Short Term Bioassays in the Analysis of Complex Environmental Mixtures II


EPA-600/9-82-004
March 1982
Environmental Mixtures II
Edited by
MICHAEL D. WATERS
SHAHBEGS. SANDHU
JOELLEN LEWTAS HUISINGH
LARRY CLAXTON
and
STEPHEN NESNOW
U. S, Environmental Protection Agency
Research Triangle Park, North Carolina
PLENUM PRESS • NEW YORK AND LONDON

-------
ii
DISCLAIMER
This report has been reviewed by the Health Effects Research
Laboratory, U.S. Environmental Protection Agency, and approved for
publication. Mention of trade names or conmercial products does
not constitute endorsement or recommendation for use.

-------
iii
FOREWORD
The many benefits of our modern industrial society are
accompanied by certain hazards. Careful assessment of the relative
risk of existing and new man-made environmental hazards is
necessary for the establishment of sound regulatory policy. These
regulations serve to enhance the quality of our environment in
order to promote the public health and welfare and the productive
capacity of our nation's population.
The Health Effects Research Laboratory, Research Triangle
Park, North Carolina, conducts a coordinated environmental health
research program in toxicology, epidemiology, and clinical studies
using human volunteer subjects. These studies address problems in
air pollution, non-ionizing radiation, environmental
carcinogenesis, and the toxicology of pesticides as well as other
chemical pollutants. The Laboratory develops and revises air
quality criteria documents on pollutants for which national ambient
air quality standards exist or are proposed, provides the data for
registration of new pesticides or proposed suspension of those
already in use, conducts research on hazardous and toxic materials,
and is preparing the health basis for non-ionizing radiation
standards. Direct support to the regulatory function of the Agency
is provided in the form of expert testimony and preparation of
affidavits as well as expert advice to the Administrator to assure
the adequacy of health care and surveillance of persons having
suffered imminent and substantial endangerraent of their health.
The Second Symposium on the Application of Short-term
Bioassays in the Fractionation and Analysis of Complex
Environmental Mixtures arose out of the recent developments in the
methodology for the chemical analysis and bioassay of complex
environmental mixtures from a variety of media. The present
proceedings reflects the state-of-the-art in this promising area of
research for the identification and evaluation of potential human
health hazards.
F. Gordon Hueter, Ph.D.
Director
Health Effects Research Lahoratorv

-------
iv
PREFACE
More Chan one hundred short-term bioassays are now available
for detecting the toxicity, mutagenicity, and potential
carcinogenicity of chemicals. These bioassavs were developed and
validated with individual compounds, and their principal
application was perceived to be in evaluating the health hazard of
such materials. However, man is rarely exposed to single
chemicals; his exposure to hazardous chemicals is more commonly a
multifactorial phenomenon. Although chemical analysis can be used
to detect known hazardous compounds, it would be a staggering and
expensive task to analyze large numbers of samples for all known or
suspected hazardous constituents. Furthermore, the biological
activity of a complex mixture cannot be reliably predicted from
knowledge of its components. On the other hand, bioassays alone
cannot tell us which components of complex mixtures are responsible
for the biological activity detected. Thus, cost effectiveness and
technical feasibility dictate stepwise and perhaps iterative
application of both chemical and biological methods in evaluating
the health effects of complex environmental mixtures.
Through the coupling of reliable biological detection systems
with methods of chemical fractionation and analysis, it is
frequently possible to isolate the individual chemical species that
show biological activity. Initially, complex mixtures may be
separated and bioassayed in carefully defined chemical fractions.
The results of such short-term screening bioassays then may be used
to guide the course of further fractionation and to determine the
need for more stringent and comprehensive biological testing.
Another approach to the screening of complex environmental
mixtures for health effects involves the use of in situ bioassavs.
The biological effects of environmental chemicals are influenced by
a combination of environmental factors that cannot be completely
reproduced in the laboratory. By maintaining test organisms at
sites to be monitored, one can rapidly identify potentially
hazardous environmental mixtures that warrant further
invest igat ion.
These are the proceedings of the Second Symposium on the
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, held in Williamsburg,

-------
PREFACE
v
VA, March 4 through 7, 1980, and sponsored by the U.S.
Environmental Protection Agency. The first symposium of this
series, also held in Williamsburg, in February, 1978, combined
accounts of the latest methods for collection and chemical analysis
of complex samples with discussions of current research involving
the use of short-term bioassays in conjunction with fractionation
and analysis of such mixtures. The emphasis of the present
proceedings is on the application of these methods in testing a
variety of media, including ambient air, drinking water and aoueous
effluents, terrestrial systems, mobile-source emissions, and
stationary-source emissions and effluents. The critical problem of
human health hazard and risk assessment is also addressed.
We hope that this volume will help to consolidate our
knowledge of the techniques and applications of chemical analysis
and bioassay of complex environmental mixtures and that it will
provide direction for further research in this area.
Michael D. Waters
Shahbeg S. Sandhu

-------
vi
ABSTRACT
The present proceedings of the U.S. Environmental Protection
Agency's Second Symposium on the Application of Short-term
Bioassays in the Fractionation and Analysis of Complex
Environmental Mixtures, held in Williamsburg, VA, March &-7 , 1980,
includes 37 papers as well as the Keynote Address. The papers are
divided according to the environmental media wherein short-term
bioassays are applied—ambient air, water, and soil—and the
sources of environmental pollution—mobile source emissions,
stationary source emissions, and industrial emissions and
effluents. A separate section is devoted to the problems of health
hazard and risk assessment.

-------
vii
CONTENTS
Keynote Address
Vilma Hunt
SESSION 1: AMBIENT AIR
Bioassay of Particulate Organic Matter from Ambient Air*
Joellen Lewtas Huisingh
Collection, Chemical Fractionation, and Mutagenicity
Bioassay of Ambient Air Particulate*	21
Alan Kolber, Thomas Wolff,
Thomas Hughes, Edo Pellizzari,
Charles Sparacino, Michael Waters,
J. Lewtas Huisingh, and Larry Claxton
Evaluation of Collection and Extraction Methods for
Mutagenesis Studies on Ambient Air Pariculate*	45
R. Jungers, R. Burton, L. Claxton, and
J. Lewtas Huisingh
*Invited paper.
TContributed paper
based on poster presentation.

-------
\11l
Integration of the Ames Bioassay and Chemical Analyses
in an Epidemiological Cancer Incidence. Study*
C. Peter Flessel, Jerome J. Wesolowski,
SuzAnne Twiss, James Cheng, Joel Ondo,
Nadine Monto, and Raymond Chan
Mutagenicity of Airborne Particulate Matter in
Relationship to Traffic and Meteorological Conditions*
Ingrid Alfheim and Mona Miller
Dectection of Genetically Toxic Metals by a Microtiter
Microbial DNA Repair Assay*
Guylyn R. Warren
A Culture System for the Direct Exposure of Mammalian
Cells to Airborne Pollutantst
Ronald E. Rasmussen and
T. Timothy Crocker
SESSION 2: DRINKING WATER AND AOUEOUS EFFLUENTS
Is Drinking Water a Significant Source of Human Exposure
to Chemical Carcinogens and Mutagens?*
Richard J. Bull
Alternative Strategies and Methods for Concentrating
Chemicals from Water*
Frederick C. Kopfler
Detection of Organic Mutagens in Water Residues*
John C. Loper and M. Wilson Tabor

-------
CONTENTS
Short-term Methods for Assessing la Vivo Carcinogenic
Activity of Complex Mixtures*
Michael A. Pereira and Richard J. Bull
The Initiating and Promoting Activity of Chemicals
Isolated from Drinking Waters in the Senear Mouse:
A Five-city Study*
Merrel Robinson, John W. Glass,
David Cmehil, Richard J. Bull, and
John G. Orthoefor
Aqueous Effluent Concentration for Application to Biotest
SystemsT
William D. Ross, William J. Hillan,
Mark T. Wininger, JoAnne Gridly,
Lan Fong Lee, Richard J. Hare, and
Shahbeg S. Sandhu
SESSION 3: TERRESTRIAL SYSTEMS
Potential Utility of Plant Test Systems for Environmental
Monitoring: An Overview*
Shahbeg Sandhu
Arabidopsis Assay of Environmental Mutagens*
G.P. Re'dei
Soybean System for Testing the Genetic Effects of
Industrial Emissions and Liquid Effluents*
Baldev K. Vig
Mutagenicity of Nitrogen Compounds from Synthetic Crude
Oils: Collection, Separation, and Biological Testing*
T.K. Rao, J.L. Epler, M.R. Guerin, B.R. Clark,
and C.-h. Ho

-------
CONTENTS
The Detection of Potential Genetic Hazards in Using Plant
Cytogenetics and Microbial Mutagenesis Assays*	253
Milton J. Constantin, Karen Lowe,
T.K. Rao, Frank W. Larimer, and
James L. Epler
SESSION 4: MOBILE SOURCES	267
Short-term Carcinogenesis and Mutagenesis Bioassays of
Mobile-source Emissions*	269
Joellen Lewtas Huisingh
Tumorigenesis of Diesel Exhaust, Gasoline Exhaust, and
Related Emission Extracts on Senear Mouse Skin*	277
Stephen Nesnow, Larry L. Triplett, and
Thomas J. Slaga
Bacterial Mutagenesis and the Evaluation of Mobile-source
Emissions*	299
Larry Claxton and Mike Kohan
Comparison of the Mutagenic Activity in Carbon
Particulate Matter and in Diesel and Gasoline
Engine Exhaust*	319
Goran Lofroth
Mutagenic Effects of Environmental Particulates in the
CHO/HGPRT System!	337
G.M. Chescheir III, Neil E. Garrett,
John D. Shelburne, J. Lewtas Huisingh,
and Michael D. Waters

