCabe, 1994).
With respect to nonspecific host defence mechanisms, mice treated
with lead at doses of 5, 10, or 20 µg/kg body weight given intra-
peritoneally once or at doses of 25, 50, or 100 µg/kg body weight
given orally once showed increased clearance of colloidal carbon
(Schlick & Friedberg, 1981). Furthermore, treatment of mice with 130
or 1300 ppm of lead orally for 10 weeks impaired the phagocytosis of
sheep red blood cells (Koller & Roan, 1977). Lead also has consistent
overall effects on host resistance to infection. Thus, treatment
resulted in significantly decreased resistance of mice to Klebsiella
pneumoniae (Hemphill et al., 1971) and S. typhimurium, and decreased
resistance of rats to a bacterial endotoxin and to a challenge with
E. coli, S. epidermidis, or S. enteritidis. The increased
susceptibility of rodents to Gram-negative bacteria after exposure
to lead is likely to be due to hypersensitivity to an endotoxin of
bacterial origin (Cook et al., 1974, 1975)
Organolead compounds, such as tetrethyllead, can also be
immunotoxic (Luster et al., 1992).
2.2.6.3 Mercury
Mercuric salts have been shown repeatedly to depress both humoral
and cellular immunity and nonspecific host defences in animals. For
instance, B6C3F1 mice given mercuric chloride orally for seven weeks
had decreased thymus and spleen weights, an impaired plaque-forming
cell response, and inhibition of lymphocyte proliferation at a daily
dose of 75 mg/litre of drinking-water (Dieter et al., 1983).
Methylmercury was reported to decrease humoral immunity in mice
treated orally for three weeks with 0.5, 2, or 10 mg/litre drinking-
water (Blackley et al., 1980).
2.2.6.4 Organotins
Several organotins have been shown to be markedly immunotoxic and
are considered as prototype immunotoxicants (Penninks et al., 1990),
even though no human data are available.
Di- n-octyltin dichloride at 50 or 150 mg/kg of diet for six
weeks induced a dramatic, dose-related decrease in the weight of the
thymus in rats, associated with a less severe decrease in spleen and
lymph node weights (Seinen & Willems, 1976). The numbers of cells in
the thymus and spleen, but not the bone marrow, were decreased.
Histologically, lymphocyte depletion was seen in the thymus and in
thymus-dependent areas of the spleen. Interestingly, thymic atrophy
recovered quickly after cessation of exposure (Seinen et al., 1977).
It was later shown to be associated with a 25% decrease in peripheral
blood lymphocytes, with a preferential loss of Th lymphocytes. As
expected, T-cell functions, such as the delayed-type hypersensitivity
response and T-lymphocyte proliferation, were depressed. Inhibition of
humoral immunity was also seen, with reduced numbers of plaque-forming
cells and decreased circulating antibody titres. NK cell activity was
not affected, whereas susceptibility to L. monocytogenes infection
was markedly increased.
Immune function is not impaired in guinea-pigs or mice fed
di- n-octyltin dichloride, which correlates with the absence of
thymic atrophy (Seinen & Penninks, 1979). Mice treated intravenously
or intraperitoneally develop thymic atrophy, however, suggesting
interspecies variability in the disposition of dialkyltins after oral
intake, although other, poorly understood mechanisms may account for
this variability (Penninks et al., 1990). No interspecies differences
in lymphocyte functions were noted after exposure in vitro.
Generally similar findings were made with the trialkylorganotin,
tri- n-butyltin oxide (TBTO). As trisubstituted organotins are
rapidly metabolized to disubstituted derivatives, the latter are
considered to be involved in the reported thymic effects (Snoeij et
al., 1988). In a short-term study in rats, pronounced effects were
found on the lymphoid organs: thymus (Figure 29), spleen, and lymph
nodes. These effects were most pronounced in thymus-dependent areas
(Figure 30) (Krajnc et al., 1984). Interestingly, thymus atrophy also
occurred in fish, as guppies exposed to organotin compounds showed
severe thymic atrophy (Figure 31). In function tests (Vos et al.,
1984), rats that were exposed to TBTO for six weeks after weaning had
suppressed delayed-type hypersensitivity responses to ovalbumin and
tuberculin and suppressed IgG responses to sheep erythrocytes. In
vitro mitogen responses to concanavalin A in thymus and spleen and
NK cell activity in both the spleen and the lungs were decreased (Van
Loveren et al., 1990b). Exposure to TBTO at 20 or 80 mg/kg of diet for
six weeks led to decreased resistance to infection with
L. monocytogenes or Trichinella spiralis. The latter effect was
evidenced by increased numbers of adult worms in the gut as a result
of impaired worm expulsion, increased numbers of muscle larvae in the
striated tissue, decreased inflammatory responses around these larvae,
and decreased antibody responses to T. spiralis, especially in the
IgE class (Vos et al., 1984). After long-term exposure (15-17 months)
to 5 or 50 mg/kg of diet, delayed-type hypersensitivity was not
suppressed, but assays for NK cell activity and resistance to
infection indicated suppression.
As the immune responsiveness of older animals can be expected to
be less strong than that of younger rats, the effects of exposure to
immunotoxic chemicals may become evident less easily; however, tests
for function still indicated immunotoxicity. In experiments in which
exposure to TBTO was begun only at 12 months of age, both infection
models showed immunotoxicity to TBTO. Very few studies have focused on
the immunotoxic effects of chemicals on the gut immune system, but the
studies of TBTO showed both a decreased capacity of the host to expel
adult T. spiralis worms from the gut and increased production of
serum IgA specific for this parasite (Vos et al., 1990a).
The mechanism of the immunotoxicity of organotin compounds has
been investigated extensively (Penninks et al., 1990). A direct
influence on the synthesis of thymic hormones is uncertain, as
conflicting results have been obtained in different experiments.
Interference with the influx of prothymocytes can be ruled out, as
thymic atrophy develops too rapidly. Interestingly, organotins reduced
the proliferative activity of thymocytes and the number of
proliferating thymoblasts within 24h after exposure was begun, at a
time when thymic atrophy was not evident. This selective effect on
thymoblasts and the physiological destruction of most cortical
thymocytes would result in marked depletion and, finally, in thymic
atrophy.
2.2.6.5 Gallium arsenide
Gallium arsenide is an intermetallic compound used widely in the
electronics industry, primarily in the manufacture of transistors and
light-emitting diodes. A single intratracheal instillation of 50, 100,
or 200 mg/kg body weight into female B6C3F1 mice resulted in a dose-
related decrease in the IgM and IgG antibody response to sheep
erythrocytes. Similarly, cell-mediated immunity, as evaluated by the
delayed-type hypersensitivity reaction to keyhole limpet haemocyanin
and the mixed lymphocyte response, was also decreased in a dose-
dependent way. Increases were observed in complement C3 levels,
mitogen response to lipopolysaccharide, and NK cell activity. No
effects were observed on response to T-cell mitogens, total complement
CH50 activity, or host resistance to Plasmodium yoelii or
Streptococcus pneumoniae; however, a significant decrease in host
resistance was observed to L. monocytogenes and B16F10 tumour
challenge (Sikorski et al., 1989).
2.2.6.6 Beryllium
Beryllium induces a variety of diseases, including granulomatous
lung (chronic beryllium disease) and skin conditions. These
granulomatous reactions involve a lymphocyte response to beryllium
salts. The major lymphocyte population consists of Th cells (CD4). The
T-cell response to beryllium is IL2-dependent (Saltini et al., 1989).
The antigen has not been identified, but may be a beryllium-protein
complex. There appears to be a genetic predisposition, as the majority
of patients with beryllium lung disease share a particular HLA-Dp
allele (HLA-DpB1) (Richeldi et al., 1992). The development and
maintenance of lung and skin granulomas depend on the presence of
antigen, antigen-presenting cells, and memory T lymphocytes and the
release of proinflammatory cytokines by macrophages and lymphocytes
(Boros, 1988; Kunkel et al., 1989).
2.2.7 Air pollutants
Pollutants characteristic of occupational and urban environments
may cause or aggravate pulmonary diseases. Pulmonary defence
mechanisms to pathogens comprise mechanical defences, nonspecific
defences (ingestion by phagocytic cells, lysis of virus-infected
cells), and specific immunity. A number of studies in experimental
animals have shown that exposure to air pollutants, including ozone,
nitrogen dioxide, sulfur dioxide, some volatile organic compounds, and
metal particulates, adversely affects pulmonary defences, and
primarily nonspecific defences important in clearing certain Gram-
positive bacteria from the lung (Graham & Gardner, 1985; Jakab &
Hmieleski, 1988; Selgrade & Gilmour, 1994).
In dogs, exposure to ozone at 3 ppm for 2 h per day for three
days markedly increased the number of epithelial neutrophils, whereas
the number of circulating neutrophils was decreased (O'Byrne et al.,
1984). A significant decrease in absolute thymocyte numbers was also
observed in mice continuously exposed to 0.7ppm ozone for three to
seven days (Li & Richters, 1991). Decreased spleen and thymus weights
were reported in mice exposed to ozone alone or in combination with
nitrogen dioxide (Fujimaki, 1989; Goodman et al., 1989). The numbers
of neutrophils and alveolar macrophages in bronchoalveolar lavage
fluid were found to be increased in rats, and T-lymphocyte
infiltrations were seen in ozone-induced lesions of mice.
Accumulations of macrophages are located mainly at the bronchoalveolar
junction and in alveoli (Figures 32 and 33).
Modulation of nonspecific defence mechanisms by ozone has also
been described (Goldstein et al., 1971; Holt & Keast, 1977; Van
Loveren et al., 1988a, 1990b). Thus, phagocytic activity in alveolar
macrophages is suppressed, but this depends on the concentration and
duration of exposure; enhanced phagocytic activity was also observed.
Alterations in the macrophage production of arachidonic acid
metabolites, resulting in increased prostaglandin 2 production, have
been suggested to be involved (Gilmour et al., 1993). NK cell activity
is either unaffected or stimulated by low ozone concentrations,
whereas high concentrations decreased both the number and the activity
of splenic and pulmonary NK cells (Burleson et al., 1989; Van Loveren
et al., 1990b). Ozone also affects T cells (Dziedzic & White, 1986;
Van Loveren et al., 1988a; Bleavins & Dziedzic, 1990; Dziedzic et al.,
1990). Ozone-induced systemic dysfunction has been reported in animals
and probably contributes to impaired host defences (Aranyi et al.,
1983). Humoral immunity, e.g. circulating antibody titres to a variety
of antigens and the plaque-forming cell response to sheep
erythrocytes, is depressed after exposure to ozone; cellular immunity
is also inhibited. The numbers of all major T lymphocyte subsets,
mitogen-induced T lymphocyte proliferation, and delayed-type
hypersensitivity responses were all shown to be decreased. Numerous
studies with infectivity models show that exposure to ozone has an
adverse influence on the host defences to respiratory infections (Van
Loveren et al., 1994), and most of the studies demonstrate that the
primary targets are the alveolar macrophages (Selgrade & Gilmour,
1994).
Although the influence of other air pollutants such as nitrogen
dioxide and sulfuric acid on host defences has been the subject of
fewer studies, the available data suggest that they have similar
adverse effects (Graham & Gardner, 1985). In view of the numerous
possible targets of air pollutants on respiratory defences and because
of the intricate mechanisms involved, infectivity models in animals
are particularly relevant for ascertaining the likely consequences of
air pollution for exposed human populations.
2.2.8 Mycotoxins
Mycotoxins are structurally diverse secondary metabolites of
fungi that grow on feed. Mycotoxin-induced immunosuppression may be
manifested as depressed T- or B-lymphocyte function, decreased
antibody production, or impaired macrophage activity. Immuno-
stimulation may also be observed with the tricothecenes under
some experimental conditions. Similar effects have been found on the
proliferative responses of human and rodent lymphocytes in vitro
(Lang et al., 1993). Most of the data have been obtained in vivo or
in vitro in animal systems, and there is only limited evidence that
mycotoxins are immunosuppressive in humans (Lea et al., 1989).
Dietary exposure to various mycotoxins resulted in decreased
antibody production, T-lymphocyte proliferative response, delayed-type
hypersensitivity, and NK cell activity (Pestka & Bondy, 1990).
Interestingly, dietary intake was associated with increased
susceptibility to experimental infections.
Aflatoxin is markedly immunosuppressive in cattle and poultry
(see below). Thymic atrophy, suppression of mitogen-induced T- and
B-lymphocyte proliferation, and decreased antibody responses to
various microbial antigens and sheep erythrocytes have been observed
(Corrier, 1992). Cell-mediated immune responses appear to be affected
at lower concentrations than antibody responses. The mechanism of
action seems to be related to impaired protein synthesis.
Ochratoxin, a mycotoxin produced by several species of
Aspergillus and Penicillium, causes depletion of lymphoid cells in
the spleen and lymph nodes of dogs, swine, and mice (Corrier, 1992).
The dose, the route of administration, and the animal species appear
to be critical factors, however; for instance, administration of 13 mg
of ochratoxin to mice in six intraperitoneal injections did not impair
T-lymphocyte proliferation (Luster et al., 1987), whereas
intraperitoneal injections of 5 mg/kg body weight for 50 days did
(Prior & Sisodia, 1982). Ochratoxin also impairs NK cell activity and
increases tumour cell growth in mice (Luster et al., 1987).
The trichothecenes, including T-2 toxin and deoxynivalenol
(vomitoxin), are a structurally related group of mycotoxins produced
by Fusarium. T-2 toxin has been studied extensively and has been
shown to induce lymphoid depletion in the thymus, spleen, and lymph
nodes of numerous laboratory animals (Pestka & Bondy, 1990; Corrier,
1992). In addition, mitogen-induced T- or B-lymphocyte proliferation,
antibody production, and macrophage activation have been found to be
depressed after exposure to either T-2 toxin or vomitoxin. The
impaired immune responsiveness is associated with increased
susceptibility to a variety of experimental infections. As the
tricothecenes are currently considered to be the most potent small-
molecule inhibitors of protein synthesis in eukaryotic cells, the
immunosuppression associated with exposure to these mycotoxins is
likely to be directly or indirectly linked to inhibition of protein
synthesis.
2.2.9 Particles
2.2.9.1 Asbestos
Exposure to asbestos is associated with the development of
inflammatory, fibrotic, and malignant (i.e. pleural mesothelioma and
bronchogenic carcinoma) diseases in humans. Although the pathogenesis
of asbestos-induced lung diseases is complex, a number of observations
indicate that immune processes influence the development and
resolution of both the inflammatory response and fibrotic lesions. For
example, exposure to asbestos is associated with alterations in
cellular and humoral-mediated immunity, including reduction of
lymphocyte mitogenesis, delayed hypersensitivity responses, and
primary antibody responses (Hartmann et al., 1984; Hartmann, 1985;
Bissonette et al., 1989; Miller, 1992). In addition, immunodeficient
mice resolve asbestos-induced inflammatory and fibrotic responses only
with difficulty (Corsini et al., 1994), suggesting that immune
mediators with anti-inflammatory or anti-fibrotic activity (e.g. IL-4
or INF gamma) are involved. Furthermore, it is well established that
alveolar macrophages and type II epithelial cells secrete inflammatory
cytokines, chemokines, and growth factors in response to asbestos
(Driscoll et al., 1990; Rosenthal et al., 1994), and these mediators
are directly involved in the inflammatory responses (e.g. inflammatory
cell recruitment) and fibrogenesis (e.g. fibroblast proliferation).
2.2.9.2 Silica
Experimental animals have been used extensively to define the
pathogenesis of silicosis (Uber & McReynolds, 1982). Several immune
changes have been demonstrated in guinea-pigs, including depression of
humoral and cellular immunity and increased susceptibility to
infectious agents. Similarly, mice exposed to silica showed decreased
lipopolysaccharide-induced proliferation of B lymphocytes and
depressed plaque-forming cell responses (Scheuchenzuber et al., 1985).
Antibody responses to T-independent antigens, however, were less
markedly depressed than responses to T-dependent antigens, suggesting
an additional effect on T-cell control of humoral immunity. The
effects of silica on cellular immunity depend on the dose and route of
entry of antigens. The concanavalin A-induced proliferation response
of spleen T lymphocytes was increased, whereas that of mesenteric
lymph node T lymphocytes was depressed. The aberrations of humoral and
cellular immunity induced by silica are thus complex, and it remains
to be established how these immune changes correlate with the
induction of lung fibrosis or autoantibodies, the major adverse
consequences of exposure to silica. In addition, silica is markedly
toxic to macrophages and activates alveolar macrophages, granulocytes,
and monocytes (Gusev et al., 1993). Infectivity models consistently
show an increased susceptibility of silica-exposed rodents to
infectious pathogens.
2.2.10 Substances of abuse
The immunotoxic consequences of exposure to substances of abuse
are difficult to ascertain in most instances as confounding factors,
such as intercurrent infections secondary to intravenous injection,
may contribute to the observed changes. Recent research has provided
evidence, however, that substances of abuse can directly affect the
immune system (Descotes, 1986; Watson, 1990; Friedman et al., 1991a;
Watson, 1993).
In rodent lymphocytes in vitro, D9-tetrahydrocannabinol
depressed the proliferative responses of T lymphocytes in a dose-
dependent manner (Friedman et al., 1991b). Further to the early
findings that opiates adversely affect immune competence (Cantacuzene,
1898), an increasing body of evidence shows that exogenous opioids
have a variety of effects on cells of the immune system (Rouveix,
1993). At pharmacological concentrations, opiates suppress antibody
production, lymphocyte proliferation, and delayed-type
hypersensitivity and decrease NK cell activity in various animal
models. In addition, phagocytosis is impaired. Opioid peptides can,
however, also have a stimulatory effect on the immune system,
depending on the experimental conditions. ß-Endorphin affects cytokine
production in rat and mouse T-cell cultures in vitro; e.g. it
stimulates the synthesis of IL-2, IL-4, and INF gamma, thereby
inducing MHC class II expression on B cells (van den Bergh et al.,
1993a,b, 1994).
In general, short-term exposure of mice, rats, and guinea-pigs to
mainstream tobacco smoke either produces no significant immuno-
modulatory effect or a slight immunostimulation, which returns
to normal shortly after cessation of exposure (Johnson et al., 1990).
In contrast, subchronic or chronic (more than one year) exposure is
generally immunosuppressive: cellular immunity, e.g. mitogen-induced
lymphocyte proliferative response, and NK cell activity are impaired
after long-term exposure to tobacco smoke. The humoral immune response
is also depressed, as shown by decreased antibody titres, and animals
exposed to cigarette smoke for extended periods are more susceptible
to tumour and infectious challenge than naive animals.
2.2.11 Ultraviolet radiation
The earliest indication that ultraviolet radiation (UVR) affects
the immune system came from studies of host resistance to UVR-induced
tumours in mice (Kripke, 1974). Subsequent studies showed that low
doses of UVR suppress contact hypersensitivity responses to chemical
sensitizers (Toews et al., 1980) and that systemic immunosuppression
(depressed contact hypersensitivity in unirradiated skin) occurs after
exposure to higher doses (Jessup et al., 1978). Irradiated mice were
also found to be less resistant to infection (Giannini, 1990). Other
studies (Noonan & De Fabo, 1990) have determined that systemic
suppression of immunoreactivity is not a function of the dose of UVR
but rather of the interval between irradiation and immunization of the
mice. Thus, induction of contact hypersensitivity responses in mice
exposed to low doses of UVR was not affected when the animals were
immunized through unirradiated skin immediately after exposure to UVR;
however, sensitization was suppressed if three days were allowed to
elapse between irradiation and immunization. It has also been shown
that the dose of UVR required to induce 50% suppression of the immune
response depends on the strain of mouse and the type of antigen used
(Noonan & De Fabo, 1990; Noonan & Hoffman, 1994).
The mechanism of UVR-induced suppression of cellular immunity has
not been elucidated, nor has a single initial event been identified
that leads to suppression of immunoreactivity. Currently, induction of
pyrimidine dimers in DNA (Kripke et al., 1992) and isomerization of
urocanic acid (Noonan & De Fabo, 1992) are the leading contenders.
Increased suppressor cell activity (Brodie & Halliday, 1991) and
efferent lymphatic blockade, which inhibits lymphocyte homing, may be
responsible for the UVR-associated accumulation of lymphocytes in
lymph nodes in UVR-exposed areas (Spangrude et al., 1983) and have
been proposed as possible causes of immunosuppression. Exposure to UVR
has also been shown to alter the pattern of cytokine production by
T cells, from a response dominated by Th1 (i.e. favouring delayed
hypersensitivity responses) to one dominated by Th2 (i.e. favouring
antibody production) (Araneo et al., 1989; Simon et al., 1990).
Exposure to UVR has been reported to affect Langerhans cells directly,
such that their interaction with T cells induces specific antigen
tolerance in the Th1 subpopulation (Simon et al., 1990) and
preferential activation of the Th2 population (Simon et al., 1991).
This may be the reason that mice exposed to UVR are more susceptible
to infection with the protozoan Leishmania major (Giannini, 1992),
since resistance to infection with this intracellular parasite is
dependent on the magnitude of the Th1 response of the host (Reed &
Scott, 1993). In addition, reduced resistance to T. spiralis was
found in rats exposed to UVR on days 5-10 of infection (Goettsch et
al., 1993). Altered cytokine production profiles may also be
responsible for increased sensitivity to Mycobacterium lepraemurium,
an intracellular pathogen that induces a chronic and eventually fatal
infection in susceptible mice. In a comparison of susceptible (BALB/c)
and resistant (C57Bl/6J) mice, Brett & Butler (1986) determined that
resistance to infection is correlated with the ability of mouse
lymphocytes to elaborate cytokines that activate macrophages, rather
than with the actual development of a delayed hypersensitivity
response to bacterial antigens. Jeevan & Kripke (1990) and Jeevan et
al. (1992) reported that irradiation of BALB/c mice resulted in
decreased resistance to infection, as measured by bacterial counts and
length of survival after infection. Elevated bacterial counts were
seen in animals exposed to doses of UVR that did not suppress the
delayed hypersensitivity response to bacterial antigens, suggesting
that the underlying mechanism of UVR-induced suppression of resistance
to infection is independent of suppressed delayed hypersensitivity.
2.2.12 Food additives
There is little information about the effects of food additives
on the immune system. An early study showed that the preservative
methylparaben and the antioxidants butylated hydroxyanisole, butylated
hydroxytoluene, and propylgallate suppress the in-vitro T-dependent
antibody response, whereas vanillin and vanillic acid stimulate it
(Archer et al., 1978). The significance of these findings in vivo
has yet to be established.
