United States
Environmental Protection
Agency
EPA-600/8-83-002
March 1983
Research and Development
Short-Term Tests for
Environmentally Induced
Chronic Health Effects
-------
EPA-600/ 8-83-002
March 1983
Short-Term Tests for
Environmentally Induced Chronic
Health Effects
Avril D. Woodhead
Editor
Brookhaven National Laboratory
Upton, New York 11973
Michael D. Waters
Technical Editor
Health Effects Research Laboratory
Research Triangle Park, North Carolina 27711
Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
DISCLAIMER
Mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
FOREWORD
The formation of the U.S. Environmental Protection Agency in 1970 ushered in
the first decade of environmental awareness as a total national phenomenon. It was a
decade punctuated by major Congressional mandates to restore the nation's waters,
to reduce air pollution, and to find a comprehensive approach to other environ-
mental problems—those associated with pesticide use, hazardous waste disposal and
toxic substances. It was a decade underscored by the demand for new technology and
better science to answer environmental questions and to solve environmental
problems.
As the scientific and technical arm of the Agency, The Office of Research and
Development is responsible for advancing the state of knowledge about the environ-
ment such that critical issues and questions can be addressed and answered
effectively, based on the application of state-of-the-art science and technology. In the
years since 1970, The Office of Research and Development has produced manifold
increases in the data base from which environmental decisions are made and in the
sophistication of the understanding which has provided the basis for decisions.
This volume represents our effort to take stock of scientific advances in the use of
short-term tests for environmentally induced chronic health effects since the in-
ception of the Agency and to gauge what progress has been made and what remains
to be accomplished. The essays in this volume present a range of perspectives on the
subject, from the vantage points of the scientific and technical disciplines which have
been carrying out relevant research. The points of view represented are varied and
sometimes conflicting. But scientific progress depends on just such diversity. The
authors at times have speculated about emerging problems and research needs. Such
attempts require extrapolation based upon informed scientific judgment. The
outcome of that process must, in the final analysis, be recognized as opinion and not
fact.
-------
EDITOR'S NOTE
This book, "Short-Term Tests for Environmentally Induced Chronic Health Ef-
fects," is a part of the Environmental Protection Agency, Office of Research and
Development's "Decade Document" Program. The main objective was to provide
an assessment of the potential of short-term bioassays as tools in predicting chronic
health effects resulting from exposure to environmental agents. It was also intended
to guide future developments in these test systems, particularly in areas relevant to
the needs of the EPA.
With these guidelines in hand, I approached my distinguished authors for their
contribution. Their response was mixed. All were enthusiastic about the prospect of
writing for the book, but several pointed out that there were so few short-term tests
available in their fields that it would be difficult to write a full chapter. I assured them
that the value of the document would lie not only in highlighting the spectacular
development of short-term tests in some areas over the past ten years, but equally in
drawing attention to the many large gaps in our knowledge, and to areas where there
is a great need for the validation of short-term assays. The chapters are therefore
somewhat diverse in their content. In fields in which short-term tests exist, the
authors have critically reviewed their status, they have discussed how the tests might
be more precisely defined and have pointed the way to future programs of research.
In areas in which there are very few short-term tests, the contributors have provided
excellent evaluations of the problems that hamper development of such tests; this
may encourage others to investigate any possible biological screening methods.
Each chapter was reviewed by 3-4 reviewers drawn from EPA staff and other experts.
I am grateful to the authors for their forebearance, and for their readiness to accede
to my suggestions for making deletions and for modifying their chapters to ensure
some continuity of format. In several cases they allowed me to take out sections that
were near and dear to their hearts.
I want to pay special tribute to Michael Waters, the EPA Project Officer on this
volume, who has cheerfully and wholeheartedly devoted a great deal of time and
effort to its production. He was instrumental in developing close and successful
working relationships between myself, the EPA staff, the reviewers, and the authors.
Richard Setlow, Chairman of the Biology Department at Brookhaven National
Laboratory, and Alexander Hollaender, A.U.I. Washington, have given me much
sound advice, active aid, and encouragement at all stages during preparation.
I also owe particular thanks to the volume editor, Dr. Jessica Barren, who did a mas-
terful job guiding the book through the later stages of preparation and production.
Finally, my especial thanks to two members of the Biology Department, Virginia
Pond and Kathy Kissel, for their hard work and outstanding support. Virginia Pond
did a tremendous task of carefully and painstakingly editing the manuscript and
preparing the index. Kathy Kissel typed, and retyped, the manuscript with speed,
accuracy and unfailing good humor.
A.D. Woodhead
-------
PREFACE
Roger Cortesi
INTRODUCTION
The purpose of this preface is to give the reader a general understanding of the
regulatory process as it pertains to health issues, and to discuss the role that health
data play in that process. I hope to make clear at which stage data from short-term
tests are considered and to put into perspective the benefits as well as the
shortcomings of their use.
WHAT HEALTH DATA CAN PROVIDE
The purpose of health-effects data in the regulatory process is to determine the
levels of risk that are associated with exposure to different concentrations of a pollu-
tant. This requires a knowledge of the dose-response curve and the level and extent
of individual and population exposures. The outputs of a risk assessment can be
manifold, but two have been shown to be generally useful, namely: (1) an estimate
of the increased probability of disease or death of an individual as a function of
increased exposure; (2) an estimate of the morbidity or mortality in a population for
any given exposure. Ideally, what would be provided by health data is a dose-re-
sponse curve at levels of concern. For the Environmental Protection Agency (EPA),
dose-response curves are needed at low levels of exposure (indeed, by toxicologic
standards, very low levels) and for chronic as well as for acute exposures.
This situation, while simple and understandable in theory, is very much more
complicated in practice. The derivation of a precise dose-reponse curve at low levels
requires a clear definition of the health endpoint, the ability to extrapolate the data
and the obtaining of sufficient data.
Health Endpoint Uncertainty
Uncertainties in the health endpoint arise from questions as to what sector of the
population is to be protected, and from uncertainties and variability in the clinical
manifestation of the disease.
Determination of the part of the population at risk from exposure to a particular
pollutant is far from straightforward, as consideration must range from its effects
upon the germ cells to those upon the aged. Obviously, the critical response to a pol-
lutant can vary with the target population. Symptoms of chronic disease can differ
markedly from person to person, making evaluation of the response far from precise.
Also, the assessment of the dose that an individual has received is difficult, since ex-
posure estimates in retrospective epidemiological studies are usually not much better
than an order of magnitude.
The Author: Roger Cortesi is Acting Director of the Office of Health Research, United States Environmen-
tal Protection Agency, Washington, D.C.
-------
Extrapolation of Data
' For the EPA's purpose, extrapolation of data is virtually always needed in two
areas. The first is extrapolation from relatively high doses where direct cause-effect
relations can be observed to the much lower doses typical of environmental
exposures. The second is from the test system or animal to man. This is, of course,
the well-known "mouse to man" problem. In short-term testing we would like to
make even more extreme projections, from cell to man or from biochemical marker
to man. Extrapolation is an inaccurate process, even when the kinetics of the rela-
tionship are known. When the functional form of the relationship is unknown, as in
much of the dose-response data, the estimation of the dose-response curve at
environmental levels in humans from data obtained at higher doses or in nonhuman
systems is an educated guess at best.
Generation of Data
It is very difficult in many cases to obtain data even if one is willing to accept the
obstacles of endpoint definition and extrapolations. The problem arises even for the
effects of acute exposures and is worse for chronic exposures. Short-term high lev-
els of air pollution can be clinically tested on healthy subjects but not on the very
young nor on the elderly or infirm, where the most serious effects are to be
anticipated. Clinical testing for chronic exposures is not permissible, and therefore
data for the dose-reponse curve must be extracted from epidemiological studies or
from nonhuman test systems.
Epidemiology has successfully highlighted causal relationships between particular
enironmental pollutants and human cancer, but when it comes to estimating dose-
response relationships it has serious and fundamental drawbacks. The major diffi-
culty is that the method can rarely detect reliably a 50 percent increase in risk even in
a large population because the intrinsic biases of epidemiological studies, in them-
selves, are about this value. Yet a relative risk of this level or smaller is important
when the effect is common, such as cardiovascular disease or kidney failure.
The problems of epidemiology, particularly with respect to chronic effects, are
many. They include the insidious nature of the diseases; the determination of expo-
sure levels; the control of confounding variables; and the large number of people
needed to detect a minor effect. These problems and others are treated extensively in
the literature, so they need not be expanded upon here. They make the derivation of
a dose-response relationship possible only in a general and imprecise sense at best. In
most cases, it is extremely fortunate if a robust association can be made between a
pollutant and a chronic disease.
With epidemiology of limited use and clinical studies for chronic effects not allow-
able, we have to try to get the most we can from derived data from nonhuman test
systems. Besides the problems of extrapolation to humans and from high to low ex-
posure, which I have discussed above, there are a number of other factors making it
difficult to obtain a dose-response curve. For example, the appropriate animal
models may not be available and, when they are, then a lifetime exposure is often
needed. In many cases, the exposure regime is not easy to define; for instance, should
it be a series of spiked exposures or a continuous low-level one? Finally, often it is
impossible to reproduce the ambient conditions of the pollutant.
First, lifetime tests on animals are time-consuming and costly procedures. More-
over, the cumbersomeness and expense of such tests militate against their use when
widespread testing is desirable, such as trying to find, by use of effects testing, the op-
timal operating conditions of an engine so that the exhaust is least toxic. Lifetime
assays on animals do not readily lend themselves to routine testing because quality
assurance is difficult over a two- or three-year period or more.
VI
-------
DISCUSSION
The purpose of what has been said above is not to make the task of health research
for standard-setting* appear hopeless. The purpose is to point out that the concept of
a "scientific" assessment of the health risks imposed by exposure to pollutants at am-
bient levels is generally not possible because we cannot produce dose-response data.
This being the case, one might think that trying to improve the situation by turning
to short-term tests is misdirection. We are hopeful, however, that short-term tests
will be able to produce data that are quantitatively useful in the regulatory process.
We believe that short-term tests must be pursued for several reasons.
Further, in many regulatory situations, if health data are to be used at all, they
must be available quickly. These objections are overcome by the use of short-term
assays.
We think that in the process of developing and validating short-term tests,
mechanisms of toxicity will become much better defined, thereby greatly improving
the precision of the dose-response curves and hence the risk assessments. Similarly,
short-term tests may be developed that can be used in large-scale surveys of human
populations, such as rapid and accurate measures of enzyme activity or of the mor-
phology and motility of spermatozoa, thereby again making epidemiology a more
powerful and useful source of data.
THE REGULATORY PROCESS
I have discussed above why the scientific component for setting exposure stan-
dards is uncertain. In this part, I will describe the regulatory process itself and try to
extract some key factors in deciding upon health-based exposure levels.
It is important to understand that the standard-setting process is profoundly polit-
ical (in the nonpejorative sense). The question of where to set a limit cannot be con-
sidered separately from such issues as what it costs, who benefits, and who pays.
(The fact that it cannot be considered apart from these factors does not, of course,
mean that one can always accurately define such things.) This consideration implies,
among other things, that a health standard can be set at two different levels in two
different places without either standard being "wrong." Not only is the standard-
setting process not entirely a scientific one, it is also not entirely an economic one
either, in that one cannot balance cost versus benefit in any rigorous way. This is, of
course, to be expected from the imprecision of the health data, but it is driven further
in this direction by the imprecision of the economic data as well. We also have to take
into consideration major differences in philosophy, as to what is "right" and "rea-
sonable," and what will be the consequences of failure to pursue a version of right
and reasonable.
I will now describe a way of looking at the standard setting and regulatory process
that I believe to be instructive even though many will profoundly disagree. I believe
there is enough truth in what I describe to provide insight into why the process is
indeed political.
Obviously, it is very difficult for regulations to be set less stringently than the regu-
latees are essentially willing to accept. This means there is a tendency for regulations
to be as stringent as possible considering the political pressure that increasing strin-
gency generates. That the process should proceed in this way can be repugnant to
many with a strong belief in order and method, be that belief scientific, economic, or
philosophic. The situation is not, however, as purely political as it appears. Scientif-
ic, economic, and philosophic considerations are certainly taken into account, but
such data are not used mechanically to regulate. If, for instance, a substance is
"Standard is not used in its physical or chemical sense, that is, as an acknowledged measure for quantitative
or qualitative value. The standard may be set at different levels, as I will describe further in this section.
Vll
-------
known to be very harmful and to cause especially damaging effects, regulations can
be very stringent before political pressure stops the process. Conversely, if the threat
is accompanied by widely perceived benefits, the political pressure will become un-
bearable with muchless stringency. A clear example of this is the loss of life weallow
from such things as automobiles, cigarettes, and alcohol. Alternatively, a very strin-
gent standard that adds only a small percentage to the unit cost may be tolerable,
even though the overall added cost greatly exceeds any perceived benefits. While
such a procedure does, in some sense, lack intellectual tidiness, it effectively copes
with the competing interests of the most seriously concerned parties.
One more important inference can be made from the regulatory process described
above. Namely, some kind of assessment of risk is always made, whether it is
admitted or not. This assessment must be done consciously or unconsciously, openly
or covertly, individually or collectively, in deciding just how tight a standard to set.
This procedure, as you might expect, does not produce consistent results in such
narrow terms as incremental lifetime risk, or in terms of dollars per life saved. The
level of a pollutant that will be tolerated is very much influenced by what can be done
about it. If the benefits from the production of a pollutant all lie in the future, strict
standards are more likely to be set. When the benefits are large and immediate, less
stringent standards are likely. The range of increased lifetime risk to an individual
from various standard-setting procedures has extended from more than one part in a
hundred down to less than one part in a million.
Setting exposure limits is a legal process and not a scientific one. Science can only
specify what the limit should reflect, and only then to the extent that a consensus ex-
ists in the scientific community. When there is no agreement in the interpretation of
the scientific data, then the regulatory agency must decide what to do in the face of
this scientific uncertainty. In order to do this in an acceptable way, the decision has
to be made according to precepts of fairness embodied in the law. These precepts bas-
ically define what information must be considered, and the steps to be taken in doing
so. This means, in essence, that every submission impacting upon the standard-set-
ting process is made public, so that people who disagree are free to attempt to dis-
credit them. Submissions such as scientific papers or data which have major implica-
tions for setting a limit are subjected to a great deal of scrutiny, usually under the
direction of lawyers. If a paper has implications that could cost a lot, it is going to be
reviewed line by line.
The significance of all this for short-term tests is as follows: short-term tests need
a high probability of identifying a potentially harmful effect (that is, a high detection
probability or, equivalently, a low fraction of false negatives) and a low probability
of falsely detecting a harmful effect (that is, a low false-alarm probability or, equiva-
lently, a low fraction of false positives).* Finally, they should yield results that can be
expressed quantitatively, giving a dose-response curve that is applicable to humans.
The certainty with which one must be able to assert that a test meets these three cri-
teria is not clear at present. Recent vigorous challenges in court of assertions of the
carcinogenicity of a substance based upon cancers or tumors induced in laboratory
animals indicate that short-term tests will have to meet the above conditions at least
as well as currently used long-term animal assays.
PROBLEMS ASSOCIATED WITH DATA RELEASE
Situations occur where the specter of health damage to people and their children
arises; Love Canal and Three M ile Island are just two examples. When this happens,
there is a great desire to know, virtually day by day, the results of any tests that would
*lt is worth observing in passing that high detection and low false-alarm probabilities are not independently
achieved. A high detection probability can be achieved at the cost of a high false-alarm probability and vice
versa. For example, a test which calls all substances toxicants will have a detection probability of one, but
also have unacceplably high false-alarm probability.
-------
be useful for taking vigorous or drastic remedial actions. In fact, if there are relevant
tests in the processes of development, there will be very strong pressure to try the
tests before they have been properly validated. In such a situation, there are several
specific problems. First, there is pressure, often successful, to release the results of
tests before they have met adequate peer review. Second, the caveats accom-
panying the results are often ignored. And third, there is a marked propensity for
fallacious reasoning—a post hoc ergo propter hoc—thinking by both public and
press, resulting in conclusions being reached about cause and effect which may be
completely wrong.
Such an atmosphere is clearly not one which fosters objective and disinterested as-
sessment. In such an atmosphere, statements about what the results "mean" will
have to be made sooner and with less review and discussion than is normal for the
scientific process. Any delay in announcing results causes great problems in terms of
public demand that something be done in that delay itself is often perceived as bu-
reaucratic "cover-up."
The successful development and application of short-term tests will certainly face
this dilemma. I believe that the conflict is an intrinsic one and hence can only be mol-
lified, but not resolved. The question of how to mollify this conflict is not one we
treat here.
-------
CONTENTS
Page
Foreword iii
Editor's Note iv
Preface v
Figures xii
Tables xvi
Quantitative Assessment of Risks, Richard B. Setlow 1
Short-Term Tests for Genetic Toxicity, Herman E. Brockman
and Frederick J. de Serres 11
Short-Term In Vitro Tests for Identifying Carcinogens:
Transformation of Mammalian Cells in Culture, Arthur C. Upton 38
Short-Term Tests for Identifying Teratogens, Thomas H. Shepard 69
Short-Term Tests for Renal Toxicity, David S. Miller and Anthony
M. Guarino 89
Early Warning Systems for Behavioral Toxicology, Bernard Weiss 108
The Role of the Environment in Chronic Cardiovascular Disease,
C. Richard Minick and Carl G. Becker 136
Frontiers in Short-Term Testing of Pneumotoxicants, D.B. Menzel,
E.D. Smolko, D.E. Gardner, and J.A. Graham 175
Index 211
XI
-------
FIGURES
Number Page
Setlow
1 - Linear (L), linear quadratic (LQ), and quadratic (Q) dose-response
curves for mortality from leukemia or bone cancer induced by
gamma radiation 3
2 - Two ethylation products of deoxyguanosine in DNA 6
3 - Two possible relations between the dose of an environmental
agent and the relative level of specific DNA adducts formed in
cells 6
4 - Hypothetical dose-response curves relating a particular biological
effect with concentration of specific DNA adducts 8
Brockman and de Serres
1 - The long latent period of cancer illustrated by the association
between cigarette smoking and lung cancer 17
2 - DNA damage test using the bacterium Escherichia coli 24
3 - The mouse spot test for detection of a forward mutation at
one of the heterozygous loci determining coat color 25
4 - The detection of chromosomal aberrations by the formation of
micronuclei 26
5 - Sister chromatid exchange (SCE) test in human lymphocytes
in culture 27
6 - The use pf the higher plant Tradescantia as an in situ
environmental monitor 33
Upton
\ - Neoplastic transformation of mouse cells in culture 40
2 - Outline of procedure of standard colony assay for in vitro
transformation of Golden Syrian hamster embryo cells 42
3 - Schematic illustration of the microscopic appearance of
transformed fibroblasts as compared with normal fibroblasts 43
4 - Outline of procedure of mass culture assay for in vitro
transformation of Rauscher leukemia virus-infected Fischer
rat embryo cells • 47
5 - Variation in the probability that a compound will be correctly
identified as a carcinogen (true positive result) by the combined use
of the in vitro transformation test and the Salmonella mutagenicity
test, in relation to the percentage of carcinogens among the
samples being tested 51
-------
Number Page
6 - Variation in the probability of true positive and false positive
results provided by the combined use of discordant results, from
the in vitro transformation test and the Salmonella mutagenicity
test, in relation to the percentage of carcinogens among the
samples being tested 52
Shepard
1 - Curve illustrating the susceptibility to teratogenesis from
fertilization throughout intrauterine development 72
2 - Diagram of the factors that influence embryonic dose of a
foreign chemical present in the maternal blood stream 72
3 - Diagram showing that each drug tested has a dose spectrum
ranging from no effect to one of maternal lethality 73
4 - Diagram giving an example of an in vitro test of a teratogen,
using a growing rat embryo 76
5 - Diagram showing the relationship between cause, mechanism
and manifestation of a teratogenic agent 77
6 - Perspective of our major defenses against teratogenic agents 79
7 - Diagram of a scheme whereby three data bases (individual,
agents and clinical syndromes) might be used to determine the
mechanisms which cause congenital defects and their
prevention 82
8 - Events between the introduction of a new agent causing
widespread disease and the ending of the epidemic 83
Miller and Guarino
1 - Diagrammatic representation of the relationships between the
component parts of the nephron and associated blood
vessels 91
2 - Simplified diagram of the glomerular filtration apparatus 92
3 - Simplified summary of the movement of water and solutes in
the different regions of a model nephron 93
4 - Concentration of total solutes in tubular fluid as a function of
segment of the nephron 95
Weiss
\ - Procedure for plotting visual fields in monkeys 113
2 - The response of a patient with Minamata Disease and a
normal control 115
3 Mice being tested on a rotarod 116
4 A system designed to measure fine motor control 117
5 Conduction velocities, for two worker groups, of the slower
motor fibers (CVSF) of the ulnar nerve 119
6 Scheme for testing copulatory performance in male rats 120
-------
Number Pa8e
7 - System for evaluating the effect of microwave exposure on
thermo-regulatory behavior ................................. '"
8 - Irritant potency of airborne substances is measured
behaviorally with a response that terminates irritant flow ....... 124
Minick and Becker
1 - Blood vessels in the muscularis of the gallbladder of a dog
injected intravenously with ellagic acid ....................... 147
2 Section of lung treated in a similar fashion to that
illustrated in Fig. 1 ........................................ 147
3 Atherosclerosis of the right coronary artery of a 35-year-old
man with marked coronary artherosclerosis, old and recent
myocardial infarcts, and aortic and mitral stenosis ............. 149
4 - Atherosclerosis of the right coronary artery of rabbit that
received semisynthetic lipid-rich diet and seven injections of
horse serum over a period of nine months .................... 149
5 - Atherosclerosis of the left coronary artery of a 72-year-old
man who died with disseminated rheumatoid disease ........... 150
6 Atherosclerosis of mesenteric artery of rabbit that received
semisynthetic, lipid-rich diet and six intravenous injections of
bovine serum albumin over a period of one year ............... 150
7 Lumen of rabbit mesenteric artery is narrowed by musculoelastic
intimal thickening (bar) resembling diffuse intimal thickening
in man [[[ 151
8 Atherosclerosis of rabbit artery representative of that induced
in rabbits repeatedly injected with foreign serum protein and
subsequently fed a cholesterol-supplemented diet 40 to 80 days
after last injection ......................................... 151
9 Cardiac allograft and recipient's heart and aorta from rabbit
fed a cholesterol-supplemented diet and treated with immuno-
suppressives ............................................... 152
10 Proximal left main coronary artery of cardiac allograft that
functioned for approximately two weeks ...................... 152
1 1 Atheromatous change in gastric artery of normocholesterolemic
chicken infected with Marek's disease herpesvirus and fed a diet
low in cholesterol .......................................... 153
12 Higher magnification of artery shown in Fig. 11 ............... 153
13 - Cross-section of heart of a chicken infected with Marek's
disease herpesvirus and fed cholesterol-supplemented diet for
1 5 weeks ................................................. 154
14 Major coronary artery of chicken infected with Marek's disease
herpesvirus and fed a cholesterol-supplemented diet for
1 5 weeks ................................................. 154
15 Photomicrograph of endothelial surface of rabbit that received
one injection of horse serum and was sacrificed 15 days later .... [55
16 Luminal surface of large coronary artery of transplanted
-------
Number Page
Menzel, Smolko, Gardner, and Graham
1 Three major regions of the respiratory tract 178
2 - Some common lung diseases in man 180
3 - Spirometric measurements of lung volumes 185
4 - Examples of forced expiratory volume (FEV), forced vital
capacity (FVC), and maximal midexpiratory flow rates (MMFR)
in normal individuals and patients with chronic obstructive and
restrictive lung diseases during forced expiration into a
spirometer 188
5 - Flow-volume curves illustrating the different patterns existing
in patients with obstructive and restrictive diseases 188
6 - Comparison of normal lung volumes to those observed in
restrictive and obstructive disease states 189
-------
TABLES
Number Pa8e
Brockman and de Serres
1 - An outline of twenty-four short-term tests for genetic toxicity
selected for systematic review and evaluation in Phase 1 of the
EPA's GENE-TOX Program 20
2 - Possible use of tests from Table 1 for hazard identification,
hazard evaluation and risk estimation 29
Upton
1 - Some properties of transformed cells 41
2 - Comparative effects of different classes of chemicals with
respect to transformation of hamster embryo cells in vitro,
mutagenicity in Salmonella, and carcinogenicity in vivo 44
3 - Results with the in vitro transformation assay (colony formation
in soft agar), as compared with five other short-term tests for
carcinogenicity 45
4 - Cell culture systems investigated for assaying the carcinogenic
potential of chemicals 48
5 - Biological properties of promoting agents, as compared with
initiating agents 49
6 - Comparative results of short-term tests when used alone or in
combinations 50
7 - Comparative reliability of results from in vitro transformation
test and Salmonella mutation test, used alone or in combination,
on various classes of chemicals 53
Miller and Guarino
1 - Nephrotoxic metallic compounds 98
2 - Nephrotoxic organic compounds 98
3 - Portions of the nephron that can be studied by the use of available
in vitro procedures 106
Weiss
1 - Adverse neurobehavioral responses ascribed to metals 109
2 - Developmental progress in mice 125
3 - Comparison of children with high and low tooth lead
concentrations on a teacher rating scale J2g
4 Intelligence test scores of children with high and low tooth lead
concentrations
-------
Number Page
Menzel, Smolko, Gardner, and Graham
1 - Inflammatory and immune related protein constituents of normal
human alveolar epithelial fluid 194
2 - Indicators of injury in lung lavage fluids 195
3 - Properties and functions of normal human alveolar
macrophages 196
xvn
-------
QUANTITATIVE ASSESSMENT OF RISKS*
Richard B. Setlow
INTRODUCTION
Assessment of risks involves an estimation of the extent of the biological effects
that result from exposure to particular levels of environmental agents. I use the word
"effects" in a general sense and, even though most of the examples cited are in the
field of carcinogenesis, the description should be applicable to most biological
endpoints. The difficulties in making such estimates led one National Research
Council Committee (National Research Council, 1980a) to recommend that the "Of-
fice of Pesticide Programs of the Environmental Protection Agency (EPA) should
abandon its attempts to produce numerical estimates of the effects of the use of pes-
ticides on human mortality and morbidity except when reliable human epidemiolog-
ical data are available." But Harris (1981) states:
A robust epidemiological study of the effect of an environmental agent on
the risk of human cancer should have the following characteristics: Well-
defined groups of exposed subjects and comparable control subjects
should be identified. The magnitude and duration of individual exposures
should be measured. Possible significant confounding factors should be
evaluated. Potential biases caused by the nonrandom selection or follow-
up of subjects should be avoided. The duration of the follow-up should be
sufficient to observe a significant increase in the incidence of the suspected
cancer. The number of persons or person-years at risk should be sufficient
to detect a statistically significant difference in cancer rates between ex-
posed and control subjects.
Clearly, few studies can satisfy these criteria.
At low doses of environmental agents, the effects of concern are either small ones,
such as minor headaches, that affect large numbers of people or large ones, such as
cancer, that affect small numbers of people. The degree of risk is estimated by ex-
trapolation from animal tests or from accidental exposure of individuals to high
doses. But before we can distinguish the extent of the risk, we need to be able to
measure the background level of the effect in the human population in the absence of
any known exposure. The fact that background levels vary from one population to
another is taken as an indication of the role of the environment in ill health. Never-
theless, there are few detailed explanations for the magnitude of background levels.
However, if one could show (which one cannot at present) that the cause of the back-
ground effect was the same as from an environmental agent, one could safely infer
that no threshold dose existed. The dose-response curve would have a finite slope at
low doses because any presumptive threshold dose will have been exceeded by the
cause of the background effect (Hoel, 1980). If the background is high, a relatively
large increased incidence is needed to be detectable, implying that individuals must
•Research carried out under the auspices of the United States Department of Energy.
The Author: Richard B. Setlow is Chairman of the Department of Biology at Brookhaven National Labora-
tory, Upton, Long Island, New York 11973.
-------
have been exposed to relatively high doses. On the other hand, if the background
level is low, relatively small changes in incidence are readily observable and hence
may be associated easily with the responsible environmental agent(s). Mesothe-
liomas of the lung arising from exposure to asbestos and adenocarcinomas of the va-
gina in young women resulting from exposure in utero to diethylstilbestrol are
examples of such cause-and-effect association (Herbst et al., 1977). Of course, a pre-
cise estimate of risk from such epidemiologic studies requires an adequate knowl-
edge of the exposure levels.
The assessment should in principle use human data but fortunately not many such
data exist. When they do, uncertainties in exposure levels and in the number of
people affected mask the shape of the dose-response curve at low doses. A typical
example is the effects arising from exposure to ionizing radiation. The lifetime risk
to males of mortality from leukemia or bone cancer is approximately 100 per 104
individuals (Fig. 1). The'acute radiation dose that would double this risk is between
150 and 200 rads (National Research Council, 1980b). Judging from animal experi-
ments, the doubling dose would be appreciably higher at low dose rates. Hence, the
background frequency of mortality from leukemia cannot be ascribed to the back-
ground of ionizing radiation—natural plus manmade—of approximately 0.1 rad per
year. The problem in extrapolating dose-response data from high doses to low doses
is illustrated by the curves shown in Fig. 1. Human data obtained at relatively high
doses (50 to 300 rads), as a result of the atomic bombs dropped on Japan, have been
analyzed extensively and three possible dose-response models have been described
for such data.* These models, discussed in a National Research Council report (Na-
tional Research Council, 1980b), are called the linear (L), linear quadratic (LQ) and
quadratic (Q) models. The curves drawn in Fig. 1 are based on the equations derived
from these data. Because of the high frequency of background mortality, any sample
of 10,000 individuals, unirradiated or irradiated with a low dose, will show a large
statistical uncertainty about the average mortality figure of approximately 100. The
confidence interval shown is twice the standard deviation, the latter being the square
root of the average number of events. Obviously, data obtained from the exposure
of 104 males to 30 rads cannot be used to distinguish among the three models. Pop-
ulations appreciably greater than 104 would have to be used. Clearly, as the dose goes
down, the population examined must increase. Thus, the shape of a dose-response
curve cannot be determined at low doses unless one uses tremendous numbers of
humans or experimental animals. Even if one could do such an experiment on peo-
ple, the uncertainty in the doses and the possibility that the responses of individuals
may vary greatly make such experiments more or less useless for risk assessment for
the average person. Therefore, the emphasis in this volume is on short-term tests.
RISK ASSESSMENT FROM SHORT-TERM TESTS
The problem is how to extrapolate quantitatively from short-term tests to assess-
ments of risk to humans. An intermediate step in this process is to perform chronic
experiments with animals. Such experiments also have their shortcomings because
they use limited numbers of animals and, to obtain observable results, they require
high exposures. Thus, the interpreter of such experiments is also faced with the prob-
lem of extrapolating to low exposure levels. A knowledge of the dose-response
curve is needed, but, as we saw in Fig. 1, the correct shape of the curve at low doses
cannot be derived from data at high doses. Hence, the extrapolation is made using
mathematical models of one type or another that are felt to give reasonable upper
limits to the risk and therefore to give safe environmental standards. It is not our
aim to examine these many mathematical possibilities. They are discussed well and
extensively elsewhere (Cornfield etal., 1980; Mantel, 1980;Peto, 1977; Preussmann,
1980).
•The accuracy of the dosimetry is in question (Marshall, 1981).
-------
250
c
-------
Extrapolations from short-term tests to human risks involve three major un-
resolved difficulties: (1) the projection from acute, short-term to chronic, long-term
exposures; (2) the complicated interconnected kinetics of metabolic activation, de-
activation and repair; and (3) the fact that a number of important biological end-
points require multiple environmental insults. (For example, the effects of tumor
promoters may be all important at low dose levels.) We take up these difficulties be-
low.
Acute Versus Chronic Exposure
It is possible to determine dose-response relation for acute exposures as the many
short-term tests described in this volume illustrate. How should one extrapolate
such data to chronic exposures? There is no single answer to this question without a
knowledge of the pharmacokinetics of the various processes involved (Anderson et
al., 1980). Unfortunately, in many instances the details of the processes are not
known and therefore an elucidation of the kinetics may not be possible. It is known,
however, that chronic exposure to a number of chemicals, not necessarily
carcinogens, can change the level of enzyme systems that convert procarcinogens to
ultimate carcinogens and degrade the ultimate carcinogens to unreactive, excretable
compounds. Moreover, low chronic exposure of animals to alkylating agents renders
them better able to remove DNA damage from a large challenge dose (Montesanoet
al., 1979). Thus, how one should extrapolate from acute to chronic exposures or
from high to low rates of exposure depends upon the rates of various reactions, the
lifetimes of reaction products, and the biological effects of such products.
Metabolic Pathways
The analysis of the biological effects of environmental agents is complicated
because many agents classified as hazardous do not react directly with cell
macromolecules such as protein, RNA and DNA or with membranes. They often
must be metabolically activated to form reactive intermediates. Such intermediates
may affect crucial structures in cells or, before they do so, may be inactivated to
innocuous compounds. Hence, the ability of an agent to damage a particular tissue
depends critically on the activation and deactivation pathways in the organism. An
additional complication, hidden in the above analysis, is the implication that one
knows which are the important activated compounds, and which are the important
reaction products in macromolecules. In this context, the shape of the dose-response
curve depends not on the concentration of the ingested agent but on the internal level
of reactive compounds. For example, Bailey et al. (1981) have reported that analysis
of the in vivo production of methylcysteine in rat hemoglobin shows a linear dose-re-
sponse curve after injection of various concentrations of methyl methanesulfonate;
on the other hand, after injection of dimethylnitrosamine a curvilinear dose-re-
sponse curve was obtained. In this system, a high background level of methylcysteine
was observed.
Not only is it essential to know which products are the deleterious ones in order to
measure their kinetics of appearance and disappearance, but one must have accurate
knowledge of the relative rates of removal of the macromolecular adducts, not just
their rates of formation. The rate of repair or removal and the chemical instability of
particular adducts determine their half-life, and this value, when measured against
the rates of replication, transcription, and translation of genetic material, will
determine the extent of the harmful biological effects. The half-lives of some prod-
ucts also depend on the tissue in which the product resides.
One of the better examples to illustrate these repair considerations involves the
products resulting from treatment of cells with alkylating agents (Singer, 1979). Sup-
pose that the only important products are those that affect DNA. Even at this level of
investigation, there are many products formed. One of the more numerous reactions
is the one with guanine. But there are a number of different guanine adducts. For ex-
ample, 7-alkylguanine is the most plentiful and 06-alkylguanine is formed in a much
-------
smaller amount (Fig. 2). However, the latter product is felt to be the more important
biologically, and there are reasonable indications that it is the one that is associated
with mutagenesis and carcinogenesis. In any event, 06-alklguanine is removed more
rapidly from many tissues than is 7-alkylguanine (Rajewsky et al., 1977). Such tis-
sues are less likely to be the targets for carcinogenesis. For example, 06-ethylguanine
has a very long half-life in the neural tissue but a very short half-life in the liver of
newborn rats, and the rats develop neural, not liver, tumors.
A complication in this simple description is that the cellular machinery that re-
moves 06-alkylguanine may become saturated by high levels of the product(Medcalf
and Lawley, 1981; Pegg and Hui, 1978). As a result, the level of DNA adducts at high
doses and over short periods of time would be expected to increase faster than a
linear power of the dose. (See Fig. 3.) If a low concentration (Fig. 3B) is adminis-
tered for a long time, one would expect that particular repair systems might be able
to keep up with the production of DNA adducts and hence their level would be less
than those observed after treating cells at high concentrations as illustrated in Fig.
3 A. Moreover, there are good indications in prokaryotic systems, and some in
eukaryotic ones, that low chronic doses of alkylating agents may stimulate the
production of a repair system to remove 06-alkylguanine. Clearly a knowledge of the
saturation level of enzymic repair systems and of the properties of any inducible re-
pair systems as a function of dose is necessary for an extrapolation from high to low
doses and from acute to chronic ones. Likewise, comparable kinetic data are needed
for human tissues before any valid extrapolation can be developed from cells to ani-
mals to humans. One of the most important future lines of investigation is in the
direction of determining the concentration of DNA adducts in target tissue. This is
not an easy task at low chronic exposure levels but is being approached for a number
of specific products by ultrasensitive enzymic radioimmunoassays that are capable
of detecting close to 10~15 moles of the product (Hsu et al., 1981; Hsu et al., 1980;
Muller and Rajewsky, 1980).
One well-documented case of the quantitative role of DNA repair in carcinogenesis
is the effects of ultraviolet radiation in the induction of skin cancer in humans (Scott
and Straf, 1977). The most common photoproducts in cellular DNA are intrastrand
pyrimidine dimers. They are responsible for killing and mutagenesis in microorga-
nisms and for tumor production in a model fish system. Because pyrimidine dimers
are known to have adverse biological consequences, they have been extensively stud-
ied for their repairability (Setlow, 1980). Individuals with the disease xeroderma pig-
mentosum have a cancer risk from exposure to sunlight that is 103-to 104-fold greater
than that of the average population because these individuals have a defect in one or
more DNA repair pathways. Despite the high rate of dimer formation by sunlight
(~105 dimers per hr per exposed skin cell), normal DNA repair reduces the effect of
the ultraviolet dose to the skin by an approximate factor of 7 to 20 (Setlow, 1980).
The big difference in the skin cancer prevalence between individuals with xeroderma
pigmentosum and normal individuals is observed at the high levels of DNA damage,
damage from sunlight exposure that results in the formation of more than 104 dimers
per skin cell. What would be the effect of DNA repair at lower UV doses and at very
low dose rates? The effect would be less because, on the average, individuals with
xeroderma pigmentosum are not absolutely deficient in DNA repair so that at lower
dose rates their repair system might not be saturated so readily. This question of sat-
uration of repair and dose-rate levels is an important one to answer if we are to ad-
dress the problem of risk assessment. Cairns (1980) has extrapolated from the
observation that people with xeroderma pigmentosum get no more internal cancers
than expected for the average population to suggest that U V-mimetic chemicals, such
as polycyclic aromatic hydrocarbons, are not important in environmental carcino-
genesis. Such carcinogens in the environment probably produce much less DNA
damage per exposed cell than does a short sunlight exposure. Hence it is important
to know how the relative sensitivities of xeroderma pigmentosum and normal cells
depend on dose and on dose rate. If the relative sensitivities are more or less
-------
H2N
HN
H,N ' " N
deoxyribose
deoxyribose
06-ethylguanine
7-ethylguanine
Figure 2. Two ethylation products of deoxyguanosine in DNA. Ethylation at the O6-
position leads to mispairing more often than ethylation at the 7-position.
(A)
HighC
Low C
1 2 3
Dose (concentration x time)
Figure 3. Each panel shows two possible relations between the dose of an
environmental agent and the relative level of specific DNA adducts
formed in cells. If adducts are unstable, are repaired, or are diluted by
cell division, one would expect steeper dose-response curves at high
concentrations, A, than at low concentrations, B. The background level
of adducts may be greater than zero, as shown, as a result of other,
not measured, chemical reactions.
-------
independent of dose and dose rate, an extrapolation such as Cairns made would
have high credibility. If, instead, the sensitivity of xeroderma pigmentosum cells
decreased to normal values at very low dose rates, the extrapolation would not be
valid. The dose-response curves at high dose rates for killing or mutagenesis of xero-
derma pigmentosum cells by polycyclic hydrocarbon derivatives indicate that they
are much more sensitive than normal cells even at low total doses (Maher et al.,
1977). Nevertheless, the question remains an open one because these data have not
been obtained directly at the dose rates that would be equivalent to environmental
exposures; furthermore, ultraviolet carcinogenesis is more than just the result of a
single exposure to light.
Multiple Insults
Although changes in DN A can result in mutation and may induce transformation
or tumorigenesis in a number of simple model systems, the effects observed in
animals usually depend on a repetitive stimulus by a carcinogen or on the effects of a
carcinogen followed by a promoter (Slaga et al., 1978). Also, individuals are usually
exposed to a variety of agents at the same time or sequentially and such agents might
act synergistically so that what could be a threshold in experiments dealing with one
compound disappears in the real world.
What is needed, to extend short-term animal data to low doses and then to ap-
ply these data to humans is not only extrapolation rules (Rail, 1977) but also a basic
understanding of the relationship between macromolecular damage and biological
effects (Bridges, 1980; Campbell, 1980; Comptroller General's Report to Congress,
1981; McElheny and Abrahamson, 1979). For example, if damage to DN A were the
important initiating effect in carcinogenesis, the data in Fig. 3 could be used to pre-
dict the possible levels of biological effects from the level of DNA adducts (Fig. 4).
The background levels of adducts or biological effects in both graphs could result
from the action of chemicals other than the one under consideration or from the nor-
mal metabolic wear and tear in biological systems in the absence of specific environ-
mental chemical insults. Of course, Fig. 4 is a great oversimplification since the bio-
logical effect would depend not only on the level of particular damage but on how
such damage changes with the time scale of the biological system. Other dimensions
should be included in Fig. 4 indicating the effects of changing the rate of production
of adducts and of other environmental agents, such as promoters or protectors. In
the absence of such data, extrapolation from animal experiments at high dose rates
to low dose rates necessitates somewhat arbitrary dose-response relations (Cornfield
et al., 1980; Mantel, 1980); moreover, the linear extrapolation usually cannot be
excluded (Guess et al., 1977). Thus, a rational approach requires a constant interplay
between fundamental studies at the molecular level and animal data, and an
interplay between animal pharmacokinetics and human pharmacokinetics
(Anderson et al., 1980; Bridges, 1980; Campbell, 1980). A basic part of these interac-
tions is a knowledge of which macromolecular changes are important in producing
biological effects and of the experimental techniques that are required for measuring
such changes in animals and humans. It is the rare instance, such as nonmelanoma
skin cancer, in which there are crude epidemiologic data that are sufficient to predict
approximately the increase in the biological effect that would result from an increase
in solar ultraviolet. The extrapolation is relatively easy because almost all non-
melanoma skin cancer seems to arise from sunlight exposure and because basic bio-
logical data indicate which range of solar wavelengths is carcinogenic (Scott and
Straf, 1977).
It is not sufficient to know that some particular agent is bad. It is necessary to
know how bad it is and to whom, and what is the extent of the risk to one segment of
the population against the benefit to another. We are also faced with the fact that dif-
ferent segments of the human population react differently to environmental agents.
Thus, enhanced ultraviolet radiation is not a real problem for highly pigmented indi-
viduals and is a much more serious problem for fair-skinned individuals with a Celtic
-------
1 2 3
[DNA adducts]
Figure 4. Hypothetical dose-response curves relating a particular biological effect with
concentration of specific DNA adducts. The animal data represent the typical
ones obtained at high doses and, by themselves, give no indication of the
possible curves joining them to the point at zero adducts. The background
biological effect could arise from causes other than the specific adducts
under consideration.
mcestry. Moreover, there exist small, seemingly sensitive subsets of the population,
subsets that are defective in one or more mechanisms of DNA repair (Setlow, 1978).
In fact, initial experiments indicate that there may be a relatively wide variance in
DNA repair capacity even among the seemingly normal population. Certainly there
ire many indications that the ability to metabolize potentially hazardous
:nvironmental chemicals may vary between individuals as a function of age and as a
function of other agents in the environment or in the diet (Harris et al., 1980).
Without a rational approach based on fundamental knowledge, we are in danger
of being led to absurdities of regulation, especially because the sensitivity of analyt-
ical techniques to detect deleterious agents is increasing rapidly. For example, there
is little doubt that high concentrations of diethylstilbestrol are dangerous, but to ban
such a compound from meat at the present sensitivity of detection seems to suggest
that the benefits at this level do not outweigh the estimated one death per United
States population per 100 years (Jukes, 1977). A ban implies that there are no natu-
rally occurring biologically active analogs that might have effects orders of magni-
tude greater.
-------
PERSPECTIVE
Risk assessment should involve three parameters: background, threshold, and ex-
trapolation. Background refers to the existence of an environmental level of a potent-
ially deleterious agent independent of the activities of humans and includes a level of
biological effects ascribable to unknown causes or to causes by some agent other
than a chemical under consideration. Even in the absence of threshold, a knowledge
of the shape of the dose-response relation at low dose levels is crucial for predicting
the relative risk. Complications arise because the environment and its interaction
with biological systems are made up of nonindependent terms although the labora-
tory analyses usually assumes they are so. Most of the factors involved in risk assess-
ment are not determinable directly by experiment but must be extrapolated on the
basis of theory or of fundamental biological knowledge of the systems under
investigation. That knowledge is imperfect, but, just as an assessment of risk at low
doses of ionizing radiation can come about only from a fundamental understanding
of the interaction between radiation and cells in animals, so an assessment of risk
from low levels of chemicals can come about only from a fundamental under-
standing of the interactions of such agents with cells and with tissues of animals and
humans. The direct experiment on humans is not possible and, if it were, its existence
would indicate the failure of present risk assessment procedures.
REFERENCES
Anderson, M.W., D.G. Hoel, and N.L. Kaplan. A general scheme for the incorporation of
pharmacokinetics in low-dose risk estimation for chemical carcinogenesis: Example-vinyl
chloride. Toxicol. Appl. Pharmacol. 55: 154-161, 1980.
Bailey, E., T.A. Connors, P.B. Farmer, S.M. Gorf, and J. Rickard. Methylation ofcysteinein
hemoglobin following exposure to methylating agents. Cancer Res.47(3514-2517, 1081.
Bridges, B.A. An approach to the assessment of risk to man from DNA damaging agents.
Arch. Toxicol. Suppl. 5:271-281, 1980.
Cairns, J. The origins of human cancers. Nature 259:353-357, 1980.
Campbell, T.C. Chemical carcinogens and human risk assessment. Fed. Proc. 39:2467-2484,
1980.
Comptroller General's Report to Congress. Problems in Assessing the Cancer Risk of Low-
Level Ionizing Radiation Exposure, Vol. 1, General Accounting Office, Washington, D.C.,
1981.
Cornfield, J., K. Rai, and J.V. Ryzin. Procedures for assessing risk at low levels of exposure.
Arch. Toxical. Suppl. 5:295-303, 1980.
Guess, H., K. Crump, and R. Peto. Uncertainty estimates for low-dose-rate extrapolations of
animal carcinogenicity data. Cancer Res. 37:3475-3483, 1977.
Harris, C.C., J.J. Mulvihill, S.S. Thorgeirsson, and J.D. Minna. Individual differences in can-
cer susceptibility. Ann. Intern. Med. 92:809-825, 1980.
Harris, J. E. Potential Risk of Lung Cancer from Diesel Engine Emissions. Report to the Diesel
Impacts Study Committee, National Academy Press, Washington, D.C., 62p.,1981.
Herbst, A.L., R.E. Scully, S.J. Robbey, W.R. Welch, and P. Cole. Abnormal development of
the human genital tract following prenatal exposure to diethylstilbestrol. In Origins of Hu-
man Cancer, edited by H.H. Hiatt, J.D. Watson, and J.A. Winsten. Conferences on Cell
Proliferation, Vol. 4A, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., pp.
399-412, 1977.
Hoel, D.G. Incorporation of background in dose-response models. Fed. Proc. 39:73-75, 1980.
Hsu, I.C., M.C. Poirier, S.H. Yuspa, D. Grunberger, I.B. Weinstein, R.H. Yolken, and C.C.
Harris. Measurement of benzo(a)pyrene-DNA adducts by enzyme immunoassays and ra-
dioimmunoassay. Cancer Res. -#7:1091-1095, 1981.
Hsu, I.C., M.C. Poirier, S.H. Yuspa, R.H. Yolken, and C.C. Harris. Ultrasensitive enzymatic
radioimmunoassay (USERIA) detects femtomoles of acetylaminofluorene-DNA adducts.
Carcinogenesis 7:455-458, 1980.
-------
Jukes, T.H. DBS in beef production: Science, politics, and emotion. In Origins of Human Can-
cer, edited by H.H. Hiatt, J.D. Watson, and J.A. Winsten. Conferences on Cell Prolifera-
tion, Vol. 4C, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., pp. 1657-1664,
1977.
Land, C.E. Estimating cancer risks from low doses of ionizing radiation. Science 209:1197-
1203, 1980.
Maher, V.M., J.J. McCormick, P.L. Grover, and P. Sims. Effect of DNA repair on the cyto-
toxicity and mutagenicity of polycyclic hydrocarbon derivatives in normal and xeroderma
pigmentosum fibroblasts. Mutat. Res. 43: 117-138, 1977.
Mantel, N. Limited usefulness of mathematical models for assessing the carcinogenic risk of
minute doses. Arch. Toxicol. Suppl. 3:305-310, 1980.
Marshall, E. New A-bomb studies alter radiation estimates. Science 212:900-903, 1981.
McElheny, V.K., and S. Abrahamson (eds.) Banbury Report 1. Assessing Chemical Mutagens:
The Risk to Humans, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., 367
p., 1979.
Medcalf, A.S.C.,and P.D. Lawley. Time course of 06-methylguanine removal from DNA of N-
methyl-W-nitrosourea-treated human fibroblasts. Nature 259:796-798, 1981.
Montesano, R., H. Bresil, and G.P. Margison. Increased excision of 06-methylguanine from
rat liver DNA after chronic administration of dimethylnitrosamine. Cancer Res. 39:1798-
1802, 1979.
Muller, R., and M.F. Rajewsky. Immunological quantification by high-affinity antibodies of
06-ethyldeoxyguanosine in DNA exposed to A'-ethyl-A'-nitrosourea. Cancer Res. ^0:887-
896, 1980.
National Research Council. Committee on Prototype Explicit Analyses for Pesticides. Regu-
lating Pesticides, National Academy of Sciences, Washington, D.C., 288p., 1980a.
National Research Council. Committee on the Biological Effects of Ionizing Radiations. The
Effects on Populations of Exposure to Low Levels of Ionizing Radiation: 1980, National
Academy of Sciences, Washington, D.C., 524p., 1980b.
Pegg, A.E., and G. Hui. Removal of methylated purines from rat liver DNA after administra-
tion of dimethylnitrosamine. Cancer Res. 38:20\ 1-2017, 1978.
Peto, R. Epidemiology, multistage models, and short-term mutagenicity tests. In Origins of
Human Cancer, edited by H.H. Hiatt, J.D. Watson, and J.A. Winsten, Conferences on Cell
Proliferation, Vol. 4C, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., pp.
1403-1428, 1977.
Preussmann, R. The problem of thresholds in chemical carcinogenesis. Some views on theoret-
ical and practical aspects. J. Cancer Res. Clin. Oncol. 97:1-14, 1980.
Rajewsky, M.F., L.H. Augenlicht, H. Biessman, R. Goth, D.F. Hulser, O.D. Laerum, and L.
Ya Lomakina. Nervous-system-specific carcinogenesis by ethylnitrosourea in the rat: Mo-
lecular and cellular aspects. In Origins of Human Cancer, edited by H.H. Hiatt, J.D.
Watson, and J.A. Winsten, Conferences on Cell Proliferation, Vol. 4B, Cold Spring Harbor
Laboratory, Cold Spring Harbor, N.Y., pp. 709-726, 1977.
Rail, D.P. Species differences in carcinogenesis testing. In Origins of Human Cancer, edited by
H.H. Hiatt, J.D. Watson, and J.A. Winsten, Conferences on Cell Proliferation, Vol. 4C,
Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., pp. 1383-1390, 1977.
Scott, E.L., and M.L. Straf. Ultraviolet radiation as a cause of cancer. In Origins of Human
Cancer, edited by H.H. Hiatt, J.D. Watson, and J.A. Winsten, Conferences on Cell Pro-
liferation, Vol. 4A, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., pp.
529-546, 1977.
Setlow, R.B. Repair deficient human disorders and cancer. Nature 277:713-717, 1978.
Setlow, R.B. Different basic mechanisms in DNA repair. Arch. Toxicol. Suppl. 3:217-228, 1980.
Singer, B. W-Nitroso alkylating agents: Formation and persistence of alkyl derivatives in mam-
malian nucleic acids as contributing factors in carcinogenesis. J. Nail. Cancer Inst. 62:1329-
1339, 1979.
Slaga, T.J., A. Sivak, and R.K. Boutwell (eds.) Carcinogenesis. Vol. 2, Mechanisms of Tumor
Promotion and Cocarcinogenesis, Raven Press, New York, 558p., 1978.
10
-------
SHORT-TERM TESTS FOR GENETIC TOXICITY
Herman E. Brockman and Frederick J. de Serres
INTRODUCTION
The genetic material of the human species and of all other species is our most pre-
cious heritage from long and still ongoing evolutionary processes. The great respon-
sibility of safeguarding this legacy is thrust upon each generation. The discipline of
genetic toxicology deals most closely with this awesome responsibility of ensuring
that the genetic material is transmitted to future generations in a form that will not
lead to increased incidences of ill health. This responsibility is difficult because of the
increasing number and diversity of man-made (synthetic) chemicals that are entering
our environment. The growth of the chemical industry burgeoned after World War
II, particularly the petrochemical industries. Presently, over 50,000 synthetic chemi-
cals are rather commonly used, and about 1,000 new chemicals or derivatives of ex-
isting ones enter the market place each year. Most of these chemicals eventually
become part of our environment, and yet many of them have not been tested for their
genetic toxicity before being released for general use.
Currently, there is great concern that toxic substances being introduced into our
environment are contributing factors in various classes of human ill health,
including genetic diseases, cancer, birth defects, heart attacks, strokes, and aging.
We have come to realize that toxic substances often are in the air we breathe, the
food we eat, and the water we drink. Furthermore, we are often exposed to toxic sub-
stances in our occupations and homes.
In the last decade, a major concept about the mechanism by which many toxic
substances cause certain adverse health effects has gained increasing acceptance.
This concept is that the primary or initial cause of certain diseases is a mutation in
the genetic material of a cell. The development of this concept has resulted in the
field of genetic toxicology, a hybrid discipline that draws upon the traditional
sciences of genetics and toxicology. Genetics pursues an understanding of how
biological traits are inherited, how information is encoded in the genetic material,
and how this information is used in cells. Toxicology deals with the identification of
toxic substances that cause illness and death, and with the mechanisms of their ac-
tion. Genetic toxicology is concerned with toxic effects that are due to an agent's
ability to alter the information encoded in the genetic material of an organism.
In this review, we will discuss the new field of genetic toxicology. First, we will
provide a brief historical perspective (1900-1970), introducing and defining many of
the terms and concepts that we will use; then we will present the major advances that
have taken place in the last ten years. In this connection, we will discuss the federal
legislation and actions that have accompanied the development of this field, and fi-
nally we will speculate on future emphases in genetic toxicology.
The Authors: Herman E. Brockman works in the Department of Biological Sciences, Illinois State Univer-
sity, Normal, Illinois 61761. Frederick J. de Serres works in the Mutagenesis Branch of the National Insti-
tute of Environmental Health Sciences, Research Triangle Park, North Carolina 27711.
-------
HISTORICAL PERSPECTIVE
Genetics is one of the oldest of all sciences, and at the same time it is a relatively
young science. When humans started to cultivate plants and domesticate animals
about 10,000 years ago, they also began to practice artificial selection—that is, they
selected for breeding those members of a species that had desirable characteristics.
Thus, early humans chose certain phenotypes (what the individuals looked like) and,
simultaneously, the underlying genotypes (the individuals' genetic constitutions).
However, a sound experimental basis for genetics was not established until 1866,
when the Augustinian monk Gregor Mendel published his results and interpreta-
tions from crosses between varieties of the common garden pea Pisum sativum
(Mendel, 1866). The significance of this publication was not appreciated until 1900,
when the same kinds of results and interpretations were published by others, at
which time genetics became recognized as a formal science.
Mendel and others concluded that in diploid organisms, such as humans, each
gene is present in one copy in the gametes (eggs and sperm) but is present in two
copies in all other cells in the body. Furthermore, a gene can occur in two different
states (alleles): a dominant allele and a recessive allele. For example, in the garden
pea, purple and white are two different phenotypes of one trait, flower color. The
allele /"(purple) is dominant to the allelep(white)—therefore, the genotype Pp, as
well as the genotype PP, is purple, whereas the genotype pp is white. Today, we
know that there can be many different alleles, some of which can have intermediate
states of dominance.
Some Major Discoveries Concerning Mutation
One of the rediscoverers of Mendel's results, de Vries, coined the terms muta-
tion and mutant (de Vries, 1901). A mutation is an abrupt change in phenotype that
is inherited, and a mutant is the individual resulting from such a change. In 1910,
Morgan reported that a single white-eyed male had appeared in a culture of red-eyed
fruit flies (Drosophila melanogaster) that had reproduced for many generations in
the laboratory (Morgan, 1910). The red-eyed flies had been collected originally in
nature (in the "wild"); hence, they were called wild type. Morgan showed that the
mutant allele for white eyes is recessive to the wild-type allele for red eyes and is lo-
cated on one of the sex chromosomes referred to as the X chromosome. By 1915,
many mutants of Drosophila had been recognized by their different phenotypes, and
each mutation had been mapped at a specific place (locus) on a particular chromo-
some by Morgan and his collaborators. Because these mutants had arisen abruptly
in cultures of wild-type flies, they were referred to as spontaneous mutants.
Many investigators in the early 1900s considered whether mutations could be in-
duced by certain physical and chemical agents. Using Drosophila, Muller( 1927) first
devised a test to detect induced mutation. This test is referred to as the sex(X)-linked
recessive lethal test because the mutations being measured are located on the X chro-
mosome, are recessive, and cause death. The test is unambiguous, quantitative,
sensitive, and rather simple. Muller showed that the physical agent X ray, at the dose
used, caused a 140-fold increase over the spontaneous frequency of sex-linked re-
cessive lethal mutations. Muller had discovered the first mutagen, and we also can
credit him with devising the first short-term test for mutagenicity. Muller's paper is
the real beginning of mutation research within the science of genetics and led
eventually to the field of genetic toxicology. His paper also marks the origin of envi-
ronmental mutagenesis, which is the study of mutagens that are known or suspected
to be in our environment. Muller also clearly stated that the environmental mutagen
X ray is a potential hazard to the genetic material of humans. He indicated that
their indiscriminate use in medical practice might result in an increased frequency of
mutations having lethal or detrimental effects in future generations. Unfortunately,
his message was largely ignored until much later.
12
-------
The induced mutations studied by Muller and others were mainly sex-linked reces-
sive lethals or morphological (visible) mutations. Beadle and Tatum became inter-
ested in the underlying biochemistry of these mutations, and in 1941 they reported
the first induction of biochemical mutants (Beadle and Tatum, 1941). They used the
pink bread mold Neurospora crassa, which is a simpler organism than the garden
pea or fruit fly. Wild-type Neurospora grows on a very simple, mainly inorganic,
medium; therefore, it must synthesize a large array of organic compounds and mac-
romolecules needed for its growth and reproduction. They reasoned correctly that
they could induce mutations with X ray or ultraviolet light (UV) in genes determining
the synthesis of enzymes that catalyze biosynthetic steps. Their techniques were ex-
tended in the 1940s and 1950s to isolate biochemical mutants in several other micro-
organisms, including bacteria.
In the 1940s, certain chemicals were shown to be mutagenic. Using the sex-linked
recessive lethal test in Drosophila, Auerbach and Robson (1946) demonstrated that
mustard gas is mutagenic. Research on the mutagenicity of chemicals has expanded
tremendously since that pioneering study.
Certain biochemical mutants of Neurospora isolated by the methods of Beadle
and Tatum were used to devise the reverse-mutation test in Neurospora (Jensen et
al., 1951). Mutations in the direction of wild type—mutant are called forward muta-
tions, whereas mutations in the direction of mutant—wild type are referred to as re-
verse mutations. The reverse-mutation test in microbes is much simpler and less time
consuming than most forward-mutation tests, because a selective (minimal)
medium is used on which only the revertant colonies grow. Reverse-mutation tests
became a popular way to test forthe mutagenicity of large numbers of chemicals. As
data from mutagenicity tests accumulated in the 1950s and 1960s, mutagens were
discovered among numerous classes of chemicals such as food additives, medicinals,
pesticides, cosmetics, and industrial chemicals. Furthermore, it became known that
mutagens are found as highly complex mixtures in our environment.
In experiments where two or more doses of a mutagen are used, dose-response
curves for percent survival and for mutation frequency can be constructed. For most
mutagens, these two activities go together—that is, as the dose of an agent is increas-
ed, the percent survival decreases and the mutation frequency increases. These re-
sults led many people to assume that mutagenic chemicals would be detected by clas-
sical toxicologic tests, because concentrations that caused mutation also caused kill-
ing. But, in the 1950s and 1960s, rather startling examples of chemicals termed "su-
permutagens" were discovered. These chemicals cause high frequencies of mutations
at doses that cause little or no killing. An early example was TV-methyl-N'-nilro-N-
nitrosoguanidine in the bacterium Escherichia coli. The discovery of these mutagens
focused attention on the fact that because existing classical toxicologic tests would
not detect certain chemical mutagens there was a need for short-term tests for
mutagenicity as part of a standard set of toxicologic tests.
A major concern in the area of chemical mutagenesis has been whether chemical
carcinogens also are mutagens. Almost all carcinogens have been identified from
studies in human epidemiology or from tests with laboratory animals, usually mice
or rats. It was noted in the 1950s and 1960s that many chemicals that are carcinogens
in man and laboratory animals are not mutagenic in microbial tests. This discrep-
ancy was found to be due to the fact that these chemicals are not carcinogenic per se;
rather, they are metabolized in the animal to one or more carcinogenic intermediates.
The original noncarcinogenic chemical is termed a procarcinogen, and the carcino-
genic intermediate is termed an ultimate carcinogen. If the conversion is due to en-
zymes, the process is called metabolic activation, and the enzymes are termed acti-
vating enzymes. Similarly, we also refer to promutagens and ultimate mutagens. The
earlier results in microbial tests were due to microbes not having nearly as extensive
a repertoire of activating enzymes as cells of animals, especially those in the major
detoxifying organ, the liver. In order to make the microbial mutagenicity tests more
13
-------
"animal-like," various approaches were used to combine a promutagen, indicator
cells (for example, bacteria), and activating enzymes from animal cells,
usually from rat or mouse liver. Using this important modification, studies in the
1970s suggested that most carcinogens are mutagenic and conversely, that most non-
carcinogens are nonmutagenic.
The Environmental Mutagen Society (U.S.) was organized in 1969. This society
serves as a focal point for scientists from many fields who are concerned about the
adverse effects of environmental mutagens on the health of humans and other
species.
Some Major Discoveries Concerning the Molecular Biology of Mutation
Geneticists soon realized that an understanding of the mutational process was
impossible without answers to the questions: What is the molecular nature of the ge-
netic material? What is the structure of this molecule? How does this molecule
replicate? The answers to these questions emerged dramatically. Avery, MacLeod,
and McCarty (1944) provided the first direct evidence that the genetic material is de-
oxyribonucleic acid (DNA). The basic building blocks of DNA are nucleotides,
which are covalently bonded together in a specific way to form a polynucleotide
chain. Watson and Crick (1953a) proposed a three-dimensional model for the struc-
ture of DNA in which two polynucleotide chains are wound around each other and
around an imaginary central axis in a regular way to give a helical structure called
the double helix. They stressed that the novel feature of their model is what later was
called the complementarity of the polynucleotide chains. This complementarity is
due to the fact that, among the four nucleotide bases in DNA, there are only two le-
gitimate base pairs: the purine adenine (A) with the pyrimidine thymine (T), and the
purine guanine (G) with the pyrimidine cytosine (C). The two polynucleotide chains
are held together along their length by hydrogen bonds between these bases.
In the same year, Watson and Crick proposed models for the replication and spon-
taneous mutation of DNA, both of which are dependent on the complementarity of
the polynucleotide chains (Watson and Crick, 19536). Their model for replication is
disarmingly simple: the double helix unwinds, and each polynucleotide chain acts
as a template for the synthesis of a complementary daughter chain according to the
rules of complementarity. Their model for spontaneous mutation, which was the
first one based upon their model for the molecular structure of DNA, basically
says that the pairing of bases is not a perfect process. Occasionally, illegitimate base
pairing (other than A with T and G with C) occurs during DNA replication, and after
one or more rounds of replication a base-pair substitution mutation occurs in the
DNA molecule.
Two other closely related discoveries about the molecular biology of mutation are
important to our discussion. First, the molecular nature of the initial lesions in DNA
caused by certain mutagens has been elucidated. Second, it has been discovered that
cells possess complex and amazingly effective enzymatic mechanisms for removing
or bypassing certain of these lesions. For example, a major lesion induced by UV is
the pyrimidine dimer, in which two adjacent pyrimidines in the same polynucleotide
chain are covalently bonded together. Because a dimer blocks DNA replication, it
acts as a lethal lesion. Setlow and Carrier (1964) and Boyce and Howard-Flanders
(1964) showed that a dimer and some adjacent nucleotides are excised, and that a
correct sequence of nucleotides is laid down in the resulting gap, using the comple-
mentary chain as a template. The steps are catalyzed by enzymes that constitute a
nucleotide excision-repair pathway. This repair pathway is relatively error free, but
there are other repair pathways that are referred to as error prone, because during
the process of bypassing the dimer an altered (mutated) sequence of bases often is
synthesized. The lethal effect of the lesion is circumvented, but a mutation occurs
during the process. Therefore, error (mutation)-prone DNA repair pathways are
very important mechanisms for mutation.
14
-------
Two Major Principles about Mutation
One principle is based on the fact that higher organisms are composed of two
kinds of cells: somatic cells and germ cells. Only the germ cells may transmit genetic
material to future generations; therefore, the consequence of mutation in the two cell
types is much different. Mutations in somatic cells are thought to be the initiating
event in certain diseases, such as cancer, whereas mutations in germ cells may result
in genetic diseases, such as Down's syndrome and sickle cell anemia, in future gener-
ations.
A second principle is that there are many ways that the genetic material can be al-
tered, resulting in mutations. We will discuss three major types of alterations:
changes in the number of chromosomes, changes in the architecture of chromo-
somes (chromosomal aberrations), and changes within a gene (gene mutations). The
first two types, collectively referred to as chromosome abnormalities, generally are
detected by microscopic examination. Gene mutations usually are too small to be
seen with the light microscope, but they can be detected by genetic tests.
In humans with a normal chromosome complement, each somatic (diploid) cell
contains two of each of 23 kinds of chromosomes. In germinal tissue, diploid cells
undergo meiosis, a division process that produces the gametes. A key event in meio-
sis is that the two members of a pair of chromosomes first pair, and then disjoin, one
member of each pair becoming incorporated into a gamete. Two gametes fuse at fer-
tilization, and the diploid chromosome number is restored.
One major type of alteration in the genetic material is a change in the number of
chromosomes per somatic cell to give other than 46 chromosomes. If the number of
chromosomes is an integral multiple of 23, the condition is called euploidy (for ex-
ample, triploid - 69). If the number is not an integral multiple of 23, the condition is
called aneuploidy (for example, monosomy = 45, trisomy 47). In humans,
euploidy (other than the diploid condition) generally is lethal before birth, whereas
some kinds of aneuploidy are compatible with life but usually cause severe mental
and physical defects. Aneuploidy arises when the members of a pair of chromosomes
do not disj oin normally, so that one gamete may receive both members of the pair of
chromosomes and the other gamete may receive no members of that pair. When a
gamete that contains less or more than 23 chromosomes combines with a normal
gamete, an aneuploid individual is produced.
Loss of a chromosome in humans usually is incompatible with life, except for
monosomy (2N - 1) for the X chromosome (45,X). Most 45,X cases are aborted
spontaneously, early in pregnancy; those that live are sterile females characterized
by short stature and failure to develop female secondary sex characteristics (Turner's
syndrome). Most kinds of trisomy (2N+ 1 = 47) are lethal. About one-half of detected
spontaneous abortions that are chromosomally abnormal are trisomics; individuals
that live often have severe physical and mental defects. The best-known example of
trisomy is of chromosome 21, which results in Down's syndrome.
Another major type of alteration in the genetic material is a change in the
architecture of the chromosomes but without the loss or gain of a complete chromo-
some. There are four kinds of chromosomal aberrations: deficiencies and duplica-
tion of less than one entire chromosome, and translocations and inversions, which
are rearrangements of less than one entire chromosome. In translocations, a section
of one chromosome is moved to a new location in the genome, whereas in inversions,
a section of a chromosome is inverted 180°. Chromosomal aberrations are caused by
agents that break chromosomes (clastogenic agents). A third major type of altera-
tion in the genetic material is called gene mutations, because no more than one gene
is affected. Intermediate between gene mutations and the deficiency class of chromo-
some aberrations are multilocus deletions, which are larger than one gene but
smaller than a deficiency normally discernible through the light microscope. All
these types of alterations in the genetic material can cause ill health and death.
15
-------
Short-Term Tests for Genetic Toxicity
Traditionally, carcinogens have been identified from epidemiologic studies in hu-
mans and by cancer tests in mice and rats. A major difficulty in identifying carcinogens
in epidemiologic studies in humans is the long latent period (usually 20-30 years)
between first exposure of the person to a carcinogen and diagnosis of cancer. This is
demonstrated dramatically for the association between cigarette smoking and lung
cancer (Fig. 1). Although the incidence of cigarette smoking in males increased
steadily from 1900 to about 1940, the incidence of lung cancer did not show a similar
increase until 25 to 30 years later. The same latent period can be seen emerging in the
curves for cigarette smoking and for lung cancer in females. Epidemiologic studies
also often are limited by inadequate sample sizes and incomplete information on the
subjects. Even when these studies identify a chemical as carcinogenic, a portion of
the human population, which can be very large if the chemical is distributed widely,
will have been exposed to the chemical before it is identified as a carcinogen. The
other traditional method of identifying carcinogens is in laboratory tests with ani-
mals, but these tests unfortunately also have serious limitations: they are expensive
(at least 1/4 million dollars per chemical), lengthy (about three years), and are rel-
atively insensitive (the low number of animals used limits the detection of weak car-
cinogens). Therefore, although both human epidemiologic studies and cancer tests in
laboratory animals will continue to be useful for detecting certain carcinogens, they
cannot meet the urgent need to test a large number of chemicals. Short-term tests for
genetic toxicity are needed as indicators of mutagenicity and as predictors of carci-
nogenicity. These tests, as indicated earlier, are considered valid because mutation in
somatic cells is thought to be the initiating event in certain somatic cell diseases such
as cancer and because mutation in germ cells could cause an increase in genetic dis-
eases in future generations.
Short-term tests for genetic toxicity are to various degrees relatively inexpensive,
rapid, simple, and sensitive. There are some problems in interpretation and of ex-
trapolation to humans, as we will discuss, but the great advantage of these tests is
that they permit us to test chemicals and chemical mixtures before they are released
into our environment.
MAJOR ADVANCES IN THE DEVELOPMENT AND USE OF
SHORT-TERM TESTS FOR GENETIC TOXICITY
Further Development Since About 1970
Three main problems in the development and use of short-term tests for genetic
toxicity have been the need to develop: (1) a capability to test chemical metabolites
as well as the chemical itself; (2) strains that are highly sensitive to the mutagenic ac-
tivity of various chemicals; and (3) strains and tests that can detect a broad range of
mutational alterations.
Metabolic Activation—
A number of experiments were undertaken in the late 1960s and early 1970s to
make the microbial tests more "animal-like." Indicator cells (usually bacteria), the
chemical being tested, and an enzymatic activating system from an animal usually
are combined. One of the first approaches to testing the mammalian metabolites of
chemicals for genotoxic activity was an in vivo test, the host-mediated assay (Legator
and Mailing, 1971). Indicator cells and the chemical are injected into the host, usually
mice; as a control, cells only are injected into other mice. If the chemical is a promut-
agen, it may be metabolized by the host's enzymes to an ultimate mutagen, which
may enter the indicator cells. After specified intervals, the animals are killed, and the
indicator cells are recovered, washed, and tested appropriately for percent survival
and mutation frequency.
16
-------
5,000
1920
1940
1960
1980
Figure 1. The long latent period of cancer illustrated by the association between
cigarette smoking and lung cancer (after Cairns, 1975).
Metabolic activation of chemicals also can be tested for in vitro, in which case the
indicator cells, the chemical, and an activating system are combined in a test tube or
on a petri plate. Appropriate controls are included, and mutation is scored in the in-
dicator cells. Mailing (1966) showed that dimethylnitrosamine is activated chemi-
cally in vitro by a hydroxylating mixture to a mutagenic form. Mailing (1971) later
showed that the same chemical is activated enzymatically in vitro by a microsomal
fraction from an homogenate of mouse liver. The supernatant from a 9,000 x g
centrifugation of the liver homogenate contains the microsomes, which are especially
rich in certain of the activating enzymes. This supernatant is referred to as the S-9
fraction, and, in combination with certain cofactors, as the S-9 mix; currently, it is
the most common method of in vitro metabolic activation. This procedure is rela-
tively rapid and inexpensive, but has the limitation that the entire in vivo metabolism
of an organ or organism is not being tested.
Another important development has been the use of certain chemicals or mixtures
of chemicals to induce (stimulate increased synthesis of) the activating enzymes. The
most commonly used inducer is Aroclor-1254, a mixture of polychlorinated bi-
phenyls. It is injected into rodents, and a few days later the S-9 fraction is prepared
from the livers or other organs of these animals. The induced S-9 fractions have a
higher concentration of certain activating enzymes, especially cytochrome P-450,
than uninduced S-9 fractions. This procedure has greatly increased the utility of the
microbial (especially the bacterial) short-term tests for genetic toxicity.
Aflatoxins are examples of the many chemicals that require metabolic activation
for mutagenicity. Aflatoxins are synthesized by certain strains of the fungus Aspergillus
flavus as they grow on foodstuffs, such as peanuts. Aflatoxin Bi, which is a potent
liver carcinogen in a number of animals, is not mutagenic in bacteria in the absence
of an S-9 fraction but becomes so in the presence of an S-9 fraction from rodent liver.
Certain classes of chemicals, such as polycyclic aromatic hydrocarbons, require
17
-------
metabolic activation. An example is benzo(a)pyrene, which is found in cigarette
smoke, automobile exhaust, and many industrial byproducts, and is, therefore,
ubiquitous in air.
As one progresses along the evolutionary scale from simple organisms to higher
plants and animals, there is an increase in the amount and diversity of the activating
enzymes, which appears to reach its zenith in the liver of mammals. The common
tester strains of bacteria can activate relatively few chemicals. Therefore, if bacteria
are used in a short-term test for genetic toxicity, some kind of metabolic activating
system, such as an S-9 fraction from rat liver, usually is included in the test. Fungi,
such as Neurospora and Saccharomyces, can activate certain chemicals, for ex-
ample, aflatoxin BI and dimethylnitrosamine; nevertheless, mammalian enzymes
are necessary for the activation of other chemicals. Drosophila, on the other hand,
has a rather rich repertoire of activating enzymes. Higher plants, such as maize, have
the ability to activate certain chemicals; for example, some of the pesticides used on
cultivated plants appear to be activated by them to ultimate mutagens. If chemicals
are being tested in a mammalian mutation test, activating enzymes are, of course,
already present. An exception to this generalization is mammalian cell lines, which
have a very limited ability to activate promutagens; therefore, when such cells are
used as indicators, an S-9 fraction is added. The inclusion of methods for metabolic
activation into short-term tests for genetic toxicity has been a very important ad-
vance in their development and use.
Sensitive Strains—
The second problem in the development and use of short-term tests for genetic tox-
icity has been the need to develop sensitive strains of the indicator organism. "Sensi-
tive" means that the strains are strongly mutagenized by low doses of a wide range of
classes of chemicals. The bacterial reverse-mutation test using histidine-requiring
(his) mutants of Salmonella typhimurium is an example of the development of high-
ly sensitive strains. Hundreds of his mutants were tested for their ability to be revert-
ed by various chemical mutagens by Ames, Hartman, and coworkers (Ames, 1971;
Hartmanetal., 1971). Four mutants were selected as testers based on their sensitivity
to specific mutagens (Ames, 1972). All four mutants detect intragenic mutations:
three detect frameshift mutagens of different specificities, and one detects mutagens
that induce base-pair substitutions. Walker (1977) estimates that these mutants
would detect only about 20 percent of 175 known carcinogens.
Ames and coworkers took additional steps to increase the sensitivity of these mu-
tants. Some kinds of lesions induced by mutagens in DN A are repaired by a number
of different enzymatic repair pathways. Earlier work had indicated that microbial
mutants deficient in one of the repair pathways, a nucleotide excision-repair path-
way, are more sensitive than repair-proficient strains to the killing and mutagenic ac-
tivities of a number of mutagens. This repair pathway is so named because the lesion
and some adjacent nucleotides in one of the polynucleotide chains of the DNA mole-
cule are recognized and excised. The gap created by excision is filled in (DNA repair
synthesis) with nucleotides complementary to those in the intact polynucleotide
chain. This repair pathway is relatively error (mutation) free. Cells defective in nu-
cleotide excision repair, such as uvrB mutants of Salmonella, must use other path-
ways, at least one of which is error prone. Thus, uvrB strains are more sensitive than
uvrB* strains to the mutagenicity of chemicals that cause such lesions. Ames and co-
workers incorporated the uvrB mutation into the four Salmonella his testers in order
to increase the sensitivity of the strains (Ames, 1972). The next step in sensitization
was to increase the ability of mutagens that are large molecules to penetrate through
the outer layer that surrounds the cell membrane. In normal Salmonella, the cell
membrane is surrounded by a lipopolysaccharide (LPS)-containing layer, which
acts as a partial barrier to the penetration of certain molecules-generally, the larger
the molecule, the greater the exclusion. Certain mutations cause a defective LPS
layer. One such mutation, rfa, was incorporated into the four his, uvrB testers, re-
18
-------
suiting in another increase in sensitivity (Ames et al., 1973). The inclusion of these
two features (uvrB and rfa) is estimated to permit detection of about 65 percent of
175 known carcinogens (Walker, 1977).
A final step in the sensitization of the Salmonella testers raised the level of detec-
tion of 175 known carcinogens to its current 90 percent (McCann and Ames, 1976).
This step was the incorporation of an R plasmid into the bacterial cells. Plasmids are
small circular molecules of DN A that replicate within bacterial cells, and the R plas-
mids are so named because they carry antibiotic resistance genes. Earlier research
had shown that a his strain containing the plasmid was more sensitive than a strain
not containing the plasmid to the mutagenicity of certain chemicals. McCann et al.,
(1975b) showed that two of their his, uvrB, rfa mutants that contained the R plasmid
pKMlOl were very sensitive to many carcinogens that the nonplasmid-bearing
strains did not detect or only detected weakly. One of these plasmid-bearing strains
is a base-pair substitution mutant (TA100) and one is a frameshift mutant (TA98).
Today, these two are the most widely used strains for screening potentially mutagen-
ic chemicals.
Detection of Many Mutational Alterations—
The third problem has been the need to develop strains and tests that can detect a
broad range of mutational alterations, especially the four types discussed earlier:
changes in the number of chromosomes (aneuploidy); changes in the architecture of
chromosomes (chromosomal aberrations); changes within a gene (gene mutations);
and deletion of two or more contiguous loci (multilocus deletions). The development
of tests to detect mutations has been evolving since 1927, and, in fact, most of the
tests listed in Table 1 were developed prior to 1970 during studies on the complex
processes of many types of mutations in a variety of organisms. As concern grew
about the possible hazard that environmental chemicals might pose to the genetic
material, it was realized that many short-term tests for genetic toxicity already were
available. Some could be used as they were, but others needed refinement and
development for the purpose of testing large numbers of chemicals or for detecting
certain types of mutational alterations. In addition, some new tests were developed.
The extensive use of short-term tests for genetic toxicity in the last ten years is a good
example of how basic research can be applied to a major problem.
Major Short-Term Tests for Genetic Toxicity
The number and diversity of short-term tests for genetic toxicity are great. There is
no "best" test, because each has its strengths and limitations. Our approach here will
be to discuss the tests that the Environmental Protection Agency selected for system-
atic review and evaluation during Phase I of the GENEtic TOXicology (Gene-Tox)
Program (Green and Auletta, 1980; United States Environmental Protection Agen-
cy, 1980; Waters and Auletta, 1981). Not included in this discussion is the cell trans-
formation test, because it is reviewed fully elsewhere in this Volume by Upton. In
Table 1, we have organized the tests generally in order of test organism, from least to
greatest complexity, which reflects their degree of evolutionary distance from
humans. There are some exceptions, because we have attempted simultaneously to
list them roughly in the order that they might be used to detect different types of mu-
tational alterations.
Although the list in Table 1 is extensive, it is deceptively simple, because more
than one test is included in many of the single entries. For example, in the bacterial
DNA repair tests (No. 3), 55 pairs of repair-proficient and repair-deficient strains
among three bacterial species have been used. In the case of the four fungal species
(test No. 7 to 9), about six different tests can be done with each, whereas among high-
er plants (test No. 10) numerous tests are performed using seven major species. Be-
cause the tests were developed in part to detect a broad range of mutational altera-
tions, we have shown in Table 1 the types of mutational alterations (genetic end
points) detected by each. The genetic end points are grouped into five major types:
19
-------
Table 1. An Outline of Twenty-Four Short-Term Tests for Genetic Toxicity Selected for Systematic Review and Evaluation
Phase I of the EPA's GENE-TOX Program
m
Outline and Brief Descriptive Test
Name of the Tests ID No.
DNA
Dam.
Genetic End Points
Gene Mut.
Rev.
For.
MLD
CA
AP
Test
ID No.
Comments About the Tests or
Genetic End Points
. Prokaryote (bacterial) tests
A. Reverse-mutation tests
1. Escherichia coli VJPz and
WP2 uvrA strains
2. Salmonella typhimurium his
strains
B. DNA repair tests
II. Mammalian cells in culture;
specific-locus tests
A. Chinese hamster ovary
(CHO) cells
B. Chinese hamster lung
(V-79) cells
C. Mouse lymphoma (L5178Y)
cells
. Lower eukaryote (fungal) tests
A. Neurospora crassa tests
7.
One trp allele, usually with uvrA.
One or more his alleles; usually
TA98 and TA100, which are
uvrB, rfa, pKM101.
Differential growth inhibition in
repair-deficient/repair-
proficient pairs of strains.
6-Thioguanine resistance =
hypoxanthine-guanine phos-
phoribosyl transferase (HGPRT)
locus.
5-Br-deoxyuridine ortriFI-
thymidine resistance = thymi-
dine kinase (TK) locus.