-------
CONTENTS
A Preliminary Study of the Clastogenic Effects of Diese
Exhaust Fumes Using the Tradescant ia
Micronucleus BioassayT
Te-Hsiu Ma, Van A. Anderson, and
Shahbeg S. Sandhu
Ability of Liver Homogenates, and Proteins to Reduce th
Mutagenic Effect of Diesel Exhaust Particulatest
Yi Y. Wang and Eddie T, Wei
SESSION 5: STATIONARY SOURCES
Bioassays of Effluents from Stationary Sources:
An Overview*
R.G. Merrill, Jr., W.W. McFee, and
N.A. Jaworski
Coal Fly Ash as a Model Complex Mixture for Short-term
B ioassay*
Gerald L. Fisher, Clarenc E. Chrisp,
and Floyd D. Wilson
Possible Effects of Collection Methods and Sample
Preparation on Level 1 Health Effects Testing of
Complex Mixtures*
D.J, 3rusick
Biological Monitoring of Fluidized Bed Coal Combustion
Operations I. Increased Mutagenicity During Periods of
Incomplete Combustion*
H.E. Kubitschek, D.M. Williams, and
F.R. Kirchner

-------
xii	CONTENTS
Biological Monitoring of Fluidized 3ed Coal Combustion
Operations II. Mammalian ResDonses Following Exposure
to Gaseous Effluents'*	421
F.R. Kirchner, D.M, 3uchholz, V.A. Pahnke,
and C.A. Reilly, Jr.
In Vitro and In Vivo Evaluation of Potential Toxicity of
Industrial Particlest	431
Catherine Aranyi, Jeannie Bradof,
Donald E. Gardner, and
J. Lewtas Huisingh
Mutagenicity and Carcinogenicity of a Recently
Characterized Carbon Black Adsorbate:
Cyclopenta(cd)pyrenet	445
Avram Gold, Stephen Nesnow, Martha M. Moore,
Helen Garland, Caynelle Curtis, Barry Howard,
Deloris Graham, and Eric Eisenstadt
Mutagenicity of Coal Gasification and Liquefaction
Productst	461
Rita Schoeny, David Warshawsky,
Lois Hoi 1ingsworth, Mary Hund, and
George Moore
SESSION 6: HAZARD ASSESSMENT	477
The Role of Short-term Tests in Assessing the Human
Health Hazards of Environmental Chemicals: An Overview*	479
Michael D. Waters
The International Program for the Evaluation of
Short-term Tests for Carcinogenicity (IPESTTC)*	485
Frederick J. de Serres

-------
CONTENTS	xiii
Sperm Assays in Man and Other Mammals as Indicators of
Chemically Induced Testicular Dysfunction*	495
Andrew J, Wyrobek
Assessing Carcinogenic Risk Resulting from Complex
Mixtures*	507
Roy E. Albert

-------
XIV
ACKNOWLEDGMENTS
We would like to thank Wendy A. Martin and Karen L. Spear of
Kappa Systems, Inc., for coordinating the symposium. Our thanks
are extended to Olga Wierbicki, Susan Dakin, Leslie Silkworth,
Barbara Elkins , and Priscilla Skidmore of NorthroD Services, Inc.,
for editing the proceedings and preparing the final manuscript.

-------
KFYNfOTF ADDRFSS
JRb	JL X \>»r mMm JKLmmarn JL	jKLmmtit %¦«/ \tm0r
/•

-------
Intentionally Blank Page

-------
KEYNOTE ADDRESS
Vilma Hunt
Office of Health Research
U.S. Environmental Protection Agency
Washington, District of Columbia
I am delighted to be here today addressing this symposium on
the application of short-term bioassays to the study of complex
environmental mixtures. I can think of no other area that is more
representative of the significant advances that have been made in
the last decade in the environmental health sciences field.
The U.S. Environmental Protection Agency (EPA) has been faced
since its inception with the responsibility to control harmful
environmental compounds. One of the highest priorities of EPA's
research and development program is the protection of health
through the identification and control of toxic substances. Our
responsibility in the Office of Health Research is to identify,
qualitatively and quantitatively, the harmful health effects
associated with environmental agents. In our effort to fulfill
this responsibility, we have repeatedly been slowed by the
limitations of the available testing procedures, as well as by a
relative lack of understanding of health effects themselves.
One of the major problems we have faced and still face is the
complexity of the agents of concern. Complex mixtures—whether
industrial effluents or emissions, or ambient environmental
media—are generally entities composed of hundreds of compounds.
Further, these mixtures are often poorly defined and of
continuously changing chemical composition. As such, complex
mixtures present special challenges in quantitation and assessment
that require considerable effort to overcome.
We are also faced with limitations in our understanding of the
effects we must test for. The effects of concern—cancer,
mutagenesis, teratogenesis, and other toxic impacts—are often
3

-------
4
VILMA HUNT
chronic and slow to develop. The science behind these effects is
not well understood. And, significantly, we are often looking for
effects that are expected to be of low probability for an
individual, but of large impact for the exposed population.
The magnitude of the environmental regulatory task that EPA
and other regulatory agencies face is underscored by the
astronomical numbers that come up whenever environmental assessment
is discussed. There are currently more than 4,000,000 known
chemical compounds; thousands more are being discovered each year;
70,000 are in common use, produced and distributed by some 115,000
industries and firms. Billions of gallons of industrial effluents
are discharged into our lakes, rivers, and oceans each year. The
1977 emission of criteria air pollutants to the atmosphere was 190
million tons, a quantity that does not include the other
unregulated and potentially hazardous particles, gases, and
aerosols emitted each year. In addition, several billion tons of
unwanted solid waste—some harmful, some innocuous—are disposed of
each year.
If, to fulfill our health effects assessment responsibility,
we had "only" to establish the toxic potential of some 70,000
compounds, we would be faced with a staggering assignment. When
consideration is given to the fact that these compounds appear in
the environment in diverse combinations, in a variety of media, and
often in miniscule quantities that are difficult to collect and
analyze, the staggering assignment suddenly appears overwhelming.
Our efforts to assess toxic potential include a number of
undertakings in laboratory, clinical, and epidemiological research.
Historically, EPA and other government regulatory agencies have
favored established whole-aniraal methods as the standard of
reference to establish carcinogenicity and other toxic effects.
Our resources, however, to conduct whole-animal studies are
limited. Indeed, the world laboratory capacity for conducting
these experiments has been estimated at 500 compounds per year, a
small number when compared with the number of compounds needing
assessment. Additionally, the long-term nature of whole animal
studies poses other problems for regulators faced with the need to
make timely regulatory decisions.
Data derived from epidemiological studies are, of course,
those most adequate from a regulatory standpoint. These data are
scarce, however, and we are again faced with fiscal limitations in
collecting sufficient information on large numbers of compounds,
not to mention the inherent difficulties in collecting meaningful
information for large populations exposed to a myriad of compounds.
Recognizing our limitations in whole-aniraal and
epidemiological studies, we have devoted particular attention and

-------
KEYNOTE ADDRESS
5
considerable resources to the research and development of short-
term tests over the last several years. Our interest in short-term
bioassays was sparked by recognition of their potential value to a
health effects assessment program. Our investment has been
rewarded by highly encouraging results.
As a rapid, effective, and inexpensive means to identify the
impact of complex mixtures, short-term testing can play a critical
role in the monitoring of the environmental media for presumptive
health hazards. Today, any program aimed at identifying and
reducing production and release of large numbers of hazardous
agents must take advantage of short-term bioassays to set
priorities for further evaluation by conventional toxicological,
clinical, and epidemiological investigations. By efficiently using
short-term bioassays and through the development of approaches that
combine the use of various bioassay systems, we have begun to
screen large numbers of potentially harmful compounds in a
systematic and effective manner.
As this symposium reflects, the application of short-term
bioassays to the assessment of complex mixtures has been developed
hand-in-hand with the application of state-of-the art analytical
chemistry techniques. The iterative application of chemical and
biological analytical tools has greatly expanded the number of
environmental pollutants for which biological hazards have been
identified. Results of many of these studies will be reported on
during the next several days. However, the number of assessed
pollutants is still only a small fraction of those requiring
assessment. The lack of information on chemical composition and
biological activity continues to constitute a major barrier to the
assessment of human health hazards from complex environmental
mixtures.
At the first symposium on The Application of Short-term
Bioassays in the Fractionation and Analysis of Complex
Environmental Mixtures, Dr. Michael Waters of our Research Triangle
Park Laboratory dedicated the meeting "to the concept that the
joint application of state-of-the art biological and chemical
analytical techniques is the appropriate approach in environmental
research." I am happy to note that the multidisciplinary approach
being pursued in this area is representative of a widespread
evolution in the approach to environmental problems on many fronts.
We are moving away from the segmented, narrow approaches of the
past to joint undertakings that direct the activities of different
disciplines toward specific, unified goals.
The past two years have seen great strides in the short-term
bioassay field and in the analytical area. The symposium last year
was devoted to the "nuts and bolts" of this relatively new field.
The majority of the presentations dealt with the bioassay

-------
6
VILMA HUNT
techniques, the sampling methods, and the analytical procedures.
This year, the symposium is organized to report on the application
of these techniques—a reflection of the development of this field
from solely an object of research to a means for research.
In particular, the topics to be considered over the next
several days include the utilization and application of short-term
testing to different media: ambient air, drinking water and
effluents, terrestrial media, mobile-source emissions, and
stationary-source emissions and effluents. The Friday morning
session is devoted to an examination of the role of short-term
testing in hazards assessment, a topic of concern and interest to
all of us in the post-Love Canal era when we are confronted with
resolving the buried mistakes of the past.
In addition to an increase in the application and utilization
of short-term procedures, the past two years have also been marked
by the validation of many of the short-term bioassay procedures.
The validation of short-term testing was essential if short-term
testing was to find a meaningful role in environmental health
assessment. Considerable effort has been devoted to developing the
needed data, and although much work remains to be done, the initial
results are optimistic. EPA recently established the Gene-Tox
program to further the evaluation and validation of short-term
bioassays. This program, which is evaluating 27 different
short-term systems, will play an important role in identifying
aspects of short-term tests that require further development and
validation. Information from the Gene-Tox evaluations will be used
to direct future research programs.
In conclusion, the coupling of short-term bioassays with
state-of-the-art chemical analysis techniques is an exciting and
rapidly evolving field—one that offers the potential to resolve
questions of nontoxicity quickly and to provide a scientific basis
for properly allocating our resources among many studies of
potentially hazardous agents. If the past is an augury of the
future, the challenges that face this still young and rapidly
growing field will be met. I am enthusiastic about the recent
developments in this field and the developments that are
forthcoming. The importance of this research to the future of
environmental health assessment cannot be overstated.
Thank you.