The immunotoxicity of 'caramel colour', which covers a large
number of complex products used as food colorants, has been
investigated. One of the compounds in this group, 2-acetyl-4(5)-
tetrahydroxybutylimidazole (THI) (caramel colour III), has been found
to be immunotoxic in rodents (Kroplien et al., 1985). THI induces a
rapid reduction in the number of B and T cells in blood, spleen, and
lymph nodes and morphological changes in the thymus of rats, with an
increased number of mature medullary thymocytes and a decreased number
of cortical macrophages. THI might reduce the migration of mature
thymocytes into the periphery, as a decrease in the number of recent
ER4+ thymic emigrants was found in the spleens of exposed rats
(Houben et al., 1992). Functional studies indicate changes in Th cell
function, an increased capacity to clear the Gram-positive bacterium
L. monocytogenes, and modulation of the activity of adherent splenic
cells (Houben et al., 1993). It has been hypothesized that THI exerts
an antivitamin B6 action by competing with pyridoxal 5'-phosphate for
binding to the cofactor site of one or more pyridoxal 5'-phosphate-
dependent enzymes.
2.3 Immunotoxicity of environmental chemicals in wildlife and
domesticated species
Most of the concern about chemical contamination of wildlife
populations has been focused on the aquatic ecosystem, and a growing
body of literature has appeared on the effects of pollution on the
health status of aquatic life. These studies deal mainly with the
occurrence of tumours and infectious diseases in fish and marine
mammals. These are multifactorial diseases in which perturbations of
the immune system may play a part.
2.3.1 Fish and other marine species
2.3.1.1 Fish
Fish diseases are being monitored on a routine basis at various
sites in North America and Europe. In Europe, most of the programmes
are carried out under the auspices of the International Council for
Exploration of the Sea. National and local studies have been directed
to estuarine, marine, and brackish waters suspected of being polluted,
such as in the vicinity of industrial areas and after major oil
spills. The common diseases that are discussed in relation to
pollution are certain skin diseases, such as lymphocystis, papillomas,
fin rotting, and skin ulcers (Vethaak & ap Rheinallt, 1992), as they
are easily identified grossly and are therefore potentially useful for
biomonitoring. Since most diseases of fish have a viral or bacterial
etiology, and elevated incidences have been correlated with chemical
pollution, immunotoxicity may play a role. A causal relationship
between chemical pollution and a disease state induced by
immunosuppression cannot be finally established, however, since many
confounding factors exist in the natural environment. Liver neoplasia
and precursor lesions have been used to biomonitor environmental
pollution in flatfish (Malins et al., 1988; Vethaak & ap Rheinallt,
1992; Vethaak, 1993); however, the role of the immune system is not
evident.
Effects observed in field studies are modified or confounded by
numerous factors, in particular for feral fish, for which there are
deficient case histories and limited knowledge of their migratory
patterns and biology (Vethaak, 1993). Extensive epidemiological
surveys are required that include specific parameters that have been
validated in experiments under (more) controlled conditions (in
mesocosms or the laboratory). Changes in disease patterns may suggest
immune alterations, but this should be demonstrated. Since most
diseases have a complex etiology, it will be difficult to establish
the role of immunotoxic effects under field conditions. Circumstantial
evidence can be obtained in these instances, although in the case of
feral animals mesocosm or laboratory experiments must carried out in
order to reach a final conclusion (Secombes et al., 1992; Vethaak,
1993).
The etiological components and their role in the pathogenesis of
many diseases in fish in the field are, as yet, poorly understood, and
laboratory experiments are often indispensable for background
knowledge. Even when such scientific deficiencies are resolved,
laboratory studies will still be needed, since function tests under
controlled conditions yield the most reliable and sensitive methods of
assessing immunological stress and must often accompany field studies,
as mentioned above. Findings from laboratory situations do not
necessarily imply effects in the field, however; in particular, when
results from the laboratory are extrapolated to field situations,
there is often a discrepancy between the levels of exposure.
2.3.1.2 Marine mammals
Marine mammals are of special interest to the discipline of
immunotoxicology. As the highest predators in highly contaminated
marine environments, these animals are exposed to a large number of
environmental chemicals, some of which have been identified as
potentially immunotoxic. Persistent lipophilic halogenated compounds
such as PCBs, polychlorinted dibenzodioxins, polychlorinated
dibenzofurans, and pesticides accumulate in the marine food chain and
are thus biomagnified in marine mammals. The concentrations of PCBs in
the blubber layer of marine mammals are higher than in any other
wildlife species measured (Table 6). In times of food shortage and
other stressful circumstances, these lipids are mobilized, thereby
releasing their toxic burden.
Table 6. Concentrations of polychlorinated biphenyls (PCBs) in herring and the blubber
layer of marine mammals
Species Source Total PCBs Reference
(µg/g lipid)
Herring Atlantic Ocean 0.0003-0.001 De Swart et al.
(United Kingdom) (1994)
Herring Baltic Sea 0.0035-0.0045 De Swart et
(Sweden) al. (1993)
Weddell seal Weddell Sea 0.07-0.09 Luckas et al.
(Leptonychotes wedelli) (Antarctic) (1990)
Harbour seal Atlantic Ocean 1-13 Luckas et al.
(Phoca vitulina) (Iceland) (1990)
Harbour seal Baltic Sea 21-140 Luckas et al.
(Phoca vitulina) (Sweden) (1990)
Beluga whale St Lawrence 15-700 Muir et. al.
(Delphinapterus leucas) River (Canada) (1990);
Martineau et al.
(1987)
Striped dolphin Mediterranean 100-2600 Kannan et al.
(Stenella coeruleoalba) Sea (Spain) (1993)
Because of the high level of exposure of marine mammals, they may
be expected to be the first wild animals to suffer from
immunosuppression due to chronic exposure to environmental
contaminants. Toxicological research over the last 20 years has
identified environmental chemicals as the source of many disorders in
marine mammals. In both field studies and controlled experiments, PCBs
have been linked to reproductive problems. As early as 1976, a high
incidence of premature parturition was seen in California sea-lions
(Zalophus californianus), which was caused by a viral infection and
was suggested to be linked to higher levels of pollutants in animals
that aborted as compared with mothers that gave birth to healthy pups
(Gilmartin et al., 1976). Pathological changes in the uteri of seals
in the highly polluted Baltic Sea, in some cases leading to sterility,
could be correlated with increased levels of PCBs (Helle et al.,
1976). In addition, several studies have linked skeletal deformities
in grey seals (Halichoerus grypus) and harbour seals (Phoca
vitulina) in the Baltic Sea to high levels of organochlorines in
their environment (Bergman et al., 1992; Mortensen et al., 1992). In
porpoises (Phocoenoides dalli) living in the north-western part of
the North Pacific, a negative correlation was found between serum
testosterone levels and DDE concentrations in blubber (Subramanian et
al., 1987). In an experimental situation, two groups of harbour seals
were fed fish containing different levels of pollutants. Seals that
were fed fish from the heavily polluted western part of the Dutch
Wadden Sea had significantly lower pup production than seals feeding
on less polluted fish (Reijnders, 1986). In the same study, it was
shown that the seals fed polluted fish also had significantly reduced
levels of vitamin A and thyroid hormone in their serum (Brouwer et
al., 1989). No parameters of immune function were studied.
Such correlative observations in the natural environment, in
combination with the results of semi-field studies, suggest that the
current levels of contaminants may be adversely affecting certain
marine mammal populations. The occurrence of a large number of
epizootics among seals and dolphins inhabiting polluted coastal areas
-- among bottlenose dolphins (Tursiops truncatus) on the Atlantic
coast of the United States in 1987 and in the Gulf of Mexico in 1990,
striped dolphins (Stenella coeruleoalba) on the coast of France in
1989 and in the Mediterranean Sea in 1990-92, Baikal seals (Phoca
sibirica) in Lake Baikal in 1987, and harbour seals in north-west
Europe in 1988 -- led to both public and scientific discussions about
the possible contribution of environmental pollutants to these disease
outbreaks. Owing to the complexities of the relationships between
toxicants and the immune system and the difficulties in obtaining
samples fit for use in immunotoxicological studies, it has been
impossible thus far to conclusively demonstrate immunosuppression in
free-ranging marine mammals.
Another immunotoxicological experiment was carried out in which
two groups of juvenile harbour seals (Phoca vitulina) were fed fish
from the Baltic Sea or from the relatively unpolluted Atlantic Ocean.
The diets were analysed for concentrations of potential immunotoxic
chemicals: the estimated daily intakes of TCDD-like organochlorines
were 270 and 35 ng/day for the two groups, respectively. Immunological
function in the two groups was examined by measuring mitogen- and
antigen-induced proliferative responses of lymphocytes, NK cell
activity, serum antibody levels after immunization with primary
antigens, and delayed-type hypersensitivity reactions. These
techniques had to be validated for application to seals, as virtually
nothing was known about the cellular immune system of marine mammals
(De Swart et al., 1993, 1994). Seals fed herring from the contaminated
Baltic Sea had significantly depressed immune function, as measured by
decreased NK cell activity (Ross et al., in press) and T-cell mitogen-
induced lymphocyte proliferation (De Swart et al., 1994), and
significantly lower delayed-type hypersensitivity and antibody
responses to immunization with ovalbumin (Ross et al., 1995). The
functional changes were accompanied by increased numbers of
neutrophils in the peripheral blood (De Swart et al., 1994). Since NK
cells are an important line of defence against viruses, and
lymphocytes (especially T cells) play a major role in the clearance of
viral infections, functional suppression of these leukocytes may
contribute to the severity of epizootic episodes among marine mammals.
These experiments not only provide the first demonstration of
pollution-induced impairment of immune function in marine mammals but
indicate that mammals in general can undergo such impairment as a
consequence of chronic exposure to the levels of pollution found in
their natural habitats. It is still difficult, however, to link the
disease outbreaks among marine mammals directly to pollution-induced
impairment of immune function.
2.3.2 Cattle and swine
The effects of antimicrobial, corticosteroid, and hormonal
compounds have been investigated in cattle, mainly as lymphocyte
proliferative responses and neutrophil functions in vitro. The
results are in keeping with those obtained in humans (Black et al.,
1992).
Few studies have dealt specifically with the direct immunotoxic
effects of pesticides and environmental pollutants. No statistical
difference was found between control and polybrominated biphenyl-
exposed animals in the numbers of circulating total, T, and B
lymphocytes, serum immunoglobulin levels, mitogen-induced
proliferative responses of lymphocytes, antibody response to keyhole
limpet mitogen, or cell-mediated response to purified protein
derivative (Kateley & Bazzell, 1978). In contrast, peripheral blood
lymphocytes from sows fed polybrominated biphenyls for 12weeks had
significantly decreased responses to mitogens (Howard et al., 1980).
The mycotoxin tricothecene produced by Fusarium and several
other fungi was shown to reduce lymphoid tissue cellularity and serum
IgA, IgG, and IgM concentrations and to impair neutrophil migration,
chemotaxis, and phagocytosis in cattle exposed to high doses (Buening
et al., 1982; Mann et al., 1983). Similarly, aflatoxin was reported to
suppress the mitogen-induced proliferative response of bovine
lymphocytes (Paul et al., 1977). Other mycotoxins, e.g. ochratoxin A
and zearalenone, have been suggested to be immunosuppressive in cattle
(Black et al., 1992).
2.3.3 Chickens
Chickens have been used in a number of immunological studies, as
the unique bursa of Fabricius, the avian organ for B-cell development,
underlies the need for a separate avian model in immunotoxicology.
Exposure of adolescent chickens to cyclophosphamide decreased the
levels of antibodies to various antigens without decreasing graft-
versus-host reactivity (Lerman & Weidanz, 1970). Nutritional
deficiencies in selenium or vitamin E have also been shown to impair
the humoral immune responses of adolescent chickens (Marsh et al.,
1981). Exposure to cyclophosphamide in ovo results in decreased
antibody forming capacity, decreased responsiveness to
phytohaemagglutinin and concanavalin A, and decreased weights and
altered morphology of the thymus, spleen, and bursa of Fabricius
(Eskola & Toivanen, 1974). Peritoneal macrophages were not affected
after exposure to cyclophosphamide in ovo, as judged by their
number, superoxide anion production, and surface expression of Ia
antigen and transferrin receptor (Golemboski et al., 1992). Dietert et
al. (1985) showed that exposure to aflatoxin B1 in ovo did not alter
humoral immunity; however, two parameters of cell-mediated, graft-
versus-host, and cutaneous basophil hypersensitivity reactions were
depressed. Methyl methanesulfonate decreased the bactericidal action
of peritoneal macrophages for E. coli after exposure in vitro
(Qureshi et al., 1989). TCDD impairs B-cell development in the bursa
of Fabricius in chicken embryos (Nikolaidos et al., 1990).
2.4 Immunotoxicity of environmental chemicals in humans
Although limited, various lines of evidence derived from case
reports, clinical studies, and well-designed longitudinal studies
imply that environmental agents can affect the human immune system.
While these data raise concern about potential health effects, they
rarely refer to clinical disease, for a number of reasons. For
example, there may be sufficient redundancy or reserve in the immune
system that 'moderate' levels of immunosuppression do not result in
disease. Alternatively, the clinical changes most likely to be
associated with moderate immunosuppression, e.g. increased severity or
frequency of pulmonary infections, do not occur. Well-designed
clinical studies with adequate populations and appropriate monitoring,
follow-up, and documentation of exposure have rarely been conducted.
Examples of published reports that attribute immune changes in human
populations exposed to environmental agents are summarized below. As
the reader will note, these studies range from poorly defined to
relatively large longitudinal studies.
2.4.1 Case reports
In an unsubstantiated study, a cluster of cases of Hodgkin's
disease reported in a small town in Michigan (United States) was
ascribed to chronic immune stimulation by mitogenic substances in the
environment (Schwartz et al., 1978). Immunological studies of family
members revealed a large number of individuals with altered ratios of
T-lymphocyte subpopulations, autoantibodies, infections, recurrent
rashes, and NK cell function. A report of a four-year study of workers
engaged in the manufacture of benzidine, a human bladder carcinogen,
suggested that individuals with depressed cell-mediated immunity (as
judged by skin testing) had precancerous conditions and subsequent
neoplasms (Gorodilova & Mandrik, 1978); no cases of neoplastic disease
were registered in workers with normal immunological responses.
2.4.2 Air pollutants
The association betweeen changes in immunological parameters and
host resistance and inhalation of particulate materials and oxidant
gases is well established (Folinsbee, 1992). For example, decreases in
delayed-type hypersensitivity response, circulating T-cell numbers,
and T-cell proliferation have been observed with, and sometimes
preceding, asbestos-related diseases, i.e. fibrosis, asbestosis, and
mesothelioma (Kagan et al., 1977a; Gaumer et al., 1981; Lew et al.,
1986; Tsang et al., 1988). B-Cell responses are increased, however, as
evidenced by increased serum and secretory (primarily IgA)
immunoglobulins (Kagan et al., 1977b). Kagan et al. (1979) also
reported an association between exposure to asbestos and B-cell
lymphoproliferative disorders. Several studies have shown altered NK
cell activity after exposure to asbestos (Kubota et al., 1985; Tsang
et al., 1988). In studies of NK cell responses in asbestos workers,
Lew et al. (1986) found that immune changes may occur independently of
any early neoplastic process. Similarly, there have been multiple
observations of abnormal antibody production, decreased cell-mediated
immune responses, and decreased resistance to disease in people
occupationally exposed to silica (Uber & McReynolds, 1982).
Oxidant gases have been associated with an increased prevalence
of respiratory infections, particularly bacterial, and potential
immune effects, but the data are less convincing than those from
studies of rodents. The association between air pollution in
industrialized areas and altered health status has been well
established in epidemiological studies (French et al., 1973). A number
of studies have linked exposure to air pollutants (ozone, nitrogen
dioxide, sulfur dioxide, environmental tobacco smoke) with an
increased incidence, severity, or duration of symptoms associated with
respiratory infections (Lunn et al., 1967; French et al., 1973;
Durham, 1974; Harrington & Krupnick, 1985; Neas et al., 1991; Schwartz
et al., 1991; Schwartz, 1992; US Environmental Protection Agency,
1990), although several studies of nitrogen dioxide failed to show
such an association (Speizer et al., 1980; Ware et al., 1984; Samet et
al., 1993). In Ontario, Canada, increased air pollution from sulfur
dioxide and ozone during the summer was directly related to hospital
admissions for acute respiratory symptoms (Bates & Sizto, 1983).
Goings et al. (1989) subjected young adult volunteers to 1, 2, or
3 ppm nitrogen dioxide for 2h per day on two consecutive days and then
administered influenza virus intranasally. Although no statistical
differences were observed, the frequencies of infections in exposed
volunteers (91%) were higher than the 56-73% in healthy individuals,
suggesting that nitrogen dioxide may play a role in increasing
susceptibility to infection. In assessments of air pollution at home,
young children in households with gas stoves had a higher incidence of
respiratory disease and decreased pulmonary function than children in
households with electric stoves. This difference was related to
increased levels of nitrogen dioxide in homes with gas stoves, which
reached peak values of > 1100 mg/m3 (Melia et al., 1977; Speizer et
al., 1980). In contrast, Samet (1994) found no association between
indoor levels of nitrogen dioxide and respiratory infections in
children. A study of schoolchildren in Chattanooga (United States)
showed an increased incidence of respiratory illness associated with
atmospheric nitrogen dioxide levels (Shy et al., 1970).
Examination of hospital admissions in Massachusetts (United
States) in 1980 and 1982 revealed a positive association between 1-h
maximum ozone levels in the summer months and daily admissions for
pneumonia and influenza (Ozkaynak et al., 1990). An effect of
atmospheric pollution, including oxidant gases, was also seen on the
number of influenza cases in Sofia, Bulgaria (Kalpazanov et al.,
1976); however, no demonstrable adverse effect on the course of a
rhinoviral infection was seen in young adult male volunteers after
exposure to moderate levels of ozone (0.3ppm for 6 h/day over a period
of five days) (Henderson et al., 1988), and children living in areas
with high ozone concentrations had lowered CD4:CD8 ratios of
peripheral lymphocytes but no higher incidence of chest colds (Zwick
et al., 1991). The phagocytic activity of alveolar macrophages
(obtained by bronchoalveolar lavage) and other functions were impaired
in human volunteers exposed to ozone (Devlin et al., 1991). The
sensitivity of human and mouse macrophages to the effects of ozone is
similar (Selgrade et al., 1995).
Not all epidemiological studies have showed a correlation between
air pollution and respiratory disease. Some of the discrepancies from
experimental studies may be due to the parameters used to assess
enhancement of infection. In experimental studies, increases in viral
titres in the respiratory tract tend to be taken as an indication that
the exposure enhances infection, whereas epidemiological studies rely
on symptoms, many of which could be related to enhanced inflammatory
or even allergic responses to the virus in the absence of viral
replication.
Controlled studies (in an environmental chamber) showed that
acute exposure to ozone causes an inflammatory response in the lower
airways of human subjects (Koren et al., 1989; Devlin et al., 1991).
The inflammatory response was manifested by increases in various
inflammatory indicators including polymorphonuclear neutrophils
(Figure 34), proteins, fibronectin, IL-6, and tryptase (Koren et al.,
1989, 1994).
The proinflammatory effects of ozone raise the possibility that
it can increase the sensitivity of people with atopic asthma to
challenge with a specific allergen. Several studies have investigated
the effect of exposure to pollutants on subsequent reactivity to
antigen in atopic human volunteers. Molfino et al. (1991) reported
that exposure to 0.12 ppm ozone significantly increased bronchial
responsiveness to antigen in some individuals. Although they
acknowledged weaknesses in the design of the experiment and
recommended that the findings be confirmed, their observations are in
accordance with those of the majority of epidemiological studies. In
contrast, Bascom et al. (1990) found no alteration in the acute
response to nasal antigen challenge in allergic patients pre-exposed
to ozone in comparison with exposure to air; however, they did report
increased upper respiratory tract inflammation after exposure to ozone
in these patients in the absence of antigen challenge. Similarly, the
bronchial response to inhaled grass pollen was unaffected by prior
exposure to 0.1 ppm nitrogen dioxide (Orehek et al., 1981). The topic
of sensitization and the role of ozone in exacerbating asthma has been
reviewed (Koren & Blomberg, in press).
2.4.3 Pesticides
Pesticides can alter the human immune system. For example, women
chronically exposed to low levels of aldicarb-contaminated groundwater
had altered numbers of T cells and decreased CD4:CD8 ratios (Fiore et
al., 1986). While this finding was not confirmed in studies in animals
(see section 2.2.3.4), follow-up studies by Mirkin et al. (1990)
confirmed the immune changes in those individuals still available for
study, although the population was considerably smaller.
2.4.4 Halogenated aromatic hydrocarbons
A number of chemical accidents have resulted in human exposure to
halogenated aromatic hydrocarbons. A feed supplement for lactating
cows, inadvertently contaminated with polybrominated biphenyls, was
used in more than 500 dairy herds and poultry farms in Michigan
(United States) in 1973. Diary farm residents had reduced proportions
of circulating T Iymphocytes and reducedlymphoproliferative
responsiveness in vitro (Bekesi et al., 1978); these changes
persisted during follow-up (Bekesi et al., 1987). Silva et al. (1979),
however, were unable to detect any immune abnormalities in a similarly
exposed cohort.
In Taiwan, more than 2000 people were exposed in 1979 to rice oil
contaminated with PCBs and polychlorinated dibenzofurans. The clinical
features in many of the exposed individuals included chloracne,
pigmentation of skin, liver disease, and respiratory infections. When
the immune status of the exposed individuals was examined one year
after exposure, decreased concentrations of serum IgM and IgA, but not
IgG, were reported, in addition to decreased numbers of circulating Th
cells. The proportions of Ts/Tc cells and B lymphocytes were within
the control values. Suppression of delayed-type hypersensitivity to
recall antigens (streptokinase, streptodornase, tuberculin), enhanced
mitogen-induced lymphocyte proliferation, and increases in
sinopulmonary infections have been reported in this population (Chang
et al., 1981, 1982; Lu & Wu, 1985). Many of the effects were
transient, since two years after exposure most of the clinical
abnormalities and laboratory parameters had returned to normal. A
similar incident occurred in Japan in 1978, resulting in the 'yusho'
syndrome. The immune system was assumed to be affected because of an
increased frequency of respiratory infections and lowered serum IgM
and IgA concentrations (Shigematsu et al., 1978). Another incident of
intoxication with PCBs and polychlorinated dibenzofurans occurred
after exposure to contaminated soot of fires in electrical equipment
(Elo et al., 1985). The exposed people had serum PCB concentrations up
to 30 mg/litre. The number of blood T cells was lower five weeks after
exposure but had returned to normal values seven weeks later. Lowered
CD4:CD8 ratios and lymphocyte proliferation after mitogen stimulation
were also seen. Nine of the 15 most heavily exposed persons suffered
from at least one infection of the upper respiratory tract. No overt
long-term effects or chloracne were observed.