6-Thioguanine or ouabain resis-
sistance = HGPRT locus or
NaVK+ ATPase locus.
"Mainly two ad-3 alleles; "mainly
at the ad-3 region; "ad-3 region.
-------
Table 1. Continued
Genetic End Points
Outline and Brief Descriptive
Name of the Tests
Test
ID No.
DNA
Dam.
Gene Mut.
Rev.
For.
MLD
CA
AP
Test
ID No.
Comments About the Tests or
Genetic End Points
B. Aspergillus nidulans tests
C. Saccharomyces cerevisiae 9
and Schizosaccharomyces
pombe
IV. Higher plants: genetic and 10
cytogenetic tests
V. Insect (Drosophila melano-
gaster) tests
A. Chromosome abnormality 11
tests
B. Sex-linked recessive lethal 12
test
VI. Mammalian tests
A. Whole animal (in vivo) tests
1. Mouse spot test 13
2. Specific-locus test in mice 14
3. Micronucleus test in bone 15
marrow erythrocytes
4. Dominant lethal test 16
5. Heritable translocation 17
test in mice.
6. Sperm tests 18
Sb Sc
S,Gb S,GC
S,G
Ga
G
S
G
G
G
S 8. "Mitotic crossing over; bone arg
allele; ctwo systems; dbalanced
translocations.
S,G 9, "Mitotic crossing over and mitotic
gene conversion;b>one allele;
°>one locus.
S 10. "Mitotic crossing over; b>one
allele; °>one locus.
11. "Translocations.
12. The last two genetic end points
are not distinguished routinely.
13. Coat color mutations in embryos;
all genetic end points are not
distinguished routinely.
14. Visible mutations at seven loci;
all genetic end points are not
distinguished routinely.
15. The genetic end points are not
distinguished routinely.
16. Early (in utero) deaths.
17. Only translocations scored.
18. Sperm anomalies; genetic end
points not well defined.
-------
Table 1. Continued
Genetic End Points
Outline and Brief
Name of the
Descriptive Test
Tests ID No
DNA
Dam.
Gene Mut.
Rev.
For.
MLD
CA
AP
Test Comments About the I ests or
ID No. Genetic End Points
B. In vitro and in vivo tests
1. Sister chromatid exchange 1 9
(SCE)test
2. Cytogenetic tests 20
C. Special tests
1. Host-mediated assays 21
2. Tests of body fluids
B. Other DNA repair tests
22
VII. Tests that detect repair of DNA
A. Unscheduled DNA synthesis 23
(UDS) tests
24
S,G
19. Sister chromatid exchanges.
20. Chromosomal aberrations.
21. Appropriate host, indicator
organism, and test; genetic end
points depend on test.
22. Appropriate indicator organism
and test; genetic end points
depend on test.
23. DNA repair replication in vitro,
usually in primary rat hepato-
cytes or human fibroblasts.
24. Various techniques; not
presently useful as tests for
genetic toxicity.
Abbreviations; Dam..damage; Mut., mutations; Rev. reverse; For., forward; MLD, multilocus deletions; CA, chromosomal aberrations; AP, aneuploidy,
B, bacterial cells; S, somatic cells; G, germ cells; SCE, sister chromatid exchange; UDS, unscheduled DNA synthesis.
-------
DNA damage, gene mutations, multilocus deletions, chromosomal aberrations, and
aneuploidy. All but the first have been discussed earlier. DNA damage tests measure
different kinds of end points, none of which fits clearly into the other four types of
genetic end points. We have indicated whether the mutational alterations occur in
somatic or germ cells.
In the limited space here, we can discuss only the main features of each test. Refer-
ences will not be given because many of the tests have been reviewed in Chemical
Mutagens: Principles and Methods for Their Detection (Hollaender, 1971; 1973;
1976; Hollaender and de Serres, 1978; 1982) and because a report on each test will be
published in Mutation Research: Reviews in Genetic Toxicology under the aegis
of the EPA Gene-Tox Program. In Part I of Table 1, we have entered the bacterial
tests: two detect reverse mutation, and one detects DNA damage. We discussed the
reverse-mutation test in Salmonella (No. 2) earlier, when we used it as an example of
how sensitive strains were developed. It is currently the most widely used test for the
detection of genotoxic chemicals. The bacterial tests are relatively simple, rapid,
inexpensive, and sensitive. Fig. 2 illustrates the simplicity of the DNA damage test
using a DNA repair-deficient (pol A\) and a repair-proficient (pol A*) strain of E.
co//.7V-Hydroxyurethane is scored as genotoxic because the chemical causes a larger
zone of growth inhibition in the pol A\ than in the pol A* strain. This differential in-
hibition of growth occurs because damage caused in DNA by this chemical is
repaired less efficiently in the pol A i than in the pol A* strain. When equal zones of
growth inhibition are observed in both strains, the chemical is scored as not geno-
toxic. Therefore, a plus or minus result for a single chemical can be obtained from
only two petri plates in a few days. The effect of an S-9 fraction can be assayed with
an additional two plates.
The in vitro tests using mammalian cells in culture are entered next. The cells can
be handled with experimental procedures similar to those used with bacteria. All
three (No. 4 to 6) detect forward mutation at a specific locus by using resistance to a
drug that is an analog of a normal metabolite. This selection system permits the tests
to be rather rapid, simple, and inexpensive, but not as much so as the bacterial tests.
Mammalian cell culture tests have the advantage that the genes being mutagenized
are in eukaryotic chromosomes. Furthermore, the use of forward mutation rather
than reverse mutation permits detection of a wide spectrum of gene mutations.
The lower eukaryote (fungal) tests (No. 7 to 9) are entered next in Table 1. Because
fungi are microorganisms that are single cells during all or part of their life cycles,
some of the experimental procedures are similar to those used with bacteria. The
fungal tests are difficult to discuss, because for each species about six tests, which
detect different genetic end points, can be used. Like one of the bacterial tests, DNA
damage can be detected in two fungi, but in these cases, the damage is detected as
mitotic crossing over and gene conversion. In Saccharomyces cerevisiae, the test that
detects gene conversion is rapid, inexpensive, simple, and very sensitive. Reverse
mutation can be detected in fungal tests that are very similar to the bacterial reverse-
mutation tests. As in mammalian cell cultures, forward-mutation tests at specific
loci are used in fungi. In addition, the important genetic end points of multilocus de-
letions, chromosomal aberrations, and aneuploidy are detected in appropriate
fungal tests; it is noteworthy that aneuploidy can be detected during meiosis in two
species.
The higher plant tests (No. 10) are especially complex to discuss, because nu-
merous genetic and cytogenetic tests with many genetic end points have been used in
seven major species. One advantage of the higher plant tests is that often the genetic
end points are detected in germ cells. Another advantage is that the mutagenicity of
many pesticides often can be tested for in the plants on which they are used. Further-
more, higher plants appear to have some unique systems for metabolic activation of
certain chemicals.
23
-------
Figure 2. DNA damage test using the bacterium Escherichia coli. Effect of N-
hydroxyurethane on the growth of repair-proficientpoM* (left) and repair-
deficient polAi (right) strains. This chemical is scored as genotoxic because
there is a larger zone of growth inhibition in the pol At strain (from
Rosenkranz and Leifer, 1980).
Tests using Drosophila (No. 11 and 12) are the first whole animal tests that we
have listed in Table 1. They are very useful because there is an extensive literature on
the mutagenicity of chemicals in Drosophila and because the metabolic activation
systems of this organism appear to be similar to those of mammals. Also, a number
of genetic effects, transmitted through the germ cells, can be assayed. The tests are
somewhat more time consuming and expensive than many entered earlier. The sex-
linked recessive lethal test is a forward-mutation test that detects recessive lethalsat
~600 loci on the X chromosome; it is the most sensitive and economical in vivo pro-
cedure that measures mutation in all the germ cell stages of an eukaryote.
Within the mammalian tests, we have entered seven whole mammal (in vivo) tests
(No. 13 to 18), which are usually done in mice. Their general advantage is that the
chemical in question interacts with the mammal's activating and inactivating en-
zymes. Obviously, this situation is somewhat similar to that which humans face in
their environment; therefore, findings from these tests may be highly relevant to con-
siderations of hazards and risks that chemicals pose to humans.
In the mouse spot test (No. 13), embryos heterozygous at several loci determining
coat color are exposed in ulero at nine to ten days of development by injecting the
chemical into the peritoneal cavitity of the mother. Forward mutation of the
dominant allele at one of the heterozygous loci in an embryo produces a cell that is
homozygous recessive for a mutant phenotype. This cell gives rise through division
to a clone of cells, some of which may determine coat color. This mutation will be ex-
pressed as a spot of mutant coat color, such as a tan spot, in an otherwise black coat;
the mutations are scored about one week after birth, that is, about three weeks after
exposure in ulero (Fig. 3). This test detects forward mutation due to a number of
mutational alterations, and is the most rapid in vivo forward-mutation test in somat-
ic cells of mammals.
Although biochemical and immunological markers can be used in the mouse spe-
cific-locus test (No. 14), seven visible markers, mainly affecting coat color are used
most frequently. For example, a brown (bb) female is mated to a black (BB) male
that has been exposed to a test chemical. If a forward mutation from Blob occurs in
a sperm that fertilizes an egg, a zygote that is bb will be formed, which will develop
24
-------
into a brown mouse. Black progeny indicate no mutation. Thus, the incidence of for-
ward mutation at seven loci is scored. As in the sex-linked recessive lethal test in Dro-
sophila, the mutagenicity of a chemical at various stages of spermatogenesis can be
tested by varying the time between exposure of the male to the chemical and mating.
This test is relatively simple and easily scored, but it is debatable whether this test can
be called a short-term test. Nevertheless, the mouse specific-locus test is currently the
only practical and reliable method for determining whether a chemical induces heri-
table forward mutations in mammals. Therefore, it is an important test for estimat-
ing germ cell risk in humans.
Figure 3. The mouse spot test for detection of a forward mutation at one of the
heterozygous loci determining coat color. The pelt shows a typical type of
recessive tan spot over the right hip region (arrow). Photograph courtesy of L.
B Russell.
The micronucleus test (No. 15) has been used in several cell types from animals
and plants, but recently formed red blood cells (polychromatic erythrocytes) of
mammals (usually mice) have been used most often. Daughter chromosomes nor-
mally migrate to opposite poles during cell division and become incorporated into
two daughter nuclei. On the other hand, a chromosome fragment, which arises due
to chromosome breakage, or a whole chromosome that fails to migrate becomes in-
corporated into a structure that stains and looks like a small nucleus—a micronucle-
us (Fig. 4). This test is the least expensive and fastest means available to detect these
kinds of genetic end points in mammalian somatic cells.
In the dominant lethal test (No. 16) in mammals (usually mice or rats), males are
treated with a chemical and then mated to females according to a protocol that per-
mits all stages of spermatogenesis to be tested. The females are killed at about 14
days of pregnancy, and the number of early deaths out of the total number of im-
plants is scored. Early deaths are scored as dominant lethals, which are thought to be
due mainly to multiple chromosome breaks in the germ cells of the treated male.
Although a positive result in this test indicates that a chemical causes chromosome
breakage in germ cells, use of this test is limited by its relative insensitivity and by the
fact that events other than chromosome breakage may contribute to what are scored
as dominant lethals.
-------
The heritable translocation test (No. 17) in mice is similar to the dominant lethal
test in some respects, but in this case the genetic end point, translocations that are in-
herited, is precise. Again, males are treated with the test chemical and then mated tc
females. A stage of spermatogenesis in the male progeny is analyzed microscopically
for translocations, which were induced in germ cells of the male parent and transmit-
ted to a son. Although relatively expensive, difficult, and lengthy, the test clearly
identifies the ability of chemicals to cause one class of heritable chromosomal aber-
rations in mammals.
Figure 4. The detection of chromosomal aberrations by the formation of micronuclei.
A single chromosome is represented; more would be present. A break occurs
in one chromatid of the chromosome in the cell at the left, producing a
fragment. The two daughter chromosomes migrate to opposite ends of the
cell by action of spindle fibers attached to the centromeres, but the fragment
does not move. The daughter chromosomes become surrounded by a
nuclear membrane to become daughter nuclei, and the fragment becomes a
micronucleus. Illustration courtesy of A. S. Raj, A. B. Krepinsky and J. A.
Meddle.
Sperm tests (No. 18) are attractive because they detect spermatoxic agents in
mammals, including humans. The end points are sperm count, sperm motility, and
sperm head morphology. Unfortunately, when sperm count or motility are de-
creased, or when sperm head morphology is altered, we do not know what fraction,
if any, of the effect is due to genotoxicity of the chemical. Therefore, the tests detect
spermatoxic, but not necessarily genotoxic, agents. One test, the fluorescent Y-body
(YFF) assay, has been used occasionally in humans to detect aneuploidy of the Y
chromosome.
Two tests are listed under mammalian tests as "in vitro and in vivo tests," because
they are used on cultured mammalian cells that have been treated with a chemical or
on cells from exposed mammals, including man. This dual approach is very useful in
certain situations. Test No. 19 is the sister chromatid exchange (SCE) test. At met-
aphase, each chromosome is composed of two sister chromatids held together at the
centromere. If the sister chromatids are labelled differentially by appropriate pre-
vious DNA replications in the presence of a nucleotide analog, 5-bromodeoxy-
uridine, the sister chromatids stain differentially with certain dyes. In such cells there
is a spontaneous frequency of SCEs, that is, places at which there has been a clear,
reciprocal interchange between sister chromatids (Fig. 5 A). Certain chemicals, such
as the drug mitomycin-C, increase their frequency (Fig. 5B). The molecular mech-
anism of SCEs is not known, although they must arise from breakage and reunion of
DNA molecules, which presumably occurs, probably indirectly, as a consequence of
DNA damage. Therefore, we classify the SCE test as a DNA damage test. The assay
is relatively simple, rapid, inexpensive, and sensitive.
The cytogenetic tests (No. 20) in mammals also are done in vitro or in vivo. Six
major tests detect chromosomal aberrations due to chromosome breakage as their
genetic end points. Four of these tests are done in vivo and use bone marrow cells,
26
-------
spermatogonia, spermatocytes, or oocytes or early embryos. The fifth test is an in
vitro test in which cultured mammalian cells (No. 4 to 6), mainly Chinese hamster
ovary (CHO) cells, are used. Leukocytes, including those from humans, are used
both in vitro and in vivo.
"•*•. *
Wv» \^ X •***•»
.x^" .,'
"W
**-l'
**S
%-~ .
-------
major steps: hazard identification, hazard evaluation, and risk estimation. As we
proceed through these three steps, we inevitably will focus on the risk that mutagenic
chemicals pose to humans, but each test also should be thought of as a means of
assessing effects on other organisms as well.
Starting with data on a chemical from one or more short-term tests, we can divide
our analysis into the three steps mentioned above (Table 2). In order to analyze the
results from the different tests under one or more of these three headings, we must
consider the genetic end points that each detects, and whether they occur in somatic
or germ cells. Although we have entered most of the tests from Table 1 in Table 2 to
demonstrate how each of them could be used at one or more steps of the analysis,
we do not imply that all of the tests entered under each step would be used for a given
chemical. The choice of which tests to use at each step is influenced by factors such as
the chemical properties of the compound being tested, its distribution in the environ-
ment, and the possible route(s) of human exposure.
In the first step, hazard identification, a chemical is identified as being a genotoxic
hazard if it is positive in one or more tests. Tests used at this step should be relatively
rapid, inexpensive, simple, and sensitive; usually identification comes from the re-
sults in bacterial tests, which efficiently screen large numbers of chemicals by
detecting DNA damage or reverse mutation. The test for gene conversion in S.
cerevisiae also could be used here. The more time-consuming and expensive tests
that detect forward mutation, multilocus deletions, chromosomal aberrations, or
aneuploidy usually are delayed until the next step of the analysis. However, if suit-
able tests that detect multilocus deletions, chromosomal aberrations, or aneuploidy
are developed, they also might be included at this step. Tests for aneuploidy should
be included at the first step when they are thoroughly validated, because these tests
detect those chemicals that interfere with the disjunction of chromosomes through
interactions with proteins. To summarize the step of hazard identification, one
screens many chemicals in simple short-term tests in order to identify rapidly those
rtiat are genotoxic.
In the second step, hazard evaluation, the main goal is to determine whether chem-
icals that are genotoxic in p'rokaryotic tests also are genotoxic in relatively simple
eukaryotic tests. A positive result for DNA damage detected earlier in bacteria can
be tested for in fungal, higher plant, and mammalian DNA damage tests (No. 8 to 10,
19,23). A positive result in reverse-mutation tests in bacteria can be tested for in fun-
gal and higher plant reverse-mutation tests (No. 7 to 10). The test chemical can be
assayed in tests that detect forward mutation in eukaryotes (No. 4 to 10, 12). Other
eukaryotic tests can be used to determine if the chemical has clastogenic activity by
detecting chromosomal aberrations (No. 8, 10, 11). Aneuploidy can be detected in
certain tests in fungi, higher plants, and Drosophila (No. 7 to II). Because multi-
locus deletions often act as lethals, they are not detected in bacterial and many other
tests. Yet, the detection of this type of mutation is very important, because there is
evidence from the analysis of specific-locus mutations in mice that multilocus dele-
tions can have deleterious or lethal effects in heterozygotes. Most of the mutations
detected in reverse-mutation tests (and some forward-mutation tests) are recessive
and will not be expressed in the first generation from a treated diploid parent. On the
other hand, because multilocus deletions often act as dominant or semidominant
mutations, they are expressed in the first generation from a treated diploid parent.
The fraction of multilocus deletions can be determined in the ad-3 forward-mutation
test in Neurospora (No. 7).
From the results of the tests shown under hazard evaluation (Table 2), one verifies
or fails to verify the positive results for genetic toxicity obtained with a chemical in
the first step of hazard identification. Simultaneously, one extends the data on a
given chemical to more kinds of organisms and genetic end points. The results from
tests done on somatic cells of fungi, higher plants, and animals are relevant to evalu-
ating genotoxic hazards to somatic cells. If, however, certain of the tests shown un-
der hazard evaluation are done, information also is obtained on genotoxicity to
28
-------
Table 2. Possible Use of Tests from Table 1 for Hazard Identification, Hazard Evaluation, and Risk Estimation
Genetic End Points
Test DNA Gene Mut.
Test Name
Hazard Identification
Bacterial DNA repair tests
Bacterial reverse-mutation tests
Hazard Evaluation
SCE tests in mammals
UDS tests in mammals
Mammalian cells in culture tests
Fungal tests
Higher plant tests
Sex-linked recessive lethal test in
Drosophila
Chromosome abnormality tests in
Drosophila
Risk Estimation
Mouse spot test
Micronucleus test in mammals
Cytogenetic tests in mammals
Specific-locus test in mice
Heritable translocation test in mice
Dominant lethal test in mammals
ID No. Dam. Rev.
3 B
1,2 B
19 S
23 S
4-6
7-9 S S
10 S S,G
12
11
13 S
15
20
14
17
16
For. MLD CA
S
S S S
S,G S,G
G G
G
S S S
S
S,G
G G G
G
G
AP
S,G
S
G
S
S
Abbreviations: Dam., damage; Mut., mutations; Rev., reverse; For., forward; MLD, multilocus deletions; CA, chromosomal aber-
rations; AP aneuploidy; B, bacterial cells; S, somatic cells; G, germ cells; SCE, sister chromatid exchange; UDS,
unscheduled DNA synthesis.
-------
germ cells and/or about transmissibility (heritability) of mutational alterations in
submammalian species. Under forward-mutation tests, certain of the higher plants
detect mutations in germ cells. The sex-linked recessive lethal test in Drosophila de-
tects forward mutations that are induced in one or more stages of spermatogenesis
and transmitted to the next generation. Similarly for chromosomal aberrations, cer-
tain of the tests detect chromosomal aberrations in germ cells and/ or measure trans-
mission of the aberration through germ cells from one generation to the next. Be-
cause some of these tests are whole plant or animal tests, the chemical is subjected to
the organism's metabolism. If a chemical is positive in some of the tests shown under
hazard evaluation, it should be tested further using the tests shown under risk esti-
mation.
In the third step, risk estimation, tests are used that will provide data most relevant
to estimating the risk a chemical poses to human somatic or germ cells. Therefore,
the tests entered at this step in Table 2 are mammalian tests. The genotoxic potency
of a chemical for the different types of genetic end points is important. The mouse
spot test, the micronucleus test, and certain cytogenetic tests (No. 13, 15, 20) can be
used for estimating somatic cell risk. The specific locus test and the heritable trans-
location test in mice (No. 14, 17) and the dominant lethal test and certain cytogenet-
ic tests (No. 16, 20) can be used for estimating the risk of a given chemical for the in-
dicated genetic end points in germ cells.
The three-step approach shown in Table 2 is an example of what is sometimes re-
ferred to as a tier approach for assaying the genotoxicity of chemicals. As we indi-
cated earlier, all of the tests normally would not be used for a given chemical. If, how-
ever, a genotoxic chemical is known to be distributed widely in the environment or if
a large human population is known to be exposed, many tests, including those that
detect all of the genetic end points, should be included. By considering the potency of
a chemical for these end points as well as information such as the routes of human
exposure, the risk that the chemical poses for certain somatic cell and genetic
diseases may be estimated. For example, if a chemical is a potent mutagen in somatic
cell tests such as the mammalian spot test, but a very weak mutagen in germ cell tests
such as the mouse specific-locus test, one would estimate that the chemical poses a
greater risk for cancer than for genetic diseases. In another case, a chemical may
cause aneuploidy in all of the appropriate tests, but be negative in other tests; one
would estimate, therefore, that the chemical poses a risk for spontaneous abortion
and aneuploidy diseases such as Down's syndrome.
Large-Scale Programs to Evaluate the Efficacy of Selected Short-Term
Tests for Genetic Toxicity
Studies to evaluate the efficacy of certain short-term tests for genetic toxicity were
initiated independently in the United States and Japan in 1972. In both studies, the
Salmonella/ microsome test was used, and many of the test chemicals were known to
be prevalent in the environment and/ or were known to be carcinogens or noncarcin-
ogens from human epidemiologic studies or rodent cancer tests. In the United States
study, sponsored by the National Cancer Institute, 102 chemicals were tested (Poirier
and de Serres, 1979), and, in the Japanese survey, sponsored by the National Insti-
tute of Hygienic Sciences, 60 chemicals were assayed (Sugimura et al., 1976). After
. 1972, these studies were merged by mutual agreement under the U.S.-Jap^n Cooper-
ative Medical Sciences Program. In 1975 and 1976, McCann, Ames, and coworkers
published a tabulation of the results on almost 300 chemicals that had been assayed
in the Salmonella/microsome test (McCann et al., 1975a; McCann and Ames,
1976). The chemicals were selected because they could be classified as carcinogens or
noncarcinogens based on previous data from human epidemiologic studies or rodent
cancer tests.
In 1977, Imperial Chemical Industries, Limited, in Great Britain sponsored the
study of 120 coded chemicals in six short-term tests (Purchase et al., 1976; 1978),
and, in 1979, a large international program evaluated 42 chemicals in 30 short-term
tests (de Serres and Ashby, 1981). Two important features of the latter study are that
30
-------
14 pairs of structurally related carcinogens and noncarcinogens were tested, and that
the bacterial tests were done in 15 laboratories, which permitted an interlaboratory
comparison of the results.
Two major conclusions emerged from these studies. The first is that there is a
strong positive association between the carcinogenicity and the1 mutagenicity of
chemicals. We mean by this that a large majority of the known carcinogens are muta-
genic and that a large majority of the known noncarcinogens are not mutagenic. The
largest number of chemicals where this association can be noted is the almost 300
chemicals tabulated by McCann and Ames (1976): 90 percent (157/175) of the car-
cinogens are mutagenic in the Salmonella/ microspme test, and 87 percent (94/108)
of the noncarcinogens are not mutagenic in the same bacterial test. This strong pos-
itive association and other evidences provide the scientific basis for the concept that
short-term tests for genetic toxicity can be used to screen large numbers of chemicals
not only for genetic toxicity, but also with a high probability for potential carcino-
genicity. McCann and Ames (1976) also found that none of 46 common biochemi-
cals is mutagenic. The second major conclusion, which became evident from the
studies where more than one test was included, is that no single test detects all car-
cinogens as mutagens. Therefore, a battery of short-term tests is required for a com-
prehensive screening of chemicals for genetic toxicity. The battery should include
tests that detect all classes of mutagenic chemicals and the major kinds of genetic end
points.
Complex Mixtures of Chemicals and Environmental Monitoring
We have focused mainly on the usefulness of short-term tests to detect the geno-
toxicity of single chemicals, often those that are known to be present in the
environment and/or that are known to be carcinogens. It is rare, however, that an
organism in its environment is exposed to a single chemical; rather, they are exposed
to complex mixtures of chemicals. Complex mixtures can be obtained from a point
source or from the environment. Obviously, these two kinds of souces overlap. Ex-
amples of point-source complex mixtures are cigarette smoke condensates, hair
dyes, cosmetic formulations, charred meat, soot, and automobile emissions. Ex-
amples from our environment are particulates isolated from air samples and mix-
tures of organic chemicals isolated from drinking water, rivers, and lakes. These
chemical mixtures can be collected and then tested in the laboratory, or,
alternatively, test organisms can be placed in the environment. The mutagenicity
testing of complex mixtures and environmental monitoring are relatively new,
rapidly growing, difficult, and important areas of research.
A complex mixture of chemicals can be subjected to short-term tests for genetic
toxicity using the same tier approach as described above and outlined in Table 2. An
example is cigarette smoke condensate, which is the particulate fraction of cigarette
smoke containing over 2,000 chemicals and commonly referred to as tar. It is
mutagenic in a number of short-term tests (DeMarini, 1981). Cigarette smoke con-
densate has been fractionated in order to identify the mutagenic chemicals and to
remove competing toxic, but nonmutagenic, chemicals. When testing for air pollu-
tion, samples of air are drawn through a filter, on which the particulates are collected
and concentrated. The particulates are dissolved in a suitable chemical, such as
dimethyl sulfoxide, and tested for genotoxicity. The complex mixtures of chemicals
in air particulates from certain heavily polluted urban air samples are clearly
mutagenic (Hughes et al., 1980). The organic chemicals in samples of drinking wa-
ter can be concentrated by passing the water through a resin in a glass column; the
organic chemicals then are eluted with a small amount of organic solvent. Certain of
these also have been shown to be mutagenic (Loper, 1980).
The higher plant tests listed in Tables 1 and 2 are promising tools for in siiu envi-
ronmental monitoring, in which the test organism is placed in the environment or an
organism is used that is already present there. For example, maize seeds can be
planted in plots where various pesticides are applied to the soil (Plewa and Wagner,
31
-------
1981). The maize seeds are of special varieties in which reverse or forward mutation
can be scored in the pollen. Certain pesticides are mutagenic in these tests. Sparrow
pioneered the use of the higher plant Tradescantia as an in situ environmental mon-
itor (Sparrow and Schairer, 1971). Somatic mutations are scored in cells of the
stamen hairs in flowers of plants heterozygous for a pair of alleles determining flow-
er color. The heterozygote (Bb) has blue stamen hairs; mutation of the B allele to the
b allele produces a pink cell in the chain of 25-30 cells of a stamen hair (Fig. 6). This
short-term test is being used in a mobile laboratory to monitor air samples from
different locales (Schairer et al., 1978). The additional genetic end points of micro-
nucleic and chromosomal aberrations can be scored in appropriate tests in this
versatile plant. Another kind of in situ environmental test utilizes an indigenous
organism; Klekowski (1978) has used the royal fern Osmunda regalis as such a mon-
itor. This is a promising area of research, but much work still needs to be done.
Short-Term Tests in Human Population Monitoring
Human population monitoring is a natural extension from environmental moni-
toring. We noted that in indigenous in situ environmental monitoring an organism
that is a normal resident of a particular environment is used to monitor the mutagen-
icity of that environment. Humans can be used as indigenous in situ environmental
monitors. We have developed the concept throughout this chapter that the data
from short-term tests for genetic toxicity can be used to estimate genotoxic hazards
and risks to various organisms, including humans. Eventually, however, if we desire
to estimate risk to humans, we would like to use the human species as the test orga-
nism, either indirectly or directly, in human population monitoring. How do we
monitor human populations? We will mention two main approaches that utilize
short-term tests: the use of human body fluids and human cells. In general, a specific
population known or suspected of being exposed to genotoxic chemicals is com-
pared to a control (unexposed) population, selected to be as similar as possible to the
population under study. Often these populations are in a particular occupation.
A good example of the use of body fluids in human population monitoringcomes
from the work of Yamasaki and Ames (1977). They have shown that the urine of cig-
arette smokers is clearly mutagenic compared to the urine of nonsmokers, as assayed
in the Salmonella/microtome test. Epidemiologic studies have shown that cigarette
smokers have a higher incidence of bladder cancer than nonsmokers. There are dis-
advantages of using urine in human population monitoring, but many samples are
relatively easily obtained and tested. Another approach to human population moni-
toring is to use human cells. Blood is collected from individuals of the population un-
der study and from a control population, and lymphocytes are cultured in the labo-
ratory. Two important genetic end points that can be scored in these cells are SCEs
and chromosomal aberrations.
One interesting aspect of human population monitoring is that there are individu-
als whose genotypes make them highly sensitive to mutagenic agents in their
environment. These individuals have a high risk of cancer. The best-understood ex-
ample of this is the disease xeroderma pigmentosum, which is due to an autosomal
recessive mutant allele. Affected individuals are highly sensitive to sunlight, develop
many skin cancers, and usually die of cancer before reproductive age. Persons with
xeroderma pigmentosum are deficient in a nucleotide excision-repair pathway,
which normally removes pyrimidine dimers induced in DNA by the UV in sunlight
(McCormick and Maher, 1980). These persons are the human counterparts of the
uvrB mutants of Salmonella, which we discussed earlier. Other diseases are known
in which the risk of cancer is increased. Because the human population is genetically
highly heterogeneous, there are probably other undiscovered cancer-prone geno-
types. One use of human population monitoring could be to identify mutation-
prone individuals, who could be advised to adapt their lifestyles so as to minimize
their exposure to environmental mutagens.
32
-------
L-LU-1.1 I I I I 1
0 5 10cm
(a)
(c)
Figure 6. The use of the higher plant Tradescantia as an /ns/ru environmental monitor.
(a) Normal stock plant of Tradescantia clone 4430 showing several mature
inflorescences;
(b) Enlarged view of a single inflorescence showing the range in bud size
from the lowest buds containing meiotic stages to mature flower;
(c) Enlargement of stamen hairs with pink mutant events indicated by
shading. Mutations in the flower color locus are not usually lethal. The
chains of pink cells represent daughter cells of the initial mutated cell
(from Schairer et a/., 1979).
33
-------
Summary Comments on Short-Term Tests for Genetic Toxicity
One view of the use of short-term tests for genetic toxicity is that they are an im-
portant tool of preventive medicine. Through careful analysis of the data obtained,
one can identify genotoxic chemicals and chemical mixtures before they are put into
general use and released into the environment. For the chemicals already in our envi-
ronment, the tests can identify those that are genotoxic so that measures can be
taken to decrease their occurrence. Furthermore, for the chemicals that provide
great benefits to society, but which also are genotoxic, closely related structural
analogs can be synthesized that may retain the desired chemical property without be-
ing genotoxic. By striving to keep the number and concentration of genotoxic chem-
icals in our environment to the lowest levels possible, we hope to maintain the health-
iest environment possible.
FEDERAL LEGISLATION AND ACTIONS
The Federal Government has recognized the threat that toxic substances pose to
the environment and to human health, and has responded by passing legislation that
governs the production, testing, formulation, use, and disposal of toxic substances.
The most comprehensive act concerning toxic chemicals is the Toxic Substances
Control Act of 1976, which authorized the Administrator of the EPA to protect the
environment and the health of humans from harmful chemicals and chemical mix-
tures. The scope of this act was summarized by the EPA on October 7, 1976, as
follows:
The Toxic Substances Control Act authorizes the EPA to obtain from in-
dustry data on the production, use, health effects, and other matters con-
cerning chemical substances and mixtures. If warranted, EPA may regulate
the manufacture, processing, distribution in commerce, use, and disposal
of a chemical substance or mixture. Pesticides, tobacco, nuclear material,
firearms and ammunition, food, food additives, drugs, and cosmetics are
exempted from the Act. These products are currently regulated under oth-
er laws.
In this same document, EPA comments about the testing of chemicals:
The Administrator of EPA may require manufacturers or processors of
potentially harmful chemicals to conduct tests on the chemicals. Testing
may be directed to evaluating the characteristics of a chemical such as per-
sistence or acute toxicity, or to clarifying its health and environmental ef-
fects, including carcinogenic, mutagenic, behavioral, and synergistic ef-
fects. The manufacturers or processors of a chemical must bear costs of
testing that chemical.
The Toxic Substances Control Act is very significant because it requires manufac-
turers and processors of chemicals and chemical mixtures to test them before they
are produced commercially. The emphasis of the Act is clearly on preventing toxic
substances from entering our environment. The EPA recognized that an integral
part of implementation of the Toxic Substances Control Act would be short-term
tests for toxicity of the chemicals, and that short-term tests for genetic toxicity would
play an important role. Short-term tests for genetic toxicity have been used for over
50 years; therefore, a large and diverse scientific literature exists on the mutagenicity
of chemicals. There was a clear need for a mechanism to collect, organize, analyze,
and interpret all of these existing data, and to evaluate the relative merits of the many
different short-term tests. The Office of Toxic Substances of the EPA established
Phase I of the "Gene-Tox Program" in 1979: an evaluation of the current status of
bioassays in genetic toxicology (Green and Auletta, 1980; United States Environ-
mental Protection Agency, 1980; Waters and Auletta, 1981). The evaluations were
presented at a meeting in Washington, D.C. in December 1980, and are being pub-
lished in the open literature (for example, Brusick et al., 1980; Generoso et al 1980)
34
-------
An Assessment Panel has appointed committees, which will evaluate the efficacy
of the various short-term tests to detect the major kinds of alterations in the genetic
material, mutagenic chemicals in various chemical classes, and somatic and germ
cell mutations. Based on their conclusions, certain short-term tests will be selected as
the "best" combination to use in the future to detect mutagens and potential carcino-
gens. The chemicals identified as genotoxic most likely will be subjected to short-
term tests for carcinogenicity, and the chemicals positive in short-term tests for car-
cinogenicity most likely will be validated in rodent cancer tests. From the informa-
tion obtained from these tests, along with other information, the EPA will evaluate
the potential hazard and risk that a particular chemical poses to the environment
and the organisms in it.
THE FUTURE OF GENETIC TOXICOLOGY
We expect that the existing short-term tests for genetic toxicity will continue to be
improved and that new tests will be devised. These new and improved tests will be in-
creasingly validated for various classes of chemicals that are known carcinogens or
noncarcinogens, and that are known mutagens or nonmutagens in germ cells of
mammals. In the last decade, the question of the mutagenicity of known carcinogens
and noncarcinogens has been considered fairly extensively; we anticipate a major
emphasis on study of the reverse association. As chemicals not yet tested for carcino-
genicity are identified as mutagens or nonmutagens in short-term tests, they will be
studied for carcinogenicity.
Most of our exposure to mutagenic chemicals is not to pure compounds. Rather,
we are exposed to highly complex mixtures of chemicals, as for example, in air, wa-
ter, and food. Therefore, we expect that there will be a greater emphasis on the study
of complex chemical mixtures, environmental monitoring, and human population
monitoring. Because of the large number of samples, environments, and populations
that need to be surveyed, short-term tests for genetic toxicity are expected to have a
major role. Methods for environmental and human population monitoring are in an
early stage of development, and we expect an explosion of new techniques and find-
ings in these areas. Because the human population is highly heterogeneous geneti-
cally, we anticipate an emphasis on studies that seek to identify the genetic factors
that determine the relative sensitivity of individuals to environmental mutagens.
Recent discoveries on the molecular biology of the genetic material—such as the
organization of nucleotide sequences in eukaryotic DNA, the organization of DNA
and proteins into nucleosome structures, the presence of intervening sequences in
eukaryotic genes, and the mechanism of peripatetic genes, which move from place to
place in the genome—must be considered in the design of experiments and in the in-
terpretation of the results from short-term tests for genetic toxicity. The knowledge
from basic research in many fields of science will continue to have major influences
on the field of genetic toxicology as short-term tests become increasingly important
in the evaluation of the risk that toxic substances pose to our environment and to our
health in this and future generations.
REFERENCES
Ames, B.N. The detection of chemical mutagens with enteric bacteria. In Chemical Mutagens.
Principles and Methods for Their Detection, Vol. 1, edited by A. Hollaender, Plenum
Press, New York, pp. 267-282, 1971.
Ames, B.N. A bacterial system for detecting mutagens and carcinogens. In Mutagenic Effects
of Environmental Contaminants, edited by H.E. Sutton and M.I. Harris, Academic Press,
New York, pp. 57-66, 1972.
Ames, B.N., F.D. Lee, and W.E. Durston. An improved bacterial test system forthe detection
and classification of mutagens and carcinogens. Proc. Natl. Acad. Sci. U.S.A. 70:782-
786, 1973.
35
-------
Auerbach, C, and J.M. Robson. Chemical production of mutations. Nature (London) 157:
302, 1946.
Avery, O.T., C.M. MacLeod, and M. McCarty. Studies on the chemical nature of the sub-
stance inducing transformation of pneumococcal types. Induction of transformation by a
desoxyribonucleic acid fraction isolated from Pneumococcus type III, J. Exp. Med. 79:137-
158, 1944.
Beadle, G. W., and E.L. Tatum. Genetic control of biochemical reactions in Neurospora. Proc.
Nail. Acad. Sci. U.S.A. 27:499-506, 1941.
Boyce, R.P., and P. Howard-Flanders. Release of ultraviolet light-induced thymine dimers
from DNA in E. co/i K.-12. Proc. Natl. Sci. U.S.A. 5/-.293-300, 1964.
Brusick, D.J., V.F. Simmon, H.S. Rosenkranz, V.A. Ray, and R.S. Stafford. An evaluation of
the Escherichia coi WP2 and WP2urv/4 reverse mutation assay. Mutation Res. 76:169-190,
1980.
Cairns, C. The cancer problem. Sci. Amer. 233:64-78, 1975.
DeMarini, D.M. Mutagenicity of cigarette smoke condensate in Neurospora crassa. Mutation
Res. 55:375-388, 1981.
de Serres, F.J., and J. Ashby (eds.). Evaluation of Short-Term Tests for Carcinogens: Report
of the International Collaborative Program. Elsevier/North-Holland, New York, 1981.
de Vries, H. Die Mutationslheorie (The Mutation Theory). English translation published by
Kegan Paul, Trench, Trubner and Co., London, Vol. 1, 1910; Vol. 2, 1911.
Generoso, W.M., J.B. Bishop, D.G. Gosslee, G.W. Newell, C.J. Sheu, and E. von Halle. Heri-
table translocation test in mice. Mutation Res. 7(5:191-215, 1980.
Green, S., and A. Auletta. Editorial introduction to the reports of "The GENE-TOX Pro-
gram": an evaluation of bioassays in genetic toxicology. Mutation Res. 76:165-168, 1980.
Hartman, P.E., Z. Hartman, and R.C. Stahl. Classification and mapping of spontaneous and
induced mutations in the histidine operon of Salmonella. Adv. Genet. 16:1-34, 1971.
Hollaender, A. (ed.). Chemical Mutagens. Principles and Methods for Their Detection.
Plenum Press, New York, Vol. 1, Vol. 2, 1971; Vol. 3, 1973; Vol. 4, 1976.
Hollaender, A., and F.J. de Serres (eds.). Chemical Mutagens. Principles and Methods for
Their Detection. Plenum Press, New York, Vol. 5, 1978; Vol. 6, 1980; Vol. 7, 1982.
Hughes, T. J., E. Pellizzari, L. Little, C. Sparacino, and A. Kolber. Ambient air pollutants: col-
lection, chemical characterization and mutagenicity testing. Mutation Res. 7(5:51-83, 1980.
Jensen, K.A., I. Kirk, G. Kp'mark, and M. Westergaard. Chemically induced mutations in
Neurospora. Cold Spring Harbor Symp. Quant. Bio/. 7(5:245-261, 1951.
Klekowski, E. J., Jr. Detection of mutational damage in fern populations: an in situ bioassay
for mutagens in aquatic ecosystems. In Chemical Mutagens. Principles and Methods for
Their Detection, Vol. 5, edited by A. Hollaender, and F.J. de Serres, Plenum Press, New
York, pp. 79-99, 1978.
Legator, M.S., and H.V. Mailing. The host-mediated assay, a practical procedure for evalua-
ting potential mutagenic agents in mammals. In Chemical Mutagens. Principles and Meth-
ods for Their Detection. Vol. 2, edited by A. Hollaender, Plenum Press, New York, pp. 569-
589, 1971.
Loper, J.C. Mutagenic effects of organic compounds in drinking water. Mutation Res. 7(5:241-
268, 1980.
Mailing, H.V. Mutagenicity of two potent carcinogens, dimethylnitrosamineanddiethylnitro-
samine, in Neurospora crassa. Mutation Res. 3:537-540, 1966.
Mailing, H.V. Dimethylnitrosamine: formation of mutagenic compounds by interaction with
mouse liver microsomes. Mutation Res. 73:425-429, 1971.
McCann, J., and B.N. Ames. Detection of carcinogens as mutagens in the Salmonella/micro-
some test: assay of 300 chemicals: discussion. Proc. Nail. Acad. Sci. U.S.A. 73:950-954,
1976.
McCann, J., E. Choi, E. Yamasaki, and B.N. Ames. Detection of carcinogens as mutagens in
the Salmonella I microtome test: assay of 300 chemicals. Proc. Natl. Acad. Sci. U.S.A.
72:5135-5139, 1975a.
McCann, J., N.E. Springarn, J. Kobori, and B.N. Ames. Detection of carcinogens as muta-
gens: bacterial tester strains with R factor plasmids. Proc. Natl. Acad. Sci. U.S.A. 72:979-
McCormick, J.J., and J. M. Maher. DNA repair processes can alter the frequency of mutations
in diplo.d human cells. In DNA Repair and Mumgenesis in Eukarvotes. edited by W M.
Generoso, M.D. Shelby, and F.J. de Serres, Plenum Press, New York pp 315-321 1980
36
-------
Mendel, G. Versuch iiber Pflanzen-hybriden (Experiments on Plant Hybrids). Verh. Naturforsch.
Vereines Brilnn 4:3-47, 1866. Reprinted in The Origin of Genetics: a Mendel Source Book,
edited by C. Stern and E.R. Sherwood, W.H. Freeman and Co., San Francisco, pp. 1-48
1966.
Morgan, T.H. Sex limited inheritance in Drosophila. Science 32: 120-122, 1910.
Muller, H.J. Artificial transmutation of the gene. Science (5(5:84-87, 1927.
Plewa, M.J., and E.D. Wagner. Germinal cell mutagenesis in specially designed maize geno-
types. Environ. Health Perspect. 317: 61-73, 1981.
Poirier, L.A., and F.J. de Serres. Initial National Cancer Institute studies on mutagenesis as a
prescreen for chemical carcinogens: an appraisal. J. Natl. Cancer Inst. (52:919-926, 1979.
Purchase, I.F.H., E. Longstaff, J. Ashby, J.A. Styles, D. Anderson, P.A. Lefevre, and F.R.
Westwood. Evaluation of six short-term tests for detecting organic chemical carcinogens
and recommendations for their use. Nature (London) 2(54:624-627, 1976.
Purchase, I.F.H., E. Longstaff, J. Ashby, J.A. Styles, D. Anderson, P.A. Lefevre, and F.R.
Westwood. An evaluation of six short-term tests for detecting organic chemical carcino-
gens. Brit. J. Cancer 37:873-959, 1978.
Rosenkranz, H.S., and Z. Leifer. Determining the DNA-modifying activity of chemicals using
DNA-polymerase-deficient .Escherichia coli. In Chemical Mutagens. Principles and
Methods for Their Detection, Vol. 6, edited by A. Hollaender and F.J. de Serres, Plenum
Press, New York, pp. 109-144, 1980.
Schairer, L.A., J. Van't Hof, C.G. Hayes, R.M. Burton, and F.J. de Serres. Measurement of
biological activity of ambient air mixtures using a mobile laboratory for in situ exposures:
preliminary results from the Tradescantia plant test system. In Application of Short-Term
Bioassays in the Fractionation of Complex Environmental Mixtures, edited by M.D.
Waters, S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, Plenum Press, New York,
pp. 419-440, 1979.
Setlow, R.B., and W.L. Carrier. The disappearance of thymine dimers from DNA: an error-
correcting mechanism. Proc. Natl. Acad'. Sci. U.S.A. 57:226-231, 1964.
Sparrow, A.H., and L.A. Schairer. Mutational response in Tradescantia after accidental ex-
posure to a chemical mutagen. EMS Newsletter No. 5, 16-19, 1971.
Sugimura, T., S. Sato, M. Nagao, T. Yahagi, T. Matsushima, Y. Seino, M. Takeuchi, and T.
Kawachi. Overlapping of carcinogens and mutagens. In Fundamentals in Cancer Preven-
tion, edited by P.N. Magee, S. Takayama, T. Sugimura, and T. Matsushima, University of
Tokyo Press, Tokyo; University Park Press, Baltimore, pp. 191-215, 1976.
United States Environmental Protection Agency. Conference on Current Status of Bioassavs
in Genetic Toxicology (GENE-TOX). Abstracts (1980-740-062/353), USEPA, U.S. Gov-
ernment Printing Office, Washington, D.C., 1980.
Walker, G.C. Plasmid (pKMlOl)-mediated enhancement of mutagenesis and its role in Ames
Salmonella tester system. Texas J. Sci. 3:3-26, 1977.
Waters, M.D., and A. Auletta. The GENE-TOX program: genetic activity evaluation. J.
Chem. Inf. Comput. Sci. 27:35-38, 1981.
Watson, J.D., and F.H.C. Crick. Molecular structure of nucleic acids. A structure for deoxyri-
bose nucleic acid. Nature (London) 171: 737-738, 1953a.
Watson, J.D., and F.H.C. Crick. The structure of DNA. Cold Spring Harbor Symp. Quant.
Biol. 18: 123-131, 1953b.
Yamasaki, E., and B.N. Ames. Concentration of mutagens from urine by absorption with the
nonpolar resin XAD-2: Cigarette smokers have mutagenic urine. Proc. Natl. Acad. Sci.
U.S.A. 74: 3555-3559, 1977.
37
-------
SHORT-TERM IN VITRO TESTS FOR
IDENTIFYING CARCINOGENS:
TRANSFORMATION OF MAMMALIAN
CELLS IN CULTURE*
Arthur C. Upton
INTRODUCTION
Since the pioneering efforts of Sachs and his coworkers, who showed that mam-
malian cells in culture can be transformed reproducibly by chemicals and X rays
(Berwald and Sachs, 1963, 1965; Borek and Sachs, 1966), the transformation of cells
in vitro has become one of the most productive approaches in research on carcino-
genesis. A number of systems for studying the transformation of cells in culture have
been developed (see reviews by DiPaolo and Casto, 1977; Borek, 1979; Casto, 1979;
DiPaolo, 1979; Heidelberger and Mondal, 1979; Krahn, 1979; Pienta, 1979, 1980;
Sivak, 1979; Fisher and Weinstein, 1981).
The development of methods for studying the in vitro transformation of cells rep-
resents a major advance. It enables cellular and subcellular aspects of carcinogenesis
to be investigated directly in ways that are not feasible in the whole animal. It also
provides promising approaches for developing short-term tests for carcinogens, pro-
moting agents, and cocarcinogens. It is thus of enormous potential importance to
those concerned with the detection and characterization of cancer-causing agents in
the environment.
HISTORICAL PERSPECTIVE
The cultivation of mammalian cells in vitro goes back more than half a century,
but it has not yet become possible to grow every type of cell in long-term culture.
Even cancer cells can seldom be propagated for more than a short time after their
removal from the body. Those cells which adapt most readily to culture, tending to
outgrow other elements, are mesenchymal derivatives which are presumably derived
from blood vessels. Until recently most of the studies of cellular transformation have
dealt with such so-called fibroblasts (DiPaolo and Casto, 1977).
One noteworthy characteristic of these cells is that they tend to grow in an orderly
pattern and to stop multiplying after they have formed a continuous sheet over the
floor of the culture vessel. When detached, disaggregated and transferred into fresh
medium, they start multiplying again. The process by which cell division is normally
arrested after the cells have become confluent is known as density-dependent
inhibition. Although the inhibition clearly denotes some form of cell-to-cell interac-
tion, its precise mechanism remains to be determined.
*This manuscript was prepared under support by Grant No. ES00260 from the National Insti-
tute of Environmental Health Sciences and Grant No. CA13343 from the National Cancer
Institute.
The Author Arthur C. Upton is a faculty member of the Institute of Environmental Medicine,
New York University School of Medicine, New York, N.Y.
38
-------
Another property of normal cells is their limited proliferative potential. Most ex-
plants of human fibroblasts die out after the cells have divided about 50 times, de-
pending on the age of the donor. It is only in rare instances that a normal cell sponta-
neously undergoes a change in culture that enables it to multiply indefinitely, giving
rise to a permanent cell line. This change is usually, if not always, accompanied by
chromosomal abnormality, implying that it may signify a mutational alteration
(Hayflick, 1977).
Along with the change in proliferative potential, transformed cells may show alter-
ations in size, shape, staining properties, patterns of growth, and other features,
including acquisition of the capacity to form tumors when inoculated into appropri-
ate test animals (Fig. 1, Table 1). On acquiring these properties, transformed cells ex-
hibit many of the characteristics of cancer cells. Hence, the process of cell transfor-
mation in vitro is analagous in certain respects to carcinogenesis in vivo.
The first spontaneous neoplastic transformations of cultured mammalian cells
were observed in rat and mouse cells, the frequency of transformation being
appreciably higher in these species than in the guinea pig, hamster, or human. Since
the conditions under which cells of a given species are cultured may also influence
their proliferative behavior, including their rate of spontaneous transformation
(Sanford, 1967; DiPaolo and Casto, 1977), the methods and criteria for assaying
transformation depend on the cells in question and the conditions under which they
are grown.
MAJOR ADVANCES IN THE PAST DECADE
Development of Cell Transformation Assays
Of the various properties of transformed cells (Table 1), the ability to produce tu-
mors in vivo is the ultimate test of neoplasia. Since this property is not necessarily
associated with each of the other changes tabulated, efforts have been made to evalu-
ate different characteristics of transformed cells for their reliability as indicators of
tumorigenicity. In addition to being reliably indicative of neoplasia, the "ideal" assay
system should, of course, be reproducible, sensitive, predictive for all types of cells
(such as epithelial cells, fibroblasts), rapid, practical, and quantitative (Steuer and
Ting, 1977).
Colony Assay—
This approach, utilized in the initial studies made by Sachs and coworkers, con-
sists of seeding cells at low concentrations onto a "feeder" layer of cells previously
rendered incapable of dividing by exposure to X rays or the mitotic inhibitor, mito-
mycin-C. The cultures, usually of embryo fibroblasts, are then treated with graded
doses of the agent to be tested, incubated for one to two weeks, and scored (Fig. 2).
Colonies formed by transformed cells are distinguished from colonies of normal
cells by the way in which the transformed cells pile up into "hay stack,''or randomly
oriented, criss-cross patterns, and by changes in their size, shape, and staining prop-
erties (Fig. 3).
One of the chief problems with the colony assay has been its variability. Sources of
variability include differences in the responsiveness of cells from different embryos
of the same species; the subjectiveness of the scoring system, which requires evalua-
tion of the growth patterns of the different kinds of cells in a culture of the whole em-
bryo; variations among batches of calf serum (an essential component of the
medium) in its ability to support the growth of newly transformed cells; and changes
in transformation frequency associated with storage of media (DiPaolo and Casto,
1977; Pienta, 1979;Sivak, 1979; Casto and Carver, 1980). Although these difficulties
have yet to be eliminated, noteworthy progress has been made in efforts to standard-
ize the procedure through the use of pooled cryopreserved cells, meticulous attention
to technical details, and various types of control measures (Pienta, 1979).
39
-------
Initial cultivation
of mouse fibroblasts
Nonmultiplying
nontumorigenic
(anchorage-
dependent) cells
The appearance of colonies
of transformed cells
Multiplying, tumorigenic
(anchorage-independent)
cells
Tumorgenicity
Figure 1. Neoplastic transformation of mouse cells in culture. The transformed cells
form piled-up colonies, are capable of multiplying in semisolid suspension,
and form progressively growing tumors when inoculated into mice.
From Cancer: Science and Society by John Cairns. W.H. Freeman and
Company. Copyright 1978.
A wide variety of agents have been tested with this method, largely in newly ex-
planted or early passage Golden Syrian hamster embryo cells (DiPaolo and Casto,
1977; Pienta, 1979; Casto and Carver, 1980). In the largest series of assays sum-
marized to date, comprising I 10 chemicals of various classes, a 98 percent corre-
lation was observed between transformation in vitro and carcinogenicity in vivo (Ta-
ble 2). No false positives were obtained with the transformation test in contrast to
the Salmonella mutagenicity test; however, a small percentage of false neeatives
were obtained with both tests. The individual chemicals tested and the comparisons
are given in Appendix 1 Comparisons
40
-------
Table 1. Some Properties of Transformed Cells.
Ability to form tumors in test animals
Ability to proliferate indefinitely ("immortalization")
Ability to form multilayered clones
Ability to grow in semisolid suspension (loss of anchorage dependence)
Reduced serum requirement
Increased saturation density
Increased rate of proliferation
Increased cloning efficiency
Morphologic changes (size, shape, nucleo-cytoplasmic ratio, staining, cell junctions,
loss of cell-to-cell orientation)
Altered membrane properties (lectin build-up, agglutination)
Altered antigenic specificity
Increased fibrinolytic activity
Increased uptake of nutrients
Increased glycolysis
Release of tumor angiogenesis factor
Destruction by leukocytes
Destruction by normal cell lysates
(From Martin and Anderson, 1976; DiPalo and Casto, 1977; Steuer and Ting, 1977;
Casto and Carver, 1980).
Certain chemicals fail to transform embryo fibroblasts unless hamster liver micro-
somal homogenates (so-called S-9 homogenates), or hepatocyte feeder layers, are
added to the cultures. This is because of inadequate metabolic activation of the
chemicals by hamster embryo fibroblasts alone, which lack the activating capability
in their microsomes. An alternative ("host-mediated") method for enhancing the ac-
tivation of such precarcinogens is to expose the embryo cells in utero, by inoculating
the chemicals into mid-term pregnant hamsters two to three days before the embryo
cells are explanted (DiPaolo et al., 1973).
Advantages of this assay system, as compared with others, include the facts that
the frequency of spontaneous transformation in hamster embryo cells is virtually nil,
the transformation of such cells has been observed to follow one-hit kinetics, and the
test results can be scored within seven to ten days. Moreover, the transformed cells
have been found consistently to be tumorigenic on inoculation into young hamsters
(DiPaolo and Casto, 1977; Casto and Carver, 1980).
Focus Assay—
Cells, seeded at a relatively high density into culture medium without any feeder
layer, are allowed to grow to confluence and scored for transformation by the pres-
ence of multilayered, randomly oriented foci of dense, piled-up growth. The assay
ordinarily uses cells of permanent lines, which are aneuploid and require several
weeks to score, in contrast to the colony assay which typically uses diploid cells
which require only one to two weeks to score. Cells transformed in this system can
show the full spectrum of changes associated with transformation (Reznikoff et al.,
1973).
Various focus assay systems have been developed thus far, including those using
C3H mouse fibroblasts, BALB/3T3 mouse embryo fibroblasts, C3H/10T 1/2
mouse embryo fibroblasts, rat embryo cells, rat embryo cells expressing an endoge-
nous oncorna virus, hamster BHK-21 cells, guinea pig embryo cells, and human skin
fibroblast KD cells (Kakunaga, 1977; Heidelberger and Mondal, 1979; Borek,
1980a). A number of focus assay systems exploit the ability of carcinogenic agents to
41
-------
Day 0
Reconstitute frozen standard
feeder cells from liquid
nitrogen. Refeed after 8 hours.
Day 2
Subculture the feeder cells.
Day 3
Refeed the feeder cells.
Day 4
X-lrradiate feeder cells with
5000R, then seed only 50 mm
dishes at 8 x 104 cells/dish.
Reconstitute target cells and
refeed 6 hours later.
Day 5
Seed target cells onto feeder
layer cells at 500 cells/dish.
Day 6
Add chemicals and incubate
dishes without disturbing
or refeeding.
Day 14
Fix with methanol, stain with
Giemsa, and monitor dishes
for transformation.
Figure 2. Outline of procedure of standard colony assay for in vitro transformation of
Golden Syrian hamster embryo cells (adapted from Sivak, 1979).
cause a dose-dependent enhancement of virus-induced transformation in cultured
cells (DiPaolo and Casto, 1977; Sivak, 1979).
The focus assay method has been applied to a wide variety of chemicals, with re-
sults (Appendix 2) that are comparable, in general, with those obtained with the col-
ony assay method. However, the aneuploid cells usually used in the focus assay have
a higher rate of spontaneous transformation than the diploid cells used in the colony
assay, the number of additional transformed foci produced by a given dose of
carcinogen is inversely related to the number of cells plated, and it is not certain
whether each focus arises from a single transformed cell. Hence the rate of transfor-
mation induced in the focus method cannot be calculated precisely for comparison
with that in the colony method. Nevertheless, the focus method has certain advan-
tages, including higher cloning efficiency and greater availability of cloned non-
transformed cells for comparative studies.
42
-------
Normal
Transformed
Figure 3. Schematic illustration of the microscopic appearance of transformed
fibroblasts as compared with normal fibroblasts. The normal fibroblasts grow
in an orderly orientation, forming a single layer of relatively uniform,
confluent eel Is, whereas the transformed fibroblasts vary in size, shape, and
staining intensity, piling up on one another in a randomly oriented, criss-
cross pattern. Reprinted from Pilot (1978) by courtesy of Marcel Dekker.
Colony Formation in Soft Agar—
Following the observations by MacPherson and Montagnier (1964) that baby
hamster kidney (BHK) cells acquire the ability to form colonies in soft agar after
transformation by the polyoma virus, a number of investigators have utilized growth
in semisolid suspension as a measure of cellular transformation (DiPaolo and Casto,
1977; Purchase et al., 1978; Bouck and Di Mayorca, 1980). This property, which de-
notes loss of anchorage dependence, is closely correlated with tumorigenicity in ani-
mals (Steuer and Ting, 1977).
In this assay method, cells are incubated in the presence of the agent to be tested,
subsequently resuspended in culture medium containing 0.3-0.4 percent agar, seeded
onto an agar base layer containing nutrients, incubated for one to three weeks, and
then scored for colony formation. Because normal cells cannot grow unless
anchored to a substratum, the culture conditions select for transformed cells. The in-
clusion of a rat liver microsomal preparation in the culture medium enhances meta-
bolic activation of the chemical to be tested, generally increasing colony formation
(Styles, 1978).
In a comparative study of 120 chemicals of different classes, the method accurate-
ly identified 53 out of 58 carcinogens and 60 out of 62 noncarcinogens (Appendix 3).
Results obtained with human cells of the WI-38 and Chang lines were almost as good
as those obtained with hamster (BHK-21) cells, in spite of the fact that the frequency
of transformation was approximately 10 times lower in the human cells (Purchase et
al., 1978). The method compared favorably with any of five short-term tests utilizing
end points other than in vitro transformation (Table 3).
43
-------
Table 2. Comparative Effects of Different Classes of Chemicals with Respect to Transformation of Hamster Embryo Cells In
Vitro. Mutagenicity in Salmonella, and Carcinogenicity In Vivo.*
Clinical Class
Direct alkylating agents
Polycyclic hydrocarbons
and derivatives
Nitrosamines and amides
Aromatic amines and
aminoazo dyes
Metal salts
Solvents
Miscellaneous compounds
Totals
Carcino-
genicity Number
in wVof Tested
+ 10
1
+ 7
5
+ 6
1
+ 27
8
+ 4
6
+ 19
14
+ 73
35
Cell Transformation In Vitro
Concurrence^
10/10
1/1
7/7
5/5
5/6
1/1
26/27
8/8
4/4
6/6
18/19
14/14
70/73
35/35
False
Negatives
0
0
1
1
0
-
1
(96%) 3
(100%) 0
False
Positives
0
0
0
0
-
0
0
0
Mutagenicity in Salmonella
Concurrence§
10/10
1/1
7/7
4/5?
6/6
1/1?
25/27
6/8
4/4?
6/6?
6/19
13/14?
58/73?
30/35?
False
Negatives!
0
0
0
2
4?
-
13
19?
False
Positives!
0
1
1?
2
-
1?
3?
7?
*From Pienta (1979) (Values tabulated do not include results on two nitrosamines of undetermined carcinogenicity).
f+ = positive result, - = negative result.
^Values include results obtained when hamster liver S-9 homogenate was added to the culture system.
§Question marks denote the occurrence of one or more equivocal results among the values tabulated.
-------
Table 3. Results with the In Vitro Transformation Assay (Colony Formation in Soft Agar), as Compared with Five Other Short-
Term Tests for Carcinogenicity.*
Compounds Identified
Chemical Class
Polycyclics
Carcinogenic
Noncarcinogenic
Arylamines
Carcinogenic
Noncarcinogenic
Alkylating Agents
Carcinogenic
Noncarcinogenic
Miscellaneous
Carinogenic
Noncarcinogenic
Total All Classes
Carcinogenic
Noncarcinogenic
Number
Tested
11
9
20
13
18
0
9
40
58
62
Cell
Transformation
(Nos.)
10/11
9/9
19/20
13/13
17/18
-
7/9
38/40
53/58
60/62
<%>
91
100
95
100
94
-
78
95
99
97
Salmonella
Mutation
(%)
100
89
95
92
83
-
89
95
91
94
Correctly
Degranu-
lation
(%)
73
55
95
46
61
-
33
82
71
71
Sebaceous
Gland
Suppression
(%)
100
78
55
75
67
-
55
57
67
64
Tetrazolium
Reduction
!%>
45
55
40
83
39
-
33
74
40
73
Implant
(%)
75
100
39
92
25
-
22
95
37
95
*Based on data in Appendix 3 (Purchase et al., 1978).
-------
Mass Culture—
Confluent or subconfluent cells are incubated in the presence of the agent to be
tested, subsequently transferred for several passages at high density (~50,000
cells/cm2), and then observed for changes indicative of transformation such as
random orientation, criss-cross growth, ability to replicate at high density, ability to
grow at reduced serum concentrations and increased glycolysis (Table 1).
The method shown in Fig. 4 entails repeated subculturing of cells over compara-
tively long periods of time (weeks or months) and cannot be used to determine the
precise frequency of transformation, but its use of increased numbers of cells at risk
enchances its sensitivity to detect weak carcinogens (Casto and Carver, 1980).
Epithelial Cell Culture Systems
Although most cell cultures have thus far dealt with fibroblasts (Table 4), the ma-
jority of human cancers are epithelial in origin. Hence, culture systems for studying
epithelial cell transformation have been sought for years. The systems now under
most active study include cultures of liver cells (Borek, 19806; Montesano et al.,
1975; Williams, 1976; Weinstein et al., 1975), epidermal cells (Fusenig et al., 1973;
Colburn et al., 1978; Lillehaug et al., 1979), salivary gland cells (Brown, 1973), uri-
nary bladder epithelium (Hashimoto, 1978) and respiratory tract epithelium (Marchok
and Nettesheim, 1977; Steele et al., 1978).
To date, experience with these systems remains largely exploratory. None have
developed far enough as yet to be evaluated as a potential assay method (DiPaolo
and Casto, 1977; Sivak, 1979; Casto and Carver, 1980). The phenotypic markers of
the transformed state which are characteristic of fibroblasts in vitro (Table 1) are not
as readily distinguishable in cultured epithelial cells. Ability to form colonies in semi-
solid suspension is currently the most reliable and consistent criterion of the epitheli-
al cell transformation, although capacity to produce progressively growing tumors
in animals constitutes the final proof of malignancy (Borek, 19806).
Organ Culture Systems
The conditions under which cells are ordinarily cultivated in vitro—where cells of
a single type are propagated—differ profoundly from those in vivo, where growth
and differentiation involve complex interactions between mutually interdependent
cells of different types in organs and tissues. To investigate the transformation of
cells under conditions more closely resembling those in vivo, efforts have been made
to develop culture systems for various organs. Those currently under investigation
include organ culture systems for tissues of the respiratory tract, colon, and pancre-
atic duct (Saffiotti and Harris, 1979).
Among the potential advantages of the organ culture approach are (1) the ability
to preserve the morphological and functional integrity of tissues for long periods, in-
cluding intercellular and epithelial-stromal relationships; (2) the ability to study
effects of carcinogens on the tissue as a whole, and to compare the reactions of
component cell types concurrently; (3) the ability to use a tissue as an experimental
system in which to study the influence of microenvironmental changes; (4) the ability
to explore the process by which carcinogens are activated and detoxified in tissues,
during which their interactions with tissue elements can be localized by autoradio-
graphic, biochemical, and biophysical techniques; (5) the ability to study the effects
of carcinogens at different levels of biological organization, for example at the level
of the whole tissue, the epithelium, and stroma, individual cell types, cellular organ-
elles, and molecular components; (6) the ability to observe histological tissue reac-
tions preceding neoplasia, such as hyperplasia, metaplasia, and dysplasia; (7) the
ability to transplant the tissue, after its in vitro manipulation, into suitable recipient
animals for further observation in vivo; (8) the ability to use the tissue as a source of
cell monolayer cultures; and (9) the ability to use human tissue in organ culture for
evaluating species differences in responses to carcinogens, and to compare
46
-------
Day 0
Seed target cells inyo 25 cm2
flasks at 7 x 10s cell/flask.
Day 2
Refeed cells and add chemical
Day 9
Refeed cells and add chemical.
Day 16
Wash cells with BSS; split
into 3 new flasks.
Day 21
Refeed flasks.
Day 26
Split one culture into 3 new
flasks. Fix one flask with
methanol and stain with
Giemsa. Refeed one culture
and hold as backup.
Hold Stain Subculture
»
Continue passaging until
(1) focus formation is
observed in treated
cultures, or
(2) Cultures have been
passaged 14 times
after treatment.
Figure 4. Outline of procedure of mass culture assay for in vitro transformation
of Rauscher leukemia virus-infected Fischer rat embryo cells, (adapted
from Sivak, 1979).
experimentally induced responses in human tissue with those that occur naturally
(Saffiotti and Harris, 1979).
While the potential long-range promise of the organ culture approach as a research
tool should not be underestimated, it has not yet developed sufficiently to provide a
practical assay method at this time.
Assays for Cocarcinogens, Promoting Agents, and Other Modifying Factors
Because of the multicausal, multistage nature of the cancer process, short-term as-
says for cocarcinogens and promoting factors are needed, as well as assays for initi-
ating agents and complete carcinogens. It is noteworthy, therefore, that in vitro cell
transformation systems have recently been shown to be applicable to the study of a
growing number and diversity of incomplete carcinogens and factors that modify the
47
-------
Table 4. Cell Culture Systems Investigated for Assaying the Carcinogenic
Potential of Chemicals.*
Type of System
Assay Method
Scoririg Time
A. Primary or Secondary
Embryo Cultures
1. Syrian Hamster
2. Guinea Pig
3. Rat
4. Rat or Mouse
infected with murine
leukemia virus
5. Rat infected
with adenovirus
6. Syrian Hamster
infected with adenovirus
B. Fibroblast Cultures
1. Human Skin Fibroblasts
C. Fibroblast Cell Lines
Mouse Prostate (Adult)
Mouse 3T3t
Mouse 10T1/2 t
Syrian Hamster BHK21 §
Rat Embryo
Cell Line
(Fischer rat embryo )
Colony
Growth in Agar
Mass culture—Focus
Mass culture—Focus
Mass culture—Focus
Focus
Mass culture—Focus
Focus
Focus
Focus
Growth in Agar
Mass culture—Focus
8-10 days
4 months
130 weeks
3-8 weeks
25-42 days
25-30 days
7-14 weeks
4-6 weeks
10-28 weeks
4-6 weeks
2-4 weeks
40-43 weeks
). Epithelial Cell Cultures
1 . Rat Liver
2. Mouse Epidermis
Growth in Agar
Growth in Agar
8-36 weeks
12-16 weeks
"From'Casto (1979), Fisher and Weinstein (1981).
fCell line established from BALB mouse embryos.
JCell line established from C3H mouse embryos.
§Established from newborn Syrian hamster kidneys.
|| Transformation occurred in later passage cultures treated with carcinogens.
process of carcinogenesis. Furthermore, since promoting agents have thus far been
distinguished from initiating agents, cocarcinogens, and complete carcinogens large-
ly on the basis of operational definitions (Table 5), the availability of culture meth-
ods for identifying such agents and studying their mechanisms of action is of scientif-
ic, as well as practical, importance.
Based on the available evidence, the transformation of cells in vitro, like carcino-
genesis in vivo, has been inferred to be a multistage process (Heidelberger and
Mondal, 1979; Weinstein et al., 1979; Poiley et al., 1979). The tumor-promoting
agent, 12-0-tetradecanoyl-phorbol-13 acetate (TPA), among other agents, has been
observed to enhance the transformation of certain cells in vitro by various chemicals
(Heidelberger et al., 1978; Colburn et al., 1978; Weinstein et al., 1978), ultraviolet
light (Mondal and Heidelberger, 1976), X rays (Kennedy et al., 1978), and viruses
(Weinstein et al., 1978). Its transforming effects in vitro are associated with many of
the biochemical alterations that TPA produces in vivo (Weinstein et al., 1978), and
48
-------
Table 5. Biological Properties of Promoting Agents, as Compared with
Initiating Agents.
Promoting Agents Initiating Agents
1. Not carcinogenic alone-must be pre- 1. Carcinogenic by themselves--"solitary
ceded by an initiating agent carcinogens"
2. Require prolonged exposure(s) 2. Single exposure is sufficient
3. Action is reversible (at early stage) 3. Action is relatively irreversible and
and not additive additive
4. Probable threshold 4. No apparent threshold
5. No evidence of covalent binding 5. Yield electrophiles that bind cova-
lently to cell macromolecules
6. Not mutagenic 6. Mutagenic
(From Weinstein et al., 1978).
the effects are elicited at low concentration (in the range of 10 9M) in cells of diverse
species and types. Not all types of cells respond similarly, however, nor is the re-
sponse and its modification by cofactors the same in vitro as in vivo (Fusenig and
Samsel, 1978; Weinstein et al., 1978).
In view of the diversity of factors influencing the growth and transformation of
cells in vitro, and the fragmentary state of present knowledge about the dynamics of
cell regulation, much further research will be necessary before practical in vitro assay
systems for cocarcinogens and promoting agents will become available.
Comparative Results with Different Assay Systems
Most types of chemical carcinogens have been found to transform mammalian
cells in vitro under suitable test conditions, although some fail to do so in certain as-
say regimens because of inadequate metabolic activation in vitro. The addition of
liver microsomal extracts or an hepatocyte feeder layer has corrected this deficiency
in many such instances. The absence of transformation in vitro with other agents (for
instance, croton oil and asbestos) may signify that their carcinogenic activity in vivo
is mediated through promoting effects or effects on posttransformational stages in
the cancer process.
In general, in vitro transformation assays have not thus far materially outper-
formed the Ames test or certain hepatocyte DNA repair tests (Martin et al., 1978;
Williams, 1979) in the percentage of carcinogens identified, but they have tended to
yield fewer false positives (Tables 2 and 3). Furthermore, as noted above, in vitro
transformation systems are potentially able to identify cocarcinogens and
promoting agents of types that are not detectable by mutagenicity or DNA repair
tests. In vitro transformation tests have also identified correctly an appreciably high-
er percentage of carcinogens than has been identified by degranulation of rat liver
endoplasmic reticulum, reduction of terazolium red in mouse skin, suppression of
sebaceous glands in mouse skin, histologic reaction to chemically impregnated filter
discs implanted subcutaneously in mice (Table 3), or other short-term tests in-
vestigated to date, including transplacental blastomagenesis, piperidine alkyla-
tion, and the iodine and acridine tests (Purchase et al., 1978).
A combination of complementary tests can, obviously, be expected to identify a
higher percentage of carcinogens than any one test alone; however, it may also yield
a higher percentage of false positives (Table 6). An optimal balance between false
negatives and false positives among the test results tabulated (Table 3) is achieved by
combining the results of only two of the six tests: the cell transformation test and the
49
-------
Table 6. Comparative Results of Short-Term Tests When Used Alone or in
Combinations.*
Testf
CT
BM
Deg
SI
SG
TR
CT+BM
CT+BM+Deg
CT+BM+SI
CT+BM+Deg+SI
CT+BM+Deg+SG
CT+BM+SI+SG
CT+BM+Deg+SI+SG
CT+BM+Deg+SG+TR
CT+BM+SI+TR
All 6 Tests
% of
Carcinogens
Which Would be
Positive in at
Least One TestJ
91
91
71
37
67
40
99.19
99.77
99.49
99.85
99.92
99.83
99.95
99.95
99.90
99.97
%of
Non-carcinogens
Which Would be
Positive in at
Least One Test:):
3
6
29
5
36
27
8.8
35.3
13.4
38.5
58.6
44.6
60.6
69.8
59.5
71.33
% of
Errors
(False Negatives
plus
False Positives)
12
15
58
68
69
87
9.61
35.53
13.91
38.65
58.68
44.77
60.65
69.85
59.60
71.36
*From Purchase et al. (1978), based on the results in Appendix 3.
tCT = cell transformation; BM = bacterial mutation; Deg = degranulation; SI = subcutane-
ous implants; SG = sebaceous gland; TR = 1 tetrazolium reduction.
|The number of decimal places does not represent the likely accuracy of the figures but
demonstrates the likely size of the differences between them.
Salmonella mutation test, which together yield a total error rate (false negatives plus
false positives) of only 9.61 percent (Table 6).
The relative importance of false negatives, as compared with false positives, can be
expected to vary, depending on the ratio of carcinogens to noncarcinogens among
the compounds to be tested. The effects of variations in this ratio, analyzed by
Purchase et al. (1978) on the basis of the data summarized in Table 3, are shown in
Figs. 5 and 6. It is evident from this analysis that the transformation test and the bac-
terial mutation test can complement and reinforce each other (Fig. 5) but that one
test adds little to the reliability of the other when the results of the two are contradic-
tory (Fig. 6). The reliability of the two tests has been analyzed further on the basis of
three performance criteria suggested by Cooper et al. (1979) as (I) "sensitivity,"or
the proportion of known carcinogens that can be identified correctly; (2) "specific-
ity," or the proportion of known noncarcinogens that can be identfied correctly;
and (3) "predictive value," or the proportion of chemicals giving positive results that
are actually carcinogens (that is, the proportion of positives that are true positives).
When judged on the basis of these performance criteria, appropriate use of the two
tests in combination has been calculated to yield results of relatively high reliability
(Table 7). Since, however, the ability of the Salmonella mutagenicity test to identify
carcinogens varies appreciably among different classes of chemicals (Rinkus and
Legator, 1979), the figures shown in Table 7 must be interpreted accordingly
50
-------
1.0-t
I
20 40 60 80
Percentage of carcinogens in the sample tested (p)
100
Figure 5. Variation in the probability that a compound will be correctly identified as a
carcinogen (true positive result) by the combined use of the in vitro
transformation test and the Salmonella mutagenicity test, in relation to the
percentage of carcinogens among the samples being tested (from Purchase
eta/.. 1978).
Each curve represents a particular test result: (a) bacterial mutation and cell
transformation positive; (b) cell transformation positive; (c) bacterial
mutation positive; (d) bacterial mutation negative and cell transformation
positive; (e) bacterial mutation positive and cell transformation negative;
(f) bacterial mutation negative; (g) cell transformation negative; (h) bacterial
mutation and cell transformation negative. The curves are calculated from
the formula:
PA
P = •
pA
where P is the probability that a compound producing a particular test result
is a carcinogen, p is the percentage of carcinogens in the samples being
tested, A is the probability of obtaining the test result with carcinogens, and
B is the probability of obtaining the test result with non-carcinogens.
The values for A and B were obtained from Appendix 3.
51
-------
1.0 —k
0.8 —
0.6 —
0.4 —
0.2 —
Percentage of carcinogens in the sample tested (p)
Figure 6. Variation in the probability of true positive and false positive results provided
by the combined use of discordant results, from the in vitro transformation
test and the Salmonella mutagenicity test, in relation to the percentage of
carcinogens among the samples being tested (from Purchase ef a/., 1978).
The curves d and e (true positives) are the same as in Figure 5. The false
positives, shown in curve dd (Salmonella mutation test negative, in vitro
transformation test positive) and ee (Salmonella mutation test positive,
in vitro transformation test negative) were calculated from the formula:
P =
pA + (1-p)B
where P is the probability that a compound producing a particular test result
is a carcinogen, p is the percentage of carcinogens in the samples being
tested, A is the probability of obtaining the test result with carcinogens and
B is the probability of obtaining the test result with non-carcinogens '
The values for A and B were obtained from Appendix 3.
52
-------
Table 7. Comparative Reliability of Results from In Vitro Transformation
Test and Salmonella Mutation Test, used Alone or in Combina-
tion, on Various Classes of Chemicals.*
Performance Criterionf
Test Sensitivity Specificity Predictive Valuej
(accuracy in percent)
In Vitro Transformation 91 97 77
Salmonella Mutation (Ames) 91 94 63
Both Tests Combined
1. When both agree 96 100 100
2. When a positive result
in either test is
considered positive 99 91 55
*From Purchase (1980), based on results in Table 3.
fSee text for definition of terms.
^Values denote predictive value for a positive test when carcinogens comprise ten per-
cent of compounds being tested.
IMPLICATIONS OF ENVIRONMENTAL LEGISLATION AND
REGULATORY ACTIONS
During the past 25 years cancer has come to be viewed increasingly as a multicaus-
al disease in which environmental carcinogens play a significant etiologic role (Doll,
1978; Higginson, 1979). This concept, along with growing evidence that tumor-
forming activity in laboratory animals can be predictive of carcinogenicity for
humans has prompted regulatory agencies to make increasing use of experimental
evidence as a basis for determining whether a substance may pose a risk of cancer to
man. Thus, in the Food Additives Amendment of 1958, the Delaney Clause prohib-
ited approval of any food additive that causes cancer in test animals on ingestion.
This clause has been construed by the Food and Drug Administration to require dis-
approval of the addition of a substance to food in any amount if it has been identified
as a carcinogen through appropriate animal test (Hutt, 1978).
In the years since 1958, legislation conferring regulatory authority for carcinogens
has increased greatly in scope and complexity. Partly in an effort to avoid inconsist-
ency in regulatory policies on carcinogens, several agencies together formed the In-
teragency Regulatory Liaison Group (IRLG) in 1977 and two years later issued a
common policy statement setting forth the types of evidence on which carcinogens
are to be identified, the scientific bases for evaluating the evidence, and methods for
assessing risks to human populations (IRLG, 1979). With respect to short-term tests
for carcinogenicity, the IRLG Report (1979) concluded that in vitro mammalian
transformation systems are useful in exploring the mechanisms of carcinogenesis, in
screening suspect agents for further study, and in supporting evidence derived from
animal or epidemiological investigations, but that they are insufficient by themselves
to establish the carcinogenicity of an agent for humans or experimental animals.
These conclusions are in accord with those of the National Cancer Advisory Board's
Subcommittee on Environmental Carcinogenesis (1977), and with preliminary re-
sults of a three-year international collaborative study to evaluate in vitro cell trans-
formation among 30 short-term assay systems for their ability to identify 42 coded
53
-------
carcinogens and noncarcinogens (National Toxicology Program, Bulletin, 1980). In
view of the importance of these reports for short-term tests, their conclusions are
given in full in Appendix 4.
Although insufficient by themselves to establish the carcinogenicity of a substance,
short-term tests assumed increased importance as indicators of presumptive risk to
humans with the passage of Public Law 95-622, in 1978. This law requires the Secre-
tary of the Department of Health and Human Services to publish annually a report
containing ".. .a list of all substances (i) which either are known to be carcinogens or
which may reasonably be anticipated to be carcinogens and (ii) to which a significant
number of persons residing in the United States are exposed...." Compliance with
this mandate will inevitably focus increasing attention on the significance of data
from short-term tests in the assessment of cancer risks and in public policy directed
toward the control of environmental carcinogens.
OUTLOOK FOR THE FUTURE
In vitro cell transformation systems have only recently come under serious study
as potential short-term assay methods for detecting carcinogens, but it is clear that
they have much to offer. Although their methodology is still in its infancy, the
systems already appear capable of identifying carcinogens as reliably as any other
short-term tests now in use. In addition, they offer potential advantages over other
systems in (1) scoring oncogenic transformation in vitro, an end point that is more
closely analogous to carcinogenesis than the end points scored in other short-term
tests such as mutagenicity assays or suppression of sebaceous glands; (2) possessing
the capability to utilize cells of diverse sources and types, including human cells,
thus enabling direct analysis of species and organ differences in response to
potentially carcinogenic stimuli; and (3) adaptability to the detection and char-
acterization of incomplete carcinogens, cocarcinogens, promoting agents, and
other factors modifying the process of carcinogenesis.
While the prospects for in vitro cell transformation assays are highly encouraging,
their ultimate practical usefulness will depend on the extent to which various techni-
cal problems can be resolved. One problem encountered with all the assay systems
investigated thus far is variability among target cell populations. Cells of different
clones, as well as those of different embryos, vary in their growth characteristics and
transformation response in vitro. Because of this variability, cell populations used
for assays must be selected with care, monitored closely, and standardized, insofar as
possible, by the use of cryopreserved pools of cells which are carefully defined and
derived by expansion from rigorously maintained and uncontaminated foundation
stocks.
Additional factors affecting the reproducibility of assays include variations in
successive batches of culture medium, serum, trypsin, and other animal-derived re-
agents, glassware, duration of storage of cells and culture materials, quality of air
and water, number of times the cells are subcultured before use, and cell population
density at seeding(Pienta, 1979). Attention to these matters is essential for reproduc-
ibility of results, but despite all such efforts to date, inexplicably large variations
continue to be observed from time to time in positive and negative controls, empha-
sizing the complexity of culture systems and the need for further refinements in their
methodology (Sivak et al., 1980).