-------
SESSION 1
AMBIENT AIR

-------
Intentionally Blank Page

-------
BIOASSAY OF PARTICULATE ORGANIC MATTER FROM AMBIENT AIR
Joellen Lewtas Huisingh
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION
The influence of industrialization and consequent increased
concentration of urban particulate matter on the incidence of
cancer has long been a concern (Kotin and Falk, 1963; Carnow and
Meier, 1973). The first bioassays used to evaluate complex ambient
air samples were whole-animal carcinogenesis bioassays (Leiter
et al., 1942; Hueper et al., 1962). In these studies, organic
extracts of urban particulate matter were found to be carcinogenic
in rodents. Such organic extracts have also been shown to
transform rodent embryo cells in culture (Freeman et al., 1971;
Gordon et al., 1973). Carcinogenic polycyclic aromatic
hydrocarbons (PAH), such as benzo(a)pyrene, were detected in these
extracts; however, these compounds did not account for all of the
carcinogenic activity reported.
The development of the Ames Salmonella typhimurium mutagenesis
bioassay (Ames et al., 1975) provided a simpler, more sensitive,
and faster bioassay for potential carcinogenic activity that could
be applied to air samples collected by conventional techniques.
The initial applications of this bioassay to ambient air
particulate organic matter (Tokiwa et al., 1976; Pitts et al.,
1977; Talcott and Wei, 1977) stimulated research in the following
areas:
1) improvement in sample collection, extraction, and bioassay
methodology;
9

-------
10
JOELLEN LEWTAS HUISIXGH
2)	characterization and identification of potential classes
of carcinogens and specific carcinogens present in ambient
air particles; and
3)	evaluation of emission sources and atmospheric conditions
responsible for the observed mutagenicity in urban air
particulate.
This overview addresses these areas of research and summarizes
our current understanding of the mutagenicity of particulate
organic matter found in ambient air.
ADVANCES IN SAMPLE COLLECTION AND EXTRACTION
The first reported studies on the mutagenicity of particulate
organic matter from ambient air used high-volume samplers to
collect air particles on glass fiber filters (Tokiwa et al., 1976;
Pitts et al., 1977; Talcott and Wei, 1977). This sampler collects
both respirable particles (< 5 pm) and non-respirable large
particles. In these studies, the organics were extracted from the
particles with either methanol (Tokiwa, 1976), acetone (Talcott and
Wei, 1977), or a mixture of methanol, benzene, and dichloromethane
(1:1:1)(Pitts et al., 1977). Since high-volume samplers are
widely used in air-monitoring programs to determine total suspended
particulate (TSP) levels, these studies provided comparative
mutagenicity data for different sites.
Although high-volume samplers provide the simplest, and for
many investigators the only, method available for collection of air
particles, this method presents several serious disadvantages. The
most consequential disadvantage is that respirable particles are
collected simultaneously with larger particles. In cases where the
smaller respirable particles are considerably more mutagenic than
the larger particles, the larger particles dilute the overall
mutagenicity of the sample, thereby biasing the analysis of the
particle composition to which the human lung is exposed.
Other potential disadvantages are due to the large volume of
air being drawn continuously over collected particles. Samples
collected by this method may lose more volatile organics by
evaporation. The organics present on the particles are also
potentially subject to reactions with nitrogen dioxide (NO2), ozone
(O3), or peroxyacetyl nitrate (PAN), which are all present in urban
air. Pitts et al. (1978a, b) have shown that PAH (e.g.,
benzo(a)pyrene and perylene) directly coated onto glass fiber
filters reacted with NO2, O3/ and PAN, as well as ambient
photochemical smog, to form several direct-acting mutagens
(mutagens that do not require an exogenous microsomal activation
system). Although these reactions have not been shown to occur to

-------
PARTICULATE ORGANIC MATTER FROM AMBIENT AIR
11
PAH adsorbed on the surface of air particles, potential surface
reactions during filtration still must be taken into consideration
when interpreting studies using filtration for sample collection.
A recent modification of the standard high-volume sampler, the
size-selective inlet (SSI) high-volume sampler, collects only
particles < 15 ym (the definition of "inhalable particles"). This
sampler excludes the larger particles, which are not normally
inhaled.
A sampling device that does collect particles in separate size
fractions is the cascade impactor (Andersen, 1966). These samplers
use a series of plates with either holes or slots offset at each
stage to collect separate size fractions ranging from < 2 yn to >
7 pin. The cascade impactor is generally attached to a high-volume
sampler that collects the smallest particles by filtration.
Teranashi et al. (1977), using an Andersen high-volume cascade
impactor, found that the organics from particles < 1.1 pm,
collected on the backup filter, were significantly more mutagenic
than the organics from the larger particles. Pitts et al. (1978b)
reported similar findings using a Sierra high-volume cascade
impactor in downtown Los Angeles. This method, while providing a
size-fractionated sample for bioassay, still employs filtration to
collect the smallest particles, which contain most of the mutagenic
components•
In order to collect larger quantities of size-fractionated air
particulate matter for biological studies, the U.S. Environmental
Protection Agency (EPA) had a Massive Air Volume Sampler (MAVS)
designed and fabricated by Henry and Mitchell (1978) that does not
require filtration. The MAVS employs two impactors that collect
3.5- to 20-|jm and 1.7- to 3.5-pm particles, followed by an
electrostatic precipitator (ESP) that collects the particles < 1.7
um. In our initial studies with the MAVS at a Los Angeles freeway
site, we found the particles with mean diameters < 1.7 pm to be
significantly more mutagenic than the larger particles (Figure 1).
However, when the ESP was charged to maximize collection efficiency,
as much as 0.05 ppm of O3 was measured at the blower outlet of the
sampler (Mitchell et al., 1978). Jungers et al. (1980) evaluated
the effect of O3 under the MAVS operating conditions on both the
mutagenicity and chemical composition of the particulate organic
matter collected in the ESP. These studies found that under these
operating conditions, the O3 did not significantly affect the
mutagenicity or the PAH content of the organics.
A wide range of extraction methods have been used to remove
the organics from air particles for bioassay. Solvents employed
range from the nonpolar solvents cyclohexane (Miller and Alfheim,
1980) and benzene (Teranashi et al., 1977) to acetone (Talcott and
Wei, 1977) and the more polar solvent methanol (Tokiwa et al.,

-------
1000
800
600
UJ
a.
200
38
3.5 - 20.0
PARTICLE DIAMETER (/jmt BY MASSIVE VOLUME SAMPLER
Figure 1. Mutagenicity of air particulate at the Los Angeles
freeway site (upwind) as a function of particle size.

-------
PARTICULATE ORGANIC MATTER FROM AMBIENT AIR
13
1976). Since different solvents will preferentially remove
different constituents from the particles, the method of extraction
can significantly alter the resulting composition and mutagenicity
of the organics. Jungers et al. (1980), after evaluating seven
solvent systems, found that dichloromethane extraction resulted in
the most mutagenic extractible organics with a minimum of inorganic
anions present.
CHARACTERIZATION AND IDENTIFICATION OF POTENTIAL CARCINOGENS
The application of the Ames typhimurium plate-incorporation
assay (Ames et al., 1975) with multiple tester strains, used in the
presence and absence of a metabolic activation system, provides an
initial characterization of potential carcinogens present. The
organics from air particles generally show mutagenic activity only
in the tester strains susceptible to frameshift mutations (e.g.,
TA1538, TA1537, TA98, and TA100, but not TA1535). Mutagenic
activity is usually observed in the absence of a metabolic
activation system, indicating the presence of direct-acting
mutagens. When metabolic activation is added, certain air samples
show significant increases in mutagenicity (Talcott and Wei, 1977;
Pitts et al., 1977), while other samples show no increase in
activity, and even occasionally a decrease (Talcott and wei, 1977).
Upon fractionation, however, air samples generally have been shown
to contain both direct-acting mutagens and mutagens that require
the addition of metabolic activation. Table 1 shows the response
of five S. typhimurium tester strains to a typical air sample, with
and without metabolic activation, using the < 1.7-pm particles
collected with a MAVS in Birmingham, AL.
Table 1. Specific Mutagenic Activity of Air Particulate Extract
(< 1.7 ym)
Revertants/100 ug Organics
Tester
Strain
Without
S-9 Activation
With
S-9 Activation
TA1538
TA1537
TA98
81
59
121
176
121
34
188
245
TA100
TA1535