The existing data also suggest that neonates are particularly
sensitive to the immunotoxic effects of PCBs. Thus, higher incidences
of colds and gastrointestinal (vomiting, abdominal pain) and
dermatological (eczema, itchy skin) manifestations were observed in
breast-fed infants of women occupationally exposed to the PCBs
Kanechlor 500 and 300 than in infants born to unexposed women. The
incidence of these symptoms increased with increasing length of
breast-feeding (Hara, 1985). Epidemiological studies of women who
consumed contaminated fish from the Great Lakes indicated that the
maternal serum PCB level during pregnancy was positively associated
with the number and type of infectious illnesses suffered by their
breast-fed infants, especially during the first four months of life.
The incidence of infections in the infant was correlated strongly with
the highest rate of maternal fish consumption (Swain, 1991).
A number of studies have been conducted of the immune status of
people exposed to TCDD. In 1976, an accident occurred at a chemical
plant in Seveso, Italy, in which high concentrations of TCDD were
released into the local environment. An evaluation of 44 exposed
children (20 with chloracne) showed no overt changes in immune status
(Reggiani, 1980), although the adequacy of the control populations
used has been questioned. In a study of residents of an area of
Missouri (United States) with long-term exposure (average, three
years) to low levels of TCDD in contaminated soil, no clinical
symptoms were recorded, although a number of individuals showed
changes in cell-mediated immunity, manifested as altered delayed-type
hypersensitivity (Hoffman et al., 1986). In the follow-up
investigation, however, the skin anergy was not confirmed (Evans et
al., 1988). The serum concentration of the thymic hormone thymosin-a1,
which has been related to the toxic action of this compound on the
thymus, was reduced (Stehr-Green et al., 1989; Hoffman, 1992).
Jennings et al. (1988) found an increased frequency of circulating
antinuclear antibodies and immune complexes in TCDD-exposed workers.
2.4.5 Metals
Exposure to metals may also affect the immune system. Workers
with elevated blood lead levels (30-90 µg/100 ml) had increased
suppressor cell activity (Cohen et al., 1989), lowered lymphocyte
proliferation after mitogen stimulation in vitro (Jaremin, 1983),
decreased IgA concentrations in saliva, and lowered complement C3
levels (Ewers et al., 1982). These individuals also had an enhanced
prevalence of respiratory infection (Ewers et al., 1982). The
immunotoxic effects of lead may be dose-dependent, since neither
humoral nor cellular parameters were affected after long-term, low-
level exposure (Reigart & Graber, 1976; Kimber et al., 1986). In
unrelated studies, the cationic heavy metal mercury was associated
with immune complex disease in humans (Makker & Aikawa, 1979).
In contrast to the database on the immunotoxic effects of
cadmium, lead, and mercury in experimental animals in vivo and the
results of mechanistic studies in vitro, the data on the effects of
heavy metals on the human immune system are scanty and refer mainly to
occupational exposure. These studies nevertheless provide evidence
that at least mercury and lead affect the immune system (Moszczynski
et al., 1990a; Bernier et al., in press).
Significantly decreased levels of serum IgG and IgA, but not IgM,
IgD, or IgE, were reported in workers occupationally exposed to
metallic mercury vapours for 20 years in comparison with unexposed
controls (Moszczynski et al., 1990b). These workers had blood mercury
levels of < 50 µg/litre. Similarly, significantly decreased IgG and
IgA levels were observed in workers with urinary mercury levels of
0.029-0.545 mg/litre (Bencko et al., 1990). Studies of a small number
of people exposed to mercury in dental amalgams have shown increased
levels of IgE (Anneroth et al., 1992), an increased incidence of
asthma (Drouet et al., 1990), and development of contact dermatitis
(Gonçalo et al., 1992). Total lymphocyte, CD4, and CD8 levels were
higher in exposed people than in controls (Eedy et al., 1990).
Epidemiological data indicate that the main effects of
occupational exposure to lead are on cellular aspects of the immune
system and that humoral parameters remain relatively insensitive to
such exposure. Thus, serum IgG, IgM, and IgA levels remained within
the normal range in workers exposed for 4-30years, with a mean blood
lead level of 38.4 µg/dl, in comparison with a mean control level of
11.8 µg/dl (Kimber et al., 1986). Similarly, no effects were noted on
serum IgG or IgA levels in a cohort exposed to lead oxides at a
reported concentration within the plant of 266 µg/m3, who had an
estimated blood lead level of 64 µg/dl; however, the response of
lymphocytes from the exposed group to stimulation with
phytohaemagglutinin and concanavalin A in vitro was significantly
lower than that of controls (Alomran & Shleamoon, 1988). Decreased
serum Ig levels were reported in occupationally exposed workers with a
mean blood lead level of 46.9 µg/dl, but the duration of exposure was
not reported (Castillo-Mendez et al., 1991). In another study, no
significant effects were noted on serum immunoglobulin levels after
exposure to lead, but the levels of secretory IgA, which plays a major
role in the defence against respiratory and gastrointestinal
infections, was significantly decreased in workers with blood lead
levels of 21-90 µg/dl. The incidence of influenza infections per year
was significantly higher in these workers than in the control group
(Ewers et al., 1982).
Studies on the effects of lead on lymphocyte levels in
occupationally exposed workers have produced inconclusive results. One
study found an increase in absolute B lymphocyte counts and CD8 cells
(Coscia et al., 1987), while a decrease in total T lymphocytes and the
CD4 subset was reported in another set of workers (Fischbein et al.,
1993).
2.4.6 Solvents
Certain organic solvents may induce immune changes in humans.
Benzene-induced pancytopenia with associated bone-marrow hypoplasia, a
classical sign of chronic exposure to benzene, results in an
immunodeficiency state due to the reduced numbers of immunocompetent
cells (Goldstein, 1977). Alterations in the numbers of certain
lymphocyte subsets, e.g. CD3 and CD4 lymphocytes, have also been
reported in workers exposed to solvents (Denkhaus et al., 1986),
suggesting that the effects may be somewhat specific.
2.4.7 Ultraviolet radiation
Numerous reports have shown that UVR inhibits contact
hypersensitivity of the skin to sensitizers such as
dinitrochlorobenzene (DNCB) (O'Dell et al., 1980; Halprin et al.,
1981; Hersey et al., 1983a,b; Kalimo et al., 1983; Sjovall et al.,
1985; Friedmann et al., 1989; Yoshikawa et al., 1990; Vermeer et al.,
1991; Cooper et al., 1992). The dose required to produce
immunosuppressive effects in humans is similar to that in C57Bl/6
mice, the strain phenotypically most sensitive to UVR-induced immune
suppression (Noonan & Hoffman, 1994).
Human buttock skin was exposed to four daily doses of
144 mJ/cm2 UVR, and the irradiated site was sensitized with DNCB
immediately after the last exposure; the inner surface of the forearm
was challenged 30days later with DNCB and contact hypersensitivity
assessed. Forty percent of the volunteers failed to develop contact
hypersensitivity and were designated sensitive, suggesting that, as in
mice, susceptibility to UVR is genetically controlled. The sensitive
phenotype also appeared to be a risk factor for the development of
skin cancer (Yoshikawa et al., 1990). Suppression of contact
hypersensitivity is seen in a similar percentage of black-skinned
individuals, indicating that melanin cannot protect against this
phenomenon (Vermeer et al., 1991). In another study, human buttock
skin was exposed to 0.75 or two minimal erythemal doses (MED) of UVB
(1 MED = 29.1-32.5mJ/cm2, depending on the individual) for four days
and sensitized through irradiated skin immediately after the last
exposure to DNCB; subjects were challenged with diluted DNCB at a
distal site three weeks later. Some subjects were also exposed to four
MED (moderate sunburn) and sensitized three days later with DNCB.
Analysis of overall individual responses revealed decreased
frequencies of fully successful immunizations in all UVB-exposed
groups. The rate of immunological tolerance to DNCB (lasting up to
four months) in the groups that were initially sensitized on skin that
had received erythemagenic doses of UVB was 31%, whereas it was 7% in
unirradiated controls (Cooper et al., 1992).
A dose-response relationship was established in the studies of
Cooper et al. (1992) in a comparison of subjects with types I-III skin
(fair to moderately fair) who received various doses or schedules of
UVR from a bank of FS20 fluorescent sun lamps (rich in UVB) with
respect to their ability to mount an immune response to DNCB. A linear
inhibition of immune responsiveness was seen, the first detectable
decrease occurring at 0.75 of the individual's MED, reaching complete
inhibition of responsiveness for 95% of subjects when two MED were
administered every day for four days before immunization. Similar
inhibition occurred when DNCB was administered through skin that had
been exposed to a single dose of fourMED three days earlier. A dose-
response curve for fair-skinned subjects was constructed by plotting
the dose in total MED administered against the degree of immune
response to DNCB (Figure 35). The 50% immune suppressive dose was
calculated to be about 100mJ/cm2 of UVB.
Unresponsiveness to a contact sensitizer applied to UV-irradiated
skin can thus be induced in a proportion of individuals after exposure
to moderate levels of UVR, and at least some individuals become
immunologically tolerant in a manner similar to experimental animals.
Taken together, these data suggest that the systemic immunosuppression
induced in mice by UVR also occurs in humans, possibly through a
similar mechanism. UVR appears to alter antigen presentation and the
expression of Langerhans (CDla+DR+) cells, which is followed by an
influx of CDla+DR+ monocytes that preferentially activate CD4+
(suppressor-inducer) cells, which induce maturation of CD8+ Tc
Iymphocytes (Cooper et al., 1986; Baadsgaard et al., 1990). The UVR
wavelengths responsible for induction of CDla-DR+ cells are
predominantly within the UVB band and to a lesser extent in the C band
(Baadsgaard et al., 1987, 1989).
UVR from solarium lamps also suppressed NK cell activity in the
blood of subjects exposed for 1h per day for 12 days and tested one
and seven days after exposure; the activity returned to normal by 21
days after exposure (Hersey et al., 1983a). The effects of UVR on NK
cell activity are attributed to A radiation (Hersey et al., 1983b).
The depressed immune function observed in irradiated rodents
reflects anecdotal observations in humans, i.e. that exposure to
sunlight exacerbates certain infectious diseases, particularly those
involving the skin. For example, it was noted at the turn of the
century that smallpox lesions were worsened by exposure to sunlight
(Finsen, 1901), and herpetic lesions and viral warts may be
reactivated or exacerbated by sunlight (Giannini, 1990). It has also
been hypothesized that sunlight affects susceptibility to infection
with the bacteria that cause leprosy (Patki, 1991). Lesions of Herpes
simplex virus type I and type II can be reactivated by exposure to
UVR (Spruanoe, 1985; Klein, 1986). Using the criteria established by
Yoshida & Streilin (1990) for the UVB-sensitive phenotype, Taylor et
al. (1994) reported that 66% of individuals who have a history of
herpes lip lesions provoked by exposure to sunlight were sensitive to
UVB, in comparison with 40-45% of the general population and 92% of
skin cancer patients. Exposure of immunosuppressed patients to
sunlight can increase the incidence of viral warts caused by
papillomavirus (Boyle et al., 1984; Dyall-Smith & Varigos, 1985). It
is also known that UVR exacerbates the clinical course of systemic
lupus erythematosus, an autoimmune disease (Epstein et al., 1965). The
effects of UVR on the risk of infectious disease have been reviewed by
Koren et al. (1994). In contrast, certain infectious diseases appear
to be cured by sunlight; the most notable are erysipelas (Giannini,
1990), a skin disease caused by Streptococcus, and skin lesions
caused by Herpes zoster virus.
2.4.8 Others
A large number of therapeutic drugs and drugs of abuse may also
alter human immune function in humans. These include:
Therapeutic agents
Alkylating agents
Nitrogen mustards: cyclophosphamide, L-phenylalanine mustard,
chlorambucil
Alkyl sulfonates: busulfan
Nitrosoureas: carmustine (BCNU), lomustine (CCNU)
Triazenes: dimethyltriazenoimidazolecarboxamide (DTIC)
Anti-inflammatory agents
Aspirin, indomethacin, penicillamine, gold salts
Adrenocorticosteroids: prednisone
Antimetabolites
Purine antagonists: 6-mercaptopurine, azathioprine, 6-thioguanine
Pyrimidine antagonists: 5-fluorouracil, cytosine arabinoside,
bromodeoxyuridine
Folic acid antagonists: methotrexate (amethopterine)
Natural products
Vinca alkaloids: vinblastine, vincristine, procarbazine
Antibiotics: actinomycin D, adriablastine, bleomycin, daunomycin,
puromycin, mitomycin C, mithramycin
Antifungal agents: griseofulvin
Enzymes: L-asparaginase
Cyclosporin A
Estrogens: diethylstilbestrol, ethinylestradiol
Substances of abuse
Ethanol
Cannabinoids
Cocaine
Opiates
Adapted from Dean & Murray (1990)
3. STRATEGIES FOR TESTING THE IMMUNOTOXICITY OF CHEMICALS IN ANIMALS
3.1 General testing of the toxicity of chemicals
The fact that substances used in various aspects of modern life
can be simultaneously beneficial and harmful to human life creates a
legislative and regulatory dilemma. In order to balance the desire to
use the many new substances that enter the market every year and the
economic benefit that is associated with their use on the one hand
with the health and safety of the population on the other is an
important challenge to governmental authorities. Legislative and
regulatory efforts to minimize and control the risk of adverse effects
on human health has resulted in a system for assessing and classifying
the potential risk of exposure to chemicals. Potential adverse effects
can be assessed in studies with experimental animals. In conducting
such studies, attention must be paid to ethical and regulatory
requirements for animal welfare and to good laboratory practice.
In assessing and evaluating the toxic characteristics of a
chemical, its oral toxicity may be determined once initial information
has been obtained by acute testing. Toxicity is routinely tested
according to guidelines, one of which is guideline No. 407 of the
Organisation for Economic Co-operation and Development (OECD) for
testing of chemicals, the 'Repeated Dose Oral Toxicity - Rodent:
28-day or 14-day Study' (Organisation for Economic Co-operation and
Development, 1981). This guideline has undergone three revisions, the
most recent of which (January 1994) includes parameters of
immunotoxicological relevance (see Table 7). Depending on the amount
of a chemical to be produced and the expected exposure to the
chemical, testing according to this guideline may in many countries
provide the only information on its safety, including potential
toxicity to the immune system. The information yielded by this type of
testing is therefore decisive in determining how chemicals are used in
society. Subsequent guidelines have been defined for use in follow-up
studies if more exposure is expected or if there is a suspicion of
toxicity on the basis of structural analogy with other known
compounds. These include 90-day studies of oral toxicity, long-term
studies, and studies of reproductive effects. Although such guidelines
include more parameters of the immune system than OECD test guideline
No. 407, detection of potential immunotoxicity may still not be
adequately addressed. In practice, the best procedure is to carry out
appropriate tests on a case-by-case basis, at increasing levels of
complexity, when concerns are raised in more general toxicological
studies.
Table 7. Parameters of OECD Test Guideline 407 that relate to the immune system
Current Guideline 407 Proposal for updating Proposal for updating Proposal for updating
(adopted 21 May 1981) Guideline 406 Guideline 407 (revision of Guideline 407 (revision of
(February 1991) January 1993) January 1994)
Total and differential Total and differential Total and differential Total and differential
leukocyte count leukocyte count leukocyte count leukocyte count
Weight of spleen and thymus Weight of spleen and thymus Weight of spleen and thymus
Histopathology of spleen Histopathology of spleen, Histopathology of spleen, Histopathology of spleen, thymus,
thymus, lymph node, and thymus, lymph nodes (one lymph nodes (one relevant to the
bone marrow relevant to route of route of administration and a
administration and a distant distant one to cover systemic
one to cover systemic effects), effects), small intestine (including
and bone marrow Peyer's patches), and bone marrow
Histopathology of target Histopathology of target Histopathology of target Histopathology of target
organs organs organs organs
OECD, Organisation for Economic Co-operation and Development
An insight into the type of information that the OECD test
guideline No. 407 yields is given below. In this test, the substance
is administered orally in daily graduated doses to groups of
experimental animals, one dose per group for 28 or 14 days. The
preferred rodent species for this test is the rat, although others may
be used. At least three doses and a control should be used. The
highest dose should result in toxic effects but not produce an
incidence of fatalities which would prevent a meaningful evaluation;
the lowest dose should not produce any evidence of toxicity and should
exceed a usable estimate of human exposure, when available. Ideally,
the intermediate dose level(s) should produce minimal observable toxic
effects.
In compliance with the guideline, the following examinations are
carried out:
(a) haematology, including haematocrit, haemoglobin concentration,
erythrocyte count, total and differential leukocyte count, and a
measure of clotting potential such as clotting time, prothrombin
time, thromboplastin time, or platelet count;
(b) clinical biochemistry of blood, including blood parameters of
liver and kidney function. The selection of specific tests is
influenced by observations on the mode of action of the
substance. Suggested determinations are: calcium, phosphorus,
chloride, sodium, potassium, fasting glucose, serum alanine
aminotransferase, serum aspartate aminotransferase, ornithine
decarboxylase, gamma-glutamyl transpeptidase, urea nitrogen,
albumin, blood creatinine, total bilirubin, and total serum
protein. Other determinations that may be necessary for an
adequate toxicological evaluation include analyses of lipids,
hormones, acid-base balance, methaemoglobin, and cholinesterase
activity. Additional clinical biochemistry may be used when
necessary, to extend the investigation of any observed effects.
(c) pathology, including gross necropsy, with examination of the
external surface of the body, all orifices, and the cranial,
thoracic, and abdominal cavities and their contents. The liver,
kidneys, adrenal glands, and testes are weighed wet as soon as
possible after dissection to avoid drying. Liver, kidney, spleen,
adrenal glands, heart, and target organs showing gross lesions or
changes in size are preserved in a suitable medium for possible
future histopathological examination. Histopathological
examination is performed on the preserved organs or tissues of
the group given the high dose and the control group. These
examinations may be extended to animals in other dosage groups,
if considered necessary to further investigate changes observed
in the high-dose group.
A properly conducted 28- or 14-day study will provide information
on the effects of repeated doses and can indicate the need for
further, longer-term studies. It can also provide information on the
selection of doses for longer-term studies.
It is clear that the 1994 guideline is not suitable for adequate
assessment of the potential adverse effects of exposure to a test
chemical on the immune system, since the immunological parameters are
restricted to total and differential leukocyte counts and the
histopathology of the spleen. An evaluation of this test (Van Loveren
& Vos, 1992) indicated that over 50% of the immunotoxic chemicals in a
series of about 20 chemicals would not have been identified as such if
the tests had strictly adhered to the guideline. In fact, it is even
doubtful if chemicals indicated as immunotoxic only on the basis of
guideline No. 407 would in practice have been picked up: For instance,
in a toxicological experiment, a small but significant change in the
percentage of basophilic leukocytes would by itself probably not be
considered to be biologically relevant in the absence of any other
parameter to suggest that an effect on the immune system might have
been present.
These data indicate that extension of OECD test guideline No. 407
is necessary in order adequately to assess potential immunotoxicity.
It is recommended that additional immunological parameters be included
in this guideline in order to increase its power (Vos & Van Loveren,
1987; Basketter et al., 1994).
Guidelines also exist for follow-up studies if greater exposure
is expected or if there is a suspicion of toxicity on the basis of
structural analogy with other compounds. In these studies, potential
toxicity to the immune system is generally addressed somewhat more
extensively than in guideline No. 407. For instance, in a 90-day study
of oral toxicity, the OECD guidelines prescribe that histopathological
examination be done on the thymus, a representative lymph node, and
the sternum with bone marrow, in addition to the spleen. Even with
these additional parameters, it is highly questionable whether
potential immunotoxicity is adequately assessed. For this purpose, a
variety of tests is available, which are described in section 4.
Depending on what is already known about the toxicity of the test
compound, different panels of tests (also referred to as tiers) are
selected for immunotoxicological evaluation. Usually, if little or no
information is available, a dose range including high doses is used;
lower, overtly nontoxic doses are chosen if some knowledge is
available about the physical and chemical properties, toxicokinetics,
structure-activity relationships, and intended use.
3.2 Organization of tests in tiers
Immunotoxicity can be assessed in a tiered approach (Luster et
al., 1988; Van Loveren & Vos, 1989). Generally, the objective of the
first tier is to identify potentially hazardous compounds (hazard
identification). If potential immunotoxicity is identified, a second
tier of tests is carried out to confirm and further characterize the
immunotoxicity.
Various approaches have been suggested for evaluating the
potential immunotoxicity of compounds. Most are similar in design, in
that the first tier is usually a screen for immunotoxicity and the
second tier consists of a more specific confirmatory set of studies or
in-depth mechanistic studies. Since the use of the tiers is usually
tailored to the goals or objectives of the organization that proposes
them, they differ in respect of the specific assays recommended and
the organization of the assays into tiers.
3.2.1 United States National Toxicology Program panel
The tiers and assays originally adopted by the NTP, based on the
proposed guidelines for immunotoxicity evaluation in mice reported by
Luster et al. (1988), are shown in Table 8.
Table 8. Panel adopted by the US National Toxicology Program for detecting immune
alterations after exposure of rodentsa to chemicals or drugs
Parameter Procedures
Screen (Tier I)
Immunopathology Haematology: complete and differential blood count
Weights: body, spleen, thymus, kidney, liver
Cellularity: spleen
Histology: spleen, thymus, lymph node
Humoral immunity Enumerate IgM antibody plaque-forming cells to
T-dependent antigen (sheep red blood cells)
Lipopolysaccharide mitogen response
Cell-medicated Lymphocyte blastogenesis to mitogens
immunity (concanavalin A)
Mixed leukocyte response to allogeneic leukocytes
Nonspecific immunity Natural killer cell activity
Comprehensive (Tier II)
Immunopathology Quantification of splenic B and T cell numbers
Humoral-mediated Enumeration of IgG antibody response to sheep red
immunity blood cells
Cell-medicated Cytotoxic T lymphocyte cytolysis; delayed
immunity hypersensitivity response
Nonspecific immunity Macrophage function: quantification of resident
peritoneal cells, and phagocytic ability (basal and
activated by macrophage activating factor)
Host resistance Syngeneic tumour cells
challenge models PYB6 sarcoma (tumour incidence)
(end-points)b B16F10 melanoma (lung burden)
Bacterial models: Listeria monocytogenes (mortality);
Streptococcus species (mortality)
Viral models: influenza (mortality)
Parasite models: Plasmodium yoelii (parasitaemia)
a The testing panel was developed using B6C3F1 female mice.
b For any particular chemical tested, only two or three host resistance models
are selected for examination.