Another important variable is the ability to convert precarcinogens into active me-
tabolites. It is evident that a supplementary system for metabolic activation, in the
form of microsomal homogenates or hepatocyte feeder layers, is necessary for the
transformation of certain types of cultured cells (especially embryo fibroblasts) by
some classes of chemicals such as aromatic amines and nitrosamines. It would
appear, moreover, that the effectiveness of such exogenous metabolic activation
systems may vary, depending on their species and organ of origin, the enzyme in-
ducer (if any) used for priming the donor animal from which they are derived, the
54
-------
type of chemical to be activated in the assay, and the target cells to be transformed
(Sivak and Tu, 1980). Further complicating the problem are variations in the
toxicity of microsomal preparations for different cells in vitro, the dependence of cell
transformation on the diffusion of active metabolites to critical sites within the
target cells, and the likelihood that both deactivation and activation are occurring
simultaneously in the system. In view of these complexities, it would appear that
different classes of chemicals call for different assay regimes, the details of which
cannot yet be specified. Hence, the design of an optimal strategy for each class of
compounds will require further research. Moreover, because S-9 homogenates differ
in their activation patterns from the intact cells whence they are derived, feeder layers
of intact cells would seem preferable to homogenates as exogenous metabolic activa-
tion systems (Bigger et al., 1980).
Cytotoxicity also constitutes a source of variation, since the frequency of transfor-
mation is a function of the number of cells surviving exposure to the carcinogen and,
in turn, cell density. No simple relation between cytotoxicity and transforming
activity is evident from studies to date. With many chemicals, transformation is elic-
ited only over a narrow range of doses, high concentrations being too toxic although
not necessarily lethal, and low concentrations requiring treatment of many replicate
cultures over extended periods of time (Casto and Carver, 1980). With such chemi-
cals, negative responses may merely reflect failure to test an appropriate range of
doses. With carcinogens that transform only at high levels of toxicity or at low fre-
quency, the use of mass cultures can be advantageous by increasing the number of
cells at risk and amplifying transformational changes through successive passages of
the affected cells. This technique has the disadvantage, however, of being unable to
indicate the precise frequency of transformation, since each of the foci that are
scored cannot be assumed to have originated from a single transformed cell. A simi-
lar limitation restricts the use of focus assays, in general, as quantitative tools. These
problems warrant continuing study.
Although it is apparent from existing data that in vitro transformation assays are
capable of identifying carcinogens of a wide variety of chemical classes, further vali-
dation is clearly called for. Efforts aimed at such validation should include attempts
to determine the most appropriate assay regimen for every class of chemicals, com-
paring responses to noncarcinogenic as well as carcinogenic analogues in each class
(Purchase et al., 1978). Only in this way will it eventually become possible to select
the most reliable test, or battery of tests, for evaluating unknown substances.
Further refinements in methodology should reduce the frequency of false negative
results in the future. Such results may never be eliminated entirely, since some
substances undoubtedly contribute to carcinogenesis in vivo by acting on posttrans-
formational stages in the cancer process, through immunosuppression or other ef-
fects on homeostasis at the tissue or organ level. It is noteworthy that some in vitro
transformation systems have been found capable of detecting the effects of various
promoting agents, cocarcinogens, and other modifiers of carcinogenesis. Since these
factors may conceivably be as important in the pathogenesis of most human cancers
as complete carcinogens or initiating agents, and since there is little prospect of
assaying for them on a large scale in whole animals with the limited resources that
are available, the possible application of in vitro cell transformation systems is one
of the most important questions to be explored.
It is noteworthy that false positives have been observed in less than one percent of
assays. Their occurrence in some instances may actually signify that the chemicals in
question have received insufficient study in vivo and would, in fact, turn out to be
carcinogens if investigated more adequately. Alternatively, it is entirely conceivable
that transforming activity in vitro will not always be correlated with carcinogenicity
in vivo, because of the different mechanisms that operate in vivo (for example, de-
activation, DNA repair, or immunological surveillance) which may alter the final
outcome (Purchase et al., 1978). To minimize the frequency of false positives, it is
55
-------
important that the criteria used for scoring transformation include the ability of the
transformed cells to form tumors in test animals.
Based on the evidence available thus far, the in vitro cell transformation assay
appears to be the most reliable and versatile short-term test for carcinogenicity. The
assay can be performed in a matter of days or weeks, at a cost ranging from a few
hundred to a few thousand dollars. It is, therefore, far less costly and time-consum-
ing than bioassays of whole animals. Pending further validation of the assay for
chemicals of diverse classes, and in view of continuing improvements in its sensitivity
and reliability, it should be considered for inclusion in any battery of short-term tests
employed for detecting and screening environmental carcinogens. Furthermore,
because of its long-range potential as a test that may be applicable to promoting
agents, cocarcinogens, and other factors modifying the carcinogenic process, further
research to explore and refine such applications should receive high priority.
ACKNOWLEDGMENTS
The author thanks Ms. Lynda Witte and Ms. Jean Smith for assistance in the
preparation of this manuscript.
REFERENCES
Berwald, Y., and L. Sachs. In vitro cell transformation with chemical carcinogens. Nature
(London) 200:1182-1184, 1963.
. In vitro transformation of normal cells to tumor cells by carcinogenic hydro-
carbons. J. Nail. Cancer Insl. 35:641-661, 1965.
Bigger, C.A., .I.E. Tomaszewski, A. Dipple, and R.S. Lake. Limitations of metabolic activa-
tion systems with in vitro tests for carcinogens. Science 209:503-505, 1980.
Borek, C. Malignant transformation in vitro', criteria, biological markers, and application in
environmental screening of carcinogens. Radiation Res. 79:209-232, 1979.
. X-ray-induced in vitro neoplastic transformation of human diploid cells. Nature
(London)'2#3:776-778, 1980a.
Metabolic activation and malignant transformation in cultured liver cells ex-
posed to chemical carcinogens. Adv. Mod. Environ. Toxicol. 7:297-318, I980/J.
Borek, C., and L. Sachs. In vitro cell transformation by X-irradiation. Nature (London)210:
276-278, 1966.
Bouck, N., and G. Di Mayorca. The mutagenic nature of in vitro transformation of BHK cells
by chemical carcinogens. Adv. Mod. Environ. Toxicol. /:447-466, 1980.
Brown, A.M. In vitro transformation of submandibular gland epithelial cells and fibroblastsof
adult rats be methylcholanthrene. Cancer Res. 33:2779-2789, 1973.
Cairns, J. Cancer and Society, W.H. Freeman and Co., San Francisco, 1978.
Casto, B. Oncogenic transformation of mammalian cells by chemicals and viral chemical inter-
actions. In Application of Short-Term Assavs in the Fraclionalion of Complex Environ-
mental Mixtures, edited by M.D. Waters, S. Ncsnow, J.L. Huisingh, S,S. Sandhu,and L.
Claxton, Plenum Press, New York, pp. 65-97, 1979.
Casto, B.C., and J.H. Carver. Carcinogcncsis and mulagenesis testing in mammalian cell cul-
tures. Adv. Mod. Environ. Toxicol. /:24I-271, 1980.
Colburn, N.H., W.F. Vorder Brueggc, J. Bales, and S.H. Yuspa. Epidermal cell transforma-
tion in vitro In Carcinogenesis, Vol. 2. Mechanisms of Tumor Promotion andCocarcino-
genesis, edited by T.J. Slaga, A. Sivak, and R.K.. Boutwell, Raven Press New York pp
257-271, 1978.
Cooper, J.A., R. Saracci, and P. Cole. Describing the validating of carcinogen screenine tests
Br. J. Cancer 39:87-89, 1979.
Department of Health and Human Services. Fir.il Annual Report on Carcinogens Vol I De-
partment of Health and Human Services, Washington, IXC., 1980.
DiPaolo, J.A. Quantitative transformation by carcinogens of cells in early passage In F >i
ronmental Carcinogenesis, edited by P,T. Emmclot and E. Krick, HIscvier Press Am'i
dam, pp. 365-380, 1979. "'
DiPaolo, J.A., and B.C. Casto. Chemical carcinogcnesis. In Recent Advances in Cancer R
search: Cell Biology, Molecular BiologvandTumor Virology, Vol. I,edited by R C (' II
CRC Press, Cleveland, pp. 17-47, 1977. '
56
-------
DiPaolo, J.A., R.L. Nelson, P.J. Donovan, and C.H. Evans. Host-mediated in vivo-in vitro as-
say for chemical carcinogens. Arch. Pathol. 95:380-385, 1973.
Doll, R. An epidemiological perspective on the biology of cancer. Cancer Res. 35:3573-3583,
1978.
Fisher, P.B., and I.E. Weinstein. In vitro screening tests for potential carcinogens. In Carcino-
gens in Industry and Environment, edited by J.M. Sontag, Marcel Dekker, New York, 1981
(in press).
Fusenig, N.E., and W. Samsel. Growth-promoting activity of phorbol ester TPA on cultured
mouse skin keratinocytes, fibroblasts, and carcinoma cells. In Carcinogenesis, Vol. 2.
Mechanisms of Tumor Promotion and Cocarcinogenesis, edited by T.J. Slaga, A. Sivak,
and R.K. Boutwell, Raven Press, New York, pp. 203-220, 1978.
Fusenig, N.E., W. Samsel, W. Thon, and P.K.M. Worst. Malignant transformation of epider-
mal cells in culture by DMBA. fXSERM Symp. /P:2I9-228, 1973.
Hashimoto, Y. In vitro and in vivo effect of nitrosamine and urea on rat urinary bladder epithe-
lial cells. In Carcinogenesis, Vol. 2. Mechanisms of Tumor Promotion and cocarcinogen-
esis. edited by T.J. Slaga, A. Sivak. and R.K. Boutwell, Raven Press, New York, pp. 533-
543. 1978.
Hayflick, L. The cellular basis for biological aging. In The Handbook of the Biology of Aging,
edited by C.E. Finch and L. Hayflick. Van Nostrand Reinhold, New York, pp. 159-186,
1977.
Heidelberger, C., and S. Mondal. In vitro chemical Carcinogenesis. In Carcinogens: Identifica-
tion and Mechanisms of Action, edited by A.C. Griffin and C.R. Shaw, Raven Press, New
York, pp. 83-92, 1979.
Heidelberger. C., S. Mondal, and A.R. Peterson. Initiation and promotion in cell cultures. In
Carcinogenesis, VoL 2. Mechanisms of Tumor Promotion and Cocarcinogenesis. edited by
T.J. Slaga, A. Sivak. and R.K. Boutwell, and Raven Press, New York, pp. 197-202, 1978.
Higginson. J. Perspectives and future developments in research on environmental carcinogen-
esis. In Carcinogens: Identification and Mechanisms of Action, edited by A.C. Griffin and
C.R. Shaw, Raven Press, New York. pp. 187-208, 1979.
Hutt, P.B. Public policy issues in regulating carcinogens in food. Food Drug Cosmet. Law J.
33:541-557. 1978.
Interagency Regulatory Liaison Group. Work Group on Risk Assessment. Scientific bases for
identification of potential carcinogens and estimation of risks./. \ail. Cancer Inst. 63:241-
268, 1979.
Kakunaga, T. The transformation of human diploid cells by chemical carcinogens. In Origins
of Human Cancer, edited by H.H. Hiatt, J.D. Watson, and J.A. Winsten, Conferences on
Cell Proliferation, Vol. 4, Cold Spring Harbor Laboratory. Cold Spring Harbor. N. Y., pp.
1537-1548, 1977.
Kennedy. A.R.. S. Mondal, C. Heidelberger, and J.B. Little. Enhancement of X-ray transfor-
mation by 12-O-tetradecanoyl-phorbol-13-acetate in a cloned line of C3H mouse embryo
cells. Cancer Res. 35:439-443. 1978.
Krahn. D.F. Cultured mammalian cell transformation systems. In Strategies for Short-term
Testing for Mutagensi Carcinogens, edited by B.E. Butterworth and L. Goldberg. CRC
Press, West Palm Beach, pp. 55-66, 1979.
Lillehaug, J.R.. S. Mondal, and C. Heidelberger. Establishment of epithelial cell lines from
adult mouse regenerating livers. In Vitro 75:910-916. 1979.
MacPherson. L, and L. Montagnier. Agar suspension culture for the selective assay of cells
transformed by polyoma virus. Virology 23:291-294. 1964.
Marchok, A.C., and P. Nettesheim. In vitro growth characteristics of epithelial cell lines de-
rived from tracheal transplants exposed in vivo to 7. 12-dimethylbenz(a)anthracene
(DMBA). In Vitro 73:193, 1977.
Martin, C.X., A.C. McDernid, and R.C. Gamer. Testing of known carcinogens and noncar-
cinogens for their ability to induce unscheduled DNA synthesis in HeLa Cells. Cancer Res.
35:2621-2627, 1978.
Martin, G., and J.L. Anderson. Death and transformation. In Biology of Radiation Carcino-
genesis. edited by J.M. Yuhas. R.W. Tennant. and J.D. Regan. Raven Press. New York.
pp. 287-300, 1976.
McCann, J., E. Choi, E. Yamasaki, and B.N. Ames. Detection of Carcinogenesis as mutagens
in the Salmonella microsome test: assay of 300 chemicals. Proc. \ail. Acad. Sci. L'.S.A.
72:5135-5139. 1975.
Mondal, S.,andC. Heidelberger. Transformation of 10T1 2C1 8 mouse embryo fibroblasts by
ultraviolet irradiation and phorbal ester, \arure (London) 260:710-711, 1976.
57
-------
Montesano, R., L. Saint Vincent, C. Drevon, and L. Tomatis. Production of epithelial and
mesenchymal tumors with rat liver cells transformed in vitro. Int. J. Cancer /<5:55U-i;>o,
1975.
National Cancer Advisory Board. General criteria for assessing the evidence for carcinogemcity
of chemical substances: Report of the Subcommittee on Environmental Carcmogenesis,
National Cancer Advisory Board. J. Nail Cancer Inst. 58:461-465, 1977.
National Toxicology Program. Technical Bulletin, Vol. 1, Issue 2, edited by J.E. Huff, Public
Health Service, Department of Health and Human Services, Research Triangle Park, pp.
4-5, 1980.
Pienta, R.J. A hamster embryo cell model system for identifying carcinogens. In Carcinogens:
Identification and Mechanisms of Action, edited by A.C. Griffin and C.R. Shaw, Raven
Press, New York, pp. 121-141, 1979.
. Transformation of Syrian hamster embryo cells by diverse chemicals and corre-
lation with their reported carcinogenic and mutagenic activities. In Chemical Mutagens,
Vol. 6, edited by F.J. de Serres and A. Hollaender, Plenum Press, New York, pp. 175-202,
1980.
Pilot, H.C. Fundamentals of Oncology, Marcel Dekker, New York, 1978.
Poiley, J.A., R. Raineri, and R.J. Pienta. Two-stage malignant transformation in hamster em-
bryo cells. Br. J. Cancer 39:8-14, 1979.
Price, P.J., and N.K. Mishra. The use of Fischer rat embryo cells as a screen for chemical car-
cinogens and the role of the non-transforming type "C" RNA tumor viruses in the assay.
Adv. Mod. Environ. Toxicol. /:213-239, 1980.
Purchase, I.F.H. Validation of tests for carcinogenicity. In Molecular and Cellular Aspects of
Carcinogen Screening Tests, edited by R. Montesano, H. Bartsch, and L. Tomatis, Interna-
tional Agency for Research on Cancer, Lyon, pp. 343-349, 1980.
Purchase, I.F.H., E. Longstaff, J. Ashby, J.A. Styles, D. Anderson, P.A. LeFevre, and F.R.
Westwood. An evaluation of 6 short-term tests for detecting organic chemical carcinogens.
Br. J. Cancer 37:873-959, 1978.
Reznikoff, C.A., J.S. Bertram, D.W. Brankow, andC. Heidelberger. Quantitative and qualita-
tive studies of chemical transformation of cloned C3H mouse embryo cells sensitive to post-
confluence inhibition of cell division. Cancer Res. 33:3239-3249, 1973.
Rinkus, S.J., and M.S. Legator. Chemical characterization of 465 known or suspected carcino-
gens and their correlation with mutagenic activity in the Salmonella typhimurium system.
Cancer Res. 39:3289-3318, 1979.
Saffiotti, U., and C.C. Harris. Carcinogenesis studies on organ cultures of animal and human
respiratory tissues. In Carcinogens: Identification and Mechanisms of Action, edited by
A.C. Griffin and C.R. Shaw, Raven Press, New York, pp. 65-82, 1979.
Sanford, K.K. "Spontaneous" neoplastic transformation of cells in vitro: some facts and theo-
ries. Nat I. Cancer Inst. Monogr. 2(5:387-408, 1967.
Sivak, A. Overview and status of in vitro transformation. /. Assoc. Off. Anal. Chem. 62:889-
899, 1979.
Sivak, A., M.C. Charest, L. Rudenko, D.M. Silveira, I. Simons, and A.M. Wood. BALB/c-
3T3 cells as target cells for chemically induced neoplastic transformation. Adv. Mod.
Environ. Toxicol. 1:133-180, 1980.
Sivak, A., and A.S. Tu. Factors influencing neoplastic transformation by chemical carcinogens
in BALB/C-3T3 cells. In The Predictive Value of Short-term Screening Tests in Carcino-
genicity Evaluation, edited by G.M. Williams, R. Kroes, H.W. Waaijera, and K..W. vande
Poll, Elsevier, Amsterdam, pp. 171-190, 1980.
Steele, V.E., A.C. Marchok, and P. Nettesheim. An organ culture-cell culture system for study-
ing multistage Carcinogenesis in respiratory epithelium. In Carcinogenesis, Vol. 2, Mecha-
nisms of Tumor Promotion and Cocarcinogenesis, edited byT.J. Slaga, A. Sivak, and R.K.
Boutwell, Raven Press, New York, pp. 289-300, 1978.
Steuer, A.F., and R.C. Ting. Current methods for monitoring in vitro Carcinogenesis In Re-
cent Advances in Cancer Research: CellBiologv, Molecular Biologv, and Tumor Viroloev
Vol. 1, edited by R.C. Gallo, CRC Press, Cleveland, pp. 67-77, 1977.
Styles, J.A. Appendix III. Mammalian cell transformation in vitro. Br. J. Cancer 37-93 I 936
1978.
Weinstein, I.B., J.M. Orenstein, R. Gebert, M.E. Kaighn.and U.C. Stadler. Growth and st
tural properties of epithelial cell cultures established from normal rat liver and che ' II
induced hepatomas. Cancer Res. 35:253-263, 1975. ' y
Weinstein, I.B., M. Wigler, P.B. Fisher, E. Siskin, and C. Pietropaolo. Cell culture studies
the biologic effects of tumor promoters. In Carcmogenesis, Vol. 2, Mechanisms of T^ °"
58
-------
Promotion and Cocarcinogenesis, edited by T.J. Slaga, A. Sivak, and R.K. Boutwell,
Raven Press, New York, pp. 313-333, 1978.
Weinstein, I.B., H. Yamasaki, M. Wigler, L.S. Lee, P.B. Fisher, A. Jeffrey, and D. Grunberger.
Molecular and cellular events associated with the action of initiating carcinogens and
tumor promoters. In Carcinogens: Identification and Mechanisms of Action, edited by
A.C. Griffin and C.R. Shaw, Raven Press, New York, pp. 399-418, 1979.
Williams, G.M. The use of liver epithelial cultures for the study of chemical carcinogenesis.
Am. J. Pathol. 55:739-754, 1976.
. Review of in vitro test systems using DNA damage and repair for screening of
environmental carcinogens. J. Assoc. Off. Anal. Chem. 62:857-863, 1979.
APPENDIX 1
Transformation of Golden Syrian Hamster Embryo Cells By Chemicals of
Different Classes.*
Chemicals
Tested
Reported
Carcino- Transfor-
genicity mation
in wVof in vitro^
Reported
Mutagen-
icity in
Salmonella^
Direct Alkylating Agents
-Caprolactone
Glycidol
1,4-Butane sultone
1,3-Propane sultone
Methyl iodine
Dimethylcarbamyl chloride
Benzyl chloride
Glycidaldehyde
Ethyl-p-toluenesulfonate
1,2,3,4-Diepoxybutane
Propyleneimine
Polycyclic Aromatic Hydrocarbons and
Derivatives
Pyrene
Benzo(a)pyrene
Benzo(e)pyrene
Phenanthrene
3-Methylcholanthrene
Anthracene
Benz(a)anthracene
7,12-Dimethylbenzanthracene
1,2,3,4-Dibenzanthracene
1,2,5,6-Dibenzanthracene
Chrysene
1,8,9-Trihydroxyanthracene
Nitrosamines and Nitrosamides
/V-Nitrosodiethylamine
/V-Nitrosodimethylamine
/V-Nitrosoethylurea
/V-Nitrosopiperidine
/V-Nitrosodiphenylamine
/V-Methyl-/V'-nitro-/V-nitrosoguanidine
Methylazoxymethanol acetate
+§
+
+
+
+
+
59
-------
Aromatic Amines, Aminoazo Dyes,
and Nitro Compounds
1-Naphthylamine - +
2-Naphthylamine + + +
1-Anthramine - +
2-Anthramine + + +
Benzidine + + +
2,4-Toluenediamine + + +
2-Nitro-p-phenylenediamine + + +
4-Nitro-o-phenylenediamine U + f
m-Phenylenediamine U + +•
4-Methoxy-m-phenylenediame + + +•
4-Aminoazobenzene + +
4-Dimethylaminoazobenzene + <- +
3-Methoxy-4-aminoazobenzene + +§ +
2-Methyl-4-dimethylaminoazobenzene
3-Methyl-4-dimethylaminoazobenzene + + +
7,9-Dimethylbenz(c)acridine + + +
Acridine orange +1- +
Aniline
o-Chloroaniline - -
p-Chloroaniline
p-Rosaniline + +§
4-(o-Tolylazo)-o-toluidine + + t-
Azaserine +• + +
2-Aminobiphenyl
4-Aminobiphenyl +• + +
2,3'-Dimethyl-4-aminobiphenyl . HCI + + +
4-Nitrobiphenyl + + +
2-Nitronaphthalene t + +
2-Nitrofluorene + +§ +
4-Nitroquinoline-1-oxide <• + +
4-Hydroxylaminoquinoline-1-oxide + + +
2-Fluorenamine *• + +
/V-4-Acetylaminofluorene
/V-2-Acetylaminofluorene <• + +
A/-Hydroxy-2-acetylaminofluorene + + +
/V-Acetoxy-2-acetylaminofluorene <- + +
Bis(p-dimethylamino)-diphenylmethane + + -
Heavy Metal Salts
Lead acetate + f ?
Beryllium sulfate . 4H20 + + ?
Titanocene dichloride +• + ?
Nickel sulfate hexahydrate + + ?
Solvents
Acetone
Ethyl alcohol
Methyl alcohol
Dimethyl sulfoxide
Dimethyl formamide -
1,2-Propanediol - ;>
Miscellaneous Classes
Acetamide + +
Thioacetamide + +
Thiourea +• +
60
-------
5-lododeoxyuridine
5-Bromodeoxyuridine
5-Fluorodeoxyuridine
Bromobenzene
Hydrazine sulfate + +
1,2-Dimethylhydrazine + +
Safrole + +
Aflatoxin 61 +• +
Aflatoxin B2 t +
/V-[4-(5-Nitro-2-furyl)-thiazolyl]-formamide + +
2-(2-Furyl)-3-(5-nitro-2-furyl)-acrylamide + +
a-Napththylisothiocyanate
Saccharin +
Methyl carbamate
Ethyl carbamate (Urethan) + +§
Succinic anhydride + +
Auramine + +§
Ethionine + +
3-Amino-1,2,4-triazole + +
1-Phenyl-3,3-dimethyltriazene + +
Natulan . HCI (Procarbazine) + +§
12-O-Tetradecanoylphorbol-13-acetate P
Aroclor 1254 r -
Limonene P
Caffeine
Methotrexate
Hycanthone methanesulfonate + +
Sodium nitrite -
Hydroxylamine HCI
Methoxychlor .
'From Pienta (1979).
t+ = positive result; - = negative result; P = promoter; U = undertermined.
JData summarized by Pienta (1979) from McCann et al. (1975).
§Chemical transformed cells only when hamster liver S-9 homogenate was added to the
system.
61
-------
APPENDIX 2
Transforming Effects of Various Chemicals on Murine Leukemia Virus-
infected Fischer Rat Embryo Cells.*
Carcinogenic Transforming
Activity Activity
Chemical Class in wVof in
A zo Dyes
Aminoazobenzene
4-Dimethylaminoazobenzene +
4-Dimethylazobenzene +
2-Methyl-4-dimethylaminoazobenzene
3'-Methyl-4-dimethylaminoazobenzene +
6-(4-Dimethylaminophenylazo)quinoline +
Polycyclic Compounds
Anthracene
Naphthalene
Phenanthrene
Fluoranthrene
3-Methylcholanthrene t-
7,9-Dimethylbenz(c)acridine +•
Dibenz(a,/))acridine +
Dibenz(a,/')acridine +
Benzanthracene +
Dibenz(a,/?)anthracene +
7,12-Dimethylbenz(a)anthracene +
Benzo(a)pyrene +
Benzo(e)pyrene +
1,2,5,6-Dibenzanthracene +
Acromatic Amines
Aniline
/V-Acetoxy-/V-2-fluorenylacetamide +•
/V-Hydroxy-/V-2-fluorenylacetamide +
/V-2-Fluorenylacetamide +
2-Aminofluorene +
3-Aminofluorene ±
4-Aminofluorene
1-Naphthylamine
2'-Naphthylamine +
2-Aminobiphenyl
4-Aminobiphenyl +
0-Toluidine +
3,3'-Dichlorobenzidine ±
4-Nitroquinoline oxide +
Nitrosamines
Diethylnitrosamine +•
Dimethylnitrosamine +
Diphenylnitrosamine
A/-Methyl nitrosoguanidine +
Nitrosoethylurea +
Nitrosodiphenylamine +
62
-------
Miscellaneous Compounds
Glyceraldehyde 4- +
Dimethylcarbamylchloride + +
Methyl carbamate 4- 4-
Ethyl carbamate (urethane) 4- +
Methylazoxymethanol acetate f +
Propane sultone + +
Uracil mustard + 4-
Cyclophosphamide +• +
Thioacetamide + +•
Acetamide ± +
Formamide
Acetone
Propyleneamine + f
1,2-Epoxybutane + ±
Beryllium sulfate + +•
Lead acetate + 4-
Nickel sulfate + +
Natulan + 4-
Thiourea + 4-
*From Sivak(1979); Price and Mishra (1980).
t+ = positive result; - = negative result;
±= variable or borderline results.
APPENDIX 3
Comparative Activities of Various Chemicals as Evaluated by In Vitro Cell
Transformation and Five Other Short-Term Tests for Carcinogenicity.*
Compoundf
Acridine
2-Acetylaminofluorene
4-Acetylaminofluorene
Aflatoxin B
4-Aminoazobenzene
2-Aminobiphenyl
4-Aminobiphenyl
2-Aminochrysene
6-Aminochrysene
3-Aminopyrene
2-Aminonaphthalene-
1-sulfonic acid
Aniline
p-Anisidine
Anthracene
2-Aminoanthracene
Anthranilic acid
Anthraquinone
Chemicc
P
AA
AA
M
AA
AA
AA
AA
AA
AA"
AA
AA
AA
P
AA
AA
M
1.1 1 11 1
i- ^ — £ 0 " U>
CO -. CD CO CO 5 *
O • £ O t: OQ E Q
_ _
4-4-4-4-
- 4-
4- 4- + +
4-4-4-4-
4-4-4-
4-4-4-4-
4-4-4-4-
4-4-4-4-
4-4-4-4-
_ 4-
-
- 4-
- - 4-
4- 4- + 4-
-
-
?! a °'£. c
£ k S « ™
J2 D. t -i5 °-
w io K £ —
_
-
Nt Nt Nt
4-
4- 4-
+
+
+ 4-4-
4-4-4-
4- 4-
-
-
+
-
4-
-
4-
63
-------
Anthrone M - +
1,2-Benzanthracene P 4- 4- + 4- +• f ~
Benzanthrone M +
Benzidine AA + *• + + +
Benzimidazole P - - +
Benzole acid M
3,4-Benzpyrene p4-4- + 4-<--++
6-Benzoyl-2-naphthol M - + + +
Biphenyl M - *
Bis azo compound7 M - +
Bis(Chloromethyl)ether Alk t- +• + + + +
/V,A/'-Bis(2-naphthyl)-p
-phenylenediamine AA - *• - +
Butanesultone Alk <- + + + + + Nt
Caffeine M +
Calmagite1 M - + f
Camphor M - + + +
Carbazole P + -
Chlorambucil Alk + + + - + - Nt
Chloramine-T M
Cholesterol M
Colchicine M - - + + Nt
Croton oil M + + + + + +
Cyanocobalamin (B1Z) M - *
Cycasin acetate2 Alk + + + * +•
Cyclohexylamine M - t + -
Cyclophosphamide Alk + + + + +
3,3'-Diaminobenzidine AA 4- + +
2,7-Diaminofluorene AA + + + + +
3,4,5,6-Dibenzacridine P + + + 4-
1,2,3,4-
Dibenzanthracene p 4- + <• - + ++
3,4,9,10-Dibenzpyrene P + + t + + + 4-+
3,3'-Dichlorobenzidine AA + + + +
2,4-Dichloro-
phenoxyacetate M - - +
Dicyclohexylamine M +
DDT3 M
Dieldrin4 M Nt
Diethylnitrosamine Alk +• + + + f
Diethlstilbestrol M +• +
3,3'-Dimethoxybenzidine AA 1- + + *• + -
4-Dimethlaminoazo-
benzene AA + + + +• + +
9,10-Dimethylanthra-
cene p 4- 4- + 4- 4- +
p- Dimethyl ami nobenz-
aldehyde AA
7,9-Dimethylbenz-
acridine P 4- + + 4- + Nt
7,10-Dimethylbenz-
acridine P + 4- + 4- f Nt
9,10-Dimethyl-1,2-
benzanthracene P + 4- + 4-4-4-+4-
1,1 '-Dimethyl-4,4'-
bipyridinium
dichloride M +
3,3'-Dimethylbenzidine AA + + 4- 4- +
Dimethylcarbamoyl
chloride Alk 4-4-4- 4- _ ++
Dimethylformamide M
64
-------
Dimethylnitrosamine Alk + + _ + _ + _
2,3-Dimethylquin-
oxaline P - + +
Dinitrobenzene M - - +
2,4-Dinitrofluoro-
benzene M + + <-- +
2,4-Dinitrophenol M - +
Dinitrosopentamethly-
ene tetramine M - - - -
DL-Ethionine Alk + + + +
1,1'-Ethylene-2,2'-
bipyridinium dibro-
mide M + -
Ethylenethiourea M + f +
Ethyl methane-
sulphonate Alk + + + +
Hexachlorocyclohexane M - + - +
Hexamethylphosphor-
amide M + + +
Hydrazine M + + + +
Hydrocortisone M - + - +
Indole P - - + +
Merchlorethamine5 Alk + + + +
20-Methylcholanthrene P + + + + + +++
Methylene bis
(2-chloroaniline) AA + + + + + -
2-Methylindole P - + +
MNNG6 Alk + + + + + +
3-Methyl-4-nitroquin-
oline-/V-oxide AA -
Mitomycin C Alk + + + + + _
Morgan's base P + <- + - + - Nt
Naphthalene P -
1-Naphthol M - +
2-NaphthoI M - +
1-Naphthylamine AA -
2-Naphthylamine AA + + + + +
2-Naphthylamine-1,5-
disulfonic acid di-
sodium salt AA - +
Nitrobenzene M - - +
2-Nitrobiphenyl AA + + <- + _ Nt
4-Nitrobiphenyl AA f + + + +
2-Nitrofluorene AA + + + + Nt
4-Nitroquinoline-
/V-oxide AA + + + + + + +
/V-Nitrosodiphenyl-
amine M ____ + +
/V-Nitrosoephedrine Alk + + + <- + -
/V-Nitrosofolic acid Alk + + + _
4-Nonylphenol/ethylene
oxide condensate M - -
Orotic acid M - - +
Perylene P + +
Phenobarbital M - + +
/V-Phenyl-2-
naphtnylamine AA - + + +
Propanesultone Alk + + + + - + +
/3-Propiolactone Alk + + + + f
Resorcinol M - - - - -
Riboflavin M -
65
-------
Safrole M + + + + +
3,3',5,5'-Tetramethyl-
benzidine AA - +
Toluene M
Toluene-2,4-diisocya-
nete M +
2,4,5-Trichlorophen-
oxyacetate M - + - -
Trimethylphosphate M - _ + - - -Nt
Urethane M + + +• + +
Vinyl chloride Alk + +
From Purchase et al. (1978).
Compound:
(1) Calmagite: 2-hydroxy-1-(2-hydroxy-5-methylphenylazo) naphthalene-4-sul-
fonic acid.
(2) Cycasin (acetate): methyloxymethanol acetate.
(3) DDT: 1,1,1-trichloro-2,2-bis(4-chlorophenyl)ethane.
(4) Dieldrin: 1,2,3,4,10,10-hexachloro-6,7-epoxy-1,4,4a,5,6,7,8,8a-octahydro-
endo-1,4-exo-5,8-dimethanonaphthalene.
(5) Merchlorethamine: bis(2-chloroethyl)methylamine hydrochloride.
(6) MNNG: /V-methyl-/V'-nitro-/V-nitrosoguanidine.
(7) Bis azo compound: 2,2'-bis[1-(3-octadeclyaminopropylimino)ethyl]-2,2'-[3,3'-
dichloro-4,4'-biphenylylene)bis(azo)]bis(acetanilide).
fChemical class: AA = arylamines and related compounds; Alk = alkylating agents; P =
polycyclic aromatic hydrocarbons; M = miscellaneous compounds.
§Carcinogenicity in vivo: + = carcinogen; - = noncarcinogen.
||Cell transformation: growth of BHK-21 cells in soft agar.
#lmplant test: ++ = tumor at site of implant; Nt = not tested.
APPENDIX 4
In addition to the Federal Food, Drug, and Cosmetic Act (1938), a variety of other
laws has since come into being, conferring regulatory responsibilities on several
agencies (Appendix 4, Table 1).
To facilitate coordinations of regulatory policies, the CPSC, EPA, FDA and
OSHA in 1977 formed the Interagency Regulatory Liaison Group. In their 1979 Re-
port on short-term tests for carcinogenicity, they concluded:
The study of carcinogenesis at the cell level presently offers an effective
means to identify carcinogenic effects and mechanisms. In vitro mammali-
an cell transformation systems are simple models for the study of the
mechanisms of chemical and physical carcinogenesis.
As these systems become more widely used as test methods, they will
lead not only to better development and definition of screening techniques
but also to better understanding of the underlying mechanisms of
carcinogenesis.
Short-term tests for chemical carcinogens presently do not, in the ab-
sence of animal bioassays and epidemiology data, constitute definitive evi-
dence that a substance does (or does not) pose a carcinogenic hazard to hu-
mans. However, positive responses in these tests are considered suggestive
evidence of a carcinogenic hazard.
66
-------
Table A of Appendix 4. Major Legislation Providing Regulatory Authority
for the Control of Environmental Carcinogens.*
Statute Regulatory Agencyf
Federal Food, Drug, and Cosmetic Act June 25, 1938 FDA, EPA
Public Health Service Act P.L. 91 -696 (1912) FDA
Federal Insecticide, Fungicide, and Rodenticide Act P.L.
92-516(1947) EPA
Federal Hazardous Substances Act P.L. 92-516 (1972) CPSC
Occupational Safety and Health Act P.L 91-596 (1970) OSHA
Clean Air Act P.L 88-206 (1963) EPA
Poison Prevention Packaging Act P.L. 92-516 (1972) CPSC
Clean Water Act P.L. 95-217(1977) EPA
Consumer Product Safety Act P.L. 92-573 (1972) CPSC
Safe Drinking Water Act P.L. 93-523 (1974) EPA
Fair Packaging and Labelling Act P.L. 93-608 (1975) FDA
Resource Conservation and Recovery Act P.L. 94-580(1976) EPA
*From Dept. of Health and Human Services. First Annual Report on Carcinogens (1980).
fEPA = Environmental Protection Agency; FDA = Food and Drug Administration; CPSC =
Consumer Products Safety Commission; OSHA = Occupational Safety and Health Ad-
ministration, Department of Labor.
Such positive results also supply supporting evidence to positive animal
bioassays or epidemiology results. In some instances results from short-
term tests may conflict with animal bioassay data. If an animal bioassay
shows a positive response, it cannot be dismissed because a negative re-
sponse was observed in these tests. However, positive responses in such
short-term tests are ordinarily sufficient to provide suggestive evidence of
carcinogenicity, even if the substance tested has shown only negative re-
sponses in some animal bioassays. As the degree of certainty attached to
the negative responses in animal bioassays increases because the observa-
tion is reproduced in other animal species and strains or under more rigor-
ous test conditions, the suspicion about the chemical as a result of short-
term tests may be reduced and eventually eliminated. These conclusions
are in accord with those of the National Cancer Advisory Board's Subcom-
mittee on Environmental Carcinogenesis (1977);
At the present, none of the short-term tests can be used to establish
whether a compound will or will not be carcinogenic in humans or experi-
mental animals. Positive results obtained in these systems suggest exten-
sive testing of the agent in long-term animal bioassays, particularly if there
are other reasons for testing. Negative results in a short-term test,
however, do not establish the safety of the agent.
These conclusions are in accord with the preliminary results of an international
collaborative study, which have been interpreted as follows by the National Toxicol-
ogy Program (1980):
Short-term tests can be used to predict carcinogenic activity, but no sin-
gle assay or battery of assays is readily apparent as best suited for this pur-
pose.
Reliability of data from any assay system depends on thorough under-
standing of the system, awareness of pitfalls, and careful conduct of experi-
ments.
67
-------
Most assays have produced false negatives and false positives; thus, to in-
clude any assay in a battery of tests requires a tradeoff between these two
classes of errors.
Specific conclusions about test systems performance and relative utility
must await more detailed analysis of the data base.
Strong evidence for the use of a test battery has been obtained, yet recom-
mendations on test batteries must await a clear definition of their applica-
tion and the relative importance of false positive and false negative results.
68
-------
SHORT-TERM TESTS FOR IDENTIFYING
TERATOGENS*
Thomas H. Shepard
INTRODUCTION
The biological survival of the human race is dependent upon the success of three
major processes: (1) maturation of the female and male germ cells and their
favorable union (gametogenesis and fertilization), (2) intrauterine development of
the conceptus, and (3) postnatal growth and development. Teratology is the science
dealing with the causes, mechanisms, manifestations and prevention of intrauterine
developmental defects of either structural or functional nature. A teratogenic agent
(teratogen) may be a chemical, a drug, a virus, a physical condition or a deficiency
state.
Scope of Societal Problem
Incidence of Reproductive Loss—
Although it is recognized clinically that from 15 to 25 percent of human pregnan-
cies end in spontaneous abortion, the majority during the first trimester, there are
lines of evidence that over 50 percent of all conceptuses are lost. In what is now a
classic series of studies published in the 1940s and 1950s by Hertig, Rock and others
(see Hertig, 1967), a series of gravid uteri of less than 4 weeks' gestation were exam-
ined following hysterectomy. Because coital and menstrual dates were known, it was
possible to study the loss of the conceptus as related to actual gestational age. Histo-
ries could be compared with endometrial histology, acting as a check on menstrual
dating. The criteria in these studies for immediate or impending loss consisted of
severe pathologic features such as absence of inner cell mass or trophoblastic insuffi-
ciency for the stage of development.
The total percentage of conceptuses in the process of loss in these studies ranged
from 34 to 43 percent. However, when Hertig (1967) reconsidered the findings, a
more impressive rate of loss emerged. It was found that in any single menstrual cycle
15 percent of oocytes were unfertilized despite optimal conditions for fertilization to
take place. A further 10 to 15 percent were fertilized but failed to implant. Another
30 to 35 percent loss occurred during the first week following implantation.
Therefore, between 55 to 65 percent of oocytes were lost and only a total of about 40
percent of the starting group of women missed a menstrual period.
Recent studies utilizing highly sensitive chemical tests of pregnancy have been
able to detect a postimplantation conceptual loss of 41 percent among healthy
•Supported in part by the National Institute of Child Health and Human Development (HD00836).
The A uthor: Thomas H. Shepard is on the staff of the Central Laboratory for Human Embryology and the
Departments of Pediatrics and Obstetrics and Gynecology, University of Washington, School of Medicine,
Seattle. He is also affiliated with the University's Child Development and Mental Retardation Center and
Regional Primate Center.
69
-------
women (Williamson and Miller, 1980). Based upon the use of theoretic calculations
from the rates of chromosomal defects in spontaneous abortions, Boue'et al.(1975)
have estimated that 50 percent of all conceptuses may be lost from chromosomal
errors.
In this important general area of human biology, there is an immense need for
further information. The surveillance of spontaneous abortions is made very
difficult by the absence (in most cases) of good control or baseline data. Unfortu-
nately, with existing surveillance methods, the loss rate that is found is proportional
to the intensity of the inquiry. A more detailed review of this subject has been pub-
lished (Shepard and Fantel, 1979).
Incidence of Congenital Defects—
Approximately three percent of all human newborns have a congenital anomaly
requiring medical attention, and one-third of these conditions can be regarded as life
threatening. With increasing age, over twice as many congenital defects are detected.
Close to one-half the number of children in hospital wards are there because of pre-
natally acquired malformations of one kind or another. Another three percent of our
population is mentally retarded, although a certain part of this problem is caused by
postnatal factors. A congenital defect, whether structural or functional, exists for an
individual's entire life span. It is not a year or two, or ten years. In terms of human
suffering and the financial and social dislocation of the family, this toll is
tremendous. Historically, after identifying the causative mechanism, the opportunity
for prevention in this field has proven to be outstanding. Two examples of this are
the near prevention of kernicterus by preventing maternal immunization to Rh fac-
tor and the congenital rubella syndrome by mass immunization.
Our knowledge of the cause and prevention of human malformations is extremely
limited in that approximately 70 percent are of unknown causes. About 10 percent
are associated with gene mutations, five percent with chromosomal aberrations and
less than 10 percent are known to be due to the specific teratogenic agent. Although
there are more than 600 agents known to produce congenital anomalies in experi-
mental animals, less than 25 of these are known to cause defects in the human
(Schardein, 1976; Shepard, 1980). The reasons for this apparent wide discrepancy
between the numbers of animal and human teratogens are discussed more fully later
in the chapter, but are particularly related to the immense doses which investigators
are able to give experimental animals.
HISTORICAL PERSPECTIVE
There are many archaeologic artifacts which show that men and women of many
countries and continents have been fascinated by monstrosities. Babylonian tablets
give various congenital defects which were thought to be prognostic of the political
and economic future of nations. A long history of how maternal impressions may
cause congenital anomalies has been recorded; this idea, in part, is at the root of
many of the guilt feelings still found in parents producing defective offspring.
A more descriptive aspect of teratology, along with early biologic inquiry, began
in the 19th century and continued through the early 20th century. Some of the scien-
tific descriptions by Meckel, the Saint-Hilaires, Taruffi, Schwalbe and Ballantyne
rival or better those of present day. Further details of this fascinating history can be
obtained from Warkany (1971).
Bridging between the ancient historical concept of teratology and the present-day
experimental approach, there was a period of time when most malformations were
considered as being the result of gene mutations. With the discovery by Hale (1937)
and by Warkany and Nelson (1940) that vitamin deficiencies could cause defects in
experimental animals, a new impetus to the use of experimental methods arose In
1960 the Teratology Society was founded, and some years later teratology societies
were established in Europe and Japan.
70
-------
The thalidomide tragedy of the early 1960s gave further impetus to experimental
teratology, and, in particular, pointed out the unique teratogenic dangers of our en-
vironment. Government as well as the pharmaceutical industry in particular
responded by beginning to test agents in small animals for their teratogenic poten-
tial. In the past few years with the appearance of the Ames test for mutagenicity,
much pressure has been applied to produce a similar short-term tool for teratololgy.
PRESENT STATE OF TERATOLOGY
Principles
Relationship between Teratogenic Susceptibility and Stage of Development—
To a large extent, teratogenic susceptibility depends upon the period of develop-
ment when the conceptus is exposed to an agent (Fig. 1). Damage during the implan-
tation and presomite periods (0-15 days) generally produces little altered morpho-
genesis because the ovum either dies or regenerates completely, whereas, during
major organ formation (18-60 days), the embryo is highly sensitive and exposure
may produce significant morphologic changes. During the subsequent period, the
fetus is less sensitive to morphologic alterations, but changes in functional capacity,
such as intellect, reproduction or the rate and process of aging, may develop in later
postnatal life. This time specificity has been found in nearly all cases where terato-
genesis in the human has been proved and studied in detail.
Species Variation in Response to Teratogenic Agents—
Some species of animals are much more susceptible to specific teratogenic agents
than others. Aspirin, cortisone and several vitamin deficiencies are highly terato-
genic in the rodent, but there is no solid evidence of their teratogenicity in humans.
The thalidomide epidemic would not have been prevented by testing prenatal mice
and rats, but the drug is teratogenic in rabbits, monkeys and humans. It seems likely
that there are certain common embryologic processes between experimental animals
and man. Various pharmacologic and physiologic mechanisms that operate to con-
trol the maternal blood concentration of a given teratogen are shown in Fig. 2. In ad-
dition, interspecific differences exist in placental transport and embryonic metabo-
lism. One developing hypothesis is that the species variability is due in large part to
variations in biotransformation of the administered drug. The hepatic microsomal
monooxygenase (mixed-function oxidase) system plays a role in the metabolism of
more than 70 percent of drugs and chemicals. The rate of breakdown or the type of
byproduct produced by this system in different animals might help to explain the
species specificity.
Drug Quantity and Teratogenicity—
Any drug given in large enough amounts will adversely affect fetal development.
This action usually occurs through deleterious effects on maternal health and is ex-
pressed as either embryo-fetal death, fetal growth retardation or osseous retarda-
tion. Both sodium chloride and sucrose given in sufficient amounts to experimental
animals will produce these embryo-fetotoxic effects. Many of the warnings in drug
inserts about potential teratogenicity are related to this phenomenon. When
extrapolating the dose-response curve from animal experiments to humans, it is im-
portant to take into consideration the ratio (on a per kilogram basis) between the ter-
atogenic dose in the animal and the therapeutic or exposure dose in the human (see
Setlow, this volume). Figure 3 illustrates the usual relationship between dose and the
zones for teratogenicity, embryo-fetal lethality and maternal lethality.
71
-------
Functional maturation —-3
| Organogenesis
Embryonic period Fetal period
Entire developmental span
Figure 1. Curve illustrating the susceptibility to teratogenesis from fertilization
throughout intrauterine development (from Wilson, 1973, used by permission).
\ Large or
) highly charged
\ molecules cross
Kidney
Bile
Placental
transfer
Hemeostatic
dispersal
Figure 2. Diagram of the factors that influence embryonic dose of a foreign chemical
present in the maternal blood stream. Major individual and species differences
in absorption and homeostatic dispersal control the blood concentration to
which the placenta is exposed. Variations in placental transfer exist between
species but very little work has been done on interspecific differences in
drug metabolism by the embryo (from Wilson, 1973, used by permission).
72
-------
f Embryotoxic range ^
100%
50%
Increasing dosage
CD No effect zone II Teratogenic zone [13 Embryolethal zone HJMaternalethal zone
Figure 3. Diagram showing that each drug tested has a dose spectrum ranging
from no effect to one of maternal lethality. The teratogenic zone may be
broad, very small or, in some cases, nonexistent (adapted from Wilson,
1973). Naturally occurring rates in untreated animals have been corrected
for by subtraction from all three zones.
Multifactorial Causes of Congenital Defects—
According to their etiology, congenital defects can be divided into three general
types: (1) those due to a single action of a mutated gene (for example, achondroplasia
or phenylketonuria); (2) those due to the single action of an environmental agent (for
example, congenital rubella); or (3) those produced by a combination of several or
more gene defects and/or environmental agents. These multifactorial causes proba-
bly account for the major proportion of congenital defects. Congenital dislocation
of the hip and most forms of cleft lip and palate are commonly mentioned as exam-
ples of multifactorially caused congenital malformations. Fraser and his students
have illustrated this mechanism in certain inbred strains of mice and have begun to
make practical human applications (Fraser, 1969). The A/J mouse exhibits a higher
natural incidence of cleft lip and palate than do certain other strains; this is because
of differences in the topographic relationship of the embryonic facial processes. Oth-
er genes controlling the development of the palate are known to predispose to clefts.
Examples such as shortening of the head, changes in mandibular length or mecha-
nisms by which tongue obstruction might prevent normal palatal shelf closure have
been studied using mutant genes in inbred animal models. Most of the environmen-
tal agents (for example, aspirin or cortisone) known to produce clefts are more effec-
tive in these inbred strains. It seems imperative to accelerate the application of these
principles to man. This requires quantitative methods for identifying the susceptible
human genotypes and the environmental agents that might contribute to the multi-
factorially caused defects.
Experimental Animal Tests
Common practice is to select for testing two or three small animal species, usually
rats, mice and rabbits. The dosage regime usually employed is as follows: the maxi-
mum dose used is usually in considerable excess, such as 100 times human therapeu-
tic dose; a fractional dose of the maternal LD-50 (the dose required to kill 50 percent
of maternal population); and a minimal dose, preferably one that causes minimal
maternal toxicity. Since nearly all agents can produce some embryo-fetotoxicity ex-
73
-------
pressed as weight loss or minor defects at some level, it has been recommended that
this level be determined and taken into consideration in establishing the minimum
dosage.
Three general time periods for administering the test agent have been recom-
mended (Kelsey, 1974). The first is a general test for reproductive effects where both
the male and female are started on the drug before pregnancy. Midway during preg-
nancy, one-half of the females are sacrificed and the number of corpora lutea of
pregnancy (ovulated eggs) and the number and state of implantation sites are
examined. A decrease in ovulation, fertilization or implantation as well as an in-
crease in early embryonic death can be determined. The remainder of the pregnant
females is allowed to go to term and to litter. The offspring are examined and reared,
and subsequently rebred to measure any intergenerational change in reproduction.
A second type of experiment calls for administration of the compound during
varying periods of major embryonic organ development and then sacrifice of the
mother just before parturition to measure and examine the fetus for visceral and
bone defects. A third approach is to administer the drug during the latter part of
pregnancy and to observe the offspring in the perinatal and postnatal period. A more
complete discussion of these tests is available (Wilson, 1975; Palmer, 1978).
Embryonic growth and placental function in the human are more like those in
subhuman primates than in the rodent and other small animals. Although it would
seem that these primate species would best approximate the human, their cost, the
long duration of pregnancy and the availability of sufficient numbers of pregnant
animals mitigate against their use as a standard animal. Testing in these species may
be indicated for agents which must be used during human pregnancy (antihyperten-
sives, hypoglycemics, anticonvulsants) or for agents which are likely to be inadvert-
ently taken during pregnancy. They may be used for testing agents about which a
question of safety arises after regulatory approval for widespread exposure.
In Vitro Tests
Study of Ovigenesis and Spermatogenesis—
With the identification of the very exacting nutritional requirements for mamma-
lian ova by Brinster (1975), it became possible to perform in vitro studies of the ef-
fects of specific emvironmental contaminants. For instance, Brinster and Cross
(1972) using mouse ova exposed to copper were able to establish a dose response and
a lethal concentration. This type of analysis, applied to other chemicals and coupled
with the knowledge of intrauterine concentrations found in exposed workers, could
be a great help in establishing safe exposure levels. An approximation of uterine
levels in exposed women might be obtained by analysis of curettage tissue removed
for unrelated health reasons.
The in vitro tests of mammalian eggs have been applied to the study of abnormal
chromosome division (Donahue, 1975). There is evidence that accumulated
radiation exposure or delayed ovulation may result in increased chromosome imbal-
ances in the human ova (for discussion and references see Shepard, 1980).
Sperm counts, along with the study of their morphology and motility, have been
used in monitoring men at risk. Unfortunately, these parameters are not well corre-
lated with fertility. A new test which appears to be better associated with the state of
fertility utilizes the frequency with which human sperm penetrate hamster eggs in
vitro (Rogers et al., 1979). Infertility associated with hypospermia has been linked to
workplace exposure to the pesticide, l,2-dibromo-3-chloroPropane (DBCP)
(Whorton et al.. 1977). At airborne concentrations between 0.4 to 1.0 ppm of DBCP,
chronically exposed workers had reduced sperm concentrations and were infertile!
74
-------
Cell Culture—
Individual cells are grown in much the same way as bacteria. All the organization
of the original tissue in such cultures is lost, and cell multiplication and growth in
uniform populations are the dominant interests. The technique is very useful in pre-
dicting alteration in cell replication and has been used by some laboratories to screen
new chemicals and drugs. Such a routine test led to the discovery of the important
biologic action of cytochalasins by S.B. Carter (1967) of Imperial Chemical Indus-
tries, Ltd. namely, that these chemicals prevented cellular, but not nuclear, division.
More specific questions about the growth of cells could be answered by using special
cell lines (teratocarcinoma, neuronal, and endocrine) or cultures of different tissues
such as fetal lung or myocardium.
Organ Culture—
Small pieces of differentiating organs or limb buds removed from experimental
animals or from human abortion material can be maintained for a number of days in
culture. During this time growth and differentiation can be assessed during exposure
to environmental toxicants. Mainly cartilaginous or endocrine tissues and, to a lesser
extent, renal and cardiac rudiments have been used for this technique. Explanted
thyroid tissue from human abortuses has been used to determine the period of devel-
opment when iodide can be concentrated. This is an important question since thera-
peutic amounts of I31I, if administered to a thyrotoxic woman, can be concentrated
by the fetal thyroid and cause its destruction. This test was done by adding radioac-
tive iodide to the growth medium and measuring its incorporation into the explant
(Shepardand Pious, 1978). Concentration commenced in thyroid tissue from fetuses
of 70 to 75 days of age.
Embryo Culture—
Mammalian embryo culture has been feasible since 1967 when D.A.T. New dem-
onstrated normal growth of presomite- and somite-stage embryos maintained in a
serum medium (New, 1978). A major attraction of this method is that a dose-re-
sponse curve can be established and the no-effect level may prove valuable to regula-
tory agencies faced with the problem of safety standards. A number of agents have
been studied including hyperthermia, ethanol, trypan blue, B vitamins, anesthetics,
alkylating agents and autonomic stimulating drugs. A suggestion has been made
that the serum of at-risk workers could be tested with this system (Chatot et al., 1980).
Many drugs and chemicals require bioactivation by liver enzymes (the P450 mon-
ooxygenase system). Fantel et al. (1979) have shown that this fraction of liver can be
added directly to the embryo-culture system and will activate teratogenic activity. In
Fig. 4, the use of this bioactivation system with cyclophosphamide, an antitumor
agent, is illustrated. In addition to expanding the number of chemicals which show
in vitro activity against embryos, possibly species variation (see second principle)
may be better understood by using maternal P450 systems from teratogenically sen-
sitive and insensitive species.
Other Short- Term Tests—
Many new test schemes are being described and evaluated. Since most of these are
in preliminary stages, the reader is referred to the abstracts of the Teratology Society
(1980, 1981, 1982). Several reviews on the use of various in vitro tests in teratology
have appeared (Ebert and Marois, 1976; Wilson, 1978, Shepard and Pious, 1980).
Although the main use of in vitro techniques has been to learn about mechanisms of
action, there has been the expectation that a simple, inexpensive screen for
predicting teratogenic activity in humans could be developed. Unfortunately, this is
unlikely since the basic initiating causes of teratogenic activity involve many biologic
processes which include cell replication and cleavage, cell differentiation, muscle
75
-------
innervation, as well as cell adhesion, motility and interaction (Moscona, 1975). The
complexity of these causes of altered development is given in Fig. 5.
Even though these short-term tests may eventually be helpful in predicting terato-
genicity in humans, most teratologists believe that testing of pregnant animals will
continue to be necessary; otherwise there is little doubt, as has been sadly learned,
that the ultimate test subject for a new agent may continue to be the human.
In vitro test for teratogenicity
Bioactivation of teratogenic drugs
Remove sites on day 10 (early somite stage)
Drug
Drug + S-9
or Drug + cofactors
Drug + S-9
+ cofactors
Grow for 24 hours - Rotator - 5% 02-37°C
Normal
Normal
Defects,
Decrease in
growth and protein
Figure 4. Diagram giving an example of an in vitro test of a teratogen, using a growing
rat embryo. The drug used was activated by addition of a P450 monoxygenase
liver fraction (S-9) to the medium. In the bottle containing drug alone and in the
bottles containing S-9 alone or cofactors alone, the embryos grew in a normal
fashion over a 24-hour period. In the bottle containing the complete system (drug,
S-9 and cofactors), there was a dose-response relation with decreased growth,
protein content, and an increase in malformations. This example was drawn from
experiments using cyclophosphamide (Fantel et al., 1979).
76
-------
CAUSES
Action of an environmental
agent on embryo
Radiation
Chemicals
Infections
Hypoxia
Physical trauma
Drugs
Dietary imbalance
MECHANISMS
Reaction(s) within embryo
Chromosomal nondisjunction
Mutation
Lack of precursors, substrates
Altered energy sources
Enzyme inhibition
Changed membrane
characteristics
MANIFESTATIONS
Pathogenesis, initiated as one
or more of the following:
Cell death
Failed cell interactions
Reduced biosynthesis
Impeded morphogenetic
movement
Altered differentiation
schedules
Intrauterine death
Malformation
Growth retardation
Functional deficit
Figure 5. Diagram showing the relationship between cause, mechanism and
manifestation of a teratogenic agent (adapted from Wilson, 1973).
77
-------
PRESENT DEFENSES AGAINST TERATOGENIC AGENTS
It is possible to envision these existing defenses .against teratogenicity as walls or
hurdles (Fig. 6). The standard teratogenicity testing in pregnant laboratory animals
would represent part of the first barrier. Created only after the devastating effects of
thalidomide in the early 1960s this defense is held by a few to be effective, but the ma-
jority of scientists recognize its severe limitations. Animal tests of new drugs may
have prevented the introduction of new human teratogens. Various aspects of these
tests have been described above. The use of chemical structure and pharmacologic
activity (also a part of this first defense) as a predictor of teratogenicity is theoreti-
cally of great promise, but to date, has seen little practical application except, per-
haps, in revealing the teratogenic effect of chemotherapeutic agents. Since extreme
variations in drug clearance exist between individuals but not between identical
twins, it is very probable that an individual's pharmacologic defenses are under
strong genetic control. This knowledge could lead to identification and protection of
women who might be more teratogenetically vulnerable than others to common en-
vironmental exposures. Pharmacologic variations in the metabolism of diphenylhy-
dantoin might account for the small proportion of women in which the drug is tera-
togenic.
The second defense, early monitoring of the fetus and newborn, is necessary be-
cause the first defenses are still inadequate. Although fetal monitoring and amni-
ocentesis are able to reduce the number of severely crippling genetic and morpho-
genetic defects, the disadvantage of this form of monitoring is that it is after concep-
tion, and many people believe that therapeutic abortion is not a morally acceptable
tool. One would expect that prenatal loss might be a more dramatic and sensitive in-
dex of response to teratogenic agents. By study of the spontaneous losses during the
first trimester, an earlier warning system might be established that would provide
epidemiologic information some six months before the effects of a teratogenic agent
would be detected in newborns. Another advantage to the study of spontaneous
abortuses is that it shortens the period from the time of maternal teratogenic ex-
posure to the time of inquiry. This period may be less than a week as compared to the
seventh- to eighth-month time interval involved when a history is obtained from the
mother of a newborn.
Monitoring facilities for defects in neonates exist in many countries. Generally,
only easily recognized physical defects are recorded; the larger portion of congenital
diseases (60 percent) is identified only after the newborn period and so is not
included. Minor changes in brain function or long-term carcinogenesis would be
missed. Although this monitoring system produces variable results because of
artifacts associated with data collection, a continuous registration of time and place
of congenital defects should provide an important warning of teratogenic action by a
newly introduced agent, whether chemical, physical or infectious. In the United
States, a computerized system recording gross congenital defects from 1500 hospitals
(with an average of one million births per year) is maintained by the Center for
Disease Control and Commission of Professional Hospital Activities.
Later monitoring (defense 3) is illustrated best by the discovery of vaginal carcin-
omas in young women exposed in utero to diethylstilbestrol. The carcinomas are
thought to develop because of misplacement of genital epithelium during embryo-
genesis. This defense, later monitoring, is manned generally by the alert medical
practitioner. In Appendix 1 is listed sources of information available on reproduc-
tive hazards.
More than 2,000 new chemicals are synthesized or otherwise produced each year,
and as many as 200 of these may find their way into the human environment in
measurable amounts. These cannot all be tested. In Appendix 2, I outline the
selection of agents to be tested by government agencies and industry, and the way in
which the dangers posed by teratogens are categorized.
78
-------
Defense III
late monitoring
Congenital defects
, S&UM^ ~
[Newborni
9
Defense II
early monitoring
Defense I
testing
Figure 6. Perspective of our major defenses against teratogenic agents. Although the
in vitro test wall is currently used after animal tests, it should precede animal
testing in the future (from Shepard, 1974, used by permission).
Interpretation of Epidemiologic Findings
There is evidence that about one-half of all human conceptions terminate in spon-
taneous abortion, and the reported incidence of reproductive waste increases in pro-
portion to the sophistication and care of the interviewers or methods used in the
study. This means that any form of publicity will increase the recorded losses and
give the false impression of an epidemic.
Factors That Increase False Positive Findings—
There are many preliminary published reports that associate drugs or other agents
with congenital malformations. The conclusions drawn from these reports are either
of borderline significance or are produced by artifacts of data collection,
questionable statistical analysis or both. This incomplete imformation, released
through news reports or drug package inserts, may lead to panic in exposed pregnant
women and their doctors. Subsequent, detailed studies which often do not support
the original report are either omitted or given only brief mention. For example, sev-
eral lay publications have created concern and even panic in the minds of women
using a common antinauseant during pregnancy, through biased and inaccurate re-
porting, while considerable experience with this drug has supported the opinion by
most experts that the drug is not teratogenic. Positive rather than negative associa-
tions are naturally more likely to be written up by the investigator, and also more
likely to be accepted by scientific journal editors.
Regardless of how hard one tries to be impartial, there is a distinct tendency to be
overenthusiastic about positive associations and at the same time to be less rigorous
and critical when searching for malformations in the control groups. Collecting
79
-------
positive and interesting associations can easily create a snowball effect in that the ini-
tial observation, after being broadcast, tends to attract other case associations; in
this manner, the size of the exposed population remains untabulated and
consequently the true risk is not assessable. Double-blind studies are nearly impossi-
ble to carry out since the person who is being interviewed will generally mention the
presence of his health problems.
Other false positive biases that occur in the collection of data include improved re-
call of events by parents of malformed children; increased defect rates found by
multiple observers (for example, the hierarchy of examiners in the university hospi-
tal where at-risk pregnancies are likely to be referred); and an increased rate found
when a full autopsy is performed. These last two factors are especially important
when considering rate conditions such as diabetes which tend to be seen more often
in large medical centers. As an example, the newborn of the diabetic mother is more
likely to die and complete autopsy examinations can inflate the number of malfor-
mations in this group.
By use of computers, statistical analyses of many hundreds of associations between
agents and particular malformations can be performed easily. The great majority of
these associations are made without a prior hypothesis. By the laws of chance alone a
small proportion, perhaps five out of each 100, will fall into what is considered to be
statistically significant. These results are reported frequently in the literature and
cause another increase in false positive associations.
Another serious drawback in epidemiologic studies is the extremely large number
of cases needed to prove teratogenicity at a certain probability level. This is especially
true where only limited numbers of women are susceptible to the teratogenic action
of a drug. Given the extreme genetic heterogeneity of the human population, there is
always the possibility that a single individual or a small number of individuals will
have a heightened susceptibility that could lead to excessive accumulation of a
relatively safe drug or agent or of a toxic metabolite and consequently, to embryo-
fetal toxicity.
Several improvements in the collection and analysis of data have been made re-
cently. In large studies it is often possible to match the affected mother to a cohort
with similar parity, age and social class. In the analysis of data from the Collabora-
tive Perinatal Project*, a mathematical model was developed to adjust for con-
founding variables that are related both to rates of drug usage and to malformations
(Heinonen et al., 1977). Another technique that controls memory bias is the use of a
control group of parents who have offspring with types of congenital defects that are
known to be unrelated to the drug under study. For example, a study of an agent
may include parents of infants with Down's syndrome as controls. Another more
expensive, but not foolproof way, is illustrated by the work of Milkovich and Van
den Berg (1974) which linked computer-stored prescription data on mothers to the
findings in their newborns. Some interesting positive associations for meprobamate
and chlordiazepoxide were found, but another large prospective study based upon
history taking failed to confirm them (Hartz et al., 1975). It is of interest that in the
computerized study of Milkovich and Van den Berg (1974) all of the significant
prescriptions were written by physicians treating women for conditions other than
pregnancy, while none were given by the obstetrician. This highlights the fact that
mucl^of the teratogenic period of pregnancy may occur before definite diagnosis of
pregnancy is made.
*A large study earned out by the National Institute of Neurological Diseases and Stroke.
80
-------
FUTURE NEEDS FOR DEVELOPMENT OR EXPANSION
Nearly all the recommendations in this section are dependent upon our ability to
validate and transmit information. Cooperation of the entire spectrum of our society
from the school-age child through pregnant women, administrators and scientists is
needed to compile and digest this information (Appendix 1).
Short-Term Tests
Small-animal testing for effects on reproduction by new and other untested
agents must be continued even though they are not completely predictive for humans.
Although a number of biologic systems for short-term tests in vitro are mentioned
above, none has yet met all the criteria needed fora prescreen. These criteria are: (1)
simplicity, (2) lack of expense, (3) rapidity, (4) based upon multiple mechanisms and,
most importantly, (5) accurate confirmation of known and future human ter-
atogens. The culture of mammalian embryos in association with activation of com-
pounds by maternal enzymes seems to come closest at present to filling these criteria.
The growth period of these cultures needs to be extended beyond embryogenesis in
order to study late developmental aspects (palate and skeletal formation) and the
effect on the chorioallantoic placenta.
The development of short-term tests for teratogenicity is hampered by the exis-
tence of multiple mechanisms which contribute to the molecular and pathogenic
expression of abnormal development (see Fig. 5). Unlike mutagenesis, and probably
carcinogenesis, the cellular targets include other organelles as well as the nucleus.
Linkage or Association between Environmental Agents and Human
Disease or Disabilities
In the past 50 years, disease states with short incubations have been fairly easily
linked with causative factors. There is now a challenge to associate early exposures to
long-term health events. For instance, women who smoke heavily appear to enter
menopause at an earlier age than non-smokers. Is this also true for women who have
their only exposure to smoking during prenatal life? Are there any prenatal determi-
nants to the onset and course of atherosclerosis and hypertension in old age? What
percent of childhood malignancies are causally related to prenatal exposures? Beside
diethylstilbestrol and possibly ionizing radiation, there is some evidence appearing
which suggests that diphenylhydantoin may be a transplacental carcinogen.
Triangulation from Existing Data Bases
Triangulation is a navigational technique which allows a traveler to plot his loca-
tion and subsequent course by determination of his position in respect to certain
known points such as stars or coastal markers. Similarly, scientists are finding that
the cause and prevention of certain disease states can be determined by linkage of
three fixed, but expanding data bases (Fig. 7). These data bases can be labeled as uni-
versal identifiers since they are world-wide. They consist of (1) individuals, (2)
agents, and (3) disease syndromes. At birth, an individual should receive a unique
health identification number to be used throughout life for all health-related records.
Such a system, must, of course, incorporate proper safeguards for protection of pri-
vacy. An example of how this system could be useful would be the identification and
recording of the health numbers of offspring of women exposed to workplace agents
of unrealized long-term toxicity during pregnancy. There is a reasonable chance that
some existing workplace toxicants will be identified later as agents which can cross
the placenta and initiate changes which, over a long term, will produce cancer or
other pathology in the offspring. Such a multigenerational identification system
could foster public health measures to protect the unborn young of exposed preg-
nant women.
81
-------
Data bases for
triangular attack on
congenital defects
Individual
(parent-child)
Agent
(Chemical, physical,
genetic)
Mechanisms
\
Prevention
Clinical
Identification
Figure 7. Diagram of a scheme whereby three data bases (individual, agents and
clinical syndromes) might be used to determine the mechanisms which
cause congenital defects and their prevention.
The second universal identifier system, causative agents, is partly in place since
unique numbers are being assigned to chemicals (Chemical Abstracts Service (CAS)
Registry Numbers). There is a delay in the assignment of CAS numbers to many
physical and infectious agents.
The area of congenital syndrome identification (syndromology) is providing a
great deal of useful and important data. An example of the utility of specific syn-
drome identification is the fetal alcohol syndrome which appears to account for a
major part of our population in which a specific cause of mental retardation can be
assigned. McKusick's Catalog of Mendelian Inheritance in Man (1978) represents a
good updated annotated system for labeling human mutations and their associated
syndromes.
Some disease states have been controlled without knowledge of their cause
because of the discovery of effective treatment. An example of this is the surgical cor-
rection of congenital pyloric stenosis. Congenital rubella is partly controlled by im-
munizing young women, but we do not know the exact mechanism by which the fetus
is in some cases protected by natural defenses. Since we lack the intimate knowledge
of mechanisms of pathogenesis, these two congenital syndromes are still not com-
pletely preventable.
The concept of linkage of these three data bases was proposed and more fully
discussed in a public document produced for federal legislators, Human Health and
the Environment—Some Research Needs (Nelson and Whittenberger, 1977).
Identification and Investigation of Teratogenic Outbreaks
The Life Span of an Epidemic—
The time course of an epidemic can be plotted in sequence from the introduction of
the causative agent through hypothesis testing to removal of the agent (Fig. 8). The
time course of the thalidomide epidemic in different countries varied from two years
to over four years. The process was lengthened by the then slow international
82
-------
Life span of an epidemic
Recognition
of
epidemic
Rorschach
effect
multiple
hypotheses
Etiologic
agent
introduced
Hypothesis
testing
Acceptance by
scientific,
government
and lay
Cause communities
associated
Exposure
stopped
resolution
of epidemic
Time
Figures. Events between the introduction of a new agent causing widespread
disease and the ending of the epidemic. The vertical plot indicates the
number of cases. The Rorschach stage is named after the psychiatric test
where seemingly unstructured ink blots take on meaningful shapes when
studied. The puzzling circumstances associated with an early outbreak
are equally without meaning until hypotheses are generated and studied
(modified from Oakley, 1976).
exchange and assimilation of scientific reports and by delay in mechanisms leading
to removal of the drug from the market. Much improvement has since occurred in
these two areas. An example of the slow association between a clinical syndrome and
an agent is that of the oral anticoagulants (Warkany, 1976). These dicoumarin anti-
coagulants were associated with two case reports of infants born with very small
noses in 1968. The rare linkage between the use of anticoagulants during pregnancy
and the unusual reduction in the size of the nose should have signaled an association
or at least a concerted search for further supportive evidence. Instead, it was not
confirmed until 1973 when additional associations were finally brought to light at a
meeting dealing with malformation syndromes. Parenthetically, it is of interest that
an animal model for this human teratogen does not exist. The association between
alcoholism and the fetal alcohol syndrome was delayed literally for centuries. These
examples emphasize the continued need for sound monitoring of our population in
order to make epidemics visible against the background of malformations and also
for exchange of information between experimental teratologists, clinical
syndromologists and regulatory agencies.
A Proposal for Triage, Investigation, and Management of Toxicologic Outbreaks—
Since such a system should combine all toxicologic dangers (organ-specific
carcinogenesis, mutagenesis and teratogenesis), the phrase toxicologic is used here.
An executive secretary and small permanent staff should be located geographically
and administratively near to the National Toxicology Program, Center for Disease
Control or other federal body. These resource people should be consulted by tele-
phone for purposes of triaging new reports of hazards and for appointment of ad hoc
committees to validate, investigate and take action on toxicants. Active members of
relevant research societies should be called on to help in these processes.
83
-------
ACKNOWLEDGMENTS
Thanks are due to Ms. Barbara Brownfield for her help with editing and producing
the manuscript. The author thanks his coworkers and several teratologist friends for
their original ideas incorporated into this chapter. Drs. Alan Fantel, Philip Mirkes
and Trent Stephens offered valuable criticisms.
The manuscript is respectfully dedicated to Dr. James G. Wilson in thanks from a
student and admirer.
REFERENCES
Boue, J., A. Boue', and P. Lazar. Retrospective and prospective epidemiological studies of
1,500 karyotyped spontaneous human abortions. Teratology 72:11-26, 1975.
Brinster, R.L. Teratogen testing using preimplantation mammalian embryos. In Methods for
Detection of Environmental Agents That Produce Congenital Defects, edited by T.H.
Shepard, J.R. Miller, and M. Marois, North-Holland, Amsterdam; American Elsevier,
New York, pp. 113-124, 1975.
Brinster, R.L., and P.C. Cross. Effect of copper on the preimplantation mouse embryo. Nature
(London) 235:398-399, 1972.
Carter, S.B. Effects of cytochalasins on mammalian cells. Nature (London)213:2d\-2M, 1967.
Chatot, C.L., N.W. Klein, J. Piatek, and L.J. Pierro. Successful culture of rat embryos on
human serum: use in the detection of teratogens. Science 207:1471-1473, 1980.
Donahue, R.P. Chromosomal anomalies and the meiotic divisions of the oocyte. In Methods
for Detection of Environmental Agents That Produce Congenital Defects, edited by T.H.
Shepard, J.R. Miller, and M. Marois, North-Holland, Amsterdam; American Elsevier,
New York, pp. 125-139, 1975.
Ebert, J., and M. Marois. (eds.) Tests of Teratogenicity in vitro. Proceedings of the Woods
Hole Conference 1975, North-Holland, Amsterdam; American Elsevier, New York, 1976.
Fantel, A.G., J.C. Greenaway, M.R. Juchau, and T.H. Shepard. Teratogenic bioactivation of
cyclophosphamide in vitro. Life Sci. 25:67-72, 1979.
Federal Register 44 (124):37462-37466, June 26, 1979.
Fraser, F.C. Gene-environment interactions in the production of cleft palate. In Methodsfor
Teratological Studies in Experimental Animals and Man, edited by H. Nishimura, J.R.
Miller, and M. Yasuda, Igaku Shoin Ltd., Tokyo, pp. 34-49, 1969.
Gorlin, R.J., J.J. Pindborg, and M.M. Cohen, Jr. Syndromes of the Head and Neck, 2nd ed.,
McGraw-Hill Book Co., New York, 1976.
Hale, F. Relation of maternal vitamin A deficiency to microphthalmia in pigs. Tex. State J.
Med. 33:228-232, 1937.
Hartz, S.C., O.P. Heinonen, S. Shapiro, V. Siskind, and D. Slone. Antenatal exposure to
meprobamate and chlordiazepoxide in relation to malformations, mental development and
childhood mortality. New Engl. J. Med. 292:726-728, 1975.
Heinonen, O.P., D. Slone, and S. Shapiro. Birth Defects and Drugs in Pregnancy, Publishing
Sciences Group, Littleton, Mass., 1977.
Hertig, A.T. The overall problem in man. In Comparative Aspects of Reproductive Failure,
edited by K. Benirschke, Springer-Verlag, New York, 1967.
Hunt, V.R. Occupational Health Problems of Pregnant Women. A report and recommenda-
tions for the Office of the Secretary of the Department of Health, Education and Welfare,
1975. Copies available from C.G. Schiffer, Rm. 5139, H EW-North, 330 Independence Ave.
S.W., Washington, D.C., 20201.
Kelsey, F.O. Present guidelines for teratogenic studies in experimental animals. InCongenital
Defects: New Directions in Research, edited by D.T. Janerich, R.G. Skalko, and I.H.
Porter, Academic Press, New York, pp. 195-204, 1974.
McKusick, V.A. Mendelian Inheritance in Man. 5th ed.. The Johns Hopkins University Press,
Baltimore, 1978.
Milkovich, L., and B.J. Van den Berg. Effects of prenatal meprobamate and chlordiazepoxide
hydrochlonde on human embryonic and fetal development. New Engl. J. Med. 297:1268-
Moscona, A. A. Invited discussion: Embryonic cell and tissue cultures as test systems for terato-
genic agents. In Methodsfor Detection of Environmental Agents That Produce Congenital
84
-------
Defects, edited by T.H. Shepard, J.R. Miller, and M. Marois, North-Holland, Amsterdam;
American Elsevier, New York, pp. 103-107, 1975.
Nelson, N., and J.L. Whittenberger. Human Health and the Environment-Some Research
Needs. Report of the Second Task Force for Research Planning in Environmental Health
Science, DHEW Publication No. NIH 77-1277, 1977.
New, D.A.T. Whole embryo explants and transplants. In Handbook of Teratology, Vol. 4, ed-
ited by J.G. Wilson, and F.C. Fraser, Plenum Press, New York, pp. 95-133, 1978.
Oakley, G.P., Jr. Birth defect surveillance in the search for and evaluation of possible human
teratogens. Birth Defects: Orig. Artie. Ser. 12 (5): 1-6, 1976.
Palmer, A.K. The design of subprimate animal studies. In Handbook of 'Teratology, Vol. 4, ed-
ited by J.G. Wilson and F.C. Fraser, Plenum Press, New York, pp. 215-253, 1978.
Rogers, B.J., H. Van Campen, M. Ueno, H. Lambert, R. Bronson, and R. Hale. Analysis of
human spermatozoal fertilizing ability using zona-free ova. Fertil Sleril. 52:664-670, 1979.
Schardein, J.L. Drugs as Teratogens, CRC Press, Cleveland, 1976.
Shepard, T.H. Teratogenicity from drugs—an increasing problem. Disease-a-Month, June: 1-
32, 1974.
A Catalog of Teratogenic Agents, 3rd ed., Johns Hopkins University Press, Bal-
timore, 1980.
Shepard, T.H., and A.G. Fantel. Embryonic and early fetal loss. Clin. Perinatal. (5:219-243,
1979.
Shepard, T.H., and D. Pious. Cell, tissue and organ culture as teratologic tools. In Handbook
of Teratology, Vol.4, edited by J.G. Wilson and F.C. Fraser, Plenum Press, New York, pp.
71-93, 1978.
Smith, D.W. Recognizable Patterns of Human Malformation, 2nd ed., W.B. Saunders Co.,
Philadelphia, 1976.
Staples, R.L. Teratogens and the Delaney Clause. Teratology 10: 1, 1974.
Teratology Society. Teratology 21: A1-A77, 1980.
Warkany, J. Congenital Malformations: Notes and Comments, Year Book Medical Publish-
ers, Chicago, 1971.
Warfarin embryopathy. Teratology 14:205-209, 1976.
Warkany, J., and R.C. Nelson. Appearance of skeletal abnormalities in the offspring of rats
reared on a deficient diet. Science 92:383-384, 1940.
Whorton, D., R.M.Krauss, S. Marshall, and T.H. Milby. Infertility in male pesticide workers.
Lancet 2:1255-1261, 1977.
Williamson, E.M., and J.F. Miller. A prospective study into early conceptual loss. Clin. Genet.
17: 93, 1980.
Wilson, J.G. Environment and Birth Defects, Academic Press, New York, 1973.
Critique of current methods for teratogenicity testing in animals and suggestions
for their improvement. In Methods for Detection of Environmental Agents That Produce
Congenital Defects, edited by T.H. Shepard, J.R. Miller, and M. Marois, North-Holland,
Amsterdam; American Elsevier, New York, pp. 29-48, 1975.
Feasibility and design of subhuman primate studies. In Handbook of Teratology,
Vol. 4, edited by J.G. Wilson and F.C. Fraser, Plenum Press, New York, pp. 255-273, 1978.
Wilson, J.G., and F.C. Fraser (eds.) Handbook of Teratology, Vol. 1-4, Plenum Press, New
York, 1978.
APPENDIX 1:
SOURCES OF INFORMATION ON REPRODUCTIVE HAZARDS
Publications
An excellent general introduction to and discussion of teratology is available in
Wilson's Environment and Birth Defects (1973). The annotated A Catalog ofTera-
togenic Agents (Shepard, 1980) includes about 1100 agents that have been studied
for teratogenesis in animals and man. Schardein's 1976 book on teratogenic drugs is
another useful source. Handbook of Teratology edited by J.G. Wilson and F.C.
Fraser (1978), a comprehensive set of 4 volumes, summarizes many general and
specific topics as well as techniques important in teratology.
85
-------
Heinonen et al. (1977) reported on the outcome of more than 50,000 pregnancies in
which drug exposures were known and listed their risk rates. Hunt (1975) has ad-
dressed problems of pregnant women in the work place. Congenital Malformations:
Notes and Comments (Warkany, 1971), Recognizable Patterns of Human Malfor-
mations (Smith, 1976) and Syndromes of the Head and Neck (Gorlin et al., 1976)
help to define and extend the description of human defect syndromes. McKusick's
Mendelian Inheritance in Man (1978) includes more than 3000 congenital syndromes
produced by gene mutations.
Computerized Information Retrieval
The National Library of Medicine (NLM) maintains a number of data bases in-
cluding the computer service, Toxline, which lists references dealing with toxic
properties. These listings are on line to most regional medical libraries. The
chemicals are stored by CAS numbers (Chemical Abstracts Service Registry Num-
bers), and in one or two days it is possible to have a listing of articles dealing with the
reproductive toxicologic aspects of certain chemicals. The Environmental Tera-
togen Information Center (ETIC), maintained by the National Institute of Environ-
mental Health Services, contributes to this service (Toxline) and also maintains its
own answering service at Research Triangle Park, N.C.
Answering Services for Public and Professionals
Individuals and groups dealing with exposed persons often find computerized sys-
tems described above too slow or insufficiently detailed to provide data on reproduc-
tive dangers. These questions often deal with the advisability of pregnancy termina-
tion or transfer of pregnant women in the work place. Pertinent texts may not be
available to those asking questions. As a result, a number of telephone answering
services are maintained by teratologists who can either direct the caller to recent liter-
ature or to other professionals. Names and addresses of members of the Teratology
Society, along with their geographic location, are given in each December's issue of
the journal, Teratology.