-------
14
JOELLEN LEWTAS HUISINGH
Bioassay-directed chemical fractionation is an increasingly
powerful tool for the identification of potential carcinogens in
complex mixtures. This technique has been employed to characterize
and eventually identify potential carcinogens in cigarette smoke
(Swain et al., 1969), synthetic fuels (Epler, 1980), and diesel
emissions (Huisingh et al., 1978). Until the recent development
and improvement of air particle collection techniques, such studies
on particulate organic matter in air were severely limited by the
amount of sample collected. In spite of these difficulties,
Teranashi et al. (1977) fractionated particulate organic matter
from Kobe, Japan, into acidic, basic, aliphatic, aromatic, and
oxygenated fractions. The acidic, aromatic, and oxygenated
fractions accounted for most of the mutagenic activity of the total
sample. Studies by Kolber et al. (1980), who used a somewhat
different fractionation method, also showed significant mutagenic
activity in those three fractions.
EVALUATION OF EMISSION SOURCES AND ATMOSPHERIC CONDITIONS
Initial comparative studies (Tokiwa et al., 1976; Pitts
et al., 1977) showed that air particulate was more mutagenic at
industrial and urbanized sites than at rural sites. Recently,
Flessel et al. (1990) compared mutagenicity among sites in Contra
Costa County, CA, with differing amounts of industrialization and
cancer rates. These studies all indicate a higher mutagenic
activity in the more urbanized or industrialized sites.
At any one ambient sampling site, the mutagenicity appears to
vary significantly over time. A major parameter affecting airborne
mutagenicity, identified by Commoner et al. (1978), in a year-long
study at a Chicago school site, was the wind direction. In this
study, a plot of wind direction versus relative mutagenic activity
showed that wind directions of either northwest or east resulted
in air particle samples with the greatest mutagenicity. Miller
and Alfheira (1980) reported on the mutagenicity of airborne
particles from two locations in Oslo over a three-month period.
They observed higher mutagenicity in February (i.e., during the
heating season) than in March and April. They also reported
significant meteorological effects. The mutagenicity, when
calculated as revertants per cubic meter of air, was highest on
cold clear days with little wind. When revertants per milligram of
particulate matter was calculated and compared with meteorological
conditions, mutagenicity was found to be high on days with rain or
snow, when the total concentration of particles in the air was low.
Although studies have been conducted to examine the effect of
ultraviolet light (Gibson et al., 1978) and other atmospheric gases
and oxidants including O3, NO2, and PAN (Pitts et al., 1978a, b) on
PAH, the role of these in the mutagenicity of particulate organic
matter in ambient air is still uncertain.

-------
PARTICULATE ORGANIC MATTER FROM AMBIENT AIR
15
In certain cases, studies can be designed to identify specific
emission sources that contribute to the mutagenicity of the ambient
particulate organic matter. At the Los Angeles freeway site
(discussed above), samplers were situated both upwind and downwind
from the freeway. Figure 2 shows the comparative mutagenic
activity of organics from particles < 1.7 urn collected over the
same period. The particles collected downwind from the automobiles
and trucks on the freeway were significantly more mutagenic than
those collected upwind. Claxton and Huisingh (1980) have shown
that the organics from a gasoline catalyst automobile are
significantly mutagenic. It is clear from these studies and others
(Huisingh et al., 1978; Lofroth, 1980; Alfheim and Miller, 1980)
that both gasoline and diesel engine exhaust from automobiles,
buses, and trucks contribute to the mutagenicity of ambient air
particles.
300
100
50	,
ug ORGANICS 25
mg PARTICULATE 0.5
250
5.0
125
2.5
1.0
Figure 2. Mutagenicity of air particulate (< 1.7 ym) collected by
MAVS upwind and downwind of the Los Angeles freeway.
It is clear that a variety of combustion sources could
contribute organic mutagens to the ambient air. Claxton and
Huisingh (1980) compared the mutagenic activity of organics from
particles emitted from residential heaters as well as diesel and

-------
16
JOELLEN LEWTAS HUISINGH
gasoline vehicles. A study by Miller and Alfheim (1980) of two
locations in Oslo, Norway, suggested that the mutagenicity they
observed was due in part to both automotive traffic and residential
heaters, with residential heating probably contributing more in the
winter months. Industrial sources of mutagens that may be
significant are coke oven emissions (Claxton, 1980) and coal
combustion emissions (Chrisp et al., 1978).
SUMMARY
The mutagens present in ambient air particulate possess the
following characteristics:
1)	They show both direct-acting and indirect-acting mutagenic
activity. The proportions of these two classes of
activity vary with the sample location.
2)	They show mutagenic activity primarily in the tester
strains that respond to frameshift mutagens.
3)	They appear to be present in higher concentrations in the
smallest particles (< 2 pra) than in larger particles.
4)	They appear to result from specific emission sources (such
as combustion sources).
REFERENCES
Alfheim, I., and M. Miller. Mutagenicity of airborne particulate
matter in relation to traffic and meteorological conditions.
Presented at the U.S. Environmental Protection Agency Second
Symposium on the Application of Short-term Bioassays in the
Fractionation and Analysis of Complex Environmental Mixtures,
Williamsburg, VA.
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian microsome mutagenicity test. Mutation Res.
31:347-364.
Andersen, A.A. 1966. A sampler for respiratory health hazard
assessment. Am. Ind. Hyg. Assoc. J. 27:160-165.
Carnow, B.W., and P. Meier. 1973. Air pollution and pulmonary
cancer. Arch. Environ. Hlth. 27:207,

-------
PARTICULATE ORGAx\IC MATTER FROM AMBIENT AIR
17
Chrisp, C.E., G.L. Fisher, and J.E. Lambert. 1978. Mutagenicity
of filtrates from respirable coal fly ash. Science 199:73-75.
Claxton, L. 1980. Mutagenic and Carcinogenic Potency of
Extracts of Diesel and Related Environmental Emissions:
Salmonella typhlmurium Assay. In: Proceedings of the
International Symposium on Health Effects of Diesel Engine
Emissions, December 1979. EPA 600/9-80-057ab.
U.S. Environmental Protection Agency: Cincinnati, OH.
Claxton, L. , and J.L. Huisingh. 1980. Comparative mutagenic
activity of organics from combustion sources. In: Pulmonary
Toxicology of Respirable Particles. C.L. Sanders, F.T. Cross,
G.E. Dagle, and J.A. Mahaffey, eds. CONF-791002. U.S.
Department of Energy: Washington, DC. pp. 453-465.
Commoner, B., P. Madyastha, A. Bronson, and A.J. Vithayathil.
1978. Environmental mutagens in urban air particulates. J.
Toxicol. Environ. Hlth. 4:59-77.
Epler, J.L. 1980. The use of short-term tests in the isolation
and identification of chemical mutagens in complex mixtures.
In: Chemical Mutagens: Principles and Methods for Their
Detection, Vol. 6. F.J. de Serres, ed. Plenum Press: New
York. pp. 239-270.
Flessel, C.P., J.J. Wesolowski, S. Twiss, J. Cheng, J. Ondo, N.
Monto, and R. Chan. 1980. Integration of the Ames bioassay
and chemical analyses in an epidemiological cancer incidence
study. Presented at the U.S. Environmental Protection Agency
Second Symposium on the Application of Short-term Bioassays in
the Fractionation and Analysis of Complex Environmental
Mixtures, Williamsburg, VA.
Freeman, A.E., P.J. Price, R.J. Bryan, R.J. Gordon, R.V. Gilden,
G.J. Kelloff, and R.J. Hueber. 1971. Transformation of rat
and hanster embryo cells by extracts of city smog. Proc. Kat.
Acad. Sci. USA 68:445-449.
Gibson, T.L. , V.B. Smart, L.L. Smith. 1978. Non-enzymatic
activation of polycyclic aromatic hydrocarbons as mutagens.
Mutation Res. 49:153-162.
Gordon, R.J., R.J. Bryan, J.S. Rhim, C. Demoise, R.G. Wolford, A.E.
Freeman, and R.J. Huebner. 1973. Transformation of rat and
mouse embryo cells by a new class of carcinogenic compounds
isolated from city air. Int. J. Cancer 12:223-227.

-------
18
JOELLEN LEWTAS HUISINGH
Henry, W.M., and R.I. Mitchell. 197E3. Development of a Large
Sample Collector of Respirable Particulate Matter.
EPA-600/4-7B-009. U.S. Environmental Protection Agency:
Research Triangle Park, NC.
Hueper, W.C., P. Kotin, E.C. Tabor, W.W. Payne, H. Falk, and E.
Sawicki. 1962. Carcinogenic bioassays on air pollutants.
Arch. Pathol. 74:89-116.
Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R. Zweidinger,
S. Tejada, J. Bumgarner, F. Duffield, V.F. Simmon, C. Hare,
C. Rodriguez, L. Snow, and M. Waters. 1978. Application of
bioassay to the characterization of diesel particle emissions.
Part I. Characterization of Heavy Duty Diesel Particle
Emissions. Part II. Application of a mutagenicity bioassay
to monitoring light duty diesel particle emissions.
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures. M.D. Waters,
S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
Plenum Press: New York. pp. 381-418.
Jungers, R., R. Burton, L. Claxton, and J. Lewtas Huisingh. 1980.
Evaluation of collection and extraction methods for
mutagenesis studies on ambient air particulate. Presented at
the U.S. Environmental Protection Agency Second Symposium on
the Application of Short-term Bioassays for the Fractionation
and Analysis of Complex Environmental Mixtures, Williamsburg,
VA.
Kolber, A., T. Wolff, T. Hughes, E. Pellizzari, C. Sparacino, M.
Waters, J. Lewtas Huisingh, and L. Claxton. 1980. Collection
chemical fractionation, and mutagenicity bioassay of ambient
air particulate. Presented at the U.S. Environmental
Protection Agency Second Symposium on the Application of
Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures, Williamsburg, VA.
Kotin, P., and H.L. Falk. 1963. Atmospheric factors in
pathogenesis of lung cancer. Adv. Cancer Res. 7:475-514.
Leiter, J., M.B. Shimkin, and M.J. Shear. 1942. Production of
subcutaneous sarcomas in mice with tars extracted from
atmospheric dusts. J. Natl. Cancer Inst. 3:155-165.
Lofroth, G. 1980. Comparison of the mutagenic activity in carbon
particulate matter and in diesel and gasoline exhaust.
Presented at the U.S. Environmental Protection Agency Second
Symposium on the Application of Short-term Bioassays for the
Fractionation and Analysis of Complex Environmental Mixtures,
Williamsburg, VA.