In this approach, the tier 1 assay is limited; it includes assays
for both cell-mediated and humoral-mediated immunity and for innate
(nonspecific) immunity with the inclusion of NK cell assays. It also
includes immunohistopathology, which is part of the standard protocol
for studies of subchronic toxicity and carcinogenicity conducted by
the NTP. Tier II represents a more extensive evaluation and includes
additional assays for assessing effects on cell-mediated, humoral, and
innate immunity, in addition to host resistance. In this approach,
animals are usually evaluated at only one time, so that the
possibility for recovery or reversibility of immunological changes is
not evaluated. A 14-day exposure period is employed routinely;
however, 30- and 90-day exposures have been used, depending on the
pharmacokinetic properties of the chemical being tested. The dose used
in this tier system tends to be lower than those in several of the
other approaches followed. In the NTP approach, dose levels are
selected whenever possible that have no effect on body weight or other
toxicological end-points. The approach has therefore focused on
compounds for which the immune system is the most sensitive target.
This is in marked contrast to other approaches, in which the highest
dose is usually the maximum tolerated dose.
The assays that make up the NTP tier approach have undergone
various revisions, partly on the basis of an immunotoxicological
review of compounds evaluated in this tier structure (Luster et al.,
1992). The mitogen assays were first moved from tier I to tier II and
have now been dropped altogether: They were found to be insensitive
and to add little when run in conjunction with other assays that have
a proliferative component, such as the mixed leukocyte response and
the plaque assay. Furthermore, the only macrophage phagocytic assay
routinely carried out in immunotoxicological studies conducted for the
NTP is evaluation of the functional activity of the mononuclear
phagocyte system, which is an in-vivo assay for phagocytosis.
Studies by Luster et al. (1992) show that the potential
immunotoxicity of a compound can be reasonably predicted with a few
selected assays. As additional data become available, further changes
to the NTP tiers will most likely be forthcoming.
3.2.2 Dutch National Institute of Public Health and Environmental
Protection panel
The tier approach for immunotoxicological evaluation followed at
the National Institute of Public Health and Environmental Protection
(RIVM) in the Netherlands (Vos & Van Loveren, 1987) is shown in
Table 9. This approach is based essentially on OECD test guideline
No. 407, which suggests that the maximum tolerated dose be used as the
high dose in the study. As a result, significantly higher doses are
used than in the NTP approach in evaluating compounds for
immunotoxicity. Additionally, the standard exposure period is 28 days,
and the animal species routinely used is the rat instead of the mouse.
This type of testing can therefore be performed in the context of
studies in rats to determine the toxicological profile of a compound.
At least three doses should be used, the highest having a toxic effect
(but not mortality) and the lowest producing no evidence of toxicity.
Moreover, immunotoxicity tests carried out in the context of such
testing should not in any way influence the toxicity of the chemical
(e.g. immunization or challenge with an infectious agent). In the NTP
panel, the highest dose to which mice are exposed is chosen so that no
overt toxicity, i.e. changes in body weight or gross pathological
effects, is observed. As tests for immunotoxicity must be fairly
sensitive in order to preclude false negatives, the NTP tier I
includes functional assays. With a broader dose range that includes
overt toxicity, potential immunotoxicity is more likely to be
observed, without the inclusion of functional tests. If functional
assays are to be included in the first tier, those tests that require
sensitization of animals would require inclusion of satellite groups.
In OECD test guideline No. 407 for testing chemicals, none of the
other systems is approached functionally.
It has been suggested that the NK cell assay be added to tier 1 (Van
Loveren & Vos, 1992). Since the assay does not require animals to be
sensitized or challenged, the same animals can be used without
affecting other toxicological parameters, and thus an additional
satellite group of animals is unnecessary.
3.2.3 United States Environmental Protection Agency, Office of
Pesticides panel
The United States Environmental Protection Agency has proposed a
tiered approach to the evaluation of biochemical pest control agents,
which fall under the subdivision M guidelines for pesticides (Sjoblad,
1988). The proposed tiers are shown below. Tier 1 of this approach
includes functional assays for evaluating humoral immunity, cell-
mediated immunity, and innate immunity. Thus, while Tier 1 is
considered by the Agency to be an immunotoxicity screen, it is much
more encompassing than the first tier of the other approaches. By
providing options in the selection of assays for tier 1, the approach
can easily accommodate both the rat and the mouse as the laboratory
animal species used. In this approach, the tier 2 studies provide
information sufficient for risk evaluation, including information on
the time course of recovery from immunotoxic effects and host
resistance to infectious agents and tumour models. Additional
functional tests would be required if a dysfunction were observed in
tier 1 tests or if data from other sources indicated the compound
could produce an adverse effect on the immune response.
Table 9. Methods for detecting immunotoxic alterations in the rat evaluated by the
Dutch National Institute of Public Health and Environmental Protection,
Bilthoven, Netherlands
Parameters Procedures
Tier 1
Nonfunctional Routine haematology, including differential cell counts
Serum IgM, IgG, IgA, IgE determination; lymphoid organ
weights (spleen, thymus, local and distant lymph nodes)
Histopathology of lymphoid tissues, including mucosa-
associated lymphoid tissue
Bone-marrow cellularity
Analysis of lymphocyte subpopulations in spleen by flow
cytometry
Tier 2
Cell-medicated Sensitization to T-cell dependent antigens (e.g. ovalbumin,
immunity tuberculin, Listeria), and skin test challenge
Lymphoproliferative response to specific antigens (Listeria)
Mitogen responses (concanavalin A, phytohaemagglutinin)
Humoral Serum titration of IgM, IgG, IgA, IgE responses to
immunity T-dependent antigens (ovalbumin, tetanus toxoid,
Trichinella spiralis, sheep red blood cells) by ELISA
Serum titration of T-cell-independent IgM response to
lipopolysaccharide by ELISA
Mitogen response to lipopolysaccharide
Macrophaand Phagocytosis and killing of Listeria by adherent spleen
function and peritoneal cells in vitro
Cytolysis of YAC-1 lymphoma cells by adherent spleen
and peritoneal cells
Natural killer Cytolysis of YAC-1 lymphoma cells by non-adherent
function spleen and peritoneal cells.
Host resistance Trichinella spiralis challenge (muscle larvae counts and
worm expulsion)
Listeria challenge (spleen and lung clearance)
Cytomegalovirus challenge (clearance from salivary gland)
Endotoxin hypersensitivity;
Autoimmune models (adjuvant arthritis, experimental
allergic encephalomyelitis)
Ig, immunoglobulin; ELISA, enzyme-linked immunosorbent assay
Subdivision M guidelines: proposed revised requirements by the US
Environmental Protection Agency for testing the immunotoxicity of
biochemical pest control agents
AI. Tier 1
A. Spleen, thymus, and bone-marrow cellularity
B. Humoral immunity (one of the following)
1. Primary and secondary IgG and IgM responses to antigen; or,
2. Antibody plaque-forming cell assay
C. Specific cell-mediated immunity (one of the following)
1. One-way mixed lymphocyte reaction assay; or,
2. Effect of agent on normal delayed-type hypersensitivity
response; or,
3. Effect of agent on generation of cytotoxic T-lymphocyte
response
D. Nonspecific cell-mediated immunity
1. Natural killer cell activity and
2. Macrophage function
II. Tier 2
A. Required if:
1. Dysfunction is observed in tier 1 tests
2. Tier 1 test results cannot be definitively interpreted
3. Data from other sources indicate immunotoxicity
B. General testing features:
1. Evaluate time course for recovery from immunotoxic effects.
2. Determine whether observed effects impair host resistance to
infectious agents or to tumour cell challenge.
3. Perform additional specific, but appropriate, testing
essential for evaluation of potential risks.
This Agency has also suggested that immunotoxicological screening
be conducted in evaluating conventional chemical pesticides
(subdivision F guidelines; see below); however, unlike those of
subdivision M, these guidelines are not designed as a tiered testing
scheme. If the immunotoxicity screen listed in subheading I were added
to subchronic and/or chronic studies in subdivision F, it would be a
more effective screen for immunotoxicity than is currently available.
If this proposed screen indicates that the immune system is a
sensitive target, the Agency considers that it may be necessary to
evaluate the risk for immunotoxic effects as under subheading
II. Currently, these suggestions have not been promulgated as official
guidelines or regulations.
Evaluations suggested by the US Environmental Protection Agency as
appropriate additions to Subdivision F guidelines for immunotoxicity
screening (subheading I) and possible additional data appropriate for
risk evaluation of chemical pesticides (subheading II)
I. Immunotoxicity screen
A. Serum immunoglobulin levels (e.g. IgG, IgM, and IgA)
B. Spleen, thymus, and lymph node weights
C. Spleen, thymus, and bone-marrow cellularity and cell
viability
D. Special histopathology (e.g. enzyme histochemistry,
immunohistochemistry)
E. More complete evaluation of 'premature' mortality of test
animals, as possibly related to immunosuppressive effects
II. Immunotoxicity risk evaluation
A. Host resistance to challenge with infectious agent and/or
tumour cells
B. Specific cell-mediated immune responses (e.g. mixed
leukocyte response, delayed-type hypersensitivity
response,cytotoxic T lymphocyte assays)a
C. Nonspecific cell-mediated immune responses (i.e. natural
killer cell activity, macrophage function)a
D. Time course for recovery from adverse immunological effects
a Measures of specific and nonspecific cell-mediated immune
responses that also may be considered useful in an immunotoxicity
screen
3.2.4 United States Food and Drug Administration, Center for Food
Safety and Applied Nutrition panel
The United States Food and Drug Administration is considering
testing guidelines for evaluating the immunotoxic potential of direct
food additives (Hinton, 1992). The multifaceted approach is included
in the draft revision of the Toxicological Principles for the Safety
Assessment of Direct Food Additives and Color Additives Used in Food
is (US Food and Drug Administration, 1993). The testing requirements
are based on the 'concern level' of the substance, assigned on the
basis of the available toxicological information or the substance's
structural similarity to known toxicants and on estimated human
exposure from its proposed use. A compound with high toxic potential
and high exposure would be assigned a high initial 'concern level'
(3), and one with low toxic potential and low exposure would be
assigned a low initial level (1).
In general, substances will be evaluated for immunotoxic
potential on a case-by-case basis. Two types of immunotoxicity tests
and procedures are defined in this approach: Type 1 tests are those
that do not involve perturbation of the test animal (i.e.
sensitization or challenge). These are further divided into 'basic'
tests, which include haematology and serum chemistry, routine
histopathological examination, and determination of organ and body
weights, and 'expanded' tests, which are logical extensions of the
'basic' tests and include those that can be performed retrospectively.
Type 2 tests include injection of or exposure to antigens, infectious
agents, vaccines, or tumour cells. In general, type 2 tests require a
satellite group of animals for immunological evaluation. The sets of
'basic' and 'expanded' type 1 tests are defined as level-I
immunotoxicity tests, and the sets of type 2 tests are defined as
level-II tests. Some level-I tests can be used to screen for
immunotoxic effects, while others focus on the mechanism of action or
the cell types affected by the test substance. Level-II tests are
conducted to define the immunotoxic effects of food and colour
additives more specifically. The recommended testing scheme is shown
below.
The functional tests generally require sensitization of exposed
rats and controls and subsequent analysis of the responses to the
sensitizing antigens. For this reason, functional tests are not
readily conducted in the first tier of immunotoxicity testing. As
guidelines for routine toxicology experiments preclude compromising
the experiment by any agent other than the test chemical, the second
tier of immunotoxicity testing, with immune function tests, requires a
separate set of experiments. The antigens that are used to sensitize
the exposed and control rats may be relatively simple antigens, such
as ovalbumin or tetanus toxoid, or more complex antigens, such as
sheep red blood cells, bacteria, or parasites. The responses can occur
in various arms of the immune system, which consequently must be
measured with different assays. For instance, humoral responses can be
measured by determination of specific antibodies in serum; the
appropriate tests for cellular responses are proliferative responses
of lymphocytes to the specific antigens ex vivo/in vitro or delayed-
type hypersensitivity responses to injection with antigen in vivo.
Recommendations of the United States Food and Drug Administration for
testing the immunotoxicity of direct food additives
Basic testing (rat model)
Complete blood count, differential white blood cell count;
Total serum protein, albumin:globulin ratio;
Histopathology, gross and microscopic (spleen, thymus, lymph
nodes, Peyer's patches, and bone marrow);
Lymphoid organ and body weights
Retrospective level-I testing (possible in a standard toxicology
study)
Electrophoretic analysis of serum proteinsa (when positive or
marginal effect is noted in basic testing);
Immunostaining of spleen and lymph nodes for B and T cellsa
(quantification of total immunoglobulins);
Serum autoantibody screen and deposition of immunoglobulins
(micrometry for semiquantification of the proliferative
response)
Enhanced level-I testing (possible for more complete screening in the
standard toxicology study core group, with a satellite animal group,
or in a follow-up study)
Cellularity of spleen (lymph nodes and thymus when indicated);
Quantification of total B and T cells (blood and/or spleen);
Mitogen stimulation assays for B and T cells (spleen);
Natural killer cell functional analysis (spleen);
Macrophage quantification and functional analysis (spleen);
Interleukin-2 functional analysis (spleen);
When indicated or for more complete analysis, other end-points
such as total haemolytic complement activity assay in serum
Level-II testing with a satellite group or follow-up study for
screening of functional immune effects
Kinetic evaluation of humoral response to T-dependent antigen
(primary and secondary responses with sheep red blood cells,
tetanus toxoid, or other);
Kinetic evaluation of primary humoral response to a
T-independent antigen such as pneumococcal polysaccharides,
trinitrophenyl-lipopolysaccharide, or other recognized
antigens;
Delayed-type hypersensitivity response to known sensitizer of
known T effector cell;
Reversibility evaluation
Recommendations (cont'd)
Enhanced level-II testing with a satellite group or follow-up study
for evaluation of potential immunotoxic risk
Tumour challenge (MADB106 or other in rat);
PYB6 sarcoma (in mouse);
Infectivity challenge (Trichinella, Candida or other in rat;
Listeria or other in mouse) a Recommended for inclusion in
basic testing
a Recommended for inclusion in basic testing
Not all functional assays require prior sensitization of the test
animals, e.g. proliferative responses of lymphocytes ex vivo/in vitro
to mitogens which are either specific for T cells, giving information
on cellular immunity, or for B cells, providing data on humoral
immunity. The phagocytic and lytic activity of macrophages and the
nonspecific cytotoxic activity of NK cells can also be measured
ex vivo/in vitro, without prior sensitization of the test animals.
Both types of activity are examples of nonspecific defence mechanisms,
directed to bacteria and certain tumour cells and to tumour cells and
virally infected cells, respectively. Since measurement of these types
of activity does not require prior sensitization of the host, such
functional tests can be considered for inclusion in the first tier of
testing for immunotoxicity in routine toxicology.
3.3 Considerations in evaluating systemic and local immunotoxicity
3.3.1 Species selection
Selection of the most appropriate animal model for
immunotoxicology studies has been a matter of great concern. Ideally,
toxicity testing should be performed with a species that responds to a
test chemical in a pharmacologically and toxicological manner similar
to that anticipated in humans, i.e. the test animals and humans
metabolize the chemical similarly and have identical target organs and
toxic responses. Toxicological studies are often conducted in several
animal species, since it is assumed that the more species that show a
specific toxic response, the more likely it is that the response will
occur in humans. Data from studies in rodents on target organ toxicity
at immunosuppressive doses for most immunosuppressive therapeutic
agents have generally been predictive of later clinical observations.
Exceptions to the predictive value of rodent toxicological data are
infrequent but occurred in studies of glucocorticoids, which are
lympholytic in rodents but not in primates (Haynes & Murad, 1985).
Although certain compounds exhibit different pharmacokinetic
properties in rodents and in humans, rodents still appear to be the
most appropriate animal model for examining the non-species-specific
immunotoxicity of compounds, because of established toxicological
knowledge, including similarities of toxicological profiles, and the
relative ease of generating data on host resistance and immune
function in rodents. Comparative toxicological studies should be
continued and expanded, however, as novel recombinant biological
compounds and natural products that enter safety testing will probably
have species-specific host interactions and toxicological profiles.
The quantitative and possibly the qualitative susceptibility of
an individual animal to the immunotoxicity of a selected agent can be
influenced by its genetic characteristics, indicating not only a need
to consider species but also strain. Rao et al. (1988) described two
approaches for selecting appropriate genotypes for toxicity studies.
The first is to select genotypes that are representative of an animal
species, which by virtue of similar metabolic profiles may also
exhibit a sensitivity similar to that of man, such as random-bred
mice. A second approach is to attempt to identify genotypes that are
uniquely suitable for evaluating a specific class of chemicals, such
as the use of Ah-responsive rodent strains in studies with
polyhalogenated aromatic hydrocarbons. In many cases, however, this
approach requires considerable knowledge of the mechanisms of toxicity
of the compound, which may not be available. One compromise has been
to use Fl hybrids which have the stability, phenotypic uniformity, and
known genetic traits of an inbred animal, yet have the vigour
associated with heterozygosity. The description of the genetic
relationships between inbred mouse strains on the basis of the
distribution of alleles at 16 loci (Taylor, 1972) has made possible
rational selection of appropriate Fl hybrids, such as the B6C3Fl
mouse.
3.3.2 Systemic immunosuppression
Because of this complexity, the initial strategies devised by
immunologists working in toxicology and safety assessment were to
select and apply a tiered panel of assays in order to identify
immunosuppressive or, in rare instances, immunostimulatory agents in
laboratory animals (US National Research Council, 1992). Although the
configuration of these testing panels varies according to the
laboratory conducting the test and the animal species used, they
include measurements of one or more of the following: (i) altered
lymphoid organ weights and histology, including immunohistology; (ii)
quantitative changes in the cellularity of lymphoid tissue, peripheral
blood leukocytes, and/or bone marrow; (iii) impairment of cell
function at the effector or regulatory level; and/or (iv) increased
susceptibility to infectious agents or transplantable tumours.
A variety of factors must be considered in evaluating the
potential of an environmental agent or drug to adversely influence the
immune system. These include appropriate selection of animal models
and exposure variables, consideration of general toxicological
parameters and mechanisms of action, as well as an understanding of
the biological relevance of the end-points to be measured. Treatment
conditions should be based on the potential route, level, and duration
of human exposure, the biophysical properties of the agent, and any
available information on the mechanism of action. Moreover,
toxicokinetic parameters, such as bioavailability, distribution
volume, clearance, and half-life, should be measured. Doses should be
selected that will allow establishment of a clear dose-response curve
and a no-observed-effect level NOEL). Although, for reasons explained
earlier, it is beneficial to include a dose that induces overt
toxicity, any immune change observed at that dose should not
necessarily be considered to be biologically significant, since severe
stress and malnutrition are known to impair immune responsiveness.
Many laboratories routinely use three doses but generally conduct
studies to define the range of doses before a full-scale
immunotoxicological evaluation. If studies are being designed
specifically to establish reference doses for toxic chemicals,
additional exposure levels are advisable. In addition, inclusion of a
'positive control' group, treated with an agent that shares some of
the characteristics of the test compound, may be advantageous when
experimental and fiscal constraints permit.
The selection of the exposure route should reflect the most
probable route of human exposure, which is most often oral,
respiratory, or dermal. If it is necessary to deliver an accurate
dose, a parenteral exposure route may be required; however, this may
significantly change the metabolism or distribution of the agent from
that which would occur following natural exposure.
3.3.3 Local suppression
Local immune suppression has received less attention than
systemic immune suppression, and this is noteworthy, since the surface
that is exposed to the environment, i.e. the skin, the respiratory
tract, and the gastrointestinal tract, are the major ports of entry of
antigens and pathogens. While a variety of validated methods are
currently available to detect chemical skin sensitizers in humans and
experimental animals, there is no standard method to assess the
potential of chemicals to induce local immunosuppression in the skin.
Furthermore, although increasing evidence suggests that the
consequence of skin immunosuppression would be an increase in
neoplastic and infectious diseases of the skin, definitive data are
still lacking. In contrast, considerable efforts are being deployed to
develop sensitive models for monitoring skin irritants. For example,
human keratinocyte cultures and keratinocyte-fibroblast co-cultures
have been examined for end-points ranging from changes in cell
viability to production and loss of various bioactive products. Few
test systems are available for the gut and the respiratory tract.
4. METHODS OF IMMUNOTOXICOLOGY IN EXPERIMENTAL ANIMALS
This section comprises general descriptions of methods used for
evaluating immunotoxicity.
4.1 Nonfunctional tests
4.1.1 Organ weights
It is routine practice in toxicology to weigh organs that are
potentially affected by the compound that is being investigated. The
immunological organs that are suitable for weighing in screening for
potential immunotoxicity are: the thymus, which plays a decisive role
in the development of the immune system and which is affected by many
immunotoxicants; the spleen, which is the repository for many
recirculating lymphocytes; and the lymph nodes, which are important
for the induction of immune processes. Determination of the weight of
draining lymph nodes (depending on the route of exposure, i.e.
mesenteric nodes for oral exposure and bronchial nodes for inhalation)
in addition to distant lymph nodes (such as popliteal lymph nodes for
determining systemic effects) is the best. Mesenteric nodes, in
particular, occur in a string within non-lymphoid fatty tissue, and
care must be taken to remove this non-lymphoid tissue so that the
weight can be adequately determined. The cellularity of these organs
is another indication of the effects of chemicals on the immune
system. Furthermore, cell suspensions can be prepared from lymphoid
organs in order to assess the distribution of subpopulations of
lymphoid cells and to test their functionality within the organs.
Under OECD guideline No. 407, histopathological examination of
lymphoid organs and tissues is crucial for detecting the effects of
chemicals on the immune system. Therefore, upon termination of
exposure to a compound in a toxicological experiment, organs such as
the spleen should first be weighed; subsequently, they are divided
into parts which are also weighed, and one or more parts are used for
histopathological examination and the remainder to prepare cell
suspensions that can be evaluated for distribution of lymphocyte
subpopulations or can be assessed functionally.
4.1.2 Pathology
The histopathology of the thymus, spleen, and draining and
distant lymph nodes, of the mucosal immune system (Peyer's patches in
the gut or bronchus and nose-associated lymphoid tissue in the
respiratory tract), and of the skin immune system should be evaluated,
depending on the route of exposure. The first level of evaluation
should be of haematoxylin-eosin stained, paraffin-embedded slides. A
more sophisticated level of evaluation is immunoperoxidase staining of
special cell types.
Many monoclonal antibodies are available for mice, rats, and
humans to detect differentiation antigens, cell adhesion molecules,
and activation markers on haematolymphoid and stromal cells involved
in immune responses. A list of some monoclonal antibodies that can be
used in the identification of leukocytes and stromal cells in (frozen)
sections of lymphoid tissue is presented in Table 10; a selection of
these is reviewed below.