Exchange of Scientific Information
A large and inaccessible area of scientific data is that of negative findings. Pharma-
ceutical companies often fail to release results of safety tests of their new products. In
addition, scientific journals do not publish many manuscripts on agents which are
not toxic in some way. In Japan, industries and others are encouraged by the
government to publish their safety tests and at least one journal, Oyo Yakuri,
publishes these results.
Scientific exchanges with other countries are recognized as important. As a result
of this, the computerized retrieval of literature references on toxicologic substances
(Toxline) of the National Library of Medicine is available through a number of out-
lets in Western Europe, Central and South America and Africa. The library tapes are
also accessible in Japan and Australia. These countries contribute in a significant
way to collection and validation of these data bases, particularly with scientific arti-
cles originating in their own countries. Printing of texts by computer and other
means of reducing the publication time and cost of scientific information should be
fostered.
Training
Most medical students have only modest exposure to toxicology and over one-half
of the medical schools in the United States have no specific courses or trained faculty
lor teaching teratology. It is interesting that those schools with very strong programs
in developmental biology seem to be the weakest in teratology, which in itself is a
form of applied developmental biology. Human embryology should be reinstated as
a medical school requisite.
86
-------
APPENDIX 2:
SELECTION OF AGENTS TO BE TESTED
Government Selection
A number of existing federal regulatory agencies have their own investigative ca-
pabilities. The larger of these agencies include the Center for Disease Control (CDC),
Environmental Protection Agency (EPA), Food and Drug Administration (FDA),
National Cancer Institute (NCI), the National Institute of Child Health and Human
Development (NICHD), National Institute of Environmental Health Sciences
(NIEHS) and the National Institute of Occupational Safety and Health (NIOSH).
Other departments of government, such as Transportation, Labor, Energy, Defense
and Agriculture, also foster some teratologic work. The National Center for Toxi-
cologic Research (NCTR) in Arkansas, funded through the FDA and EPA, has a
mandate to develop new tests for toxic substances and, in particular, to do long-term
low-dose experiments in large numbers of experimental animals (referred to by some
as "mega mouse" studies). All of these agencies choose agents which they believe to
be important to their regulatory or mandated programs. A recent development in
toxicologic testing has been the formation of a National Toxicology Program (NTP)
for selecting and testing chemicals and developing new test methods. A chemical
selection unit located at NCTR prepares summaries of data on agents which may be
nominated by government, industry or private individuals. The assigning of priori-
ties and testing of these chemicals is performed under the NTP.
The Toxic Substances Control Act (TSCA) charges the EPA to identify and test a
specific number of chemicals and some of these tests will be of teratogenic nature.
Training money for toxicologists was included in this legislation.
Industry and Academia
Many pharmaceutical and larger chemical producers have elaborate facilities for
teratologic testing of their products. Although many of these results are published,
some degree of secrecy enters due to the competitive nature of the enterprise. The
faculty of the universities and other research units contribute many of the new obser-
vations and techniques to teratology as well as acting as consultants to both indus-
try and government.
Categorization of the Danger of Teratogenic Agents
Some methods of regulating exposure to known or suspected teratogens have
been developed. The FDA has published five categories of danger for drugs that may
be taken by pregnant women (Federal Register, 1979). The classificatioin is based
upon animal and human exposure and takes into account the absence of toxicologic
information (Table 1). The FDA recommends that where data are available they be
summarized and referenced in the accompanying package insert for the drug.
The International Agency for Research in Cancer (IARC) of the World Health
Organization has started to review reproductive and teratologic data and assess the
hazards of selected chemicals. This organization produces many useful publications
based upon ad hoc committees drawn from experts in the field of carcinogenesis,
mutagenesis and teratogenesis.
Since the Delaney Clause was passed by Congress in 1958 forbidding the use of
known animal carcinogens in food substances, a similar exclusion of teratogens was
considered. Because nearly every agent in sufficiently large doses can produce
embryo-fetal toxicity, it seemed unwise to have a similar regulation for teratogens.
In one of the very few political activities taken by the Teratology Society, the mem-
bership was surveyed on this question and the overwhelming opposition was
recorded in several scientific journals (Staples, 1974).
87
-------
Table 1. Categories of Danger of Teratogenic Agents (Federal Register,
1979).
Category
a
b
c
d
X
Results of
Animal Study
Negative
Positive
Positive
Negative or
unknown
Positive
Results of Human
Pregnancy Exposure
Negative
Unknown
Negative
Unknown
Positive
Positive
Use During
Pregnancy
Use if necessary
Use if necessary
Use if necessary
Use if the benefit
outweighs the
risk
Use for life-
threatening
situations only
Contraindicated
for use: risk
clearly outweighs
benefit.
-------
SHORT-TERM TESTS FOR RENAL TOXICITY
David S. Miller
and
Anthony M. Guarino
INTRODUCTION
In "Seven Gothic Tales," Dinesen (1972) defined man as a ingenious machine de-
signed to turn, with "infinite artfulness," red wine into urine. Human kidneys pro-
cess about 1600 liters (400 gallons) of blood per day and produce about one liter of a
highly specialized concentrated fluid known as urine. The kidney accomplishes some
of the most important functions necessary for survival: it excretes waste products of
metabolism and xenobiotics (chemical substances foreign to the body); it regulates
levels of salts, water and, to some extent, nutrients in body fluids; it maintains the
appropriate acid-base balance of blood; and it serves as an endocrine organ (secre-
ting hormones such as erythropoietin, renin, and prostaglandins). The elaborate
physiologic functions that the kidney performs require a high degree of structural
and functional complexity. Certain widespread and chronic environmental pollu-
tants—for example, heavy metals—clearly affect renal function, and simple,
short-term tests of renal toxicity are needed to protect human health. In order to
design appropriate tests, we must possess detailed knowledge of normal renal physi-
ology and biochemistry so that appropriate indices of renal function can be estab-
lished. Also, we must recognize the sites and mechanisms of action of nephrotoxic
chemicals so that sensitive targets can be identified. Ideally, in both cases, our knowl-
edge would extend from the whole organ down to molecular levels of organization.
Such is not the case at present. Indeed, little attention has been focused on the devel-
opment of simple, widely applicable tests for renal toxicity, although some fairly so-
phisticated new procedures, such as those using isolated cells and subcellular frac-
tions, show great potential as short-term tests.
KIDNEY FUNCTION
In order to appreciate fully the difficulties encountered in designing test systems
for renal toxicity studies, one must consider the complex anatomy, biochemistry and
phsiology of the normal kidney. The vertebrate kidney contains renal tubules, blood
vessels, lymphatics and interstitial tissue, and each plays a role in the various pro-
cesses that ultimately result in the production of urine. The composition of final
urine is a result of the (1) functional interactions between anatomical elements; (2)
water and solute transport characteristics of the cellular membranes of glomerular,
The Authors: David S. Miller and Anthony M. Guarino both worked at the Mount Desert Island Biological
Laboratory, Salsbury Cove, Maine. Dr. Guarino is also affiliated with the National Center for Drugs and
Biologies, U.S. Food and Drug Administration, Rockville, Maryland.
89
-------
vascular and tubular tissue; and (3) control mechanisms both from within the kidney
and from without (see reviews by Orloff and Berliner, 1973; Brenner and Rector,
1976).
The human kidneys are paired, bean-shaped organs that are located in the rear of
the peritoneal cavity, alongside the vertebral column. Each kidney is supplied by a
single renal artery and drained by a single renal vein and a ureter. Two distinct
regions can be identified: a darker outer region, the cortex, and a pale inner region,
the medulla. The structural and functional unit of the kidney is called the nephron,
and consists of the glomerulus-Bowman's capsule, proximal tubule, loop of Henle,
distal tubule and collecting duct and their respective blood supplies (Fig. 1). There
are about one million nephrons in each kidney. The glomerular apparatus and
proximal and distal segments lie within the renal cortex; the loop of Henle and col-
lecting ducts lie within the medulla.
Urine formation by the kidney involves the movements of water and solutes
through selectively permeable cellular membranes. The permeability characteristics
vary with the region of the nephron and are a direct result of the arrangement of
lipids and specialized proteins within cellular membranes. These proteins perform
structural, enzymatic (metabolism or transport) or receptor functions. Since the lipid
core of the membranes has a hydrophobic character, small water soluble molecules,
such as ions, sugars and amino acids, can only pass slowly through the lipid regions
of the membrane, but move more readily through protein-lined aqueous pores or
with the aid of specific carrier proteins. Certain substances may be"pumped"across
the membrane against their electrochemical potential ("uphill" transport) by
transport proteins that utilize metabolic energy stored in adenosine triphosphate
(ATP) or in transmembrane ion gradients. Thus, an exothermic process (splitting of
ATP by transport ATPase or "downhill" transport event) is used to drive a coupled
endothermic process (uphill cellular accumulation of K+, glucose or amino acids).
At the molecular level, evidence indicates that certain widespread environmental
contaminants, such as heavy metals, can selectively modify passive and metabolically
driven permeability characteristics of renal cell membranes, thus impairing the
membranes' ability to conserve proteins and nutrients, excrete metabolites and
maintain salt and water balance.
Blood Supply
Total blood supply to the kidneys is extremely high; in a 70 kg man the 300 g of
renal tissue receives 20-25 percent of the cardiac output. Under normal conditions,
the intrarenal distribution of blood flow is not uniform; it decreases progressively
from the outer cortical to the inner medullary regions. Thus, the outer regions of the
kidney arc better oxygenated than the inner regions and the proportion of oxidative
to glycolytic metabolism of the cells within these regions reflects the differences in
blood supply. In each kidney, the renal artery divides several times to form smaller
arteries, arterioles and eventually one or more capillary beds at the glomerulus and
around the renal tubules. This double capillary bed arrangement is unusual, but not
unique to the kidney. From the tubular capillary beds, blood flows into the renal
venous system and returns to the systemic circulation at the inferior vena cava.
Glomerular Filtration
One anastamosing capillary bed is contained within the Bowman's capsule. The
outer walls ol'the porous capillary endothelium are lined with a basement membrane
which, in turn, is covered by the epithelium that lines the urinary space. The entire
apparatus functions as a submicroscopic sieve, producing an ultrafiltrate of blood in
the urinary space (Fig. 2). Ultrafiltration is powered by hydrostatic pressure in the
glomerular capillaries, a direct result of work done by the pumping heart. The ultra-
filtraie in Bowman's space is nearly identical in composition to plasma (cell-free
90
-------
blood), except for the nearly complete absence of proteins. The glomerular sieve per-
mits some low molecular weight proteins to pass into the urinary space, but these
proteins are generally reabsorbed in the proximal tubule, so that, normally, the final
urine is essentially protein-free. Pathological conditions which might affect blood
pressure (hypertension) or decrease plasma oncotic pressure (caused, for example,
by liver disease) clearly can alter glomerular filtration rates; increased filtration pres-
sures may alter the ultrastructure and filtration characteristics of the glomerular
sieve.
Arcuate
artery
Glomerulus
Bowman's
capsule
Proximal tubule
Distal
tubule
Collecting
tubule
Loop of
Henle
Thin
segment
Vasa
recta
Pelvis
Figure 1. Diagrammatic representation of the relationships between the component
parts of the nephron and associated blood vessels. Actual spatial
relationships have been greatly simplified. Adapted from Guyton (1976).
91
-------
Capillary endothelial cell
I
Basement membrane
1
Fenestra
Slit-pore
Figure 2. Simplified diagram of the glomerular filtration apparatus. Adapted
from Guyton (1976).
Tubular Reabsorption and Secretion
The proximal tubule modifies tubular fluid primarily by (1) reabsorbing essential
substances and either utilizing them intracellularly or passing them on to the peritu-
bular capillary nets; and (2) clearing the peritubular circulation of excretory pro-
ducts and secreting them into the tubular fluid (Fig. 3). In the proximal tubule,
nearly all the filtered organic nutrients are reabsorbed, most by specific, metabol-
ically powered transport proteins that are embedded in the luminal plasma
membrane. Two-thirds of the filtered NaCl and water are reabsorbed in this segment,
92
-------
Glomerulus
Proximal tubule
Ultrafiltration
NO+,C|-,H2O,HCO-3, K+,PO4-3
sugars, amino acids,
proteins, uric acid
H + , organic acids,
and bases
\
Loop of Henle
NaCI
Distal tubule
Na+,CI-
H2O
H+.K+
NH4+
Collecting duct
Na+.Ch
H20
NH4+
Figure 3. Simplified summary of the movement of water and solutes in the different
regions of a model nephron. Arrows show direction of the transfer from or to
luminal fluid. Modification of the composition of luminal fluid as it passes
through each segment of the nephron is indicated.
93
-------
as are substantial amounts of K+, Ca2+, and phosphate ions, whilst H ions are secret-
ed into the urine to maintain the body's acid-base and nitrogen (NHs) balance. NaCl
transport across the tubular epithelium is driven by an ATPase (Na, K-activated
ATPase) and the accumulation of NaCl near the basal and lateral surfaces of the
cells draws water from the tubular lumen towards the interstitial spaces (osmotically
coupled salt and water transport). Proximal tubular secretion of organic acids and
bases removes normal metabolites, drugs and other xenobiotic compounds and their
metabolites from the blood.
The loop of Henle is a consistent anatomical feature of the kidney of all animals
that produce a urine more concentrated than their body fluids. Because of the
hairpin bend at the end of the loop and the arrangement of passive and metabolically
driven salt transport processes within the loop, this portion of the nephron functions
as a countercurrent multiplier, creating an osmotic gradient in the tubules, interstitial
fluid and blood vessels (Fig. 4). The primary interstitial solute is urea and the
osmotic gradient increases deeper into the medulla. This gradient is the driving force
for the removal of water from the fluid in the collecting ducts thus producing a con-
centrated final urine. The osmotic gradient, hence, the final urine concentration, can
be modified at the "loop" level by diuretic drugs and certain heavy-metal compounds
which act on specific transport proteins in the cellular membranes of the thick as-
cending limb.
Tubular fluid that passes from the ascending limb of Henle to the distal tubule is of
low osmolality (Fig, 4). In the distal segment, NaCl and water are reabsorbed and K+,
H+ and N H/ are secreted. Dilute urine entering the collecting ducts then passes once
more through the mudulla where the reabsorption of water occurs. This process is
promoted by an antidiuretic hormone (ADH) from the pituitary which increases the
permeabilities of the collecting duct epithelium to water and urea. Water is drawn
osmotically from the collecting-tubule fluid by the steep medullary osmotic gradient
created by the countercurrent multiplier system in the loop of Henle. Urea, the major
solute of the final urine, equilibrates with the medullary interstitium. In well-hy-
drated individuals, circulating ADH levels are low and a dilute urine is produced. In
dehydrated individuals, circulating ADH levels are high, and a more concentrated
urine is produced (Fig. 4).
RENAL TOXICOLOGY
Target Organ and Critical Organ Concepts
Over the past decade, increasing attention has been focused on the detection and
evaluation of toxicity as manifested by alteration of individual organ function. One
must recognize the individual role each organ plays in maintaining homeostasis, that
is, constant extracellular and intracellular levels of ions, nutrients, excretory pro-
ducts and gases, so that enzymes, cells, tissues, and whole organisms can function
optimally. At each organizational level, the role of excretory organs, especially the
kidney, is pivotal (see reviews byOppelt, 1970; Beliles, 1975; Foulkesand Hammond,
1975; Maher, 1976; Foulkes, 1977; Magos and Clarkson, 1977). Through homeo-
static mechanisms the organism usually is able to adapt to environmental changes.
Indeed, it is remarkable that the kidneys have a reserve that allows normal renal
function even when compromised to greater than 50 percent of original capacity.
One early clue to the potential site of toxicity of a particular agent is its localization
in certain cells and organs; this is a signal to focus further on the association between
tissue accumulation and adverse effects. Particularly in studies of heavy-metal con-
taminants, it has often been shown that undesirable functional changes appear to-
gether with renal accumulation of the toxicant. These effects are generally associated
with a critical concentration of toxicant within the cell; increases beyond this critical
level cause cell death, thus, defining the lethal concentration for that cell Analogous
terminology is applied to organs, with the critical organ concentration being the
94
-------
1200 _
900 _
600-
300
200 _
100 _
0-1
Proximal
tubule
Collecting
tubule
Figure 4. Concentration of total solutes in tubular fluid as a function of segment
of the nephron. Increases in osmolarity indicate net secretion of solutes into
the tubular fluid; decreases indicate net reabsorption of solutes. Iso-osmotic
movements of fluid are not shown. The shaded area indicates that portion of
the final urine osmolarity that is under ADH control. Adapted from Guyton
(1976).
mean concentration in the organ at the time that any of its cells reach their critical
concentration. The concept of a critical organ refers to that particular organ which
first attains its critical concentration of the pollutant; the organ of greatest
accmulation is not necessarily the critical organ.
It is rare to find a chemical which affects the function of a single organ. Overall
toxicity must be considered to be a result of the complex relationships between expo-
sure conditions (uptake, distribution, and excretion of toxicant) and cellular and
nutritional factors which might affect toxicity in several organs. Furthermore,
subcellular partitioning among various organelles and intracellular fluids may also
be of considerable importance. For example, it is known that specific proteins whi-'h
95
-------
have preferential affinity for some metals, such as cadmium, are present in cells.
These proteins bind the metal so that the cadmium is unable to interact with normal
cellular constituents; this may constitute a major, natural detoxification mechanism.
Because of these relationships, it is currently extremely difficult to formulate
practical standards for exposure or tissue levels with regard to renal toxicity alone.
Indeed, interspecific and individual differences in normal biochemical and
physiological processes, as well as in patterns of toxicant uptake and distribution, can
be so great as to make one organ critical for one individual and a different organ crit-
ical for another. In setting standards in certain areas, the Environmental Protection
Agency (EPA) has therefore placed in high-risk categories individuals who might be
extra sensitive to certain pollutants. These considerations show that it is of impor-
tance to determine which organs and which groups of cells within them are most sus-
ceptible.
The kidney is clearly one important target of chemical toxicity because (1) renal
blood-flow rates are especially high, thus exposing tissues rapidly to a large fraction
of circulating toxicant; (2) filtered toxicants may be concentrated in tubular fluid as
salts and water are reabsorbed, exposing luminal membranes to high concentrations
of toxicants; (3) secreted toxicants may be concentrated in the interstitium, tubular
cells and tubular fluid; and (4) renal tissue contains many chemically reactive sites,
such as sulfhydryl groups, that are functionally important.
Nephrotoxic agents can be divided into those that affect the four major morpho-
logic components of kidney: glomeruli, tubules, interstitial tissues, and blood vessels.
This approach can be useful in that the early manifestations of lesions affecting each
of these sites tend to be distinct. Further, there often is site specificity such that
glomerular lesions are most frequently mediated through immunologic mechanisms,
whereas tubular and interstitial effects are more commonly associated with toxic or
infectious agents. On the other hand, it must be emphasized that because of the
anatomic interdependence of the components of the kidney, damage to one com-
ponent will invariably induce secondary changes to the others.
Glomerular Toxicity
Few environmental chemicals affect the glomerulus compared with the large num-
ber which affect the other three sites. Nonetheless, there are some agents which can
cause glomerulonephritis as a result of immunologic-type reactions. A few drugs
such as hydralazine, phenylbutazone, and sulfonamides cause glomerulonephritis,
and a number of biologic products can also institute acute renal failure by this mech-
anism. Vaccines, plant poisons, and antigens, elaborated as a result of drug-induced
hypersensitivity, induce anaphylactic allergic reactions and, thus, a proportion of
their nephrotoxicity is likely to be caused through glomerular lesions.
Tubular and Interstitial Toxicity
Toxicities to the tubules tend to involve local cytotoxic action, enzyme inhibition
and osmotic disruption. The structure of renal tubular cells varies throughout the
nephron, and Correlates somewhat w.ith the functional role of the tubular segment in
question. Thus, the proximal tubular cells have a highly developed structure, with
long microvilli, many mitochondria and extensive intercellular interdigitations,
which are correlated with the major transport functions of this segment, namely re-
absorption of about 70 percent of filtered sodium and water as well as a large
proportion of filtered glucose, potassium, proteins, phosphate and amino acids.
There is considerable evidence that a sodium pump is located in the basal lateral
membranes of the proximal tubules, whose operation is mediated through the mem-
brane-bound Na,K-ATPase and the adjacent mitochondria (which supply the re-
quired energy through oxidative phosphorylation). Against this specialized
morphologic and biochemical background, it is no surprise that the proximal tubule
96
-------
is highly vulnerable to a large number of xenobiotics. Furthermore, since a number
of toxicants are reabsorbed or secreted by the proximal tubule, cells may be exposed
to particularly high levels of toxicants.
The interstitium is an additional component of the kidney which can be affected
by nephrotoxic agents. It can be considered as those tissue structures remaining after
discussion of the glomeruli, tubules, and blood vessels; the interstitial space contains
capillaries, lymphatics, and some fibroblast-like cells. Interstitial fluid contains the
solute gradients which are essential for the countercurrent production of urine. Thus,
the roles of the tubule and interstitium are highly cooperative and they both are the
common sites of the effects of toxic xenobiotics.
Tubular and Interstitial Nephrotoxins
We will briefly discuss current knowledge on the renal effects and mechanisms of
toxicity for a selected group of nephrotoxic substances. A more complete listing is
given in Tables 1 and 2.
Analgesics—
While of no direct concern to EPA, this class of nephrotoxic drugs is mentioned
for two reasons. First, the lesions are quite typical of those seen with a number of
other agents. Second, since these drugs are used extensively, one must anticipate that
actual nephrotoxicity in a particular individual may be the result of exposure to not
only environmental contaminants but also to analgesics and other drugs. Thus, we
must consider substances in the working environment and in food and water, as well
as nephrotoxic solvent exposure from hobby activities. The actual lesions induced
by analgesics involve both the tubules and interstitium; they are classified as
tubulointerstitial toxic agents. Morphologically, there are cortical changes, consist-
ing of loss or atrophy of tubules, interstitial fibrosis, and inflammation, which seem
to be secondary to obstructive atrophy caused by tubular damage in the papillae.
There seems to be more extensive lesions in patients or animals where combinations
of analgesics are used. Some recent studies have shown that the phenacetin metabo-
lite, acetaminophen, is an oxidant, and since it is concentrated in the renal medulla
it is thought to cause oxidative damage to renal cells. Not only are each of these
chemicals available as analgesics but they usually are taken in combination with as-
pirin. Aspirin is known to inhibit prostaglandin synthesis; it may potentiate neph-
rotoxic effects by blocking the usual vasodilatory effects of prostaglandin, and there-
by predisposing renal tissues to ischemia. The papillary damage is likely to be due to
a combination of direct cytotoxic effects of phenacetin and its metabolites, as well as
to the indirect effects of ischemic injury to both tubular cells and blood vessels.
The list of over-the-counter analgesic ingredients includes the following subtances
which may be implicated in causing or being associated with nephropathies:
phenacetin, salicylamide, acetaminophen, caffeine and, in many cases, metabolites
of these drugs. Obviously, to this list, one must also add aspirin although its toxicity
may not be as severe as some of the others. Interestingly, aspirin may potentiate
toxicity through yet another mechanism since it can uncouple oxidative phosphoryla-
tion, thus decreasing the energy available for important transport functions of the
tubules, for example, to energize Na,K-ATPase.
Mercurials—
A number of drugs and environmental contaminants, including diuretics, several
inorganic salts of mercury, and fungicides, fall under this category. Mercuric chlo-
ride (bichloride of mercury or corrosive sublimate) was used in mediaeval times to
poison one's enemies, death being induced by acute renal failure. The most impres-
sive microscopic changes occur rather specifically in the proximal convoluted tu-
bules; they are dilated with degeneration, fragmentation, and necrosis of epithelial
cells. Less striking changes occur in the ascending limbs of Henle, distal convoluted
tubules and collecting tubules. Earlier signs of toxicity include cytoplasmic vacuola-
97
-------
Table 1. Nephrotoxic Metallic Compounds
Metal
Environmental
Significant Hazard
Occupational
Boron
Cadmium
Chromium
Lead
Mercury
Platinum
Silver
Thallium
Uranium
Vanadium
Table 2. Nephrotoxic Organic Compounds.
Aniline
Camphor
Carbon tetrachloride
Chlordane
Chloroform
Creosote
Ethylene dichloride
Ethylene glycol
Formaldehyde
Guaiacol
Naphthalene (mothballs)
Oxalates
Parathion
Pentachlorophenol
Phenol
Phosphorus (yellow)
Resorcinol
Thiourea
Thymol
tions and loss of the brush border, while later signs include rounding of the
mitochondria. At the cellular level, inorganic mercurials appear to disrupt both tu-
bular secretion (organic anions) and reabsorption (sugars and amino acids), and cell-
volume regulation; they are potent inhibitors of cytoplasmic, mitochondrial, and
plasma membrane enzymes by binding of Hg to essential sulfhydryl groups. The or-
ganomercurial fungicides are more lipid soluble than are the inorganic Hg salts and
hence their distribution in the body is dissimilar; differences in renal toxicity are
more quantitative than qualitative. Exposure to inorganic salts of mercury may arise
from numerous diverse sources, including drinking water (particularly from wells),
food, and industrial activities such as during manufacture of scientific instruments,
electric meters, mercury arc or vapor lamps, alloys, soldering compounds, mercury-
based antifouling paints, disinfectants, and pesticides and during bronzing, photoen-
graving, and mirror silvering.
Uranium Salts—
Uranyl acetate is used as a model tubular poisoning agent in animal studies. The
lesion is focused in the proximal cells, as in the case of mercury salts, and the more
soluble salts, such as chloride and nitrate, are more nephrotoxic than are the less
soluble ones. Early signs of toxicity include impaired tubular function, such as de-
creased transport of p-aminohippuric acid (PAH), amino acids and glucose as well
as decreased Na transport in the ascending limb of Henle's loop, which results in im-
pairment of urinary concentrating ability. Experimental uranium poisoning has pro-
vided interesting information concerning the early hemodynamic changes which are
sometimes associated with nephrotoxicity and precede the appearance of tubular
lesions. Thus, following low doses of uranium salts there are decreases in renal corti-
98
-------
cal blood flow, total renal blood flow, and glomerular filtration rate, with coincident
increases in plasma renin, an enzyme which liberates angiotensin which, in turn, ex-
erts changes on blood pressure. Because of the toxicity of uranium salts to other tar-
get organs, it is difficult to attribute their actions to any specific enzymatic site. Fur-
thermore, heavy metals often exert their effects via general cytotoxic actions.
Nonspecific enzymatic inhibition can be expected, since uranium can form covalent
metal-salt complexes with a number of enzymes having functional groups such as
-SH,-PO3H2, -COOH; uranyl salts inhibit renal ATPase activity.
Cadmium—
As is the case with most toxic metal compounds, the effects of Cd differ in target or-
gans depending on how it entered the body and whether exposure is acute or chronic.
For example, when foods are stored or cooked in cadmium-galvanized containers,
violent gastroenteritis can occur shortly after eating. Renal toxicity is found more
frequently after low-level, long-term Cd exposure (Friberg et al., 1971). Cd can enter
the atmosphere through mining and manufacturing activities. Exposures from inha-
lation involve the lungs ("fume fever") primarily, but the kidneys are involved in an
important way, since they are both the major excretory organ and a target organ for
this metal. In animals and man, the morphological changes are mainly confined to
the proximal tubules, whereas the glomeruli are less affected. There are unusual
mechanisms involved in the nephrotoxic action of Cd. During initial exposures only
small amounts of Cd are excreted in the urine while accumulation occurs in the kid-
neys, but after a certain renal level of Cd is reached, urinary Cd excretion increases
and then the amounts in the kidneys will decrease or stay about the same. Researchers
have found that a low molecular weight (10,000 daltons) protein, metallothionein,
binds Cd; this Cd-metallothionein is filtered at the glomerulus and reabsorbed in the
proximal tubule. With increasing exposure, more Cd will accumulate in the kidneys
than can be bound by the available metallothionein and so will exert toxic actions on
the tubules. Thus, urinary Cd levels may at certain times give an erroneous view of
what is happening in the kidney.
Lead—
While there is some occupational exposure to the compounds of lead, most of the
current cases of chronic lead poisoning come primarily from children who chew paint
from their furniture and home. The other major source of lead is the 200,000 tons per
year emitted into the air from vehicles. The kidneys are one of the three target organ
systems affected by lead; the central nervous and blood-forming systems are the oth-
er two. Signs of the renal lesion include increased urinary excretion of amino acids,
glucose, and phosphates, indicating impaired proximal tubular function. Studies
with animals and with cells grown in culture have shown that lead salts are accumu-
lated by mitochondria and inhibit their respiration. At the electron microscopic level,
one also sees inclusion bodies in the cell nuclei of the proximal tubules, and these
have become the current diagnostic hallmark for lead poisoning. While the increased
lipid solubility of tetraethyl lead causes some differences in body distribution com-
pared with inorganic compounds of lead, the major target organs are the same.
Platinum—
Compounds of Pt are used in the exhaust systems of automobiles as the catalyst
designed to decrease emission of certain air pollutants. When disposed of, these anti-
pollution devices may introduce quantities of Pt into the terrestrial and aquatic en-
vironments, which begets a "second-generation'1 contamination problem. Platinum
also is used in several new drugs where the need to develop water-soluble forms of
this heavy metal for therapeutic uses has led to important answers as to the mecha-
nisms of toxicity of this metal (Guarino et al., 1980). In animals and man, Pt com-
99
-------
pounds caused nephrotoxicity, which was localized in the proximal tubules, and
available evidence favors inhibition of Na,K-ATPase as a major mechanism of
toxicity.
Organic Compounds—
Carbon tetrachloride (CCU) is extensively used as an industrial solvent and is
widely distributed throughout air and water. Although liver damage is dramatic af-
ter ingestion of this solvent, acute renal failure is usually the cause of death. Morpho-
logically, tubular necrosis in the loops of Henle is most obvious. The extensive stud-
ies of how CCU exerts its toxic actions have shown that it forms highly reactive, free
radicals, which, in turn, cause disruption of membranes. In the case of mitochondria,
such effects result in swelling, loss of function, and other effects similar to those
caused by hypoxia, that is, disruption of cellular energetics.
Ethylene glycol and diethylene glycol (antifreeze) exert their nephrotoxic actions
by being metabolized to oxalate crystals which accumulate and clog the proximal
and distal convoluted tubules. However, there may be other mechanisms occurring
to impair tubular function in addition to the obstruction caused by the crystals.
Two major classes of nephrotoxic insecticides, chlorinated hydrocarbons (such as
chlordane) and organophosphates (such as parathion), cause acute tubular necrosis
with subsequent renal failure. Some of the major nephrotoxic organic compounds
are listed in Table 2.
ASSESSING RENAL FUNCTION
Ideally, predictive or diagnostic tests should be simple to perform, sensitive, spe-
cific, easy to interpret and applicable to a wide range of animals, including man. In
actual practice, no single procedure can fully meet all criteria and substantial trade-
offs must be accepted, such as specificity for simplicity. Indeed, many techniques can
be used only in animal studies; and, for example, microperfusion is currently limited
to a few selected species. Comparative studies of renal structure and function indi-
cate that differences exist even between closely related mammals. One would also ex-
pect to find species differences in the distribution of and cellular sensitivities to
nephrotoxic chemicals. Thus, when evaluating data from animal experiments, one
must consider how predictive the model chosen will be for human toxicity.
The nature of the measurement that is being made in the tests must also be consid-
ered. Many tests of renal function are designed to measure the maximal capacity of
the organ to perform a specific function. The kidney does not normally function at
this high level, and impairment would indicate a reduction in the reserve capacity
which might not affect day to day function. Such tests tend to be conservative, that
is, they exaggerate toxicities, but their use is important because reductions in func-
tional capacity may become physiologically significant during times of stress.
In addition to choosing a preparation for study, the processes used to evaluate re-
nal function must also be carefully selected. To measure filtration rates, inulin or
polyethylene glycol clearances suffice. Quantitative changes in the glomerular
filtration apparatus (altered equivalent pore size) can be evaluated by the use of, for
example, dextrans of various molecular weights. Organic substrates, such as glucose,
amino acids, and organic anions and cations, are transported in the proximal tubule.
Use of nutrients and model excretory products as substrates for transport allows the
evaluation of proximal tubule reabsorptive and secretory function, respectively.
Since such transport is dependent on cellular oxidative metabolism and on
transmembrane ion gradients, malfunction in both reabsorptive and secretory trans-
port may indicate a generalized metabolic disruption. On the other hand, if only one
process is affected, the specific locations of defects in major transport functions (rec-
ognized by the accumulation of nutrients on luminal membranes or excretory pro-
ducts on basal and lateral membranes) would suggest toxicant action at one surface,
100
-------
but not the other. Conveniently measured transport processes that are specific to
other nephron segments are difficult to identify. For ascending thick limb cells, di-
uretic-sensitive coupled Na and Cl transport could be measured; for collecting ducts,
ADH- and cyclic AMP-sensitive fluid transport might do.
In Vivo Methods
Since the kidneys bring their major product, urine, to the outside of the body, sam-
pling procedures are simple and tests are relatively easy to conduct in vivo. In gen-
eral, in vivo methods include analyses of blood and urine samples to detect altera-
tions in solute or enzyme levels resulting from renal dysfunction. Traditionally,
blood urea nitrogen (BUN) and urinary enzymes, protein, glucose, ketones and
amino acids have been used. Urea, the end product of protein catabolism, is elimi-
nated from the body primarily through the kidneys; since urea is a major component
of the BUN value, it is the most frequently employed renal function test. In certain
species, however, BUN does not become elevated until at least 75 percent of the re-
nal function has been lost. Serum creatinine levels provide essentially the same in-
formation as does BUN, but the former may be a more sensitive indicator of nephro-
toxicity. In addition, urine sediment is examined microscopically for red blood cells,
white blood cells, epithelial cells and casts. Such procedures have proved to be of
more value to physicians than to experimental investigators.
Measurement of certain enzymes in the urine has recently been of some value in
diagnosing specific aspects of renal toxicities, since they show selective increases de-
pending on whether the glomeruli or tubules have been damaged. These enzymes
include lactate dehydrogenase(LDH), alkaline phosphatase, aspartate amino-trans-
ferase, creatine phosphokinase, catalase, /3-glucuronidase, JV-acetyl-/J-glucosa-
minase, renin, and muramidase. The more widespread use of techniques of measur-
ing enzymuria is prevented by the lack of sensitivity of existing methods, the presence
of inhibitors and bacterial enzymes in urine, the lack of organ specificity for many
enzymes and problems related to the manner of data expression. Moreover, isoenzy-
matic patterns may be a more specific index of renal tissue damage than are total
urinary enzyme levels. For example, the isoenzymatic pattern for LDH may be of
more diagnostic value in differentiating between a renal and a hepatic lesion.
In other procedures, test substances are administered and the renal handling of
these compounds indicates the functional status of specific regions of the nephron.
Such procedures generally measure either the renal clearance of the test substance
using plasma concentration, urine concentration and urinary secretion rate, or the
maximum ability of the kidney to reabsorb or secrete the test substance. Clearance
experiments indicate the virtual volume of plasma that would be completely cleared
of the test substance in one minute; the specific renal function that is measured is de-
termined by the choice of test substance. For example, with inulin or polyethylene
glycol (PEG), substances that are filtered but not reabsorbed or secreted, the clear-
ance value is equal to the glomerular filtration rate; with p-aminohippuric acid
(PAH), which is filtered and strongly secreted, the clearance value is roughly equal to
the renal plasma flow, being an indicator of both blood flow and proximal tubule
function.
These in vivo tests tend to be technically simple and they can provide nearly con-
tinuous monitoring of test animals over relatively long periods of time. Their major
problem is one of specificity, since altered blood or urine chemistry is not necessarily
an indicator of impaired renal function. Indeed, most nephrotoxic chemicals clearly
affect more than one organ system and the observed effects on body fluids must be
considered to be a result of action at all target organs. Although these in vivo tests
may be of only limited value when used alone, their usefulness increases many-fold
when they are considered in conjunction with more specific in vitro tests.
101
-------
In Vitro Techniques
Each of the groups of techniques discussed below can be used for biochemical,
physiological and, in some cases, morphological studies of normal and impaired re-
nal function. In general, pathological changes in biochemical and physiological pa-
rameters tend to precede changes in morphology (seen at light or electron micro-
scope levels), so the focus of this section will be on the functional rather than the
structural assessment of renal damage. Since so much of renal function is related di-
rectly to membrane transport and cellular energy metabolism, probes of these
processes will be emphasized.
Isolated Perfused Kidney—
With perfused kidney preparations, a renal artery of an anesthetized animal is can-
nulated, and physiological saline, usually with glucose, amino acids, and bovine al-
bumin, is pumped in at constant pressure (for review, seeMaack, 1980). Urine is col-
lected and the clearance of substances added to the perfusate is determined. This is a
simple procedure which allows changes in perfusion pressure (roughly comparable
to a change in filtration pressure) and in the composition of the perfusate to be made,
and the effects of these changes on overall renal function to be determined. Thus, sys-
temic influences on renal function can be eliminated experimentally. Ideally, direct
toxic effects can be assessed after chronic or acute in vivo dosing or by addition of
toxicant to the perfusate. Unfortunately, when renal function data from perfused
kidneys are compared to corresponding whole-animal data, one finds altered hemo-
dynamic characteristics, reduced concentration and diluting ability and abnormal
distal tubule function. Since glomerular and proximal tubule functions are remark-
ably well preserved, the preparation appears to be most useful for the study of these
regions of the nephron.
Micropuncture—
This methodology has been historically useful in localizing transport functions to
distinct regions of the nephron and, more recently, in studying specific mechanisms
of transport (see review by Gottschalk and Lassiter, 1973). One or more micropipets
are inserted into the lumen of a surface nephron of an intact anesthetized animal and
either tubular fluid is sampled at a given location or the changes that occur in the
composition of the fluid as it passes from one sampling location to another are deter-
mined. Without involved surgical procedures, the sites available for micropuncture
are generally surface glomeruli (which give data on single nephron glomerular filtra-
tion rate, GFR) and portions of the proximal and distal segments; only limited in-
formation on tubular function can be obtained. Except for the anesthetic, the con-
ditions are generally physiologically realistic with tissue geometry being preserved.
Micropuncture experiments require a great deal of technical skill on the part of the
investigator as well as much specialized equipment for taking and analyzing nanoliter
(10~9 1) volumes of fluid.
Isolated Tubular Segments—
Over the last twenty years, procedures have become available for isolating tubular
fragments, perfusing them with solutions of known composition and measuring the
transfer of substances from the perfusate (luminal) to the bathing medium (peritu-
bular) and vice versa (Burg and Orloff, 1973). Tubules can be dissected consistently
by hand from slices of rabbit renal tissue, but care is required to avoid damaging
them. When a suitable length of tubule has been isolated, the ends are drawn into
pipets by suction and the tubule is sealed in the pipets. Additional concentric pipets
are inserted into the lumen of the tubule, through which perfusion fluid is introduced.
Data on reabsorptive and secretory transport of fluid, electrolytes, and organic com-
pounds have been obtained for perfused segments from nearly every distinct tubular
102
-------
region of the nephron, segments which would otherwise be inaccessible. The micro-
dissection procedures tend to be tedious and rigid control of perfusion pressures,
flow rates, and temperature are critical. At present, the low data output (a few tu-
bules perfused per day) and the relatively large variation in transport data from one
preparation to the next limit the use of the technique.
Cells in Culture—
In the past few years, cell culture techniques have been utilized to a significant de-
gree for studies of renal cell function (see Handler et al., 1980). Epithelial cells from
renal tissue form monolayers with similar orientations when grown in artificial me-
dium on a suitable support; their basal-lateral surfaces face the support and apical
(luminal) surfaces face the media. Many of these polar cultures transport water and
solutes and respond to hormones in much the same manner as the epithelial tissue
from which they are derived. Tissue culture offers potentially unique advantages,
since it could provide large amounts of material from individual cell types for long-
term in vitro studies. New mutant cell lines and renal cell cultures from a wide range
of species could be developed and maintained. There are two major drawbacks,
however, limiting immediate use. First, epithelial cultures currently contain several
types of epithelial cells which represent multiple segments of the nephron, as well as
non-epithelial cells. Second, the changes in cellular physiology that occur when renal
cells are grown in culture have not been documented thoroughly. The first problem
is largely one of technology and it will probably be solved in time; the second is of
major concern, since it bears directly on the usefulness of cell culture techniques in re-
nal toxicology.
Tissue Slices—
The preparation and experimental use of tissue slices is technically the simplest of
the in vitro procedures available for the study of renal function (for review, see
Berndt, 1976). When thin slices from grossly defined regions of the kidney, such as
the cortex, are incubated in a well-defined physiological saline, cellular metabolism,
ion and volume regulation, and transport of specific substrates can be studied con-
veniently in the absence of factors affecting renal hemodynamics and GFR. Since
many slices can be obtained from a single animal, the effects of many in vitro treat-
ments can be determined and compared directly to paired controls. The limitations
of slice procedures are numerous. First, slices must be cut less than about 0.5 mm
thick in order to provide for adequate oxygenation of cells through their entire thick-
ness. Since the rate of penetration of substrate, nutrients, and toxicants is dependent
on slice thickness, slices must be uniform to give consistent data. Second, renal mor-
phology is complex and slices cut at any level of the cortex or medulla will contain
tissue from several tubular segments, blood vessels, and interstitial cells. Careful se-
lection of the process to be studied may allow evaluation of the function of one type
of tubular cell. For example, the secretion of organic anions, such as p-aminohippu-
ric acid (PAH), has been localized by micropuncture and microperfusion studies in
the proximal segment. Accumulation in proximal cells is dependent on oxidative
metabolism; therefore, PAH uptake by cortical slices can be used to determine the
functional status of these cells. Third, since cross-sections of tubular tissue are usual-
ly obtained and tubular lumens appear to be closed, luminal or transepithelial trans-
port events cannot be studied. Indeed, most investigators consider the transport
properties of slice preparations to be primarily indicative of the characteristics of the
basal and lateral tubular cell membranes. Because of these limitations, the prepara-
tion is most appropriate for the study of secretory rather than reabsorptive transport,
where only the bath-to-cell step in the transport process is expressed.
Some of the problems found with mammalian renal slices can be bypassed by care-
ful choice of a simpler model from among the lower vertebrates. In certain species of
103
-------
teleost fish, like flounder, the nephron contains few glomeruli, no loop of Henle and
little, if any, recognizable distal segment. Most of the tubular tissue appears to be
morphologically and functionally identical with the mammalian proximal segment.
The tubules are embedded in a matrix of friable hematopoetic tissue and thus can be
teased free easily; they show open lumens and seal off at the broken ends, forming a
cyst. Because of their long viability at the lower physiological temperatures of these
animals, teased tubular masses from teleost kidneys have provided a useful in vitro
preparation for the study of secretory transport in the proximal tubule (Pritchard
and Miller, 1980).
Subcellular Fractions—
Since the early 1960s, procedures have been available for separating nuclei, mito-
chondria, plasma membranes, microsomes and cytosol from homogenates of renal
tissue by differential centrifugation (Hopfer, 1977). In some cases, as with mitochon-
dria, both functional and morphological integrity is preserved and isolated mito-
chondrial preparations are clearly of value for studying inhibitors of oxidative me-
tabolism.
New procedures in the field of membrane biology now permit the isolation of mem-
brane fragments from distinct regions of the plasma membranes of polar epithelial
cells, such as the brush border membrane or basal-lateral membrane of proximal
tubule cells. M ore importantly, it is now possible to study transport in vesicles which
form spontaneously from those membrane fragments; such vesicles are essentially
free of metabolic activity associated with cytoplasmic or mitochondria! enzymes.
These preparations have been instrumental in characterizing fundamental membrane
transport mechanisms, especially in cases where complex interrelationships among
several transport processes and cellular metabolism had previously confounded the
situation. Vesicle procedures permit the functional isolation and the study of
discrete transport events in specific membranes, and thus provide a powerful tool for
rapid testing of pollutant effects.
As with any technique, there are limitations which should be considered. Subcellu-
lar fractions should be well characterized with regard to specific marker enzymes
and, if appropriate, ultrastructure. Unstated in every tissue fractionation scheme is
the assumption that the methods used do not substantially change the biochemical
and physiological characteristics of the subcellular fractions obtained; this may not
be true of all cases. Because of damage caused by tissue homogenization or by chem-
icals used in the isolation procedure, mitochondria or membrane vesicles may lose
labile enzymes. In addition, assay media tend to be nonphysiological, and activating
factors, such as hormones or secondary messengers, are generally absent. Moreover,
drug treatment in vivo may affect the physical and biochemical characteristics of the
subcellular elements and thus modify the fractions collected, as well as the size,
shape, surface charge and "sidedness"(that is, inside out versus right side out), of the
membrane vesicle populations. These relationships must be further explored.
Tissue Enzymes—
Measurement of enzyme activities or substrate metabolism in whole-tissue homo-
genates or subcellular fractions is a convenient way to obtain potentially important
data on renal cell function and on subcellular sites of toxicant action. Normal renal
metabolism clearly varies qualitatively and quantitatively with the segment of the
nephron under study. In most segments, overall metabolic rate, as measured by 02
consumption or lactate production (depending on whether aerobic or anaerobic
metabolism predominates there), is directly related to major transport functions,
such as NaCl transport in the proximal tubule and thick ascending limb, and can
provide a reasonable but relatively nonspecific index of transport capacity. In con-
trast, caution should be exercised in assuming that changes in activities of single en-
104
-------
zymes, measured in vitro, reflect impaired renal function in vivo. Most in vitro en-
zyme assays measure activities under maximum velocity (V max) conditions.
Enzyme activities in situ are dependent on actual intracellular concentrations of sub-
strates, cofactors, and regulators, which may be different from those in the in vilro
assay medium. In addition, some enzymes, such as carbonic anhydrase, are present
in cells in great excess and a large fraction of enzyme activity must usually be inhib-
ited before any physiological effect is observed. Other enzyme activities, notably
those of ATPase, appear to be directly correlated with specific transport events.
Na.K-ATPase appears to be a pivotal transport-related enzyme in the renal tubule,
driving transepithelial salt and water transport directly and reabsorptive and
secretory transport indirectly. This enzyme also plays a major role in cell volume and
ion regulation, thus maintaining conditions that are optimal for overall cell function.
Na.K-ATPase activity provides a potentially important indicator of renal effects af-
ter in vivo exposures to toxicants. However, caution should be exercised in the use of
this or any other enzyme assay as an in vitro screen, since exposure conditions in
homogenates (both surfaces of membrane fragments exposed to toxicant in artificial
medium) may result in inhibitory effects that are not possible in intact cells. More-
over, the use of any enzyme assay for in vitro screening should be supported by data
showing that the enzyme used is, indeed, a primary target of toxicant action in vivo.
FUTURE DIRECTIONS FOR RESEARCH
In Vitro Screening Procedures
In the past, little attention was focused on the development of widely applicable,
short-term tests for renal toxicity. We believe that certain of the procedures discussed
in the preceding section could provide a methodological framework for such test
protocols. Because no single group of related pathological processes underly the ef-
fects of nephrotoxic chemicals, it is unlikely that any simple screening procedure
would by universally predictive. Indeed, a battery of several simple procedures may
prove to be of greatest value.
The regions of the nephron that can be studied using each of the previously dis-
cussed i/i vilro procedures are shown in Table 3. Because of technical considerations,
perfused kidney, micropuncture and perfused tubule procedures do not lend them-
selves well to screening protocols; slices are limited in applicability to only the prox-
imal and, possibly, distal segments. The two remaining types of preparations, cells in
culture and isolated organelles, offer the greatest promise for success as specific, rap-
id screening procedures at the cellular and subcellular levels, respectively. However,
considering the present state of knowledge, significant methodological refinements
are needed before either group of procedures can be used for routine screening. For
both cases, parallel problems exist in obtaining relatively pure and functionally re-
plete preparations of cells or organelles (for example, membrane vesicles) from spe-
cific regions of the nephron.
Developmental Renal Toxicology
Young organisms require special attention with regard to pollutant toxicity. First,
additional routes of toxicant exposure must be considered. Fetal exposure is a re-
flection of both maternal body burden and current exposure levels since the placenta
is not an absolute barrier to toxicants. After birth, nursing neonates may be exposed
to high levels of lipid-soluble toxicants that have accumulated in breast milk. After
weaning, growing animals may encounter toxicants in food or water. In addition,
lead poisoning due to the ingestion of nonfood items (pica) is a significant problem
in children; other nephrotoxic substances could conceivably be ingested at this time.
Second, when compared to other organ systems, kidney function develops late, with
changes still occurring weeks after birth, so that renal lesions which occur as a result
of exposure early in life may not be expressed until weeks later. Third, our under-
105
-------
Table 3. Portions of the Nephron that can be Studied by the Use of Avail-
able In Vitro Procedures.
Procedure
Perfused kidney
Micropuncture
Perfused tubular
fragments
Tissue culture
Slice
Subcellular fraction/
membrane vesicle
Glomerulus
X
X
X
Proximal
tubule
X
X
X
X
X
X
Loop of
Henle
X
X
X
Distal
tubule
X
X
X
X
Collecting
duct
X
X
X
standing of the changes that normally occur in renal function during development is
largely incomplete. Additional studies are clearly needed to identify critical periods
before toxic effects in young animals can be understood completely and suitable test
procedures can be devised. This problem is further confounded by the fact that renal
systems in different species develop at different rates, so no single animal model com-
pletely correlates with man.
REFERENCES
Bellies, R.P. Metals. In Toxicologv, edited by L.J. Casarett and J. Doull, Macmillan, New
York, pp. 454-502, 1975.
Berndt, W.O. Use of the tissue slice technique for evaluation of renal transport processes.
Environ. Health Perspect. 75:73-88, 1976
Brenner, B.M., and F.C. Rector, Jr. The Kidney, Vol. 1, W.B. Saunders, Philadelphia, 1976.
Burg, M.B.,and J.Orloff. Perfusionof isolated renal tubules. Handb. Physiol Sect. 8:145-159,
1973.
Dinesen, I. Seven Gothic Tales, Vintage Books, New York, 1972.
Foulkes, E.G. Mechanism of renal excretion of environmental agents. Handb. Phvsiol. Sect.
9:495-502, 1977.
Foulkes, E.G., and P.B. Hammond. Toxicology of the kidney. In Toxicology, edited by L.J.
Casarett and J. Doull, Macmillan, New York, pp. 190-200, 1975.
Friberg, L., M. Piscator, and G. Norgber. Cadmium in the Environment, CRC Press, Cleve-
land, Ohio, 1971.
Gottschalk, C.W., and W.E. Lassiter. Micropuncture metholology. Handb. Phvsiol. Sect. 8'.
129-143, 1973.
Guarino, A.M., D.S. Miller, S.T. Arnold, M.A. Urbanek, M.K. Wolpert, and M.P. Hacker.
Approaches to studies of platinate toxicities. In Cisplatin-Current Status and New Devel-
opments, edited by A.W. Prestayko, S.K. Carter, and S.T. Crooke, Academic Press, New
York, pp. 237-248, 1980.
Guyton, A.C. Textbook of Medical Physiology, W.B. Saunders, Philadelphia, 1976.
Handler, J.S., F.M. Perkins, and J.P. Johnson. Studies of renal cell function using cell culture
techniques. Am. J. Physiol. 238: F1-F9, 1980.
Hopfer, U. Isolated membrane vesicles as tools for analysis of epithelial transport. Am. J.
Physio/. 233: E445-E449, 1977.
Maack, T. Physiological evaluation of the isolated perfused rat kidney. Am. J. Phvsiol. 238:
F71-F78, 1980.
Magos, L., and T.W. Clarkson. Renal injury and urinary excretion. Handb. Phvsiol. Seel. 9:
503-512, 1977.
Maher, J.F. Toxic nephropathy. In The Kidnev, edited by B.M. Brenner and F C Rector,
W.B. Saunders, Philadelphia, pp. 1355-1595, 1976.
106
-------
Oppelt, W.W. Toxicity from exposure to solvents. In Laboratory Diagnosis of Diseases
Caused by Toxic Agents, edited by F.W. Sunderman and F.W. Sunderman, Jr., W.H.
Green, St. Louis, pp. 296-304, 1970.
Orloff, J., and R.W. Berliner (eds.). Renal Physiology, Handb. Physiol. Seel. S:1082 p., 1973.
Pritchard, J.B., and D.S. Miller. Teleost kidney in evaluation of xenobiotic toxicity and
elimination. Fed Proc. 39:3207-3212, 1980.
107
-------
EARLY WARNING SYSTEMS FOR BEHAVIORAL
TOXICOLOGY
Bernard Weiss
INTRODUCTION
Traditional definitions of toxicity epitomized the tangible criteria of death and tis-
sue damage, but newer public concerns have enlarged their scope. Now toxicity in-
cludes the consequences of low-level, lifetime exposures and impairment of function
rather than outright poisoning and gross pathology. The role assigned to behavioral
measures is consonant with this new emphasis. Behavioral toxicology is the discipline
that emerged in response (Weiss and Laties, 1975).
Why Choose Behavior?
Behavior and the nervous system mediate our exchanges with the environment,
and their capacities determine how we adapt to its challenges. Behavioral measures,
moreover, do not invade or destroy tissue so that they permit repeated testing, al-
lowing us to follow the progression of an intoxication and to evaluate cumulative or
reversible effects. Behavior, finally, may be the only visible consequence of toxicity.
Consider the parallel with schizophrenia. Despite the nearly universal belief that it
arises from a chemical substrate and despite a massive research effort, a cogent
chemical basis for it has yet to be defined. It remains a behavioral disorder.
Agents Associated with Behavioral Toxicity
Behavior is the most multifaceted aspect of biology, spanning an immense array of
structures, systems, mechanisms, and processes. It should not surprise us that many
chemicals can be singled out as sources of behavioral disturbances. Heavy metals,
such as lead and mercury, have been recognized since antiquity as origins of neuro-
logical and behavioral toxicity. Lead remains troublesome because of its wide dis-
persal in the environment and its threat to the developing nervous system. At high
exposure levels it can cause brain damage, while at lower levels, without provoking
any visible symptoms, it may be responsible for depressed school performance and
for disturbed conduct. Mercury is recognized as a powerful central nervous system
poison, and one of its organic forms, methylmercury, has been responsible for several
serious epidemics of poisoning. It is capable of causing severe, irreversible brain
damage and of disrupting the pattern of normal brain development. Table 1 surveys
the array of neurobehavioral consequences ascribed to excessive exposure to various
metals. Volatile solvents and fuels can also engender severe behavioral toxicity.
Toluene, trichloroethylene, dichloromethane and carbon disulfide can produce
behavioral actions ranging from mild impairment of alertness to frank intoxication
like that produced by alcohol.
The Author: Bernard Weiss is on the staff of Ihe Division of Toxicology and Environmental Health Sciences
Center. Department of Radiation Biology and Biophysics, School of Medicine and Dentistry University of
Rochester, Rochester, New York.
108
-------
Table 1. Adverse Neurobehavioral Responses Ascribed to Metals (Weiss,
1979).'
E
D
C
'E
D
<
Antimony
Arsenic
1
o
m
Cadmium
•o
CD
0>
_J
Manganes*
Mercury
-------
Examining Behavior in Different Settings
The flexibility and variety of behavior yield both an immense luxury of choice and
a bewildering expanse of possibilities for tests.
The Natural Environment—
Behavior need not be brought into the laboratory to observe and quantify it. We
can, for example, place trained observers in the classroom to record, by standardized
procedures, the behavior of individual children. Parents can complete rating scales,
which require estimates of the magnitude, frequency or applicability of specified be-
havioral characteristics, and workers can select, from a list of adjectives, those that
describe their current moods, feelings, or complaints.
Laboratory Testing—
Sensory processing, learning and memory, and motor coordination in humans are
best studied in a controlled environment. Certain issues demand the precision af-
forded by the laboratory, especially if the alternative may be an improperly high or
low exposure standard with its consequent economic and health implications. Non-
human species, however, dominate in toxicology because we not only need to
maximize control over exposure conditions and ancillary factors such as diet, but
also over behavioral history. Moreover, the issues arising out of delayed effects and
fetal exposure cannot await resolution by epidemiology that may require decades for
completion.
HISTORICAL PERSPECTIVE
Behavioral toxicology now describes an increasingly coherent discipline, but its
importance had been recognized earlier. Industrial hygienists have long viewed atyp-
ical behavior as an important guide to adverse effects, and took account of
behavioral aberrations in establishing exposure standards. In a more formal manner,
the USSR, guided by its scientific tradition, elevated behavior and central nervous
system function into pre-eminent roles in toxicity assessment. Exposure standards
derived from such criteria spurred the United States and other Western countries to
examine behavior more carefully as a toxic end point.
A century of experimental psychology and a surgent neuroscience provided the
basic tools for behavioral toxicology, while behavioral pharmacology offered the
most direct guidance. The advent of potent drugs, especially antipsychotic agents,
for treating behavioral disorders inaugurated this new discipline. It emphasized be-
havior as the focus of drug action and aimed its resources at three problems: screen-
ing for new drugs, the behavioral mechanisms of drug action, and the relationships
between brain chemistry and behavior. It was an easy sidestep to toxicology.
PRESENT STATE OF BEHAVIORAL TOXICOLOGY
Behavioral toxicology began in response to questions about the health risks from
agents already recognized as hazards. What was needed for both resolution and doc-
umentation was the extent of the threat. For example, how much methylmercury
was safe to ingest? What blood levels of lead in children can be considered nontoxic?
With the passage of legislation such as the Toxic Substances Control Act of 1976,
the impact of such issues widened to include new agents, but also posed a dual set of
questions. One set stems from the need to detect whether a new or previously unstud-
ied agent has potential for behavioral and neurological impairment. The other de-
rives from the need to estimate the exposure levels of a known toxic agent that might
pose a health threat. The first requirement might be met by a collection of tests pre-
viously shown to be sensitive to neurotoxic agents. Those of the second ordinarily
will have to be met by more refined and specific tests. These dual responsibilities are
110
-------
reflected by two distinct definitions of what constitutes an "Early Warning System"
In one instance, it comprises identification. In the other, it signifies quantification.
The efforts of the past ten years reflect the dominance of the latter — the risk assess-
ment — in behavioral toxicology.
Evaluation of Sensory Function
Information about our immediate environment reaches the brain through a com-
plex sequence of steps. The process begins with the response of a receptor or sense
organ. The eye, for example, converts variations in light into nerve messages, and the
ear translates fluctuations in sound pressure. Defects in the receptor or in the nerve
pathways that transmit and process this information impair our ability to react to
our surroundings. Perhaps because they act as sentinels themselves, sensory systems
often reflect the early impact of toxic processes.
No sensory system is immune. Deficits in the skin senses, impaired hearing, and
tunnel vision, can all result from methylmercury poisoning, which damages pertinent
brain areas. Methanol (wood alcohol), often proposed as a fuel or fuel additive, in-
jures the optic nerve. Certain organic solvents damage peripheral nerves and mar
sensitivity to touch. Cadmium may injure the sense of smell. Such widespread sus-
ceptibility argues for the inclusion of sensory tests as measures of toxicity.
Reflex Assessment of Sensory Function—
Simple reflex testing is usually carried out during the early phases of toxicity as-
sessment and is commonplace in the drug industry. Rodents may be examined for re-
sponsiveness to a sharp hand clap, or their paw pads may be stroked to gauge skin
sensitivity. Beyond identifying gross toxicity, such testing is not a useful guide to
hazard identification or mechanisms of action. More precise, instrumented tech-
niques are preferred.
A simple, instrumented test is the startle response. A sudden loud sound elicits
both in humans and in animals a stereotyped posture (like a sharp wince) and associ-
ated physiological activity. By placing a rat on a force-sensitive platform, we can
measure the amplitude of its startle response. Hoffman and Ison at Bryn Mawr and
Rochester, respectively, found that the amplitude of the startle response is exquisitely
sensitive to accessory stimuli during or preceding the one evoking startle (Hoffman
and Ison, 1980). The startle amplitude can be enhanced or depressed by an amount
determined by the amplitude of these accessory stimuli. Presume an interval, say, of
100 milliseconds between the modifying and the startle stimuli. A loud or bright or
other intense modulating stimulus reduces the startle amplitude markedly. A
barely detectable modulating stimulus reduces amplitude by a small amount. During
a 30-minute test period, a wide range of modulating stimulus intensities can be
presented and a function subsequently plotted. Triethyl tin damages the fatty
(myelin) sheaths of nerves. To define its toxic end points further, Squibb, Carmichael,
and Tilson (1980) administered the compound to rats and measured the startle re-
sponse both to a loud sound and to an air puff to the rat's face. Responses to both
stimuli were reduced by triethyl tin poisoning, but the rats were severely impaired in
many other ways as well; for example, they sharply decreased their food intake.
Psychophysical Methods—
Psychophysics, the oldest branch of psychology, studies the relationship between
specified physical dimensions of stimuli and the responses they evoke, typically rely-
ing on cooperative humans or highly trained animals. The subject indicates if a stim-
ulus has been detected, or assigns some index of its magnitude. Psychophysical meth-
ods usually are not viewed as elements of a preliminary test battery, but as alterna-
tives when questions arise about the earliest, subtlest consequences of toxic exposure.
Psychophysical studies of vision can provide precise information for toxicity as-
sessment; I will discuss two such studies which have been concerned with the toxic ef-
111
-------
fects of methylmercury. Both experiments used monkeys because their visual
systems and visual capacities correspond to those of humans. The retina, located at
the rear of the eye, is the screen upon which visual scenes are displayed. But the ret-
ina, and its representation in the brain, is not homogeneous in structure or function.
Its central area, the fovea, is comprised of closely packed elements termed cones.
Functions such as the delicate discriminations of reading and distinguishing colors
are served by the cones, which require at least moderate levels of illumination. The
more peripheral areas of the retina are served by the rods, which, because of their
sensitivity and interconnections, can be thought of as light-gathering instruments.
We depend on the rods for night vision. Damage to central vision can be produced
by methanol, whose metabolic products cause optic nerve lesions due to swelling
(edema). Quinine damages mostly the peripheral retina. Methylmercury selectively
damages neurons in areas of the cerebral cortex that subserve the peripheral visual
fields.
Methylmercury became an important concern in North America in 1970 with the
discovery of contaminated fish in the Great Lakes and the high body burdens of oce-
anic fish such as tuna and swordfish. Episodes of methylmercury poisoning in Japan
— Minamata in the 1950s and Niigata in the 1960s — cogently documented earlier
reports that vision provided an early warning system. Evans, Laties and Weiss (1975)
trained monkeys to discriminate geometric forms (circle, square, triangle) projected
onto translucent response keys that the monkey pushed to indicate its choice. Cor-
rect choices yielded a few drops of fruit drink from a nearby spout. After prolonged
treatment with methylmercury, the monkeys began to show decreases in accuracy,
especially when the stimuli were illuminated with very dim light. This deficit, con-
gruent with the night blindness complained of by some human victims, was consis-
tent with the pattern of damage to the brain (Weiss, 1978).
Another approach is to plot the visual fields, a technique that opthalmologists call
perimetry. With the center of gaze fixed, stimuli are presented in various parts of the
field, and the patient indicates whether or not these have been seen. The resulting
chart shows those portions of the field unresponsive to stimuli. This test was recently
adapted to monkeys (Merigan and Weiss, 1980). It depends on their insatiable ap-
petite for marshmallows. The monkey sits in a special chair, facing a screen contain-
ing a central aperture of variable diameter (Fig. 1). With the monkey's gaze fixed at
the center of the aperture, a white marshmallow on another rod is moved to the inner
border of the test aperture while the monkey's gaze is monitored. The monkey's alert-
ness is maintained by allowing it to eat the stimuli. Central apertures of different di-
ameters allow the peripheral fields to be mapped with surprising consistency. A sin-
gle monkey can be mapped in about 30 minutes by two observers. The bottom
portion of Fig. 1 shows the aftermath of methylmercury poisioning—severe
constriction of the range of vision. Small depressions in the boundary, however, ap-
pear early in the course of intoxication and at that point are reversible.
Elect rophysiology—
The activities of nerve cells can be measured as electrical signals. The nervous sys-
tem, in fact, is in continuous electrical flux, a feature that has made it possible for
clinicians to rely upon "brain waves" for diagnosis. Epilepsy, for instance, is charac-
terized by massive coordinated discharges. Electrical activity is detectable by placing
electrodes (such as small metal cups) on the skull, amplifying the signals, and record-
ing the output to give an electroencephalogram (EEG). In animals, stiff wire elec-
trodes, buried directly in brain tissue, furnish clearer electrical signals than surface
electrodes and allow the local activity of different brain areas to be compared. Most
investigators have access to devices that convert the EEG to a representation of fre-
quency against amplitude (a power spectrum).
The incessant electricial activity of the brain, however, except in unique situations
such as epilepsy, does not afford a distinct index of abnormal conditions, since too
112
-------
Left eye
Right eye
No. 605 Control
Left eye Right eye
No. 82 MeHg
Figure 1. Procedure for plotting visual fields in monkeys. Top: The visual subtense of this aperture can be changed in 10 degree steps
from 10 to 70 degrees. The experimenter places a marshmallow or small piece of fruit on a thin rod at the fixation point, and
observes the monkey through his hole. While the monkey gazes at the central target, a 1 -cm white marshmallow is moved
inside the aperture at one of the light test points located at 45 degree intervals. If the monkey glances toward the second
marshmallow, a detection is marked at the appropriate location on the visual field chart. Only one eye is tested at a time. The
other eye is occluded with an opaque contact lens. Bottom: The visual field boundaries of a normal monkey, No. 605, extend
both vertically and horizontally much farther than the boundaries of a methyl-mercury (MeHg)-poisoned monkey. No. 82. The
latter monkey, long since recovered from the peak of the intoxication, shows tunnel vision. (The second boundary drawn on its
left eye plot shows the results of a second determination.) This kind of visual loss also appears in human victims. Incipient
visual field contraction seems to be the earliest adverse sign in monkeys (Merigan and Weiss, 1980).
-------
many things are happening at once. Yet, there are underlying regularities, some of
which are detectable as rhythmic fluctuations in amplitude. An example is the indi-
cation that organophosphate pesticides may change the shape of the power spectrum
in monkeys and humans (Duffy and Burchfiel, 1980). Although extensively applied
to central nervous system drug research, such representations have not wielded
much of an impact in toxicology.
One way to reduce the electrical chaos is to sharpen the focus of action by trig-
gering a particular population of nerve cells into synchronous discharge, and sepa-
rating their unique pulse from its incoherent background. The outcome is the evoked
potential. Its use was clearly demonstrated, again, when methylmercury poisoning
swept through Niigata Prefecture, Japan, in 1963-65. Ophthalmologists were among
the first clinicians to be alerted because of the previous experience at Minamata 10
years earlier. Among the many diagnostic methods they applied was the visually
evoked cortical response. One of their stimulus conditions consisted of a flickering
light presented at various frequencies. A normal control subject (Fig. 2) evinced
well-defined fluctuations in the amplitude of the cortical electrical response at fre-
quencies as high as 42 Hertz, but the patient record showed disorderly fluctuations
even at 28 Hertz. These data corresponded to reports by patients of the frequency at
which the flickering light appeared steady, indicating that electrophysiological
measures may be candidates for toxicity assessment in similar visual disturbances.
Assessing Motor Dysfunction
Almost every class of chemical agents induces characteristic deficits in coordi-
nated movement. Mercury vapor's unique toxic action is to cause tremor. Children
poisoned by methylmercury show disorders of coordination resembling cerebral pal-
sy. Volatile solvents impair the ability to execute delicate movements such as those
required to assemble small mechanical parts, and manganese dust and fumes induce
a syndrome resembling Parkinson's disease.
Motor Reflexes and Simple Instrumented Tasks—
Preliminary pharmacologic and toxicologic screens typically assay several simple
reflexes. The righting reflex, for example, is the tendency of animals to return to a
normal position after being turned on their backs. Placing reactions refer to the way
an animal extends its paws when it is brought forward to approach the edge of a sur-
face. The best-known instrumental technique for measuring reflexes is the rotarod
(Fig. 3). The animal is placed on a grooved, rotating rod to which it clings in order to
avoid a fall. At high rates of revolution, clinging for an extended time is impossible.
The number of revolutions usually is counted automatically, with the subject's fall
tripping a switch that stops the counter. Kaplan and Murphy (1972) used the rotarod
to study the neurotoxic effects of acrylamide, a highly reactive vinyl monomer, used
since the 1950s to produce various polymers, to strengthen paper products, to sepa-
rate solids in water, and as a grouting agent. Acrylamide damages both sensory and
motor nerves in a pattern characterized as "dying-back" neuropathy. The nerves first
begin to degenerate at their distal ends (farthest from the cell body), a process that
spreads gradually inwards. Rats treated by moderate doses of acrylamide began to
show deficits in performance after about four days of treatment. Visible loss of coor-
dination did not appear until about day seven. Edwards and Parker (1977) soughta
simpler method. When a mouse or rat is dropped from a height, it extends its paws in
a characteristic posture which can be measured by inking the paws beforehand.
Splaying of the hind legs was exaggerated by acrylamide and proved about as sensi-
tive as the rotarod. Later, investigators at the National Institute of Environmental
Health Sciences in North Carolina developed a technique to measure forelimb and
hmdhmb strength; the hindlimb response proved more sensitive to acrylamide
(Tilson, Mitchell and Cabe, 1979). In all three studies, the treated animals lost weight
compared to controls, a confounding problem.
114
-------
Patient
Flicker
22Hz
VER
28Hz
VER
'VWVWWWXAAA
Normal
VER
Figure 2. The response of a patient with Minamata Disease and a normal control. The
patient VER (visual evoked response) showed inability to follow the light
fluctuations even at relatively low frequencies, a marked contrast with the
control subject.
Advanced Methods—
We should not expect the comparatively simple acts assayed by preliminary toxic
screens to predict the subtle deficits incurred by human victims after moderate ex-
posures. For example, excessive tremor accompanies many types of neurological
disorders, including those induced by toxic agents such as mercury vapor. It was one
of the earliest signs of toxicity in workers exposed to the organochlorine pesticide,
Kepone, in a Virginia plant. Wood, Weiss and Weiss (1973) followed the progress of
workers who has been poisoned by mercury vapor (Fig. 4). The patient inserted a
finger in a lucite trough attached to a force sensor whose output was amplified and
processed by a computer. Tracings showed marked tremor at the time the patients
first came to the hospital, and a gradual decline over a period of months without fur-
ther exposure. Quantitative measures of tremor revealed more subtle changes, espe-
cially the presence of multiple peaks in the frequency spectrum. These cannot be ob-
115
-------
Figure 3. Mice being tested on a rotarod. They are placed on the grooved rod, which
then begins to turn. Control circuitry gradually increases the rotation speed.
When a mouse falls, it strikes a paddle that turns off the counter beneath it.
The counter is incremented once with each revolution.
served clinically, which is why industry has now turned to such a system to monitor
exposed workers in the chlor-alkali industry, where mercury electrodes are used to
convert brine into chlorine and caustic soda.
Elect rophysiological Methods—
Nerve conduction velocity is the most widely studied electrophysiological index of
toxicity, and is measured by electrically stimulating first one site on the nerve and re-
cording the electrical response at the muscle, and then stimulating another site and
recording the response again. To measure the conduction velocity of the ulnar nerve,
for example, stimulating electrodes are placed at the wrist and elbow on the inner
forearm, and recording electrodes are placed on the muscle near the base of the little
finger. A short duration stimulus pulse (about 100 milliseconds) propagates an im-
pulse along the nerve and causes muscle contraction. The time elapsing between the
stimulus pulse and the onset of contraction represents the latency of the response.
The difference in latency between elbow and wrist electrode positions is used to cal-
culate the velocity of the impulse along the nerve. If the latency from the elbow posi-
tion is six milliseconds, that from the wrist electrode is three milliseconds, and the
distance between electrodes is 21 cm, the conduction velocity is calculated as 70 me-
ters per second.
Lead-exposed humans and animals suffer depressed motor nerve conduction ve-
locities. Solvents such as methylbutyl ketone also reduce conduction velocity. The
test does not seem to bean especially sensitive measure, however. For individual di-
agnosis, the range of normal values is so wide that only the presence of ancillary signs
and symptoms is definitive. Temperature must be carefully controlled because low-
116
-------
4-9-70
100-
70:
40
10
1-14-71
-1 5 sec.
Figure 4. A system designed to measure fine motor control. The patient inserts a finger
into the slot and tries to maintain a steady force between the specified limits
(such as 10 and 40 grams) Excessively high or low forces are signalled by
lights on the nearby display box. The tracings show the improvement in
amplitude of tremor and accuracy of motor control between the time when
mercury vapor poisoning was first diagnosed in this patient and nine months
without further exposure.
ered limb temperatures decrease velocity. The young and the elderly show reduced
velocities. Empirically, such measures have not proven fruitful in animal studies
because other indices seem able to detect toxicity at earlier times or lower exposure
levels. Seppalainen (1974) in Helsinki surveyed a group of workers in storage bat-
tery plants who had been exposed to lead, but who revealed no clinical signs of in-
117
-------
toxication. They displayed about six percent slower maximal motor con-
duction velocity (MCV) of the median and ulnar nerves than an unexposed control
group. Since the MCV only reveals damage to large diameter nerve fibers, the inves-
tigators also tested the conduction velocity of narrow diameter, slower motor fibers
(CVSF) of the ulnar nerve. They found significant slowing due to lead (Fig. 5).
Complex Behaviors
Since the natural environment is complex, the patterns of behavior by which or-
ganisms respond to and manipulate the environment must be correspondingly
complex. Consider what is involved when humans exposed to a potentially toxic
agent complain of memory difficulties, apathy, inability to concentrate, and similar,
often vague symptoms.
Naturalistic Behaviors—
Even rather complex, extended behavioral processes may be controlled largely by
hereditary mechanisms. Despite their often stereotyped expression, they represent
the outcome of many component systems and processes, giving them the potential to
reveal adverse consequences that might be overlooked in a survey of simpler mech-
anisms.
Reproductive processes—Mating and maternal care are central to species surviv-
al, a role consistent with the elaborate mechanisms that have evolved for their sup-
port. Flawed mating behavior may reflect disorders of endocrine function, of neuro-
chemistry, of central nervous system structure, of sensory function, or of other, per-
haps less direct influences. Madlafousek, Hlinak, and Parizek (1971) asked whether
cadmium-induced testicular atrophy and necrosis were correlated with sexual
behavior in male rats. They placed male and sexually active female rats into a spe-
cial enclosure and recorded quantitative measures of the males' copulatory
performance (Fig. 6). After about nine days of treatment, most of the males began to
show incomplete mounts and failure to copulate. To determine whether this be-
havioral deficit arose from the destruction of testicular endocrine tissue, one group
was treated with the male hormone, testosterone. Seventy days after hormone
treatment, most of the rats had recovered copulatory function except for a pro-
longed postejaculatory interval. Testicular tissue also had regenerated, confirming
the correlation between morphology, endocrinology and behavior.
Activity—Many experimenters have tried to use activity as a quick and inexpen-
sive index of total function. Rats and mice adapt to many different kinds of enclo-
sures ranging from simple boxes to circular alleys. A simple technique is to mount
photocell beams in an enclosure and count the number of interruptions during a
specified time. The residential maze, a more recent type, is designed like a figure
eight, and contains food and water in specific locations, and photobeams positioned
at several points. Since, however, such techniques detect mostly coordinated, gross
movement, newer methods have been introduced that provide measures of smaller
amplitudes. Some rely on disturbances of an electromagnetic field, and are especially
appropriate for measuring the activity of neonatal rodents not yet capable of walk-
ing.
Animals typically display a daily rhythm to their activity. Robins are most active
in the morning and at dusk, and become even more active when they consume PCBs
(Ulfstrand, Sodergren and Rabol, 1971). Rodents are more active within the hours
of darkness. Running wheels attached to living cages are one traditional means of
monitoring daily activity. Most rodents will run several thousand revolutions per
night, and females run more during oestrus than at other times during their repro-
ductive cycle.
118
-------
0
c
o
o
CD
v_
CA
0
Q.
X
CD
ra 50
CD
»
O
C
o
o
CD
c
1 40
o
a
•o
CO
CD
0
0
CA
E
0)
£ 30
ID
c
CO
c
3
CD
.C
'o
u_
C/3
>
0 20
Lead
Poisoning
— S
t
t
t
I
•
0
I
^
%
—
Controls
O
O
O
O
<$>
&
• t= 4.90
• p <0.001
nT
Lead
Exposure
9
99
9%
9
99
— 9
A
(K93
»
9
—
Controls
0
00
8
o
0
ogo
— 8 —
1
o
_ t= 3.57
p <0.001
T
Figure 5. These charts plot conduction velocities, for two worker groups, of the slowe
motor fibers (CVSF) of the ulnar nerve. The group on the left, labelled Lead
Poisoning, consisted of 32 workers who displayed clinical symptoms or
markedly elevated blood lead concentrations. Each worker's reading is
represented by a black dot. Age-matched controls, with no lead exposures,
showed higher and more normal values. The horizontal lines represent the
means. Note the clusters of three lead-poisoned workers at the bottom. The
charts to the right compare 28 lead-exposed workers who were clinically
normal with another group of age-matched controls. Even here, the difference
between the groups is greater than would be expected by chance.
(Seppalainen, 1974).
119
-------
Time
Behavior
0 5 10 15 20 25 min.
I I I I I I I I I I I I I M I I I I I I I I I I I
1
ll
4
III
r^
1
1
T
I
..
|
Figure 6. Scheme for testing copulatory performance in male rats. The male is
introduced at the first arrow and the female at the second. Intervals are (1)
intromission latency (the time from the entrance of the female into the
testing boxtothefirst intromission); (2)ejaculation latency(thetimefromthe
first intromission to the appearance of the ejaculatory behavior pattern);
(3) mean interval between intromissions (that is, between two consecutive
complete mounts); and (4) post-ejaculatory interval (the time from the
ejaculatory behavior pattern to the next complete mount). The investigators
also measured : (5) intromission frequency (the number of intromissions
preceding an ejaculatory behavior pattern); (6) incomplete mount frequency
(the number of incomplete mounts between the first intromission and an
ejaculatory behavior pattern); and (7) ratio of incomplete mount frequency to
intromission frequency (Madlafousek et a/., 1971).
Reasoning that air pollutants might hamper physical activity, Konigsberg and
Bachman (1970) measured spontaneous running in rats exposed to ozone. Ozone is
the principal oxidant in photochemical smog, producing systemic as well as local ef-
fects on the lung. Even at ozone concentrations that appear frequently in urban
centers such as Los Angeles and Denver, activity fell off. This reduced inclination to
run has its human counterpart in runners poisoned by increased levels of ozone in the
course of exercise.
Different techniques for assessing activity yield different conclusions, probably
because they feature different acts. In addition, minute variations in enclosure design
and the environment can produce enormous variations in results. No one method is
demonstrably superior (Reiter and MacPhail, 1979).
Learned Behaviors—
Many complex functions depend on individual histories and experience. They are
learned. The dominant role of learning and learned behaviors in humans led experi-
mental psychology to feature learning as one of its stoutest pillars. Impaired learning
ability in humans has such devastating consequences that learning is an almost uni-
versal component of proposed test batteries.
A situation used extensively in behavioral pharmacology is avoidance condition-
ing. Most typically, a stimulus such as a light or a tone serves as a conditioned (warn-
ing) stimulus, providing several seconds delay before a painful electric shock circuit
is closed. A specified response by the subject allows it to avoid the shock. Bignami
(1975) relied on avoidance performance as an index of toxicity for a class of com-
pounds called anticholinesterase agents (for example, organophosphate pesticides
and nerve gases), which block an enzyme crucial to the function of the neurotrans-
mitter, acetylcholine. He trained rats to cross from one section to another of the ex-
perimental chamber at the appearance of a light or the sound of a tone to avoid an
120
-------
electric shock delivered several seconds later. His most important finding was the
confirmation of the rapid tolerance that develops to such agents. The first few in-
jections of the agents disrupted avoidance reactions markedly. Later, the compounds
exerted far less effect even though brain levels of the enzyme fell to about 10 percent
of normal. The implications of such a discrepancy pose a significant problem, be-
cause the long-term consequences of such chronic reductions on human health are
unknown.
More elegant analyses can be built on what are called Schedules of Reinforce-
ment. This term designates a scheme for manipulating operant behavior, or behavior
controlled by its consequences. In nature, the actions of organisms affect their en-
vironment, and determine its reactions. A laboratory model is exemplified by a rat in
an experimental chamber who is kept alert to the possibilities for food by mainte-
nance on a diet that holds its body weight at about 80 percent of the free-feeding val-
ue. It quickly learns to press a small lever which discharges a pellet of food. Food is
the reinforcer and the events of food delivery comprise reinforcement. Food need not
be delivered after each response, but instead can be provided intermittently. The way
in which food delivery is governed is called the Schedule of Reinforcement. For
example, a Fixed-Interval Schedule provides reinforcement for the first response
after a specified lapse of time, usually measured from the previous reinforcement.
Responses occurring before the interval ends are usually recorded but are not re-
quired for reinforcement.
Willes, Rice and Truelove (1980) found Fixed-Interval performance to be a sen-
sitive index of lead exposure. They fed infant monkeys, from birth to 200 days, with
a milk formula to which they added lead in various concentrations. One group of
four animals, treated each weekday with 500 micrograms for each kilogram of body
weight, attained blood lead concentrations in the upper range of urban American
children. After the treatment period, the monkeys were trained to press a lever for
juice delivery on an eight-minute Fixed-Interval Schedule of Reinforcement. During
the first 20 to 30 sessions, treated monkeys pressed much more frequently than con-
trols. They also tended to respond in bursts, in contrast to the controls, who showed
the typical pause after reinforcement followed by an increasing rate of response as
the opportunity for juice delivery drew nearer. Later, the average response rates for
the lead group declined to that of the controls, but the bursting pattern persisted.
These findings, coupled with earlier rat data, suggest that the period before
performance stabilizes may be a sensitive index of aberrant function, with the con-
comitant advantage of not requiring extended training.