-------
PARTICULATE ORGANIC MATTER FROM AMBIENT AIR
Mitchell, R.I., W.M. Henry, and N.C. Henderson. 1978.
Fabrication, Optimization, and Evaluation of a Massive Air
Volume Sampler of Sized Respirable Particulate Matter.
EPA-600/4-78-031. U.S. Environmental Protection Agency:
Research Triangle Park, NC.
Miller, M., and I. Alfheim. 1980. Mutagenicity and
PAH-analysis of airborne particulate natter. Atmos. Environ
14:83-88.
Pitts, J.N., D. Grosjean, J.M. Mischke, V.F. Simmon, and D. Poole
1977. Mutagenic activity of airborne particulate organic
pollutants. Toxicol. Lett. 1:65-70.
Pitts, J.N., K.A. Van Cauwenberghe, D. Grosjean, J.P. Schmid, D.R
Fitz, W.L. Belser, G.B. Knudson, and P.M. Hynds. 1978a.
Atmospheric reactions of polycylic aromatic hydrocarbons:
Facile formation of mutagenic nitro derivatives. Science
202:515-519.
Pitts, J.N., K.A. Van Cauwenberghe, D. Grosjean, J.P. Schmid, D.R
Fitz, W.L. Belser, G.B. Knudson, and P.M. Hynds. 1978b.
Chemical and microbiological studies of mutagenic pollutants
in real and simulated atmospheres. In: Application of
Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures. M.D. Waters, S. Nesnow, J.L
Huisingh, S.S. Sandhu, and L. Claxton, eds. Plenum Press:
New York. pp. 353-379.
Swain, A.P., J.E. Cooper, and R.L. Stedman. 1969. Large-scale
fractionation of cigarette smoke condensate for chemical and
biological investigations. Cancer Res. 29:579-583.
Talcott, R., and E. Wei. 1977. Airborne mutagens bioassayed in
Salmonella typhimurium. J. Natl. Cancer Inst. 58:449-451.
Teranashi, K., K. Hamada, N. Tekeda, and H. Watanabe. 1977.
Mutagenicity of the tar in air pollutants. Proc. 4th Int.
Clean Air Congress, Tokyo. pp. 33-36.
Tokiwa, H., H. Takeyoshi, K. Morita, K. Takahashi, N. Sorutar and
Y. Ohnishi. 1976. Detection of mutagenic activity in urban
air pollutants. Mutation Res. 38:351-359.
Waters, M.D. 1980. An overview of the use of short-term
bioassays in evaluation of atmospheric pollutants. In:
Sampling and Analysis of Toxic Organics in the Atmosphere.
ASTM STP 721. American Society for Testing and Materials:
Philadelphia. pp. 156-165.

-------
Intentionally Blank Page

-------
COLLECTION, CHEMICAL FRACTIONATION, AND MUTAGENICITY
BIO ASSAY OF AMBIENT AIR PARTICULATE
Alan Kolbcr, Thomas Wolff, Thomas Hughes, Edo Pellizzari,
and Charles Sparacino
Research Triangle Institute
Research Triangle Park, North Carolina
Michael Waters, Joellen Lewtas Huisingh, and Larry Claxton
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION
Our industrial society has created thousands of synthetic
xenobiotics to support our modern lifestyle. Many of these
substances, and their by-products, enter our atmosphere in the form
of vapor-phase and particulate pollutants that can be ingested by
respiration and skin contact. The insults to human health and the
ecosystem from these airborne organic pollutants are due mainly to
polycyclic organic matter, specific industrial emissions such as
halogenated hydrocarbons, and end-use chemicals such as pesticides
(Fishbein, 1976). Polycyclic organics associated with air
particulates are believed to result primarily from incomplete
combustion of organic matter (Keller and Alfheim, 1980). These
carcinogenic components, (benzo(a)pyrene (B[a]P), benz(a)
anthracene) and other polycyclics have been well characterized
chemically (Shubik and Hartwell, 1969). As population and
industrial activity increase, the growing health hazards associated
with ambient air pollution must be further assessed and evaluated.
Air pollutants are chemically complex environmental mixtures,
whose compositions vary geographically with local industrial
activity and weather conditions. Seasonal variations in chemical
composition and mutagenicity of organic extracts of air particulate
have been documented in Oslo, Stockholm, and New York City (Miller
and Alfheim, 1980; Lofroth, 1980; Daisey et al., 1979). All three
communities exhibited higher mutagenicity during the winter months,
and in each case, a major contributing factor was polynuclear
aromatic hydrocarbons (PAH). These indirect-acting mutagens, which
require metabolic activation, were produced from combustion of
heating oil.
21

-------
22
ALAN KOLBER ET AL.
Ambient-air particulate matter has been estimated at
56 x 1013 g/yr, total global load, including natural pollutants
such as dusts, agricultural particulate, and others. The man-made
(anthropogenic) global load is estimated to be 28 x	g/yr.
Vapor-phase pollutant organics are believed to represent 60 to 90%
of the total organic load; the remainder is organics adsorbed onto
particulate matter (Hidy and Brock, 1970; Duce, 1978).
Chemical monitoring and analysis of air pollutants have been
conducted for many years, but only within the last decade has
biotesting technology improved enough to permit simultaneous
chemical and biological characterization. However, chemical-
analytical expertise exceeds the present biological capabilities,
both qualitatively and quantitatively (Hughes et al., 1980).
In this study, air particulate was collected, size-
fractionated, solvent-extracted and fractionated into chemical
classes, which were then characterized by GC/MS/computer analysis.
These chemical class fractions were then biotested for mutagenic
activity using the Ames/Salmonella bacterial mutagenicity assay.
Vapor-phase organic pollutants were also collected and tested.
However, no quantitative method was available to adequately measure
the mutagenicity of the vapor-phase components; consequently, we
are presently developing a bioassay protocol capable of quantifying
the mutagenicity of vapor-phase substances. It should be noted
that mutagenicity or carcinogenicity is by no means the only
significant health hazard that could be presented by air
pollutants. As additional in vitro bioassay capabilities are
developed, other potential toxicity parameters will be examined,
such as neurotoxicity and lung toxicity.
METHODS
Collection of Air Particulate and Vapors
Ambient air particulate was collected by the Maxisampler, a
high-volume sampling device constructed by the Battelle Corporation
after the design by Henry and Mitchel (1978). This device can
sample 20,000 m of air in a 24-h period and can collect particulate
matter within the respirable range in three size fractions (< 1.7
um, 1.7 to 3.5 pm, and > 3.5 ym), using impactor plates and
electrostatic precipitation. A mechanical diagram of the sampler
is shown in Figure 1. After collection, the plates were sealed in
a transportable container; and the particulate material was removed
and characterized morphologically by scanning electron microscopy.
Figure 2 is a typical scanning electron micrograph of ambient air
particulate. Air particulate was sampled at five U.S. locations:
Elizabeth, N.J.; Upland, CA; Lake Charles, LA; Houston and

-------
impacroR stage
PRECIPITATOR STAGE
BLOWER
UAGNtHH IC
INTAKE
CONTROL PANEL
HOUHMFTFtt
KILOVOLTS
KV ADJUST
SCALPING STAGE.
CUT OFF 20*j
>3.5y STAGE
> 1.7m STAfiF •
CLECTROS1 AMC
PRtClPH ATOH
: t
f i
—i-
M
*20y
IMPACT OH PLATES
CLtAN AIR
£
ca
NN
M
X
H
>
3
>
33
H
NN
O
C
r1
>
H
W
Figure 1. Schematic of the Battelle massive air volume sampler.
to
CO

-------
24
ALAN KOLBER ET AL.
Figure 2. Scanning electron micrograph of ambient air particulate
from Lake Charles, LA (> 3.5 ym).
Beaumont, TX. Table 1 summarizes the sampling variables for
Elizabeth, N.J.
Vapor-phase substances were sampled with cartridges containing
Tenax polymeric sorbent. Each cartridge was sealed and later
thermally desorbed into cryogenic traps (see Figure 3). This
frozen sample was then tested in the Ames/Salmonella assay.
Negative results were always obtained, but this may be due to an
inadequate testing procedure.
Chemical Fractionation of Ambient Air Particulate
An initial chemical fractionation scheme developed by
Pellizzari et al. (1978) that generated 13 polar and nonpolar
chemical classes was utilized to chemically fractionate the size-
fractionated air particulate. However, the chemical differences
among the fractions were small, and dividing the crude particulate
into 13 fractions resulted in sample sizes of less than 5 mg, which

-------
>
Table
1 . SamplIns
Variables Cor
ParLleu late
Organlcs In Ambient Air
In Elizabeth, NJ







Remarks

-

Temp.
Sample Vol.
Duration of

RelatIve
Wind Dir/Vol.
Sample Type
Date
°C
<°F)
(1)
Sampling (mln)
Time
Huml (1i ty
kts (m/s)
Vapor-pliase
9/19-20/78
17
(62)
646
1 ,390
1515-1425
902
NE/5-10 kts
organlcs







(2.5-5)

9/20-21/78
23
(74)
1,211
1,455
1445-1500
91%
SW/1U kts








(5)

9/21-22/78
20
(68)
1,044
1,265
1515-1220
90%
NK/10 kts








(5)

9/22-23/78
20
(68)
1 ,093
1,335
1245-1100
88*
NE/ 5 kts








(2.5)

9/23-24/78
16
(60)
1 ,279
1 ,690
1120-1530
7 5%
SK/10 kts








(5)
I'ar t lculates
9/19-22/78


81,211,000
4,560
0805-1200



9/22-25/78


78,892,000
4,430
1210-1400



9/25-28/78


70,246,000
4,060
1405-0745


M
3
H
>
SO
>
H
O
d
r1
5
w
bO
Qi

-------
26
ALAN KOLBER ET AL.
FLOW
METER
GLASS
CARTRIDGE
PUMP
FILTER
GAS METER
NEEDLE
VALVE
VAPOR COLLECTION SYSTEM
PURGE
GAS
THERMAL
DESORPTION
CHAVBER
TWO
POSITION
VALVE
GLASS
JET
SEPARATOR
HEATED
BLOCKS
EXHAUST
CARRIER
GAS
CARRIER
GAS
CRYOGENIC
TRAP
MAGNETIC
TAPE ,
ANALYTICAL SYSTEM
PLOTTER
MASS
SPECTRO-
METER
ION
CURRENT
RECORDER
COMPUTER
CAPILLARY GAS
CHROMATO-
GRAPH
Figure 3. Vapor collection and
of organic vapors in
analytical systems for analysis
ambient air.