For a further description of these markers, and the cells that
express them, reference may be made to the introductory section and to
descriptions in the literature (Brideau et al., 1980; Bazin et al.,
1984; Dallman et al., 1984; Dijkstra et al., 1985; Joling et al.,
1985; Vaessen et al., 1985; Joling, 1987; Hünig et al., 1989; Kampinga
et al., 1989; Portoles et al., 1989; Schuurman et al., 1991a).
These markers are usually stained in frozen tissue sections of
6-8 µm, fixed in acetone. A three-step immunoperoxidase procedure is
most suitable: the first step includes the monoclonal antibody
specific for the determinants to be studied (see above), the second
step, rabbit anti-mouse immunoglobulin, and the third step, swine
anti-rabbit immunoglobulin, the latter two antibodies conjugated to
horseradish peroxidase. The peroxidase activity can be developed by
3,3-diaminobenzidine tetrahydrochloride with hydrogen peroxide as
substrate, and the sections can be counterstained with Mayer's
haematoxylin to facilitate evaluation. Negative controls are prepared
by omitting the antibody in the first step or replacing it with an
irrelevant one. Under these conditions, only the peroxidase activity
of polymorphonuclear cells, when present, is visualized, and no
immunolabelling is found.
In general, histopathological evaluation provides a semi-
quantitative estimation of effects. The experienced pathologist can do
this adequately in studies carried out 'blind', especially if the
effects are clear. For more subtle effects, morphometric analysis is a
valuable addition, especially when supported by software for assessing
the values of parameters such as size, surface, and intensity of
staining. The compartments of the immune system, i.e. specific T and B
lymphocyte areas in spleen and lymph nodes and cortical and medullary
areas of immature and differentiated thymocytes within the thymus, and
the numbers of specialized cells per surface unit are parameters that
are well suited for morphometric analysis.
4.1.3 Basal immunoglobulin level
Serum immunoglobulin levels are often altered after exposure of
rats to immunotoxic chemicals (Vos, 1980; Vos et al., 1982, 1984,
1990a; Van Loveren et al., 1993a). This is not surprising, as the
total levels measured are a function of the humoral aspects of the
immune system, which react to the antigens that the host encounters.
For this reason, measurement of antibody levels is potentially
valuable in screening for immunotoxicity. Since the amount of antibody
Table 10. Some monoclonal antibodies to leukocytes and stromal cells used in immunohistochemical studies of tissue sections and flow
cytofluorography on cell suspensions
CD Relative Mouse Rat Human Reactivity
mol. mass
T Cells
CD1 gp43,45, Ly-38 OKT6, Lymphocytes in thymic cortex, Langerhans cells in skin,
49,12 a-Leu-6 interdigitating cells
CD2 gp50 Ly-37, MRC OX-34, a-Leu-5, All T cells in thymus and peripheral lymphoid organs, subset
NSM46.7, MRC OX-54, OKT11 of macrophages (rat). Sheep erythrocyte receptor, leukocyte
RM2-5 MRC OX-55 function antigen:-2 (LFA-2). Ligand f or LFA-3 (CD58)
CD3 gp19-29 CD3-1,KT3, IF4, G4.18 a-Leu-4, T Cells in thymic medulla and peripheral lymphoid organs
145-2C11 OKT3 (T-cell receptor-associated, cytoplasmic in precursor T cells
in thymus)
CD4 gp65 Ly-4, L3T4, MRC OX-35, a-Leu-3, Lymphocytes in thymic cortex, about two-thirds of T cells in
YTS 177.9 MRC OX-38, OKT4 peripheral lymphoid organs, subset macrophages, microglia
(ER2), W3/25 T helper/inducer and delayed-hypersensitivity phenotype.
MHC class II binding, receptor for human immunodeficiency
virus
CD5 gp65-62 Ly-1,Lyt-1 MRC OX-19, a-Leu-1 Lymphocytes in thymic cortex (faint). All T cells in thymic
HIS47 medulla and peripheral lymphoid tissue, subset of B cells
CD6 gp120 Tü 33 T Cells in thymic medulla and peripheral lymphoid organs
CD Relative Mouse Rat Human Reactivity
mol. mass
CD7 gp41 WT1,B-F12, Prethymic T-cell precursors, all T cells in thymus and fewer
a-Leu-9 in peripheral lymphoid organs
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
T Cells (contd)
CD8 gp32-33 Ly-2,3,Lyt-2,3, MRC OX-8 a-Leu-2, Lymphocytes in thymic cortex, about one-third of T cells in
YTS 105.8 OKT8 peripheral lymphoid organs, splenic sinusoids (T cytotoxic/
suppressor phenotype, NK cells). MHC class I binding
CD24 p45,55,65 J11d,M1/69 SRT1 BA-1 B Cells in germinal centres and corona, myeloid cells,
thymic cortex cells (rodents). Heat-stable antigen (HSA)
CD43 gp115 Ly-48 W3/13,HIS17 DFT-1, (Pro)thymocytes, T cells, plasma cells, cells in bone
WR-14 marrow, polymorphonuclear granulocytes, brain cells.
Leukosialin, sialophorin
CDw p25-30 Thy-1 (ER4), 5F10 Thymocytes, T lymphocytes, connective tissue structures,
90 MRC OX-7, epithelial cells, fibroblasts, neurons, subset of bone-marrow
HIS51 cells, plasma cells, stem cells (T-activation molecule)
Thy-2 Thymocytes
p40-55 H57-597 R73,HIS42 WT31, T-Cell receptor a-b chain. Mature T cells in thymic medulla
TalphaF1, and peripheral lymphoid tissue
TßF1
p40-55 GL3,GL4, V65 CgammaM1, T-Cell receptor gamma-delta chain
UC7-13D5 11F2,
TCR delta1,
deltaTCS1
p41-55 MRC OX-44 Prothymocytes, lymphocytes in thymic medulla, T and B
cells
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
T Cells (contd)
p41,47 MRC OX-2 Thymocytes, dendritic cells, B cells, brain cells
ER3,ER7, Subset of thymocytes and peripheral T cells, subset of
ER9,ER10 myeloid cells
HIS44 Most lymphocytes in thymic cortex, small subset of
medullary lymphocytes, erythroid cells, cells in germinal
centre
HIS45 Some lymphocytes in thymic cortex, most medullary
thymocytes and peripheral T cells, subset of B cells.
Quiescent cell antigen (QCA-1)
MHC class I (Various antibodies to polymorphic and All nucleated cells, including leukocytes and stromal
non-polymorphic epitopes) cells; for T cells absent on thymic cortex cells (human)
B Cells
CD9 gp24 BA-2 Germinal centres (faint); some cells in thymic cortex. Late
pre-B cells
CD10 gp100 BA-3, Germinal centres (faint); some cells in thymic cortex
W8E7 Common acute lymphoblastic leukaemia antigen (CALLA)
CD19 gp95 a-Leu-12, B Cells in germinal centres and mantles, follicular dendritic
B4, FMC63 cells
CD20 p35 Ly-44 B1, a-Leu-16 B Cells in germinal centres and mantles, follicular dendritic
cells
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
B Cells (contd)
CD21 gp140 B2,BL13, B Cells in germinal centres and mantles (faint), follicular
HB-5 dendritic cells (C3d receptor, CR2, receptor for Epstein-Barr
virus)
CD22 gp135 Lyb-8.2, a-Leu-14,To B Cells in germinal centres and mantles, cytoplasmic in
Cy34.1 To 15,RFB4, precursor B cells
SHCL-1
CD23 p45 Ly-42 a-Leu-20, Some B cells in marginal centres and mantles, activated B
Tü 1 cells, subset of follicular dendritic cells (IgE Fc receptor)
CD24 p45,55,65 J11d,M1/69 SRT1 BA-1 B Cells in germinal centres and corona, myeloid cells, thymic
cortex cells (rodents). Heat-stable antigen (HSA)
CD37 gp40-45 BL14 B Cells in germinal centres and mantles
CD38 gp45 a-Leu-17, Lymphocytes in thymic cortex, cells in germinal centres,
OKT10 plasma cells (immature lymphoid cells, plasma cells)
CDw75 p53? LN1, OKB4 B Cells in germinal centre, in corona (faint), macrophages,
epithelium
CD79a p33,40 mb-1 B Cells, Ig alpha chain
CD79b p33,40 B29 B Cells, Igß chain
p200 (HIS14) All B cells, including TdT+ precursors
p200 (HIS22) All B cells in corona, pre-B cells
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
B Cells (contd)
MHC class II (Various antibodies to polymorphic and B Lymphocytes, activated T cells, monocytes/macrophages,
non-polymorphic epitopes) interdigitating cells, Langerhans cells, epithelia, endothelia
B Cells (surface); in germinal center IgM+IgG+IgA+ and in
anti-immunoglobulin corona IgM+IgD+, plasma cells
(cytoplasmic)
Monocytes/macrophages, myeloid cells
CD13 p130-150 ER-BMDM-1 a-Leu-M7, Monocytes, granulocytes, dendritic reticulum cells
My7 (aminopeptidase N)
CD14 p55 ED9 UCH-M1, B-A8, Monocytes, some granulocytes and macrophages
a-Leu-M3
CD15 p170-190 a-Leu-M1 Granulocytes, some monocytes (lacto-N-fucose pentaosyl)
CD16 p50-70 a-Leu-11 NK cells, subset of T cells, neutrophilic granulocytes,
activated macrophages. IgG-FcRIII, low affinity, complexed IgG
CD33 p67 a-Leu-M9, (Precursor) granulocytes, macrophages, Langerhans cells.
My9 Myelin-associated protein
CD68 p110 Ki-M6,Ki-M7 Macrophages (specific)
p160 F4/80 Monocytes-macrophages
p32 Mac-2 Thioglycollate-elicited peritoneal macrophages
p92-110 Mac-3 Peritoneal macrophages
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
Monocytes/macrophages, myeloid cells (contd)
4F7 Dendritic cells in skin, bone marrow
ED1 Monocytes/macrophages
ED2,HIS36 Subset of macrophages (F4/80-like)
ED3 Subset of macrophages, restricted, negative in thymus
MRC OX-41 Granulocytes, macrophages, dendritic cells
MRC OX-62 Dendritic cells (integrin-like)
(IF119) Dendritic cells
HIS48 Granulocytes
Mac-387 Macrophages
Natural killer cells
CD16 p50-70 a-Leu-11 NK cells, subset of T cells, neutrophilic granulocytes, activated
macrophages. IgG-FcRIII, low affinity, complexed IgG
CD56 p220/135 a-Leu-19, NK cells, monocytes, neuroectodermal cells NKH-1, isoform of
B-A19 neural cell adhesion molecule (NCAM)
CD57 p110 a-Leu-7, NK cells, subset of T cells, some B cells, some epithelial cells,
VC1.1 monocytes, neuroendocrine cells, NKH-1
a-asialo-GM1 NK cells, stromal components
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
Natural killer cells (contd)
NK-1.1,2B4, 3.2.3 NK cells (NKR-P1 gene family)
3A4, 5E6
Follicular dendritic cells
ED5,ED6, Ki-M4,DRC-1 Follicular dendritic cells
MRC OX-2
Epithelial cells (thymus)
(Various anti-keratin antibodies) Epithelium
(ER-TR4),4F1 HIS38 TE-3,(MR3, Thymic cortex epithelium
MR6)
(ER-TR5),IVC4 (HIS39) TE-4,(MR19), Thymic subcapsular or medullary epithelium
RFD4
Complement receptors
CD11b p160 Ly-40, MRC OX-41, Mac-1, Granulocytes, macrophages, CD5+ B cells, C3b1R, CR3
M1/70 MRC OX-42, a-Leu-15
WT.5
CD21 gp140 B2,BL13 B Cells in germinal centres and mantles (faint), follicular
dendritic cells (C3d receptor, CR2, receptor for Epstein-Barr
virus
CD35 p220 To 5 Follicular dendritic cells, macrophages, B cells in corona
(faint), renal glomerular epithelium. C3bR, CR1
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
IgG-Fc receptors
CD16 p50-70 a-Leu-11 NK cells, subset of T cells, neutrophilic granulocytes, activated
macrophages; IgG-FcRIII, low affinity, complexed IgG
CD32 gp140 Ly-17 3E1,CIKM5 B Cells, myeloid cells, macrophages; IgGFcRII, low affinity,
complexed IgG
CD64 p75 32.2 Monocytes; IgG-FcRI, high affinity, monomeric IgG
ß1-Integrin (CD29-CD49) family
CD29 p130 9EG7 B-D15 Ubiquitous, not on erythrocytes; ß1 chain of all CD49 antigens
CD49a p200 TS2/7 Activated T cells, monocytes, smooth muscle cells. Very late
antigen-1 (VLA-1), ligand of collagen, laminin
CD49b p155 31H4,AK7, T Cells, B cells, thrombocytes, fibroblasts, endothelium.
P1E6 Very late antigen-2 (VLA-2), ligand of collagen I, II, III,
and IV, laminin
CD49c p145 11G5,P1B5 B Cells, renal glomeruli, basal membranes. Very late antigen-3
(VLA-3), ligand of collagen, laminin, fibronectin, and invasin
CD49d p150 R1-2, P12520, HP2/1,44H6, Thymocytes, lymphocytes, monocytes, NK cells, eosinophilic
MFR4.B MR?4 L25.3 granulocytes, erythroblasts. Very late antigen-4 (VLA-4),
ligand of VCAM-1, fibronectin
CD49e p160 MFR5, SAM-1 Monocytes, leukocytes, œmemoryœ T cells, fibroblasts,
P12750 thrombocytes and muscle cells. Very late antigen-5 (VLA-5),
ligand of fibronectin
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
ß1-Integrin (CD29-CD49) family (contd)
CD49f p150 GoH3 Go-H3,4F10 T Cells, thymocytes, monocytes, thrombocytes. Very late
antigen-6 (VLA-6), ligand of laminin and invasin
ß2-Integrin (CD11-CD18) family
CD11a p180 Ly-15, WT.1 YTH-81.5, T and B cells, NK cells, erythroid and myeloid stem cells.
2D7, B-B15, Leukocyte function-associated antigen-1 (LFA-1) involved in cell
M17/4 G-25.2 adhesion, ligand for intercellular adhesion molecule (ICAM)-1
(CD54), ICAM-2 (CD102), ICAM-3 (CD50)
CD11b p160 Ly-40, MRC OX-41, Mac-1, Granulocytes, macrophages, CD5+ B cells. C3b1R, CR3
M1/70 MRC OX-42, a-Leu-15
WT.5
CD11c p150 a-Leu-M5, Monocytes, macrophages, granulocytes (faint), activated
S-HCL-3 lymphocytes. CR4
CD18 p95 YTS213.1, WT.3 BL5 All lymphocytes. ß-Chain of CD11 antigens
C71/16,
M18/2
p160-95 ED7,ED8 CD11-CD18 molecule
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
Others
Terminal deoxynucleotidyl transferase (TdT) Immature (lymphoid) cells in bone marrow and thymic cortex
(nuclear staining)
CD25 p55 Ly-43, AMT13, MRC OX-39 Tac,a-IL2-R Activated lymphocytes at scattered locations in thymus and
7D4,3C7 T-cell areas in peripheral lymphoid organs. Interleukin-2
receptor alpha chain
CD122 p75 5H4, CF1,Mik-ß2, NK cells, T cells, B cells, monocytes. Interleukin-2 receptor
TM-ß1 Mik-ß3 ß chain
CD26 p120 H194-112 MRC OX-61 134-2C2 (Activated) T cells. Dipeptidyl peptidase IV, in mouse T-cell
activation molecule (THAM)
CD30 p105 Ki-1,Ber-H2 Sporadic cells in thymic (cortex) and T cell areas in
peripheral lymphoid organs, some plasma cells. Activated
lymphocytes,Hodgkin cells
CD34 MY 10,8G12 Haematopoietic progenitor cells, capillary endothelium
QBEND/10 Human progenitor cell antigen (HPCA)
CD44 p65-85 Ly-24 MRC OX-49, a-Leu-44, Prothymocytes, T cells, small B cells. Lymphocyte homing
IM7 MRC OX-50 F10-44-2 receptor. Phagocytic glycoprotein-1 (PgP-1), HCAM
CD45 p180-210 Ly-5 MRC OX-1, T29/33 All leukocytes. Common leukocyte antigen
HIS41
Table 10 (cont'd)
CD Relative Mouse Rat Human Reactivity
mol. mass
Others (contd)
CD45R p190-220 B220 MRC OX-22, a-Leu-18, All B cells, subset of T cells. Common leukocyte antigen.
MRC OX-32, MB1,MT2 HIS24 restricted to strains of the RT7.2 allotype and labels
HIS24 all peripheral B cells except cells in marginal zone, pre-B
cells
CD45RA p205-220 14.8 MRC OX-33 HI100 B Cells, T cytotoxic-suppressor cells (faint), subset of
thymocytes.In humans, also CD4+ subset (naive-virgin
T-cells)
CD45RO p190-220 UCH-L1 T Cells in immature and memory stage. Common leukocyte
antigen
CD54 p90 KAT-1 1A29 84H10,B-C14 Endothelial cells, many activated cell types. Intercellular
3E2 HA58 adhesion molecule-1 (ICAM-1)
CD71 p95 YTA74.4,C2 MRC OX-26 B3/25 Proliferating cells in germinal centres, some cells in thymus
and T-cell areas in peripheral lymphoid organs, stromal
cells. Transferrin receptor
Ki-67 Proliferating cells in germinal centres, some cells in thymus
and T-cell areas in peripheral lymphoid organs. Proliferation
antigen present in late G1, S, G2, and M phases
PCNA PCNA PCNA Proliferating cells in germinal centres, some cells in thymus
and T-cell areas in peripheral lymphoid organs
MEL14 Recirculating T and B cells. Lymphocyte homing receptor
MHC, major histocompatibility complex; NK, natural killer; Ig, immunoglobulin
CD nomenclature from: Clark & Lanier (1989); Knapp et al. (1989); Schlossman et al. (1994, 1995)
Antibodies within parentheses are not commercially available.
in serum is a function of the antibody's half-life, the longer the
study the more likely it is that an effect will be observed.
Immunoglobulin levels can be influenced by the cleanness of the
facility: studies conducted in facilities with excellent husbandry
will have lower basal levels than those conducted in 'dirty'
facilities where animals are constantly exposed to foreign antigens,
including pathogens.
Measurement of basal immunoglobulin levels is useful only after
subchronic or chronic exposure, i.e. with sufficient time for normal
metabolic elimination. Basal levels of immunoglobulin decrease only
when synthesis is reduced or prevented such that metabolized
immunoglobulins are not replaced. The parameter therefore yields
little information about possible mechanisms of immunotoxicity, and
should rather be regarded as a screening parameter; this is in fact
true for most non-functional tests. The IgM and G classes have usually
been measured; however, since the two other classes (A and E) are
biologically very important (for instance in mucosal immunity and
allergic manifestations), they should also be measured.
Total IgM and IgG concentrations in serum can be analysed by
means of a 'sandwich' enzyme linked immunosorbent assay (ELISA), as
described by Vos et al. (1982). Total IgA and IgE concentrations can
be analysed in an essentially similar way, except that the microtitre
plates are coated with monoclonal anti-rat IgA (Van Loveren et al.,
1988b) or monoclonal anti-rat IgE antibodies (MARE-1), respectively,
and immunoglobulins bound to these antibodies in serum samples are
detected by sheep anti-rat IgA or monoclonal anti-kappa chains of rat
immunoglobulins (MARK-1), conjugated with peroxidase.
Data from the ELISA are usually reported as percentages of
control values, and a titration curve based on pooled sera is
prepared; optimal dilutions of exposed and unexposed groups are then
plotted from this curve. The deviation of the dilution of the test
groups from the control groups is expressed, with the dilution in the
control group set at 100%. While studies in rats indicate that
measurement of basal immunoglobulin levels is useful in predicting the
immunotoxic effects of compounds, studies conducted in mice at the NTP
do not, and measurement of basal immunoglobulins is not included in
either tier of their testing panel (Luster et al., 1988). There are
several possible reasons for the difference in the usefulness of basal
immunoglobulin levels in rats and mice. First, in the studies of Vos
and colleagues, cited above, the exposure period was routinely longer
than the 14-day studies conducted within the NTP; since serum antibody
level is a function of the antibody half-life, longer studies are more
likely to detect an effect. Furthermore, the doses used at the NTP
were often lower, by design, than those used in rats. A final possible
explanation, which remains to be confirmed, is that immunoglobulin
synthesis in rats is more sensitive than that in mice.
4.1.4 Bone marrow
Bone marrow is an important haematopoietic organ and a source of
precursors for lymphocytes and other leukocytes. Changes in the bone
marrow are therefore likely to result in alterations of
immunocompetent cell populations, which may be long lasting or
permanent and thus serve as an indicator of potential immunotoxicity.
In a study to validate immunotoxicological parameters, bone-marrow
cellularity was shown to be an indicator of the immunotoxicity of
cyclosporin A, used as the model compound (Van Loveren et al., 1993a).
Determination of cellularity in stained slides of bone marrow,
evaluation of smears, and actual counts of the numbers of cells within
bone marrow are practical. For this purpose, both ends of a femur are
cut off, and bone-marrow cells are collected by flushing balanced salt
solution through the femur with a 21-gauge needle. The concentration
of nucleated cells is determined in a Coulter counter; a differential
count of cells can be done visually in May-Grunvald Giemsa-stained
cytospin preparations.
4.1.5 Enumeration of leukocytes in bronchoalveolar lavage fluid,
peritoneal cavity, and skin
Mononuclear phagocytes in the alveoli of the lung play an
important role in clearing inhaled particles, including
microorganisms, from the lung. The numbers of cells and alterations in
their function can be end-points of the toxicity of inhaled chemicals.
In order to study these parameters, methods for harvesting the cells
from the lungs should be easy to perform, guarantee the sterility of
the cell harvest, and be standardized. Methods involve use of either a
syringe (Blusse Van Oud Alblas & Van Furth, 1979) or a complex system
of syringes, tubes, and valves (Moolenbeek, 1982). These methods often
result in contamination of the harvested cell population; moreover,
they are laborious and cannot easily be standardized since the syringe
is operated manually. In a more recently developed method (Van
Soolingen et al., 1990), an excised lung is placed in a pressure
chamber and connected to a cannula through which lavage fluid can be
introduced into the lung and transferred from the lung into a test
tube. This procedure is repeated several times to obtain an optimal
yield.
Enumeration of mononuclear cells in the peritoneal cavity can
also best be performed by harvesting these cells by lavage. Because
of the architecture of this organ, three or four cycles of
intraperitoneal injections of lavage fluid, followed by gentle
massaging of the abdomen, and aspiration of the fluid with the syringe
that was also used for injection suffice.