Homeostatic Functions
Disturbances in homeostasis frequently appear during the incipient stages of a
toxic process. Weight loss may be detected even before subtle neurological signs ap-
pear, as in acrylamide and methylmercury poisoning in rodents. Loss of appetite is a
common complaint among victims of toxic exposure. Metabolic disturbances might
be reflected even more quickly than disruption of body temperature regulation. A
web of simple reflex adjustments maintains central body temperature within rather
narrow limits by continuous changes in blood flow to the skin, metabolic rate and
sweating. These are microadjustments, however. The major contribution to temper-
ature regulation is behavior, for instance, wearing appropriate clothing during the
winter. Hormonal disturbances and nutritional deficiencies that impair metabolic
heat production can be detected by measures of behavioral thermoregulation. Rats
deprived of essential vitamins or lacking full thyroid function work for external heat
to replace the heat they cannot produce metabolically (Weiss and Laties, 1961).
Although microwaves heat tissue, some investigators have claimed that there are
behavioral effects at intensities too low to elevate deep body temperature. Stern and
his coworkers (1979) trained shaved rats to press a lever for a burst of heat from a
121
-------
heat lamp, a task rats readily perform at low temperatures (Fig. 7). They then com-
pared alternating 15-minute periods with and without microwave exposure. During
exposures as low as five milliwatts per square centimeter of surface (half the current
human standard) the rats reduced their frequency of responses. This experiment is
also a useful example of how it is possible to measure presumably subjective states
(feeling cold) by behavioral methods.
Measures of Aversion and Attraction
The bilious miasma that clings to certain cities and the choking fumes emitted by
some industrial processes are not simply unaesthetic. Smarting eyes and raw throats
imply toxic actions. The aversive potencies of such irritants need to be established in-
dependently of overt tissue damage. Especially with new chemicals, we cannot
always adopt the easy solution of asking human subjects to evaluate or estimate irri-
tant potency. We turn to animals, instead.
Estimates of respiratory irritation obtained from various parameters of lung func-
tion are limited by the number of feasible measures and by reactions outside the di-
rect scope of the lung. Tests relying on the response of the whole animal have been
adopted for this reason. Mice or rats are put into a sealed cylinder, and variations in
breathing are recorded as pressure variations. Irritant gases, fumes, or vapors intro-
duced into the cylinder shift the pattern of respiration. Surveying a range of concen-
trations allows the experimenter to calculate relative indices of irritant potency. This
technique, like other reflex measures, has the virtue of not requiring extended train-
ing, although it could sometimes be misleading if animals respond to low concentra-
tions because they have been conditioned to the smell of the agent. It also is not in-
trinsically a measure of aversiveness, which is gauged by escape or avoidance re-
sponse (Wood, 1979). Such a measure is obtained in the same way that we estimate
the aversiveness of electric shock. In this case, a noxious gas or aerosol is injected
into a chamber and the subject makes a defined response so that the gas is replaced
by a stream of clean air (Fig. 8 A). The technique has been used to measure the aver-
sive properties of ammonia. At low concentrations, the mice waited 20 to 30 seconds
before turning off the flow, but, at high concentrations, they waited only about five
seconds (Fig. 8B).
Behavioral Teratology
There is a growing awareness of the special, enhanced vulnerability of the fetus
and neonate, particulary its nervous system. What is especially troublesome is that
toxic substances may cause derangement of central nervous system development
which will be manifested, not as visible deformities (teratologies), but as behavioral
disorders later in life. Environmental contaminants are of particular concern be-
cause they may be stored in the female's body long before conception and because
explicit exposure sources may not be known. Most studies in behavioral teratology
attempt some assessments in peri nates and neonates, seeking not just evidence of ab-
normalities, but of delayed development. The progression in Table 2 shows the mat-
urational stages in mice often used as markers of development.
However, the earliest periods of development may not reflect the impact of a toxic
process. Damage in human infants may go unnoticed because of the neonate's undif-
ferentiated behavior. Only later, when the child fails to progress normally, does the
damage become apparent. The same problems arise in laboratory animals. In fact,
certain kinds of damage may not manifest themselves until well into maturity, when
the reserve capacity and compensatory mechanisms of the brain, reduced because of
the original deficit, are no longer adequate to cope with the progressive erosion of
function that accompanies aging.
One of the most significant studies in the past decade was carried out by Crarimer
(1975) after suspicions about the elevated sensitivity of the fetus to methylmercury
had been stirred by events at Minamata. Cranmer treated pregnant mice withmeth-
122
-------
Blower
Figure 7. System for evaluating the effect of microwave exposure on thermo-
regulatory behavior (adapted from Stern et a/., 1979). The rat was shaved
before the session, then placed in the chamber, which was constructed
largely of materials transparent to microwaves. The ambient temperature
was about 4-5° C. Each press of the lever by the rat turned on the heat lamp
for two seconds. Fifteen-minute periods without microwaves alternated with
15-minute periods during which the rat was exposed to microwave radiation
from a feeder horn. During the microwave exposures, which ranged from a
power density of 5 to 20 milliwatts per square centimeter, the rat turned on
the heat lamp less often.
ylmercury at various times before they gave birth. Some of the offspring died before
weaning; others showed reduced weight gain, and even those prenatally exposed off-
spring who were grossly indistinguishable from controls showed behavioral devia-
tions. At 30 days of age, they were tested in an open-field device where they not only
proved less venturesome than controls but also tended to take backward steps, an
unusual occurrence. They also swam in peculiar ways, showing frequent episodes of
incoordination and motionless floating. Even more significant, as the treated ani-
mals aged, they began to show generalized debilitation, evidence of immune defi-
ciencies, and neuromuscular lesions. Disorders that earlier had required tests to be
disclosed erupted into clinical disease in old age. Later research showed that even
much lower dose levels can produce later behavioral abnormalities.
123
-------
(A)
(B)
Irritant
60
50
40
1 30
u_
'o
c
o
20
10-
0-,
i
Figure 8.
0.0 .05 .10 .15
NH3 Concentration (%)
Irritant potency of airborne substances is measured behaviorallv with a response that terminates irritant flow (A) Room air
flowing through the exposure chamber evokes no response by the mouse (left). When room air is replaced by an aversive
concentration of ammonia (right), the mouse inserts its snout into one of the attached cones, an action that interrupts the
photobeam. Selection of the correct cone is programmed to pour clean air into the chamber and through the cone The other cone
is used to evaluate response specificity; responses are counted, but do not stop the flow of ammonia (B) The higher the
concentration of ammonia, the more quickly the mouse responds to replace it with air (Wood, 1979).
.20
-------
Table 2. Developmental Progress in Mice (adapted from Fox. 1965).*
Period
Duration, Days
Main reflexological criteria
1. Perinatal
2. Neonatal
3. Postnatal
Transition
4. Postnatal
infantile
(orPre-juvenile)
5. Juvenile
Birth - 3
3 -9
9- 15
15-26
26 - onward
until sexual
maturity
Weak rooting, righting, geotaxis
and limb placing.
Strong and stereotyped reflexes and
limb placing reactions, pivoting and
circling locomotion.
Primitive reflexes disappearing (for
example, rooting). Adult locomotor
activities appear and organs of
special sense are functional at the
end of this period. Hyperkinesis
disappears. Overgeneralized sensory
responses seen.
Refinement of locomotor abilities
and sensory capacities. Over-
generalized responses to sensory
stimuli disappear.
Further refinement of activities
and sensory responses as the over-
generalized responsiveness is no
longer seen and manipulative
abilities are adult-like.
'Characteristic behavior patterns appear (and fade) within relatively circumscribed peri-
ods, which are labelled on the basis of time since birth. Workers in behavioral teratology
typically select a restricted subsample of these criteria for assessment.
Such studies supplement the growing epidemiologic evidence of the developing
nervous system's enhanced sensitivity and emphasize the critical role that longi-
tudinal studies, extending into old age, must play in behavioral teratologic research.
Data from the 1971-72 epidemic of methyl mercury poisoning in Iraq now suggest
that the fetus is at least 10 times more sensitive than the adult to this agent and that
increasingly subtle deficits will appear as the children mature (Marsh et al., 1980).
These studies have a counterpart in the debate about what constitutes a safe ex-
posure to lead. In children, lead concentrations in blood are the traditional indices,
but blood lead is a labile measure, reflecting only recent exposure, and cannot reveal
effects on school performance of excessive exposure during earlier periods of life.
Needleman (1980) therefore measured lead in deciduous teeth collected from over
3000 asymptomatic children in the first and second grades. Teeth, like bone, store
lead so that the amount in the dentine of baby teeth serves as an integrated measure
of early exposure. He also collected performance rating scale measures from the
teachers. In addition, he examined 58 children with the highest and 100 children with
the lowest tooth lead concentrations by means of various psychological tests. Even
in clinically asymptomatic children, differences in lead levels correlated with differ-
ences in classroom behavior (Table 3) and IQ scores (Table 4). Perhaps the most in-
triguing finding was the clear relationship between tooth lead and scores of class-
room behavior based on a teachers' rating scale, a finding that makes the assumption
of a threshold questionable.
125
-------
TableS. Comparison of Children with High and Low Tooth Lead Concen-
trations on a Teacher Rating Scale (adapted from Needleman,
1980).*
Low Lead High Lead
PValuef
Distractible
Not persistent
Dependent
Disorganized
Hyperactive
Impulsive
Easily frustrated
Daydreamer
Does not follow:
Simple directions
Sequence of directions
Low overall functioning
14
9
10
10
6
9
11
15
4
12
8
36
21
23
20
16
25
25
34
14
34
26
0.003
0.05
0.05
0.14
0.08
0.01
0.04
0.01
0.05
0.003
0.003
*The percentage of each group receiving a negative evaluation is tallied here.
fine probability that these differences could have occurred by chance.
Table 4. Intelligence Test Scores of Children with High and Low Tooth
Lead Concentrations (adapted from Needleman, 1980).*
Full-scale IQ
(normalized)
Verbal IQ
Information
Vocabulary
Digit span
Arithmetic
Comprehension
Similarities
Performance IQ
Picture completion
Picture arrangement
Block design
Object assembly
Coding
Mazes
Low Lead
(Mean)
106.6
103.9
10.5
11.0
10.6
10.4
11.0
10.8
108.7
12.2
11.3
11.0
10.9
11.0
10.6
High Lead
(Mean)
102.1
99.3
9.4
10.0
9.3
10.1
10.2
10.3
104.9
11.3
10.8
10.3
10.6
10.9
10.1
P Valuef
0.03
0.03
0.04
0.05
0.02
0.49
0.08
0.36
0.08
0.03
0.38
0.15
0.54
0.90
0.37
"The test employed was the Wechsler Intelligence Scale for Children (Revised), which
yields, besides a total (full-scale) IQ score, individual scores for the component
subtests. The latter are divided into two categories: Verbal and Performance.
tjhe probability that these differences could have occurred by chance.
126
-------
OUTLOOK FOR THE FUTURE
Behavioral toxicology, having achieved recognition as an essential component of
toxicity evaluation, now has to mature as a discipline. In growing as a science, it has
drawn from psychology, psychiatry, epidemiology and the vast spectrum of neuro-
science. It cannot afford, however, simply to graft the methods of these more estab-
lished disciplines onto toxicologic issues. Toxicologic questions may not respond to
tactics designed for other problems. Much of physiological psychology, for
example, was aimed at resolving relationships between brain and behavior by trying
to localize specific brain functions. A frequent approach was to damage an area sur-
gically or electrically, and then determine the consequences for a specific behavior. A
poison, however, is likely to act on a wide array of structures and chemical systems
within the brain, to damage peripheral nerves as well, and even to impair the func-
tion of other organ systems. Lack of specificity, coupled with the emphasis of envi-
ronmental toxicology on early warning signs, mandates new approaches. This is
where the challenge lies.
Choices and Strategies
There are a galaxy of possibilities from which we can select approaches and specif-
ic tests and then formulate a strategy. For hazard detection, two alternatives can be
pursued. One is to devise a cluster (or battery) of specific tests affording a compre-
hensive assessment of critical functions such as sensory processes, coordination and
strength, learning and memory, activity and other naturalistic behaviors, and apply
this approach in all instances. Such a standardized vehicle is helpful to both manu-
facturers and regulatory agencies because it sets boundaries to disagreement, but it
may waste resources. Not every agent or question demands a full treatment. The
other alternative is to proceed in a branching sequence, like a computer program-
mer's flow chart, beginning with simple assays of behavior, then assimilating more
complex or specific ones at successive steps dictated by the earlier data.
The Environmental Protection Agency (1980) recognizes the virtues of combining
both approaches. Once a hazard is identified by screening tests, the sequential scheme
could then pursue the questions posed by the preliminary data, conducting detailed,
intensive surveys of selected functions. The coming decade will also continue
another kind of scenario. Problems will arise that are a legacy of past neglect, such as
chemical dumps. For example, at least 50, and probably many more, of the sub-
stances identified in the Love Canal are neurotoxic. The Health and Human Services
Committee's report (1980) on the potential health effects of toxic chemical dumps
recognizes that "Clinical assessment of the presence of disease in individuals is less
efficient than comparisons of quantitative health measures in populations."It also
recognized that we lack techniques for detecting incipient toxicity.
If behavioral toxicology is to thrive and to fulfill its potential, it must evolve effi-
cient, feasible and sensitive testing strategies that can reflect the impact of chemicals
on the highest levels of human function.
Development of Methods
Techniques should be selected on the basis of their congruence with function. A
behavioral analysis should determine the selection of tests, not the reverse. The most
critical task is not to evaluate many different tests, but to evolve a logic or strategy
of methods development. Each of the components offers its own problems and pos-
sibilities.
Sensory Processes—
Reflex testing—Quick and simple tests of sensory function are among the most
eagerly sought technologies in behavioral toxicology. A promising technique, reflex
127
-------
modulation, has been developed during the past few years. In mature humans, audi-
bility has been tested by means of the "startle" response (a puff of air to the eyelid, for
example). Sounds presented as modulating stimuli change the amplitude of the re-
sponse in accordance with their loudness. Such audibility functions closely parallel
verbal reports, giving us reason to rely on these data. Therefore, these tests are use-
ful for organisms that cannot report to us: namely, infants, stroke victims, retardates
and animals.
Psychophysics— Even with refinements in reflex modulation, reflex methods are
unlikely to yield subtle details of dysfunction. Although objections frequently are
lodged against psychophysical procedures because of their apparent cost and com-
plexity, occasions are certain to arise for which there is no alternative. No matter
how many simple tests are performed, with no matter how many animals, they still
will fail to provide an adequate guide to human parameters. In the end, the costs of
studying a few highly trained primates may be far less than any alternative proce-
dure. Further, the precision with which psychophysical studies can be conducted
and the stability that can be achieved in individual animals mean that from each
animal a large harvest of information is obtained. During the next decade, some of
these procedures may be shortened, or at least made less expensive, by exploiting the
possibilities of automation derived from advances in computer technology. For ex-
ample, the visual fields of monkeys, as described under Psychophysical methods,
can be plotted by many of the same procedures now employed for automated testing
in humans which rely on devices that can sense the direction of gaze.
Naturalistic behaviors—The hope is often voiced that behavioral toxicology can
bypass complex studies of learned behavior and investments in complex
instrumentation by observing the behavior of animals under natural or quasi-natural
conditions. But this route offers problems of its own in deciding which behaviors are
to be observed and how they are to be specified (Reiter and MacPhail, 1979). Control
over important sources of variation is often impossible, especially when (1) animals
are sampled from a population composed of several distinct subpopulations(forex-
ample, dominant and subordinate animals); (2) numerous behavioral measurements
are recorded on each subject such as frequency counts of behaviors exhibited during
agonistic interactions (Aspey and Blankenship, 1977); and (3) behaviors also are
composed of sequences, and only rarely are these transition relationships analyzed
adequately and reproducibly.
Naturalistic behaviors are not a shortcut. In the absence of standardized instru-
mentation, highly trained observers and repeated checks of observer consistency are
needed to achieve repeatability from one laboratory to another. Behavioral lexicol-
ogists cannot neglect the potential of ethological approaches, however. But this will
first require refinement of behavioral taxonomies for different species in different
situations, and also familiarity with time series analysis and multivariate statistics.
Multivariate techniques are especially critical because several dimensions of be-
havior are inevitably measured concurrently (Aspey and Blankenship, 1977). The
relationships among these measures and the ability to compress many measures into
a smaller number are provided by such techniques. They also can help to define
clusters of individuals, an especially important role when confronted by subpopu-
lations of susceptible individuals.
Complex learned processes—Enhanced sensitivity to what is appropriate for
ethological observations should enhance sensitivity to what is appropriate for studies
of complex learned behaviors. The main difference between them is that experimen-
tally oriented behavioral research focuses on a single dependent variable while
holding extrinsic sources of variation constant (Aspey and Blankenship, 1977). They
converge in such questions as what governs the way in which animals distribute their
128
-------
foraging behavior? Baum (1981) points out that "laboratory experiments on food-
producing behavior ought to relate to foraging for food in the wild." Foraging
theory deals with how an animal makes transitions from one source of food to the
other, but the same kind of analysis can be applied to behavior allocated to different
Schedules (payoffs) of Reinforcement. Schedules of Reinforcement, because they
specify some relationship between behavior and its consequences, are a means for
modulating the distribution of behavior in time and among alternatives. The proto-
type experiment is one in which the organism responds on one of two alternative re-
sponse devices, each having a different frequency or probability of paying off with a
reinforcer such as food. Comparing the rates of responding on the two yields a ratio
that can be related mathematically to the frequency of reinforcement associated with
each (Hursh, 1980). This situation has the potential to be an elegant tool in
behavioral toxicology and mimics an important human activity.
Behavioral toxicology can also exploit more fully the possibilities for microanal-
ysis made available by digital computer technology (Weiss, 1970). A relatively
simple extension, such as recording the intervals between successive responses, per-
mits more subtle and comprehensive analyses than more typical measures; for
example, the application of time series measures (Weiss, 1981; Weiss and Gott, 1972).
The drug amphetamine, a stimulant, changed the pattern of responding in pigeons
that pecked a key to gain occasional brief access to grain. The drug increased the in-
cidence of aborted responses, that is, responses that failed to strike the key or that
struck it with insufficient force to close the contacts. Microanalysis plays an obvious
role in behavioral toxicology because of our interest in subtle modifications of be-
havior, especially those that render the organism less able to adj ust to the shifting de-
mands of the environment.
Behavioral Teratology—
Specific contributions of behavioral teratology to environmental toxicology are
still rather modest. Behavioral teratologic research is intrinsically limited by a num-
ber of different problems. One, noted by Dews and Wenger (1979), is that each sub-
ject can only contribute one piece of data, which can be swamped in a mass of varia-
bility. Structural deformities occur only occasionally; most individuals are at least
grossly normal, even after exposure to potent teratogens. Unless prenatal and ne-
onatal exposures have affected an array of mechanisms in a substantial fraction of
the exposed organisms, only a huge number of subjects and rather precise tests can
detect the aftermath of a toxic exposure.
Even so, current procedures could be enhanced. The immature organism is capable
of more complex behavior than typically is assessed in screening. Deficits in integra-
tive functions rather than reflexes may be better clues to problems as the organism
matures. The ten-day-old rat, guided by smell, is capable, for example, of orienting
toward and finding a clump of bedding from its own nest (Leon, 1978). Infant mam-
mals also seek heat as early as one day after birth, a behavior far more effective in
temperature regulation than reflexes (Satinoff, 1980) and easily assessed. Neonates
can also learn to suck an instrumented nipple. Newborn dogs, for example, respond
appropriately when milk delivery is governed by intermittent reinforcement sched-
ules (Stanley, 1970). Behavioral teratologic assessment optimally extends to the
organism's lifetime. Longitudinal studies are important because, as the brain ages, it
undergoes numerous structural and functional losses. If such changes are
accelerated by environmental chemicals, these exposures represent a hazard rather
inadequately weighted. Only a restricted number of agents can be studied in this way
because of the cost, but certain agents and questions require such depth.
129
-------
Behavioral Epidemiology—
Behavioral assessment in human populations poses special problems. Consider the
episode of polybrominated biphenyl (PBB) poisoning that took place several years
ago in Michigan. Dairy cattle and other farm animals were fed highly toxic PBBs (a
fire retardant formulation) instead of the nutritional supplement with a similar
name. Over thirty thousand contaminated dairy cattle were slaughtered as a result.
Many of the farm families that consumed the tainted milk and meat complain of
nonspecific symptoms, such as depression, memory loss, lethargy, and insomnia,
which are unaccompanied by clear clinical indications of disease. Simple, direct
questionnaires and conventional clinical medical examinations are intrinsically un-
satisfactory. More suitable psychometric techniques such as forced-choice inven-
tories may be used to refine subjective complaints. These typically are made up of
multiple choice items, none of whose alternatives are clearly preferable to the others,
so that subjects cannot choose answers designed to convey a specific impression.
Multivariate techniques will again play a central role. Studies at the Institute of
Occupational Hygiene in Helsinki of workers in battery plants where they are ex-
posed to lead and in rayon manufacturing where they are exposed to carbon disulfide
support the potential of multivariate analyses (Seppalainen, 1974). Any one test
singly may provide only the most tentative discrimination between an exposed
group of workers and controls, but a group of tests is much more effective in re-
vealing differences among populations with different exposure histories (Hanninen,
1979).
Ancillary Measures—
Ancillary measures of central nervous system status, such as electrophysiology,
neurochemistry, and histopathology, also play crucial roles in toxicologic assess-
ment. If behavioral deficits appear, it is certain that some underlying system has been
disturbed. But even if by behavioral criteria the organism seems fit, it is no guarantee
that adverse consequences have not been imposed on one or more ancillary
mechanisms. The nervous system possesses impressive reserve capacity, both chem-
ically and morphologically, and also compensatory mechanisms that exploit that
capacity. During methylmercury intoxication, for example, nerve cells in the brain
may be destroyed in limited quantity so that the consequences do not erupt into
overt neurological or behavioral disturbances, a phenomenon that has led to the
concept of "silent damage." Both to weigh such damage and to provide correlative
measures with behavior, other sources of information about the organism's status
need to be tapped.
Electrophysiology—The time and costs required to train animals for the precise
assessment of sensor capacity and other complex functions have encouraged
speculation that electrophysiological techniques may prove to be quicker and
cheaper as well as a source of additional information. The averaged response
method has been proposed as an attractive alternative, and the digital computer
makes this an easy tactic. Averaging depends on a time-locked event such as a light
flash. After each flash, subsequent electrical activity is recorded from the scalp or
deep brain electrodes for 500 milliseconds. This period is divided into 500 one-milli-
second intervals represented by 500 computer memory locations. The amplitude of
the electrical response at each millisecond is added to the corresponding memory lo-
cation. Although the tracing of any single response to the stimulus event is unlikely to
be distinctive, the average of, for example, 100 flashes may yield a clear waveform
because the random, uncorrelated activity ("noise") is cancelled out. The averaged
response represents a profile of stimulus-triggered electrical activity. In theory, this
technique bypasses adaptation and training and permits direct evaluation of a par-
ticular sensory system. The method has some drawbacks. Suppose the intent is to
measure the functional capacity of the visual system. In typical experiments, the trig-
130
-------
gering stimulus is a brief, bright flash of light. Such a stimulus may not reveal much
information about incipient toxicity: early methylmercury intoxication seems to be
characterized by localized dysfunction of small areas at the boundaries of the visual
field. Illumination of the entire visual field by a light flash will not produce changes
detectable in the evoked potential.
Evoked response methods are also burdened by complex problems of variability.
Estimation of background "noise" is essential before assumptions can be made about
waveform shape (Glaser and Ruchkin, 1976). Tests of the homogeneity or consist-
ency of waveforms can be conducted with relatively little cost and computer time.
Simple tests of statistical significance, not restricted by assumptions about the noise
statistics and easily programmed, are also available (Lowy and Weiss, 1968). Prob-
lems of variability are not confined to waveform analysis but also extend to compar-
isons among subjects (Snyder, Becak, and Dustman, 1979). By addressing these is-
sues, toxicology could advance electrophysiology in parallel with its own concerns.
Histopathology— Morphology is the best entrenched criterion of toxicity. Wheth-
er the pathology conventionally practiced in toxicology is adequate for the nervous
system might be questioned, however. For example, to discern depletion of a nerve
cell population in a particular area of the brain, 30 percent of the cells would have to
have been destroyed. Moreover, deranged morphology may be expressed in other,
more subtle ways. Suppose that a toxic substance administered during early devel-
opment reduces the number of nerve cell processes (dendritic spines) receiving con-
nections from other cells. Even if the number of cells remains intact, the functional
potential of the brain will have been compromised. Such a flaw in development
might not be manifested behaviorally until advanced age or disease or trauma, when
the reserve capacity of the brain has been depleted.
Quantitative neuroanatomy will augment conventional histopathology. Such as-
says rely on the phenomenon that certain silver compounds penetrate into selective
nerve cells and outline them with remarkable clarity. With proper optics and elec-
tronics, the nerve cell image can be converted into computer-readable form, and
then be subjected to pattern-recognition routines to quantify the length and number
of branches (Coleman, West and Wyss, 1973).
Pathology often is also a question of interpretation as exemplified by the contro-
versy over brain damage from monosodium glutamate (MSG), the source of the
Chinese Restaurant Syndrome. It has now been established that MSG kills cells in
the developing brain (Olney, 1980), at least in mice and rats, which is why it has been
removed from baby foods. Olney's early results were at first greeted with skepticism
but have been upheld by later studies. The disagreement arose because specialized
cells in the brain, the glia, consumed the debris left by the dead neurons and because
the locus of destruction was confined to a small area on the borders of the internal
fluid system (the ventricles) of the brain and spinal cord. In addition, the critics used
a traditional cell stain that is relatively insensitive to this kind of damage. By turning
to special techniques that differentially mark individual neurotransmitter systems
and by conducting coordinated assays of brain chemistry, the breadth of damage
from MSG and allied neurotoxic amino acids has been demonstrated (Kizer,
Nemeroff and Youngblood, 1978). With such data as guides, it was found that lower
doses alter the levels of certain critical hormones without producing overtly detect-
able cell pathology.
Neurochemistry—Like its morphology, the chemistry of the brain is remarkably
heterogeneous, and, like its interconnected web of structures, its chemical entities act
as components of an integrated, precisely orchestrated network. It is inevitable that
some environmental contaminants will interfere with one or more of the myriad
processes and mechanisms through which the brain's chemical functions are
131
-------
conducted. Consider a few of the possible steps at which such agents might interfere
with normal function: (1) dietary uptake and transport to the brain of required nu-
trients such as amino acids; (2) enzymatic conversion, by one or more steps, of amino
acid precursors into functional neurotransmitters; (3) storage of neurotransmitters
at specific sites within the nerve cell; (4) release of the neurotransmitter into the syn-
apse when the nerve is stimulated; (5) metabolism and removal of the transmitter
from the synapse; and (6) excitation responses of the postsynaptic neuron. Multiply
these actions by the 40 or so identified neurotransmitters and neuromodulators in
the central nervous system (Lipton, DiMascio and Killam, 1978) and the result is
almost as staggering an array of possibilities as that provided by behavior itself. Un-
less some specific information is available about the impact of a given chemical, the
neurochemist will have to proceed by systematic search, perhaps focusing on those
chemical constituents and processes extensively studied in psychopharmacology.
Metabolism and synthesis of the transmitter serotonin, for example, is increased by
DDT, and its concentration in the brain decreased by nickel, manganese and perhaps
cadmium as well (Rastogi, Merali, and Singhal, 1977).
Neurochemical methods, however, may find their widest application through test
systems that rely on special preparations of brain tissue. Nerve cell cultures, clumps
of nerve endings, brain minces and slices, and excised nerve-muscle preparations
might prove especially apt for the phase of early hazard screening and its emphasis
on identification.
Nonspecific measures—Behavioral toxicology legitimately focuses on the central
nervous system as the source of the mechanisms that determine behavior. Suchafo-
cus, however, should not block awareness that other systems within the body make
significant contributions to behavior through indirect actions on the central nervous
system. Pronounced falls in body weight, which are common when toxic exposures
occur, have been noted in association with lead, methylmercury, acrylamide, carbon
disulfide, and triethyl tin poisoning. Disturbances of energy regulation profoundly
influence many bodily processes and inevitably will be reflected in behavior. For ex-
ample, undernutrition affects brain chemistry (Michaelson, 1980). Any report of
neurobehavioral deficits accompanied by significant changes in body-weight
regulation or food intake must be tempered with that realization, and should move
the investigators to establish the independence and significance of their behavioral
observations apart from a profound interference with the body's vital economy.
The impact of other organ systems cannot be ignored either. A poison whose pri-
mary effect is impairment of liver function can easily exert profound effects on be-
havior; the neurological consequences of "hepatic coma" have been recognized since
antiquity. Liver impairment is associated with increased brain synthesis of the neuro-
transmitter, serotonin, probably because its precursor amino acid, tryptophan,
enters the brain more readily (Nutrition Reviews, 1980). Clinical descriptions of the
aftermath of hepatic coma feature a variety of nervous system disorders such as
tremor, grimacing, slurring of speech, incoordination, and impairment of intellectu-
al function (Victor, Adams and Cole, 1965). Behavioral and neurological disorders
have often been noted in clinical descriptions of uremia and other kidney diseases.
SHORT-TERM TESTS: SOURCES AND DEVELOPMENT
The ideal test is short, sensitive, easily quantified, simple to perform, and cheap.
Most of all, it is a valid predictor of long-term consequences to humans. Such a test,
given such an aim, is a self-contradiction. There are too many barriers to such an
ideal blending. Behavior and the central nervous system are viewed as especially irk-
some targets of chronic exposure because the cumulative impact of a toxic agent is so
difficult to distinguish from other influences, especially if the objective is to detect
incipient damage and to arrest its course. Chronic and acute exposures may also
132
-------
foster qualitatively different consequences. If specifications call for chronic evalua-
tion, why focus on short-term tests obliquely related to the main issues?
A question typically buried in the search for short-term assays is their reliability,
or reproducibility. Techniques that seem optimal for short-term tests, such as nat-
uralistic observation and activity measures, are often enormously variable from one
laboratory to another and from one occasion to the next (Bornschein, Pearson and
Reiter, 1980). Many investigations of lead-induced changes in motor activity were
spurred by somewhat insubstantial reports that the childhood syndrome of "Hyper-
activity" was significantly associated with the body burden of lead. Such a simplistic
translation neglects the most salient features of the clinical syndrome now called
"Attention Deficit Disorder" by the American Psychiatric Association. Given the
vastly different techniques of measurement and dosage regimens, the association
between motor activity and lead resists generalization. Even more bothersome is
what the survey by Bornschein, Pearson and Reiter (1980) could not tell us. First, it
could not estimate the reproducibility of the measure because the reports rarely
provide a quantitative estimate of consistency from one occasion to the next. If the
measures are highly variable, what is their predictive value? A second problem is
consistency from one laboratory to another. Activity measures are sensitive to many
variables. How many of these must be uniform to establish consistency among lab-
oratories? Further, if automated activity measures are subject to so many extraneous
influences, can naturalistic observations, which multiply these influences by varia-
tions among observers, show any greater consistency?
With more complex measures, reproducibility can be enhanced enormously.
Fixed-Interval Operant performance in Boston looks no different than Fixed-Inter-
val performance in Rome, and it is equally responsive to drugs. Individual animals
are also consistent. The same features of stability hold for psychophysics, which is
why we can study individual animals and make sense of the effects. To claim that
simple, "short-term" tests are cheaper than more complex procedures is specious.
What is the ultimate cost of a false positive or false negative? How many subjects
have to be studied and how much time has to be involved in order to reach a believ-
able conclusion? Perceptive and accurate cost accounting should guide our
decisions.
Short-term tests are obligatory for initial hazard identification because invest-
ments in detailed evaluation are unlikely unless some preliminary guides are avail-
able (Tilson, Mitchell and Cabe, 1979). They also offer estimates of the appropriate
range of exposure levels. Given these limitations of role, is there any way to improve
the sensitivity and selectivity of short-term test methods? Dews and Wenger (1979)
believe that enough information may be extracted from lethality determinations and
measures of spontaneous locomotory activity (and I would add energy regulation) to
make additional increments of information unreasonably expensive. Such a point of
view does not conflict with that voiced by Tilson, Mitchell and Cabe (1979). In
essence, find out, with minimal resources, if the job needs to be done (hazard identi-
fication). Then attack, with the best available measures, the main, crucial, enduring
issue: What is the risk?
REFERENCES
Aspey, W.P., and J.E. Blankenship. Spiders and snails and statistical tales: application of mul-
tivariate analyses to diverse ethological data. In Quantitative Methods in the Study of An-
imal Behavior, edited by B.A. Hazlett, Academic Press, New York, pp. 75-120, 1977.
Baum, W. M. Changing over and choice. In Quantification of Steady-State Operant Behavior,
edited by C.M. Bradshaw, E. Szabadi and C.F. Lowe, Elsevier, Amsterdam, pp. 67-78, 1981.
Bignami, G. Behavioral toxicity of anticholinesterase agents: methodological, neurochemical,
and neuropsychological aspects. In Behavioral Toxicology, edited by B. Weiss and V.G.
Laties, Plenum Press, New York, pp. 155-215, 1975.
133
-------
Bornschein, R., D. Pearson, and L. Reiter. Behavioral effects of moderate lead exposure in
children and animal models: Part 2, animal studies, Crit. Rev. Toxicol. 7:101-152, 1980.
Coleman, P., M. West, and U. Wyss. Computer-aided quantitative neuroanatomy. In Digital
Computers in the Behavioral Laboratory, edited by B. Weiss, Appleton-Century-Crofts,
New York, pp. 379-426, 1973.
Cranmer, J.S. Behavioral teratology and toxicology. In Behavioral Toxicology, edited by B.
Weiss and V.G. Laties, Plenum Press, New York, pp. 311-349, 1975.
Dews, P.B., and G.R. Wenger. Testing for behavioral effects of agents. Neurobehavioral
Toxicol. 1, Suppl. 7:119-127, 1979.
Duffy, F.H., and J.L. Burchfiel. Long-term effects of the organophosphate Sarin on EEGs in
monkeys and humans. Neurotoxicol. 7:667-689, 1980.
Edwards, P.M., and V.H. Parker. A simple, sensitive, and objective method for early assess-
ment of acrylamide neuropathy in rats. Toxicol. Appl. Pharmacol. 40:589-591, 1977.
Environmental Protection Agency. Chloromethane and chlorinated benzenes proposed test
rule: amendments to proposed health effects standards. Fed. Regist. 45 (740j:48524-48566,
July 18, 1980.
Evans, H.L., V.G. Laties, and B. Weiss. Behavioral effects of mercury and methylmercury.
Fed. Proc. 34:1858-1867, 1975.
Fox, W.M. Reflex-ontogeny and behavioral development of the mouse. Anim. Behav. 5:234-
241, 1965.
Glaser, E.M., and D.S. Ruchkin. Principles of Neurobiological Signal Analysis. Academic
Press, New York, 1976.
Hanninen, H. Psychological test methods: sensitivity to long-term chemical exposure at work.
Neurobehav. Toxicol. 1, Suppl. 7:157-161, 1979.
Health and Human Services Committee. DHEW CCERP Subcommittee Report on the Po-
tential Health Effects of Toxic Chemical Dumps. Released by the Eckhardt-Moffett
Hearing, May 22, 1980.
Hoffman, H., and J. Ison. Reflex modification in the domain of startle: I. Some empirical find-
ings and their implications for how the nervous system processes sensory input. Psvchol.
Rev. £7:175-189, 1980.
Hursh, S.R. Economic concepts for the analysis of behavior. J.Exp. Anal. Behav. 34:219-238,
1980.
Kaplan, M.L., and S.D. Murphy. Effects of acrylamide on rotarod performance and sciatic
nerve /3-glucuronidase activity of rats. Toxicol. Appl. Pharmacol. 22:259-268, 1972.
Kizer, J.S., C.B. Nemeroff, and W.W. Youngblood. Neurotoxic amino acids and structurally
related analogs. Pharmacol. Rev. 29:301-318, 1978.
Konigsberg, A.S., and C.H. Bachman. Ozonized atmosphere and gross motor activity of rats.
Int. J. Biometeorol. 74:261-266, 1970.
Leon, M. Filial responsiveness to olfactory cues in the laboratory rat. In Advances in the Study
of Behavior, edited by J.S. Rosenblatt, R.A. Hinde, C. Beer, and M.C. Busnel, Academic
Press, New York, pp. 117-153, 1978.
Lipton, M.A., A. DiMascio, and K.F. Killam. Psychopharmacology: A Generation of Pro-
gress Raven Press, New York, 1978.
Lowy, K., and B. Weiss. Assessing the significance of averaged evoked potentials with an on-
line computer: the split-sweep method. Electroencephalogr. Clin.Neurophysiol. 25:177-
180, 1968.
Madlafousek, J., Z. Hlinak, and J. Parizek. Sexual behavior of male rats sterilized by cadmi-
um. J. Reprod. Fertil. 26:189-196, 1971.
Marsh, D.O., G.J. Myers, T.W. Clarkson, L. Amin-Zaki, S. Tikriti, and M.A. Majeed. Fetal
methylmercury poisoning: clinical and toxicological data on 29 cases. Ann. Neural. 7:348-
353, 1980.
Merigan, W.H., and B. Weiss. Neurotoxicitv of the Visual System. Raven Press, New York,
1980.
Michaelson, I. A. An appraisal of rodent studies on the behavioral toxicity of lead: The role of
nutritional status. In Lead Toxicity, edited by R.L. Singhal and J.A. Thomas, Urban and
Schwarzenberg, Baltimore, pp. 301-365, 1980.
Needleman, H.L. Human lead exposure: Difficulties and strategies in the assessment of neuro-
logical impact. In Lead Toxicity, edited by R.L. Singhal and J.A. Thomas, Urban and
Schwarzenberg, Baltimore, pp. 1-17, 1980.
Nutrition Reviews. Hepatic encephalopathy: A unifying hypothesis Nutr Rev 35-371-373
1980.
134
-------
Olney, J.W. Excitotoxic mechanisms of neurotoxicity. In Experimental and Clincal Neurotox-
icologv, edited by P.S. Spencer and H.H. Schaumburg, Williams and Wilkins, Baltimore,
pp. 272-294, 1980.
Rastogi, R.B., Z. Merali, and R.L. Singhal. Cadmium alters behavior and the biosynthetic ca-
pacity for catecholamines and serotonin in neonatal rat brain. J. Neurochem. 25:789-794,
1977.
Reiter, L.W., and R.C. MacPhail. Motor activity: A survey of methods with potential use in
toxicity testing. Neurobehav. Toxicol. 1, Suppl. 1:53-66, 1979.
Satinoff, E. Independence of behavioral and automatic thermoregulatory responses. In Neural
Mechanisms of Goal-Directed Behavior and Learning, edited by R.F. Thompson, L.H.
Hicks, and V.B. Shvyrokor, Academic Press, New York, pp. 189-194, 1980.
Seppalainen, A.M. Peripheral nervous system in lead-exposed workers. In Behavioral Toxi-
cology. Early Detection of Occupational Hazards, edited by C. Kintaras, B.L. Johnson,
and I. de Groot, U.S. Department of Health, Education and Welfare, Washington, D.C.,
pp. 240-247, 1974.
Synder, E.W., E.G. Beck, and R.E. Dustman. Visual evoked potentials in monkeys. E/ectroen-
cephalogr. Clin. Neurophysiol. 47:430-440, 1979.
Squibb, R.E., N.G. Carmichael, and H.A. Tilson. Behavioral and neuromorphological effects
of triethyl tin bromide in adult rats. Toxicol. Appl. Pharmacol. 55:188-197, 1980.
Stanley, W.C. Feeding behavior and learning in neonatal dogs. In Second Symposium on Oral
Sensation and Perception, edited by J.F. Bosma, Thomas, Springfield, 111., pp. 242-290,
1970.
Stern, S., L. Margolin, B. Weiss, S.-T. Lu, and S.M. Michaelson. Microwaves: effects on
thermoregulatory behavior in rats. Science 206:1198-1201, 1979.
Tilson, H. A., C.L. Mitchell, and P.A. Cabe. Screening for neurobehavioral toxicity: The need
for and examples of validation of testing procedures. Neurobehav. Toxicol. I, Suppl.
/:137-148, 1979.
Ulfstrand, S., A. Sodergren, and J. Rabol. Effect of PCB on nocturnal activity of caged robins
(Erithaeus rubecula L.). Nature (London) 237:467^168, 1971.
Victor, M., R.D. Adams, and M. Cole. The acquired (non-Wilsonian) type of chronic hepato-
cerebral degeneration. Medicine(Baltimore) 44:345-396, 1965.
Weiss, B. The fine structure of operant behavior during transition states. In The Theory of
Reinforcement Schedules, edited by W.N. Schoenfeld, Appleton-Century-Crofts, New
York, pp. 277-311, 1970.
Weiss, B. The behavioral toxicology of metals. Fed. Proc. 37:22-21, 1978.
Weiss, B. Behavior as a sentry of metal toxicity. In Trace Metals Exposure and Health Effects,
edited by E. Di Ferrante, Pergamon Press, Oxford, pp. 185-198, 1979.
Weiss, B. Microproperties of operant behavior as aspects of toxicity. In Quantification of
Steady-State Operant Behavior, edited by C.M. Bradshaw, E. Szabadi, and C.F. Lowe,
Elsevier, Amsterdam, pp. 249-265, 1981.
Weiss, B.,and T. Gott. A microanalysis of drug effects on fixed-ratio performance in pigeons.
J. Pharmacol. Exp. Ther. ISO: 189-202, 1972.
Weiss, B., and V.G. Laties. Behavioral thermoregulation. Science 755:1338-1344, 1961.
Weiss, B., and V.G. Laties. (eds.). Behavioral Toxicology. Plenum Press, New York, 1975.
Willes, R.F., D.C. Rice, and J.F. Truelove. Chronic effects of lead in nonhuman primates. In
Lead Toxicity, edited by R.L. Singhal and J.A. Thomas, Urban and Schwarzenberg, Bal-
timore, pp. 213-240, 1980.
Wood, R.W. Reinforcing properties of inhaled substances. Neurobehav. Toxicol. 1, Suppl.
7:67-72, 1979.
Wood, R.W., A.B. Weiss, and B. Weiss. Hand tremor induced by industrial exposure to inor-
ganic mercury. Arch. Environ. Health 26:249-252, 1973.
135
-------
THE ROLE OF THE ENVIRONMENT IN CHRONIC
CARDIOVASCULAR DISEASE
C. Richard Minick and Carl G. Becker
INTRODUCTION
Cardiovascular diseases account for over one-half of all deaths in the United
States and are of similar importance in other developed Western countries. More
than 80 percent of these deaths can be attributed to complications of atherosclerosis.
Atherosclerosis is a change in the walls of musculoelastic and large muscular arteries
characterized by lipid-rich, fibrocellular thickening of the innermost or intimal coat
of the arterial wall. Atherosclerosis produces its effects by encroachment of the lu-
mens of arteries as a consequence of intimal thickening, thereby compromising
blood flow and decreasing the supply of oxygen and nutrients to organs or tissue.
The resulting lack of oxygen, or ischemia, may lead to death of tissue or infarction.
Atherosclerosis may also result in severe damage to the medial coat of arteries with
the consequent loss of elasticity or compliance of the arterial wall and formation of
areas of weakness or aneurysms. In many instances, the diminished blood supply
due to intimal thickening is acutely accentuated by sudden occlusion of the artery by
formation of a blood clot or thrombus on the altered luminal surface.
Atherosclerosis tends to involve preferentially certain arterial beds, coronary ar-
teries, cerebral arteries and arteries of the lower extremities with the result that the
effects are particularly striking in the organs and tissues supplied by these vessels.
The related clinical manifestations include angina pectoris due to ischemia of the
heart, death of myocardial tissue or myocardial infarct and sudden death or heart at-
tack, stroke or cerebral infarction, and loss of blood supply to the lower extremities
resulting in gangrene. In 1975, atherosclerosis resulted in 643,000 deaths from coro-
nary heart disease, 155,000 deaths from cerebrovascular disease and 55,000 deaths
from other diseases of the arteries. In the same year, the direct and indirect cost of
cardiovascular disease mostly due to atherosclerosis was estimated to be 50.4 billion
dollars—16.0 billion in direct costs, and 34.4 billion in indirect costs due to loss of
productivity as a result of either illness (8.7 billion) or death (25.7 billion).
RISK FACTORS IN ATHEROSCLEROSIS
The causes of atherosclerosis are poorly understood. Although atherosclerosis
has been known to occur since antiquity and is found in all populations, manifesta-
tions of atherosclerosis have reached epidemic proportions in the Western
industrialized societies in the twentieth century. Investigators have attempted to cor-
relate changes in the environment of particularly susceptible populations with the
increased incidence of the disease. These environmental risk factors can generally be
divided, albeit somewhat artificially, into changes in the personal environment and
changes in the overall environment. Alteration in the fat content of the diet which
The A uihors: C. Richard Minick and Carl G. Becker are staff members of the Department of Pathology, The
New York Hospital-Cornell Medical Center, New York, New York.
136
-------
may result in hyperlipemia is an example of change in personal environment while
changes in water hardness or availability of trace minerals are examples of changes
in the overall environment.
In general, these two categories of change also require a different approach to
their control. Personal risk factors, that is, diet-induced hypercholesterolemia, cig-
arette smoking and hypertension, may at least in part be modified by individual ac-
tion. Alterations in the general environment, such as air pollution, water hardness,
toxic chemicals, are more likely to require legislative action for their control. Thus
far, the environmental factors shown to enhance risk significantly are all personal
ones; they are diabetes mellitus, maleness, postmenopausal state, familial or genetic
factors and stress or "coronary prone" behavior patterns in addition to the ones
mentioned above. Obesity and physical inactivity are alo suspected risk factors.
Personal Risk Factors
High concentrations of blood lipid, particularly cholesterol, have been shown to
be significantly correlated with the increased risk of ischemic heart disease (IHD) in
comparisons between populations with a low incidence of arteriosclerotic heart dis-
ease and those with a high incidence and also in studies within populations that com-
pared those persons with increased risk to the remaining population (Keys, 1970;
Kannel et al., 1971). Serum cholesterol levels have been found to be related to intake
of animal protein and saturated fat (Connor and Connor, 1972). Increased
cholesterol and cholesterol ester constitute the major lipid component of athero-
sclerotic plaques, and there is incontrovertible evidence that increased serum cho-
lesterol concentration contributes to the development of atherosclerosis in experi-
mental animals. Cholesterol carried in certain blood proteins which transport cho-
lesterol and other blood lipids are also correlated with IHD. The quantity of low-den-
sity-lipoprotein cholesterol correlates positively with the incidence of IHD in a pop-
ulation. In contrast, the quantity of high-density-lipoprotein cholesterol correlates
inversely with the incidence of IHD (Miller and Miller, 1975; Gordon et al., 1977).
Finally, studies in nonhuman primates suggest that low-density lipoproteins are
heterogeneous and that a high molecular weight low-density lipoprotein is particu-
larly associated with increased risk of coronary atherosclerosis (Rudel et al., 1979;
1981). Other studies, however, indicate that differences in serum cholesterol cannot
entirely account for increased risk either between or within populations (Gordon et
al., 1974; Garcia-Palmieri, 1973). Further, dietary and other intervention studies
have not had the anticipated effect in reducing mortality from IHD (Dayton et al.,
1969; Oliver, 1980). High blood pressure is also significantly correlated with risk of
development of IHD and the risk increases directly with blood pressure levels even in
the so-called normal range. Blood pressure was found to be one of the strongest
factors in the Framingham Study with almost a ten-fold range of IHD morbidity
over the measured range of systolic blood pressure (Kannel et al., 1976).
Cigarette smoking is associated with an increased risk of IHD, including Sudden
Death; data from the International Atherosclerosis Project indicate a relationship
between coronary artery-raised lesions and a history of cigarette smoking (Strong et
al., 1969). The risk of death for men at all ages and for women below the age of sixty-
five is directly correlated to the number of cigarettes smoked and is about 60 to 70
percent greater for cigarette smokers than nonsmokers. The younger smoker is
generally at greater risk than are older smokers. The excess mortality for cigarette
smokers decreases when smoking is stopped, and the mortality rate gradually
returns toward the normal rate over a number of years.
Several complications resulting from atherosclerosis including coronary heart
disease, sudden death, peripheral vascular disease and stroke occur more commonly
among diabetics than nondiabetics, and the difference appears particularly striking
in younger age groups. The association with peripheral artery disease is a particularly
137
-------
strong one and the vast majority of patients with this disease can be shown to be
overtly diabetic, have abnormal glucose tolerance tests or to be cigarette smokers.
Although diabetes is often associated with increased incidence of hyperlipidemia,
hypertension, and obesity, the changes in these factors are not believed to be of suffi-
cient magnitude to explain the increased incidence of cardiovascular morbidity and
mortality. It has been suggested that increased mortality may be mediated by other
factors, perhaps involving arterial wall metabolism (Wolinsky, Goldfischer et al.,
1978).
Other factors have been associated with increased risk of arteriosclerotic heart dis-
ease. Among these are sex, family history, personality type and physical inactivity.
Premenopausal women have a significantly lower incidence and severity of coronary
artery disease than women of the same age with artificial menopause. The use of oral
contraceptives has been found to increase the risk of IHD disease, particularly in
young women (Hennekens and MacMahon, 1977; Lancet Editorial, 1977). This risk
is further increased when associated with cigarette smoking (Jain, 1976).
Inactivity has been receiving increased attention as a risk factor particularly be-
cause of recent findings which indicate that physical activity, specifically long-dis-
tance running, is associated with a proportional increase in the high-density lipopro-
tein fraction, lipoproteins that we previously noted were inversely correlated with
increased risk of atherosclerosis. In addition, there is a decrease in total serum cho-
lesterol (Paffenbarger and Hyde, 1980). Obesity is also a risk factor although many of
its effects may be mediated through other risk factors such as hyperlipidemia, hyper-
tension and diabetes. Obese individuals also have lower levels of high-density lipo-
protein and are less physically active.
Finally, there are families in which heart attacks occur prematurely and with in-
creased frequency. In some instances, familial predisposition is clearly associated
with readily identifiable factors such as various types of familial hyperlipoprotein-
emia (Frederickson and Ferrans, 1978). However, in others, it is difficult to be cer-
tain what the relative contribution of life style, environment and dietary habits are as
compared to genetic factors. It is also uncertain in some instances how these genetic
factors operate although it is becoming increasingly apparent that there are hered-
itary components to a number of risk factors including hyperlipoproteinemia, hy-
pertension and diabetes mellitus (Department of Health, Education and Wel-
fare, 1971). Mathews (1975) has suggested a relationship between
population frequencies of the histocompatibility antigen HLA-8 and haplotypes 1-8
and the frequency of IHD.
Environmental Risk Factors
Although certain personal factors have been shown to be correlated with increased
risk of ischemic heart disease, these factors are believed to account for less than 50
percent of the increased risk in certain Western populations (Stamler and Epstein,
1972; Gordon et al., 1974; McGill, 1978). Large numbers of individuals have compli-
cations of coronary artery disease even though known risk factors are lacking.
Severe atherosclerosis and its complications also occur in a few individuals within
low-risk populations who share the characteristics leading to decreased risk in that
population. Taken together, these findings suggest that there are other risk factors,
acting either alone or in synergy with the known ones. There may be legitimate con-
cern, therefore, about the effects of substances that are encountered either occupa-
tionally or in our general living situation. The recognition that some environmental
factors may be important in initiation or growth of neoplasms has further strength-
ened concern with regard to their influence on other chronic disease processes.
However, it is important to point out that, even in the instance of neoplasia, personal
factors are believed to contribute much more to increased risk than those in the
general environment (Waugh, 1979). General environmental factors that have been
considered to be of possible importance include carbon monoxide, water hardness
138
-------
and intake of trace metals, occupational exposure to certain inhalants, noise, radio-
frequency irradiation, and physical and psychophysical stress (Rosenman, 1979;
Harlanet al., 1981).
Carbon Monoxide—
Carbon monoxide (CO) is one of the more common pollutants in most industri-
alized societies. Worldwide CO production from technological and natural sources
is estimated to be at least 250 million tons per year (Astrup and Kjeldsen, 1974). Al-
though fuel consumption by motor vehicles is a major source of carbon monoxide,
the most frequent source of inhaled CO is in tobacco smoke. Cigarette smoke con-
tains between 2.7 and 6 percent CO and during inhalation and dilution with air aver-
ages about 440 ppm of CO, a level eight times greater than the maximum level per-
mitted in industry. Consumption of one to two packs of cigarettes daily may lead to
alveolar CO concentrations of approximately 50 ppm which, at equilibrium with
blood, equals a carboxyhemoglobin (COHb) concentration of about seven percent
(Ball and Turner, 1974). Ayres et al. (1969) demonstrated COHb concentrations in
nonsmokers of 0.5 percent as compared to concentration in smokers of one or two
packs of 1 and 3 percent, respectively. Certain occupational groups such as fire
fighters and vehicular tunnel workers also experience elevated CO exposure (Ayres
et al., 1969; Sammons and Coleman, 1974).
Increasing concentrations of carboxyhemoglobin (COHb) have been found to in-
fluence myocardial oxygen utilization in patients with ostensibly normal hearts as
well as those with IHD when patients are subjected to exercise. Vogel and Gleser
(1972) found that exposure to 225 ppm CO, resulting in COHb saturations of 18 to
21 percent, had little effect on normal resting myocardial oxygen consumption but
resulted in a decrease in total oxygen consumption of 23 percent at maximum exer-
cise. Increased COHb saturation to 9 percent in patients with no evidence of coro-
nary heart disease was found to result in increased coronary blood flow and decreased
myocardial oxygen extraction (Ayres etal., 1969). In contrast, in patients with coro-
nary heart disease, a rapid increase in COHb did not result in a significant increase in
coronary blood flow indicating that they have decreased ability to compensate for
increasing COHb concentrations (Ayres et al., 1970; 1973).
As might be expected from these findings, Anderson etal. (1973) found that adult
men with stable angina pectoris had significantly shorter treadmill exercise times be-
fore onset of pain and that the pain lasted for a longer period of time when exposed
to CO and stressed with exercise. EKG changes also occurred earlier and lasted for
longer periods. Similarly, Aronowetal. (1972) demonstrated that exposure to vehic-
ular CO in heavy automobile traffic resulted in a significant decrease in exercise per-
formance before the onset of angina pectoris and EKG changes. Breathing
increasing concentrations of CO has also been found to shorten exercise tolerance in
patients with significant iliofemoral occlusive arterial disease (Aronow et al., 1974).
A study in Copenhagen indicated that increased levels of COHb correlated with
increased tendency to development of atherosclerosis; those with COHb levels of
five percent or more were found to be 21 times more likely to be affected than matched
controls with COHb below three percent (Wald et al., 1973). Studies of Cohen,
Deane and Goldsmith (1969) have indicated increased fatality rate from myocardial
infarction in areas of high pollution during periods of relatively increased CO con-
centration. However, others have attempted unsuccessfully to correlate the number
of myocardial infarcts with levels of ambient CO including those associated with
sudden death (Kuller et al., 1975). Exposure to increased atmospheric concentra-
tions of CO has also been found to increase myocardial irritability in humans and to
lead to myocardial fibrosis in rabbits (Kjeldsen et al., 1974).
Unfortunately, after many years of experimentation, the findings with respect to
atherosclerosis are unclear. Astrup (1974) and Kjeldsen etal.(1968, 1972)found that
exposure to increased concentrations of CO led to increased deposition of lipid in
139
-------
aortas of rabbits fed cholesterol-supplemented diets. Interpretation of the results of
these experiments is complicated by the finding that serum cholesterol concentra-
tions in animals exposed to CO were significantly higher than diet-fed controls.
Further, the same investigators were unable to demonstrate an effect in rabbits with
normal blood cholesterol levels, a result which was at variance with their earlier ex-
periments (Hugod et al., 1978). In nonhuman primates, exposure to increased CO
has been found to result in anatomical damage to coronary arteries (Thomsen and
Kjeldsen, 1974).
Increased concentrations of CO in the atmosphere has also been found to have an
effect on fetal development, vascular permeability and cerebral function
(McFarland et al., 1944; Lilienthal and Fugitt, 1945; Schulte, 1961, 1963; Siggard-
Andersen et al., 1967; Pauli et al., 1968; Siggard-Andersen etal., 1968; Astrupetal.,
1972; Parving et al., 1972). Impairment of cerebral function occurs at the levels of
COHb found in the smoking population.
The results of the above studies do not support the concept of a safe threshold level
of COHb. In the presence of regional vascular insufficiency, even the lowest levels of
COHb may critically interfere with adequate oxygen delivery. A safe level of CO ex-
posure appears to be the lowest level which can be achieved (Turino, 1981).
Water Hardness and Trace Metals—
Since the original observations of Kobayoshi (1957) noting a relatively high death
rate from stroke among Japanese with soft water supplies, many investigators have
attempted to correlate water hardness with mortality from cardiovascular diseases,
particularly arteriosclerosis. Water hardness is defined as a condition when the con-
centration of calcium carbonate is equivalent to the dissolved calcium and magne-
sium present (Sharrett, 1981). Data with respect to water hardness may be confound-
ed with trace mineral intake. However, overall, none of the trace elements is consist-
ently correlated with water hardness, but in some areas at least the concentration of
magnesium appears to be independently associated (Sharrett, 1981). Several large
studies made in England, Wales, Canada, North America and an International study
have failed to resolve the question of a consistent correlation between water hardness
and mortality rates from cardiovascular disease in general and atherosclerotic
cardiovascular disease in particular (Comstock, 1979). The association, a weak one
at best, tends to become insignificant when smaller included areas are examined
separately or when other related circumstances such as climate or rainfall are con-
sidered (Sharrett, 1981).
Magnesium is often found in relatively high concentrations in hard water, and
intake from drinking water may account for as much as 10 to 20 percent of the total
magnesium ingestion (Sharrett, 1977). There is experimental and clinical evidence to
indicate that magnesium deficiency can cause disturbances of cardiac rhythm
(Polimeniand Page, 1973; Iseri et al., 1975) and epidemiologic evidence that at least
some increased cardiovascular mortality associated with soft water may be due to
Sudden Death (Peterson et al., 1970; Anderson and LeRiche, 1971; Crawford etal.,
1977). Taken together with the finding of low myocardial magnesium levels in acci-
dent victims in these areas (Anderson et al., 1975) and in association with Sudden
Death (Behrand Burton, 1973; Anderson etal., 1975; Chipperfield and Chipperfield,
1978), this latter finding has been interpreted as evidence that inadequate magne-
sium is a causal factor in cardiovascular disease, possibly through its effect on cardi-
ac rhythm. However, an independent association has not been established. Some
British studies indicate insignificant or trivial associations with magnesium content
of drinking water even though cardiovascular disease is significantly associated with
water hardness in the same areas (Morris et al., 1961; Crawford etal., 1968;Elwood
et al., 1977). Further studies on magnesium metabolism and magnesium require-
ments in man would be useful (Sharrett, 1981).
140
-------
Other investigators have attempted to correlate changes in incidence of athero-
sclerosis with trace minerals and not necessarily with those minerals that determine
water hardness although in many studies the two are confounded (Borhani, 1981).
Clinicopathologic, epidemiologic and experimental data suggest that excessive ex-
posure to several trace elements or insufficient dietary intake may be associated with
changes in the cardiovascular system which may lead to coronary artery disease. Al-
though an effect of industrial exposure has yet to be documented, exposure to cobalt
in beer has been shown to lead to a cardiomyopathy similar to that resulting from
thiamine deficiency, presumably by interfering with mitochondrial energy metabo-
lism (Alexander, 1969). Changes in the electrocardiogram have been noted following
exposure to antimony (Brieger et al., 1954), and copper smelter workers exposed to
arsenic have been reported to have increased cardiovascular mortality (Axelson et
al., 1978).
Increased lead concentrations have been observed in tissues and blood of patients
with hypertension (Schroeder and Kraemer, 1974; Wedeen et al., 1975; Beevers,
Erskine et al., 1976) and in the aortas of patients with coronary heart disease (Voors
et al., 1973). Increased cadmium concentrations have also been found in tissue and
blood of some patients with hypertension (Schroeder and Buchman, 1962; Axelson
and Priscator, 1966; Glauser et al., 1976) although others have not been able to con-
firm these findings (Beevers, Campbell et al., 1976).
Increased zinc to copper ratios have been reported in association with coronary
heart disease (Klevay, 1975), and there has been speculation that decreases in zinc,
with a decreased zinc to copper ratio, may be part of the protective effect that is seen
with physical exercise and exposure to hard water. It is also suggested that increased
ischemic heart disease in patients with hypertension is due to a change in zinc to cop-
per ratios as a result of lowered copper concentrations following increased urinary
excretion of copper.
Decreased concentrations of chromium have also been observed in aortas of
patients dying of arteriosclerotic cardiovascular disease as compared to those dying
of accidents or other diseases (Schroeder et al., 1970). Chromium concentrations in
aortas of North Americans were found to be low and declined with age (Schroeder,
1968; Schroeder et al., 1970) whereas those of African or Oriental populations were
much higher. Aortic tissue of Thailanders had more chromium than any other groups
and they have very little aortic atherosclerosis (Schroeder, 1958). Chromium defi-
ciency has also been associated with abnormal glucose tolerance and elevated serum
cholesterol concentrations (Schroeder and Balassa, 1965; Schroeder, 1966).
Abraham and co-workers (1980a) were unable to establish a relationship between
serum chromium levels and ischemic heart disease or abnormal blood glucose levels.
However, they did observe an elevation in patients with acute myocardial infarction
suggesting the possibility that large amounts of chromium may be liberated acutely
from myocardial tissue following injury. These same investigators obtained data to
suggest that intraperitoneal injections of chromium salts may lead to enhanced
regression of diet-induced aortic atherosclerosis in rabbits (Abraham et al., 19806).
A cause-and-effect relationship has not been established, however.
In conclusion, the evidence for implication of zinc, copper, lead, cadmium and
chromium in coronary artery disease is perhaps more plausible than that of other
elements, but in no case has a strong or consistent association been established. Even
assuming a consistent association, their role as independent risk factors has not been
established nor has the mechanism of action of these elements in normal biological
processes and in cardiovascular disease been determined.
Inhalants—
With respect to coronary artery disease, only two occupational inhalants have
been found to have a definite effect. A connection has been reported between an in-
141
-------
creased incidence of ischemic heart disease and exposure to carbon disulphide in the
viscose rayon industry (Nurminen, 1976) and glycerol nitric acid esters in the explo-
sives manufacturing industry (Weill, 1981). In the latter instance, the risks appear to
be greatest following cessation of exposure suggesting that the mechanisms may re-
sult from a rebound phenomenon perhaps with vessel spasm.
Halogenated hydrocarbons and fluoro-alkanes have been implicated in
myocardial irritability and cardiac toxicity (Lehnert et al., 1974; Ravens et al., 1974;
Flowers et al., 1975; Speizer et al., 1975). The overuse of bronchodilators for the
treatment of asthma has also been cited in cardiovascular disease. In addition,
benzene derivatives, pesticides, the organosilicon monomer methyl methacrylate,
butyl isocyanate, cadmium, nitrogen oxide, cobalt and arsenic have all been sug-
gested as cardiotoxins in individual cases following occupational exposure. None-
theless, only carbon disulfide and glycerol nitrates have been well documented as oc-
cupational inhalants that have a decided effect in cardiovascular disease (Weill,
1981).
Physical Agents—
Vibratory stress is a well-known environmental hazard for those using pneumatic
drills, chain saws, hand and pedestal grinding and copper swaging. Stress associated
with continued vibration leads to transient paling and paraesthesis of the fingers. If
exposure is continued, eventually persistent paleness of the finger, the so-called vi-
bration white finger, results due to permanent changes in blood supply and vessels
(Walton, 1974; Department of Health, Education and Welfare, 1977).
Moderate increases in blood pressure have been observed in response to high or
low noise levels (Ortiz et al., 1974). Increases in adrenal secretion, particularly in
medullary hormones (Ortiz et al., 1974), as well as increases in serum cholesterol
(Friedman et al., 1967; Ortiz, 1974) and platelet aggregation (Maas, 1973) have also
been noted in experimental animals following exposure to noise. Considering the
prolonged exposure of workers in various industries to increased noise levels, further
studies appear especially warranted. Exposure to radiofrequency radiation has also
been suggested to have possible acute and chronic effects in the cardiovascular
system (Resnekov, 1981). Unfortunately, most of the available data is far from con-
clusive, and the mechanism of action of these environmental stimuli are unknown
(Harlan, 1981).
Psychological and Social Stress—
Certain personality or behavioral patterns have been shown in several studies to be
related independently to the development of coronary heart disease (Brand et al.,
1976; Jenkins, 1976; Haynes, Feinleibetal., 1978; Haynes, Levine etal., 1978; Weiss,
1981). Although the results of these studies are highly interesting, they are compli-
cated by a lack of clear definitions of the relative contribution of behavior patterns
and psychological stress on the cardiovascular system.
Changes in sociocultural factors have also been cited as being important in cardio-
vascular disease, primarily high blood pressure and IHD. Changes in incidence of
these diseases are particularly striking for groups moving between societies or cul-
tures but have also been observed in association with cultural changes following
moves within a society or country (Gampel et al., 1962; Page et al., 1974). It is not
clear to what degree the effect of these moves may be directly attributed to social
stress as compared to that portion of the increased risk that is contributed by other
environmental changes such as diet, physical activity and level of health care.
The role of occupational stress in cardiovascular disease is also difficult to assess.
Even though there are differences in cardiovascular mortality and morbidity in dif-
ferent occupations, there is no direct evidence that links these conditions to job-re-
lated stress. Further, the assignment of stress levels to various job classifications is
142
-------
externally perceived and may have no validity with respect to stress on various organ
systems.
Overall, the role of psychosocial stress in cardiovascular disease is an important
area for future investigations, especially since the interaction between stress and be-
havioral or personality traits, if shown to be important, may be amenable to various
types of modification.
In summary, general environmental risk factors may be potentially important be-
cause of (1) the need to explain the increase in atherosclerosis and related diseases in
Western industrialized societies; (2) the magnitude of exposure to many of these
general environmental factors; (3) preliminary data available concerning the con-
tribution of some factors; and (4) the particular pertinence of some of these factors to
societies at high risk. However, for the most part the available epidemiologic data is
at best only suggestive. As is emphasized above, there is need for additional rigorous
studies to investigate these areas further. Since evidence to implicate any of these
factors is so weak, it would be inappropriate at this time to suggest legislative action
except in those few areas where the data are persuasive.
MECHANISMS OF DEVELOPMENT OF ATHEROSCLEROSIS
It is important to understand more about the manner by which well-established
risk factors may contribute to the genesis and progression of atherosclerosis. An
understanding of the mechanism(s) may suggest possible interventions in lesion ini-
tiation and development even if the particular risk factor responsible cannot be en-
tirely eliminated or significantly modified. For example, evidence that platelets may
be a factor in atherogenesis by causing smooth muscle cell proliferation and intimal
thickening had led several investigators to attempt to inhibit the development of
atherosclerosis by interfering with platelet function either by platelet depletion or by
pharmacologic inhibition of platelet function.Inhibition of platelet reaction with the
vessel wall has been found to decrease significantly the extent of intimal thickening
induced in experimental animals by mechanical injury, chemical injury associated
with homocystinemia, and immunologic injury due to graft rejection (Kincaid-
Smith, 1969; HarkeretaL, 1974; 1976; Friedman etal., 1977; Lurieetal., 1981). In
the instance of immunologic injury due to graft rejection, these findings may be par-
ticularly pertinent since they suggest that graft-induced atherosclerosis in organ
transplants may be almost completely eliminated by pharmacologic inhibition of
platelet function. In this regard, it is well to remember that accentuated atheroscle-
rosis induced in allografts has been found to be a primary long-term complication of
organ transplantation in humans, particularly heart transplants (Rider et al., 1975).
In subsequent discussions we will consider how some of the above risk factors may
accentuate arteriosclerotic cardiovascular disease either by interaction with cur-
rently accepted mechanisms of atherogenesis or by other mechanisms.
There are at least three major hypotheses that are considered to be important in ex-
plaining the pathogenesis of atherosclerosis: the lipid filtration hypothesis, the arte-
rial injury hypothesis, and the recently proposed monoclonal hypothesis. Each has
potentially important implications with respect to our environment. The lipid hy-
pothesis is environmentally important with respect to diet and other environmental
factors that may elevate blood lipids, in particular cholesterol. The injury hypothesis
relates importantly to the environment through factors that injure arteries and may
predispose to atherosclerosis. Finally, chemical or viral agents in the environment
may lead to mutation of proliferating smooth muscle cells resulting in focal areas of
intimal thickening—arteriosclerotic plaques—that are monotypic. Obviously these
hypotheses are not mutually exclusive. Hypercholesterolemia could result in arterial
injury, and the response of the arterial wall is probably modified by hyperlipemia.
Further injury could result in proliferation of a monotypic population of smooth
muscle cells through a process of selection.
143
-------
According to the lipid filtration hypothesis, lipid accumulation is a result of filtra-
tion of increased quantities of cholesterol-bearing lipoprotein across the endothelial
lining of the arterial wall as a consequence of increased serum levels. This increased
filtration leads to lipid accumulation in the arterial wall and secondarily to intimal
proliferation and atherosclerosis. If this hypo'thesis were the sole explanation, one
would anticipate that lipid accumulation would occur in a short period of time fol-
lowing the onset of hypercholesterolemia, and atherosclerosis would be similar in all
arterial beds and within all portions of an arterial bed. However, it is well established
that certain portions of the arterial tree are more prone to lipid accumulation and
atherosclerosis than other portions. In experimental animals, it has also been repeat-
edly observed that lipid accumulation and lesion development are not initiated im-
mediately following the onset of hypercholesterolemia but may have their onset
several weeks later, suggesting that hyperlipemia leads to other events in the arterial
wall that are necessary for the deposition of lipid.
Studies of human autopsies and of naturally occurring and experimentally in-
duced atherosclerosis in animals indicate that injury to arteries and the reaction of
the arterial wall are fundamental processes in the pathogenesis of atherosclerosis.
Evidence obtained during the last two decades has led to a reformulation of the arte-
rial injury hypothesis. These findings include the observation that endothelial
damage is an important common feature of many injuries including those resulting
from mechanical, chemical and immunological stimuli (Ross et al., 1974; Minick,
1976; Ross and Harker, 1976). Further, the blood and blood elements have been
shown to contain mitogens that stimulate smooth muscle cell proliferation. Impor-
tant mitogens are contained within the alpha granule of the platelet, the mononu-
clear cell and low-density lipoproteins in hypercholesterolemic serum (Ross et al.,
1974; Fisher-Dzoga et al., 1974; Ross and Vogel, 1978; Martin etal, 1981). Thus, in
its modern form, the injury hypothesis proposes that endothelial injury and desqua-
mation lead to a reaction of platelets and other blood elements with the damaged
arterial wall. Platelets and perhaps also macrophages then release mitogens that
stimulate smooth muscle cell proliferation and deposition of connective tissue re-
sulting in thickening of the innermost or intimal coat of the artery. Loss of the barrier
function of the endothelium enhances the filtration of macromolecules such as
cholesterol-rich lipoproteins into the wall resulting in intimal liquid accumulation
and atherosclerosis.
The concept that injury and reactive changes in the arterial wall may lead to ather-
osclerosis is not new. Virchow hypothesized that inflammation in arteries led to the
development of atherosclerosis and suggested that injury to the innermost coat in-
cited this inflammation (Virchow, 1858; Long, 1967; Saphir, 1967). The fact that
arteriosclerosis shares important features with inflammation continues to be at-
tractive to pathologists (Joris and Majno, 1978).
Rokitansky (1856) and, later, Duguid (1949) suggested that encrustration of
blood components on the lining of blood vessels at sites of injury, that is, formation
and organization of thrombi, resulted in the development of arteriosclerotic plaques.
Thus, although the injury and thrombogenic hypothesis may be viewed as different,
they in fact do share important features. Moreover, a large body of observations in-
dicates that events leading to inflammation and to coagulation of blood are often the
same or at least cooperative. This relationship becomes especially apparent when in-
flammation is induced via mechanisms related to immune reactions, both humoral
and cellular. Recent studies have implicated lipoproteins in changes in function of
components of inflammation and blood coagulation. Very low-density, intermedi-
ate, and high-density lipoprotein were found to stimulate a procoagulant activity by
blood monocytes; low density lipoprotein suppressed this stimulation (Schwartz et
al., 1981).
Many environmental substances may contribute to arterial injury and result in
atherosclerosis. In the instance of immunological injury, antigens in foreign pro-
144
-------
teins, drugs or microbes, ingested as foodstuffs or inhaled, as in tobacco smoke,
could lead to immunologic injury to blood vessels. Although we have a better under-
standing of these mechanisms today in molecular terms and have clearly demon-
strated that immunologic injury to arteries and reactive changes can lead to athero-
sclerosis, the concept that immunologic injury to vessel walls might contribute to
atherosclerosis was appreciated nearly 50 years ago. Autopsy studies of patients who
had experienced multiple attacks of rheumatic fever indicated that some had
developed precocious arteriosclerosis which was especially apparent in coronary
arteries (Zeek, 1932a, b). Fahr (1921) had drawn attention to the resemblance be-
tween the arterial lesions of rheumatic fever, polyarteritis nodosa and dermatomyo-
sitis. Similar arterial lesions involving the intima, media and adventitia of blood ves-
sels were also described in tissues of patients suffering from serum sickness following
injection of horse antipneumococcal serum (Klinge, 1933;Schmitt, 1935-1936; Rich,
1946; Rich and Gregory, 1947). These studies also commented on the predilection of
coronary arteries for acute immunologically mediated injury.
More recently, precocious development of atherosclerosis has been described in
association with systemic lupus erythematosus and rheumatoid arthritis (Bywaters,
1957; Tsakraklides et al., 1974), both diseases mediated at least in part by deposition
of circulating antigen-antibody complexes. Precocious atherosclerosis in cardiac
and renal homografts is presumed to be mediated by cellular or perhaps humoral
immune reactions (Rider et al., 1975).
Taken together, these studies of human autopsy material indicated that: (1) dis-
eases mediated by immunologic mechanisms caused arterial injury characterized by
inflammation of the walls of arteries and often thrombus formation; (2) these sites of
injury could acquire lipid and evolve into atherosclerotic plaques or into primarily
proliferative lesions containing little lipid; and (3) the evolution of these lesions
could be rather rapid, perhaps because of protracted or recurrent episodes of injury,
leading to the precocious development of atherosclerosis.
Early studies of experimental serum sickness indicated that arteries were targets
of injury (Klinge, 1933; Schmitt, 1935-1936; Rich, 1946; Rich and Gregory, 1947). It
was later demonstrated that formation of antigen-antibody complexes in slight to
moderate antigen excess in the circulation and their subsequent deposition in the
walls of blood vessels led to the development of arteritis (Dixon et al., 1958). Such
depositon in rabbits could be inhibited by administration of antagonists of hista-
mine or serotonin or by depletion of platelets (Kniker and Cochrane, 1968). It was
further demonstrated that release of histamine from platelets could be mediated by
complement components or by a mechanism requiring IgE and basophilic leuko-
cytes. Study of the latter phenomenon revealed that when antigen reacted with
specific IgE antibodies on the surface of basophils or mast cells the release of hista-
mine was triggered. In addition, the stimulated cells also released a lipid initially
called platelet activating factor (PAF) and recently identified as acetyl glyceryl ether
phyophorylcholine which caused platelets to aggregate transiently and to release
vasoactive amines (Benveniste, 1974; McManus et al., 1980). Release of PAF also
causes transient disappearance of platelets and leukoytes from the circulation
(Pinckard et al., 1977). Further, PAF-induced aggregation of platelets can cause
accumulation of platelets at sites of immune complex deposition (Kravis and
Henson, 1977). In addition, it has been shown that challenge with antigen of animals
preferentially synthesizing IgE antibodies to that antigen also results in activation of
the clotting system (Pinckard et al., 1975). In this connection it has been demon-
strated that challenge of sensitized basophils also results in release of enzymes capa-
ble of cleaving and activating prekallikrein or Fletcher factor, which in turn, can ini-
tiate activation of the intrinsic pathway of coagulation leading to clot formation,
fibrinolysis, and the generation of bradykinin, thus potentiating the inflammatory
response and increasing vessel permeability (Newball, Berninger et al., 1979;
145
-------
Newball, Talamo et al., 1979). This set of reactions might also enhance permeability
of the vessel wall to other macromolecules not involved in the initial process,
including plasma lipoproteins.
It must be pointed out that exact mechanisms are unknown by which IgE-
mediated release of vasoactive amines enhances deposition of circulating immune
complexes. It is conceivable that release of vasoactive amines and/ or generation of
bradykinin stimulates endothelial contraction as occurs in venules (Majno et al.,
1969), but arterial endothelial contraction has not been demonstrated unequivocally
though the endothelium of systemic arteries contains contractile protein (Becker and
Murphy, 1969; Becker and Nachman, 1973), and cultured umbilical vein endothelial
cells contain smooth muscle myosin (Moore et al., 1977). Further, it is not known if
immune complex deposition is potentiated in all arteries by these mechanisms. For
example, coronary arteries are commonly injured in experimental serum sickness
whereas cerebral vessels rarely are.
The surfaces of endothelial cells themselves may play a role in localization of im-
mune complexes. There is some disagreement as to whether normal endothelial cells
have receptors for the Fc portion of immunoglobulin molecules or for complement
components. Immunoglobins bind to cells with Fc receptors which are avid for the
Fc of the immunoglobulin molecule. This site is exposed after reaction with the
specific antigen. A product of cleavage of the third component of complement,
C3b binds to specific receptors and mediates activation of other components of com-
plement. One report indicates that these receptors are present (Shadforth et al.,
1979). Other investigators could not demonstrate Fc or C3b receptors on normal
human pulmonary and umbilical vein endothelial cells in culture (Ryan et al., 1980;
Lyss et al., 1980). However, when these cells were infected with Herpes simplex virus
(Type 1), they synthesized Fc and C3 receptors (Lyss et al., 1980).
Synthesis of Fc and C3 receptors may be an adaptive function serving to enhance
the clearance of foreign antigen associated with infection. However, it may be adap-
tation at a price, increasing the chances of vascular injury, especially if the quantity
of circulating complexes is large. These findings may be especially pertinent to de-
generative vascular disease since virus infections have recently been shown to lead to
atherosclerosis (Fabricant et al., 1978; Minick et al., 1979).
Whether immune complexes will deposit in the walls of blood vessels is also influ-
enced by the size of the complex (those weighing 106 daltons or more are likely to de-
posit), and by the quality of antibody, that is, its avidity for antigen (Cochrane and
Hawkins, 1968; Germuth et al., 1972).
The quality of antigen may also be important. The high avidity of DN A for vascu-
lar collagen may play a role in the deposition of DNA-antiDNA complexes in vessel
walls in patients with systemic lupus erythematosus. Furthermore, it has been dem-
onstrated that a glycoprotein antigen present in cured tobacco leaves and cigarette
smoke is not only an allergen, but because of its polyphenol hapten groups is capable
of activating the factor XII- or Hageman factor-dependent pathways of coagulation,
fibrinolysis, and kinin generation, events which might enhance focal inflammation
(Becker and Dubin, 1977). In this connection, it was found that activation of factor
XII— dependent pathways by intravenous injection of ellagic acid or rutin (known
polyphenol activators of factor XII) results in inflammation in the vessel of the gall-
bladder and to a lesser extent of the lung, but not in vessels of other viscera (Figs. 1
and 2). These latter findings emphasize the selectivity of various types of injury for
certain vascular beds.
Although it is not precisely known how circulating immune complexes deposit in
or on vessel walls, depletion of complement prevents all but extremely mild changes
in artery walls in animals injected with foreign protein (Henson, 1977). Circulating
immune complexes activate the complement cascade and certain components
become associated with the complex. Chemotactic factors derived from complement
components, especially C5a and C567 complex, attract neutrophils to sites of com-
146
-------
Figure 1. Blood vessels in the muscularis of the gallbladder of a dog injected
intravenously with ellagic acid. The walls and veins of an artery are
necrotic (arrowheads) and infiltrated by polymorphonuclear neutrophils
(hematoxylin and eosin; X500). (From Becker et al., 1980).
Figure 2. Section of lung treated in a similar fashion to that illustrated in Fig 1
Increased numbers of polymorphonuclear neutrophils (arrowheads) are
present in alveolar capillaries that are focally injured, allowing extravasa-
tion of erythrocytes (arrows) (hematoxylin and eosin, XI,250). (From Becker
et al., 1 980.) Similar changes are often seen in lungs of rabbits following
inhalation of tobacco glycoprotein.
plement activation (Henson and Oades, 1975; Fehr and Jacob, 1977). Neutrophils
are then induced by C5a to produce toxic oxygen radicals which may be injurious to
vascular endothelium (Sacks et al., 1978). Furthermore, when the attracted neutro-
phils attempt to phagocytize deposited immune complexes, exocytosis occurs along
the stimulated portion of cell membrane, resulting in the release of enzymic constit-
uents, such as elastase and collagenase, destructive to vessel wall constituents. In ad-
dition, C3a and C5a, the anaphylatoxic components, of complement can cause dis-
charge of inflammatory mediators from basophils and mast cells also contributing
to intensification of the vascular lesion (Henson, 1972; Becker and Henson, 1973).
Activation of the alternative pathway of complement which can be achieved by a
wide variety of environmental and endogenous substances, including cobra venom,
components of microorganisms, and aggregated immunoglobulins, can presumably
trigger the same events without the need for antibody. Furthermore, complement
components accelerate thrombin-induced aggregation of platelets (Polley and
Nachman, 1978). This induced aggregation may play an important role in normal
hemostasis, but it might also potentiate platelet accumulation at sites of vessel in-
jury. The role of complement may not be limited to immunologically or viral-in-
147
-------
duced atherosclerosis. Severe atherosclerotic changes that characteristically develop
in the aortas of rabbits fed cholesterol are very much milder than in rabbits geneti-
cally deficient in the sixth component of complement, indicating that activation of
the complement cascade and subsequent events described above may play a role in
the pathogenesis of diet-induced as well as immunologically mediated atherosclero-
sis (Geertinger and Sorensen, 1975).