-------
AMBIENT AIR PARTICULATE
27
limited adequate bioassay and chemical identification. The scheme
was modified to generate six chemical classes: acids, bases, PAHs,
polar neutrals, nonpolar neutrals, and insolubles (see Figure A).
Air particulate (~1.0 g) was subjected to ultrasonic treatment
in 100 ml cyclohexane (Burdick-Jackson) for 30 min and then
filtered through a Teflon (DuPont) filter (0.5-pm pore size). The
filtrate was evaporated to dryness using a rotary evaporator; the
solids retained by the filter were dissolved in 100 ml methanol
(Burdick-Jackson), and sonicated and filtered through the Teflon
filter; and the filtrate was taken to dryness. The filtrates were
combined, dried, weighed, and redissolved in methylene chloride
(CK2CI2). The solution was spiked with internal standards to allow
for quantification and to serve as a quality control parameter for
the overall partition/analysis scheme. The standards used were
quinoline-d7 (organic base), phenol-d5 (organic acid), and
anthracene-d10 (PAH).
This CH2CI2 solution was extracted twice with equal volumes
each of 10% sulfuric acid (H2SO4) and then once with 20% H2SO4.
The aqueous phases were combined and washed with CH2CI2, and this
CH2CI2 phase was combined with the original CH2CI2 solution. The
aqueous phase was cooled (ice bath), adjusted with 25% sodium
hydroxide (NaOH) to pH 10, and extracted three times with CH2CI2 to
generate the "organic bases." The aqueous phase was discarded.
The original CH2CI2 solution was extracted three times with 5%
NaOH, and the aqueous phases were combined and washed with CH2CI2.
The CH2CI2 solution then was combined with the original CH2CI2
solution. The NaOH phases were placed in an ice bath and acidified
to pH 3 with 20% H2SO4 and extracted three times with CH2CI2 to
generate "organic acids." The remaining aqueous phase was
discarded.
The original CH2CI2 phase was evaporated to dryness,
reconstituted in cyclohexane, and filtered through a Teflon filter
(0.5 um). The cyclohexane filtrate was extracted three times with
an equal volume of methanol:water (4:1) solution; the methanol:
water extract was then concentrated, and the water extracted three
times with ethyl acetate. The solvent was removed to generate
"polar neutrals." The cyclohexane phase was extracted three times
with an equal volume of nitroraethane; the nitromethane phases were
then combined and evaporated to dryness to generate the PAHs.
The cyclohexane phase was evaporated to dryness to generate the
"nonpolar neutrals" fraction.

-------
28
ALAN KOLBER ET AL.
") SONICAT= IN MeQ-<
2> F LTER
"1
"RATE
COMBINE sl»"RATES
1' REMOVESOLVENT
INORGANIC
aesn-j*
3) AC D WASH SFOuENCE*
AQt'fcCwS
(DISCARD)
AOLF.OUS
1! 3ASIPV IC *0
;h3c4
1) REMOVE SOLVENT
2) W£lG-<
ORGANIC EASES
CH,Clj
1} BASF WASH SEQUENCE*
Acueous ch,cl2
1) ACIO FV 'OpH 3
?) iX'RAC 3X AITH
C-t-CU
CH-,
cl2