Langerhans cells in the skin can be enumerated with
histopathological techniques. Frozen tissue sections are used, stained
with immunoperoxidase techniques including markers for MHC class II
antigens or specific markers, as indicated above. Morphometric
analysis may provide a quantitative basis for this type of evaluation.
4.1.6 Flow cytometric analysis
Evaluation of phenotypic markers has proved to be one of the most
sensitive indicators of immunotoxic compounds. The availability of
fluorescent activated cell sorter (FACS) analysis units and
fluorescent cell counter units in immunotoxicology laboratories has
made analysis of cell populations routine. Determination of the
phenotype of lymphoid cells is a non-functional assay, although it has
often been inappropriately grouped with functional tests. The presence
or absence of a particular marker on the surface of a cell does not
reveal the functional capability of the cell. The usefulness of
surface marker analysis for predicting potential immunotoxicity has
been demonstrated. In studies conducted by Luster et al. (1992), a 91%
concordance was found for correct identification of immunotoxic
compounds on the basis of studies of surface markers alone.
As indicated above (section 4.1.2), numerous markers are
expressed on the cells of the immune system. Essentially, the same
reagents as used on tissue sections are applied on cells that have
been isolated from tissues, body fluids, or lavage fluids in
suspension (see above). Furthermore, both polyclonal and monoclonal
antibodies are available for detecting these surface markers. While
many of the markers have been used in immunological investigations,
very few have been evaluated with a large number of immunosuppressive
compounds. The markers that have been routinely used in studies of
immunotoxicity conducted for the NTP in mice and the cell types they
identify are shown in Table 11. The CD4:CD8 ratio in spleen has been
shown to concord best with the immunotoxicity of these surface markers
(Luster et al., 1992).
Table 11. Phenotypic markers on lymphocyte subpopulations used in
studies of immunotoxicity by the United States National
Toxicology Program
Surface marker Cell type
sIg+ Pan B cells
Thy 1.2+ or CD3+ Pan T cells
CD4+CD8- T Helper/delayed-type
hypersensitivity cells
CD8+CD4- T Suppressor-cytotoxic cells
CD8+CD4+ Immature T cells
The identification of phenotypic markers in rats has not
developed as rapidly as in mice; however, antibodies to rat cell
surface markers are now becoming available commercially and are being
used in immunotoxicological assessments (Smialowicz et al., 1990). The
monoclonal antibodies currently used for this purpose are: OX4 or
MARK-1 for B cells, W3/13 or OX 19 for T cells, R79 for the T cell
receptor, W3/25 for CD4 cells, and OX 8 for CD8 cells.
In enumerating the cell types in lymphoid tissue, both
percentages and absolute cell numbers should be reported. Of the two,
absolute cell numbers are by far the most meaningful. Compounds that
affect all populations equally and thus do not change the relative
percentages of the various cell types may be missed if only
percentages are evaluated. In addition, significant differences in the
magnitude of an effect on one or more of the populations can be
observed when the data are evaluated as absolute numbers and not as
percentages. As indicated above, the absolute changes more closely
reflect the events occurring in the animal and should thus be given
priority in interpreting data.
FACS analysis is also being used to determine the activation
state of various cell types, on the basis of changes in detectable
activation markers. Some of the activation markers that have been
studied are F4/80 (Austyn & Gordon, 1981), Mac-1 (Springer et al.,
1979), Mac-2 (Ho & Springer, 1984), transferrin receptor (Neckers &
Cossman, 1983), and IL-2 receptor (Cantrell et al., 1988). While
activation markers are of value in studying the mechanism of action of
compounds, their usefulness as predictors of immunotoxicity has yet to
be firmly established.
4.2 Functional tests
4.2.1 Macrophage activity
Phagocytic activity is the first line of defence against many
pathogens. Macrophages can phagocytose many particles, including
bacteria, and can lyse and inactivate them. Alterations in phagocytic
activity are therefore important potentially adverse effects of
chemicals on the immune system. The capacity to ingest particles
in vitro can be measured, and activity in vivo can be measured by
determining the clearance of bacteria, such as L. monocytogenes.
This test is dealt with in section 4.2.10.1.
Several assays have been developed for evaluating various types
of phagocytosis in mice and can also be used in rats, with slight
modifications. Innate and non-immune-mediated phagocytosis by
macrophages can be evaluated by determining the uptake of fluorescent
latex covaspheres (Duke et al., 1985). Macrophages and peritoneal
exudate cells are placed on a tissue culture slide and incubated with
the covaspheres for 24h on a rocking platform. The slides are then
fixed with methanol. The slide chambers are evaluated under a
fluorescent microscope, and macrophages with more than five latex
covaspheres are counted as positive for phagocytosis. The results are
expressed as percentage of phagocytosis, which is calculated as the
ratio of macrophages positive for phagocytosis to total macrophages
counted. In order to distinguish phagocytosed latex covaspheres from
those that are merely associated with the macrophage surface, the
cells are exposed for 30-60s to methylene chloride vapour. By
immersing the slides in this manner, the covaspheres that have not
been phagocytosed are dissolved, while those inside the macrophage
remain intact (Burleson et al., 1987). Phagocytosed fluorescent latex
particles can easily be quantified under the fluorescence microscope.
While this assay is straightforward, it is labour intensive, and
reading the slides, shifting back and forth from the fluorescent
field, and counting the macrophages is time-consuming.
A radioisotopic procedure, the chicken erythrocyte assay, can be
used to evaluate both adherence to and phagocytosis of particles by
macrophages. The phagocytic capacity is measured as an immunologically
mediated (Fc receptor) response. Macrophages are added to each well of
a 24-well tissue dish and allowed to adhere for a 2-3-h incubation
period. Nonadherent cells are washed, and chicken erythrocytes
labelled with 51Cr are added to each well; then a subagglutinating
dilution of antisera to chicken erythrocytes is added to each well and
the plate incubated for 1h. The plates are then washed to remove
unbound erythrocytes; an ammonium chloride solution is added to lyse
adhered erythrocytes, and the supernatant is collected and counted to
determine adherence of the erythrocytes to the macrophages. Next, both
the macrophages and the phagocytosed chicken erythrocytes are lysed by
addition of 0.1 N sodium hydroxide, and the solution is counted to
determine the amount of phagocytosis. Three to six wells in each group
do not receive 51Cr and are used to evaluate the DNA content
(Labarca & Paigen, 1980). The data are expressed as adherence counts
per minute, phagocytosed counts per minute, and specific activity for
adherence and phagocytosis. Specific activity is determined by
dividing the number of adhered or phagocytosed counts per minute by
the DNA content per well. The data must be expressed in terms of
specific activity, since compounds that affect the macrophages'
ability to adhere to the 24-well culture dish will significantly alter
the results obtained.
While both the nonspecific and immune-mediated phagocytosis
assays are useful for understanding the potential mechanisms of action
of compounds, changes in phagocytic activity in these in-vitro assays
have not been found to be predictive of immunotoxicity. For example, a
single intratracheal exposure to gallium arsenide resulted in
increased adherence and phagocytosis by chicken erythrocytes but
decreased phagocytosis of latex covaspheres (Sikorski et al., 1989).
The phagocytosis assay that is most predictive of altered
macrophage function is evaluation of the functional ability of the
mononuclear phagocyte system. This is a holistic assay for measuring
the capacity of the fixed macrophages of the mononuclear phagocyte
system, where macrophages provide the first line of defence against
both pathogenic and non-pathogenic blood-borne particles. The fixed
macrophages of the mononuclear phagocyte system line the liver
endothelium (Kupffer cells), the spleen, the lymph nodes (reticular
cells), the lung (interstitial macrophages), and other organs such as
the thymus and bone marrow. When the assay is conducted in mice, the
animals are injected intravenously with 51Cr-labelled sheep
erythrocytes, and a 5-µl blood sample is taken from the clipped tail
at 3-min intervals over a 15-min period. A final 30-min blood sample
is taken, and 1h after injection the animals are sacrificed and the
liver, spleen, lungs, thymus, and kidneys are removed, weighed, and
counted in a gamma counter. The 60-min time interval after injection
of sheep erythrocytes was selected as the time of sacrifice since it
represents the plateau for particle uptake by the selected organs
(White et al., 1985). Blood clearance of the radiolabelled cells is
expressed as vascular half-life and as a phagocytic index, which is
determined by the slope of the clearance curve. Organ distribution is
expressed as percent organ uptake and counts per minute per milligram
of tissue (specific activity). The assay can detect both stimulation
and inhibition of the mononuclear phagocyte system. Bick et al. (1984)
reported marked stimulation of the mononuclear phagocyte system after
treatment with diethylstilbestrol; more recently, morphine sulfate was
shown to decrease vascular clearance and hepatic and splenic
phagocytosis significantly (LeVier et al., 1993).
4.2.2 Natural killer activity
NK activity against neoplastic and virus-infected targets has
been clearly demonstrated in vitro and is thought to play an
important role in vivo in providing surveillance against neoplastic
cells and as a first line of defence against viruses (Herberman &
Ortaldo, 1981). In humans, rats, and mice, most cells with NK activity
can be identified by morphological (although the definition is not
morphological) and functional characteristics (Timonen et al., 1981).
Most of the cells that show NK activity are nonadherent, non-
phagocytic lymphocytes and are morphologically associated with large
granular lymphocytes (Timonen et al., 1982). Although cells with NK
activity do not strictly belong to the T-cell lineage, they can
express T cell-associated markers and express surface receptors, such
as those for the Fc portion of IgG and the ganglioside asialo GM1.
Some of these markers are also expressed by monocytes, macrophages,
and polymorphonuclear leukocytes (Herberman & Ortaldo, 1981). Within
4h, the cells can show nonantigen-specific cytotoxic activity
in vitro and in vivo against certain (NK-sensitive) tumour cell
lines and virus-infected cells.
The cells have enhanced cytolytic function after activation with
a variety of stimuli, including viral infection (Stein-Streinlein et
al., 1983), BCG (Tracey et al., 1977), IL-2 (Henney et al., 1981;
Domzig et al., 1983; Lanier et al., 1985; Malkovsky et al., 1987),
interferon, and interferon inducers (polyI:C) (Tracey et al., 1977;
Oehler & Herberman, 1978; Djeu et al., 1979a,b). NK activity in vitro
can be stimulated with IL-2 and interferon (Tracey et al., 1977; Djeu
et al., 1979b). Anti-asialo GM1 antibody can strongly inhibit
cytotoxic NK activity both in vitro and in vivo (Kasai et al.,
1980, 1981; Yosioka et al., 1986). This antibody binds to the cell
surface glycolipid GM1 and suppresses the lytic activity of effector
cells. Large granular lymphocytes are found in several lymphoid
organs. Many cells with high NK activity are found in spleen and
peripheral blood (Rolstad et al., 1986); lymph nodes have less NK
activity, and thymus and bone marrow show only marginal activity. NK
activity can also be demonstrated in the bronchus-associated lymphoid
tissue in the lungs. Moreover, large granular lymphocytes can migrate
from the circulation into the extravascular tissue and can even be in
contact with the lumen of the alveoli (Timonen et al., 1982; Reynolds
et al., 1984; Rolstad et al., 1986; Prichard et al., 1987). The
presence of large granular lymphocytes associated with NK activity in
the lungs is probably of great importance, because the lungs
constitute a major site for neoplastic disease (metastatic spread) and
viral infections. NK cells may also operate in certain types of
bacterial infections. In experimental animals, suppression of NK cell
activity increased the numbers of metastases after transplantation of
tumours.
The clinical significance of altered NK cell activity in humans
has not clearly been established. Asymptomatic individuals with low NK
cell responses may be at some risk for developing upper respiratory
infections and for increased morbidity (Levy et al., 1991); and
extreme susceptibility to severe and repeated herpes virus infection
was reported in an individual without NK cells (Biron et al., 1989).
It is obvious therefore that exposure to toxic substances that alter
NK activity can have biological consequences, and testing this
activity is important in assessing potential immunotoxicity.
The procedure for determining NK activity is as follows: cell
populations that exert NK activity (usually enriched peripheral blood
mononuclear cells or spleen cells) are cultured with NK-sensitive
target cells. A cell type frequently used for this purpose is the YAC
lymphoma cell line, which has been applied to mice, rats, humans, and
even seals. YAC lymphoma target cells are radiolabelled with 51 Cr,
and lysis of the cells, resulting in release of chromium, within 4h is
used to estimate the cytolytic activity of the NK cells within the
cell population. This assay has been used to demonstrate the effects
of numerous compounds on NK activity in rats (e.g. TBTO, ozone, and
HCB: Vos et al., 1984; Van Loveren et al., 1990c), mice (Luster et
al., 1992), and harbour seals (Ross et al., in press).
4.2.3 Antigen-specific antibody responses
Most antibody responses require not only B cells, which, after
maturation into plasma cells, produce antibodies, but also the help of
T lymphocytes. A variety of T cell-dependent antigens can be used for
this purpose, and an excellent one is tetanus toxoid. A typical
immunization schedule in rats comprises intravenous immunization on
day 0 followed by a booster on day 10. Primary and secondary IgG and
IgM responses can then be measured in serum, taken on day 10 (just
before the booster) and day 21, respectively. The primary IgM response
is the immunoglobulin response that is least under the control of T
cells. As tetanus toxoid is also used for human immunization, the
responses to this antigen may be useful in extrapolating experimental
data to humans. The responses can be determined in an ELISA (Vos et
al., 1979b).
Another widely used T cell-dependent antigen is ovalbumin. This
antigen can be and has been used to induce all classes of antibody
responses, i.e. IgM, IgG, IgA, and IgE, that can be measured with the
ELISA (Vos et al., 1980; Van Loveren et al., 1988b). The classical
assay of specific IgE responses is the passive cutaneous anaphylaxis
reaction. Serial dilutions are injected into the skin of rats,
sensitizing local mast cells; the specific antigen is then injected
intravenously, simultaneously with Evans blue. Mast cell products are
released where IgE meets the antigen, and IgE is cross-linked on the
membranes of the mast cells, leading to extravasation of Evans blue.
The titre can be determined from the magnitude of the reaction at each
dilution of IgE. ELISA techniques and the specific reagents to detect
IgE in an ELISA that are now available make this test preferable.
Ovalbumin induces not only humoral responses but also delayed-
type hypersensitivity. Sensitization to ovalbumin in Freund's complete
adjuvant enhances responses and makes it possibile to assay both
responses in one animal. Delayed-type hypersensitivity can also be
directed to purified protein derivative, with responses induced by the
adjuvant (Vos et al., 1980). At least in mice, however, immunization
in complete adjuvant skews responses in the direction of Th1
responses, i.e. delayed hypersensitivity, and hence suppresses Th2-,
IgE-, and IgA-dependent immune responses.
A few antigens can induce humoral immune responses without
involvement of T lymphocytes. One example is trinitrophenol-Ficoll
(lipopolysaccharide). Sensitization of animals to this antigen yields
immunoglobulin responses that can be measured in an ELISA. This is a
useful test for use in mechanistic studies to separate the effects of
compounds on B and T cells.
4.2.4 Antibody responses to sheep red blood cells
4.2.4.1 Spleen immunoglobulin M and immunoglobulin G plaque-forming
cell assay to the T-dependent antigen, sheep red blood cells
A widely used particulate T cell-dependent antigen is sheep red
blood cells. Antibody titres induced after sensitization can be
assayed with various techniques; one that is widely used is the
plaque-forming cell assay, or antibody-forming cell response. This
assay is relatively simple and can be conducted with inexpensive
equipment found in most laboratories, but the optimal concentration of
sheep red blood cells must be injected. As the antigenicity of red
blood cells varies significantly from sheep to sheep, time must be
invested to obtain cells from a sheep that repeatedly gives a high
response (>= 1500 plaque-forming cells/106 spleen cells). The
number of cells administered (about 2 × 108) should also be
optimized for both mice and rats in the laboratory conducting the
assay. The intravenous route is that preferred for sensitization;
intraperitoneal injections can be used but significantly increase the
potential for nonresponding animals as a result of a poor injection.
Animals are sacrificed on day 4 after injection, and spleen cells are
prepared by mincing the spleen between two frosted microscope slides,
teasing it apart with forceps, or passing it through a small mesh
screen; all of these methods are satisfactory, and that used to
prepare single splenocyte cultures varies from laboratory to
laboratory. An aliquot of cells is added to sheep erythrocytes and
guinea-pig complement; these are placed in a microscope slide chamber
when the Cunningham assay method is used (Cunningham & Szenberg,
1968), or, in the Jerne method, cells and guinea-pig complement are
added to a test tube containing warm agar and after thorough mixing
the test tube mixture is plated in a petri dish and covered with a
microscope cover slip (Jerne et al., 1974). In either case, the
preparations are then incubated at 37°C for 3-4h to allow plaques to
develop. The plaques are counted under a Bellco plaque viewer. A
plaque results from the lysis of sheep erythrocytes and is elicited as
a result of the interaction of complement and antibodies directed
against sheep erythrocytes, which are produced in response to the
intravenous sensitization. As each plaque is generated from a single
IgM antibody-producing plasma cell, the number of IgM plaque-forming
cells present in the whole spleen can be calculated. The data are
expressed as specific activity (IgM plaque-forming cells/106 spleen
cells) and IgM plaque-forming cells per spleen.
By incorporating rabbit anti-mouse or anti-rat IgG antibody into
the preparation of spleen cells, complement, and sheep red blood
cells, the number of IgG antibody-forming cells present in the spleen
can also be determined. This number is calculated by subtracting the
number of IgM plaque-forming cells from the total number of both IgM
and IgG plaque-forming cells. The optimal IgG primary response is
observed five days after sensitization (Sikorski et al., 1989).
The T-dependent IgM response to sheep red blood cells is one of
the most sensitive immunotoxicological assays currently in use. Luster
et al. (1992) reported that the individual concordance of the plaque-
forming cell assay for predicting immunotoxicity was the highest of
all the functional assays (78%). Furthermore, use of this assay in
combination with either NK cell activity or surface marker analysis
resulted in pairwise concordances for predictability of more than 90%.
While the plaque-forming cell assay has been shown to be
sensitive and predictive, the procedure does have its limitations. As
indicated earlier, the effect of the test compound on the immune
system is evaluated only in spleen cells, and effects on other
antibody-producing organs and tissues are not determined. The assay is
somewhat laborious, and it is preferable that several people
participate, to help in removing spleens, preparing cell preparations,
counting cells, and adding preparations to either microscope slide
chambers or agar dishes. An additional drawback is that the assay must
be conducted on the same day as the animals are sacrificed. This is in
marked contrast to the ELISA, in which sera can be frozen and
evaluated at a later date. While the slides and petri dishes can be
placed in a cold room or refrigerator and counted the next day, this
procedure is not recommended, as they tend to dry out to some extent,
making viewing and discerning plaques more difficult.
4.2.4.2 Enzyme-linked immunosorbent assay of anti-sheep red blood
cell antibodies of classes M, G, and A in rats
An alternative to the plaque-forming cell assay is ELISA of anti-
sheep red blood cell antibody titres in serum. Antigen preparations
made from ghosts of sheep erythrocytes by extraction with potassium
chloride are used to coat the bottoms of the wells of 96-well
microtitre plates. Serum samples from rats immunized with sheep
erythrocytes are titrated onto these plates using specific polyclonal
antibodies to rat IgM or IgG, to which peroxidase is conjugated. IgA
has also been assayed, using monoclonal anti-rat IgA antibodies and
polyclonal rat anti-mouse IgG conjugated with peroxidase. The ELISA of
serum titres of IgM, IgG, and IgA to sheep erythrocytes is an easy,
reliable method that can be used to detect the effects of chemicals on
the immune system of the rat (Van Loveren et al., 1991; Ladics et al.,
1995).
The assay measures titres of specific antibodies, in contrast to
the plaque-forming cell assay which determines the number of cells
that are actually responsible for production. The ELISA assesses the
production of antibodies, either per cell or in terms of the total
capacity of the host to produce these antibodies in vivo. In
interpreting the effects of exposure to chemicals, account must be
taken of the fact that the cells used in the assay are derived from
specialized parts of the body, such as the spleen, and alterations in
the numbers of antibody-producing cells in such an organ in rats
immunized with sheep red blood cells cannot give information on other,
inaccessible pools of antibody-producing cells. In the ELISA,
alterations in titres due to exposure to chemicals indicate changes in
the immune potential of the exposed animals. In screening for the
effects of chemicals on the immune system, therefore, ELISAs may be
preferable, but for studies on specific immunosuppressive mechanisms,
the plaque-forming cell assay, although labour- and time-intensive, is
a powerful tool for obtaining information complementary to the data
provided by the ELISA. Unfortunately, it is not always possible to
perform the two assays with material from the same animal. The peak
response in the plaque-forming cell assay in both rats (Fischer 344)
and mice (B6C3F1) occurs on day 4 after sensitization, while the peak
response in the ELISA occurs on day 6 for rats and day 4-5 for mice
(Temple et al., 1993). In order to detect the effects of chemicals on
the immune response to sheep red blood cells, it is preferable to
choose the optimal conditions, or to follow the kinetics, of the
response.
4.2.5 Responsiveness to B-cell mitogens
Responsiveness to lipopolysaccharide is another estimate of
humoral immune response, as solely B cells respond to this mitogen.
Although the responses of rats to this mitogen are less pronounced
than those of mice, good results can be obtained, and the
immunosuppressive effects of chemicals can be detected (Vos et al.,
1984).
An alternative B-cell mitogen is S. typhimurium mitogen (STM),
a water-soluble, proteinaceous extract derived from the cell walls of
S. typhimurium; it is a more potent mitogen for rat B lymphocytes
than lipopolysaccharide (Minchin et al., 1990). In both mice and rats,
the polyclonal activation of B lymphocytes is a multistep process. In
mice, mitogens alone can provide all the signals necessary for
proliferation and differentiation; in the rat, STM stimulation induces
B lymphocytes to proliferate without differentiating. The addition of
lymphokines to STM-stimulated B cells also failed to stimulate them to
differentiate (Stunz & Feldbush, 1986). Nevertheless, this mitogen is
useful for evaluating effects on the proliferative ability of rat B
lymphocytes. Smialowicz et al. (1991) showed a decrease in the STM
response in Fischer 344 rats after oral exposure to 2-methoxyethanol.
Unlike the bell-shaped mitogen dose-response curves observed with
T-cell mitogens, the proliferative response of B lymphocytes to both
lipopolysaccharide and STM rises quickly at low concentrations of the
mitogens and plateaus at higher concentrations. As a result, a single
concentration on the plateau phase of the mitogen response curve is
sufficient to evaluate the effects of a test compound on B-cell
mitogen-driven proliferation. One of the reasons that the mitogen
assays appear to be insensitive is that the cells must remain in
culture for several days in order to obtain a peak response. As a
result, they may recover from the immunomodulatory effects of the test
compounds during this in-vitro phase. This is a common problem with
many ex-vivo/in-vitro assays, including the cytotoxic T lymphocyte and
mixed leukocyte response assays; because of the short, 4-h period of
the NK cell assay, this is less of a concern.