The attraction and adherence of polymorphonuclear neutrophils, monocytesand
platelets to sites of immune complex deposition either because of complement
components, Fc receptors on these cells, or participation of the clotting system
through either the extrinsic system via release of tissue factor or the intrinsic
pathway can also result in acute vascular injury. Indeed, the various pathways may
act synergistically. However, the attraction of platelets and monocytes to the site of
injury influences more than just the development of acute injury. Both platelets and
monocytes have been demonstrated to release mitogens capable of stimulating the
proliferation of vascular smooth muscle cells and fibroblasts, and in the case of
monocyte-derived mitogen, endothelial cells as well (Ross et al., 1974; Polverini et
al., 1977; Ross and Vogel, 1978; Martin et al., 1981). Since these mitogens lead to
proliferation of mesenchymal cells, they contribute to the healing of the injury but
also may stimulate the local proliferation of cells that results in the atherosclerotic
plaque.
Experiments were conducted over the last several years in our laboratory to test
the hypothesis that immunologically induced injury of blood vessels can act syner-
gistically with diet-induced hyperlipidemia to induce atherosclerosis. Arterial injury
was induced in rabbits by repeated intravenous injection of foreign serum approxi-
mately every four to eight weeks. Certain rabbits were fed a cholesterol-poor, lipid-
rich diet for as long as 17 months; this elevated serum cholesterol levels from the
normal range of 40 to 80 mg/dl to between 150 and 400 mg/dl, a range comparable to
that of adult humans in developed Western countries. Changes that developed in the
arteries of these rabbits which strikingly resemble those alterations in chronic
human atherosclerosis with respect to distribution and morphologic characteristics
in the arterial tree (Figs. 3,4,5 and 6). Earlier experiments in which rabbits were fed a
cholesterol-supplemented diet and subjected to repeated immunologic challenge
over a shorter period of time resulted in less chronic lesions (Minick et al., 1966).
In other experiments, sites of injury induced immunologically as above were al-
lowed to heal for 40 to 80 days prior to induction of hypercholesterolemia. It was ob-
served that lipid accumulation occurred preferentially at sites of intimal thickening,
that is, the sites of previous injury (Figs. 7 and 8). These findings indicate that sites of
intimal thickening may have an increased avidity for lipid long after the initial
injury, permitting them to evolve into atherosclerotic plaques. In humans sites of
intimal thickening resulting from arterial injury induced many months or even years
previously may preferentially accumulate lipid and evolve as atherosclerosis later in
life when plasma lipids increase.
In recent experiments, vasectomized monkeys with circulating immune com-
plexes consisting of sperm-derived antigens and antibodies also were observed to
develop accentuated atherosclerosis in the presence of hyperlipidemia as compared
to sham-operated controls (Alexander and Clarkson, 1978). Subsequently, it was
shown that vasectomy resulted in accentuated atherosclerosis even in monkeys fed a
monkey-chow diet low in lipid (Clarkson and Alexander, 1980). It has been observed
that thrombotic and obliterative changes occur in medium and small myocardial
arteries leading to myocardial necrosis in the hearts of mice with lupus-like syn-
dromes (Accinni and Dixon, 1979). In this connection, restriction of caloric intake in
these mice has been shown to increase their life span apparently by reducing the
number of circulating immune complexes (Fernandes et al., 1978). Campbell et al.
(1978) also found that, in mice infected with Coxsackie virus, persistence of viral
antigen in aortas and aortic injury occurred only in hyperlipemic mice.
148
-------
Figure 3. Atherosclerosis of the righl nary artery of a 35-year-old man with
marked coronary atheroscle , old and recent myocardial infarcts,
and aortic and mitral steno: 3tal cholesterol in serum one month
before death was 279 mg per dl Lumen of artery is narrowed by
musculoelastic intimal thickening characterized by fatty-proliferative
changes with pooling of lipid and very little elastic tissue (intima identi-
fied by bar). Note similar changes in the rabbit atherosclerosis shown
in Fig 4 (Weigert's hematoxylin and eosin, X25). (From Minick and
Murphy, 1973).
Figure 4. Atherosclerosis of the right coronary artery of rabbit that received
semisynthetic lipid-rich diet and seven injections of horse serum over a
period of nine months. Average total cholesterol in serum, 157 mg per
dl. Atheromatous intimal thickening with straightening, fragmentation,
and reduplication of internal elastic membrane (intima identified by
bar). Fatty-proliferative changes with very little elastic tissue overlie a
layer of musculoelastic intimal thickening Note marked similarity to the
human atherosclerosis shown in Fig. 3. In the right lower quarter of
the picture, the atheromatous intima overlies a thinned media, as
commonly occurs in human atherosclerosis The atheromatous change
seen here did not occur in coronary arteries of control rabbits that
received semisynthetic lipid-rich diets without injections of foreign
serum protein (Weigert's hematoxylin and eosin, X75). (From Minick
and Murphy, 1973)
Clinico-pathologic observations and experimental results indicate that immuno-
logically mediated vascular injury leading to atherosclerosis can also be induced in
transplanted tissue perhaps by cellular mechanisms. This alteration has been ob-
served clinically in the unexpectedly severe and rapidly developing atherosclerosis
found to occur in some human cardiac homografts (Thompson, 1969; Rider et al.,
1975). Coronary atherosclerosis in association with graft rejection has been investi-
I49
-------
Figure 5. Atherosclerosis of the left coronary artery of a 72-year-old man who
died with disseminated rheumatoid disease. Serum cholesterol
concentration not known. Fatty-hyaline intimal thickening with fibro-
muscular cap and deep intimal deposits of calcium is associated with
clusters of foam cells. Note the striking similarity to the rabbit athero-
sclerosis shown in Fig. 6 (hematoxylin and eosin, X105). (From Minick
and Murphy, 1973).
Figure 6. Atherosclerosis of mesenteric artery of rabbit that received semisyn-
thetic, lipid-rich diet and six intravenous injections of bovine serum
albumin over a period of one year. Average total cholesterol in serum,
280 mg per dl. Fatty-hyaline intimal thickening with fibromuscular cap
(C) and deep intimal deposits of calcium (arrowheads) is associated
with foam cells (arrows). Note the striking similarity to the human
atherosclerosis shown in Fig. 5 (hematoxylin and eosin, X165). (From
Minick and Murphy, 1973).
gated in experimental animals. In one study, heterotopic cardiac homotransplants
were performed in the necks of 60 rabbits. Addition of cholesterol to the diet accel-
erated and intensified the development of atherosclerosis in coronary and myocar-
dial arteries and the segment of aorta attached to the transplanted heart. Similar
atherosclerotic lesions were found in the arteries in hearts transplanted to rabbits
which received no lipid supplement, and no lesions were seen in the vessels of graft
recipients (Figs. 9 and 10). Thus, both humorally and cellularly mediated immune
150
-------
Figure 7. Lumen of rabbit mesenteric artery is narrowed by musculoelastic
intimal thickening (bar) resembling diffuse intimal thickening in man.
Arterial change characteristic of those seen in rabbits fed diets low in
lipid, injected with foreign serum protein, and sacrificed 40 to 80 days
after the last injection (Weigert's hematoxylin and eosin, X95). (From
Hardin et al., 1973).
Figure 8. Atherosclerosis of rabbit artery representative of that induced in rabbits
repeatedly injected with foreign serum protein and subsequently fed a
cholesterol-supplemented diet 40 to 80 days after last injection.
Morphologic and statistical evidence indicates that areas of fibromus-
cular intimal thickening (bar) like those illustrated in Fig. 7 preferen-
tially accumulated lipid and evolved as atherosclerosis like that
illustrated in this artery (hematoxylin and eosin, X60). (From Hardin et
al., 1973).
reactions may lead to atherosclerosis. In both instances endothelial injury appears to
be an important early change (Figs. 15 and 16).
Cellular immune mechanisms may be involved in the development of atheroscler-
osis associated with viral infection. Infection of genetically appropriate strains of
specific pathogen-free chickens with Marek's disease herpes virus will lead to occlu-
sive atherosclerosis even in animals with normal levels of serum cholesterol (Fabricant
et al., 1978; Minick et al., 1979). Although these lesions evolve in a short period of
time, they bear a striking resemblance to chronic atherosclerosis in man (Figs.
11,12,13 and 14). Results of these experiments may have implications with respect
to human atherosclerosis since man is persistently infected with up to five herpes
viruses.
151
-------
Figure 9. Cardiac allograft and recipient's heart and aorta from rabbit fed a
cholesterol-supplemented diet and treated with immunosuppressives.
Rabbit was sacrificed 92 days after transplantation. Note marked
atherosclerosis of transplanted aortic valve (arrows) and ascending
aorta (*). Atherosclerotic change ends abruptly at line of anastomosis
with recipient's common carotid artery (C). (From Alonso et al., 1977).
Figure 10. Proximal left main coronary artery of cardiac allograft that functioned
for approximately two weeks. The lumen is markedly narrowed by
atheromatous intimal thickening (bar). Cellular cap covers lipid
deposits deep in intima. Recipient rabbit was fed cholesterol-supple-
mented diet and received no immunosuppressive agents (Weigert's
hematoxylin and eosin, X35). (From Alonso et al., 1977).
INSET Serial section of artery illustrated in Fig. 10. Atheromatous gruel
containing cholesterol clefts (arrowheads) is present deep in intima
(Weigert's hematoxylin and eosin, X38). (From Alonso et al., 1977).
Since it has been demonstrated that human endothelial cells display histocompat-
ibility antigens and are susceptible to infection with a number of viruses, it is con-
ceivable that these cells can become the target of cell-mediated cytotoxic reactions in
much the same way that infected lymphocytes became target cells for uninfected
lymphocytes bearing the same histocompatibility determinants (Zinkernagel and
Doherty, 1979). Finally, it is known that monocytes release tissue factor in delayed
hypersensitivity reactions resulting in activation of the extrinsic coagulation path-
way possibly worsening vascular injury (Edwards et al., 1979). As noted above in
some instances release of tissue factor by macrophages may be regulated by various
lipoproteins (Schwartz et al., 1981).
152
-------
Figure 11. Atheromatous change in gastric artery of normocholesterolemic
chicken infected with Marek's disease herpesvirus and fed a diet low
in cholesterol. Lumen (*) is occluded by thickened intima. Athero-
matous change is present deep in intima and media (arrowheads)
(Weigert-van Gieson, X40) (From Fabricant et al., 1978).
Figure 12. Higher magnification of artery shown in Fig. 11. Note foam cells,
extracellular lipid, and cholesterol clefts (arrowheads) in media
(Weigert-van Gieson, X190). (From Fabricant et al., 1978).
It must be pointed out that once an arteriosclerotic plaque forms the endothelial
surface may no longer be normal, characteristics of flow are certainly altered, and
evidence has been presented indicating that secretion by and consumption of blood
platelets is enhanced. Thus some of the same pathways that are involved in the ini-
tiation of atherosclerotic plaques may be involved in its continued growth. On the
other hand, additional pathogenic mechanisms may play a part in lesion progression.
Recently, a new and very different theory about the initiation and growth of arte-
riosclerotic plaques has been put forth by Benditt (Benditt and Benditt, I973;
Benditt, 1976; Pearson et al., 1977; Benditt and Gown, 1980). This theory, now
called the monoclonal theory of plaque formation, is derived from a highly innova-
tive set of experiments in which small tissue samples of individual artenosclerotic
plaques from the aortas of women who were heterozygous for isoenzymes of glucose-
6-phosphate dehydrogenase (G6PD), a sex-linked gene, were analyzed clectro-
phoretically for their content of specific G6PD variant. The results indicated that
plaques contained one or another variant (A or B), but seldom both. Analysis of
uterine leiomyoma, a benign proliferation of smooth muscle cells, in the same exper-
imental design have yielded similar data (Under and Gartler, 1965). The following
interpretation of these data has been constructed: (1) uterine leiomyomata are con-
sidered benign neoplasms and are apparently monotypic or monoclonal; (2) arterio-
sclerotic plaques appear to be monotypic or monoclonal; (3) therefore, arteriosclerot-
153
-------
(13)
Figure 13. Cross-section of heart of a chicken infected with Marek's disease
herpesvirus and fed cholesterol-supplemented diet for 15 weeks. The
lumens of major coronary arteries are occluded by atherosclerotic
change as seen in the two arteries in the upper portion of this photo-
graph (arrowheads) (X2). (From Minick et al., 1979).
Figure 14. Major coronary artery of chicken infected with Marek's disease
herpesvirus and fed a cholesterol-supplemented diet for 15 weeks.
Cholesterol clefts (arrows) are present in atheromatous intima (bar)
adjacent to the media. The subjacent media is markedly thinned.
There is no cellular reaction in adventitia and little adventitial thick-
ening (Weigert-van Gieson, X38). (From Minick et al., 1979).
ic plaques are also benign neoplasms; and (4) if so, then the arteriosclerotic plaque
has arisen as a consequence of neoplastic transformation of smooth muscle cells.
This construction has very important implications for our understanding of the
pathogenesis of atherosclerosis especially in relation to the nature of environmental
factors which may contribute to it. This theory raises the possibility that a large
number of environmental mutagens, such as those in cigarette smoke or tryptophan
derivatives in cooked meat and an unidentified host of promoting substances, may
contribute to the initiation and continued growth of arteriosclerotic plaques. It is
also implied that the mechanisms described above, immunologic, inflammatory,
and thrombogenic factors, may have something to do with intimal thickening in
arteries, but not with the formation of the arteriosclerotic plaque. However the data
leading to this conclusion are reproducible in some but not all laboratories (Pearson
et al., 1977; Thomas et al., 1979). The interpretation can be challenged.
The question is whether evidence of monoclonality or monotypia necessarily im-
plies mutation or transformation. In other systems, the answer is no. Immunization
of rabbits with streptococcal teichoic acid stimulates the production of monoclonal
antibody to teichoic acid in a significant number of recipients (Osterland et al., 1966).
The cells producing the monoclonal antibody do not constitute clones of neoplastic
154
-------
(15)
(16)
Figure 15. Photomicrograph of endothelial surface of rabbit that received one
injection of horse serum and was sacrificed 15 days later Endothe-
lium is partially denuded and red cells and platelets adhere to the
surface (arrowheads). Note rolled edge of endothelial cell where it is
separating from the aortic surface (arrows) (XI,290). (From Minick et
al., 1978)
Figure 16. Lummal surface of large coronary artery of transplanted rabbit heart
that functioned for 18 days. There is complete loss of endothelium
with platelets adherent to subendothehal basement membrane deep
within the media (arrowheads) (X7.600) (From Alonso et al., 1977)
plasma cells. They are cells which differ from neighboring lymphocytes and plasma
cells only in terms of their capacity to multiply and secrete a unique product in re-
sponse to a certain stimulus. The responders are selected in response to the stimulus.
The question of selection as applied to the monotypic character ol certain arterio-
sclerotic plaques has been raised by Thomas and colleagues (I979).
Further, suggesting that mutational events account for the monotypic character
of certain plaques appears to ignore the evidence indicating that areas ol diffuse in-
timal thickening evolve into arteriosclerotic plaques. Data of Thomas et al. (1979)
indicate that such areas are ditypic. Evidence cited earlier indicates that such diffuse
155
-------
intimal thickening induced experimentally by immune injury can, in the presence of
diet-induced hyperlipidemia, evolve into atherosclerotic plaques closely resembling
those in humans (Hardin et al., 1973). On the other hand, areas of intense cellular
proliferation may be areas in which mutational events become more likely, leading to
the emergence of cancer cells in ulcerative colitis and from a pool of atypical, but not
malignant cells in the lungs of cigarette smokers.
Assuming, however, that some or perhaps even many plaques are monotypic, is it
necessary to invoke neoplastic transformation as an explanation? Scanning electron
microscopic studies of the aorta in rabbits made serum sick by injection of foreign
protein indicate that in some instances sites of injury of the endothelium are very
small, involving only a few cells per site (Minick, 1976; Sharma and Geer, 1977).
Such injury also involves focal damage to the underlying intima and media. If we can
assume that at the site of injury some smooth muscle cells may be killed outright and
that another small number are capable of responding to mitogenic stimuli, then it is
possible for the population of smooth muscle cells at this site to be skewed such that
the evolving plaque is monotypic. This explanation does not require selective ad-
vantage of one monotypic cell over the other.
Finally, it must be pointed out that neither the monoclonal hypothesis interpreted
as representing neoplastic transformation nor the injury hypothesis is exclusive.
Perhaps, from the standpoint of examining the effect of environmental factors on
the pathogenesis of atherosclerosis it is best at the present time to consider both
hypotheses.
PATHOGENESIS OF ATHEROSCLEROSIS AND RISK FACTORS
The next question to be considered is how theories of the pathogenesis of athero-
sclerosis relate to known risk factors for death from IHD.
Hyperlipemia
Data from epidemiologic studies, chemical analyses of diseased tissue and animal
experiments substantiate the importance of plasma lipids, particularly cholesterol in
the pathogenesis of the atherosclerotic lesion. Despite the importance of hyper-
lipemia, the mechanism by which it leads to atherosclerosis is not known. Serum
lipids and lipoproteins may be the primary factor. It is possible that, in the presence
of increased plasma concentrations, increased filtration of lipoprotein into the
arterial wall may lead to proliferation of smooth muscle cells and enhanced con-
nective tissue synthesis resulting in formation of a fibrous plaque. Low-density lipo-
protein from hyperlipemic sera has been shown to be mitogenic for smooth muscle
cells (Fischer-Dzoga et al., 1974). However, this hypothesis does not explain the
distribution of atherosclerosis within the arterial tree and its segmental localization
in a particular artery.
Experiments completed over the past several years have delineated an elaborate
physiologic system important to cholesterol metabolism of the entire organism as it
relates to the cholesterol economy of the cell. These findings have provided interest-
ing insights into the control of concentrations of low-density lipoproteins particu-
larly with regard to the possible interaction between genetic and dietary factors.
However, they have yet to furnish more specific insights into lipid transport and ac-
cumulation in the arterial wall other than to function as a scavenger in response to
increased concentration of low-density lipoproteins (Brown etal., 1981). Recent ex-
periments of Davies (1981) have provided interesting new insights into the influence
of growth on endocytosis and of platelet-derived growth factor on high-affinity re-
ceptors for low-density lipoprotein.
On the other hand, it is possible that many substances in the environment lead to
injury to the arterial wall, selectively damaging certain portions of the arterial bed
more than others. Depending on the environment and the injurious stimulus, these
156
-------
sites of injury in the arterial wall will evolve either as fibrocellular atherosclerotic
lesions or as areas of fibromuscular intimal thickening. In the presence of increased
serum lipids, fibromuscular intimal lesions will accumulate various quantities of cel-
lular and extracellular lipid. In especially severe and sustained injury, appreciable
lipid may accumulate even in the presence of normal serum lipid levels. Alternatively,
sites of injury evolving in the presence of normal serum lipid levels may evolve as
areas of fibromuscular intimal thickening resembling diffuse intimal thickening in
human arteries. Such areas can be shown to contain increased lipid as compared to
uninvolved arteries (Falcone et al., 1980).
It is widely accepted that injury favors lipid accumulation within the arterial wall
by altering the barrier function of endothelium, thus allowing increased quantities of
cholesteryl ester-rich low-density lipoprotein to gain access to the cells. Although
appealing in its simplicity, this theory does not explain the preferential lipid accumu-
lation in sites of fibromuscular intimal thickening that have evolved from previous
injury (Hardin et al., 1973). Our experiments indicate that lipid accumulates prefer-
entially in areas of fibrocellular intimal thickening covered by endothelium and not
in such areas lacking endothelium, as would have been anticipated if lipid accumula-
tion were simply a function of the presence or absence of the endothelial barrier
(Minicketal., 1979; Falcone etal., 1980). We also demonstrated cholesterol accumu-
lation in an injured re-endothelialized aorta was correlated with serum cholesterol
concentration; in contrast, there was no correlation over a range of 900 mg per dl in
areas devoid of endothelium (Falcone et al., 1980).
Our experiments indicate that there is greater accumulation of radiolabeled lipo-
protein in re-endothelialized areas of an injured aorta as compared to adjacent de-
endothelialized areas (Falcone et al., 1981). This may result from the trapping of
lipoprotein there since the re-endothelialized neointima also accumulates increased
quantities of glycosaminoglycan (Minick et al., 1977; Wight et al., 1979), and cer-
tain glycosaminoglycans bind to low-density lipoproteins (Iverius, 1972). Finally,
results of other experiments indicate that injury leads to an alteration of a number of
cellular enzymes responsible for the synthesis and degradation of cholesteryl esters,
and that the endothelium modulates those activities (Hajjar et al., 1981). In sum-
mary, it appears that in this model system injury leads to alterations in pro'teoglycan
metabolism that favor accumulation of cholesterol and cholesteryl ester-rich lipo-
proteins in the previously injured re-endothelialized aorta. At the same time, other
metabolic perturbations in the enzymes responsible for cholesteryl ester degradation
further enhance its accumulation in this area.
Lipids and lipoproteins by themselves may be an important source of vascular in-
jury and may damage the endothelium. Both low-density lipoproteins and hypertri-
glyceridemic very low-density lipoproteins have been shown to damage endothelial
cells in vitro (Hendrikson et al., 1979; Gianturco et al., 1980). Experiments of Ross
and Harker (1976) indicate that modest diet-induced hypercholesterolemia in non-
human primates will lead to endothelial injury and denudation. Others have not
been able to demonstrate this (Goode et al., 1977; Taylor et al., 1978). However, it
should be emphasized that the denudation is an extreme form of endothelial injury.
Small areas of the endothelial desquamation rapidly re-endothelialize primarily due
to migration of adjacent endothelial cells (Reidy and Schwartz, 1981). The possibil-
ity exists that hyperlipemia may induce more subtle forms of endothelial injury that
lead to endothelial desquamation without denudation.
The response of the arterial wall to injury may result in increased exposure to var-
ious mitogens, which may be important in uptake of lipoproteins and lipid accumu-
lation. Experiments of Daviesand colleagues (Davies and Ross, 1978; 1980; Davies
et al., 1980; Davies, 1981) and Vlodavsky et al. (1978) indicate that in some in-
stances cell growth stimulated by platelet-derived growth factor and other mitogens
may increase bulk phase endocytosis in smooth muscle cells and endothelial cells. In
addition, platelet-derived growth factor stimulated high affinity binding, uptake and
157
-------
degradation of low-density lipoprotein by smooth muscle cells and binding by endo-
thelial cells that was independent of its effect on cell growth.
Cigarette Smoking
Cigarette smoking is one of the best defined risk factors for atherosclerosis and
heart attack. It not only is a major risk factor itself, but it acts synergistically with
others such as hyperlipemia. For reasons that are unclear, the younger smoker is
very much more at risk of heart attack than the older smoker. Recent experimental
data suggest that nicotine can injure endothelial cells of umbilical arteries ex vivo, as
assessed by decreased production of prostacyclin (Stoel et al., 1980), and also may
lead to endothelial injury in vivo (Booyse et al., 1981). Carbon monoxide has also
been shown to potentiate the deposition of lipid in arterial walls, perhaps by damage
to vascular endothelium (Thomsen, 1974). However, the experimental data are
equivocal. Moreover, the animal experiments that were designed to measure the ef-
fect of carbon monoxide on vessels were of relatively short duration compared to the
number of pack-years of the average smoker.
Recently, a highly allergenic glycoprotein has been isolated from cured tobacco
leaves and from cigarette smoke to which approximately one-third of individuals,
smokers and nonsmokers, exhibit immediate cutaneous hypersensitivity. Repetitive
immunologic challenge with tobacco antigen might contribute to arterial injury and
the development of atherosclerosis (Becker et al., 1976; Becker and Dubin, 1977).
The same antigen also contains polyphenol haptens which enable it to activate the
coagulation factor XII or Hageman factor-dependent pathways of coagulation,
fibrinolysis, and kinin generation. Intravenous injection of these polyphenols can
also result in vascular injury and thrombosis (Figs. 1 and 2). Both allergically medi-
ated reactions and factor XH-dependent reactions might act synergistically to
induce vascular damage.
It is also important to point out that the heart in the hypersensitive host is a target
organ for anaphylactic reactions and responds to antigenic challenge with a fall in
contractility and coronary artery perfusion and the development of a variety of
arrhythymias,. including ventricular fibrillation (Capurro and Levi, 1975). These
changes have been induced experimentally with tobacco antigen raising the possibil-
ity that Sudden Cardiac Death in some smokers may be mediated by allergic mech-
anisms (Levi et al., 1982). These data may explain the apparently greater risk of cig-
arette smoking for the younger smoker, that is, this group includes those hypersen-
sitive to constituents of smoke. If they die of heart attacks, they are no longer repre-
sented in the population and smoking decreases as a risk factor.
Hypertension
With regard to hypertension, mechanisms of increased risk are also unclear.
Experimental evidence from hypertensive animals indicates that the walls of arteries
have increased permeability to several plasma proteins (Duncan et al., 1965; Weiner
et al., 1969; Bretherton et al., 1976), which may be important in explaining both the
intimal thickening and lipid accumulation that are accentuated by hypertension.
The mechanism of the increased permeability is not clear and experimental data are
conflicting. It has been attributed to increased quantities of angiotensin II (Giese,
1964; Robertson and Khairallah, 1972; Giacomelli et al., 1976). Some observers
have noted that the incidence of cardiovascular complications is higher in hyperten-
sive patients with high plasma renin concentrations than in those with low plasma
renin concentrations (Brunner et al., 1972; 1973), but the statistical evidence for this
has been challenged (Doyle et al., 1973; Mroczek et al., 1973; Genestetal, 1974;
Kaplan, 1975). Further, Gabbiani et al. (1979) were unable to confirm a relationship
between high circulating renin levels and endothelial permeability. Overturf et al.
(1981) recently observed that the amount of atherosclerosis induced by low renin
158
-------
and high renin hypertension was not significantly different in rabbits fed a cholester-
ol supplemented diet.
Altered hemodynamic factors have also been considered to be of importance
(Stehbens, 1975). It has been suggested that hypertension leads to endothelial dam-
age and increased cell turnover (Schwartz and Benditt, 1977). On the other hand, in-
creased replication could represent an attempt to adapt to dilatation and increased
vessel size resulting from hypertension (Wolinsky, 1972). In this regard, Schwartz
and Standaert (1982) found that endothelial replication increased following acute
increases in blood pressure but was not sustained as the blood pressure elevation be-
came chronic. Endothelial turnover may be used as an indirect measure of injury but
this finding suggests that it is associated with acute change in blood pressure and not
with chronic hypertension. Finally, activities of several lysosomal degradative
enzymes within the aortic wall have been shown to change following the onset of ex-
perimental hypertension in rats (Wolinsky, Capron et al., 1978). It may seem diffi-
cult to reconcile an increase in a lysosomal degradative enzyme with the increased
cholesteryl ester accumulation seen in atherosclerosis due to hypertension.
However, it should be emphasized that the increase in activity may not be sufficient
to degrade the increased quantities of lipoprotein entering the wall. As a result, cho-
lesteryl esters may accumulate in association with hypertension. We have observed
similar phenomena with regard to arterial injury, lipid accumulation, and choles-
teryl ester hydrolase activity (Hajjar et al., 1981).
Diabetes Mellitus
Even though studies in humans may clearly identify diabetes mellitus as one of the
risk factors associated with atherosclerosis (Garcia et al., 1974; Ostrander et al.,
1965) the explanation(s) for the increased risk is not clearly defined. Several lipopro-
tein abnormalities have been identified in diabetes (Stout, Bierman and Brunzell,
1975), but they generally are considered to be insufficient to account for the
magnitude of increased risk.
The diabetic patient also has a well-defined abnormality of the microvasculature,
characterized by thickening of the basement membranes(Maueretal., 1976). More-
over, several of the glycoproteins in basement membranes may be excessively glyco-
sylated (Spiroand Spiro, 1971; Biesswengerand Spiro, 1973). Abnormalities associ-
ated with changes in the microvasculature include diabetic retinopathy, glomerulo-
sclerosis and diabetic cardiomyopathy. Based on observations of decreased in vitro
life spans of diabetic as compared to nondiabetic fibroblasts, Vracko and Benditt
(1974) have suggested that the basis for the basement membrane thickening is an in-
creased susceptibility of the endothelium of diabetics to injury, and, with increased
endothelial turnover in vessels, each new population of endothelial cells contributes
its layer of basement membrane to the overall thickness of the blood vessel. It is not
known whether similar endothelial damage and abnormalities of glycoprotein and
basement membrane synthesis or degradation may contribute to the accentuated
atherosclerosis seen in larger blood vessels.
Results of recent experiments of Wolinsky and his colleagues indicate that there
are abnormalities of the activity of catabolic enzymes in the arterial wall of animals
with experimentally induced diabetes; they were able to show that there is a decrease
of several lysosomal enzymes, including acid cholesteryl ester hydrolase, in both
streptozotocin- and alloxan-induced diabetes in rats. Abnormalities of lysosomal
enzymes were restored to normal following administration of insulin. They specu-
late the decrease in these catabolic enzymes may account for some lipid accumu-
lation in the wall of the diabetic human artery, especially considering the possibility
of a need for capacity to degrade increased amounts of cholesteryl ester bearing lipo-
protein in the diabetic (Wolinsky, Capron et al., 1978; Wolinsky, Goldfisher et al.,
1978). Further experiments revealed that the increased activity of acid cholesteryl
ester hydrolase seen in arteries of hypertensive rats was decreased in the presence of
159
-------
experimental diabetes, suggesting that in the presence of hypertension and possible
increased filtration of lipoprotein the ability to degrade cholesteryl ester is decreased
(Wolinsky, Capron et al., 1978). However, these findings should be interpreted with
caution since, as the authors point out, experimental models of diabetes in rats with
severe insulin deficiency, glucagon excess, and absent ketosis are not a faithful
counterpart of the complicated metabolic abnormalities in human diabetes.
Stout, Bierman, and Ross (1975) have suggested that the relative excess of insulin
seen in many diabetics may make important contributions to increased risk. This
hypothesis is based on the finding that insulin enhances growth of vascular smooth
muscle cells in vitro and hence could be of importance in proliferation of smooth
muscle cells and intimal thickening in vivo. Finally, the association of diabetes with
autoimmune diseases and virus infections suggests intriguing possibilities with re-
gard to vessel injury in diabetes and the interaction of viral, immune and immuno-
genetic factors.
Sex, Age, and Physical Activity
Being a male also appears to predispose one to heart attacks. The mechanisms un-
derlying this association are unclear, and a large number of experimental and clinical
studies have not solved this puzzle (Lancet Editorial, 1977). In fact, administration
of estrogen potentiates heart attacks in males and birth control pills can apparently
precipitate thrombotic and ischemic heart disease events in females (Hennekens and
MacMahon, 1977). As noted earlier, smoking has been shown to interact signifi-
cantly with use of birth control pills (Jain, 1976). Recent findings demonstrating the
presence of androgen receptors in hearts and aortas of male baboons may provide an
interesting lead for further studies (Lin et al., 1980; McGill et al., 1979).
The likelihood of heart attack, aortic aneurysm and clinical symptoms of periph-
eral vascular disease all increase with age. The most obvious explanation for this is
that arteries become progressively narrowed to the point that ischemia occurs, or to
the point that a clot or thrombus can occlude the vessel. Changes in the patterns of
blood flow over plaques may perturb blood platelets, and also potentiate further
endothelial cell injury. Also, a wide variety of substances which are not demonstrably
thrombogenic when injected or absorbed into the blood stream become so in the pre-
sence of stasis which makes their weak capacity to activate clotting manifest. The re-
action of other body systems that may bear on atherosclerosis may change with ag-
ing, for example, the immune system, and many older people have increased levels of
circulating immune complexes.
On the other hand, every pathologist has seen examples of atherosclerotic but
ectatic coronary arteries in hearts of older patients. A similar phenomenon is partic-
ularly striking in the Masai in Africa. Masai have a low incidence of ischemic heart
disease although coronary artery intimal thickness is as great as that of men in the
United States (Mann et al., 1972). Bond and co-workers (Bond, Adams and Bullock,
1981; Bond, Adams, Kaduck et al., 1981) showed that the coronary arteries of cy-
nomologus monkeys (crab-eating macaques) fed lipid-rich diets developed arterio-
sclerotic plaques. However, in some of these macaques the lumen was larger, even
though the intimal surface area involved by disease was increased. This is explained
by the finding that the overall size of the artery is increased. Reasons for the ectatic
atherosclerotic arteries seen in man are obscure. They are possibly genetically deter-
mined since longevity is familial.
There is also evidence that physical inactivity may lead to increased cardiovascular
mortality, particularly these deaths associated with ischemic heart disease. Epidemi-
ological studies of London transport workers and British civil servants, San Fran-
cisco longshoremen and Harvard University alumni have all indicated that regular
exercise can promote health and cardiovascular fitness (Morris et al., 1966; 1973;
Paffenbargeretal., 1977; Chave et al., 1978; Paffenbarger et al., 1978). This effect is
160
-------
partly independent of other risk factors such as cigarette smoking, hypertension,
obesity, and genetic factors implicated in coronary heart disease. There are several
possible explanations. Exercise has been found to increase fibrinolytic activity in-
duced by experimental venous occlusion (Williams et al., 1980). The implication of
this finding seems obvious since thrombus formation appears to be important in the
pathophysiology of myocardial infarction, stroke and pulmonary embolism. Exer-
cise has also been found to lower plasma triglyceride, low- and very low-density lipo-
protein cholesterol while increasing the proportion of high-density lipoprotein cho-
lesterol. The mechanisms leading to these changes is poorly understood; however,
they appear to be related to the level of exertion and not entirely related to weight re-
duction (Huttenen et al., 1979; Wood and Haskell, 1979; Hartung et al., 1980).
Wolinsky and colleagues (1979) showed that exercised rats had remarkable
changes in the activity of catabolic enzymes in their aortas. Activities of four of six
hydrolases, including acid cholesteryl ester hydrolase, were significantly increased;
acid cholesteryl ester hydrolase was the most effected, showing an increase of 25 per-
cent to 30 percent above control levels. Such findings suggest that in humans regular
exercise may lead to an increased capacity of the arterial wall to catabolize chol-
esteryl ester. Finally, exercise has well-known salutory effects on cardiac function
(Astrand and Rodhal, 1970).
The question may now be raised as to how the pathogenetic mechanisms can be
used to understand the effect of the environment on development and progression of
atherosclerotic cardiovascular disease and its complications and to develop hypoth-
eses concerning the recent, unexplained decrease in deaths from arteriosclerotic car-
diovascular disease. We have discussed several diseases mediated by immune mech-
anisms that are associated with precocious atherosclerosis. It is reasonable to hy-
pothesize that other disease states in which similar mechanisms are operative might
permit the development of atherosclerosis at an accelerated rate. For example, viral
hepatitis, type B, can, in some humans, result in a condition in which circulating im-
mune complexes consisting of viral antigen and host antibodies deposit in the walls
of blood vessels inducing injury and inflammation and resulting in a disease called
periarteritis nodosa (Michalak, 1978). Not all cases of periarteritis are associated
with hepatitis indicating that response to other endogenous or exogenous antigens
can also result in this severe form of vascular injury. Also, only a small fraction of
people with viral hepatitis develop periarteritis. However, as many as 15 percent of
patients with hepatits develop arthralgias or joint pain early in their illness indicating
that transient circulation and deposition of immune complexes in tissues have oc-
curred. It may be that in years hence retrospective study of such patients will reveal
that they have a higher incidence of atherosclerosis and its complications. Viral in-
fection of vascular tissue may produce local injury and initiate the proliferative
change that can lead to the development of atherosclerosis. Further, viruses have
been shown to lead to diabetes in experimental animals and humans (Yoon et al.,
1979; Notkins, 1979). It is also conceivable that oncogenic viruses might contribute
to growth of atherosclerotic plaques. In this connection it is important to remember
that Marek's disease virus which can induce atherosclerosis in some strains of chick-
ens can induce malignant tumors of lymphoid tissue in others. Of course, it is not
known whether the induction of atherosclerosis or of lymphomas in the infected
chickens proceeds via similar mechanisms.
Immune injury may also be triggered by ingested or inhaled antigens. We have dis-
cussed allergens in cigarette smoke that can provoke anaphylactic responses and
activate the coagulation system. Can inhalation of similar products from other
plants in large quantity in the working place, for example, in cotton mills, also result
in vascular injury? Is there sufficient absorption of dietary antigens to produce vascu-
lar injury in the sensitized host? A recent report indicates that in a form of glomeru-
lonephritis, the surface immunoglobulin form IgA2is present selectively in immune
deposits. IgA2is present in far less quantity than IgA in plasma. One way of inter-
161
-------
preting this observation is that antigen-IgA2 complexes were formed in the lungs or
gut, reabsorbed, and then deposited in the renal vasculature (Andre et al, 1980). If
this interpretation is valid, then the antigens concerned probably were derived from
the environment, possibly the diet.
Chemical initiators or promoters of tumor growth may also contribute to the
growth of atherosclerotic plaques (Albert et al., 1977). This interesting suggestion,
based on the monoclonal hypothesis set forth by Benditt, has provided an important
new direction in atherosclerosis research.
Finally, since the formation of a thrombus within an arteriosclerotic vessel is what
initiates a heart attack or stroke in most instances, identification of environmental
substances which singularly or acting synergistically potentiate thrombus formation
is an important area that needs further study. The fact that such substances have
been described in cigarette smoke raises the possibility that other commonly in-
gested materials may also potentiate thrombogenesis. Genetic or acquired defects in
the regulation of clot formation might act synergistically with these substances to
potentiate clot formation. The association between taking oral contraceptives,
smoking, and the increased occurrence of heart attacks and strokes is an example of
such an interaction.
SHORT-TERM TESTS
There are, at present, no short-term tests by which agents in the external environ-
ment can be assayed for a contributing role in atherogenesis. The development of
such tests might take the direction of measuring the effect of various agents on vas-
cular cells in tissue culture. However, this approach would be rather simplistic given
the complicated nature of the disease in vivo, and generalizations from results in
such models might be misleading. Alternatively, the effect of environmental agents
on the induction of atherosclerosis can be measured in nonhuman primates and
other animal models. Such experiments are exceedingly costly and time consuming
and their applicability to human disease can always be questioned. Although there
are no short-term tests, there are a number of relatively new direct and indirect
methods for clinical evaluation of the extent and severity of atherosclerosis in vivo
such as intravenous angiography, cardiac imaging and Doppler methods in ultra-
sonography and scintigraphy. These new methods may be useful in evaluating the ef-
fect of various types of intervention on the progression or regression of human
arterial disease.
With regard to short-term tests, measurements of personal factors that can con-
tribute to the development of atherosclerosis and its complications are performed
regularly. These tests include the measurement of blood pressure, degree of obesity,
blood glucose and serum lipids (cholesterols, triglycerides, and lipoproteins). To
these tests might be added monitoring of cardiac rhythm. Depending on the results,
a variety of therapeutic or preventive approaches can be taken. Detection of
ventricular premature contractions and appropriate therapy can reduce the inci-
dence of Sudden Cardiac Death.
RECENT TRENDS IN CARDIOVASCULAR MORTALITY
Myocardial infarct (heart attacks) will result in over half a million fataltites in the
United States this year. However, despite the seriousness of the problem, there is
evidence indicating that there has been a general and continuing decrease in cardio-
vascular deaths in the United States which began in 1968. In real figures, 20 percent
or 200,000 fewer deaths occurred in 1977 than would have been predicted in 1968.
Unfortunately, understanding the nature of this decline may be as elusive as un-
raveling the nature of the disease process itself; a decrease in hospital mortality from
myocardial infarction and a decline in the known risk factors may have contributed
(Wolinsky, 1981). As a result of more effective acute care of myocardial infarction
162
-------
there has been a decline in hospital mortality of about one-third during this period of
time (to place this in proper perspective it is well to remember that over one-half of
deaths occur outside of the hospital). Increase in efficiency of identification of hyper-
tensive patients and effective therapy have increased dramatically since 1972.
Further, in the past decade, a nutritional change has occurred in the United States,
and there has been a reduction in the consumption of fluid milk and cream of 19.2
percent, butter 31.9 percent, animal fats and oils 56.7 percent, and eggs 12.6 percent.
Overall, the average intake of saturated fat has dropped about four percent and
cholesterol 10 percent. Consumption of polyunsaturated fats has increased about 55
percent. There has been a general decrease in serum cholesterol for middle-aged men
of approximately seven percent. There has also been a decrease in cigarette smoking
by adult males although the overall cigarette consumption has not decreased.
From known data regarding the overall consumption of each of these substances,
it is possible to make predictions of the effect of changes of this magnitude on cardio-
vascular mortality. A decrease of mean cholesterol levels of 5 mg per 100 ml would
be expected to result in a decline of 4.3 percent in cardiovascular deaths in six years.
A 2 mm drop in blood pressure that can be attributed to the reduction cholesterol
would be expected to triple the decline in cardiovascular mortality. The reduction in
cigarette smoking by middle-aged men (age 35 to 75) would be expected to decrease
further the predicted mortality by an overall figure of 18 percent as compared to the
actual decrease of 21 percent observed in the United States from 1968-1976 (Wolinsky,
1981). Obviously, the validity of the available data also has implications with regard
to the diet-heart hypothesis and proposed dietary modification in a large segment of
the American population, a topic of recent controversy. Although these arguments
for the involvement of known risk factors seem persuasive, others have interpreted
the data quite differently and are not convinced that alterations in these factors ex-
plain the improvement in mortality (Stallones, 1980). They suggest that the improve-
ment rate is related to changes in unrecognized risk factors.
Perhaps the recent unexplained decrease in deaths from atherosclerosis and its
complications is related to immunization against an increasingly greater number of
childhood diseases. Similarly, better socioeconomic conditions may have resulted in
greater resistance to infection and its potential sequela of immunologically mediated
vascular injury. In this connection rheumatic fever, a sequela of Group A strepto-
coccal infection and known to predispose to the precocious development of athero-
sclerosis, is now a relatively uncommon disease in this country. The reasons for this
are obscure, but they may well be related to improved nutrition and living condi-
tions, since the incidence of rheumatic fever had already started to fall precipitously
before the advent of antimicrobial drugs. The decline in cardiovascular mortality
has taken place in an environment in which many of the general environmental fac-
tors have not been purposefully modified while personal ones have probably been
modified considerably. Thus, personal factors appear to be more important as risks
in cardiovascular mortality in Western industrialized societies.
Finally, when considering the relationship of environmental factors to vascular
disease, it must be reiterated that known risk factors can account for only about one-
half of patients who suffer from heart attacks. Clearly other environmental risk fac-
tors must exist. An important effort in the study of cardiovascular disease must be
the identification of these factors and the characterization of mechanisms by which
they induce disease.
ACKNOWLEDGEMENTS
We gratefully acknowledge the assistance of Carol Ibsen in the preparation of this
manuscript.
163
-------
REFERENCES
Abraham, A.S., M. Sonnenblick, M. Eni, O. Shemsh, and A.P. Batt. Serum chromium in pa-
tients with recent and old myocardial infarction. Am. Heart J. 99:604-606, 1980a.
The effect of chromium on established atherosclerotic plaques in rabbits. Am.
J. din. Nutr. 33:2294-2298, 19806.
Accinni, L., and F.J. Dixon. Degenerative vascular disease and myocardial infarction in mice
with lupus-like syndrome. Am. J. Pathol. 96:477-492, 1979.
Albert, R.E., M. Vanderlaan, F.J. Burns, and M. Nishizumi. Effect of carcinogens on chicken
atherosclerosis. Cancer Res. 37:2232-2235, 1977.
Alexander, C.S. Cobalt and the heart. Ann. Intern. Med. 70:411-413, 1969.
Alexander, N.J., and T.B. Clarkson. Vasectomy increases the severity of diet-induced athero-
sclerosis in Macaco fascicularis. Science 20/:538-541, 1978.
Alonso, D.R., P.K. Starek, and C.R. Minick. Studies on the pathogenesis of atheroarterio-
sclerosis induced in rabbit cardiac allografts by the synergy of graft rejection and hyper-
cholesterolemia. Am. J. Pathol. 57:415-442, 1977.
Anderson, E.W., R.J. Andelman, J.M. Strauch, N.J. Fortuin, and J.H. Knelson. Effect of
low-level carbon monoxide exposure on onset and duration of angina pectoris. Ann.
Intern. Med. 79:46-50, 1973.
Anderson, T. W., and W.H. LeRiche. Sudden death from ischemic heart disease in Ontario and
its correlation with water hardness and other factors. J. Can. Med. Assoc. 705:155-160,
1971.
Anderson, T.W., L.C. Neri, G.B. Schreiber, F.D.F. Talbot, and A. Zdrojewski. Ischemic heart
disease, water hardness and myocardial magnesium. J. Can. Med. Assoc. 7/3:199-202,
1975.
Andre, C., F.C. Berthoux, F. Andre, J. Gillon, C. Genin, and J-C. Sabatier. Prevalence of
IgA2 deposits in IgAnephropathies: a clue to their pathogenesis. N. Engl. J. Med. 303:1343-
1346, 1980.
Aronow, W.S., C.N. Harris, M.W. Isbell, S.N. Rokaw, and B. Imparato. Effect of free-way
travel on angina pectoris. Ann. Intern. Med. 77:669-676, 1972.
Aronow, W.S., E.A. Stemmer, and M.W. Isbell. Effect of carbon monoxide exposure on inter-
mittent claudication. Circulation 49:415-417, 1974.
Astrand, P.O., and K. Rodhal. Textbook of Work Physiology. McGraw Hill, New York, 1970.
Astrup, P. Tobacco smoking and coronary disease. Ada Cardiol. Suppl. 29:105-117, 1974.
Astrup, P., and K. Kjeldsen. Carbon monoxide, smoking and atherosclerosis. Med. Clin.
North Am. 55:323-350, 1974.
Astrup, P., D. Trolle, H.M. Olsen, and K. Kjeldsen. Effect of moderate carbon-monoxide ex-
posure on fetal development. Lancet 1972, ii:1220-1222, 1972.
Axelson, O., E. Dahlgren, C.D. Jansson, and S.O. Rehnlund. Arsenic exposure and mortality.
A case-referent study from a Swedish copper smelter. Br. J. Ind. Med. 35:8-15, 1978.
Axelson, O., and M. Priscator. Renal damage after prolonged exposure to cadmium. Arch.
Environ. Health 12:360-373, 1966.
Ayres, S.M., S. Giannelli, Jr., and H. Mueller. Myocardial and systemic responses to carboxy-
hemoglobin. Ann. N.Y. Acad. Sci. 174:268-293, 1970.
Ayres, S.M., S. Giannelli, Jr., H. Mueller, and A. Criscitiello. Myocardial and systemic vascu-
lar responses to low concentrations of carboxyhemoglobin. Ann. Clin. Lab Sci. 3:440-447,
1973.
Ayres, S.M., H.S. Mueller, J.J. Gregory, S. Giannelli, Jr., and J.L. Penny. Systemic and myo-
cardial hemodynamic responses to relatively small concentrations of carboxyhernoglobin
(COHb). Arch. Environ. Health 18:699-709, 1969.
Ball, K., and R. Turner. Smoking and the heart. The basis for action. Lancet 1974, ii:822-826,
1974.
Becker, C.G., and T. Dubin. Activation of factor XII by tobacco glycoprotein. J. Exp. Med.
146:457-467, 1977.
Becker, C.G.,T. Dubin, and F. Glenn. Induction of cholecystitis by activation of factor XII./.
Exp. Med. /5/:81-90, 1980.
Becker, C.G..T. Dubin, and H.P. Wiedemann. Hypersensitivity to tobacco antigen. Proc. Nail.
Acad. Sci. U.S.A. 73:1712-1716, 1976.
Becker, C.G., and G.E. Murphy. Demonstration of contractile protein in endothelial cells of
the heart valves, endocardium, intima, arteriosclerotic plaques and Aschoff bodies of
rheumatic heart disease. Am. J. Pathol. 55:1-37, 1969.
164
-------
Becker, C.G., and R.L. Nachman. Contractile proteins of endothelial cells, platelets and
smooth muscle. Am. J. Pathol. 71:1-22, 1973.
Becker, E.L., and P.M. Henson. In vitro studies of immunologically induced secretion of
mediators from cells and related phenomena. Adv. Immunol. 77:93-193, 1973.
Beevers, D.G., B.C. Campbell, A. Goldberg, M.R. Moore, and V.M. Hawthorne. Blood-
cadmium in hypertensives and normotensives. Lancet 1976, ii: 1222-1224, 1976.
Beevers, D.G., E. Erskine, and M. Robertson. Blood-lead and hypertension. Lancet 1976, ii:l-
3, 1976.
Behr, G., and P. Burton. Heart muscle magnesium. Lancet 1973, ii:450, 1973.
Beisswenger, P.J., and R.J. Spiro. Studies on the human glomerular basement membrane:
composition, nature of the carbohydrate units and chemical changes in diabetes mellitus.
Diabetes 22:180-193, 1973.
Benditt, E.P. Implications of the monoclonal character of human atherosclerotic plaques.
Ann. N.Y. Acad. Sci. 275:96-100, 1976.
Benditt, E.P., and J.M. Benditt. Evidence for a monoclonal origin of human atherosclerotic
plaques. Proc. Natl. Acad. Sci. U.S.A. 70:1753-1756, 1973.
Benditt, E.P., and A.M. Gown. Atheroma: the artery wall and the environment. Int. Rev. Exp.
Pathol. 27:55-118, 1980.
Benveniste, J. Platelet-activating factor, a new mediator of anaphylaxis and immune complex
deposition from rabbit and human basophils. Nature (London) 249:581-582, 1974.
Bond, M.G., M.R. Adams, and B.C. Bullock. Complicating factors in evaluating coronary
artery atherosclerosis. Artery 9:21-29, 1981.
Bond, M.G., M.R. Adams, J.M. Kaduck, and B.C. Bullock. Effects of atherosclerosis on
coronary artery size: implications for myocardial infarction. Fed. Proc. Fed. Am. Soc.
Exp. Med. 40:173, 1981.
Booyse, P.M., G. Osikowicz, and A.J. Quarfoot. Effects of chronic oral consumption of nic-
otine on the rabbit aortic endothelium. Am. J. Pathol. 702:229-238, 1981.
Borhani, N.O. Exposure to trace elements and cardiovascular disease. Circulation <53:260A-
263A, 1981.
Brand, R.J., R.H. Rosenman, R.I. Sholtz, and M. Friedman. Mutivariate prediction of coro-
nary heart disease in the Western Collaborative Group Study compared with the findings of
the Framingham Study. Circulation 53:348-355, 1976.
Bretherton, K.M., A.J. Day, and S.L. Skinner. Effect of hypertension on the entry of I25I-
labelled low-density lipoprotein into the aortic intima in normal fed rabbits. Atheroscler-
osis 24:99-106, 1976.
Brieger, H., C.W. Semisch, J. Statsney, and D. A. Piatnek. Industrial antimony poisoning. Ind.
Med. Surg. 23:521-523, 1954.
Brown, M.S., P.T. Kovanen, and J.L. Goldstein. Regulation of plasma cholesterol by lipopro-
tein receptors. Science 272:628-635, 1981.
Brunner, H.R., J.H. Laragh, L. Baer, M.A. Newton, F.T. Goodwin, L.R. Krakoff, R.H. Bard,
and F.R. Buhler. Essential hypertension: renin and aldosterone, heart attack and stroke. TV.
Engl. J. Med. 286:441-449, 1972.
Brunner, H.R., J.E. Sealey, and J.H. Laragh. Renin as a risk factor in essential hypertension:
more evidence. Am. J. Med. 55:295-302, 1973.
Bywaters, E.G.L. Peripheral vascular obstruction in rheumatoid arthritis and its relationship
to other vascular lesions. Ann. Rheum. Dis. 76:84-103, 1957.
Campbell, A.E., R.M. Laria, and G.E. Madge. Coxsackie virus B cardiopathy and angiopathy
in the hypercholesterolemic host. Atherosclerosis 37:295-306, 1978.
Capurro, N., and R. Levi. The heart as a target organ in systemic allergic reactions: comparison
of cardiac anaphylaxis in vivo and in vitro. Circ. Res. 3(5:520-528, 1975.
Chave, S.P.W., J.N. Morris, S. Moss, and A.M. Semmence. Vigorous exercise in leisure-time
and the death rate: a study of male civil servants. J. Epidemiol. Commun. Health 32:239-
243, 1978.
Chipperfield, B., and J.R. Chipperfield. Differences in metal content of the heart muscle in
death from ischemic heart disease. Am. Heart J. 95:732-737, 1978.
Clarkson, T.B., and N.J. Alexander. Long-term vasectomy. Effect on the occurrence and
extent of atherosclerosis in Rhesus monkeys. J. Clin. Invest. (55:15-25, 1980.
Cochrane, C.G., and D. Hawkins. Studies on circulating immune complexes. III. Factors
governing the ability of circulating complexesto localize in blood vessels./. Exp. Med. 127:
137-154, 1968.
165
-------
Cohen, S.I., M. Deane, and J.R. Goldsmith. Carbon monoxide and survival from myocardial
infarction. Arch. Environ. Health 79:510-517, 1969.
Comstock, G. W. Water hardness and cardiovascular diseases. Am. J. Epidemiol. 7/0:375-400,
1979.
Connor, W.E., and S.L. Connor. The key of nutritional factors in the prevention of coronary
heart disease. Prev. Med. /:49-83, 1972.
Crawford, M.D., D.G. Clayton, F. Stanley, and A.G. Shaper. An epidemiological study of
sudden death in hard and soft water areas. /. Chronic Dis. 50:69-80, 1977.
Crawford, M.D., M.J. Gardner, and J.N. Morris. Mortality and hardness of local water-sup-
plies. Lancet 1968, i:827-831, 1968.
Davies, P.P. Growth regulation of endocytosis in cultured vascular cells. Thromb. Haemostas
46:221, 1981.
Davies, P.P., and R. Ross. Mediation of pinocytosis in cultured arterial smooth muscle and
endothelial cells by platelet-derived growth factor. J. Cell Biol. 79:663-671, 1978.
Growth-mediated, density dependent inhibition of endocytosis in cultured
arterial smooth muscle cells. Exp. Cell Res. 729:329-336, 1980.
Davies, P.P., C.J. Selden, and S.M. Schwartz. Enhanced rates of fluid pinocytosis during
exponential growth and monolayer regeneration by cultured arterial endothelial cells. J.
Cell Physiol. 702:119-127, 1980.
Dayton, S., M.L. Pearce, S. Hashimoto, W.J. Dixon, and U. Tomiyasu. A controlled clinical
trial of a diet high in unsaturated fat in preventing complications of atherosclerosis. Cir-
culation 40, Suppl. 11:1, 1969.
Department of Health, Education and Welfare. Arteriosclerosis: a report by the National
Heart and Lung Institute Task Force on Arteriosclerosis. DHEW Pub. No. (NIH) 72-219,
1971.
Department of Health, Education and Welfare. Proceedings of the International Occupational
Hand-Arm Vibration Conference. DHEW (NIOSH) Pub. No. 77-170. U.S. Government
Printing Office, Washington, D.C., 1977.
Dixon, F.J., J.J. Vasquez, W.O. Weigle, and C.G. Cochrane. Pathogenesis of serum sickness.
Arch. Pathol. 65:18-28, 1958.
Doyle, A.E., G. Jerums, C.I. Johnston, and W.J. Louis. Plasma renin levels and vascular
complications in hypertension. Br. Med. J. 2:206-207, 1973.
Duguid, J.B. Pathogenesis of atherosclerosis. Lancet 1949, ii:925, 1949.
Duncan, L.E., Jr., K. Buck, and A. Lynch. The effect of pressure and stretching on the passage
of labeled albumin into canine aortic wall. J. Atheroscler. Res. 5:69-79, 1965.
Edwards, R.L., F.R. Richles, and A.M. Bobrove. Mononuclear cell tissue factor: cell of origin
and requirements for activation. Blood 5*359-370, 1979.
Elwood, P.C., A.S. St. Leger,and M. Morton. Mortality and concentration of elements in tap
water in county, boroughs in England and Wales. Br.J. Prev. Soc. Med. 31:178-182, 1977.
Fabricant, C.G., J. Fabricant, M.M. Litrenta, andC.R. Minick. Virus-induced atherosclerosis.
J. Exp. Med. 148:335-340, 1978.
Fahr, T. Zurfragederpolymyosites(Dermatomyositis). Arch. Dermatol. Syphilol. 130:1,1921.
Falcone, D.J., D.P. Hajjar, and C.R. Minick. Enhancement of cholesterol and cholesteryl
ester accumulation in re-endothelialized aorta. Am. J. Pathol. 99:81-104, 1980.
. Endothelium enhances accumulation of I25I-APO-B lipoprotein in injured aorta.
Fed. Proc. Fed. Am. Soc. Exp. Med. 40:328, 1981.
Fehr, J., and H.S. Jacob. In vitro granulocyte adherence and in vivo margination: two associ-
ated complement-dependent functions. J. Exp. Med, 146:641-652, 1977.
Fernandes, G., P. Friend, E.J. Yunis, and R.A. Good. Influence of dietary restriction on
immunologic function and renal disease in (NZB x NZW) F. mice. Proc. Nail. Acad. Sci.
U.S.A 75:1500-1504, 1978.
Fischer-Dzoga, K., R. Chen, and R.W. Wissler. Effects of serum lipoproteins on the morphol-
ogy, growth and metabolism of arterial smooth muscle cells Adv Exp Med Biol 43-299-
311, 1974.
Flowers, N.C., R.C. Hand, and L.G. Horan. Concentrations of fluoralkanes associated with
cardiac conduction system toxicity. Arch. Environ. Health 30:353-360, 1975.
Fredrickson, D.S., and V.T. Ferrans. Acid cholesteryl ester hydrolase deficiency. In The
Metabolic Basis of Inherited Disease, edited by J.B. Stanburg, J.B. Wyngaarden, and D.S.
Frederickson, 4th ed., McGraw Hill, New York, pp. 670-687, 1978.
Friedman, M.,O.B. Sanford, and A. E. Brown. Plasma lipid responses of rats and rabbits to an
auditory stimulus. Am. J. Physiol. 2/2:1174-1178, 1967.
166
-------
Friedman, R.J., M.B. Stemerman, B. Wenz, S. Moore, J. Gauldie, M. Gent, M.L. Tiell, and
T.H. Spaet. The effect of thrombocytopenia on experimental arteriosclerotic lesion forma-
tion in rabbit. Smooth muscle cell proliferation and re-endothelialization. J. Clin. Invest.
60:1191-1201, 1977.
Gabbiani, G., G. Elemer, C.H. Guelpa, M.B. Balotton, M-C. Badonnel, and I. Huttner. Mor-
phologic and functional changes of the aortic intima during experimental hypertension.
Am. J. Pathol. 9(5:399-422, 1979.
Gampel, B., C. Slome, N. Scotch, and J.H. Abramson. Urbanization and hypertension among
Zulu adults. J. Chronic Dis. 15:61-10, 1962.
Garcia, M.J., P.M. McNamara, T. Gordon, and W.B. Kannel. Morbidity and mortality in
diabetics in the Framingham populations. Sixteen year follow-up study. Diabetes 23:105-
112, 1974.
Garcia-Palmieri, M.R. Precursors of coronary artery disease in Puerto Rico. Am. J. Clin.
Nutr. 26:1133-1137, 1973.
Geertinger, P., and H. Sorensen. On the reduced atherogenic effect of cholesterol feeding in
rabbits with cogenital complement (C6) deficiency. Artery 7:177-184, 1975.
Genest, J., R. Boucher, O. Kuchel, J.M. Rojo-Ortega, et al. The renin-angiotensin system:
some new aspects. Trans. Am. Clin. Climatol. Assoc. S<5:139, 1974.
Germuth, F.G., Jr., L.B. Senterfit, and G.R. Dreesman. Immune complex disease. V. The
nature of the circulating complexes associated with glomerular alterations in the chronic
BSA-rabbit system. Johns Hopkins Med. J. 130:344-351, 1972.
Giacomelli, F., P. Anversa, and J. Wiener. Effect of angiotension-induced hypertension on rat
coronary arteries and myocardium. Am. J. Pathol. 84:111-138, 1976.
Gianturco, S.H., S.G. Eskin, L.T. Navarro, C.J. Lahart, L.C. Smith, and A.M. Gotto. Ab-
normal effects of hypertriacylglycerolemic very low-density lipoproteins on 3-hydroxy-3-
methyl-glutaryl-CoA reductase activity and viability of cultured bovine aortic endothelial
cells. Biochem: Biophys. Acta 618:143-152, 1980.
Giese, J. Acute hypertensive vascular disease. 2. Studies on vascular reaction patterns and
permeability changes by means of vital microscopy and colloidal tracer technique. Acta
Pathol. Microbiol. Scand. 6"2:497-515, 1964.
Glauser, S.C., C.T. Bello, and E.M. Glauser. Blood-cadmium levels in normotensive and un-
treated hypertensive humans. Lancet 1976, i:717-718, 1976.
Goode, T.B., P.p. Davies, M.A. Reidy, and D.E. Bowyer. Aortic endothelial cell morphology
observed in situ by scanning electron microscopy during atherogenesis in the rabbit.
Atherosclerosis 27:235-251, 1977.
Gordon, T., W.P. Castelli, and M.E. Hjortland. High-density lipoprotein as a protective factor
against coronary heart disease. The Framingham Study. Am. J. Med. 62:101-114, 1911.
Gordon, T., M.R. Garcia-Palmieri, A. Kagan, W.E. Kannel, and J. Schiffman. Differences in
coronary heart disease in Framingham, Honolulu and Puerto Rico. /. Chronic Dis. 27:329-
344, 1974.
Hajjar, D.P., D.J. Falcone, S. Fowler, and C.R. Minick. Endothelium modifies the altered
metabolism of the injured aortic wall. Am. J. Pathol. 102:28-39, 1981.
Hardin, N.J., C.R. Minick, and G.E. Murphy. Experimental induction of atheroarteriosclerosis
by the synergy of allergic injury to arteries and lipid-rich diet. III. The role of earlier
acquired fibromuscular intimal thickening in the pathogenesis of later developing athero-
sclerosis. Am. J. Pathol. 73:301-326, 1973.
Barker, L.A., R. Ross, S.J. Slichter, and C.R. Scott. Homocystine-induced arteriosclerosis:
the role of endothelial cell injury and platelet response in its genesis. J. Clin. Invest. 58:131-
741, 1976.
Marker, L.A., S.J. Slichter, C.R. Scott, and R. Ross. Homocystinemia. Vascular injury and
arterial thrombosis. N. Engl. J. Med. 297:537-543, 1974.
Harlan, W.R. Physical and psychosocial stress and the cardiovascular system. Circulation 63:
266A-271A, 1981.
Harlan, W.R., A.R. Sharrett, H. Weill, G.M. Turino, N.O. Borhani, and L. Resnekov. Impact
of the environment on cardiovascular disease; report of the American Heart Association
Task Force on environment and the cardiovascular system. Introduction. Circulation 63:
243A-246A, 1981.
Hartung, G.H., J.P. Foregt, R.E. Mitchell, I. Vlasek, and A.M. Gotto. Relation of diet to high-
density-lipoprotein cholesterol in middle-aged marathon runners, joggers, and inactive
men. N. Engl. J. Med. 502:357-361, 1980.
167
-------
Haynes, S.G., M. Feinleib, S. Levine, N. Scotch, and W.B. Kannel. The relationship of psycho-
social factors to coronary heart disease in the Framingham Study. II. Prevalence of
coronary heart disease. Am. J. Epidemiol. 107:384-402, 1978.
Haynes, S.G., S. Levine, N. Scotch, M. Feinleib, and W.B. Kannel. The relationship of psycho-
social factors to coronary heart disease in the Framingham. Study. I. Methods and risk
factors. Am. J. Epidemiol. 707:362-383, 1978.
Hennekens, C.H., and B. MacMahon. Oral contraceptives and myocardial infarction. N. Engl.
J. Med. 29(5:1166-1167, 1977.
Henrikson, T., S.A. Evensen, and B. Carlander. Injury to human endothelial cells in culture
induced by low density lipoproteins. Scand. J. Clin. Lab. Invest. 39:361-368, 1979.
Henson, P.M. Complement-dependent adherence of cells to antigen and antibody. Mecha-
nisms and consequence. In Biological Activities of Complement, Karger, Basel, 1972.
Henson, P.M. Immune complex diseases. Cellular mediators and the pathogenesis of inflam-
matory tissue injury produced by immune complexes. In Bayer Symposium VI, Experi-
mental Models of Chronic Inflammatory Diseases, edited by L.E. Glynn, and H.O.
Schlumberger, Springer Verlag, Berlin-New York, pp. 94-106, 1977.
Henson, P.M., and Z.G. Oades. Stimulation of human neutrophils by soluble and insoluble
immunoglobulin aggregates. J. Clin. Invest. 56:1053-1062, 1975.
Hugod, C, L. Hawkins, K. Kjeldsen, H.K. Thomsen, and P. Astrup. Effect of carbon monoxide
exposure on aortic and coronary intimal morphology in the rabbit. A revaluation. Athero-
sclerosis 30:333-343, 1978.
Huttenen, J.K., E. Lansimies, E. Voultilainen, and C. Ehnholme. Effect of moderate physical
exercise on serum lipoproteins: a controlled clincal trial with special reference to serum
high-density lipoproteins. Circulation (50:1220-1229, 1979.
Iseri, L.T., J. Freed, and A.R. Bures. Magnesium deficiency and cardiac disorders. Am. J.
Med. /5«:837-846, 1975.
Iverius, P.H. The interaction between human plasma lipoproteins and connective tissue glyco-
saminoglycans. /. Biol. Chem. 247:2601-2613, 1972.
Jain, A.K. Cigarette smoking, use of oral contraceptives, and myocardial infarction. Am. J.
Obset. Gynecol. 126:301-301, 1976.
Jenkins, C. D. Recent evidence supporting psychologic social risk factors for coronary disease.
N. Engl. J. Med. 294:987-994, 1033-1038, 1976.
Joris, I., and G. Majno. Atherosclerosis and inflammation in the thrombotic process in athero-
genesis. Adv. Exp. Med. Biol. 104:221-233, 1978.
Kannel, W.B., W.P. Castelli, T. Gordon, and P.M. McNamara. Serum cholesterol, lipopro-
teins, and the risk of coronary heart disease. The Framingham Study. Ann. Intern. Med.
74:1-21, 1971.
Kannel, W.B., D. McGee, and T. Gordon. A general cardiovascular risk profile: The
Framingham Study. Am. J. Cardiol. 35:46-51, 1976.
Kaplan, N.M. The prognostic implications of plasma renin in essential hypertension. J. Am.
Med. Assoc. 231:161, 1975.
Keys, A. (ed.). Coronary Heart Disease in Seven Countries. Circulation 41-42 (Suppl.):l-2\ 1,
1970.
Kincaid-Smith, P. Modification of the vascular lesion of rejection in cadaveric renal allografts
by dipyridamole and anticoagulants. Lancet 1969, ii:920-922, 1969.
Kjeldsen, K., P. Astrup, and J. Wanstrup. Ultrastructural intimal changes in the rabbit aorta
after a moderate carbon monoxide exposure. Atherosclerosis 7(5:67-82, 1972.
Kjeldsen, K., H.K. Thomsen, and P. Astrup. Effects of CO on myocardium Ultrastructural
changes in rabbits after moderate chronic exposure. Circ. Res. 34:339-348, 1974.
Kjeldsen, K., J. Wanstrup, and P. Astrup. Enhancing influence of arterial hypoxia on the
development of atheromatosis in cholesterol-fed rabbits. J. Atherosclerosis Res. 5:835-854,
1968.
Klevay, L. M. Coronary heart disease. The zinc/copper hypothesis. Am. J. Clin. Nuir. 28:164-
774, 1975.
Klinge, F. Der Rheumatismus: palhologisch analomische und experimented pathologische.
Tataschen und ihre Auswertung fur des artzliche Rheumaproblem. Ergeb. Allg. Pathol.
Pathol. Anat. 27:1-235, 1933.
Kniker, W.T., and C.G. Cochrane. The localization of circulating immune complexes in exper-
imental serum sickness. J. Exp. Med. 727:119-136, 1968.
Kobayoshi, J. Geographical relationship between the chemical nature of river water and death
rate from apoplexy. Ber. Ohara Inst. Landwirtsch. Biol., Okayama Univ. 77:12, 1957.
168
-------
Kravis, T.C., and P.M. Henson. Accumulation of platelets at sites of antigen-antibody
mediated injury: a possible role for IgE antibody and mast cells. /. Immunol 118:1569-
1573, 1977.
Kuller, L.H., E.P. Radford, D. Swift, J.A. Perper, and R. Fisher. Carbon monoxide and heart
attacks. Arch. Environ. Health 50:477-482, 1975.
Lancet Editorial. Coronary artery disease in young women. Lancet 1977, ii:282-283, 1977.
Lehnert, G., A. Morgan, D. Szadowski, and R.L. Zielhuid. Halogenated hydrocarbon solvents:
long term effects and biological sampling in human beings. Int. Arch. Occup. Health
33:251-255, 1974.
Levi, R., J.H. Zavecz, J.A. Burke, and C.G. Becker. Cardiac and pulmonary anaphylaxis in-
duced in guinea pigs and rabbits by glycoprotein isolated from tobacco leaves and cigarette
smoke condensate. Am. J. Pathol. 70(5:318-325, 1982.
Lilienthal, J.J., Jr., and C.H. Fugitt. The effect of low concentrations of carboxy hemoglobin on
the "altitude" tolerance" of man. Am. J. Physiol. 745:359-364, 1945.
Lin, A.L., H.C. McGill, and A.S. Shain. Androgen receptors of the baboon aorta and myo-
cardium. Fed. Proc. Fed. Am. Soc. Exp. Biol. 40:333, 1980.
Linder, G., and S.M. Gartler. Glucose-6-phosphate dehydrogenase, mosaicism: utilization as a
cell marker in the study of leiomyoma. Science 150:61-6%, 1965.
Long, E.R. Development of our knowledge of arteriosclerosis. InCowdry's Arteriosclerosis. A
Survey of the Problem, edited by H.T. Blumenthal, Charles C. Thomas, Springfield, Illi-
nois, pp. 5-20, 1967.
Lurie, K.G., M.E. Billingham, S.W. Jamieson, D.C. Harrison, and B.A. Reitz. Pathogenesis
and prevention of graft arteriosclerosis in an experimental heart transplant model. Trans-
plantation 37:41-47, 1981.
Lyss, A.P., H. Friedman, R. Corkey, N. Kefalides, M. Bina, S.T. Widirstky, and D.B. Cines.
Viral induction of receptors for immunoglobulin and complement on human endothelial
cells. Circulation 62 (Suppl.) 111:272, 1980.
Maas, B. Platelet adhesiveness during exposure to noise. Ger. Med. 3:111, 1973.
Majno, G., S.M. Shea, and M. Levanthal. Endothelial contraction induced by histamine-type
mediators. An electron microscopic study. /. Cell Biol. 42:641-672, 1969.
Mann, G.V., A. Sporry, M. Gray, and D. Jarashow. Atherosclerosis in the Masai. Am. J.
Epidemiol. 95:26-37, 1972.
Martin, B.M., M.A. Gimbrone, Jr., E.R. Unanue, and R.S. Cohan. Stimulation of mono-
lymphoid mesenchymal cell proliferation by a macrophage-derived growth factor. /.
Immunol 72(5:1510-1515, 1981.
Mathews, J.D. Ischemic heart-disease: possible genetic markers. Lancet 1975, ii:681-682, 1975.
Mauer, S.M., J. Barbosa, R.L. Vernier, C.M. Kjellstrand, T.J. Buselmeier, R.L. Simmons,
J.S. Najarian, and F.C. Goetz, Development of diabetic vascular lesions in normal kidneys
transplanted into patients with diabetes mellitus. TV. Engl. J. Med. 295:916-920, 1976.
McFarland, R.A., F.J.W. Roughton, M.H. Halperin, and J.I. Niven. The effects of carbon
monoxide and altitude on visual thresholds. /. Aviat. Med. 75:381-394, 1944.
McGill, H.C., Jr. Risk factors in atherosclerosis. Adv. Exp. Med. Biol. 104:213-219, 1978.
McGill, B.C., Jr., V.C. Anselmo, J.M. Buchanan, and P.J. Sheridan. The heart is a target
organ for androgen. Science 207:115-116, 1979.
McManus, L.M., D.J. Hanahan, C.A. Demopolous, and R.N. Pinckard. Acetyl glyceryl ether
mediator (AGEM): intravascular pathobiology. Fed. Proc. Fed. Am. Soc. Exp. Biol. 39:
694, 1980.
Michalak, T. Immune complexes of hepatitis B surface antigens in the pathogenesis of periar-
teritis nodosa. Am. J. Pathol. 90:619-632, 1978.
Miller, G.J., and N.E. Miller. Plasma high-density lipoprotein concentration and development
of ischemic heart disease. Lancet 1975, i:16-19, 1975.
Minick, C.R. Immunologic arterial injury in atherogenesis. Ann. N. Y. Acad. Sci. 275:210-227,
1976.
Minick, C.R., D.R. Alonso, M.M. Litrenta, M.F. Silane, and M.B. Stemerman. Regenerated
endothelium and intimal proteoglycan accumulation. Circulation 42 (Suppl,):l44, 1977.
Minick, C.R., D.R. Alonso, and L. Rankin. Role of immunologic arterial injury in atherogen-
esis. Thromb. Haemostas. 39:304-311, 1978.
Minick, C.R., C.G. Fabricant, J. Fabricant, and M.M. Litrenta. Atheroarteriosclerosis in-
duced by infection with a herpes virus. Am. J. Pathol. 96:613-106, 1979.
Minick, C.R., and G.E. Murphy. Experimental induction of atheroarteriosclerosis by the
synergy of allergic injury to arteries and lipid-rich diet. II. Effect of repeatedly injected for-
169
-------
eign protein in rabbits fed a lipid-rich cholesterol-poor diet. Am. J. Pathol. 73:265-300,
1973.
Minick, C.R., G.E. Murphy, and W.G. Campbell. Experimental induction of atheroarterio-
sclerosis by the synergy of allergic injury to arteries and lipid-rich diet. I. Effect of repeated
injections of horse serum in rabbits fed a dietary cholesterol supplement. J. Exp. Med. 124:
635-654, 1966.
Moore, A., E.A. Jaffe, C.G. Becker, and R.L. Nachman. Myosin in cultured human endothelial
cells. Br. J. Haematol. 35:71-79, 1977.
Morris, J.N., S.P.W. Chave, C. Adam, C. Sirey, L. Epstein, and D.J. Sheehan. Vigorous exer-
cise in leisure-time and the incidence of coronary heart-disease. Lancet 1973, i:333-339,
1973.
Morris, J.N., M.D. Crawford, and J.A. Heady. Hardness of local water supplies and mortality
from cardiovascular disease. Lancet 1961, i:860-862, 1961.
Morris, J.N., A. Kagan, D.C. Pattison, M.J. Gardner, and P.A.B. Raffle. Incidence and pre-
diction of ischemic heart-disease in London busmen. Lancet 1966, ii:553-559, 1966.
Mroczek, W.J., F.A. Finnerty, and K.J. Catt. Lack of association between plasma-renin and
history of heart-attack or stroke in patients with essential hypertension. Lancet 1973, ii:464-
468, 1973.
Newball, H.H., R.W. Berninger, R.C. Talamo, and L.M. Lichtenstein. Anaphylactic release of
a basophil kallikrein-like activity. I. Purification and characterization. J. din. Invest. 64:
457-465, 1979.
Newball, H.H., R.C. Talamo, and L.M. Lichtenstein. Anaphylactic release of a basophil
killikrein-like activity. II. A mediator of immediate hypersensitivity reactions. J. Clin.
Invest. 64:466-475, 1979.
Notkins, A.L. The causes of diabetes. Sci. Am. 241:62-12, 1979.
Nurminen, M. Survival experience of a cohort of carbon disulphide exposed workers from an
eight-year prospective follow-up period. Int. J. Epidemiol. 5:179-185, 1976.
Oliver, M.F. Dietary prevention of coronary heart disease: the role of essential fatty acids. In
Proceedings of the Vih International Symposium on Atherosclerosis, edited by A.M.
Gotto, L.C. Smith, and B. Allen, Springer Verlag, New York, pp. 235-243, 1980.
Ortiz, G.A., A.E. Arguelles, H.A. Crespin, G. Sposari, and C.T. Villafane. Modifications of
epinephrine, norepinephrine, blood lipid fractions and the cardiovascular system produced
by noise in an industrial medium. Harm. Res. 5:57-64, 1974.
Osterland, C.K., E.J. Miller, W.W. Karakawa, and R.M. Krause. Characteristics of strepto-
coccal group-specific antibody isolated from hyperimmune rabbits. J. Exp. Med. 123:599-
614, 1966.
Ostrander, L.D..T. Francis, Jr., N.S. Hayner, M.O. Kjelsberg, and F.H. Epstein. The relation-
ship of cardiovascular disease to hyperglycemia. Ann. Intern. Med. (52:1188-1198, 1965.
Overturf, M.L., C. Aschenbrenner, R.E. Druilhet,and W.M. Kirkendall. Renin as a risk factor
for atherogenesis. Effects of hypercholesterolemia and two-kidney-one-clip hypertension
in the rabbit. Atherosclerosis 35:97-119, 1981.
Paffenbarger, R.S., Jr., W.E. Hale, R.J. Brand, and R.T. Hyde. Work-energy level, personal
characteristics, and fatal heart attack. A birth cohort effect. Am. J. Epidemiol. 105:200-
213, 1977.
Paffenbarger, R.S., Jr., and R.T. Hyde. Exercise as protection against heart attacks. Editorial.
N. Engl. J. Med. 302:1026-1027, 1980.
Paffenbarger, R.S., Jr., A.L. Wing, and R.T. Hyde. Physical activity as an index of heart attack
risk in college alumni. Am. J. Epidemiol. 108:\6\-\75, 1978.
Page, L.B., A. Damon, and R.C. Moellering, Jr. Antecedents of cardiovascular disease in six
Solomon Islands Societies. Circulation 49:1132-1146, 1974.
Parving, H.H., K. Ohlsson, H.J. Buchardt-Hansen, and M. Rorth. Effect of carbon monoxide
on capillary permeability to albumin and a2-macroglobulin. Scand. J. Clin. Lab. Invest. 29:
381-388, 1972.
Pauli, H.G., B. Truniger, J.K. Larsen, and R.O. Mulhausen. Renal function during prolonged
exposure to hypoxia and carbon monoxide. Scand. J. Clin. Lab. Invest. 103(Suppl.):55-60,
1968.
Pearson, T.A., E.G. Kramer, K. Solez, and R.H. Heptinstall. The human atherosclerotic
plaque. Am. J. Pathol. 86:657-664, 1977.
Peterson, D.R., D.J. Thompson, and J.M. Nam. Water hardness, arteriosclerotic heart disease
and sudden death. Am. J. Epidemiol. 92:90-93, 1970.
170
-------
Pinckard, R.N., M. Halonen, J.D. Palmer, C. Butler, J.O. Shaw, and P.M. Henson. Intravas-
cular aggregation and pulmonary sequestration of platelets during the IgE-induced syste-
temic anaphylaxis in the rabbit: abrogation of lethal anaphylactic shock by platelet deple-
tion. /. fmmunol. 7/9:2185-2193, 1977.
Pinckard, R.N., C. Tanigawa, and M. Halonen. IgE-induced blood coagulation alterations in
the rabbit: consumption of coagulation factors XII, XI, and IX in vivo. J. Immunol 115-
525-532, 1975.
Polimeni, P.I., and E. Page. Magnesium in heart muscle. Circ. Res. 33:367-374, 1973.
Polley, M., and R.L. Nachman. The human complement system in thrombin-mediated platelet
function. J. Exp. Med. 737:1713-1726, 1978.
Polverini, P.J., R.S. Cotran, M.A. Gimbrone, Jr., and E.R. Unanue. Activated macrophages
induce vascular proliferation. Nature (London) 2(59:804-806, 1977.
Ravens, K.G., H. Steer, R. Wrenski, and C. Pepe. Transitory myocardial damage by halo-
genated hydrocarbons. Dtsch. Med. Wochenschr. 99:1364-1367, 1974.
Reidy, M.A., and S.M. Schwartz. Endothelial regeneration. III. Time course of intimal changes
after small defined injury to rat aortic endothelium. Lab. Invest. 44:301-308, 1981.
Resnekov, L. Noise, radio frequency radiation and the cardiovascular system. Circulation 63:
264A-266A, 1981.
Rich, A.R. Hypersensitivity in disease with special reference to periateritis nodosa, rheumatic
fever, disseminated lupus erythematosus, and rheumatoid arthritis. Harvey Led. 42:106-
147, 1946.
Rich, A.R., and J.E. Gregory. Experimental anaphylactic lesions of the coronary arteries of
the "sclerotic" type, commonly associated with rheumatic fever and disseminated lupus
erythematosus. Bull. Johns Hopkins Hasp. 57:312-324, 1947.
Rider, A.K., J.C. Copeland, S.A. Hunt, J. Mason, M.J. Spector, R.A. Winkle, C.P. Bieber, M.
E. Billingham, E. Dong, R.B. Griepp, J.S. Schroeder, E.B. Stinson, D.C. Harrison, and
N.E. Shumway. The status of cardiac transplantation. Circulation 52:531-539, 1975.
Robertson, A.L., and P.A. Khairallah. Effects of angiotension II and some analogues on vas-
cular permeability in the rabbit. Circ. Res. 37:923-931, 1972.
Rokitansky, C. Lehrbuch der Pathologischen Anatomie, Vol. 2, 3rd ed., Wm. Bronmueller,
Vienna, 310 p., 1856.
Rosenman, K.D. Cardiovascular disease and environmental exposure. Br. J. Ind. Med. 3(5:85-
97, 1979.
Ross, R., J.A. Glomset, B. Kariya, and L.A. Harker. A platelet-dependent serum factor that
stimulates the proliferation of arterial smooth muscle cells in vitro. Proc. Natl. Acad. Sci.
U.S.A. 77:1207-1210, 1974.
Ross, R., and L.A. Harker. Hyperlipidemia and atherosclerosis. Chronic hyperlipidemia initi-
ates and maintains lesion by endothelial cell desquamation and lipid accumulation. Science
793:1094-1100, 1976.
Ross, R., and A. Vogel. The platelet-derived growth factor. Cell 74:203-210, 1978.