1- REMOVESCLVENT

2< WEIGH
AQUEOUS
OISCAPC-)
ORGANIC ACIDS !
II REMOVE SOLVENT
2) RSOlSSOLVC in CfiH.
3! =ILTER
I
C6H12
o
"12

Ul
nr

2- WE
NONPOLAR
M £U"
RA^S
'} .VASH 3X WITh MeNr;^
—I
KN»NO,
IPA>-' 1
i
MsC'H/HjO

') EXTRAC

2: pevov

3! WEIGH
| KJlAH NbU^RALS j
FN T
'Acic Wash Sequence. 2 X with I0°i Ho504, 1 X wth 20% H?3Q«
**3
-------
AMBIENT AIR PARTICULATE
29
were included in the mixture, and the recovery of mutagenic
activity before and after fractionation was monitored by the
Ames/Salmonella bacterial mutagenicity bioassay.
Mutagenicity Bioassay
The plate-incorporation Aines/Salmonella mutagenicity assay was
performed as described by Ames et al. (1975), with and without S-9
rat liver microsomal activation preparation. All assays were
conducted with standard mutagens (positive controls) and solvent
(negative) controls (samples were exchanged into dimethylsulfoxide
[DMSO]). For revertant selection, minimal Vogel-Bonner medium E,
supplemented with 1.6% Difco bacto agar and 2% dextrose, was used
for base agar layers (25 ml, poured automatically). A 2.5 ml soft
agar overlay containing bacteria, minimum amounts of histidine and
biotin, and sample was routinely used.
To prepare S-9 from rat liver, the following procedure was
used: Male Charles River rats, strain CDj (200 ± 20 g), served as
the source of liver material. The animals were housed in suspended
cages for one week before induction. Food and water were given ad
libitum. Induced rat liver (from at least three rats) was obtained
from rats injected intraperitoneally on day one with 500 mg/kg of
Aroclor 125A in Mazola Corn Oil (0.5 ml of a 200 mg/ml for a 200 g
rat). On day five, the animals were sacrificed and their livers
removed. These were immersed in cold, sterile 1.15% KC1, washed
two times, and blotted dry with sterile paper. The livers were
weighed, ninced, brought to 200 mg/ml with 0.25 M sucrose, and
homogenized (on ice) with four strokes of a cold Potter-Elvenhjem
apparatus with a Teflon pestle. The homogenate was centrifuged for
10 min at 9,000 x g, the lipoprotein layer aspirated off, and the
protein concentration of the liver supernatant adjusted (after
protein assay) to 30 mg protein/ml with 0.25 M sucrose stock
solution. The stock solution was checked for sterility, quick-
frozen in small aliquots (2 to 5 ml), and stored for a maximum of
two months at -80°C. For the assay, an aliquot of stock solution
was slow-thawed and adjusted to an appropriate concentration of
protein in 0.25 M sucrose. NADPH was added at 320 ug/plate. Each
batch was checked against known positive controls using the
plate-incorporation assay. The optimal S-9 protein concentration
for mutagenicity was determined by testing with B(a)P and 7-12
dimethylbenzanthracene (DMBA).
Agar diffusion well technique. A modification of the
bacterial mutagenicity test was developed to screen for chemical
fractions of air particulate organics when sample size was limited.
In this test, the sarople(s) and S-9 microsomal preparation were
placed in wells cut in the agar. Soft agar then was added to fill
the well, and after solidification, the bacterial overlay
-------
30
ALAN KOLBER ET AL.
in 2.5 ml soft agar was poured over the entire surface. Results
were interpreted as follows: a direct acting mutagen was observed
as a ring of colonies around the sample well; a promutagen was
observed as a line of colonies between the sample and S-9 wells,
and toxicity resulted in a clear zone around the sample well.
This qualitative bacterial mutagenicity test was verified with
known mutagens, and its sensitivity was shown to be similar to that
of the spot test. Optimum size, number, configuration, and spacing
of wells, as well as amount of S-9 rat liver microsomal preparation
required per well, were determined. A configuration suited to
general screening included a center S-9 well and four surrounding
sample and control wells, as illustrated in Figure 5. Advantages
of this test over the spot test included ability to 1) measure
multiple parameters (dosage, activation requirements, toxicity, and
mutagenicity) on a single plate, thereby greatly extending the
amount of information obtainable from the available sample; 2)
determine positive, negative, and solvent controls on the same
plate as the test compound; 3) conduct dose-response and multiple-
compound testing on a single plate; 4) eliminate adsorption of
test compound by the filter disc; 5) eliminate runoff of liquids
onto the agar; and 6) reduce potential loss of mutagenic activity
by chemical hydrolysis.
A priority scheme for biological testing was also developed,
and was based on total size of each chemical-class fraction:
1) For > 10 mg of sample, the plate-incorporation method was
used, with five nontoxic dose levels (at least in
duplicate) in the following order: TA98, TA100, TA1535,
TA1537, and TA1538. Toxicity was tested at the highest
dose only. Preferred initial dose ranges were 1000, 500,
250, 100, and 10 ug/plate. The assay was performed first
with, and then without, Aroclor-induced S-9 (3.0 mg/
plate).
2) For < 10 mg of sample, the agar well diffusion test was
used at two concentrations (500 and 100 ug/plate), except
for PNA fractions which were tested using poured plates
(500 ug/plate, with Aroclor-induced S-9 and TA98.
Duplicate plates were assessed in the following order:
TA98, TA1535, TA1537, TA1538, and TA100. Sample was
diluted to one concentration for all tests; to obtain
lower test concentrations, a smaller sample volume per
well was used. Toxicity was tested at 500 ug/plate with
TA98 and TA1535, using 1000 colonies/plate.
-------
AMBIENT AIR PARTICULATE
31
Figure 5. Well testing configuration for bacterial mutagenesis
screening. The wells contain: 1) positive control
promutagen (e.g. , 100 \sg of 2-anthramine in 100 ul
solvent); 2) DMSO, solvent control (100 pi); 3) low
concentration of sample (100 ug sample in 20 ul DMSO);
4) high concentration of sample (500 ug sample in 100
ul DMSO); 5) Aroclor-induced S-9 microsomal preparation
([3.0 rag protein] + NADPH [480 pg in 300 ul])• The
distances are: a) 2.5 mm; b) 5.0 mm; c) 10.0 mm.
The wells are 18.0 mm in diameter.
RESULTS
Collection, Extraction, and Fractionation
Sampling of 81 x 10^ 1 ambient air required 76 h and generated
1.347 g of particulate, of which 180.1 mg were extractable
organics, or 2.3 ng organic/1 ambient air. Table 2 shows the total
particulates collected for each size fraction from each sampling
trip. Extractable organics varied from 2.5 to 15.3% of the
particulate mass, and the smaller size fractions did not
consistently contain the greatest amounts of organics (the 53%
extractable organics for Beaumont, TX [< 1.7 urn] is probably
incorrect, due to a weighing error). Table 3 illustrates the
distribution of the extracted organic after chemical fractionation;
12 to 15% loss of organic matter was common during fractionation.
-------
Table 2. Total Particulate Samples Collected
co
to
Particulate Particulate Total Organlcs
Sample Location Size Range (ym) Weight (nig) Extracted (mg) % Extracted
Upland, CA
Houston, TX 1tla
Lake Charles, LA #1
Elizabeth, NJ
Lake Charles, LA it2
Beaumont, TX
Houston, TX //2
<1 .
i .7 to
>3.
<1.
>1.
<1 .
>3.
<1.
1.7 to
>3.
<1.
1.7 to
>3.
3.")
3.5
3.5
<1
1.7 to 3.5
>3.5
<1.7
1.7 to 3.5
>3.5
666
64 1
286
337
790
747
113
1347
1302
1435
2618
488
548
285
882
1701
3043
1079
1353
33
36
18
19
86
25
12
180
120
99
401
15
27
151
31
42
110
5.0
5.6
6.3
5.6
10.9
3.3
10.6
13.4
9.2
6.9
15.3
3.1
4.9
53.0
3.5
2.5
3.6
al'art I c. les > I . 7 ym were combined to generate sufficient sample for fractionation and
hlotftfit1np.
W
O
r1
00
tn
po
tn
H
>
tr>
-------
Table 3. Fractionation of Organlcs Extracted from Ambient Air Particulate Samples
Amount of Organlcs In F,ach Fraction (mg)
Sample Location
Total. Organlcs
Size Range Extracted
(liR) (mg) Acid
Nonpolar Polar %
Base Neutrals PAIIb Neutrals Recovery
Upland, CA
Houston, TX If 1
Lake Charles, LA #1
F.l Izaheth, NJ
Lake Charles, LA II2
Beaumont, TX
Houston, TX #2
<1.7
1.7 to 3.5
>3.5
<1.7
>1.7
<1 .7
>1 .7
<1 .7
1.7 to 3.5
>3.5
<1 .7
1.7 to 3.5
>3.5
<1 .7
1.7 to 3.5
>3.5
<1.7
1.7 to 3.5
>3.5
32.7
17.5
35.9
19.3
86.4
24 .6
11 .9
180.1
119.7
99.4
401 .2
15.2
26.8
150.6
30.7
42.1
110.2
NDa
ND
2.9
1.8
1.1
1 .0
0.8
7.2
1 .7
8.5
10.4
5.5
170.2
3.9
16.2
15.5
6.8
10.4
20.4
ND
ND
1 .0
0.6
0.7
0.6
1.2
0.1
0.1
0.5
0.4
0.2
16.3
1.9
3.1
6.3
0.6
0.4
2.8
ND
ND
6.5
4.7
9.4
4 .0
3.2
6.1
3.9
86.2
47 .6
52.6
49.2
4.9
5.3
50.3
12.2
21.3
20.9
ND
ND
1 .6
2.0
2.5
3.2
2.4
2.1
2.1
11.1
9.6
6.0
7.1
1 .5
1.0
H.5
2.2
4.8
17.2
Nl)
ND
15.8
6.3
16.0
9.9
33.2
3.7
2.4
34 .6
26.1
13.4
152.6
1.3
0.8
6.7
8.9
3.R
38.0
ND
ND
B5.0
88.0
82.7
96.9
47.2
78.0
87 .4
78.2
78.6
78.2
98.6
88.8
98.5
58.0
100.0
96.7
85 .6
ND
ND
W
M
5-;
H
>
l-M
SO
"8
ta
H
l-H
O
c
c-1
>
8NU = not determined.
CO
co
-------
34
ALAN KOLBER ET AL.
Mutagenicity
Significant mutagenicity was observed for various chemical
fractions of the extracted particulate organics from some sites. A
sample was considered mutagenic when it generated at least two
times the number of revertants over spontaneous background for that
bacterial strain. The PAH fraction was mutagenic at alL five
sites. The polar neutral and organic acids were mutagenic at four
of the five sites, and the organic bases at three of the sites.
The mutagenicity data on fractions available in quantities too low
for the plate incorporation data were determined using the well
test. Both indirect- and direct-acting mutagens were present
in the complex mixtures and chemical fractions. An example of the
mutagenicity data obtained during this study is given in Table 4.
Chemical Analysis
Chemical fractions from each of the organic extracts of the
five geographical sites were analyzed by gas chromatography-mass
spectrometry (GC/MS). Compounds identified from the analysis were
matched against known toxic chemicals from the U.S. Health,
Education, and Welfare Chemical Registry and other sources. Table
5 lists known mutagenic, carcinogenic, neoplastic, and teratogenic
compounds identified. Many of these toxic compounds were found in
the extracted organics from all sites for which chemical analysis
was performed. Toxic PAH compounds were best represented.
DISCUSSION
This study was designed to develop and validate an integrated
multidisciplinary approach to the study of genotoxic effects of
ambient air pollutants. Engineering principles were employed to
develop and test the massive air volume sampler used to collect and
size-fractionate ambient air particulate. Analytical chemists
developed and validated the cyclohexane-methanol extraction scheme
for particulate organic matter and the acid-base extractive
fractionation scheme to separate the crude organic extract into
substituent chemical classes, which were then submitted to the
biologist for mutagenesis testing. Such multidisciplinary
approaches to the identification of bioactive chemical substituents
of various complex environmental mixtures, including organics
extracted from ambient air particulate matter, have been employed
in recent studies, including those of Epler et al. (1978), Rao et
al. (1980), Tokiwa et al. (1977), and others.
-------
Tnhle 4. Mutagenenis l>at;i of Parti cul ;i tt? Kroct Inuti f rom Lulu1 Clinrlrn, l.A (<1.7 |«n)fi
T A100
TA98
TA1538
TA1537
MA Protein
Sample
FracC ion
Doses
(ug/plate)
Concent rat ion
(rag) +MA
b

-1
HA

+MA
b

-MA

+MA
b

-HA

+MAb
-MA

Tnt;i 1
1000
3.0
507*
~
17
450*
+
21
127*
±
28
210*
~
70
152*

28
125*
t
12

or p,;in 1 i'fi
!>00
3.0
32/
~
15
341
+
30
7'.
+
24
169*
t
12
60
±
5
80
+
0

I'olar
500
3.0
'.01
f
18
387
+
2
129*
i
51
741*
t
6
117
i
6
191*
*
12
67 i 2
21 ±
11
neutrals''
100
3.0
186

63
256
i
73
94
t
23
155*
+
9
56

K
74
i
1?
24 1 10
23 i
9
Ac 1i1kc
500
3.0
326
~
9
221
+
91
82
±
15
77
t
17
59
1
21
67
+
11
62 ± 17
19 t
10

100
1.0
150
!
95
168
+
57
63
+
20
H9
+
22
61
+
5
52
+
12
4/ t 21
39 t
4
Bases
500
1 .5
163
*
106
94
~
63
105*
t
17
51
~
16
69
t
1
51
*
8

100
1.5
197
h
71
276
»
25
77
t
21
66
+
19
56
t
3
46
+
0

500
3.0

103*
t
40

100
3.0

53
i
24

500
6.0

58
i
19

100
6.0

b2
+
1/

Nonpn!a r
500
3.0
398
±
13
224
X
21
122*
+
7
107*
+
45
69
t
5
53
+
3

neut ra Is
100
3.0
365
t
7
279
*
25
93
±
9
88
*
10
60
±
7
37
*
0

I'olynuclear
500
1 .5

1007*
*
159

a romatIc
100
1 .5

394*

hydroc.'irhons 500
3.0

755*
t
144

100
3.0

283*
*
58

500
6.0

257*
f
198

1'os I t Ive
100

2404*
~
218
280
f
13
2309*
£
1?
1848*
t
41
2086*
i
5
1260*
*
55
39 i 20
174* »
38
cone rol
10

2197*
~
682
1241*
i
215
2074*
~
31
1117*
t
47
I 529*
i
13
623*
i
27
69 ! 20
70 t
6