4.2.6 Responsiveness to T-cell mitogens
The proliferative ability of T lymphocytes after stimulation with
mitogens can be measured by the uptake of 3H-thymidine in a manner
similar to that used to measure B-cell proliferation (Anderson et al.,
1972). Concanavalin A and phytohaemagglutinin are T-cell mitogens in
both rats and mice; pokeweed mitogen stimulates the proliferation of
both T and B cells and thus lacks specificity. Although both
concanavalin A and phytohaemagglutinin stimulate T lymphocytes,
T cells responsive to concanavalin A have been reported to be less
mature than those responsive to phytohaemagglutinin (Stobo & Paul,
1973). Multiple concentrations of these mitogens should be used to
ensure that a peak response is obtained: both produce a bell-shaped
dose-response curve, and too high a concentration can result in a
suboptimal response.
Historically, mitogens have been included in the battery of tests
for evaluating potential immunotoxicity, because the assay is one that
can also be carried out in humans. Human studies, however, are
conducted on peripheral blood, while most studies of rodent lymphocyte
transformation are conducted using spleen or lymph node cells. Thus,
the argument that the assay has clinical relevance is not well
founded. Furthermore, as the response of lymphocytes is extremely
robust, the assay lacks sensitivity. After a significant number of
compounds were evaluated for potential immunotoxicity in mitogen
assays, use of this assay was shifted from the tier 1 screen
originally described by Luster et al. (1988) to the tier 2
comprehensive evaluation. Use of the mitogen assay has now been
removed completely from studies conducted for the NTP, since other
assays in which cellular proliferation is required (e.g. plaque-
forming cell assay, mixed leukocyte reaction) were considered to be
more sensitive, and the data obtained from the mitogen assays add
little if any to an evaluation of the potential immunotoxicity of test
compounds.
4.2.7 Mixed lymphocyte reaction
In the mixed lymphocyte reaction (also known as mixed lymphocyte
culture), suspensions of responder T lymphocytes from spleen or lymph
nodes are co-cultured with allogeneic stimulator cells. The foreign
histocompatibility antigen (MHC class I or class II molecules)
expressed on the allogeneic stimulator cells serves as the activating
stimulus for inbred populations. In noninbred populations, a pool of
allogeneic cells can be used as stimulators. The assay analyses the
ability of T cells to recognize allogenic cells as 'non-self' as a
result of the presence of different MHC class II antigens on their
surface. In response to the class II antigens, the spleen or node
cells proliferate. Because a sufficiently large number of T cells in
the mixed lymphocyte population respond to the stimulator population,
the responder T cells need not be primed. Proliferation of the
responder cells is one of the parameters for T-cell responsiveness to
cellular antigens. If the allogeneic stimulator cell suspension
contains T cells, their uptake of 3H-thymidine must be prevented by
gamma-irradiation or mitomycin C, in order to preclude background
thymidine uptake.
4.2.8 Cytotoxic T lymphocyte assay
The Tc lymphocyte assay is a continuation of the mixed lymphocyte
reaction response in which the T lymphocytes further differentiate
into cytotoxic effector cells under the influence of various
cytokines. In mice, the assay is usually conducted using P815
mastocytoma cells as the sensitizer and target cell (Murray et al.,
1985). Mice are exposed in vivo to the test agent, and spleen cells
are then removed and placed in culture flasks with the P815
mastocytoma cells. After a five-day co-culture period, the spleen
cells are harvested and added to fresh P815 mastocytoma cells which
have been radiolabelled with 51Cr as sodium chromate. After a 4-h
incubation, the percentage cytotoxicity is determined by measuring the
specific release of 51Cr into the supernatant. The five days of
culture are necessary for the T lymphocytes to differentiate into
cytotoxic effector cells. Unfortunately, this extended period in
culture may give the spleen cells sufficient time to recover from any
adverse effects of the test compound, although such effects may have
been present at the time the spleen cells were removed from the
animal. This inherent limitation of the assay detracts from its
usefulness in assessing the immunotoxicity of test compounds.
A holistic Tc lymphocyte assay has been described, in which the
animal is sensitized after injection of the irradiated target cells
(Devens et al., 1985). Inhibiting the ability of the sensitizing cells
to proliferate either through irradiation or mitomycin C treatment
before injection prevents development of Tc lymphocytes in the animal.
Smialowicz et al. (1989) developed an assay in rats in which effector
cells are generated in culture by incubating cells with lymph node
cells from Wistar/Furth rats, and 51Cr-labelled W/Fu-G1 tumour cells
are used as the target cells. The assay requires four days in culture
and can be run simultaneously with the rat mixed lymphocyte reaction,
thus providing information on the test compound's ability to affect
proliferation and differentiation into effector cells.
4.2.9 Delayed-type hypersensitivity responses
Delayed-type hypersensitivity responsiveness is a reflection of
the capacity of the cellular immune system to execute immune responses
and especially those dependent on IL-2 and INF gamma, which include
attraction and activation of nonspecific mononuclear leukocytes
(macrophages-monocytes). Many systems can be used, depending on the
antigen. One is sensitization to BCG, followed by challenge with
purified protein derivative, to which sensitivity is induced. Another
example is ovalbumin, to which sensitization is most efficient if the
ovalbumin is emulsified in complete Freund's adjuvant. In this system,
delayed hypersensitivity can be measured to both purified protein
derivative and ovalbumin (Vos et al., 1980). Another antigen is
L. monocytogenes: This system is particularly interesting since it
can be used in the context of experiments in which host resistance to
this pathogen is also measured (Van Loveren et al., 1988a).
Delayed hypersensitivity responses can be measured after
sensitization to Listeria by subcutaneous injection of the test
antigen into the ears. Prior to and 24 and/or 48h after challenge, the
increment in ear thickness can be measured with a micrometer by a
person unaware of the experimental group. The background ear swelling
responses of similar, unimmunized control animals are subtracted from
the swelling responses found in immunized animals.
Several delayed-type hypersensitivity assays have been developed
and used for evaluating immunotoxicity in the mouse. Most have
involved measuring swelling in either the footpad or the ear after
sensitization and challenge with a protein antigen. Studies by
LaGrange et al. (1974) demonstrated that sheep erythrocytes could
elicit a delayed-type hypersensitivity response after a single
injection into the foot pad; however, more sheep erythrocytes were
needed to elicit the delayed-type hypersensitivity response than to
produce the optimal humoral immune response. Foot pad swelling can be
measured with a micrometer, as described for rats or by a more
objective, isotopic procedure, as described by Paranjpe & Boone (1972)
and Munson et al. (1982). The delayed-type hypersensitivity response
to sheep erythrocytes was previously considered to be a good assay for
detecting effects on cell-mediated immunity; however, the lack of
persistence of the response (LaGrange et al., 1974; Askenase et al.,
1977) and the possible contribution of antibody to the response raised
concern about the specificity of the assay when sheep erythrocytes are
used as the eliciting antigen. Benzo[ a]pyrene, a compound that
selectively affects humoral but not cell-mediated immunity in adult
mice, appears to decrease cell-mediated immunity when measured in the
sheep erythrocyte assay but has no effect on delayed-type
hypersensitivity when evaluated in the keyhole limpet haemocyanin
assay. The effect in the sheep erythrocyte assay is observed at doses
of benzo[ a]pyrene that decrease antibody production, suggesting a
significant antibody component of the swelling observed (White, 1992).
Keyhole limpet haemocyanin is another protein antigen used in
evaluating delayed-type hypersensitivity responses. Holsapple et al.
(1984) characterized the response to this antigen in the mouse,
showing that it produced the classical delayed-type hypersensitivity
response both with and without adjuvant. Two immunizations with
keyhole limpet haemocyanin were required, however, to produce a
response equivalent to one obtained with complete Freund's adjuvant.
In these studies, animals were sensitized with subcutaneous injections
of keyhole limpet haemocyanin in the shoulder area, with seven days
between the sensitizations. They were then challenged with the same
antigen injected intradermally into the central portion of the pinna
of one of the ears. Increases in ear thickness were evaluated by both
micrometer readings and radioisotopically. The unchallenged ear was
used as an individual control for each animal, and a group of
unsensitized but challenged animals was used to control for
nonspecific and background effects. The results indicated that,
whenever possible, the use of adjuvant in delayed-type
hypersensitivity studies should be avoided. Despite the fact that
complete Freund's adjuvant boosted the responses to keyhole limpet
haemocyanin, it partially masked the dexamethasone-induced suppression
of the response. In some cases, however, delayed-type hypersensitivity
responses are difficult to induce without adjuvant.
The studies currently conducted in mice and rats with this assay
are holistic assays for evaluating cell-mediated immunity. Since
sensitization and challenge occur in the intact animal, all components
of the immune system are present to respond in a physiologically
relevant manner. This type of assay is much more valuable for
evaluating the effects of compounds on cell-mediated immunity than are
in-vitro assays such as the mixed leukocyte response or Tc cell assay.
Luster et al. (1992) reported that the delayed-type hypersensitivity
response assay in mice was highly predictive (100% concordance) of
immunotoxicity when used in combination with the NK cell assay and the
plaque-forming cell assay.
4.2.10 Host resistance models
4.2.10.1 Listeria monocytogenes
Relevant mechanisms of defence against L. monocytogenes include
phagocytosis by macrophages and T cell-dependent lymphokine production
which enhances phagocytosis (Mackaness, 1969; McGregor et al., 1973;
Takeya et al., 1977; Pennington, 1985; Van Loveren et al., 1987).
Humoral immunity is not relevant in protection against infection in
this model. Clearance of Listeria after infection by, for instance,
the intravenous or the intratracheal route can be assessed at various
times after infection by determining the numbers of colony forming
units in the spleen or lungs, respectively. This can be done by
classical methods (Reynolds & Thomson, 1973) that involve the
following steps: Serial dilutions of homogenates of the organs,
prepared in mortars with sterile sea sand, are plated onto sheep blood
agar plates; after a 24-h incubation at 37°C, the colonies are counted
to determine the number of viable bacteria in the organ. Differences
in the numbers of bacteria retrieved from the organs are an indication
of the clearance of the bacteria, i.e. the rate at which the host
disposes of the bacteria after infection.
Histopathology after a Listeria infection can also be valuable.
For instance, exposure to ozone before an intratracheal infection with
Listeria affects pathological lesions due to the infection (Van
Loveren et al., 1988a): Pulmonary infection with Listeria induces
histopathological lesions characterized by foci of inflammatory cells,
such as lymphoid and histiocytic cells, accompanied by local cell
degeneration and influx of granulocytes. If rats are exposed to ozone
for one week before infection, the lesions are much more severe than
in unexposed animals and persist at times when either ozone-associated
or infection-associated effects alone would have resolved. The quality
of the lesions is also influenced by prior exposure to ozone: mature
granulomas were found in Listeria-infected rats that were also
exposed to ozone.
When mice are challenged with Listeria, mortality is the usual
end-point monitored; however, clearance and organ bacterial colony
counts can also be determined. L. monocytogenes is a Gram-positive
bacterium. The resistance of mice to the organism is genetically
regulated, and the susceptibility of the B6C3F1 strain, the strain
designated by the NTP for immunotoxicity studies, comes from the C3H
parent, since the C57Bl/6 mouse is resistant (Kongshavn et al., 1980).
Listeria can easily be stored at -70°C at a stock concentration of
approximately 108 colony forming units per ml. In studies in mice,
three challenge levels are routinely selected to produce 20, 50, and
80% mortality in the vehicle control animals. Mortality is recorded
daily for 14 days. Treatment groups consisting of 12 mice per group
have been found to be useful for obtaining statistically meaningful
data on host resistance. This assay is extremely reproducible when the
organism is administered intravenously.
The Listeria assay can detect both protection from and
increased susceptibility to chemicals and drugs. Morahan et al. (1979)
used the model to demonstrate a dose-dependent decrease in host
resistance after exposure to delta-9-tetrahydrocannabinol, the major
psychoactive constituent of marijuana. The Listeria host resistance
assay is that most often used in immunotoxicological assessment of
compounds, and numerous examples of its use can be found in the
literature. It is one of the primary models used by the NTP for
evaluating immunosuppression. Since Listeria is a human pathogen,
appropriate precautions are needed in conducting the assay.
4.2.10.2 Streptococcus infectivity models
Two species of Streptococcus have been used widely in bacterial
host resistance models for immunotoxicological assessment.
S. pneumoniae has been used primarily for evaluating systemic
immunity. S. zooepidemicus has also been used to evaluate systemic
immunity but is used extensively to evaluate the effects of drugs and
chemicals on the local immunity of the pulmonary system.
S. pneumoniae is a Gram-positive coccus to which host
resistance is multifaceted (Winkelstein, 1981). The first line of
defence against this organism is the complement system. Activation of
the complement system can result in direct lysis of certain strains of
S. pneumoniae; however, owing to the nature of their cell wall, some
strains are resistant to lysis by complement. Complement can still
participate directly in the removal of these bacteria as a result of
deposition of complement component C3 on their surface, which
facilitates phagocytosis by polymorphonuclear leukocytes and
macrophages. In the later stages of the infection, antigen-specific
antibody plays a major role in controlling the infection. Thus,
compounds that affect complement, polymorphonuclear leukocytes, B-cell
maturation and proliferation, or the production of antibody can be
evaluated in this system. S. pneumoniae is an excellent model for
evaluating immunotoxicity, since it elicits multiple immune components
which participate in host resistance, each of which can be a potential
target for an adverse effect of a xenobiotic. To date, this model has
had limited success in rats.
Preparation of S. pneumoniaefor challenge is slightly more
complicated than the procedures used for Listeria; however, the
potential of the model for detecting immunotoxic compounds makes the
additional steps worthwhile. Stock preparations of S. pneumoniae
(ATCC 6314) are easily maintained at -70°C in defibrinated rabbit
blood, and aliquots of the stock preparation can be removed and grown
in culture at various dilutions to obtain the desired challenge
concentration. An alternative approach is to grow the organism in
culture and to monitor the bacterial concentrations by measuring the
turbidity of the culture. A 5-µl aliquot of the stock preparation is
used to inoculate 50ml of brain-heart infusion broth, which is
incubated at 37°C, and the turbidity of the overnight culture is
determined with an Abbott Biochromatic analyser system or another
instrument that can sensitively measure changes in culture turbidity.
The overnight culture is diluted with fresh brain-heart infusion broth
to yield an absorbence difference of 0.020-0.025. The turbidity of the
subculture is monitored periodically, and when the optimal density
reaches an absorbence difference of 0.080, the subculture is rapidly
cooled in an ice bath and diluted to the desired inoculum level. The
turbidity of each inoculum is checked in the analyser, and adjustments
are made to obtain the preselected differences in absorbence.
Routinely, one day after the last exposure, female mice are challenged
intraperitoneally with 0.2ml of the S. pneumoniae inoculum. If the
inoculum is administered intravenously, extremely high challenge
levels must be used, which may reflect the efficiency of the
mononuclear phagocyte system to clear and kill the organism. Three
innocula, each at a different concentration, are prepared to give a
range of lethality (e.g. 20, 50, and 80%), and a sample of each is
serially diluted and placed on blood agar plates to determine the
number of colony-forming units administered to the animals. Owing to
the rapid onset of infection, mortality is recorded twice daily for
seven days. In studies by White et al. (1986), when female B6C3F1 mice
were exposed daily for 14 days to 1,2,3,6,7,8-hexachlorodibenzo-
para-dioxin or TCDD by gavage, they were found to have decreased
host resistance to S. pneumoniae, which is consistent with the
decrease in complement activity caused by these compounds.
Another species of Streptococcus that has been used as a host
resistance model is S. zooepidemicus, a group C streptococcus
(Fugmann et al., 1983). Exposure to N-nitrosodimethylamine was shown
to decrease host resistance to this strain significantly. Infection
with S. zooepidemicus may be dependent on an antibody-mediated
response, since the time to death after challenge is considerably
longer than with S. pneumoniae. Numerous studies have demonstrated
that aerosolized S. zooepidemicus is one of the most sensitive
indicators of the toxicity of air pollution: Mice exposed for short
periods to single or mixed pollutants before infection with an aerosol
of S. zooepidemicus and then assessed for mortality over 20 days had
increased mortality with increasing concentrations of ozone (Coffin &
Gardner, 1972; Ehrlich et al., 1977), nitrogen dioxide (Ehrlich &
Henry, 1968; Sherwood et al., 1981), sulfur dioxide (Selgrade et al.,
1989), metal particulates (Gardner et al., 1977; Adkins et al., 1979,
1980; Aranyi et al., 1985), phosgene (Selgrade et al., 1989), and
other volatile organic compounds (Aranyi et al., 1986). With many of
these compounds, enhanced susceptibility to infection has been
demonstrated at concentrations at or below the United States national
ambient air quality standards or threshold limit values. With all of
these compounds, enhanced mortality has been associated with failure
to clear bacteria from the lung and suppression of alveolar macrophage
phagocytic function. This model has recently been adapted to rats. In
this species, both ozone (Gilmour & Selgrade, 1993) and phosgene (Yang
et al., 1995) delayed clearance of bacteria from the lungs and
enhanced the inflammatory response (polymorphonuclear leukocytes in
lavage fluid) at concentrations that do not themselves produce
inflammation; however, mortality does not occur in this species. While
the bacteria have generally been administered as aerosols, Sherwood et
al. (1988) showed that similar results could be obtained when they
were administered intratracheally or intranasally. Since some strains
of Streptococcus are pathogenic to humans, appropriate precautions
must be taken when using this host resistance model.
4.2.10.3 Viral infection model with mouse and rat cytomegalovirus
Cytomegalovirus infections are widely distributed in humans, with
about 60-90% of the population infected. Human cytomegalovirus
infections occur in several forms, the most serious being congenital
and perinatal infection and infection of immunosuppressed individuals.
Less serious forms include post-perfusion syndrome and some cases of
infectious mononucleosis; however, the vast majority of postnatal
infections in immunocompetent individuals are clinically asymptomatic.
More severe disease may occur in immunodeficient hosts, such as
transplant patients (Naraqi et al., 1977; Pass et al., 1978; Marker et
al., 1981; Rubin et al., 1981). Primarily on the basis of
morphological considerations, cytomegaloviruses are classified as
members of the family Herpesviridae. Because these viruses have a
relatively protracted replication cycle, a slowly developing
cytopathology characterized by cytomegaly, and a relatively restricted
host range, they are grouped into the beta Herpesviridae subfamily
(Roizman et al., 1981; Roizman, 1982). The roles of several arms of
the immune system in resistance to cytomegalovirus have been studied
extensively in mice. The role of humoral immunity is not well
understood. It was suggested initially that neutralizing antibodies do
not play a pivotal role in recovery from cytomegalovirus infection in
mice (Osborn et al., 1968; Tonari & Minamishima, 1983); however, the
role of antibodies in neutralization of murine cytomegalovirus and in
antibody-dependent cell-mediated cytotoxicity is now recognized
(Manischewitz & Quinnan, 1980; Quinnan et al., 1980; Farrell &
Shellam, 1991). Cytomegalovirus-specific Tc cells can be detected in
cytomegalovirus-infected mice (Ho, 1980; Quinnan et al., 1980). NK
cell activity appeared to be the most effective, especially during the
initial stages of infection (Bancroft et al., 1981; Selgrade et al.,
1982; Bukowski et al., 1984). Enhanced susceptibility to infection has
been demonstrated in mice when macrophage function was blocked by
silica, and transfer of syngeneic adult macrophages to suckling mice
significantly increased their resistance to mouse cytomegalovirus
infection (Selgrade & Osborn, 1974). An inverse correlation is seen
between the virulence of mouse cytomegalovirus and its infectivity for
peritoneal macrophages (Inada & Mims, 1985), suggesting that
attenuated virus may be controlled, in part, by macrophages. Since the
rat virus acts very much like the attenuated mouse virus, macrophages
may be even more important in rats. Macrophages may facilitate the
generation of latent infection (Booss, 1980; Yamaguchi et al., 1988).
Enhanced susceptibility to mouse cytomegalovirus has been
demonstrated after treatment of mice with cyclophosphamide,
cyclosporin A, nickel chloride, or DMBA. Treatment with benzo[ a]-
pyrene or TCDD did not affect susceptibility to this infection
(Selgrade et al., 1982). Enhanced susceptibility was correlated with
chemical suppression of virus-augmented NK cell activity during the
first week of infection. In rats, exposure to immunotoxic agents such
as organotin compounds led to altered resistance to rat
cytomegalovirus (Garssen et al., 1995).
Experimentally, rodents can be inoculated intraperitoneally with
a species-specific cytomegalovirus, and the concentration of the virus
in tissue can be determined in a plaque-forming assay, which is a
modification of the method described by Bruggeman et al. (1983, 1985).
Rat embryo-cell monolayers are prepared in 24-well plates. Different
organs (salivary gland, lung, kidney, liver, spleen), obtained at
various times after infection, are homogenized in a tissue grinder and
stored as 10% weight/volume samples at -135°C until use. The confluent
monolayers are then infected with 10-fold serial dilutions of the
organ suspensions. After centrifugation, the suspension is removed,
and 1 or 0.6% agarose is added. After incubation at 37°C in 5% carbon
dioxide for seven days, the cells are fixed in 3.7% formaldehyde
solution, the agarose layer is removed, and the monolayer is stained
with 1% aqueous methylene blue. Plaques are counted under a
stereoscopic microscope.
In PVG rats, cytomegalovirus is detectable eight days after
infection, although the virus load is much higher on days 15-20. The
viral load in the salivary gland is higher than that in other organs,
i.e. spleen, lung, kidney, and liver. In contrast, in Lewis rats and
BN rats the viral load in e.g. the kidney was higher than that in the
salivary gland during the first week after infection (Bruning, 1985),
perhaps due to a strain difference (Bruggeman et al., 1983, 1985).
Total body irradiation of PVG rats with 60Co one day before
infection with cytomegalovirus increased the viral load in the
salivary gland, lung, kidney, spleen, and liver over that in
unirradiated PVG rats; histological analysis also indicated a higher
viral load in the salivary gland of infected rats. The mucosal
epithelium of the salivary gland contains enlarged cells with nuclear
inclusion bodies; these could be detected in irradiated, infected rats
only if the salivary gland was dissected and fixed 15 days after
infection. These results are in agreement with those of Bruggeman et
al. (1983), who found that gamma irradiation also induced higher viral
loads in the salivary gland of BN rats. Taken together these results
indicate a role for cellular immunity in resistance to this virus in
rats.