Rudel, L.L., R. Shah, and D.G. Greene. Study of the dislipoproteinemia induced by dietary
cholesterol in Rhesus monkeys (Macaco mulatto). J. Lipid Res. 20:55-65, 1979.
Rudel, L.L., C.W. Leathers, M.G. Bond, and B.C. Bullock. Dietary ethanol-induced modifi-
cations in hyperlipoproteinemia and atherosclerosis in nonhuman primates (Macaco
nemestrina). Arteriosclerosis 7:144-155, 1981.
Ryan, U.S., D.R. Schultz, R.J. Del Vecchio, and J.W. Ryan. Endothelial cells of bovine pul-
monary artery lack receptors for C3b and for the Fc portion of immunoglobulin. Science
203:748-749, 1980.
Sacks, T., C.F. Moldow, P.R. Craddock, R.K. Bowers, and H.S. Jacob. Oxygen radicals
mediate endothelial cell damage by complement-stimulated granulocytes. J. Clin. Invest.
(57:1161-1171, 1978.
Sammons, J.H., and R.L. Coleman. Firefighters'occupational exposure to carbon monoxide.
J. Occup. Med. 7(5:543-546, 1974.
Saphir, O. Inflammatory factors in arteriosclerosis. In Arteriosclerosis: A survey of the Pro-
blem, edited by H.T. Blumenthal, Charles C. Thomas, Springfield, Illinois, 1967.
Schmitt, H. Experimental Untersuchungen zur Pathogenese der Rheumatischen Athero-
sklerose. Virchows Arch. 29(5:603-626, 1935-1936.
Schroeder, H.A. Chromium deficiency in rats: a syndrome simulating diabetes mellitus with
retarded growth. J. Nutr. 58:439-445, 1966.
. Degenerative cardiovascular disease in the Orient. 1. Atherosclerosis. J. Chron.
Dis. 5:287-311, 1958.
171
-------
The role of chromium in mammalian nutrition. Am. J. Clin. Nutr. 21:230-244,
196f[
Schroeder, H.A., and J.J. Balassa. Influence of chromium, cadmium and lead on rat aortic
lipids and circulating cholesterol. Am. J. Physiol. 209:433-437, 1965.
Schroeder, H.A., and J. Buchman. Cadmium hypertension. Arch. Environ. Health 14:693-
699, 1962.
Schroeder, H.A., and L.A. Kraemer. Cardiovascular mortality, municipal water, and corro-
sion. Arch. Environ. Health 25:303-311, 1974.
Schroeder, H. A., A.P. Nason, and I.H. Tipton. Chromium deficiency as a factor in athero-
sclerosis. J. Chronic Dis. 23:123-142, 1970.
Schulte, J.H. The medical aspects of closed cabin atmosphere control. Mil. Med. 726:40-48,
1961.
Effects of mild carbon monoxide intoxication. Arch. Environ. Health 7:524-
530, 1963.
Schwartz, B.S., G.A. Levy, L.K. Curtiss, D.S. Fair, and T.S. Edgington.Plasma lipoprotein
induction and suppression of the generation of cellular procoagulant activity in vitro. Two
procoagulant activities are produced by peripheral blood mononuclear cells. J. Clin.
Invest. 67:1650-1658, 1981.
Schwartz, S.M., and E.P. Benditt. Aortic endothelial cell replication. I. Effects of age and
hypertension in the rat. Circ. Res. 41:248-255, 1977.
Schwartz, S.M., and D.M. Standaert. Effect of acute vs. chronic hypertension on endothelial
cell replication. In Frontiers in Hypertension, edited by J. Laragh, Springer Verlag, New
York, 1982 (in press).
Shadforth, M.F., P.H. Cunningham, and B.S. Andrews. The demonstration of Clq and Fc-
IgG receptors on the surface of human endothelial cells. Fed. Proc. Fed. Am. Soc. Exp.
Biol. 35:1075, 1979.
Sharma, H.M., and J.C. Geer. Experimental aortic lesions of acute serum sickness in rabbits.
Am. J. Pathol. 88:255-266, 1977.
Sharrett, A. R. The role of chemical constituents of drinking water in cardiovascular diseases.
Am. J. Epidemiol. //0:401-419, 1979.
. Water hardness and cardiovascular disease-elements in water and human
tissues. Sci. Total Environ. 7:217-226, 1977.
. Water hardness and cardiovascular disease. Circulation <53:247A-250A, 1981.
Siggard-Andersen, J., F. Bonde Peterson, T.I. Hansen, and K. Mellemgard. Plasma volume
and vascular permeability during hypoxia and carbon monoxide exposure. Scand. J. Clin.
Lab. Invest. 22 (SuppI.):39-48, 1968.
Siggard-Andersen, J., K. Kjeldsen, F. Bonde Peterson, and P. Astrup. A possible connection
between carbon monoxide exposure, capillary filtration rate and atherosclerosis. Ada
Med. Scand. 782:397-399, 1967.
Speizer, F.E., D.H. Wegman, and A. Ramirez. Palpitation rates associated with flurocarbon
exposure in a hospital setting. N. Engl. J. Med. 292:624-628, 1975.
Spiro, R.G., and M.J. Spiro. Effect of diabetes on the biosynthesis of the renal glomerular
basement membrane. Studies on the glycosyltransferase. Diabetes 20:641-648, 1971.
Stallones, R.A. The rise and fall of ischemic heart disease. Sci. Am. 243:53-60, 1980.
Stamler, J., and F.H. Epstein. Coronary heart disease: risk factors as guides to preventive
action. Prev. Med. 7:27-48, 1972.
Stehbens, W.E. The role of hemodynamics in the pathogenesis of atherosclerosis. Prog.
Cardiovasc. Dis. 7S:89-103, 1975.
Stoel, I., W.J. Giessen, v/d, E. Zwolsman, J.F.A. Quadt, F. ten Hoor, F. W.A. Verheugt, and P.
G. Hugenholtz. Effect of nicotine on prostacylin production in human umbilical arteries.
Circulation 62 (Suppl. ///J:97, 1980.
Stout, R.W., E.L. Bierman, and J.D. Brunzell. Atherosclerosis and disorders of lipid metabo-
lism in diabetes. In Diabetes: Its Physiological and Biochemical Basis, edited byJ. Vallance-
Owen, University Park Press, Baltimore, Md., pp. 125-169, 1975.
Stout, R.W., E.L. Bierman, and R. Ross. Effect of insulin on proliferation of cultured primate
arterial cells. Circ. Res. 3(5:319-327, 1975.
Strong, J.P., ML. Richards, H.C. McGill, Jr., D.A. Eggen, and M.T. McMurry. On the asso-
ciation of cigarette smoking with coronary and aortic atherosclerosis. J. Atheroscler. Res.
70:303-317, 1969.
Taylor, K., S. Glagov, J. Lamberti, D. Vesselinovitch, and T. Schaffner. Surface configuration
of early atheromatous lesions in controlled pressure, perfusion fixed monkey aortas. In
172
-------
Scanning Electron Microscopy, Vol. II, edited by O. Johari, SEM, Inc., A.M. F. O'Hare,
111., pp. 449-458, 1978.
Thomas, W.A., J.M. Reiner, K. Janakidevi, R.A. Florentine, and K.T. Lee. Population
dynamics of arterial cells during atherogenesis. X. Study of monotypism in atherosclerotic
lesions of black women heterozygous for glucose-6-phosphate dihydrogenase (G-6-PD).
Exp. Mol. Pathol. 57:367-386, 1979.
Thompson, J.G. Production of severe atheroma in a transplanted human heart. Lancet 1969,
ii:1088-1092, 1969.
Thomsen, H.K. An electron-microscopic study on the coronary arteries of Macaco irus
monkeys. Atherosclerosis 20:233-240, 1974.
Thomsen, H.K.., and K. Kjeldsen. Threshold limit for CO-induced myocardial damage: an
electron microscopic study in rabbits. Arch. Environ. Health 29:73-78, 1974.
Tsakraklides, V.G., L.C. Blieden, and J.E. Edwards. Coronary atherosclerosis and my.ocardial
infarction associated with systemic lupus erythmatosus. Am. Heart J. 57:637-641, 1974.
Turino, G.M. Effects of carbon monoxide on the cardiorespiratory system. Carbon monoxide
toxicity: physiology and biochemistry. Circulation 63:253A-259A, 1981.
Virchow, R.L.K. Cellular Pathology as Based Upon Physiological Pathological Histology
(1858), translated by F. Chance, Dover, New York, pp. 232-504, 1971.
Vlodavsky, I., P.E. Fielding, C.F. Fielding, and D. Gospodarowicz. Role of contact inhibition
in the regulation of receptor-mediated uptake of low-density lipoprotein in cultured
vascular endothelial cells. Proc. Natl. Acad. Sci. U.S.A. 75:356-360, 1978.
Vogel, J.A., and M.A. Gleser. Effect of carbon monoxide on oxygen transport during exercise.
J. Appl. Physiol. 32:234-239, 1972.
Voors, A.W., M.S. Shuman, G.P. Woodward, and P.N. Gallagher. Arterial lead levels and
cardiac death. Environ. Health Persp. 4:97-103, 1973.
Vracko, R., and E.P. Benditt. Manifestations of diabetes mellitus—their possible relationships
to an underlying cell defect: a review. Am. J. Pathol. 75:204-221, 1974.
Wald, M., S.M. Howard, P.O. Smith, and K. Kjeldsen. Association between atherosclerotic
diseases and carboxyhaemoglobin levels in tobacco smokers. Br. Med. J. 7:761-765, 1973.
Walton, K..W. The pathology of Raynaud's phenomenon of occupational origin. In The Vibra-
tion Syndrome, edited by W. Taylor and P.H. Pelmear, Academic Press, London, New
York, pp. 109-119, 1974.
Waugh.T.H., II. Cancer and the environment. Higginson speaks out. Science 205:1363-1366,
1979.
Wedeen, R.P., J.K. Maesaka, B. Weiner, G.A. Lipat, M.M. Lyons, L.F. Vitale, and M.M.
Joselow. Occupational lead nephropathy. Am. J. Med. 59:630-641, 1975.
Weill, H. Cardiorespiratory effects of inhalant occupational exposures. Circulation 63:250A.-
252A, 1981.
Weiss, S.M. (ed.). Coronary-prone behavior and coronary heart disease: a critical review (by
Review Panel). Circulation 63:1199-1215, 1981.
Wiener, J., R.G. Lattes, A.B. Barnet, B.G. Meltzer, and D. Spiro. The cellular pathology of
experimental hypertension. IV. Evidence for increased vascular permeabiility. Am. J.
Pathol. 54:187-207, 1969.
Wight, T.N., K.D. Curwen, W.P. Homan, and C.R. Minick. Effect of regenerated endothelium
on glycosaminoglycan accumulation in the arterial wall. Fed. Proc. Fed. Am. Soc. Exp.
Biol. 38:1075, 1979.
Williams, R.S., E.E. Logue, J.L. Lewis, T. Barton, N. W. Stead, A.G. Wallace, and S.V. Pizzo.
Physical conditioning augments the fibrinolytic response to venous occlusion in healthy
adults. N. Engl. J. Med. 302:987-991, 1980.
Wolinsky, H. Long-term effects of hypertension on the rat aortic wall and their relation to
concurrent aging changes. Morphological and chemical studies. Circ. Res. 30:301-309,
1972.
. Taking heart—atherosclerosis is down in America. We must be doing something
right. The Sciences 27:6-9, 1981.
Wolinsky, H., L. Capron, S. Goldfischer, F. Capron, B. Coltoff-Schiller, and L. Kasak.
Hydrolase activities in the rat aorta. II. Effects of hypertension alone and in combination
with diabetes mellitus. Circ. Res. 42:831-839, 1978.
Wolinsky, H., S. Goldfischer, L. Capron, F. Capron, B. Coltoff-Schiller, and L. Kasak.
Hydrolase activities in the rat aorta. I. Effects of diabetes mellitus and insulin treatment.
Circ. Res. 42:821-831, 1978.
173
-------
Wolinsky, H., S. Goldfischer, D. Katz, R. Markle, L. Gidez, S. Wassertheil-Smoller, and B.
Colthoff-Schiller. Hydrolase activities in the rat aorta. III. Effects of regular swimming
activity and its cessation. Circ. Res. 45:546-553, 1979.
Wood, P.D., and W.L. Haskell. The effect of exercise on plasma high-density lipoproteins.
Lipids 14: 417-427, 1979.
Yoon, J., A. Marshall, T. Onodera, and L.A. Notkins. Virus-induced diabetes mellitus. N.
Engl. J. Med. 300:1173-1179, 1979.
Zeek, P. Studies in Atherosclerosis. I. Conditions in childhood which predispose to the early
development of arteriosclerosis. Am. J. Med. Sci. /cW:350-356, 1932a.
. Studies in Atherosclerosis. II. Atheroma and its sequelae in rheumatic heart
disease. Am. J. Med. Sci. 184:356-364, \932b.
Zinkernagel, R.M., and P.C. Doherty. MHC-restricted cytoxic T cells: studies on the biolog-
ical role of polymorphic major transplantation antigens determining T cell restriction-
specificity function and responsiveness. Adv. Immunol. 27:51-177, 1979.
174
-------
FRONTIERS IN SHORT-TERM TESTING
OF PNEUMOTOXICANTS
D.B. Menzel, E.D. Smolko, D.E. Gardner,
and J.A. Graham
INTRODUCTION
Pulmonary disease is a health concern for some 47 million people in the United
States, representing about 24 percent of the entire population. Interstitial fibrosis
and chronic obstructive lung disease, which includes emphysema, bronchitis, and
asthma, are the major lung diseases. The exact contribution of environmental tox-
icants to the etiology and incidence of these diseases is not known, but ozone, nitro-
gen oxides, and sulfur oxides are likely to be contributing factors. Lung cancer is
mostly due to cigarette smoking, but the interaction between cigarette smoking and
air pollution may also exacerbate lung disease. Because lung diseases are difficult to
treat and are generally irreversible, increasing our understanding of the relationship
of pollutants to the causation of lung disease is an urgent concern.
The lung has approximately 40 cell types, all of which theoretically can be dam-
aged by air pollutants; the damage can be acute or chronic. Our focus here is on the
chronic effects of chemicals and on potential early diagnostic techniques predictive
of human chronic lung injury. It will become evident that no validated predictive
short-term tests now exist. However, there are some promising methods that, with
further study, may become adequate early detection systems. Clearly, the develop-
ment of such methods is a highly meritorious objective.
Chronic lung diseases have been the subject of intense study for many years. The
precise mechanisms of the pathogenesis of human lung diseases are unknown in
many instances, but the search for such mechanisms has revealed important corre-
lations between the biochemistry, function, and structure of the entire lung and cer-
tain cell types, particularly alveolar macrophages, fibroblasts, epithelial cells, and
mucous-secreting cells. The validity of animals as proper surrogates for man in the
study of lung disease has become increasingly sustained as the basic mechanisms of
lung disease are uncovered. Because the precise mechanisms of human lung disease
are not known, the selection of a few critical test parameters for study is not now
possible. Lacking such information, one is constrained to choose techniques that re-
flect some aspect(s) of the chronic disease itself and that are sufficiently sensitive to
detect early disease manifestations.
Within these limits, what are the special criteria for short-term tests predictive of
chronic pulmonary disease?
1. Specificity: What is the extent of false positive and false negative determina-
tions, and to what degree are such determinations acceptable? No widely ap-
The Authors: Daniel B. Menzel and Elaine D. Smolko are on the staff of the Departments of Pharmacology
and Medicine, Duke University Medical Center, Durham, North Carolina. Donald E. Gardner is on the statf
of Northrop Services, Incorporated-Environmental Sciences, Research Triangle Park, North Carolina.
Judy A. Graham is on the staff of the Inhalation Toxicology Division, Health Effects Research Laboratory,
United States Environmental Protection Agency, Research Triangle Park, North Carolina.
175
-------
plied short-term test is assumed to be 100 percent accurate, but such a test is
usually used to identify chemicals for further in-depth examination. Thus,
false positive findings are more acceptable than false negative results. Since
the objective of the short-term test is to identify chemicals that cause chronic
diseases of a severely adverse nature, such as emphysema and fibrosis, false
negative results are undesirable. For ubiquitous pollutants to which large seg-
ments of the population are exposed, the impact of false negative results could
be catastrophic, so their acceptable level should be extremely low. A low level
of false negative results is acceptable for the myriad of remaining, less com-
mon pollutants to which the population at large is exposed only infrequently,
since rapid testing will enable the toxicity of more chemicals to be identified.
2. Validity: The short-term test must be an accurate predictor of the chronic ef-
fect of the pollutant in the vast majority of the human population. But first,
knowledge is needed of the pathogenesis of the human disease process to en-
sure an outcome analogous to the chronic disease. Careful consideration must
also be given to unusually susceptible subpopulations. For example, observa-
tions of young cigarette smokers showed that men responded to the smoke
differently than women (Becklake and Permutt, 1979). Men were more sus-
ceptible to the development of chronic obstructive lung disease (COLD), and
they primarily exhibited an increased resistance in the small airways, whereas
women principally exhibited an increased resistance in the large airways and a
greater pulmonary vascular response to smoking. Therefore, a screening test
developed in only one sex may not be predictive of the disease process in the
other sex. Genetic differences may be similarly important, but are largely un-
investigated.
3. Cost-effectiveness: With finite resources, priorities must be carefully assessed.
If screening tests are of poor specificity and validity, then money would be
better spent on chronic studies with animals in which the experimental
outcome is more clearly associated with a human disease. For highly suspect
chemicals, chronic studies must still be undertaken to validate the short-term
results.
With these three broad objectives in mind, a brief review of some of the most
promising avenues of research is presented. A short orientation to the special prop-
erties of the respiratory system and the consequent limitations of inhalation toxicol-
ogy is essential to our discussion.
MORPHOLOGY OF THE LUNG IN TOXIC RESPONSES
The deposition and retention of inhaled gases and aerosols are influenced not only
by the physicochemical properties of the chemicals, but also by the anatomical fea-
tures of the respiratory tract, including the lung volume, the alveolar surface area,
and the structural and spatial relationships of conducting airways into the alveoli
themselves. Extensive reviews of regional deposition of particles and gases are avail-
able (U.S. Environmental Protection Agency, 1982; Raabe, 1982; National Research
Council, 1977). The distribution of deposited material as a function of time along
with the location of the different cell types within the respiratory tract determine
particular cells or regions of the lung that are at risk. The development of short-term
tests of pulmonary diseases is complicated by regional specializations of the lung and
by the large number of different cell types that exist within these regions. The influ-
ence of structure on function is so fundamental that most chronic lung diseases are
defined primarily by their morphology and secondarily by their functional
changes. A clear understanding of the morphology of the normal lung is therefore
essential to insights into both the distribution of toxicants and the responses of the
lung to injury.
176
-------
The Normal Anatomy of the Mammalian Lung
Three major regions comprise the respiratory tract: the nasopharyngeal, the
tracheobronchial, and the pulmonary regions (Fig. 1). The nasopharnyx starts at the
anterior nares or buccal-nasal cavity and extends downward to the level of the
larnyx. The nasal passages are lined with a well-perfused mucous epithelium com-
posed of a ciliated columnar epithelium with many scattered mucous glands. The
nasopharnyx in humans and animals filters out large inhaled particles and increases
the relative humidity and temperature of inhaled air. The trachea, bronchi, and
bronchioles conduct air from the nasopharnyx to the alveolar region where gas ex-
change occurs. The airways of the tracheobronchial region are lined with a ciliated
epithelium, which is coated with a thin layer of mucus that is continuously secreted
from goblet cells and mucous-secreting glands. The mucous covering terminates at
the surfactant film covering the alveolar membranes. This mucous covering is con-
stantly propelled upwards to the nasopharnyx, thereby serving as a moving escalator
to transport particles out of the lung. Any particles that are deposited in the alveoli
may be engulfed by macrophages and removed from the respiratory tract. The
branching patterns and physical dimensions of the airways influence the deposition
of inhaled particles and the absorption of inhaled gases.
The physical dimensions of the airways of human and animal lungs have been de-
scribed by a number of mathematical models (Weibel, 1963; Davies, 1961). In hu-
mans, the airways tend to be equal and bifurcating, decreasing in diameter as they
divide. As bifurcation increases, the cross-sectional area increases. Since the cross-
sectional area increases as the gas diffuses downwards in the lung, a second phenom-
enon of radial diffusion also occurs. Gases tend to be diluted by this combination of
anatomical features, independent of mixing with inspired gases. While human and
animal lungs differ, the anatomies of both respiratory systems are sufficiently similar
for equivalent divisions to be made. Generally, these divisions, or so-called gener-
ations, are based upon the number of times that the major airway divides to reach a
particular anatomical site.
The acinus is the basic functional unit of the mammalian lung and is the primary
location of gas exchange between the environment and the blood. The acinus con-
sists of structures, or acini, starting at the respiratory bronchiole. About 200,000
acini exist in the human lung, including three or four generations of respiratory
bronchioles, several generations of alveolar ducts and alveolar sacs, and hundreds of
alveoli and associated blood vessels, lymphatic tissues, supportive tissues, and nerve
endings. Quantitative anatomical information on the lung is available, including
estimates of airway tube numbers, diameters and lengths; alveolar numbers and di-
ameters; and surface areas and mean thicknesses for the air-to-blood barrier (Weibel,
1963; Kliment, 1973). This information is important to the quantitative estimation
of dose in humans on the basis of dose received by animals. The dose-response rela-
tionship for various regions within the lung is highly important to the eventual devel-
opment of short-term tests.
Because structure is so intimately associated with function in the lung and, in turn,
with the deposition and the dose of toxicants reaching specific lung segments, short-
term tests need to be specialized to regions of the lung. Chronic diseases in humans
are often localized in specific segments rather than existing throughout the lung.
Bronchitis, for example, is confined to those airways containing ciliated and mucous
secretory glands. Emphysema, on the other hand, is confined to the alveolar or gas
exchange regions. The anatomical separation of diseases can be analyzed in more de-
tail by using the pulmonary function tests and morphological examinations that are
discussed below.
Tests that are designed to sample cells from the various regions of the lung have
been developed. Fiberoptic bronchoscopes allow biopsy procedures to sample
major conducting airways in humans, although such devices are too large to extend
177
-------
Upper respiratory tract
Anterior nares
rachea
Lower vxlxjxixlxxivxix;:;:
respiratory •|:jX:X;:;Xx:;::Xjx;:;X;:
tract :*::x:x::::::::::::*:*
Pulmonary
Bronchus
Aveoli
Figure "\. Three major regions of the respiratory tract. The upper respiratory tract is
known as the nasopharyngeal region. The lower respiratory tract is
divided into the conducting airways (tracheobronchial) and the gas
exchange (pulmonary) regions.
beyond the major bronchi of animals used routinely in toxicology. An alternative
means of sampling is lung lavage (Section VIII). This procedure samples all of the
cells and secretions that can be removed by mild agitation with solutions in the lung.
One cell type readily recovered by lavage is the pulmonary macrophage, which nor-
mally resides within the alveoli. The macrophage is motile and arises from the bone
marrow; in the lung, macrophages serve an important function by collecting soluble
and insoluble material from the alveoli and transporting the material up to the junc-
tion with the mucociliary escalator. There, the dead or dying macrophages are trans-
ported slowly out of the lung on the blanket of mucus covering the conducting air-
ways. Another important function is to engulf and kill bacteria that otherwise might
cause lung infections. The death of macrophages is not without consequences in the
lung because they contain powerful degrading enzymes which, when released, act
upon other cells to produce a number of pathophysiological conditions. When pur-
ified, concentrated, and instilled into the lung, these degrading enzymes can produce
experimental emphysema and bronchitis.
Many of the common chronic pulmonary diseases of humans result in changes in
the numbers or anatomical relationships of cells within the lung. Interstitial fibro-
sis results from a thickening of the distance between the gas exchange region and the
pulmonary capillary bed. Since the lung is designed to function with an exquisite
balance between gas exchange and blood flow, changes in either condition result in
inappropriate gas exchange and in major compromises in the functional capacity of
the lung.
The Anatomical Pathology
Although humans suffer from a large number of pulmonary diseases, chemicals
have been im plicated in the etiology of only a few of the more common diseases. One
178
-------
large class is chronic obstructive lung disease (COLD), which includes primarily
bronchitis and emphysema. Bronchogenic cancer is also associated with cigarette
smoking and, in animals, with experimental exposure to chemical carcinogens.
Bronchitis and asthma may be exacerbated by exposure to chemicals, and interstitial
fibrosis may be initiated or aggravated by chemical exposure. Pulmonary fibrosis is
associated with fibrous mineral exposure, and mesothelioma, a tumor of the pleura,
is particularly associated with asbestos exposure in both humans and animals. Pneu-
moconiosis is associated with exposure to mineral dusts in humans and
with formaldehyde in animals. Some examples of these diseases, as well as normal
tissue, are shown in Fig. 2.
COLD is particularly important in terms of long-term exposure to air pollutants
and can be induced in rodents by both ozone and nitrogen dioxide at concentrations
greater than those occurring in urban air (Freeman et al., 1972, 1974). The charac-
teristics of obstructive and restrictive lung diseases have been reviewed by West
(1977) and Thurlbeck (1979). COLD is of two major types—emphysema and chronic
bronchitis—but the two types are generally classed together because small airway ob-
struction (<2 mm), leading to gas trapping which may make expiration difficult in
advanced cases, is a common hallmark. In some patients, both emphysema and
chronic bronchitis are present.
The classical definition of emphysema is the destruction of the alveoli distal to the
terminal bronchiole, with distension of their walls. These anatomical changes may
be associated with measurable changes in pulmonary function. Pulmonary func-
tion tests are widely used in humans as a diagnostic aid. Although anatomical
changes have been used more extensively than pulmonary function tests to estimate
the progression of pathophysiology, there is an insufficient number of detailed quan-
titative analyses in animals that relate anatomical changes to pulmonary function
measurements. The issue, then, for a chemically induced chronic disease is which is
the more sensitive and practical technique of potential application to a short-term
test. Recent advances in measuring pulmonary physiology in small animals (O'Neil
and Raub, 1982) suggest that this issue could be evaluated shortly. Minor to extreme
levels of chronic diseases, such as emphysema and fibrosis, can be chemically in-
duced with elastase, bleomycin, or silica. At various times during pathogenesis,
physiologic, morphologic, and morphometric measurements could be made and
compared. A few such studies have already been performed (Raub et al., 1982a;
Snider and Sherter, 1977), but the doses of elastase used have caused both functional
and structural alterations. A further complication is that human emphysema pres-
ents as several different types (centrilobular, panlobular, panseptal, and bullous),
depending on the site and distribution of the lesion. However, emphysema in
animals has some differences from emphysema in man (Port etal., 1977). In another
recent study, neonatal rats were exposed intermittently for 6 weeks to 490 /j.g of
ozone perm3 (0.25 ppm) (Raub etal., 1982b;Raub, 1982; Barry etal., 1982a,b). Vital
capacity and total lung capacity were increased, while other pulmonary function
parameters were not altered. Qualitative morphology detected no damage to the
gaseous exchange region of the lung, although sophisticated morphometric analyses
showed a number of alterations in the alveolar epithelium. At present, data are insuf-
ficient to state confidently whether structural or functional changes occur first, thus
making np single technique preferred for short-term testing. When such information
is available, the level of difficulty of the test procedure will likely be a significant
determinant of the choice of short-term test (i.e., functional or structural).
Chronic bronchitis is clinically defined as excessive bronchial mucus production,
causing excessive expectoration of sputum. Histological examination of human
bronchitic airways shows hypertrophy of mucous glands in the large bronchi and
chronic inflammatory changes, including cellular infiltration, edema, and some
increase in smooth muscles in the small airways. In COLD, the increased resistance
to airflow can be caused by conditions that affect the diameter of the functional air-
179
-------
Figure 2. Some common lung diseases in man The authors are grateful to Phillip C.
Pratt, M D., Department of Pathology, Duke University Medical Center, for
providing the photomicrographs.
(a) Interstitial fibrosis. Note the honeycombed appearance of the lung.
180
-------
(b) Primary carcinoma. The white mass in the upper left portion of the lung has grown
around the major airways and vessels and has even reached the lower margin
of the aorta.
181
-------
«^*;"'« »-*:.-• "* 7-is^'-~'#*5l£.-:
•«^L:-:,I*.StJ•:-".' • S&»R>4ifc^S?
-------
(d) The wall of the bronchus of a chronic bronchitic patient. The mucous glands are greatly enlarged, and the fine boundary of cilia and mucous
cells has been lost. (4Ox original). ' . > u ,.
-------
way lumen. In bronchitis, excessive mucus secretions or thickening of the airway
wall by edema or muscular hypertrophy effectively narrow the lumen. In emphyse-
ma, the loss of alveolar walls and parenchyma, which normally exert a radial trac-
tion on the airways, allows the airway to narrow.
In restrictive pulmonary diseases, the ability to expand the lung is decreased,
making inflation of the lung more difficult; this can be caused by alterations in lung
parenchyma, pleura, chest wall, or neuromuscular function. Pulmonary fibrosis is
the major disease group in this category. In fibrosis, the interstitium, i.e., the tissue
between the alveolar epithelium and the capillary endothelium, becomes thickened.
The process begins with an inflammatory infiltration of lymphocytes and plasma
cells, followed by an increase in the number of fibroblasts that produce collagen. In
its most severe form, the alveolar structure is destroyed, and excessive deposition of
collagen or scarring occurs. In fibrosis, the airway diameter is enlarged due to exces-
sive radial traction, as opposed to a narrowing in COLD.
PULMONARY EXPOSURE METHODS
Since the respiratory tract is the main portal of entry and one of the first surfaces
to come in contact with airborne chemicals, toxicological studies must be designed
and conducted to mimic as closely as possible this natural route of exposure. The test
substance, therefore, must be generated in clean air in a physical and chemical state
that most nearly duplicates that found in the environment. The choice of technique
for conducting inhalation studies with experimental animals must take into con-
sideration the duration of the exposure, the chemical and physical state of the test
substance, the cost and availability of the exposure facility, and the need for the pre-
vention of exposure through nonrespiratory pathways, such as skin, eyes, or food. A
number of excellent reviews describe different exposure systems (i.e., whole-body,
head-only, nose- or mouth-only, intratracheal instillation, and isolated ventilated
perfused lung), operational procedures, generation of test atmospheres, and sam-
pling techniques (Willeke, 1980; Phalen, 1976; Drew, 1978; Rampy, 1981; Jacoby
and Barthold, 1981). Depending on the objective of the experiment, animals are ex-
posed completely or partially to the test atmosphere. Whole-body exposures are the
commonest and are reviewed in detail by Hinners et al. (1966), Leach (1958), and
Drew and Laskin (1973). Exposures restricted to the head (Thomas and Lie, 1963;
Mauderly et al., 1971; Lippman, 1980) or to the nose or mouth (Blair et al., 1969;
Raabe, 1973; Braretal., 1975) have the advantage of reducing the simultaneous in-
gestion of material deposited on the body. Test material can also be instilled via the
trachea (Hatch etal., 1981; Phalenand Morrow, 1973; Auerbachetal., 1970; Bianco
et al., 1974; Gardner etal., 1972). Effects of toxicants on the lung can be measured ex
vivo in perfused lung preparations (Niemeier, 1976; Anderson and Eling, 1976;
Charles and Menzel, 1979; O'Neil and Tierney, 1974; Young 1976) from exposed or
normal animals. The technology of each method is complex.
FUNCTION TESTS FOR CHRONIC PULMONARY DISEASE
A number of pulmonary function tests, which can be used on a routine basis with
small laboratory animals for general toxicological testing, have been developed.
Several excellent books, reviews, and monographs describe the principles of respira-
tory physiology and the specific techniques for making such measurements (O'Neil
and Raub, 1982; Horvath, 1977; West, 1977; Slonim and Hamilton, 1976; Becklake
and Permutt, 1979). This section will briefly discuss some of the measurements that
were specifically developed to be similar to tests that are being used in humans to
diagnose lung disease.
The maximum volume of gas that the lung can contain is referred to as the total
lung capacity (TLC). The TLC can be conveniently subdivided into four indepen-
dent volumes and four lung compartments or capacities that can be measured by
pulmonary function tests. Fig. 3 shows these subdivisions.
184
-------
£
u
13
z
* I J J
INSPIRATORY RESERVE VOLUME
RESTING TIDAL
VOLUME
FUNCTIONAL
, RESIDUAL .
CAPACITY
t-A-A-
TIDAL VOLUME
(ANY LEVEL OF
ACTIVITY)
EXPIRATORY RESERVE VOLUME
v t r t
t
RESIDUAL
VOLUME
1
SPECIAL DIVISIONS FOR
PULMONARY FUNCTION TESTS
PRIMARY SUBDIVISIONS
OF LUNG VOLUME
-O
J
Figure 3. Spirometric measurements of lung volumes. The resting tidal volume is the volume of air
inspired and expired during normal breathing. By inspiring and expiring maximal amounts of air,
the vital capacity is measured. Other divisions are apparent and are described in pulmonary
function tests (from West, 1977; Slonim, 1976).
-------
Another category of static mechanical measurements made in laboratory animals
is the volume-pressure relationship of the respiratory system. Compliance is a mea-
sure of the distensibility of the chest wall and lungs, i.e., the ease with which lung vol-
ume is changed. Static compliance is the slope of a static volume pressure curve at a
given point or the slope of a linear approximation to a nearly straight portion of such
a curve in the tidal volume range. Compliance of the respiratory system (Crs) in-
cludes that of the chest wall (Cw) and the lung (Cl); assessment of both Cland Cwis
important when examining the elastic recoil of the system. Cw values vary with the
size and species of the animal, but Cl values are determined largely by the elastic
recoil of the lung. The elastic recoil is the sum of the recoil forces of the lung tissues in
the alveolar walls and smallest airways, and the surface tension forces of the liquid-
air interface lining in the gas exchange area of the lung. Diseases that change the
elastic properties of the lung or the surface tension within the lung can affect lung
compliance. Certain diseases decrease lung compliance, and the lungs become stiff;
other diseases increase compliance, with a loss of normal elasticity.
In contrast to the static measurements of lung function, dynamic measurements
describe and quantify the mechanics of the respiratory system under conditions of
respiratory gas flow. These include measurements of pulmonary and airway resist-
ance, flow-volume relationships, and frequency dependence of compliance and re-
sistance. Measuring dynamic pulmonary mechanics in small mammals involves a
number of difficulties: (1) the high frequency at which the animal breathes; (2) the
small dimensions of the airways; (3) the relatively large dead space of conventional
measuring equipment; and (4) the relatively poor frequency-response characteristic
of equipment measuring flow, volume, and pressure in the respiratory system (Koo
et al., 1976; Sinnett, et al., 1981).
Minute ventilation is calculated as the product of tidal volume (TV) and respir-
atory frequency (f). Dead space ventilation (VD) includes both the anatomical and
the physiological dead space. In terms of gas exchange, this dead space is wasted
ventilation and increases with certain lung diseases. Anatomical dead space is in-
creased with exercise, age, or emphysema, but is decreased in asthmatics as a result
of airway narrowing. In patients with lung disease, the physiological dead space may
increase because of an inequality in blood flow and ventilation within the lung. Pul-
monary resistance (Ri_), a measurement of the frictional resistance in the airway
(RAW) and the tissue viscocity resistance (Ry), is relatively low in the healthy animal.
In cases where alterations are found in the cross-sectional area of the large airways
(bronchoconstriction), RT can change substantially. In laboratory animals,
RAW can be substantially altered following exposure to chemicals.
A very useful test of pulmonary function is the measurement of the maximal
volume of air that can be exhaled forcefully (FVC) after a maximal inspiration. The
FEVi is the volume of air that can be forcibly exhaled during the first second of ex-
piration; this measurement has been regarded as one of the most useful overall spiro-
metric measurements in humans. The measurement of maximum respiratory flow at
different volumes can also be used to examine the relative contribution of small and
large airways to pulmonary resistance in injured lungs. When airway disease or in-
jury is present, or if lung recoil is reduced, flow at any volume can be reduced, com-
pared to that of the normal lung. Mauderly et al. (1979) found that rats with emphy-
sema have lower flows at all volumes and lower initial peak flows than control ani-
mals.
A number of techniques are used to detect maldistribution of ventilation following
mild injury to the small airways by inhaled toxicants (O'Neil and Raub, 1982; Drill
and Thomas, 1980). Measurement of the impedance to gas flow in and out of the
lung may be the most sensitive functional test for mild, small-airway lesions. The
single-breath nitrogen washout test gives an index of the gas distribution in the lung
and has been shown to be sensitive for detecting airway disease (O'Neil and Raub,
1982; Drill and Thomas, 1980). This test has not been widely used in small animal
186
-------
toxicology. Multiple breath nitrogen washout has also been used. Raub et al. (1982a)
observed that hamsters that have experimental emphysema show marked changes in
the slope of the nitrogen washout and in the breath index.
The lung's primary function is the exchange of oxygen and carbon dioxide, which
move between air and blood by simple diffusion. The transfer of respiratory gases
between the alveolar spaces and the pulmonary capillary blood depends on the level
of alveolar ventilation, the evenness of ventilation, the volume and distribution of
pulmonary blood flow, and the distance needed for diffusion between the alveolar
wall and the pulmonary capillary blood. Abnormal gas exchange can be estimated
by measuring the rate at which a test gas, such as carbon monoxide, is absorbed into
the blood from the airspaces. Such measurements (diffusion capacity for carbon
monoxide, DLCO) can be accomplished in small animals and appear sensitive to
subtle changes in pulmonary structures (Takezawa et al., 1980).
Measurement of arterial blood-gas partial pressures can give a general overview of
the respiratory gas exchange and total pulmonary function. Pulmonary damage re-
sults in an abnormally low partial pressure or concentration of oxygen and, in some
cases, an abnormally high concentration and partial pressure of carbon dioxide.
Low arterial oxygen partial pressure (hypoxemia) can result from: (1) hypoventila-
tion, (2) low-inspiratory oxygen, (3) pulmonary alveolocapillary diffusion impair-
ment, (4) mismatching of ventilation and perfusion, and (5) right-to-left pulmonary
vascular shunt. The measurement of blood gases can be useful in evaluating both
acute and chronic conditions associated with pulmonary dysfunction. Acid-base im-
balances, increases or decreases in the partial pressure of carbon dioxide or oxygen,
and differences in the oxygen partial pressure between the alveolus and the arteri-
alized blood have all been used to compare the blood gases in normal and emphy-
semic laboratory animals (Lucey et al., 1978).
PULMONARY FUNCTION TESTS FOR CHRONIC OBSTRUCTIVE
AND RESTRICTIVE LUNG DISEASES
The most common and severe obstructive (emphysema and chronic bronchitis)
and restrictive (fibrosis) lung diseases are associated with functional as well as struc-
tural alterations (West, 1977; Thurlbeck, 1979). Rapid tests predictive of chronic
lung disease in humans have been sought (Becklake and Permutt, 1979), and the
ones most described have been pulmonary function measurements. A patient with
COLD will not be able to exhale air forcefully after maximal inhalation as quickly as
a normal person; i.e., FEVi is reduced, and the total volume (FVC) expelled is less
(Fig. 4). However, considerable morphological changes occur before FEVi is altered
(Thurlbeck, 1979). In patients with restrictive lung disease, FVC is also reduced, but
a larger percentage of air is exhaled in the first second compared to patients with
COLD. Maximum midexpiratory flow rate (MMFR) is also altered in these disease
states. In COLD, MMFR is below normal; in restrictive disease, MMFR can be
above normal. Flow-volume curves (Fig. 5) and lung volumes (Fig. 6) are also altered
characteristically in restrictive and obstructive diseases. The distribution of ventil-
ation within the lung is altered as measured by nitrogen washout techniques. In pa-
tients with COLD, small airways are enlarged, causing a lower concentration of in-
spired nitrogen in more distal airways; alveolar units can receive inspired gas from
adjacent units, which can then be detected. Closing volumes can be a sensitive mea-
surement of alveolar dysfunction in patients with COLD. In apparently healthy
smokers with normal vital capacity (VC), closing volumes can be increased. Changes
in closing volume may be useful in the early identification of people likely to develop
COLD, but its use in animals is highly speculative at present. Inflammatory changes,
hyperplasia of mucous glands, and excess mucus secretions of chronic bronchitis
narrow airways and can cause a premature airway closure. Destruction of alveolar
walls and the resulting loss of radial traction on small airways can also cause
increased closing volumes.
187
-------
A. NORMAL
B. OBSTRUCTIVE
C. RESTRICTIVE
FEV = 4.0
FVC = 5.0
% = 80
MMFR = 3.5 I/sec
FEV = 1.3
FVC = 3.1
% = 42
MMFR = 1.4 I/sec
FEV = 2.8
FVC = 3.1
% = 90
MMFR = 3.7 I/sec
Figure 4. Examples of forced expiratory volume (FEV), forced vital capacity (FVC),
and maximal midexpiratory flow rates (MMFR) in normal individuals
and patients with chronic obstructive and restrictive lung diseases
during forced expiration into a spirometer. The percent is the ratio of
FEV to FVC (from West, 1977).
8 -i
FLOW
RATE
(I/sec)
6 -
4 -
2 -
Normal
8642
LUNG VOLUME (I)
i
0
Figure 5. Flow-volume curves illustrating the different patterns existing in
patients with obstructive and restrictive diseases (from West, 1977).
Other techniques like the single-breath nitrogen test, DUO, frequency depend-
ence of compliance, closing volume, and difference in flow rates (MMFR 50 and 25
while breathing air and a helium-oxygen mixture) have promise, but experience is
still sparse due to the recent introduction of these tests (Becklake and Permutt, 1979;
Tattersall et al., 1978; Ranga and Kleinerman, 1978).
Pulmonary function testing in animals is based on the same principles as in hu-
mans and has been reviewed by O'Neil and Raub (1982) and by Leith( 1976). Most of
these tests assume a commonality of physiological principles, and this is generally
true. For example, functional changes observed in elastase-induced emphysema in
hamsters and rats (Raub et al., 1982a; Mauderly et al., 1979; Radinsky and Snider,
1978) are similar to those occurring in humans. Nevertheless, there are differences,
and one should exercise caution in automatically assuming such similarities (O'Neil
and Raub, 1982).
188
-------
LUNG VOLUMES
OBSTRUCTIVE DISEASE
NORMAL
1C
IRV
TV
ERV
RV
vc
RV
o
Figure 6. Comparison of normal lung volumes to those observed in restrictive and obstructive disease states. Abbreviations are as follows:
TLC, total lung capacity; 1C, inspiratory capacity; FRC, functional residual capacity; VC, vital capacity; TV, tidal volume; IRV,
inspiratory reserve volume; ERV, expiratory reserve volume; RV, residual volume.
-------
The chronic effects of air pollutants are illustrative of the difficulties likely to
occur when attempting to develop and use short-term tests. These difficulties arise
out of the complexity of environmentally induced pulmonary disease. An extensive
chronic study of air pollutant toxicology has been summarized by Stara et al. (1980).
Beagle dogs were exposed for 14 hours per day for 68 months to air or to one of seven
treatments with raw or photochemically reacted auto exhaust, oxides of sulfur or
nitrogen, or their combinations. After exposure ceased, the dogs were allowed to re-
main in clean air for 32 to 36 months. Pulmonary function was studied after 18, 36,
and 61 months of exposure and at 2 years postexposure; pulmonary structure was
examined 32 to 36 months postexposure. Lung morphology was studied using light
microscopy, scanning electron microscopy (SEM), and transmission electron micro-
scopy (TEM) (Hyde et al., 1978). The observed structural changes were interpreted
as being analogous to an incipient stage of human proximal centrilobular
emphysema. Some pulmonary function changes, which were correlated to the
structural changes in the postexposure period, also occurred.
Could the morphological changes have been observed earlier by pulmonary func-
tion tests? Compared to controls, no significant effects on pulmonary function were
observed afer 18 or 36 months of exposure (Vaughanetal., 1969; Lewis etal., 1974).
However, in an analysis of the incidence of dogs having abnormal values based on
clinical criteria, the oxides of nitrogen group had a lower Duo/TLC after 36
months of exposure. After 61 months of exposure, an analysis of incidence of dogs
with normal values showed that the only major change was in the oxides of nitrogen
group, which had a decrease in peak expiratory flow rate (Lewis et al., 1974). A wide
battery of pulmonary function tests was applied, but sensitive tests of small airway
function in animals were not available at that time. About 2 years after exposure
ceased, pulmonary function was examined again by others and included some newer
techniques, such as frequency dependence of compliance, that should be helpful in
detecting changes in small airways (Gillespie, 1980). Additional changes were ob-
served in the oxides of sulfur or nitrogen groups, including decreased Duo/TLC,
decreased chest wall compliance, increased TLC, and greater changes in dynamic
compliance with increasing frequency. When the results for the immediate and 2-
year postexposure studies were compared, residual volume, vital capacity, func-
tional residual capacity, inspiratory capacity, and total lung capacity of the experi-
mental dogs increased compared to the controls. Thus, pulmonary function con-
tinued to deteriorate after exposure ceased. Since the functional changes in all ex-
posure groups correlated well with structural changes, it can be hypothesized that
structural changes also progressed during the postexposure period.
Transient changes in the breathing mechanics of experimental animals have also
been used as a short-term test of the acute effects of pulmonary irritants. Two sys-
tems have been developed for general use. Amdurand Mead (1955,1958) developed
a method to measure the pulmonary mechanics of the unanesthetized guinea pig.
The method has been evolved into a highly sophisticated procedure which monitors
pulmonary function. Most irritant gases, such as sulfur dioxide, produce an increased
resistance to air flow. Dose-response curves can be developed and related quantita-
tively to exposures. Generally, the Amdur method using unanesthetized guinea pigs
is more sensitive to pulmonary irritants than other methods using anesthetized
animals or animals other than guinea pigs. The general correlation between guinea
pig and human responses is good, and the method has been used extensively in air
pollution research.
Alarie and coworkers (1973) developed a method using mice confined in a modi-
fied whole-body plethysmograph. They found good correlations between a number
of industrial chemicals that are irritants for man and changes in respiratory rate in
mice. Alarie (1973) presented an extensive review of the physiological basis for the
assay and compared the relative potency of these irritants in the mouse assay with
reports from the workplace.
190
-------
At present, pulmonary function measurements are not more predictive of injury
than morphological changes. More research including state-of-the-art techniques
for pulmonary physiology and morphometry is needed before this issue can be re-
solved. To date, such comparisons have indicated that structure is measurably altered
before function (Raub et al., 1982b,c; Barry et al., 1982a,b). However, the morpho-
metric techniques used in these studies are too sophisticated for use in routine test-
ing, further complicating the issue. At issue, also, is the relevancy of transient
changes in pulmonary function measured by the Mead and Amdur, and Alarie tech-
niques. No correlation has been attempted between these techniques and the chronic
effects in animals or occupational disease in man. Alarie (1973) addresses this issue
in a more detailed discussion of his technique.
EARLY DETECTION OF CHRONIC BRONCHITIS
A research group (Lamb and Reid, 1968; Reid, 1970; Reid and Jones, 1980) devel-
oped an animal model of human chronic bronchitis in rats. The rat was chosen be-
cause, unlike the hamster, mouse, rabbit, and guinea pig, it has mucous glands that
in chronic bronchitis increase in size, number, and activity, and spread to deeper
parts of the lung. Characteristics of chronic bronchitis were produced by using sulfur
dioxide or cigarette smoke as the inducing agent, and the size and number of mucous
glands in different regions of the lung were measured microscopically. The magni-
tude of these changes was dependent on the dose of sulfur dioxide and cigarette
smoke. Four types of acid glycoproteins are produced by mucous glands and can be
measured biochemically. The proportion of cells producing a given type of glycopro-
tein changes in chronic bronchitis. This rat model has been used to study the toxicity
of sulfur dioxide and cigarette smoke, but not with a view toward its potential value
as an early indicator of bronchitis-producing agents. A 6-week exposure to sulfur
dioxide was indicative of progressive changes in chronic bronchitis. Given the time-
and dose-dependency of the alterations in mucous glands, more investigations
would be useful, particularly ones that correlate such biochemical changes with
pulmonary function alterations.
DETECTION OF LUNG FIBROSIS
Many industrial processes produce dusts that could be harmful if inhaled. The
pneumoconioses, parenchymal lung diseases caused by inorganic dust inhalation,
are a class of diseases characterized by massive fibrosis in their advanced states. Per-
haps the most studied form is silicosis, particularly with respect to potential early in-
dicators of chronic disease (Reiser and Last, 1979; Ziskind et al., 1976). Silicosis is a
general class of diseases produced from the inhalation of dusts predominantly
associated with mining, mineral refining and processing, but which are also found in
a wide variety of urban situations. With the recognition that pulmonary disease can
arise from the inhalation of such dusts, man-made substitutes have been sought, and
alterations have been made in industrial processes to reduce the exposure to, and the
toxicity of, dusts. Because many different types of dusts can be generated, a rapid
screen of these materials for their biological activity would provide producer, con-
sumer, and regulator with an important tool.
Silicosis results from chronic exposure to free crystalline silica (Ziskind et al.,
1976). The occupational form most studied arises from the inhalation of particles
less than 10 jum in diameter, with the assumption that smaller particles of less than 3
/um are responsible for the lesions. Environmental silicosis, if it exists, may result
191
-------
from the inhalation of free silica particles of 2.5 to 15 jum mass median aerodynamic
diameter. Three forms of silicosis have been characterized in humans: Chronic,
caused by moderate exposure levels for 20 to 40 years; accelerated, caused by higher
exposure levels for 5 to 15 years; and diffuse, from a still higher level of exposure for
less than 5 years (Ziskind et al., 1976). The years given are the time to appearance of
silicotic nodules on chest roentgenograms. The classical silicotic nodules can be
produced in rats after a 1-year exposure. In the chronic form in humans, lung
nodules have concentric rings of connective tissue around a zone of connective tissue
with interspersed free silica particles 1 to 2 /^m in diameter. Massive fibrotic lesions
composed of nodules of fibrosis with destroyed blood vessels and bronchi occur in
the upper lobes of the lung. Complications with rheumatoid arthritis,
emphysematous changes, and tuberculosis can also occur. Accelerated silicosis has
similar pathological changes, but progression is more rapid, and massive fibrosis
occurs more frequently. Acute silicosis is infrequently observed, and nodules are not
grossly observable. Lung consolidation does occur, as do edema and inflammatory
reactions.
Diagnosis of human silicosis is based upon exposure history and roentgeno-
graphic evidence. Although adequate for clinical diagnosis, such procedures are far
too insensitive to consider for short-term tests, especially in rodents. Pulmonary
function has been measured in human silicosis, but because of the complexity of the
disease, which is often accompanied by infection, bronchitis, emphysema, and
smoking, changes unique to silicosis have not been observed (Ziskind et al., 1976). In
controlled animal studies, these confounding variables are absent, but adequate pul-
monary function tests have not yet been reported. In a review of animal models for
silicosis, Reiser and Last (1979) point out that factors such as the presence of lung
pathogens and the route of administration (intratracheal versus inhalation) can have
a great influence on whether the animal model mimics the human disease (Holt,
1974; Zaidi, 1969).
Until there is an understanding of the mechanism(s) of the pathogenesis of sili-
cosis, the development of short-term tests will be impeded. Seven major events in the
pathogenesis of silicosis have been described (Ziskind et al., 1976): (1) phagocytosis
of the particles by alveolar macrophages; (2) death of the macrophages that have in-
gested the silica particles; (3) release of the intracellular contents of the injured and
dying macrophages, including silica particles and enzymes; (4) phagocytosis of re-
leased silica by other macrophages, followed by the death of these macrophages; (5)
accumulation of silica in other lung cells; (6) increased production of collagen fibers
in affected areas; and (7) hyalinization. The cytotoxicity of silica is an essential com-
ponent of these processes, and, centering on this, in vitro models have been proposed
and are described in the following section. Collagen production, being the integral
part of fibrosis, has also been studied and is discussed in the section on metabolism.
Autoimmune phenomena have been observed in the development of silicosis and
have been reviewed by Ziskind et al. (1976) and by Reiser and Last (1979). Individ-
uals with silicosis have an increased amount of serum autoantibodies, i.e., rheuma-
toid factor, antinuclear antibody and gamma globulins, as well as an increased pre-
valence of autoimmune disease. Silicotic lesions contain antibody-producing cells
and immunoglobulins. Animals injected with silica and antigens have far higher
specific antibody liters than animals injected with antigen alone. This adjutant effect
does not occur with relatively nontoxic dusts like kaolin and coal. Although specific
lung autoantibodies have not been observed in humans with silicosis, autoantibodies
that reacted primarily with collagen have been found in patients with various
chronic lung diseases. In vitro studies have shown that lung connective tissue anti-
bodies can stimulate alveolar macrophages to secrete a factor that stimulates col-
lagen production in fibroblasts. When properly activated, T-lymphocytesand other
cell types also produce these factors. The immunological involvement is interesting
because simple blood tests may have potential as screening methods.
192
-------
THE USE OF PULMONARY LAVAGE TO DETECT LUNG
DAMAGE
The lungs of humans and animals can be washed out or lavaged with various solu-
tions and the resultant fluid analyzed for alterations in fluid and cellular compo-
nents. Presently, this technique is not sufficiently validated as a short-term test for
chronic disease in animals, but several research groups (Hunninghake et al., 1979;
Henderson, 1982), especially Henderson and coworkers, are conducting investiga-
tions in this promising area.
Several protein constituents of normal human lavage fluid are listed in Table 1
(Hunninghake et al., 1979). The proteases and antiproteases may prevent the devel-
opment of emphysema and other destructive lung diseases. Table 2 depicts several
parameters in lavage fluid that may be indicators of lung damage (Henderson, 1982).
Measured changes in these parameters at any arbitrary point during exposure do not
imply that chronic disease will be the outcome. For example, elevated protein in lung
lavage, which is commonly assumed to reflect pulmonary edema, is a classical acute
response to nitrogen dioxide and ozone exposure (Selgrade et al., 1981; Hu et al.,
1982). This lavage technique has not been applied after chronic exposures to nitro-
gen dioxide and ozone, and the magnitude of protein exudation has not yet been cor-
related with the extent of chronic disease in either case.
Lavage analyses of patients with chronic lung diseases, such as alveolar protein-
osis or cystic fibrosis, have been correlated with their respective disease states. Al-
though no environmental component has been implicated in these human diseases, a
promising diagnostic role for pulmonary lavage fluid is developing. For example,
Kubich et al. (1980) are pursuing immunological methods to detect developing em-
physema. Studies by Henderson et al. (1978) and Pickrell et al. (1981) of the balance
between proteolytic and antiproteolytic activity in lavage fluid may also lead to early
tests for emphysema. Analogous experiments in animals could provide a screen for
early lung injury likely to produce COLD.
Analysis of the cellular content of lavage fluid is also important. In both humans
(Hunninghake et al., 1979) and animals (Gardner, 1982a), the predominant (more
than 90 percent) cell type is the alveolar macrophage. The remaining cell types are
lymphocytes, polymorphonuclear leukocytes, eosinophils, and basophils. Changes
in the numbers and functions of these cells can reflect lung damage and be part of the
pathogenesis of disease. Table 3 shows most of the functions of alveolar
macrophages in humans. The alveolar macrophages of animals perform similar
functions (Brain et al., 1977). Probably their most important activity is to phago-
cytize and kill microbes, thereby maintaining the sterility of the lung. Microbiocidal
activity in macrophages is inhibited by acute pollutant exposure (Gardner and
Graham, 1977). Other major roles for macrophages are as effector and accessory
cells in inflammatory and immune responses. Their involvement in emphysema is
proposed by virtue of their ai-antiprotease content (Hunninghake et al., 1979).
Human lavage studies of smokers show an increased number of inflammatory and
immune effector cells, and morphological studies have located these cell accumu-
lations at sites of lung destruction (Hunninghake etal., 1979). Smokers have an un-
usually high proportion of polymorphonuclear leukocytes in the lavage fluid.
Smokers' alveolar macrophages (also present in larger numbers) are activated, and,
in contrast to macrophages from nonsmokers, they secrete elastase that stimulates
the influx of polymorphonuclear leukocytes and the secretion of proteases. In ad-
dition, antiproteases, although found in normal quantity, do not function as well in
smokers as in nonsmokers.
EXPOSURE ASSAYS OF LUNG CELLS IN VITRO
Exposure of cells in vitro provides the most easily performed short-term test. Cell
lines of Type II alveolar cells and lung fibroblasts are readily available; primary cul-
193
-------
Table 1. Inflammatory and Immune Related Protein Constituents of Nor-
mal Human Alveolar Epithelial Fluid*
Components
Status in Normal Lung
Immunoglobulins
IgG
IgM
IgA
Monomeric
Dimeric
Free secretory piece
igE
igD
Complement
Classical pathway
Alternate pathway
Common pathway
Antiproteases
or, -antiproteinase
cr2-macroglobulin
Anti-chymotrypsin
Low-molecular-weight
bronchial inhibitor
Enzymes thought to be
capable of deranging
alveolar structures
Collagenase
Elastase
Neutral protease
Other enzymes
Lysozyme
/3-glucuronidase
Esterase
Acid protease
Phospholipase A
Glycosidase
Present
Absent
Present
9% of total
91% of total
Present
Present
Functional
Functional
Functional
Present
Very low if present
? (present in bronchial mucus)
? (present in bronchial mucus)
Absent
Absent
Present
Present
Present
Present
Present
Present
Present
*From Hunninghake et al. (1979)
tures of alveolar macrophages are simple to obtain by pulmonary lavage; tracheal
rings, lung tissue slices, and tracheal explants can be maintained in vitro for several
days. Although these cells are altered in chronic disease states, more investigation is
required before in vitro exposures can be applied widely as short-term tests for
chronic lung disease.
Macrophages
Investigations of the mechanisms of silicosis led to the development of model sys-
tems in which alveolar or peritoneal macrophages are exposed in vitro to silica, and
the resulting cytotoxicity is measured morphologically, histochemically, or bio-
chemically. Other particles have been tested, and macrophage cytotoxicity has been
reasonably well correlated with fibrogenicity of fibers (Reiser and Last, 1979;
194
-------
Table 2. Indicators of Injury in Lung Lavage Fluids*
Endpoint
Site
Possible Indication
if Altered
Lactate dehydrogenase
Glucose-6-phosphate
Acid phosphatase
and
/J-glucuronidase
Alkaline phosphatase
Glutathione peroxidase
and
Glutathione reductase
Total protein
Sialic acid
Phagocytic cells
PMNs
Macrophages
Cytosol (glycolysis)
Cytosol (hexose mono-
phosphate shunt)
Lysosomes
Plasma membranes
Type II cell lamellar
bodies
Serum
Cytosol
Extracellular
Mucous glycoproteins
Cell damage
1. Cell damage
2. Leakage from cells
undergoing repair
1. Release during phagocytosis
2. PMN and/or macrophage
damage
1. Type II cell damage or
increased secretions
2. Transudation of serum-
proteins
Protection mechanism activated
against lipoperoxidation
Transudation of proteins a'cross
alveolar-capillary barrier
1. Increased mucus secretion
2. Transudation of serum glyco-
proteins
Acute inflammation
Persistent inflammation
*From Henderson (1982)
Chamberlain et al., 1979). However, this correlation does not hold for various forms
of asbestos (Allison, 1971).
One research group (Allison, 1971, 1975; Nash et al., 1966; Allison and Davies,
1974) studied the interaction of silica and macrophages extensively. The silica par-
ticles were found to have two major types of effects: (1) at high doses, rapid cyto-
toxicity occurs, characterized by release of lysosomal and cytoplasmic enzymes into
the medium; and (2) at lower doses, toxicity is delayed for several hours. Silica is
taken into phagosomes, and lysosomes cluster around and empty their contents into
the phagosomes. A few hours later, signs of cell damage are seen, probably due to the
interaction of silica with the secondary lysosomal membrane, which makes it perme-
able to lysosomal hydrolytic enzymes. The immediate high dose effects are thought
to be characteristic of particle interactions with cell plasma membranes. The delayed
cytotoxicity appears to be a function of the interaction of hydrogen bonding
between the surface of silica and the membranes of the secondary lysosomes. Silico-
sis is only generally associated with exposure to these active forms of silica.
Red Blood Cell Hemolysis
Research on the mechanisms of interactions between toxic particles and cell mem-
branes led to the development of hemolysis assays. Mixing the fibers with a suspen-
sion of erythrocytes results in hemolysis. The hemolysis itself is unrelated to dust dis-
eases but it is a relatively easy way to study the interaction of chemicals with plasma
membranes. Most investigations in this area have focused on silica, asbestos, and re-
195
-------
Table 3. Properties and Functions of Normal Human Alveolar Macro-
phages*
Origin
Derived from bone marrow
Capable of self-replication
Surface receptors
Fc-IgG
C3B
C3d
Proteases-antiproteases
Interactions with microorganisms
Bacteria
Fungi
Viruses
Interactions with noninfectious particulates
Phagocytosis of particulates
Pinocytosis
Metabolism of ingested compounds
Effector and accessory cell in inflammatory and immune reactions
Responds to lymphokines
Responds to chemotactic stimuli
Regulates lymphocyte response to mitogens and antigens
Secretes
Colony-stimulating factor
Chemotactic factor for neutrophils
Superoxide anion
Platelet-activating factor
Elastase
Collagenase
Neutral protease
Mediates antibody-dependent cellular cytotoxicy
Protection of alveolar structures from protease attack
Contains a,—antiproteinase
Cytosol inhibits human neutrophil elastase
Ingests neutrophil elastase
*From Hunninghake et al. (1979)
lated substances. For silica and asbestos, the correlation between macrophage cyto-
toxicity and hemolytic activity is good (Harington et al., 1971).
The hemolytic activity of forms of silica correlates well with fibrogenic activity
(Allison, 1971; Harington et al.,1971). However, for forms of asbestos, the corre-
lation is less clear. For example, chrysotile, crocidolite, and amosite are very fibro-
genic when inhaled, but the latter two do not significantly affect hemolysis. Ben-
tonite and kaolin, nonfibrogenic compounds, are also more hemolytic than cristo-
balite and quartz, figrogenic compounds (O'Hery and Gornley, 1978). Various tech-
nical factors, such as pH, species of animal donating the erythrocytes, time of incu-
bation, and the precise physicochemical properties of the fiber, can affect results and
make correlations between different laboratories difficult (Harington et al., 1971;
Schnitzer and Pundsack, 1970; Palekar et al., 1979; Light and Wei, 1977;
Summerton et al., 1977).
196
-------
Although these hemolysis studies may, therefore, lead to a better understanding of
plasma membrane-fiber interactions, good correlations between silicosis and hemo-
lytic potency have not been reported.
Fibroblasts
Macrophages can stimulate collagen biosynthesis by fibroblasts, a major step in
the development of fibrosis. Many of the studies of macrophage activity were con-
ducted with silica (Reiser and Last, 1979). Briefly, when alveolar and peritoneal
macrophages are incubated with quartz, the medium supernate or macrophage
lysate will stimulate the synthesis or release of collagen from fibroblasts. The direct
addition of quartz to the fibroblasts does not cause the same response, but activated
macrophages without quartz can also stimulate fibroblasts. Complete interpretation
of these studies is again confounded by differences in technique.
Collagen and prolyl hydroxylase (the rate-limiting enzyme in collagen biosyn-
thesis) were examined in dogs 2-1/2 to 3 years postexposure to irradiated auto ex-
haust as an indicator of fibrosis (Orthoefer et al., 1976). The collagen content (hy-
droxyproline) was unchanged. Prolyl hydroxylase was increased in one group
having emphysema, but was unchanged in another group similarly diseased. Thus,
collagen biochemistry as studied here was insensitive. However, given the impor-
tance of collagen turnover in chronic lung disease and the complexity of collagen
biochemistry, it should not be dismissed without further research.
Tracheal Epithelium and Mucus
Several investigators have successfully correlated the in vitro and in vivo acute ex-
posure effects of cadmium, nickel, fly ash, sulfuric acid, and carbon on ciliary
beating frequency of tracheal rings (Adalis et al., 1977, 1978; Schiff et al., 1978,
1981). However, no attempts have been made to correlate the in vitro acute studies
with chronic in vivo experiments.
Perhaps a more promising area for short-term tests of chronic effects is in vitro
models of mucus biosynthesis, since mucus production is excessive in chronic bron-
chitis. Tracheas are incubated with labeled precursors of mucous glycoproteins, and
the incorporation of precursors and production of glycoproteins are measured (Last
and Kaizu, 1980; Sturgess and Reid, 1972; Reid and Jones, 1980). Humans, rats, and
monkeys are suitable donors. Mucus-secreting cells have been isolated from ham-
sters and cultured; these cells were capable of mucous glycoprotein synthesis
(Baseman et al., 1980). Most of these studies have been concerned with in vivo pol-
lutant exposure, but investigations of secretory mechanisms have been made in vitro
with pharmacological agents, and in vitro pollutant exposures may be of interest.
Tissue Slices
The use of lung tissue slices in toxicology has been revei wed by O'Neil and Young
(1982) and Freeman and O'Neil (1982). The sample preparation is relatively simple
and has been applied to studies of lung metabolism, particularly of phospholipids
and carbohydrates. A significant feature of this method is that because several cell
types are present, cell-cell interactions proceed. The major limitation is that normal
metabolism continues for only about 5 hours with current techniques, and the
medium is poorly mixed in the airways of the slices, so kinetic studies are difficult.
No studies have been undertaken to examine the correlation between the effects ob-
served with lung slices and chronic exposure.
CHEMICAL CARCINOGENESIS BY TRACHEAL TRANSPLANTS
Our discussion of short-term chemical carcinogenesis assays will be limited to the
tracheal transplant and the strain A mouse pulmonary tumor bioassay. In vitro car-
cinogenesis assays are presented by Upton (Chapter 3).
197
-------
Carcinogenesis in the respiratory tract is known to be induced by exogenous
factors such as cigarette smoke (Hammond, 1966; Wynder et al., 1970) and occupa-
tional or environmental exposure (Frank, 1978; Kraybill, 1977). The complex inter-
actions of the precise components of these exposures render the determination of
direct causative relationships a difficult task. In addition, various endogenous
factors are possibly involved in the induction of cancer. Some individuals may be
predisposed to the development of neoplasms due to physiological, nutritional, or
disease states (Templeton, 1975) or to genetic abnormalities (Mulvihill, 1975) that
could affect functions such as drug metabolism (Nebert et al., 1978; Paigen et al.,
1978).
One of the principal methods used in studying the etiology of bronchogenic cancer
is the tracheal transplant. Briefly, the trachea is removed from a donor animal and
transplanted in its entirety into the deep subcutaneous connective tissue of an iso-
genic recipient. After several weeks, essentially no histological difference is discern-
ible between the transplanted tissue and the host tissue. Pellets containing a carcin-
ogen are inserted into the grafted trachea to induce tumor production. Tracheal
transplants have been successful in rats, mice, and hamsters (Kendrick et al., 1974).
Initial studies with the tracheal transplant system used gelatin pellets containing the
carcinogen (Kendrick et al., 1974); however, several disadvantages were discovered.
The gelatin pellet inserted into the tracheal lumen dissolved within 10 minutes. Thus,
the carcinogen was released rapidly and moved to the proximal end of the transplant
where most tumors develop. The problem has been eliminated by using beeswax
pellets, which do not dissolve in vj'voand which allow a more even distribution of the
carcinogen throughout the tracheal epithelium. This method also permits better
control over the release of the carcinogen and the length of exposure (Griesemer et
al., 1977).
The tracheal transplant model is a valuable tool in determining the relationship
between age and susceptibility to cancer. For example, by using older animals as
donors and young adults as recipients, age-specific characteristics can be studied and
confounding factors associated with old age can be eliminated. This model is also
useful in studying the immune system's effect on the development of squamous car-
cinomas in that the host can be immunosuppressed without exposing the target
tissue (Nettesheim et al., 1981). One criticism of the procedure has been its use of
tracheal tissue rather than bronchial tissue. The trachea in rodents, however, is
similar to the bronchus in humans in that it contains cartilage, compound mucous
glands, and goblet cells. No other segment of the respiratory tract in rodents has such
a resemblance to the human bronchus (Kendrick et al., 1974).
The tracheal transplant system has several other advantages, the primary one
being that known quantities of a carcinogen can be administered to a specific tissue
for a designated period of time. This kind of control is not possible with inhalation
exposure or with intratracheal instillation (Nettesheim and Griesemer, 1978). In ad-
dition, the transplanted tissue can be easily inspected for tumor formation; vital
functions of the host continue with no interference from the carcinogen or from
developing tumors; and the target tissue and the host can be controlled independent-
ly, allowing the study of endogenous and exogenous factors affecting tumor devel-
opment (Kendrick et al., 1974).
Using the tracheal transplant method, Topping et al. (1978) studied a range of
polycyclic hydrocarbons (PCHs) both in vitro and in vivo and compared the release
rates of the compounds from beeswax pellets. The results indicated that in vitro re-
lease does not generally predict in vivo release. In vitro release was faster, implying a
slow exchange from the tracheal lumen to the circulation in the transplant. Also, in
vivo release was more variable, probably because changes occur in the tracheal wall
that affect rates of exchange and release. Later studies by the same group decreased
the variability by altering the pellet matrix (Pal et al., 1978). The results of Topping
et al. (1978) showed that a range of PCHs, including noncarcinogens, initiators, and
198
-------
carcinogens, cause significant changes in the epithelial and mesenchymal pathology
of respiratory tract mucosa. The toxic response caused by noncarcinogens was short-
lived. Carcinogens caused long-lasting responses characterized by altered epithelial
differentiation (metaplastic changes). The severity of response to a carcinogen corre-
sponded directly to its potency.
Nettesheim et al. (1977) used the tracheal transplant system to study the induction
of preneoplastic and neoplastic lesions produced by benzo(a)pyrene, B(a)P, deliv-
ered from beeswax pellets. Exposure time ranged from 1 to 6 months, and cumula-
tive doses ranged from 10 to 2,490 ug. The five highest B(a)P doses (900 to 2,490 /ug
per pellet) had a release rate constant of about 2.8±0.6 percent per day for the first 2
months. Since the amount released is determined by the amount in the pellet, more
was delivered early in the exposure. The duration of the exposure was determined by
the B(a)P dose, and exposures ranged from 0.5 to 6 months. At the three lowest
doses (10 to 300 /ug per pellet), most of the B(a)P was released in 1 to 2 weeks
(Nettesheim et al., 1977). Tracheal grafts receiving 3000 /ug B(a)P in gelatin pellets re-
tained about 30 percent of the dose 3 weeks after instillation of the carcinogen
(Kendrick et al., 1974). With B(a)P doses ranging from 480 to 2,490 ug per pellet,
massive hyperplasia developed in the first few weeks. At 1 month, atrophic-pleo-
morphic lesions were seen, with the reappearance of normal mucociliary epithelium.
This change might indicate the appearance of cells that are resistant to the toxic ef-
fects of B(a)P. A significant number of squamous metaplasias did not develop until 4
to 6 months for any exposure level, and the lesions were only focal; such lesions may
be early preneoplasia, since some of them appeared months after carcinogen release
had stopped (Nettesheim et al., 1977).
The first 2 weeks of exposure to tumorigenic doses of B(a)P produced a consistent
hyperplastic reaction throughout the tracheal graft, thus indicating a uniform ex-
posure. Later in the experiment, the response within the same trachea varied greatly.
The implication is that a heterogeneous cell population like the tracheal epithelium
has a variety of responses to a carcinogen, in spite of a uniform exposure (Nettesheim
et al., 1977). Neoplasms were induced in 10 months or less by B(a)P doses of 900 /ug.
In many cases, noninvasive and invasive carcinomas were seen in the same graft.
Dysplasias were most often found in tracheas with early neoplastic lesions. The pre-
neoplastic and preinvasive neoplastic lesions seen in the tracheal transplant are sug-
gestive of lesions in the human larnyx and bronchi (Nettesheim et al., 1977).
Grafted rat tracheas have also been used to determine the effects of 7,12-dimethyl-
benz(a)anthrancene (DMBA). DMBA doses ranged from 0 to 3,200 ug per pellet
(beeswax), and about 1.7 percent of the carcinogen remaining in the pellet was re-
leased per day (Griesemer et al., 1977). At levels of 325 to 3,200 /ug DMBA per pellet,
all grafts showed the same degree of hyperplasia, metaplasia, keratinization, and cel-
lular atypia. After 1 month, the first microinvasive lesions were seen, and at 2
months, tumor nodules were present. All cases were identified as invasive squamous
cell carcinomas, but in some cases of high dose exposure, sarcomas developed with
the carcinomas. At levels of 10 to 210 ug DMBA per pellet, lesions varied both be-
tween and within dose groups. Preneoplastic and neoplastic focal lesions developed.
A22-month tumorigenesis study at these lower doses showed that as the level of car-
cinogen decreased, tumor incidence decreased and tumor latency increased. The low-
est dose of DMBA to produce tumors with the tracheal transplant system was be-
tween 40 and 10 /ug (Griesemer et al., 1977).
Cocarcinogenic activity has been studied primarily using skin as the experimental
tissue (Van Duuren, 1969; Van Duuren et al., 1971). Inhalation studies of gaseous air
pollutants have shown no definite cocarcinogenic effect, with the possible exception
of sulfur dioxide (Nettesheim et al., 1981). Rats exposed to B(a)P aerosols and 10
ppm sulfur dioxide developed a 20-percent lung tumor incidence, but only a 3-per-
cent incidence was observed in animals exposed to B(a)P alone (Laskinetal., 1970).
Topping et al. (1981) studied the effects of benzo(e)pyrene, B(e)P, in combination
199
-------
with B(a)P on tracheal mucosa of rats; both B(e)P and B(a)P are found in tobacco
smoke and exhaust from gasoline engines. Previously, B(a)P was found to be cocar-
cinogenic for mouse skin (Van Duuren and Goldschmidt, 1976). The tracheal trans-
plant system was selected by Topping et al. (1981), because this model allows control
of exposure to the carcinogen and the suspected carcinogen and of targeting to a
specific segment of respiratory tract mucosa. Data showed no carcinogenicity in the
tracheal epithelium associated with B(e)P, but significant cocarcinogenicity in the
tracheal-peritracheal connective tissue. These results, in combination with the pre-
vious findings using mouse skin (Van Duuren and Goldschmidt, 1976), indicate that
the target tissue type is a major determinant of the cocarcinogenicity of B(e)P.
The strain A mouse pulmonary tumor bioassay has also been studied as a short-
term test for predicting potential chemical carcinogenicity (Shimkin and Stoner,
1975). Maronpot et al. (1982) used this bioassay in experiments with 60 chemicals.
These results were then compared with those of a 2-year rodent bioassay. Available
data using the 2-year test were adequate for only 54 of the chemicals, so 6 of those
used with the mouse bioassay were excluded from comparative analysis. Overall, the
results lacked agreement, including a high percent of false negative results with the
strain A mouse test. Possible reasons for the dissimilarities include differences in:
pharmacokinetics and metabolism by species or strains, routes of chemical adminis-
tration, adminstered dose, or duration of exposure to the chemicals. Another factor
might be the precise determination of what constitutes a positive result; judgments
are frequently difficult with the criteria presently used. Although the results seem to
question seriously the validity of the strain A mouse pulmonary tumor bioassay,
several points require attention. First, the 2-year bioassay has not been conclusively
validated, and the strain A bioassay may actually be more predictive in some cases.
In addition, validation must take into account considerations such as the chemical
metabolism in each bioassay model, the specific organ responding to the carcinogen,
the extent to which the response is positive, the sensitivity of the bioassay, and
genetics. Any conclusions regarding the use of the strain A mouse system in predic-
ting carcinogenicity cannot be reached until more data are available.
BIOCHEMICAL STUDIES USING_EX VIVO TECHNIQUES
In the last 5 years, considerable strides have been made in the use of isolated, ven-
tilated, and perfused lung (IVPL) preparations in toxicological studies (Smith and
Bend, 1981). The precise methodology varies between laboratories, but all apply the
same principle of ventilation occurring simultaneously with perfusion. Ventilating
the lung is essential to ensure adequate perfusion. The mechanical movement of the
lung seems also to stimulate prostaglandin production and may influence a number
of other metabolic steps as well. Usually, hemoglobin-free solutions, such as Krebs-
Henseleit, supplemented with serum albumin are used for the perfusion, and re-
quired rates entail the use of large volumes of solution even for small lungs, such as
those from rats. Perfusion rates of less than 2 ml per minute are not generally satis-
factory, and rates of 10 ml per minute or greaterare often used with rat lungs. Blood
has also been used as a general medium, but this procedure suffers from the difficul-
ties of requiring several donor rats, being highly complex and variable, being con-
taminated easily by prior treatment of the donor rats, and being subject to clotting
and hemolysis. Direct measurement of product formation in blood is also difficult,
although it is easily accomplished in hemoglobin-free media (Williams and Menzel,
1982). Because of overlapping absorption spectra, direct measurement of the oxida-
tion-reduction state of hemoproteins in the lung is not possible in blood-perfused
preparations.
The major driving force for the use of IVPL preparations is the very complexity of
the lung itself. Only fibroblasts and type II pneumocytes are now cultured easily,
with promise that Clara cells may soon become available. Primary cultures of the re-
200
-------
maining cells are difficult to produce with any purity, although strides have been
made in the sorting of separated cells from the lung. Still, cell separation by differ-
ential centrifugation, elutriation, or cell sorting remains cumbersome and unpro-
ductive. The likelihood that for the next 5 to 10 years these methods will yield suf-
ficient numbers of cells for routine short-term tests is low. Given proper perfusion
medium and careful control of the physiological parameters of the preparation, the
interrelationships between the multiple cell types of the lung can be preserved. The
IVPL, then, becomes a potential tool for the short-term evaluation of pneumotoxi-
cants.
Although there is no widely accepted method for the evaluation of pneumotoxi-
cants using the IVPL, this preparation is easily adapted to studies of the metabolism
of xenobiotic compounds (Williams et al., 1982). Covalent reactions between B(a)P
and lung macromolecules can be observed by perfusion with 3H-B(a)P, homogeniza-
tion of the lung, separation of lung macromolecules by precipitation and exhaustive
washing, and measurement of labeled covalently reacted B(a)P metabolites (Williams
et al., 1982). If covalent reaction with cellular macromolecules is presumed to lead to
either mutagenesis, carcinogenesis, or toxicity, these methods could be used to rank
chemicals quickly.
TESTS OF PARTICLE REMOVAL FROM THE LUNG
Although the primary function of the lung is to ensure an efficient exchange of
gases, this organ also provides the body with a primary protective barrier against air-
borne assaults from both viable and nonviable agents. Any breach in this defense sys-
tem would be expected to increase the host's risk of disease (Green et al., 1977;
Gardner, 1979; Proctor, 1979; Brain et al., 1977; Sanders et al., 1977; Brain and
Valberg, 1979). It is readily evident that any dysfunction can lead to disease, espe-
cially pulmonary infections, bronchial cancer, chronic bronchitis, bronchial asthma,
pneumoconioses, and possibly pulmonary emphysema (Proctor, 1979). It remains
to be determined whether or not the tests that are now available for evaluating the
adequacy of respiratory defenses are useful to detect or screen for disease suscepti-
bility, to note its inception, to study it pathogenesis, and to improve our under-
standing of the relationship between morphology and function.
The tests of host defenses that have been most widely used for examining problems
associated with infectious lung disease include measurements of mucociliary clear-
ance, functioning of alveolar macrophages, and susceptibility of experimentally in-
duced infections. The effectiveness of the mucociliary escalator can be determined
by measuring the rate of transport, the frequency of ciliary beating, and the size and
distribution of mucus-secreting cells (Bang and Bang, 1977). Any adverse effect
would be likely to reduce the amount of inhaled and deposited substances that could
be removed from the respiratory tract. Numerous gases and particles have been
shown to impair host defenses through the depression of the mucociliary escalator
(Adalis et al., 1978; Grose et al., 1980; Wolff et al., 1975; Schiff et al., 1979).
Another group of investigators (Lippmann et al., 1982; Schlesinger et al., 1978,
1979) investigated the effects of inhaled pollutants on bronchial airway function in
donkeys and humans. After various periods of exposure, the subjects inhaled an
inert test aerosol ("TC-FezOs), and tracheobronchial mucociliary clearance of the
particles was measured. Mucociliary clearance is an important defense mechanism
of the lung and is usually altered in chronic bronchitis. Generally, the same sequence
of events on mucociliary bronchial clearance was produced by sulfuric acid and
cigarette smoke after both short-term and chronic exposures in donkeys. The short-
term exposure studies of donkeys and humans were correlated for both pollutants.
Given the chronic bronchitis produced by cigarette smoke and the results of their
studies, these investigators believe that the observations in donkeys may be related
to chronic bronchitis. From this interpretation, a potential short-term animal model
201
-------
might exist. However, the technique itself is more amenable to a research situation,
as opposed to being applied to a screening test.
Within the. lung, the resident population of alveolar macrophages is a crucial de-
fense because these cells are responsible for a variety of important activities, includ-
ing the detoxification and removal of deposited airborne particles; maintaining pul-
monary sterility against invading microorganisms; interacting with lymphocytes for
immunological function; as well as acting as scavengers in removing various cellular
debris. In order to fulfill their purpose adequately, these cells must maintain an inte-
grated membrane, mobility, and phagocytic activity, and have a well-developed and
functioning enzyme system (Gardner, 1982a). A number of tests have been developed
to evaluate the "health" of these cells after either in vivo or in vitro exposure (Waters
et al., 1975; Agostra et al., 1975; Aranyi et al., 1981; Brain, 1977; Coffin and
Gardner, 1972; Goldstein, 1977; Graham etal., 1975). Parameters of number, stabil-
ity, viability, morphology, function, biochemistry, and metabolism have been
studied; the most useful measurements are those of functional capacity, such as
phagocytosis and bacteriocidal activity.