4.2.10.4 Influenza virus model
Influenza virus A2/Taiwan H2N2 has been used as a viral
challenge in evaluating alterations in host resistance of mice after
exposure to various compounds. Compounds that decrease host resistance
to the virus are N-nitrosodimethylamine (Thomas et al., 1985b) and
TCDD (House et al., 1990a). Compounds that do not to alter host
resistance to this pathogen include ozone (Selgrade et al., 1988),
benzo[ a]pyrene, benzo[ e]pyrene (Munson & White, 1990), methyl
isocyanate (Luster et al., 1986), and Pyrexol (House et al., 1990b).
Mortality is the end-point routinely measured in evaluating decreased
host resistance to influenza virus, which is usually instilled
intranasally (Fenters et al., 1979). Host resistance to this virus has
been reported to be mediated by cell-mediated immunity (Ada et al.,
1981), interferon (Hoshino et al., 1983), and antibody (Vireligier,
1975). This model had been suggested for use in evaluating compounds
that affect humoral immunity; however, its inability to detect such
compounds indicates that it is not suitable. A possible explanation
for the discrepancy is that administration of the virus by intranasal
instillation may invoke local immune mechanisms in the lung and may
not adequately reflect systemic immunocompetence. In several cases,
enhanced mortality has been demonstrated in the absence of effects on
viral titres in the lung (Selgrade et al., 1988; Burleson et al., in
press), indicating that enhanced mortality does not always reflect
effects on virus-specific immune defences.
Influenza virus has been used in evaluating immunotoxicity in
rats, after adaptation. Studies by Ehrlich & Burleson (1991) showed
that rats exposed to phosgene had significantly decreased host
resistance. TCDD was shown to affect the resistance of rats to the
adapted influenza virus RAIV (Yang et al., 1994).
As influenza virus is a human pathogen, appropriate precautions
must be taken.
4.2.10.5 Parasitic infection model with Trichinella spiralis
Resistance to infection with the helminth T. spiralis has been
evaluated in both mice and rats after exposure to a variety of
chemicals. In humans, as in other carnivores, infection occurs by
eating meat containing infectious larvae. The life cycle of the worm
is as follows: Infectious larvae excyst in the acid-pepsin environment
of the stomach, rapidly migrate to the jejunum, and penetrate host
intestinal epithelial cells. Sexually mature parasites are present
within three to four days after infection. The viviparous females
produce larvae that migrate via the lymphatic and blood vessels to
host muscle, where they encyst and are encapsulated within a host-
derived structure. Encapsulated muscle larvae can survive for years
within this structure.
An intense inflammatory response, comprised mainly of mast cells
and eosinophils, accompanies intracellular infection in the intestine.
T Cell-dependent immunity plays a crucial role in this inflammatory
response (Manson-Smith et al., 1979; Vos et al., 1983b; Wakelin,
1993), which is responsible for the expulsion of adult parasites.
Antibodies damage the reproductive structures of the female parasite
(Love et al., 1976), have a major role in the rapid elimination of
subsequent infections in rats (Appleton & McGregor, 1984), and
sensitize migrating newborn larvae for destruction by granulocytes
(Ruitenberg et al., 1983).
The number of encysted muscle larvae is typically much higher in
immunosuppressed animals than in immunocompetent animals, due perhaps
to delayed expulsion of adult worms from the intestine, decreased host
control of parasite fecundity, decreased destruction of migrating
larvae, or a combination of resistance defects. These end-points of
host resistance to T. spiralis infection and class-specific antibody
titres can be measured by standard techniques (Van Loveren et al.,
1994). Histological evaluation of the inflammatory infiltrate
surrounding encysted muscle larvae has also been described (Van
Loveren et al., 1993b).
It should be noted that direct effects of the chemical under
study can affect the outcome of infection. For example Bolas-Fernandez
et al. (1988) determined that cyclosporin A delays expulsion of adult
parasites from the intestines of rats but does not increase the number
of larvae encysted in host muscle. This was determined to be a direct
effect of cyclosporin A on the fecundity of female parasites rather
than on immunity to infection. Animals are infected by oral gavage
with known numbers of larvae, isolated from infected donor muscle.
Because infection is spread only by consumption of infected meat or
freshly isolated larvae, there is little danger of the infection
spreading to other animals housed in the same room. T. spiralis is a
human pathogen and must be handled as such; normal laboratory
practices are sufficient to prevent accidental infection.
T. spiralis infection has been used as a host resistance model
in both rats and mice. In general, chemicals that suppress T-cell
function suppress resistance to T. spiralis infection. Thus, TBTO
(Vos et al., 1990b), diethylstilbestrol (Luebke et al., 1984), TCDD
(Luebke et al., 1994, 1995), and the virustatic agent acyclovir
(Stahlmann et al., 1992) had deleterious effects on resistance.
4.2.10.6 Plasmodium model
Two strains of Plasmodium have been used to evaluate the
potential immunotoxicity of compounds. P. yoelii (17XNL) is a
nonlethal strain that produces a self-limiting parasitaemia in mice.
Resistance to this organism is multifaceted and includes specific
antibody, macrophage involvement, and T cell-mediated functions
(Luster et al., 1986). In this assay, animals are injected with 106
parasitized erythrocytes, and the degree of parasitaemia is monitored
over the course of the infection by taking blood samples. In control
animals, the peak response usually occurs 10-14 days after injection.
The degree of parasitaemia can be evaluated by a variety of methods,
e.g. manually, by counting parasitized erythrocytes in blood smears
(Luebke et al., 1991). Host resistance to P. yoelii has been used to
assess the immunotoxicity of benzidine (Luster et al., 1985b),
diphenylhydantoin (Tucker et al., 1985), TCDD (Tucker et al., 1986),
pyran copolymer (Krishna et al., 1989), gallium arsenide (Sikorski et
al., 1989), and 2'-deoxycoformycin (Luebke et al., 1991).
P. berghei is lethal to mice and certain strains of rats and
has been used in assessing immunotoxicity (Loose et al., 1978). Host
resistance depends on specific antibody production and ingestion and
destruction of antibody-coated Plasmodium by phagocytic cells such
as macrophages. T Lymphocytes may also be involved in host resistance
to the organism (Bradley & Morahan, 1982). Mortality has been
evaluated after injection of 106 Plasmodium-infected erythrocytes.
Compounds that have been evaluated for immunotoxicity in this model
system include 4,4'-thiobis(6- tert-butyl- meta-cresol) (Holsapple
et al., 1988), dietary fish-oil supplement (Blok et al., 1992), and
styrene (Dogra et al., 1992). Neither P. yoelii nor P. berghei is
infectious in humans; infection of animals can occur only through
parenteral injection of contaminated blood.
4.2.10.7 B16F10 Melanoma model
The B16F10 tumour cell line is a malignant melanoma that is
syngeneic with the C57Bl/6 mouse, which is one of the parents of the
B6C3F1 mouse. This tumour line was selected for its propensity to
metastasize to the lung. The assay is an outgrowth of the work of
Fidler and colleagues (Fidler, 1973; Fidler et al., 1978). NK cells
and macrophages have been proposed to be involved in host resistance
to this metastasizing tumour; however, T lymphocytes have also been
shown to play a role (Parhar & Lala, 1987). This host resistance assay
is referred to as an artificial metastasis model, since the tumour
cells are administered by intravenous injection, usually into the tail
vein, and lodge in the lung, which is the first capillary bed they
encounter. The B16F10 tumour cells can be stored frozen and can easily
be grown in culture before use. Routinely, 1-5 × 105 cells are
injected intravenously into sentinel mice (i.e. untreated mice
injected with the highest challenge level of tumour cells), and the
tumour burden is monitored in order to select the optimal day of
assay.
Two parameters are routinely used to assess tumour burden. One is
DNA synthesis in the lungs of mice bearing tumours. Since background
DNA synthesis in the lungs of mice without tumours is extremely low,
any detectable rate is a result of the presence of a tumour. In order
to measure synthesis, mice are pulsed intraperitoneally one day before
sacrifice with 0.2 ml of 10-6 mol/litre of 5-fluorodeoxyuridine,
followed 30 min later by 2 µCi of 125I-iododeoxyuridine administered
by the intravenous route. After sacrifice, the lungs are removed,
placed in Bouin's fixative solution, and counted with a gamma counter.
A second indicator of tumour burden is visual enumeration of tumour
nodules after fixation in Bouin's solution. The visibility of the
black nodules of the melanin-producing B16F10 tumour cells on the
yellow background of the fixed lung tissue allows enumeration of up to
200-250 nodules on the surface of the lungs. A good correlation has
been shown between number of tumour nodules and radioactivity present
in the lungs (White, 1992). Thus, if the tumour nodules become too
numerous to count, the results of the study can still be determined
from the radioassay. This system has been useful in demonstrating
decreased host resistance after systemic exposure to the tumour
promoter phorbol myristate acetate (Murray et al., 1985),
intratracheal exposure to gallium arsenide (Sikorski et al., 1989),
and exposure to nickel chloride (Smialowicz et al., 1985b); it has
also been used to show enhanced host resistance after exposure to
manganese chloride (Smialowicz et al., 1984) and 4,4'-thiobis(6- tert-
butyl- meta-cresol) (Holsapple et al., 1988).
4.2.10.8 PYB6 Carcinoma model
The PYB6 tumour cell line is a fibrosarcoma originally induced
with a polyoma virus in C57Bl/6 mice. Host resistance to the tumour
includes NK cell activity and T cell-mediated killing (Urban et al.,
1982). While PYB6 cells can easily be grown in culture, they should be
passed through an animal before use in challenge studies for
immunotoxicity (Luster et al., 1988). In studies with the PYB6 line,
mice are injected in the thigh with 1-5 × 103 viable tumour cells
and are then palpated weekly to detect the development of tumours at
the injection site. The end-points evaluated include the incidence of
tumours and time to tumour appearance; tumour size can also be
measured. This assay has been useful in detecting decreased host
resistance to many compounds, including Aroclor 1254 (Lubet et al.,
1986), DMBA (Dean et al., 1986), and benzene (Rosenthal & Snyder,
1987).
4.2.10.9 MADB106 Adenocarcinoma model
A tumour model used to evaluate host resistance in rats is the
MADB106 rat mammary adenocarcinoma, which is syngeneic with the
Fischer 344 rat. NK cells appear to play a major role in host defence
to this tumour (Barlozzari et al., 1985). In this model, survival time
after injection of the cells is the usual end-point monitored.
Compounds that decrease host resistance to the tumour can decrease
both the percentage survival and the survival time of treated animals.
Control rats begin to succumb to the adenocarcinoma two to three weeks
after an intravenous injection of 2 × 106 tumour cells. Smialowicz
et al. (1985b) showed a significant decrease in the survival of
animals treated with a single intramuscular dose of nickel chloride,
which was correlated with a decrease in NK cell activity.
4.2.11 Autoimmune models
Autoimmune models can also be used to investigate whether a
compound exacerbates induced or genetically predisposed auto-immunity.
These models are used mainly to elucidate the pathogenesis of
autoimmunity and the effect of immunosuppression in
immunopharmacology. Few studies have been reported, although the
relevance of the model for extrapolation to humans may be good. A
number of autoimmune models are available in rats and mice. Autoimmune
phenomena can be either induced or occur spontaneously. In induced
models, an autoantigen is isolated from a target organ obtained from
another species (generally bovine), and the animal is immunized with
this purified antigen in adjuvant. Examples in the rat (Calder &
Lightman, 1992) are experimental encephalomyelitis elicited by bovine
spinal cord antigen (Stanley & Pender, 1991), experimental uveitis
elicited by bovine retinal S-antigen (Fox et al., 1987), and adjuvant
arthritis elicited by Mycobacterium containing H37RA adjuvant
(adjuvant arthritis) or collagen (Holmdahl et al., 1990; Klareskog &
Olsson, 1990; Wooley, 1991). Autoimmune phenomena and associated organ
pathology normally emerge in almost all immunized animals within two
to three weeks. Depending on the effector reaction and the
reversibility of the damage, the disease either stops when the damage
is complete (e.g. uveitis, resulting in blindness of the animal) or
the autoimmune reaction is transient and animals recover (adjuvant
arthritis). In some models, animals subsequently experience a relapse
around day 30 (experimental encephalomyelitis). The induction and
development of autoimmunity in these animals are mediated by T cells
that show the cytokine expression pattern (IL-2, INF gamma) of the Th1
subset. The effector phase of disease symptoms is also mainly a T
cell-mediated process, to which CD4+ cells, CD8+ cells, and
macrophages contribute. Lewis (RT11) rats are particularly
susceptible. These autoimmune models can be induced in other species,
such as mice (Baker et al., 1990) and rhesus monkeys (Rose et al.,
1991). They are generally accepted as models of human (organ-specific)
autoimmune diseases, e.g. experimental allergic encephalomyelitis as a
model of multiple sclerosis, experimental allergic uveitis as a model
of idiopathic posterior uveitis, and adjuvant arthritis as a model of
rheumatoid arthritis.
Autoimmunity can also be induced by metals (Druet et al., 1989;
Bigazzi, 1992). A well-known example is glomerulopathy induced by
mercuric chloride in BN (RT1n) rats. The process is initiated by T
cells with a cytokine synthesis pattern (IL-4) of the Th2 subset. The
relative incidence of these cells is much higher in BB rats than in
other strains. After the cells of the Th2 subset have been stimulated,
there is polyclonal stimulation of B lymphocytes, leading to synthesis
of antibodies (including pathogenic antibodies) to the glomerular
basement membrane. These antibodies subsequently mediate autoimmune
destruction of renal glomeruli. This model is the best studied model
of 'drug'-induced autoimmunity. Mercuric chloride elicits
glomerulonephritis in other rat strains, in which glomerular
destruction is not due to anti-glomerular autoantibodies but is
mediated by immune complexes deposited in the glomerulus (Druet et
al., 1989).
In spontaneous models, predisposition to the development of
autoimmune phenomena and disease is determined by the genetic
composition of the animal strain. Well-known examples are BB rats
(Like et al., 1982; Guberski, 1994) and NOD mice (Lampeter et al.,
1989; Leiter, 1993), which develop autoimmune pancreatitis and
subsequently diabetes. Within the pancreas, the islets of Langerhans
are infiltrated by T lymphocytes and macrophages; subsequent
destruction of the islets results in diabetes. These spontaneous
models are considered to be animal models of human diabetes (Dotta &
Eisenbarth, 1989; Lampeter et al., 1989; Riley, 1989). Other examples
are the systemic autoimmunity that emerges in certain mouse strains
(Guttierez-Ramos et al., 1990) like (NZB × NZW)F1 mice (Theofilopoulos
& Dixon, 1985) and the mixed lymphocyte reaction in lpr (Matsuzawa
et al., 1990) and gld mice (Roths et al., 1984). The spontaneous
pathology in these animals resembles various disease manifestations in
human systemic lupus erythematosus and is similarly mediated by immune
complexes that deposit in tissue. In (NZB × NZW)F1 mice, mainly lupus
nephritis is induced by immune complexes; in the mixed lymphocyte
reaction in lpr mice, both joint manifestations and
glomerulonephritis are seen. The genes associated with autoimmune and
immune complex disease are not known. In comparison with induced
models, these models have the advantage of spontaneous, gradual
development of autoimmune disease symptoms; however, this is a
disadvantage in experimental design, as not all animals develop
disease, and emerging disease develops at different times or ages.
In general, treatment of animals with immunosuppressive drugs
that interfere with signal transduction (cyclosporin, FK-506,
rapamycin) or cell proliferation (cytostatics like azathioprine,
mizoribine, and brequinar), and anti-inflammatory agents
(corticosteroids) inhibit the development of symptoms in these models.
Exposure to immunotoxic chemicals may also lead to alterations in the
course of disease emergence. For instance, HCB, which leads to
immunoenhancement in rats, markedly enhanced the severity of allergic
encephalomyelitis (Van Loveren at al., 1990c). In contrast, arthritic
lesions were strongly suppressed in HCB-exposed Lewis rats, indicating
that HCB has biologically significant immunotoxic effects. Although
the contrasting effects in the two autoimmune models are not yet
understood, and clear dose-effect relationships have yet to be
established, this type of information should be obtained for risk
assessment.
4.3 Assessment of immunotoxicity in non-rodent species
While most immunotoxicological evaluations are conducted in mice
or rats, use of other species is increasing.
4.3.1 Non-human primates
Various non-human primates, including Macaca mulatta (rhesus
macaques), M. nemestrina (pig-tailed macaques), Cercocebus atys
(sooty mangabeys), M. fasicularis (cynomolgus monkeys), and
marmosets have been used in immunotoxicological studies. Many of the
assays carried out in mice or rats can be adapted for use with non-
human primates. Strategies and methods used in studies of humans have
also been introduced in studies on non-human primates. Monoclonal
antibodies generated to human leukocyte subsets can be used in
phenotyping blood mononuclear cells of e.g. marmoset monkeys
(Callithrix jacchus) (Neubert et al., 1990, 1991), although the
possibility of such use differs, depending on the evolutionary
distance of the non-human primate from humans. While most of the
assays conducted in non-human primates involve serum or peripheral
blood, some assays, such as those used to measure delayed-type
hypersensitivity, are holistic, in that the animals are sensitized
in vivo and then evaluated in vivo at the challenge site (Bugelski
et al., 1990; Bleavins & Alvey, 1991).
The effects of chronic exposure to the PCB Arochlor 1254 on the
immune response of rhesus monkeys have been evaluated by Tryphonas et
al. (1989). In these studies, the lymphocyte response to concanavalin
A and phytohaemagglutinin was evaluated, as were total serum
immunoglobulin levels, antibodies to sheep red blood cells, and
numbers of T and B cells in peripheral blood. In later studies
(Tryphonas et al., 1991a,b), one-way mixed lymphocyte cultures,
antibodies to pneumococcal antigens, phagocytic mononuclear cell
function, NK function, haemolytic complement activity, and production
of IL-1, tumour necrosis factor, thymosins, and interferon were
evaluated in monkeys exposed to Arocolor 1254. Ahmed-Ansari et al.
(1989) evaluated phenotypic markers and function in three species of
non-human primate. The functional assays included NK cell activity,
lymphocyte transformation, and antigen presentation. Extensive studies
included the evaluation of more than 20 phenotypic markers or
combinations of markers for each of the three monkey species. Use of
the monkey as a test species is likely to increase as more and more
biotechnology and recombinant products are produced.
4.3.2 Dogs
While dogs are not the species of choice for immunotoxicological
studies, they are used predominantly in assessing toxicological
safety, and virtually all of the assays used for assessing immunotoxic
potential have been adapted for use in dogs. These include evaluation
of basal levels of IgA, IgG, and IgM (Glickman et al., 1988),
allergen-specific serum IgE (Kleinbeck et al., 1989), mononuclear
phagocyte function (Thiem et al., 1988), NK cell activity (Raskin et
al., 1989), Tc cell activity (Holmes et al., 1989), and mitogen-and
cell-mediated immune responses (Nimmo Wilkie et al., 1992).
4.3.3 Non-mammalian species
Non-mammalian species are also used extensively for evaluating
the potential adverse effects of compounds and agents on the immune
system.
4.3.3.1 Fish
Because of their environment, fish are an excellent model for
studying the effects of water-and sediment-borne pollutants. There are
several other good reasons for studying immunotoxicity in fish: many
of their diseases are related to environmental quality, various
environmental pollutants have immunotoxic potential, and many of the
diseases have an immune component. Moreover, there is concern about
the health status of aquatic ecosystems in relation to pollution, and
fish will be useful target species for developing biomarkers (see
box). Fish are easy to obtain, there is an extensive body of
knowledge, and their economic interest (aquaculture) facilitates the
finding of research resources. At present, immunotoxicology in fish is
not as sophisticated as that in mammals. Screening and functional
tests are being developed in the laboratory but cannot yet be applied
in the field.
A wide range of species is used for field and laboratory studies.
The choice of species depends on its biology (migratory or local,
marine or freshwater; sediment-dwelling or pelagic) and on experience
in the laboratory. A lack of consistency, e.g. becuse of a limited
number of species (as in mammalian immunotoxicology) makes this field
of research diffuse and extended and may result in limited progress;
nevertheless, a variable or consistent effect over a variety of
species is certainly a valuable observation. Some species seem to be
preferred, such as trout, salmon, and carp, which are practical, owing
to their size, for sampling blood and tissues for laboratory studies.
Smaller species, such as guppies (Poecilia reticulata) and medaka
(Oryzias latipes), have secured a niche in aquatic toxicology owing
to the ease of husbandry and relatively low cost; moreover, because of
their small size, whole animals can be used for histopathological
examination (Wester & Canton, 1991), but their application in
immunotoxicology may be limited because of difficulty in obtaining
adequate blood and tissue samples. For studies of saltwater species,
bottom-dwelling flatfish are commonly used in field studies and, to a
certain extent, in studies of mesocosms and in the laboratory. In
Europe, the flounder (Platichthys flesus) and dab (Limanda limanda)
are popular target species since they are susceptible to certain
recognizable diseases and are commonly available.
A compehensive variety of parameters is listed by Anderson (1990)
and Weeks et al. (1992). A modified set based on those lists and
assays used in rodent immunotoxicology is presented in the box above
and classified as tier 1 (screening tests) and tier 2 (functional
assays). Parameters commonly mentioned in the literature are discussed
below.
Blood cell counts and differential counts: As leukocytes play a
major role in specific and nonspecific humoral and cellular immune
responses, this parameter is used as a measure of the status of the
defence system, in particular in tier 1 testing. It is relatively easy
to test blood samples drawn from live animals, but many environmental
factors unrelated to defence may modify leukocyte status (Anderson,
1990). The use of monoclonal antibodies directed against individual
cell types may improve their identification (Bly et al., 1990; Van
Diepen et al., 1991). Another possible parameter is the haematocrit;
however, it has no known specificity for any immune function, although
it may be considered as a general indicator of stress.
Candidate biomarkers for immunotoxicity in fish
Tier 1: Screening tests
* Conventional haematology, including total and differential blood
cell counts, surface markers (flow cytometry), and macrophage
density and morphology: easy, nonspecific
* Serum immunoglobulin concentrations in naive (unstimulated) fish:
easy, limited specificity
* Lymphoid organ weight (mainly spleen, occasionally thymus):
impracticable
* Histopathology of the thymus, spleen, and kidney: possible, can
be specific
Tier 2: Functional assays
* Humoral immune response (agglutination, enzyme-linked
immunosorbent assay): possible, can be specific
* Cellular immune response (allograft rejection in scale, skin, or
eye): possible, can be specific
* Macrophage functions (phagocytosis, bacterial killing, migration,
chemiluminescence): limited specificity
* Host resistance (bacterial infections): possible, can be
specific, relevant
Nonspecific defence: Other indicators of nonspecific defence
have been proposed as indicators of immunological stress. These
include acute-phase proteins (Fletcher, 1986), the levels of which
appear to