Primary cultures of alveolar macrophages have been recommended recently for
use in an early stage cytotoxicity assay to obtain preliminary environmental assess-
rnent information, to identify problem areas, and to prioritize testing procedures
(Duke et al., 1977). Because the macrophage represents the first line of pulmonary
defense, these cells are appropriately used to define and to "rank" the acute cellular
toxicity of airborne chemicals (Aranyi et al., 1981; Waters et al., 1975). The para-
meters most generally tested include viability, ATP, and phagocytic activity. Al-
though these test systems have been utilized primarily to evaluate acute toxicity, the
cells may also provide a key to understanding some chronic pulmonary effects that
could be attributed to the action of toxic chemicals on these defense cells. For ex-
ample, as a result of cell injury, cell death, or selective exocytosis, macrophages re-
lease into the lung a number of potent and biologically active proteolytic enzymes,
such as elastase and collagenase, that are thought to be contributing factors in the
pathogenesis of chronic lung disease (Gardner, 1982a; Janoff et al., 1977; Brain,
1980).
Because any suppression in the host's body defenses by environmental chemicals is
expected to result in prolonged microbial viability and enhanced multiplication of
the infectious organism, a number of model systems have been developed to deter-
mine the degree of efficiency of the total pulmonary defense system in a compro-
mised host. The most successful model is to introduce a viable, opportunistic
pathogen within the respiratory tract of the test subject. If the host's defenses are
functioning normally, the lungs will return to a sterile condition within a few hours.
However, if the host's defenses have been altered to a degree that permits the infec-
tious agent to establish itself and to multiply, a dysfunction of the defense system is
then easily recognized (Coffin and Gardner, 1972; Gardner, 1982b). This infectious
model system has been utilized to identify the effects of pollutants on acute pulmo-
nary infections due to Streptococcus, Diplococcus, KlebsieUa, and influenza virus
(Ehrlich, 1980; Gardner, 1981; Ehrlich and Fenters, 1973; Maigetter et al., 1976).
SUMMARY
At present, no short-term tests for chronic pulmonary disease have an acceptable
false negative error rate. In attempting the development of adequate tests, several
factors must be considered. A more rapid means to detect chronic lung disease is
desirable, but it must detect all potential major disease classes. A rapid test for only
fibrosis would not be satisfactory, because chronic studies would still be needed to
determine if the chemical caused other diseases like emphysema. In the chronic study
in search of emphysema, the presence of other diseases, like fibrosis, would be de-
tected. Thus, models for all major lung diseases must be pursued concurrently.
202
-------
The most promising short-term tests for chronic diseases would measure early
events in pathogenesis, like alterations of DN A or cellular transformation in tests for
mutagenesis and carcinogenesis. The isolated, ventilated, and perfused lung prepar-
ation is a new rapid assay system of particular promise. The ease with which cova-
lently reacted DNA, RNA, and proteins can be isolated from perfused lungs after
treatment with toxicants and the preservation of the physiological state of the lung
during measurement make this system particularly useful (Williams et al., 1982).
Toxicants can be administered rapidly by intratracheal instillation, ventilatory
gases, or aerosol. Particles are easily applied in liquid suspension. Radiolabeled, as
well as nonradiolabeled, compounds can be evaluated. Validation should be under-
taken by correlating short-term binding and metabolism with chronic toxicity.
Several early events for chronic lung disease are known, but when this knowledge
is applied to the development of model systems, correlations with chronic effects are
very weak. Perhaps part of the problem is that the available models, like collagen
synthesis and macrophage cytotoxicity, measure only a few of the total early events.
The more successful rapid screens for mutagenesis and carcinogenesis employ a bat-
tery of tests, each making a contribution in terms of detecting a single early event in
the disease process; together they cover the range of potential events reasonably ef-
fectively. To overcome this problem for pulmonary diseases will require mechanistic
studies on the pathogenesis of the major pulmonary diseases, so that a battery of
tests could be developed. To select just one test for each disease would probably lead
to excessive false positive error rates if the pathogenic event were too early in the dis-
ease process or to excessive false negative rates if the mechanism being screened were
an intermediate event.
Attempting to develop, refine, and validate a battery of short-term tests for
chronic lung disease is a vast undertaking. A typical chronic study takes 1 to 2 years
and is relatively expensive, because a large number of animals are used and extensive
examinations are performed. However, if the typical chronic study could be refined
to reduce both the time and cost of testing, then it might meet more of the criteria for
a good short-term test than would a series of in vitro studies. Refining and validating
a more rapid chronic study will require intense investigation. Shortening the length
of exposure, from even 12 to 6 months, will require the identification of early stages
of pathogenesis. Thus, more studies of mechanisms in animals are needed, both to
identify early stages and to correlate such changes with those in humans, to ensure
that the data will have value in human risk assessment. To reduce costs, the number
of endpoints examined must be reduced; mechanistic studies are also required to en-
sure that the needed endpoints remain in the protocol. In addition, improvements in
measurement techniques will reduce variability and, hence, lower the number of
animals required. Dose-rate considerations in toxicology can be complex, and the
potential recovery or progression of the effects when exposure ceases is important.
Usually, a high dose shortens the time required to observe an effect in inhalation
toxicology, but the level can be so high that additional mechanisms of toxicity which
are unrealistic will come into play. Thus, definition of a high dose is quite difficult
and might have to be arbitrarily assigned as some function of the LC5o. For some
compounds, even determining the LCso can be difficult because of engineering con-
siderations in generating the proper test atmosphere. Choosing the proper exposure
time is also critical. As currently defined, chronic diseases are just that—chronic; they
take some time to develop. If the exposure period is too short relative to the endpoint
for a particular stage in the pathogenesis of the disease, a false negative error will
occur. This question might be resolved by encouraging researchers on current or
projected long-term studies to make examinations at several times, especially when
newer techniques of pulmonary pathology, function, and biochemistry are being
used. The accumulated data can be used in determining validation procedures for
the short-term test. The validation would include chemicals of high, medium, and
low potency for inducing all major types of chronic lung disease.
203
-------
Despite the difficulties, short-term tests need to be developed for specific pulmo-
nary diseases. Inhalation is a major, efficient route of exposure to a wide variety of
chemicals. The effects of these chemicals on man are not now known precisely, and
thus the risk to the exposed population is also poorly quantified. The driving force
for short-term tests is real and urgent enough to encourage further development. Re-
search into basic mechanisms is paramount, for in this way more relevant tests can
be developed. Knowing the mechanisms of human disease, short-term in vitro tests
are more likely with cell, tissue, or organ culture. The rapid advances in lung cell and
organ culture are encouraging. A major parallel effort is underway to define the in-
haled dose to the target region of the lung. By comparing human and animal lungs,
the magnitude of the regional dose can be estimated and a better scale-up from
animals to humans can be determined. Although no quantum jumps in short-term
tests for pulmonary toxicity are on the horizon, steady progress is being made to as-
sure the applicability of animal test results to human disease processes, to reduce the
time needed to attain a clear indication of pulmonary disease in animals, and to
compare doses producing disease in animals with those required in man.
REFERENCES
Adalis, D., D.E. Gardner, and F.J. Miller. Cytotoxic effects of nickel on ciliated epithelium.
Am. Rev. Resp. Dis. 188:347-354, 1978.
Adalis, D., D.E. Gardner, F.G. Miller, and D.L. Coffin. Toxic effects of cadmium on ciliary
activity using a trachea! ring model system. Environ. Res. 75:111-120, 1977.
Agostra, P., C.A. Fisher, E.J. Kaminski, and J.C. Calendra. In vitro studies of rat pulmonary
alveolar macrophages subsequent to intrapulmonary instillation of lead. Toxicol. Appl.
Pharmacol. 55:187-188, 1975.
Alarie, Y. Sensory irritation by airborne chemicals. CRC Crit. Rev. Toxicol. 2:299-363, 1973.
Alarie, Y., I. Wakisaka, and S. Oka. Sensory irritation by sulfur dioxide and chlorobenzilidene
malononitrile. Environ. Physiol. Biochem. 5:53-64, 1973.
Allison, A.C. Lyposomes and the toxicity of particulate pollutants. Arch. Intern. Med. 128:
131-139, 1971.
Allison, A.C. Effects of silica, asbestos, and other pollutants on macrophages. In Air Pollution
and the Lung, edited by J.D. Brain, D.F. Proctor, and L.M. Reid, Israel Universities Press,
Jerusalem, pp. 114-132, 1975.
Allison, A.C., and P. Da vies. Mononuclear phagocyte activation in some pathological pro-
cesses. In Activation of Macrophages, edited by W.H. Wagner, and H. Harm, Excerpta
Medica, Amsterdam, pp. 141-156, 1974.
Amdur, M.O., and J. Mead. A method for studying the mechanical properties of the lungs of
unanesthetized animals. In Proceedings of the Third National Air Pollution Symposium.
National Air Pollution Symposium, Pasadena, CA, April 1955, pp. 150-159.
Mechanics of respirationin unanesthetized guinea pigs. Am.J. Physiol. 192:364-
368, 1958.'
Anderson, M.W., and T.E. Eling. Studies on the uptake, metabolism, and release of endo-
genous and exogenous chemicals by use of the isolated perfused lung. Environ. Health
Persp. 76:77-81, 1976.
Aranyi, C., D.E. Gardner, and J.J. Huisingh. Evaluation of potential inhalation hazard of
particulate silicous compounds by in vitro alveolar macrophages test application to indus-
trial particulates containing hazardous impurities. In Health Effects of Synthetic Silica
Particulates, ASTM STP 732, edited by D.D. Dunnam, American Society for Testing and
Materials, Philadelphia, pp. 48-61, 1981.
Auerbach, O., E.G. Hammond, D. Kirman, and L. Garfinkel. Effect of cigarette smoking on
dogs. Arch. Environ. Health 27:754-768, 1970.
Bang, E.G., and F.B. Bang. Nasal mucociliary systems. In Respiratory Defense Mechanisms,
edited by J.D. Brain, D.F. Proctor, and L.M. Reid, Marcel Dekker Inc New York pp
405-426, 1977.
Barry, B.E., F.J. Miller, and J.D. Crapo. Alveolar epithelial injury caused by inhalation of low
levels of ozone. Am. Rev. Resp. Dis. /25(II):148, 1982a.
Alveolar epithelial injury caused by inhalation of low levels of ozone. Adv.
Modern Environ. Toxicol., in press, 1982b.
204
-------
Baseman, J.B., N.S. Hayes, W.E. Goldman, and A.M. Collier. Mucus and surfactant synthesis
and secretion by cultured hamster respiratory cells. Environ. Health Persp. 35:139-146,
1980.
Becklake, M.R., and S. Permutt. Evaluation of tests of lung function for "screening" for early
detection of chronic obstructive lung disease. In The Lung in the Transition Between
Health and Disease, edited by P.T. Macklem, and S. Permutt, Marcel Dekker, Inc., New
York, pp. 345-387, 1979.
Bianco, A., F.R. Gibb, R.W. Kilpper, S. Landman, and P.E. Morrow. Studies of tantalum
dust in the lungs. Radial. 112:549-556, 1974.
Blair, W.J., N.S. Porter, D.P. Brown, and A.D. Wehner. Apparatus for direct inhalation of
cigarette smoke by dogs. J. Appl. Physiol. 26:847-850, 1969.
Brain, J.D. The respiratory tract and the environment. Environ. Health Persp. 20:113-126,
1977.
. Macrophage damage in relation to the pathogenesis of lung disease Environ
Health Persp. 35:21-28, 1980.
Brain, J.D., and P.A. Valberg. Deposition of aerosol in the respiratory tract. Am. Rev. Resp.
Dis. 720(6):1325-1373, 1979.
Brain, J.D., D.F. Proctor, and L.M. Reid, editors. Respiratory Defense Mechanisms, Parts I
and II, Marcel Dekker, Inc., New York, 1071 pp., 1977.
Brar, B.S., C.E. Traitor, C.R. Boshart, and J.F. Noble. A laboratory aerosol system foi con-
ducting inhalation toxicity studies. Toxicol. Appl. Pharmacol. 3.3(1): 134-135, 1975.
Chamberlain, M., R.C. Brown, R. Davies, and D.M. Griffiths. In vitro prediction of the patho-
genicity of mineral dusts. Br. J. Exp. Path. (50:320-327, 1979.
Charles, J.M., and D.B. Menzel. Influence of atmospheric particulates on pulmonary absorp-
tion phenomenon. In Assessing Toxic Effects of Environmental Pollutants, edited by S.D.
Lee, and J.B. Mudd, Ann Arbor Science, Ann Arbor, pp. 287-302, 1979.
Coffin, D.L., and D.E. Gardner. Interaction of biological agents and chemical air pollutants.
Ann. Occup. Hyg. 75:219-234, 1972.
Davies, C.N. A formalized anatomy of the human respiratory tract. In Inhaled Particles and
Vapours, edited by C.N. Davies, Pergamon Press, Oxford, pp. 82-91, 1961.
Drew, R.T., editor. Proceedings of the Workshop on Inhalation Chamber Technology,
Brookhaven National Laboratory, Upton, New York, October 16-17, 1978. Published by
U.S. Department of Commerce, Springfield, Virginia, 94 p., 1978.
Drew, R.T., and S. Laskin. Environmental inhalation chambers. In Methods of Animal Ex-
perimentation, Volume IV, edited by W.I. Gay, Academic Press, New York, 41 p., 1973.
Drill, S., and R. Thomas. Evaluation of Short-Term Bioassays to Predict Functional Impair-
ment, U.S. Army Bioengineering Research and Development Laboratory, Fort Detrick,
Frederick, Maryland, Contract No. AD A103766. Published by Mitre Corporation,
Mclean, Virginia, 119 p., 1980.
Duke, K.M., M.E. Davies, and A.J. Dennis. IERL-RTP Procedures Manual: Level 1 Environ-
mental Assessment Biological Tests Tests for Pilot Studies, EPA Publication No. 600/7-
77-043, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina,
103 p., 1977.
Ehrlich, R. Interaction between environmental pollutants and respiratory infections. Environ.
Health Persp. 55:89-100, 1980.
Ehrlich, R., and J.D. Fenters. Influence of nitrogen dioxide on experimental influenza in
squirrel monkeys. In Proceedings of the Third International Clean Air Congress. UDI-
Verlag, Dusseldorf, West Germany, pp. A11-A13, 1973.
Frank, A.L. Occupational lung cancer. In Pathogenesis and Therapy of Lung Cancer, edited
by C.C. Harris, Marcel Dekker, Inc., New York, pp. 25-52, 1978.
Freeman, B.A., and J.J. O'Neil. Use of tissue slices in lung toxicology studies. In Pulmonary
Toxicology, edited by G.E.R. Hook, Raven Press, New York, in press, 1982.
Freeman, G., S.C. Crane, N.J. Furiosi, R.J. Stephens, M.J. Evans, and W.D. Moore. Covert
reduction in ventilatory surface in rats during prolonged exposure to subacute nitrogen
dioxide. Am. Rev. Resp. Dis. 70(5:563-577, 1972.
Freeman, G., L.T. Juhos, N.J. Furiosi, R. Mussenden, R.J. Stephens, and M.J. Evans.
Pathology of pulmonary disease from exposure to interdependent ambient gases (NO2 and
NO3). Arch. Envrion. Health 29:203-210, 1974.
Gardner, D.E. Alteration in host bacteria interaction by environmental chemicals. In Assessing
Toxic Effects of Environmental Pollutants, edited by S.D. Lee, Ann Arbor Science, Ann
Arbor, pp. 87-105, 1979.
205
-------
Gardner, D.E. Effects of environmental chemicals on host defense responses of the lung. In
Biological Relevance of Immune Suppression, edited by J.H. Dean, and M. Padarathsingh,
Van Nostrand Reinhold Company, San Francisco, pp. 275-287, 1981.
Alterations in macrophage function by environmental chemicals. In Pulmonary
Toxicologv, edited by G.E.R. Hook, Raven Press, New York, in press, 1982a.
The use of experimental airborne infections for monitoring altered host defenses.
Environ. Health Persp., in press, 1982b.
Gardner, D.E., and J.A. Graham. Increased pulmonary disease mediated through altered
bacterial defenses. In Pulmonary Macrophages and Epithelial Cells, edited by C.L.
Sanders, R.P. Schneider, G.E. Dagle, and H.A. Ragan, CONF-760927, Energy Research
and Development Administration, Technical Information Center (available from NTIS,
Springfield, Virginia), pp. 1-21, 1977.
Gardner, D.E..T.R. Lewis, S.M. Alpert, D.J. Hurst, and D.L. Coffin. The role of tolerance in
pulmonary defense mechanisms. Arch. Environ. Health 25:432-438, 1972.
Gillespie, J.R. Review of the cardiovascular and pulmonary function studies on beagles ex-
posed for 68 months to auto exhaust and other air pollutants. In Long-Term Effects of Air
Pollutants: In Canine Species, edited by J.F. Stara, D.L. Dungworth, J.G. Orthoefer, and
W.S. Tyler, EPA-600/8-80-014, U.S. Environmental Protection Agency, Washington,
DC, pp. 115-148, 1980.
Goldstein, E. The influence of environmental toxicity on the antibacterial activity of the
alveolar macrophage. In Pulmonary Macrophages and Epithelial Cells, edited by C.L.
Sanders, R.P. Schneider, G.E. Dagle, and H.A. Ragan, CONF-760927, Energy Research
and Development Administration, Technical Information Center (available from NTIS,
Springfield, Virginia), pp. 387-394, 1977.
Graham, J.A., D.E. Gardner, M.D. Waters, and D.L. Coffin. Effect of trace metals on phago-
cytosis by alveolar macrophages. Infect. Immun. 77:1278-1283, 1975.
Green, G.M., G.J. Jakab, R.B. Low, and G.E. Davis. Defense mechanisms of the respiratory
membrane. Am. Rev. Resp. Dis. 7/5:479-514, 1977.
Griesemer, R.A., P. Nettesheim, D.H. Martin, and J.E. Caton, Jr. Quantitative exposure of
grafted rat tracheas to 7,12-dimethylbenz(a)anthracene. Cancer Res. 37:1266-1271, 1977.
Grose, E.G., D.E. Gardner, and F. J. Miller. Response of ciliated epithelium to ozone and sul-
furic acid. Environ. Res. 22:377-385, 1980.
Hammond, C.E. Smoking in relation to the death rates of one million men and women. Nail.
Cancer Inst. Monog. 19:127-204, 1966.
Harington, J.S., K. Miller, and G. Macnab. Hemolysis by asbestos. Environ. Res. 4:95-117,
1971.
Hatch, G., R. Slade, E. Boykin, F.J. Miller, and D.E. Gardner. Correlation of effects of
inhaled versus intratracheally injected metals on susceptibility to infection in mice. Am.
Rev. Resp. Dis. 724(2):167-173, 1981.
Henderson, R.F., A.H. Rebar, J.A. Pickrell, D.G. Brownstein, and H. Muhle. The use of pul-
monary lavage fluid as an indicator of lung damage: effect of species, age, and method of
lavage on baseline values. In Lovelace Inhalation Toxicology Research Institute Annual
Report LF-60, U.S. Department of Commerce (available from NTIS, Springfield, Vir-
ginia), pp. 339-345, 1978.
Henderson, R.F. The use of pulmonary washings to detect lung damage. In Pulmonary Toxi-
cology, edited by G.E.R. Hook, Raven Press, New York, ip press, 1982.
Hinners, R.G., J.K. Burkart, and G.L. Contner. Animal exposure chambers in air pollution
studies. Arch. Environ. Health 73:609-615, 1966.
Holt, P.P. Small animals in the study of pathological effects of asbestos. Environ. Health
Persp. 9:205-211, 1974.
Horvath, S.M., editor. Pulmonary Function Testing in Occupational Medicine, Navy En-
vironmental Health Center, Naval Station, Norfolk, Virginia, 29 p., 1977.
Hu, P.C., F.J. Miller, M.J. Daniels, G.E. Hatch, J.A. Graham, D.E. Gardner, and M.J.K.
Selgrade. Protein accumulation in lung lavage fluid following ozone exposure. Environ.
Res., in press, 1982.
Hunninghake, G.W., J.E. Gadek, V.J. Ferrans, and R.G. Crystal. Inflammatory and immune
processes in the human lung in health and disease: Evaluation by bronchoalveolar lavage.
Am. J. Pathol. 97:149-206, 1979. ^
Hyde, D., J.G. Orthoefer, D. Dungworth, W. Tyler, R. Carter, and H. Lum. Morphometric
and morphologic evaluation of pulmonary lesions in beagle dogs chronically exposed to
high ambient levels of air pollutants. Lab. Invest. 35:455-469, 1978.
206
-------
Jacoby, R.O.,and S.W. Barthold. Quality assurance for rodents used in toxicological research
and testing. In Scientific Considerations in Monitoring and Evaluating Toxicological Re-
search, edited by E.J. Gralla, Hemisphere Publishing Company, Washington DC pp 27-
55,1981.
Janoff, A., B. Sloan, G. Weinbaum, V. Damiano, R.A. Sandhaus, J. Elias.and P. Kimbel. Ex-
perimental emphysema induced with purified human neutrophil elastase: Tissue localiza-
tion of the instilled protease. Am. Rev. Resp. Dis. 775:461-478, 1977.
Karlinsky, J.B., and G.L. Snider. Animal models of emphysema. Am. Rev. Resp. Dis.
7/7:1109-1133, 1978.
Kendrick, J., P. Nettesheim, and A.S. Mammons. Tumor induction in tracheal grafts: A new
experimental model for respiratory carcinogenesis studies. J. Nat 1. Cancer Inst. 52:1317-
1325, 1974.
Kliment, V. Similarity and dimensional analysis, evaluation of aerosol deposition in the lungs
of laboratory animals and man. Folia Morphol. 27:59-64, 1973.
Koo, K.W., D.E. Leith, C.B. Sherter, and G.L. Snider. Respiratory mechanics in normal ham-
sters. J. Appl. Physiol. 40:936-942, 1976.
Kraybill, H.F. Conceptual approaches to the assessment of nonoccupational environmental
cancer. In Advances in Modern Toxicology, Volume 3, Environmental Cancer, edited by
H.F. Kraybill, and M.A. Mehlman, Hemisphere Publishing Company, Washington, DC,
pp. 27-62, 1977.
Kubich, U., J. Rosenbloom, P. Christner, P. Kimbel, and G. Weinbaum. Development of a
hemagglutination assay to measure elastin fragments and antielastin antibodies. Chest 77
(suppl.):278-279, 1980.
Lamb, D., and L. Reid. Mitotic rates, goblet cell increase, and histochemical changes in mucus
and in rat bronchial epithelium during exposure to sulphur dioxide. J. Pathol. Bacterial.
95:97-111, 1968.
Laskin, S., M. Kuschner, and R.T. Drew. Studies in pulmonary carcinogenesis. In Inhalation
Carcinogenesis, edited by M.G. Hanna, Jr., P. Nettesheim, and J.R. Gilbert, AEC Sympo-
sium Series, 75:321-351, 1970.
Last, J.A., and T. Kaizu. Mucus glycoprotein secretion by tracheal explants: Effects of pol-
lutants. Environ. Health Persp. 15:131-137, 1980.
Leach, L. J. A Multiple Chamber Exposure Unit Designed for Chronic Inhalation Studies, Re-
port UR-534, University of Rochester, Rochester, New York, 1958.
Leith, D.E. Comparative mammalian respiratory mechanics. The Physiologist / 9:485-510,
1976.
Lewis, T.R., W.J. Moorman, Y.Y. Yang, and J.F. Stara. Long-term exposure to auto exhaust
and other pollutant mixtures: effects on pulmonary function in the beagle. Arch. Environ.
Health 29:102-106, 1974.
Light, W.G., and E.T. Wei. Surface charge and hemolytic activity of asbestos. Environ. Res.
73:135-145, 1977.
Lippmann, M. Aerosol exposure methods. In Generation of Aerosols and Facilitiesfor Expo-
sure Experiments, edited by K. Willeke, Ann Arbor Science, Ann Arbor, pp, 443-458, 1980.
Lippmann, M., R.B. Schlesinger, G. Leikauf, D. Spektor, and R.E. Albert. Effects of sulfuric
acid on respiratory tract airways. In Inhaled Particles, edited by V.W.H. Walton, Unwin
Brothers, Ltd., Surrey, England, in press, 1982.
Lucey, E.C.,B.R. Celli,and G.L. Snider. Maximum expiratory flow and transpulmonary pres-
sure in the hamster. J. Appl. Physiol. Respirat. Environ. Exercise Physiol. 45:840-845,
1978.
Maigetter, R.Z., R. Ehrlich, J.D. Fenters,and D.E. Gardner. Potentiating effects of manganese
dioxide on experimental respiratory infections. Environ. Res. 77:386-391, 1976.
Maronpot, R.R., H.P. Witschi, L.H. Smith, and J.L. McCoy. Recent experience with the
strain A mouse pulmonary tumor bioassay model. In Short- Term Bioassays in the Analysis
of Complex Environmental Mixtures III, edited by M.D. Waters, S.S. Sandhu, L. Claxton,
N. Chernoff, and S. Nesnow, in press, 1982.
Mauderly, J.L., E.G. Damon, and R.K. Jones. Effects of intratracheally instilled elastase on
lung function of Fischer-344 rats. Fed. Proc. 38:1325, 1979.
Mauderly, J.L., W.C. Nenno, and G.A. Morrison. Stocks for holding unanesthetized dogs in
the standing position. Lab. Animal Sci. 27:263-266, 1971.
Mulvihill, J.J. Congenital and genetic diseases. In Persons at High Risk of Cancer: An Ap-
proach to Cancer Etiology and Control, edited by J.F. Fraumeni, Academic Press, New
York, pp. 3-25, 1975.
207
-------
Nash, T., A.C. Allison, and J.S. Harington. Physico-chemical properties of silica in relation to
its toxicity. Nalure(London) 2/0:259-261, 1966.
National Research Council. Respiratory transport and absorption. In Ozone and Other Photo-
chemical Oxidants, National Academy of Sciences, Washington, DC, pp. 280-322, 1977.
Nebert, D.W., S.A. Atlas, T.M. Guenthner, and R.E. Kouri. The Ah-locus: Genetic regula-
tion of the enzymes which metabolize polycyclic hydrocarbons and the risk for cancer. In
Polycyclic Hydrocarbons and Cancer, Volume 2, edited by H. Gelboin, and P. Ts'o, Aca-
demic Press, New York, pp. 1272-1278, 1978.
Nettesheim, P., and R. A. Griesemer. Experimental models for studies of respiratory tract car-
cinogenesis. In Lung Biology in Health and Disease, Volume 10, Pathogenesis and Therapy
of Lung Cancer, edited by C.C. Harris, Marcel Dekker, Inc., pp. 75-188, 1978.
Nettesheim, P., R.A. Griesemer, D.H. Martin, and J.E. Caton, Jr. Induction of preneoplastic
and neoplastic lesions in grafted rat tracheas continuously exposed to benzo(a)pyrene.
Cancer Res. 37:1272-1278, 1977.
Nettesheim, P., D.C. Topping, and R. Jamasbi. Host and environmental factors enhancing
carcinogenesis in the respiratory tract. Ann Rev. Pharmacol, Toxicol 27:133-163, 1981.
Niemeier, R. W. Isolated perfused rabbit lung: A critical appraisal. Environ. Health Persp. 16'.
67-71, 1976.
O'Hery, J., and I. P. Gornley. Some factors affecting the haemolytic activity of silicate minerals.
Ann. Occup. Hyg. 27:131-139, 1978.
O'Neil, J.J., and J.A. Raub. Pulmonary function testingin small laboratory mammals. In Pul-
monary Toxicology, edited by G.E.R. Hook, Raven Press, New York, in press, 1982.
O'Neil, J, J., and D. F. Tierney. Rat metabolism: utilization by isolated perfused lungs and tissue
slices. Am. J. Physiol. 226:867-873, 1974.
O'Neil, J.J., and S.L. Young. Tissue slices in the study of lung metabolism. In The Develop-
mental Biology of the Lung and the Pathobiology of Hyaline Membrane Disease, edited by
P.M. Farrell, Academic Press, New York, pp. 87-99, 1982.
Orthoefer, J.G., R.S. Bhatnagar, A. Rahman, Y.Y. Yang, S.D. Lee, and J.F. Stara. Collagen
and prolyl hydroxlylase levels in lungs of beagles exposed to air pollutants. Environ. Res.
72:299-305, 1976.
Paigen, B., H.L. Gurtoo, J. Minowada, E. Ward, L. Houton, K. Paigen, A. Reilly, and R.
Vincent. Genetics of aryl hydrocarbon hydroxylase in the human population and its rela-
tionship to lung cancer. In Polycyclic Hydrocrbons and Cancer, Volume 2, edited by H.
Gelboin, and P. Ts'o, Academic Press, New York, pp. 391-406, 1978.
Pal, B.C., D.C. Topping, R.A. Griesemer, F.R. Nelson, and P. Nettesheim. Development of a
system for controlled release of benzo(a)pyrene, 7,12-dimethylbenz(a)anthracene and
phorbol ester for tumor induction in heterotopic tracheal grafts. Cancer Res. 38'. 1376-1383,
1978.
Palekar, L.D., C.M. Spooner, and D.L. Coffin. Influence of crystallization habit of minerals
on in vitro cytotoxicity. Ann. NY Acad. Sci. 330:673-686, 1979.
Phalen, R.F. Inhalation exposure of animals. Environ. Health Persp. 76:17-24, 1976.
Phalen, R.F.,and P.E. Morrow. Experimental inhalation of metallic silver. Health Physics24:
509-518, 1973.
Pickrell, J.A., F.F. Hahn, A.H. Rebar, E.G. Damon, R.L. Beethe, R.C. Pfleger, and C.H.
Hobbs. Pulmonary effects of exposure to 20 ppm NO2. Chest 50:50-52, 1981.
Port, C.D., K.V. Ketels, D.L. Coffin, and P. Kane. A comparative study of experimental and
spontaneous emphysema. J. Toxicol. Environ. Health 2:589-604, 1977.
Proctor, D.F. Tests of airway defense mechanisms. In The Lung in the Transition Between
Health and Disease, Volume 12, edited by C. Lenfant, Marcel Dekker, Inc., New York, pp.
227-24i, 1979.
Raabe, O.G. An improved apparatus for acute inhalation exposure of rodents to radioactive
aerosols. Toxicol. Appl. Pharmacol. 26:264-273, 1973.
• Deposition and clearance of inhaled aerosols. In Mechanisms in Respiratory
Toxicology, Volume I, edited by H. Witschi, and P. Nettesheim, CRC Press, Inc., Boca
Raton, Florida, pp. 27-76, 1982.
Rampy, L.W. Generating and controlling atmospheres in inhalation chambers. In Scientific
Considerations in Monitoring and Evaluating Toxicological Research, edited by E.J.
Gralla, Hemisphere Publishing Company, Washington, DC, pp. 57-70, 1981.
Ranga, V., and J. Kleinerman. Structure and function of small airways in health and disease.
Arch. Pathoi Lab. Med. 702:609-617, 1978.
208
-------
Raub, J.A. Effects of low-level ozone exposure on pulmonary function in adults and neonatal
rats. Adv. Modern Environ. Toxicol, in press, 1982.
Raub, J.A., R.R. Mercer, F.J. Miller, J.A. Graham, and J.J. O'Neil. Dose response of elastase-
induced emphysema in hamsters. Am. Rev. Resp. Dis. 725:432-435, 1982a.
Raub, J.A., M.J. Wiester, F.J. Miller, J.A. Graham, J.J. O'Neil, and J.D. Crapo. Ozone-in-
duced alterations of pulmonary function in adult and neonatal rats. Am. Rev. Resp. Dis.
7 25(2): 147, 1982b.
Reid, L. Evaluation of model systems for study of airway epithelium, cilia, and mucus Arch.
Intern. Med. 726:428^434, 1970.
Reid, L.M., and R. Jones. Mucous membrane of respiratory epithelium. Environ. Health
Persp. 55:113-119, 1980.
Reiser, K.M., and J.A. Last. Silicosis and fibrogenesis: Fact and artifact. Toxicol. 13:51-72,
1979.
Sanders, C.L., R.P. Schneider, G.E. Dagle, and H.A. Ragan, editors, Pulmonary Macro-
phages and Epithelial Cells, CONF-760927, Energy Research and Development Adminis-
tration, Technical Information Center (available from NTIS, Springfield, Virginia), 228
pp., 1977.
Schiff, L.J., M.M. Byrne, and W.T. Brown. Technique for determining ciliary activity in
hamster tracheal organ culture. Tissue Culture Assoc. Manual -#7:871-874, 1978.
Schiff, L.J., M.M. Byrne, J.D. Fenters, J.A. Graham, and D.E. Gardner. Cytotoxic effects of
sulfuric acid mist, carbon particulates and their mixtures on hamster tracheal epithelium.
Environ. Res. 79:339-354, 1979.
Schiff, L.J., M.M. Byrne, and J.A. Graham. Flyash-induced changes in hamster tracheal epi-
thelium in vivo and in vitro. J. Toxicol. Environ. Health 5:431-448, 1981.
Schlesinger, R.B., M. Halpern, R.E. Albert, and M. Lippmann. Effect of chronic inhalation of
sulfuric acid mist upon mucociliary clearance from the lungs of donkeys. /. Environ.
Pathol. Toxicol. 2:1351-1367, 1979.
Schlesinger, R.B., M. Lippmann, and R.E. Albert. Effects of short-term exposures to sulfuric
acid and ammonium sulfate aerosols upon bronchial airway function in the donkey. Am.
Ind. Hyg. Assoc. J. 59:275-286, 1978.
Schnitzer, R.J., and F.L. Pundsack. Asbestos hemolysis. Environ. Res. 3:1-13, 1970.
Belgrade, M.J.K., M.L. Mole, F.J. Miller, G.E. Hatch, D.E. Gardner, and P.C. Hu. Effects of
NOz inhalation and vitamin C deficiency on protein and lipid accumulation in the lung.
Environ. Res. 26:422-437, 1981.
Shimkin, M.B., and G.D. Stoner. Lung tumors in mice: Application to carcinogenesis bio-
assay. In Advances in Cancer Research, edited by G. Klein, and S. Weinhouse, Academic
Press, New York, pp. 1-58, 1975.
Sinnett, E.E., A.C. Jackson, D.E. Leith, and J.P. Butler. Fast integrated flow plethysmograph
for small mammals. J. Appl. Physiol. Respirat. Environ. Exercise Physiol. 50:1104-1110,
1981.
Slonim,N.B.,and L.H. Hamilton. Respiratory Physiology, 3rd ed., C.V. Mosby Company, St.
Louis, 225 p., 1976.
Smith, B.R., and J.R. Bend. Lung-perfusion techniques for xenobiotic metabolism and toxicity
studies. Methods of Enzymology 77:105-120, 1981.
Snider, G.L., and C.B. Sherter. A one-year study of the evolution of elastase-induced emphy-
sema in hamsters. J. Appl. Physiol. Respirat. Environ. Exercise Physiol. 43:721-729, 1977.
Stara, J.F., D.L. Dungworth, J.G. Orthoefer, and W.S. Tyler, editors. Long-Term Effects of
Air Pollutants: In Canine Species, EPA-600/8-80-014, U.S. Environmental Protection
Agency, Washington, DC, 287 p., 1980.
Sturgess, J., and L. Reid. Secretory activity of the human bronchial mucous glands in vitro.
Exp.'lMolec. Pathol. 76:362-381, 1972.
Summerton, J., S. Hoenig, C. Butler II, and M. Chvapil. The mechanism of hemolysis by
silica and its bearing on silicosis. Exp. Molec. Palhol. 26:113-128, 1977.
Takezawa, J., F.J. Miller, and J.J. O'Neil. Single breath diffusing capacity and lung volumes in
small laboratory mammals. /. Appl. Physiol. Respirat. Environ. Exercise Physio/. 48:1052-
1059, 1980.
Tattersall, S.F., M.K. Benson, D. Hunter, A. Mansell, N.B. Pride, and C.M. Fletcher. The use
of tests of peripheral lung function for predicting future disability from airflow obstruction
in middle-aged smokers. Am. Rev. Resp. Dis. 7/5:1035-1050, 1978.
Templeton, A.C. Acquired diseases. In Persons at High Risk of Cancer: An Approach to Can-
cer Etiology and Control, edited by J.F. Fraumeni, Academic Press, New York, pp. 69-84,
1975.
209
-------
Thomas, R.G., and R. Lie. Procedures and Equipment Used in Inhalation Studies on Small
Animals, Lovelace Foundation Report No. LF-11, U.S. Atomic Energy Commission Re-
search and Development, Albuquerque, 1963.
Thurlbeck, W.M. Changes in lung structure. In The Lung in the Transition Between Health and
Disease, edited by P.T. Macklem, and S. Permutt, Marcel Dekker, Inc., New York, pp. 17-
41, 1979.
Topping, D.C., D.H. Martin, and P. Nettesheim. Determination of cocarcinogenic activity of
benzo(e)pyrene for respiratory tract mucosa. Cancer Lett. 77:315-321, 1981.
Topping, D.C., B.C. Pal, D.H. Martin, F.R. Nelson, and P. Nettesheim. Pathologic changes in-
duced in respiratory tract mucosa by polycyclic hydrocarbons of differing carcinogenic
activity. Am. J. Pathol. 93:3\ 1-324, 1978.
U.S. Environmental Protection Agency. Respiratory tract deposition and fate of inhaled aero-
sols and SCh. In Air Quality Criteria for Paniculate Matter and Sulfur Oxides, U.S. Envi-
ronmental Protection Agency, Research Triangle Park, North Carolina, in press, 1982.
Van Duuren, B.L. Tumor-promoting agents in two-stage carcinogenesis. Prog. Exp. Tumor
Res. 77:31-68, 1969.
Van Duuren, B.L., T. Blazey, B.M. Goldschmidt, C. Katz, S. Melchionne, and A. Sivak. Co-
carcinogenesis studies on mouse skin and inhibition of tumor induction. J, Natl. Cancer
Inst. 46:1039-1044, 1971.
Van Duuren, B.L., and B.M. Goldschmidt. Cocarcinogenic and tumor-promoting agents in
tobacco carcinogenesis. /. Natl. Cancer Inst. 5(5:1237-1242, 1976.
Vaughan, J.R., Jr., L.F. Jennelle, and T.R. Lewis. Long-term exposure to low levels of air pol-
lutants: effects on pulmonary function in the beagle. Arch. Environ. Health 79:45-50, 1969.
Waters, M.D., D.E. Gardner, C. Aranyi, and D.L. Coffin. Metal toxicity for rabbit alveolar
macrophages in vitro. Environ. Res. 9:32-41, 1975.
Weibel, E.R. Morphometry of the Human Lung, Academic Press, New York, 151 p., 1963.
West, J.B. Pulmonary Pathophvsio/ogy - The Essentials, Williams and Wilkins Company, Bal-
timore, 227 p., 1977.
Willeke, K., editor. Generation of Aerosols and Facilities for Exposure Experiments, Ann
Arbor Science, Ann Arbor, 573 pp., 1980.
Williams, S.J., M.A. Karis, and D.B. Menzel. Interactions of heavy metals with the pulmonary
metabolism of 3H-benzo(a)pyrene. Environ. Res., in press, 1982.
Williams, S.J., and D.B. Menzel. Modulation of pulmonary p-nitroanisole O-dimethylase by
intratracheally instilled heavy metal salts. Toxicol. Appl. Pharmcol., in press, 1982.
Wolff, R.K., M. Dolovich, P. Eng, C.M. Rossman, and M.T. Newhouse. Sulfur dioxide and
tracheobronchial clearance in man. Arch. Environ. Health 30:521-521, 1975.
Wynder, E.L., E.L. Mabuchi, and E.J. Beattie. The epidemiology of lung cancer: recent trends,
J. Am. Med. Assoc. 2/3:2221-2228, 1970.
Young, S.L. An isolated perfused rat lung preparation. Environ. Health Persp. 7(5:61-66, 1976.
Zaidi, S.H. Experimental Pneumoconiosis, Johns Hopkins University Press, Baltimore, 1969.
Ziskind, M., R.N. Jones, and H. Weill. Silicosis, Am. Rev. Resp. Dis. 7/3:643-665, 1976.
210
-------
INDEX
Abortion (see also Human abortuses)
chromosomal defects, 30,70
surveillance methods, 70
spontaneous, 30,70,78,79
therapeutic, 78
Acetaminophen
toxic effect of, 97
Achondroplasia, 73
Acinus, gas exchange, 177
Acrylamide
neurotoxic effects of, 114
weight loss from, 114,121,132
Activity, measures of
and hazard identification, 118,120,133
Acute versus chronic exposure, 4
Adenocarcinomas, vaginal
and diethylstilbestrol, 2
Adenosine triphosphate
transport functions, 90
ADH (see Antidiuretic hormone)
Aflatoxin B,
activation-requiring, 17
liver carcinogen, 17
Aging
induced by toxic substances, 11,81
of brain, 131
Air pollutants
and COLD, 179
behavioral toxicity, 109,124
carcinogenicity, 199
Air samples (particulates)
complex mixtures, 31
Alcohol syndrome, fetal, 82,83
Alkylating agents, 44,45,59,63-66,75
chronic exposures, 4
reaction with DNA, 4,6,8
Alkylguanine, 4,5
Ames test (see also "mutagenicity test"
under Salmonella)
for mutagenicity, 18,19,49,71
p-Aminohippuric acid (see PAH)
Amniocentesis
and fetal monitoring, 78
AMP, 101
Amphetamine
behavior modified by, 129
Analgesics
nephrotoxic effect, 97
Aneuploid cells
focus assay, 41
spontaneous transformation, 43
Aneuploidy
assays for, 21,26,28,29
development of tests for, 23
Down's syndrome, 30
evaluation, 21,28,29
fungal assays, 23,28
mental and physical defects, 15
spontaneous abortion, 30
Angina pectoris
and exercise performance, 139
Animal pharmacokinetics
dose-response extrapolation, 7,8
Anticoagulants
dicoumarin, 83
Antidiuretic hormone (ADH), 94,95,101
Antigens, and
glomerular lesions, 96
vascular injury, 144,145,161
Aroclor-1254
induction of enzymes, 17
Aromatic amines, 44,54,60,62
Arsenic
and cardiovascular mortality, 141
Arterial endothelium
contractile protein in, 146
Arterial injury
and tobacco antigen, 158
hypothesis, 143-144,153
lipid accumulation, 156-157
Arterial permeability
and angiotensin II, 158
Arteriosclerosis (see also Atherosclerosis),
and
inflammation, 144
thrombus formation, 162
Arteriosclerotic heart disease
personal risk factors, 137-138
Arteriosclerotic plaques (see also
Atherosclerotic plaques; Plaque formation)
development of, 144
isoenzymes, 153
transformation of smooth muscle cells,
154-155
Arylamines, 45
Asbestos
and COLD, 179
hemolytic activity, 196
lung mesothelium, 2
macrophage cytotoxicity, 194
transformation effects, 49
Aspergillus flavus
211
-------
aflatoxin synthesis, 17
Aspirin
nephrotoxic effects, 97
teratogenicity, 71
Assay systems (see also individual test
systems; In vitro tests)
ideal, properties of, viii,39
comparative results, 44,45,49-50,63-66
Asthma, and
anatomic dead space, 186
chemical exposure, 179
Atherosclerosis (see also Arteriosclerosis)
and diabetes, 137,159
and graft rejection, 149-150
and immune reactions, 151
and platelet function, 143
and trace elements, 141
and vasectomy, 148
and virus infections, 146,151
deaths from, 136,137
development of, 143
distribution in arterial tree, 148
precocious development of, 145,161
Atherosclerotic plaques (see also
Arteriosclerotic plaques;
Plaque formation)
avidity for lipid, 148
cholesterol, 137
tumor growth, 162
Atomic bombs
human mortality data, 2,3
ATP (see Adenosine triphosphate)
ATPase (see also Na, K-ATPase)
activity affected by uranium, 99
transport-related, 105
Autoimmune phenomena
and silicosis, 192
Automobile exhaust
benzo (a)pyrene in, 18,200
Aversion and attraction
toxicity assessment by, 122,124
Avoidance conditioning
toxicity assessment by, 120
Background level
as risk assessment parameter, 1,9
of DNA adducts, 6,7,8
of ionizing radiation, 2
Bacterial DNA repair tests
as short-term tests, 18,19,23
Bacterial mutation test, 50,63-66
(see Ames test; "mutagenicity test"
under Salmonella)
Base-pair substitutions
Salmonella assay, 19
Behavior, and
heavy metals, 109
effect of microwaves on, 121-122
forced-choice inventories, 129
rating scales, 109,110
Behavioral teratology, 122,125,129
Benzo(a)pyrene
and neoplasia, 199
sources, 18
Bioactivation, 41,43,46,71,75,76
(see also Hepatic microsomal
monooxygenase; Hepatocyte feeder
layer; Liver microsomal homogenates;
Metabolic activation)
Biotransformation (see Bioactivation)
Blood, in
micronucleus assay, 25
mutagenicity tests, 27
population monitoring, 32
Blood lipid
and ischemic heart disease, 137
Blood pressure
physical factors in, 142
Blood ultrafiltrate
Bowman's capsule, 90,91
Blood urea nitrogen
as renal function test, 101
Bone cancer
from atomic bomb, 2,3
Bone marrow cells
chromosome aberrations assayed in, 26
Boron
nephrotoxic, 98
Brain, and
aging effects, 122-123,125
electrical "noise", 112,114,131
5-Bromodeoxyuridine
for detection of SCEs, 26
Bronchial airway function
and inhaled pollutants, 201
Bronchitis, and
chemical exposures, 178-179
cigarette smoking, 179-191
Bronchoscopes, 177
Cadmium
brain serotonin decreased by, 132
excretion, 99
inhalation, 99
nephrotoxic, 98
protein binding of, 96,99
pulmonary assays, 197
testicular atrophy from, 118
Capillary bed, renal, 90
Carbon
in vitro assays for, 197
Carbon disulfide
and cardiovascular disease, 142
and weight loss, 132
behavioral toxicity from, 108,130
Carbon monoxide
and ischemic heart disease, 139-140, 158
Carbon tetrachloride
tissue damage from, 100
Carboxyhemoglobin
and atherosclerosis, 139
Carcinogenesis
212
-------
multicausal, multistage nature, 47,81
Carcinogenic analogues, 55
Carcinogenicity
and mutagenicity, 31
Carcinogens
correct identification, 51,52
tracheal transplants, 198-199
Carcinoma of lung, 181
Cardiac rhythm
and magnesium, 140
Cardiovascular diseases
and occupational inhalants, 141-142
and water hardness and trace metals,
140-141
deaths from, 136
Cardiovascular mortality, and
arsenic, 141
inactivity, 160
recent trends, 162
Cell culture (see also Colony assay;
Embryonic cells; Fibroblasts; Focus assay;
In vitro tests; Mammalian cells in
culture; Mass cultures; Transformed cells)
and chemical Carcinogenicity assays, 48
epithelial, 46,48,103
in teratology, 75
Cellular membranes, renal
transport mechanism, 90
Center for Disease Control, 78,83,87
Chemical Abstracts Service Registry
Numbers, 82,86
Chemicals (see also specific classes)
Carcinogenicity assays of, 39
complex mixtures, 31,32,35
miscellaneous compounds, 44,45,59-66
mutagenic effects, 13
Chinese hamster ovary cells
(see also Hamster cells)
in vitro tests, 27
Chlorinated hydrocarbons
nephrotoxic, 100
CHO cells (see Chinese hamster ovary cehX
Cholesterol, and
ischemic heart disease, 137,157
low-density lipoprotein, 156
Chromium
and cardiovascular disease, 141
nephrotoxic, 98
Chromosomal aberrations
aneuploidy, euploidy, 15
in fungi, 23,28
in vitro and in vivo tests for, 22,23,27,28,
29,30,32
micronucleus test for, 25,26
teratogenic, 69, 70, 74
translocations and inversions, 15
Chromosome complement
of humans, 15
Chronic bronchitis, and
description, 179,180
glycoprotein changes, 191
pulmonary function tests, 187-191
sulfur dioxide and cigarette smoke, 191
Chronic obstructive lung disease (see also
Bronchitis; Chronic bronchitis;
Chronic pulmonary disease; Emphysema)
flow volume curve, 188
maximum midexpiratory flow rate,
187,188
pulmonary function tests, 187-191
susceptible subpopulations, 175
types of, 179
Chronic pulmonary disease
function tests, 184
short-term tests, 175,203
Chronic restrictive lung disease
description, 184,188
flow volume curve, 188
lung volumes, 189
maximum midexpiratory flow rate,
187,188
pulmonary function tests, 187-191
Cigarette smoke
allergenic glycoprotein antigen in,
146,158
benzo(a)pyrene in, 18,200
carbon monoxide in, 139
mucociliary clearance and, 201
mutagenic, 31
respiratory tract carcinogenesis, 198
Cigarette smokers (see also Smoking)
age effect, 158
ischemic heart disease, 137-138, 158
lung cancer, 16,17
mutagenicity of urine, 32
pulmonary lavage, 193
sex differences in COLD, 176
Circulatory system
aging of, 160
Clara cells
in lung preparations, 200
Clastogenic agents, 15
Cleft lip and palate, 73
CO (See Carbon monoxide)
Cobalt
and atherosclerosis, 141
Cocarcinogens
assays for, 47,48,56
COHb (see Carboxyhemoglobin)
COLD (see Chronic obstructive
lung disease)
Collagen biosynthesis
macrophage-stimulated, 197
short-term assays, 203
Colony assay, 39-40,42,48
advantages, 40
comparisons, 43
soft agar assay, 43,45,48,63-66
Common garden pea (see Pisum sativum)
Complement cascade
development of atherosclerosis, 146-147
Complementarity
213
-------
of replication, 14
Complex behavior
assessment of, 118-121,128-129
Complex chemical mixtures
and environmental monitoring, 31,32,35
Computer technology
in behavioral toxicology, 128
Congenital defects, 70,73,78,80,82
Consumer Product Safety Commission,
66,67
Coronary heart disease
psychological stress, 142
zinc/copper ratios, 141
Cortisone, 71,73
Critical organ concentration
definition of, 94-95
Croton oil
absence of transformation, 49
Cytochalasins
and cell division, 75
Cytochrome P^tSO
in induced S-9 fractions, 17
Cytosol, renal, 104
Data collection, vi,79
extrapolation of, v,vi
false positive findings, 79-80
memory bias, 80
DDT
effect on nervous system, 109,132
Dead space ventilation, 186
Deactivation pathways, 4
Degranulation, 50,63-66
liver reticulum, 49
Delaney Clause, 53,87
Density-dependent inhibition
of cell division, 38
Deoxyguanbsine
ethylation products of, 6
Deoxyribonucleic acid (see DNA)
Diabetes, and
abnormal microvasculature, 159
atherosclerosis, 159-160
autoimmune disease, 160
viruses, 160-161
Dichloromethane
behavioral toxin, 108
Dicoumarin
teratogenic, 83
Diethylstilbestrol
transplacental carcinogen, 78,81
vaginal adenocarcinomas, 2,9
Dimers (see Pyrimidine dimers)
Dimethylbenz (a) anthracene
in tracheal transplants, 199
Dimethylnitrosamine
activation of, 17,18
dose-response curve, 3,4
Diploid cells
in colony assay, 42
Distal tubule (see also Tubular [adj],
Tubules)
anatomy, 90,91,93
and mercurials, 97
Diuretic drugs
mercurials in, 97
urine concentration, 94
DNA
structure of, 14
DNA adducts (see also Alkylguanine;
Ethylguanine)
background levels, 6,7,8
concentration vs dose, 5,6
DNA damage
assayed by SCE, 26,27
gene conversion in fungi, 23
short-term tests for, 20-22,23,24,26-27,
28,29
DNA repair
of sunlight damage, 5
quantitative role in carcinogenesis, 5,7
tests for, 20-22,23,29,49
Dominant lethal test
whole-mammal in vivo test, 25
Dosage regime, vi
maximum, fractional, minimum, 73-74
Dose-response curves, v,vi,71
and teratogenicity, 71,73,75
extrapolation of,viii,l,2,3,9
linear, linear quadratic, quadratic, 2,3,4
survival versus mutation, 13
Doubling dose
in cancer mortality, 2
Down's syndrome, 80
aneuploidy, 30
trisomy of X chromosome, 15
Drosophila melanogaster
activating enzymes in, 18,24
aneuploidy detected in, 28,29
In vivo short-term assays, 12,30
sex-linked recessive lethal test, 12,24,30
spontaneous mutations, 12
x-ray-induced mutations, 12
E. co/i (see Escherichia coli)
Electrophysiological tests
toxicity assessment by, 112-114,130,131
Embryo-fetal lethality, 71,73,80
Embryonic cells
colony assay, 41
in vitro tests, 75,76,81
mass culture assay, 44,47
Emphysema
in alveoli, 177
immunological tests, 193-194
pulmonary function tests, 187
types of, 179
Endothelial cell surface
and immune complexes, 146,151
Fc and C3 receptors, 146
nicotine effects, 158
renin effects, 158
214
-------
Endothelium
and diabetics, 159
barrier function, 157
permeability and renin, 158
Environmental monitoring
by higher plants, 32,33
of complex chemical mixtures, 31,32
Environmental Protection Agency,
v,19,34,38,67,87
Environmental risk factors
and ischemic heart disease, 138-142
Environmental Teratogen Information
Center, 86
Enzyme assays, vii, 104-105
Enzymic radioimmunoassays
ultrasensitive, 5
Enzymuria
measurement of, 101
Epidemiologic studies
limitations of, 16
risk assessment, v,vi, 1,2
Epidermal cells (see Cell culture)
Epithelial cell culture (see Cell culture)
Escherichia coli
repair capacity, 23,24
supermutagenic effects, 13
Ethology
and behavioral toxicology, 128
O '-Ethylguanine
carcinogenic, 5,6
Euploidy
and lethality, 15
Excision-repair pathway (see also Nucleo-
tide excision-repair pathway)
error-prone, 14
Exercise
cardiovascular fitness, 160
respiratory dead space, 186
Exposure limits
legal processes, viii
setting of, vii
Exposure regime (see Dosage regime)
Extrapolation, vi,vii
linear, 7
mathematical models, 2-4
Ex vivo lung techniques, 200-201
False negative findings, viii,40
complementary tests, 45,49,50
false-alarm probability, viii
in behavioral toxicology, 133
False positive findings, viii,40
complementary tests, 45,49,50
epidemiological studies, 79,80
in behavioral toxicology, 133
Fetal alcohol syndrome, 82
Fetus
monitoring of, 78
sensitivity to methylmercury, 125
vunerability, 122
Fibroblasts (see also Cell culture)
cell transformation studies, 38,40,43
collagen biosynthesis by, 197,201
colony assay, 39
focus assay, 41-42
require metabolic activation, 54-55
unscheduled DNA synthesis, 27
Fibrosis
collagen deposition, 184,194
contrasted to COLD, 184
fibrous mineral exposure, 179,191-192
interstitial, 178-180
pulmonary function tests, 187
Fish
flounder kidney preparations, 104
methylmercury contamination, 112
Fixed interval performance, 121,133
Fluoro-alkanes
and myocardial irritability, 142
Fly ash
in vitro assays for, 197
Focus assay, 48,55,62-63
comparison with colony assay, 42
method, 41-42
Food additives
behavioral toxicity of, 109
Food and Drug Administration, 53,66,67,
87,88
Forward mutation
in fungi, 23
in mammalian cell culture, 23
in Neurospora, 28
in vivo test in somatic cells, 22,24,25,29
Frameshift mutagens
and Salmonella assays, 18-19
Fruit flies (see Drosophila)
Fungi
metabolic activation, 18
short-term tests, 19,23,28-30
Fungicides, mercurial
nephrotoxic, 98
Gene conversion
in S. cerevisiae, 23,28
Gene mutations, 15
and congenital defects, 70,73
Genetic heterogeneity
mutation-prone genotypes, 32
variation in susceptibility, 73,80
Genetic toxicology
defined, 11
future emphases, 35
Gene-Tox Program, 19,20-22,34
Germ cells
dominant lethal mutation assay, 25
in vivo mammalian tests, 24,25,26,30
of higher plants, 23
sex-linked recessive lethal assay, 25
Glomerular filtration, 92, 100
dynamics, 90,91
immunologic-type reactions, 96
in vitro tests, 102,103
215
-------
Glomerular tubule function
in vitro tests, 102
Glomerulus, 90,91,92,93,96
Golden Syrian hamster cells
(see under Hamster cells)
Guinea pig cells
spontaneous transformation, 39
embryo focus assay, 41,48
Hamster cells (see also Chinese
hamster ovary cells)
baby hamster kidney (BHK-21)
colony formation in soft agar, 43
focus assay, 42
Golden Syrian, for colony assay, 40,42,
48,59-61
spontaneous transformation, 39
Hazard identification and evaluation, 27-30
Hearing
impaired by methylmercury, 111
Heart attacks
age effects, 160
familial predisposition, 138
induced by sex hormones, 160
toxic substances, 11
Heavy metals (see also Cadmium; Lead;
Mercury; Platinum; Uranium)
subcellular partitioning, 95
toxicity of, 89,90,94,98,108
Henle's Loop (see Loop of Henle)
Hepatic coma
neurological effect, 132
Hepatic microsomal monooxygenase
(see also Liver microsomal homogenates)
for bioactivation, 41,71,75
Hepatitis
and cardiovascular disease, 161
Hepatocyte feeder layer
(see also Hepatic microsomal
monooxygenase; Liver microsomal
homogenates), 55
Herpes virus
and atherosclerosis, 151
Higher plant assays, 23
in silu environmental monitors, 32
metabolic activation systems, 23
Histopathology
in toxicity assessment, 131
Homeostatic functions, 121-122
role of kidney, 94
Host-mediated assays
for genotoxic activity, 16,27
Human abortuses (see also Abortion)
thyroid tissue, 75
spontaneous, 78
Human body fluids
population monitoring, 32
Human cells
in focus assay, 41,48
population monitoring, 32
spontaneous transformation, 39
Human population monitoring
short-term tests, 32,34,35
use of body fluids and cells, 32
Humans
cancer mortality from atomic bombs, 2,3
chromosome complement of, 15
Hydralazine
and glomerulonephritis, 96
jV-Hydroxyurethane
genotoxic, 23,24
Hyperactivity
lead-induced, 133
Hypercholesterolemia
endothelial injury, 157
Hyperlipemia, 156
Hypertension
and ischemic heart disease, 141,158-159
Identical twins
and drug clearance, 78
IHD (see Ischemic heart disease)
Immune complex deposition
antigen effects, 145,161
caloric intake, 148
from drugs and microbes, 145
in arterial endothelium, 146-147
Industry
evaluation of short-term assays by, 30-31
teratologic testing by, 87
Inhalants
and ischemic heart disease, 141-142
Initiating agents (see also Carcinogens)
properties of, 49
Insecticides
nephrotoxic, 100
Instrumented tasks
and reflex testing, 114
Insulin
and heart disease, 159
Interagency Regulatory Liaison Group,
53,66
International Agency for Research in
Cancer, 87
Interstitial nephrotoxins, 96-100
Intimal thickening
and atherosclerosis, 154,157
Inulin
renal clearance, 101
In vitro tests (see also Assay systems;
Cell culture; Colony assay; Embryo
culture; Focus assay; Mass culture;
Organ culture)
for carcinogens, 42,55
for renal function, 102-105
for teratogens, 74-76,81
mammalian cell culture
chromosome aberrations, 26-27
forward mutations, 23,24
sister chromatid exchange, 26
In vivo tests
for renal function, 101
216
-------
mammalian
chromosome aberrations, 26
mouse spot test, 24,25
sister chromatid exchange, 26,27
specific locus test, 25
Ionizing radiation
carcinogenic, 2,81
IQ (see also Rating scales)
and toxicity assessment, 125,126
Ischemic heart disease, and
birth control pills, 138,160
carbon monoxide, 139-140
cholesterol, 137
cigarette smoking, 137-138
environmental risk factors, 138-142
high blood pressure, 137
zinc/copper ratio, 141
Kepone
neurological disorders, 115
Kernicterus, 70
Kidney (see also Glomerular [adj.];
Nephron; Renal [adj.]; Tubular [adj.];
Tubules)
anatomy of 89,90,91
as target organ, 99-100
functional tests, 101,105
in vitro cell assays, 102,103,105
in vitro perfusion, 102
functions, 89,90
reserve capacity of, 94,100
urine formation in, 90
Lead
and hypertension, 141
behavioral toxicity from, 108,121,133
blood levels of, 121,125
depressed nerve conduction velocity from.
116
depressed school performance from, 108,
125,126,133
in deciduous teeth, 125
nephrotoxic, 98,99
sources of, 99,105
Learned behavior
avoidance conditioning, 120,121
foraging behavior, 129
Schedules of Reinforcement of, 121,129
toxicity assessment by, 121,129
Leucocytes
and chromosome aberrations, 27,32
Leukemia
incidence, from atomic bomb, 2,3
Life span
and caloric intake, 148
Limbs of Henle (see also Loop of Henle)
and mercurial toxicity, 97
Lipid accumulation
in cellular membranes, 90
potentiated by carbon monoxide, 158
preferential sites for, 148
Lipid filtration hypothesis
in atherosclerosis, 143-144
Lipoprotein
in injured arteries, 157
mitogenicity, 157
Liver
activation enzymes in, 14
damage from carbon tetrachloride, 100
Liver cells (see Cell culture)
Liver microsomal homogenates
(S-9 homogenates) (see also Bioactivation;
Hepatic microsomal monooxygenase;
Hepatocyte feeder layer; Metabolic
activation; Rodent liver homogenates)
and colony formation, 43
for transformation, 41,49,54
Longitudinal studies
in behavior teratology, 125,129
Loop of Henle (see also Limbs of Henle)
anatomy of, 90,91
as countercurrent multiplier, 94
carbon tetrachloride effects on, 100
uranium effects on, 98
Lung architecture
elastic compliance of, 186
mathematical models of, 177
Lung cancer
and cigarette smoking, 16,17
Lung disease (see Asthma; Bronchitis;
Chronic bronchitis; Chronic obstructive
lung disease; Chronic pulmonary disease;
Chronic restrictive lung disease;
Emphysema)
Lung fibrosis (see Fibrosis)
Lung functions
measurements, 186
Lung metabolism
in tissue slices, 197
Lung preparations
as promising assay systems, 203
fibroblasts, pneumocytes, Clara cells,
200-201
in vitro assays, 194-197
metabolism of xenobiotics, 201
methodology for, 184,199,200
Lung ventilation
measurements of, 184-187
Lung volume
compliance of, 186
in chronic obstructive and restrictive lung
disease, 187
measurement of, 184-186
Lymphocytes
chemically induced SCEs in, 27
Macromolecular adducts
(see also DNA adducts)
removal of, 5
Macrophages, alveolar, 177,178,196,202
and silicosis, 192,195
elastase and collagenase release by, 202
217
-------
primary cultures of, 193,194,195,202
properties and functions, 196
short-term cytotoxicity assays, 202
Magnesium,
and cardiovascular diseases, 140
Maize
activating enzymes in, 18
as environmental monitor, 31-32
Mammalian cells in culture
(see also Cell culture)
activation ability, 18
chromosome aberrations, 27
forward mutations in, 23,24
sister-chromatid exchanges in, 25,26
Mammalian embryo culture
(see Embryonic cells)
Mammalian in vivo tests, 24-27
Manganese, and
motor dysfunction, 114
secretonin concentration, 132
Mass culture, 46,47,48,55
Maze
activity assessment in, 118
Mega mouse studies, 87
Membranes (see Cellular membranes, renal;
Plasma membranes, renal)
Mental retardation, 70
and fetal alcohol syndrome, 82
Mercury (see also Methylmercury)
behavioral toxicity from, 108,115
nephrotoxic, 97-98
Mesenchymal derivatives (see Fibroblasts)
Mesotheliomas, lung
and asbestos, 2,179
Metabolic activation (see also Bioactivation)
formation of mutagenic intermediates, 13
in vitro, 17
kinetics of, 4
of cellular macromolecules, 4
of higher plant systems, 23
Metallic compounds
nephrotoxic, 98
transformation and mutagenicity of, 44
Metallothionein
cadmium binding by, 99
Methylbutyl ketone
and nerve conduction, 116
Methylmercury (see also Mercury), and
brain and sensory development, 108,111
electrophysiological assessment of, 114
fetal sensitivity to, 123
neuromuscular lesions, 122
peripheral vision, 113
poisoning epidemics, 114,122
psychophysical assessment, 112
weight loss, 121,132
Methyl methanesulfonate
dose-response curve of, 4
Mice (see also Mouse cells), and
behavioral research with, 116,124
host-mediated assay, 27
MSG-induced brain damage in, 131
teratogenicity assays, 71,73
whole mammal in vivo tests, 24-27
Micronucleus test
for somatic cell risk estimation, 25,26,30
Microsomes (see also Liver microsomal
homogenates)
activating enzymes in, 17
renal, 104
Microwave exposure, behavioral effects,
121,123
Minamata disease (see "poisoning epidemics"
under Methylmercury)
Mitochondria, renal, 104
Mitogens, platelet-derived
in injured arterial wall, 144,148,157
Mitomycin-C
induction of SCEs by, 26,27
Monkeys, in
behavioral toxicology, 112,113,121
teratologic toxicology, 71
Monoclonal hypothesis
of atherosclerosis, 143,153-158, 161
Monocytes, and
immune complex deposition, 148
mitogens, 148
Monooxygenase system
bioactivation of, 71,75,76
species variation in, 75
Monosodium glutamate
and brain damage, 131
Monosomy
and spontaneous abortions, 15
Motor control, 117
Mouse cells
BALB/3T3 and C3H/10T1/2 embryo
fibroblasts, 41
spontaneous transformation of, 39,40
Mouse spot test
forward mutation, 24,25
somatic cell risk estimation, 28,30
Mouse, strain A
tracheal transplant bioassay, 197-199
Mucociliary clearance
and chronic bronchitis, 201
Mucous biosynthesis
in vitro lung studies, 197
Multilocus deletions, 15
for risk assessment, 28,30
in heterozygotes, 28
in Neurospora, 28
strains and tests for, 19-23,28,29
Multivariate analyses, 129,130
Mustard gas
mutagenic, 13
Mutagenicity
and carcinogenicity, 31
Mutant
defined, 12
in short-term assays, 13
Mutants, biochemical
218
-------
in Neurospora, 13
Mutation-prone individuals
identified, 32
Myocardial infarction, and
CO pollution, 139
inhalants, 141-142
Na, K-ATPase
inhibition by platinum, 100
transport-related enzyme, 94,96,105
Nares, tumors of, 179
National Cancer Institute, 30,87
National Center for Toxicologic Research,
87
National Institute of Child Health and
Human Development, 87
National Institute of
Environmental Health Service, 86,87
Toxline, 86
National Institute of Occupational Safety
and Health, 87
National Library of Medicine, 86
National Toxicology Program, 67,83,87
Naturalistic behavior
toxicity assessment by 118,120,128
Neonate
behaviorial teratology in, 122,125
Neoplastic transformation, 40
spontaneous, 39
Nephron (see also Glomerular [adj];
Glomerulus; Kidney; Renal [adj];
Tubular [adj]; Tubules)
anatomy of, 89,90-94
in flounder, 104
in vitro tests, 102-105
micropuncture of, 102
salt and water movement in, 93
Nephrotoxins
tubular and interstitial, 94-100
Nerve conduction velocity, 116-119
Nerve gases
as anticholinesterase agents, 120
Neural tumors
from O '-ethylguanine, 5
Neuromuscular lesions
methylmercury-induced, 123
Neurospora
biochemical mutants of, 13
forward mutations in, 28
metabolic activation of, 18
Newborn monitoring, 78
Nickel
and serotonin concentration, 132
pulmonary, in vitro assays for, 197
Night blindness
and methylmercury poisoning, 112
Nitrogen oxides
in pulmonary function tests, 175,190
Nitrogen washout tests, 187
Nitrosamines and amides, 44,54,59,62
Nonhuman test systems, vi,74 (see also
individual animals)
Nonmelanoma skin cancer, 7,8 (see also
Xeroderma pigmentosum)
Nucleotide excision-repair pathway (see also
Excision-repair pathway)
in Salmonella, 18,32
in Xeroderma pigmentosum, 32
Occupational Safety and Health
Adminstration, 66,67
Oocytes
chromosome aberrations in, 27
Open-field device
in behavioral tests, 123
Organ culture, 46,75
Organic compounds
nephrotoxic, 98,100
Organomercurial fungicides (see Fungicides,
mercurial)
Organophosphate pesticides
(see "organophosphate" under
Pesticides)
Organ transplantation
and atherosclerosis, 143,149
Osmunda regalis
as environmental monitor, 32
Ovigenesis
as in vitro test, 74
Ozone
and lung disease, 175,179
behavioral effects of, 109,120
PAF (see Platelet activating factor)
PAH (p-aminohippuric acid)
renal clearance of, 101
secretion of, 103
uranium-inhibited transport of, 98
Parenchyma! lung diseases
and inorganic dust inhalation, 191
PBBs (see Polybrominated biphenyls)
PCBs (see Polychlorinated biphenyls)
Pea (see Pisum sativum)
Peanut mold
as source of aflatoxin, 17
Perfused kidney
in vitro test, 102
Periarteritis
and hepatitis, 161
Perinate
in behavioral teratology, 122,125
Peripheral artery disease
and diabetes, 137
Personal risk factors
in atherosclerosis, 137-138
Pesticides
activation of, by higher plants, 18,23
and neurological disorders from, 109,115
infertility associated with, 74
organophosphate
as anticholinesterase agents, 120
effect of, on electroencephalogram, 114
219
-------
Phenacetin
toxic effects of, 97
Phenylbutazone
and glomerulonephritis, 96
Phenylketonuria, 73
Photoproduct
pyrimidine dimers, 5
Pink bread mold (see Neurospora)
Pisum sativum
and Mendelian genetics, 12
Placenta, 72,74
chorioallantoic, 81
passage of carcinogens and teratogens, 81
species variation in, 74
Plant poisons
and glomerular lesions, 96
Plants
in short-term tests, 21,28,29,31-32,33
Plaque formation (see also
Arteriosclerotic plaques; Atherosclerotic
plaques)
and monoclonal theory, 150-156
Plasma membranes, renal
subcellular fractionation of, 104
transport proteins in, 92
Platelet activating factor, 145
Platelet function, and
atherosclerosis, 143
immune complex, 146,147
release of mitogens, 144,147-148,157
Platinum
nephrotoxic, 98,99
Pneumoconiosis
and mineral dusts, 179,191
Polybrominated biphenyls
and behavioral toxicology, 130
Polychlorinated biphenyls, and
enzyme induction by, 17
motor activity, 118
Polycyclic aromatic hydrocarbons
activation requiring, 17-18
UV-mjmetic, 5
Polycyclic hydrocarbons, 44,45,59,62,63-66
effect on respiratory tract niucosa, 198
Polyethylene glycol
renal clearance of, 101
Polymorphonuclear leucocytes
in human pulmonary lavage, 193
Population monitoring (see Human
population monitoring)
Precarcinogen (see Procarcinogeh)
Primates, in
psychophysical testing, 128
terotogenicity testing, 74
Procarcinogen
activation of, 41
chemical exposures, 4
defined, 13
Promoters
of carcinogens, 7
of tumors, 4
Promoting factors
assays for, 47-48, 49,55
properties of, 49
Promutagen
defined, 13
host-activated, 16
Proximal tubular function
in vitro tests, 102
in vivo tests, 101
Proximal tubule (see also Tubular
[adj]; Tubules)
anatomy of, 89,91
and heavy metal exposures, 97-100
and organic compound exposures, 100
isolated membranes of, 104
reabsorption and secretion by,
91,93,95,100,103
vulnerability to xenobiotics, 97
Psychophysics, 111-112,128
Public laws and environmental carcinogens,
66
Pulmonary defense mechanisms
alveolar macrophages, 196,202
mucociliary clearance, 201-202
Pulmonary exposure methods, 184
Pulmonary function tests
forceful exhalation, 186
for chronic lung disease, 187
ozone effects on, 179
predictive, 190
with guinea pigs, 190
Pulmonary irritants, short-term test
for, 190
Pulmonary lavage fluids, and
alveolar macrophages, 193
detection of edema, 193
indicators of injury, 195
protein constituents of, 194
Pulmonary lavage studies
and cigarette smoking, 193
Pulmonary tumor bioassay, with
strain A mouse, 200
tracheal transplants, 197-200
Pyrimidine dimers
formation rate, 5
UV-induced, 5,14
Quartz
and collagen biosynthesis, 197
Quinine
and peripheral vision, 112
Rabbits, 71,73
Radiation exposure
and chromosome imbalances, 74
Radiofrequency radiation
and blood pressure, 142
Rat cells
embryonic, for focus assay, 41,48
spontaneous transformation of, 39
virus-infected transformation of, 47
Rat hepatocytes
220
-------
and unscheduled DNA synthesis, 27
Rating scales (see also IQ)
of behavior, 110,126
Rats
and integrative functions tests, 129
brain damage in, 131
for dominant lethal mutation test, 25
in teratogenic assessment, 71,73
Recessive lethals (see Sex-linked
recessive mutation test)
Red blood cell hemolysis assays, 195-196
Reflex testing
reflex modulation, 127,128
toxicity assessments by, 111,114-116, 122
Regulatory processes, v-viii
Renal accumulation
of heavy metals, 94
Renal cellular membrane (see Cellular
membranes; Plasma membranes, renal)
Renal clearance
of test substances, 101
Renal function
assessment of, 100-105
Renal microperfusion
limitations of, 100,102
localization of PAH secretion by, 103
Renal micropuncture
transport function tests, 102
Renal transport
isolated membrane studies of
mechanism, 104
methods of measurement, 100
Renal tubule (see Distal tubule; Proximal
tubule; Tubular [adj]; Tubules)
Repair-deficient strains, in
E. coli, 23,24
Salmonella, 18
Repair levels
saturation of, 5
Repair pathways (see Excision-repair
pathway; Nucleotide excision-repair
pathway)
Repair-proficient strains
E. coli, 23,24
Salmonella, 18-19
Repetitive stimulus
tumorigenesis from, 7
Reproductive behavior, 118,120
Reproductive loss
incidence of, 69,70
Respiratory rate
and industrial chemicals, 190
Respiratory tract carcinogenesis
and cigarette smoke, 198
Respiratory tract epithelium
(see Cell culture)
Restrictive lung disease
(see Chronic restrictive lung disease)
Reverse mutation tests with
fungi and higher plants, 28,29
Salmonella his mutants, 18,23
Rh factor
maternal immunization to, 70
Righting reflexes, 114
Risk assessment, vi-viii,3,87,88
parameters of, 9
quantitative, 3
versus benefits, 7
Risk estimation, 27,28-30
Risk factors in atherosclerosis
cigarette smoking, 158
diabetes mellitus, 159-160
diet and hyperlipemia, 136,138,156-157
hypertension, 158-159
sex, age and physical activity, 137,160
Rodent liver homogenates (see also Liver
microsomal homogenates)
in vitro activation by, 17
Rodents (see also Hamster cells; Mice;
Rats), 71
R plasmid
in Salmonella, 19
Rubella syndrome
congenital, 70,73,82
S-9 fraction
induced, 18
in vitro metabolic activation by, 17
S-9 homogenates (see Liver microsomal
homogenates; S-9 fraction)
Saccharomyces cerevisiae
gene conversion in, 23,28
metabolic activation by, 18
Salivary gland cells (see Cell culture)
Salmonella
coat-deficient mutants, 18
excision-repair deficient mutants, 18
his mutants, 18
microsome test, 31,32
mutagenicity test, 40,44,45,49,50,
53,59-61,71
plasmid-bearing mutants, 19
reverse-mutation test, 18,23
SCE (see Sister chromatid exchange)
Schedules of behavioral reinforcement,
121,129
School performance
and lead, 125,126
Scintigraphy, 162
Sense organs
and toxic effects, 111-114
Serotonin
and liver function, 132
antigen excess, 145
effect of DDT and heavy metals on, 132
Serum cholesterol
in ischemic heart disease, 137,141
Serum sickness
and arterial injury, 145,146
Sex
as atherosclerotic risk factor, 160
Sex-linked recessive lethal mutation test
221
-------
in Drosophila, 12,24,30
Sickle cell anemia, 15
Silica, and
collagen biosynthesis stimulation, 197
cytotoxicity, 192,195,196
interaction with macrophages, 195
Silicosis
animal models for, 192
autoimmune phenomena in, 190
from dust inhalation, 191
pathogenesis of, 192
Silver
nephrotoxic, 98
Sister chromatid exchange
as DNA-damage test, 26,27,32
Skin cancer (see also Nonmelanoma skin
cancer; Xeroderma pigmentosum)
and UV in sunlight, 7
Skin sensitivity
as toxicity assessment test, 111
Smoking (see also Cigarette smokers)
and early menopause, 81
Smooth muscle cell proliferation
and mitogens, 148
Sodium chloride
as teratogen, 71
Soft water
and Sudden Death, 140
Solvents
behavioral toxicity of, 108,111,116
carcinogenicity of, 44
Somatic cells
forward mutations in, 24
short-term tests with, 20,22,29
Species variation, in
development of kidney, 106
drug clearance, 72,78
maternal P450, 75
teratogenic response, 71,72
Specific locus test, and
germ-cell risk estimation, 28-30
in vivo forward mutations, 24
Spermatogenesis, in
in vitro tests, 74
in vivo tests, 25,26-27,30
Spermatoxic agents
assayed in mammals, 26
Spermatozoa
in epidemiological studies, vii,74
Sperm YFF test
for aneuploidy, 26
Spontaneous abortion (see Abortion)
Spontaneous transformation
(see Transformation)
Standard setting, vii.viii
Startle response
toxicity assessment method, 111
Strains, microbial
development of, 16-19
Stress
and coronary heart disease, 142
Strokes
induction by toxins, 11
Subcellular fractions
and heavy metals, 95
in vitro renal tests, 104
Subhuman primates (see individual
animals; Nonhuman test systems;
Primates)
Sucrose
as teratogen, 71
Sudden Death, and
cigarette smoking, 137,158
. diabetes, 138
water quality, 140
Sulfhydral groups
in kidney, 96
sulfonamides
and glomerulonephritis, 96
Sulfuric acid
in vitro, lung studies for, 197
mucociliary clearance of, 201
Sulfur dioxide
cocarcinogenic in trachea!
transplants, 199
Sulfur oxides
and lung disease, 175
pulmonary function tests for, 187,190
Sunlight
carcinogenic wavelengths of, 5,7
Supermutagens, 13
Susceptible populations, 32
sex differences and lung disease, 176
Swordfish
methylmercury body burden, 112
Syrian hamster (see under Hamster cells)
Tar
from cigarettes, mutagenic, 31
Target organ concept, 94-96
Target tissue
DNA adduct concentrations in, 5,6
Teratogenic agent (see Teratogens)
Teratogenic outbreak
epidemic, 82-83
Teratogenic susceptibility, 71,72,78
and stage of development, 71
Teratogens, 77
defenses against, 78,79
maternal blood concentrations of, 71
selection for testing, 87
species variation in response, 71
Test-agent administration
time periods for, 74
Testosterone secretion
cadmium effects on, 118
Thalidomide, 71,78,83
Thallium
nephrotoxic, 98
Three-tier approach
for assessing genotoxicity, 30
Threshold dose, 1,9
222
-------
Thrombogenesis
and heart attack, 162
Thyroid tissue
in organ culture, 75
Toluene
and behavioral toxicity, 108
Tooth lead
and teachers' rating scales, 125-126
Toxic Substances Control Act,
34,87,110
Toxline, 86
Trace metals
and cardiovascular mortality, 140-141
Tracheal instillation
pulmonary exposure method, 184
Tracheal rings
for in vitro lung studies, 197
Tracheal transplants
in carcinogenesis, 197-200
Tradescantia
as environmental monitor, 32,33
Transformation
of hamster embryo cells, 44,59-61
spontaneous, 39,42
virus-induced, enhancement of, 42
Transformed cells
colony assay, 39,43
properties of, 39,40,41
Translocation, chromosomal, 15
Translocation, heritable
in mouse germ cells, 26
in risk assessment, 30
Transmembrane ion gradients, 100
Tremor
mercury vapor-induced, 114,115
Triangulation
in teratology, 81-82
Trichloroethylene
behavioral toxicity from, 108
Triethyl tin
behavioral toxicity from, 111
Trisomy
and Down's syndrome, 15
Tubular cells
anatomy of, 96
ischemic injury to, 97
Tubular filtrate
antidiuretic hormone effect on, 94,95
heavy metal impairment by, 98-100
Tubular segments, isolated
in vitro renal function test,
102-103
Tubules (see also Distal tubule;
Proximal tubule, Tubular [adj])
reabsorption and secretion by, 92-94
Tumor promoters (see Promoters)
Tuna
methylmercury body burden, 112
Turner's syndrome
monosomic condition, 15
Ultimate carcinogen
defined, 13
Ultimate mutagen
defined, 13
formation by host's enzymes, 16
Ultrasonography, 162
Ultraviolet light
biochemical mutation induction, 13
skin cancer, 5
Unscheduled DNA synthesis assay, 27
Uranium
hemodynamic changes from, 98
tubular lesions from, 98
Urinary bladder epithelium
(see Cell culture)
Urine
analysis, 100,101
cigarette smokers, 32
final composition of, 93,94
in human population monitoring, 32
renal function test, 101
Uterine leiomyoma
and monoclonal hypothesis, 153
UV (see Ultraviolet light)
UV-mimetic chemicals
polycyclic aromatic hydrocarbons as, 5
Vaccines
and glomerular lesions, 96
Vaginal carcinoma
and diethylstilbestrol, 78
Vasectomy
and atherosclerosis, 148
Vasoactive amines
and immune complexes, 146
Vibration white finger, 142
Vigilance performance
carbon monoxide effect on, 109
Virus infections
and atherosclerosis, 146,151
Vision
perimetry testing of, 112,113
psychophysical studies of, 111
Vitamin deficiency
teratologic effects of, 71
Water hardness
and cardiovascular mortality, 140-141
Weight loss, from
acrylamide and methylmercury
poisoning, 121
carbon disulfide, 132
heavy metals, 132
Xenobiotics, excretion of, 89
Xeroderma pigmentosum
(see also Nonmelanoma skin cancer)
autosomal recessive disease, 32
repair-deficient condition, 5,7,32
Xeroderma pigmentosum cells
effect of UV-mimetic chemicals in, 5
223
-------
X rays
induction of mutations by, 12,13
Y chromosome
fluorescence test for aneuploidy, 26
Zinc/copper ratios
and coronary heart disease, 141
224
•ft-U.S. Government Printing Office:1983--659-095/0596
------- |