1061*
t
14
627*
+
in
635*
+
38
109*
f
17
683*
f
30
339'
~
78
/ 4 ~ 6
20 ~
2
So I vent
0.1 nl

2 IH
t-
16
205
+
9
51

3
53
t
5
64
±
5
44
i
3
38 t 6
24 ±
6
>
2
a
M
H
>
1-^
?0
>
SO
H
P-^
O
d
f
>
H
M
mill rol
^pi at o-iricorporat ion assays; resultfl reported In moan t standard deviation computed with triplicate plates.
dMA
ne CabolIc act i vat ion.
cThe polar neutrals and acids were tested In TAISTi; howi'vcr, l hi* aticmiu revi'rlani g w«ir«' not within acceptable limit.1;*
^Mutagenic response Is at least twofold that of Urn spontaneous rover t ant s«.
CO
o\
-------
CO
a
Table *•. MuTagcnlc, Carcinogenic , Neoplastlr, and Te r a I np.eti L c ('ompmuuhi3 1 den I 1 f i ed In Ambient Air
Compounds
Up!and
CA
Acenaphthetiti
A1 ! yl-phi-nnl
Ami no ' ant hi iicpno
Ami no-anrhraqu i ncino
An 111ne
Ant Itan t lit eiu'.
Anthracene
Ant hraqtti none
tieiiz-anth racen r
Benzo-fluora n t hone
Lgnzo-perylene
Benzo phenanthreitc
Bcnzo-^yr en«
Broi iopheno I
Hnl y 1 ben/.y ! n 11 r< is amine
da I t'eicie
Cliloruplienol
Chrysene
C r e a o I
Dicycloliocy I amine
Dlmrtliy I am hr»cene
I)i melhy I -pheininthreno
P I no rant tiene
Methyl benz-anthrneene
He thy I benz-plienant hi t nc
Lake
Charles
M (1)
Lake
Oiarlcs
I.A (2)
RUzflbtth,
N.I
Re.iumon t,
TX
Houston
TX (I)
Houstnn
TX (2)
Muta
Ron! c
C» re Ino-
KonJo
Neo- Tf*ratr»-
plant I r Ken t r
U.S. Department of Health, Kriurat tort, and Welfare, 1977 .
Indefinite.
*
*
V
+
+
+
+
(runt imted)
>
w
o
r
CO
M
fd
w
H
>
-------
Table 5. Mutagenic, Carcinogenic, Neoplastic, and Teratogenic Compounds'1 Identified 1 r» Ambient Air (continued)
Compounds
Upland
CA
Lake
Charles
I.A (I)
Lake
Charles
I.A {?.)
Elizabeth,
NJ
Rr/iitmont ,
TX
Houston
TX (I)
Houston
TX (2)
Muta-
genic
Cnrr I nir
genlc
Neo-
plastic
Terato-
genic
>
2
oa
N
H
>
?a
T3
>
PO
H
1-^
n
Methyl-chryscne
Mp t hvl cut?-- phenar»r hrc-ne
Me thyl-phenunthiene
Mel hyl — pyrem*
Mel hyl -ht earate
Na phlharene
Naphtha lone
Peryleno
Phennnrhrcne
rticiivoi
Phenyl-benzene
Phen y 1 t»no-py rone
Plif iiy 1 - f-iwipliy t h I ami nc
Phenyl phenol
Pio^esterone
Propenyl-phenol
Py rone
On 1 no 11 r»o
Sty t ene
(p-Trl r,i«r| hy 1 -butyI)phenol
Ti*t rah yd rohenzo-phenanthrId 1 ne
To lu Id 1 tie
Tr line thy l-phcna nth rene
Trlphenylethylene
Suiiptic.Lod
>
H
M
*Il.S. Dep.irLroent of Health, Education, and Welfare, 197 7 .
CO
-1
-------
38
ALAN KOLBER ET AL.
Several problems were encountered in the development and
validation of this technical approach to the investigation of the
bioactivity of air particulate matter. Technical considerations
concerning the specificity of the chemical fractionation scheme
have been discussed previously (Pellizzari et al. , 1978). The
chemical fractionation scheme separates organic acids and bases by
treatment with aqueous base and aqueous acid, followed by
partitioning into organic solvent. Lofroth (1980) contends that
such treatment may generate artifacts, especially in the organic
base (nitrogen-containing compounds), due to reaction of substances
with the aqueous acid or base during the extraction. Epler et al.
(1978; Rao et al., 1980) recommended separation of the acid and
base components with Sephadex LH-20 to avoid this problem; however,
they found no significantly different results using LH-20 vs.
acid-base extraction.
Sample size was a chronic problem in this study. The massive
air volume sampler did not collect the amounts of particulate
expected. In a 10-day sampling period, often less than 3 g
particulate matter, including all three size fractions, was
collected. Only 5 to 15% of the particulate represents extractable
organics, and this material, when partitioned between five
fractions, usually did not provide enough sample for quantitative
dose-responsive plate-incorporation mutagenicity assays. For
triplicate determinations in all five Salmonella strains, using
five doses, both with and without metabolic activation (as
suggested by deSerres and Shelby, 1979), the assay would require
50 mg. Enough material must also be available for chemical
analysis. As can be seen from Table 3, none of the samples were of
adequate size for complete bioassay and chemical analysis.
Despite the limited data available, the results of air
particulate research by various investigators at different sites
show certain similarities to this study. For example, Pitts et al.
(1977) collected air particulate at rural and urban sites in
California and extracted the organics using sonication with equal
parts of methanol, benzene, and dichloromethane. The particulate
collected from the rural site was not mutagenic, but material from
all urban locations was mutagenic without S-9 addition, exhibiting
a linear dose response for Salmonella strains TA98, TA1538, and
TA1537. Thus, the mutagenic components present were probably not
polynuclear aromatic hydrocarbons (PAH), which require activation.
In another case, mutagenic activity requiring S-9 metabolic
activation was demonstrated in a California air sample; this
activity was then quantitatively related to B(a)P concentration
(Flessel, 1980).
In Japan, ambient air was sampled at six residential and
industrial sites (Tokiwa et al., 1977). The organic fraction from
ambient air particulate (extracted with methanol) collected in
-------
AMBIENT AIR PARTICULATE
39
Ohmuta (an industrial city) was mutagenic, exhibiting a linear dose
response from 100 to 400 pg with Salmonella strain TA98. A sample
collected from a residential area (Fukuoka City) required from 370
to 1970 ug/plate to generate a positive mutagenic response. The
mutagenicity was approximately three times greater for the sample
collected from the industrial city, and a linear dose response was
observed for both samples. These researchers identified the
compounds in the methanol extract using an alumina column and GC/MS
analysis. A majority of the compounds identified were PAHs such as
benzopyrenes, anthracenes, fluorenes, and dibenzoisomers of these
compounds. These Japanese studies revealed that air particulate
near industrial sites possessed higher mutagenic potential than
that from rural sites and that both S-9-activated (PAH) and direct-
acting (non-PAH) mutagenic components were present in air
particulates, agreeing with the California studies of Pitts et al.
(1977) and Flessel (1977).
Talcott and Wei (1977) collected air particulate on glass
fiber filters and extracted the filter with acetone in a Soxhlet
apparatus. The extracted material exhibited the highest activity
with TA100 and required S-9 metabolic activation. Direct-acting
mutagens were also detected with TA98 and TA1537. Activity was
again attributed to at least two types of mutagens: a PAH and a
non-PAH fraction. The view that PAH compounds are not the sole (or
major) mutagenic factor associated with air particulate was further
reinforced by the studies of Dehnen et al. (1977). Here, mutagenic
activity was predominantly produced by compounds other than PAHs,
since the observed mutagenicity did not require S-9. Dehnen used a
chemical fractionation scheme to divide his crude air particulate
sample into cyclohexane and methanol-extractable fractions. An
alumina column was employed to further fractionate the cyclohexane
extract into a purified cyclohexane fraction (containing PAH-type
compounds) and a 2-propanol fraction (containing azo-heterocyclic
compounds). The highest mutagenic response was found in the
2-propanol and methanol extracts; neither fraction contained
PAH-like compounds.
Other investigators (Tokiwa et al., 1977) have combined a
chemical fractionation scheme with a mutagenesis-detection system.
After solvent extraction, fractionation of the crude organic
extract into chemical classes permitted identification of the
mutagenically active chemical compounds within each chemical class.
In some cases, fractionation appeared to reduce or remove toxic
effects that otherwise would have prevented expression of
mutagenicity. Thus, the mechanism of synergism or antagonism of
the individual components in the crude mixture is open to
investigation. Pelroy et al. (1978) observed such effects with
shale oil and its fractions, as did Rao et al. (1980) with fossil
fuels and their fractions.
-------
40
ALAN KOLBER ET AL.
Pellizzari et al. (1978) initially tested a West Virginia
ambient air particulate for mutagenicity using a fractionation
scheme and the qualitative spot test (Ames et al., 1975)
with the standard Araes Salmonella tester strains (TA98, TA100,
TA1535, TA1537, and TA1538), both with and without a metabolic
activation system (S-9). Mutagenic activity was observed for the
organic bases, organic acids, and aromatics. Each of the five
tester strains gave a positive response with at least one of the
active fractions, and only the aromatics required metabolic
activation (Hughes et al., 1978; Pellizzari et al., 1978). In this
West Virginia sample, 2-nitro-4,5-dichlorophenol, a direct-acting
mutagen (Pellizzari et al., 1978), was identified in the polar
neutrals, and fluoranthene, pyrene, benz(a)anthracene, and B(a)P
were identified in the aromatic fraction. All of these compounds
have been identified either as co-carcinogens (Van Duuren, 1976) or
as mutagens requiring metabolic activation (McCann et al., 1975).
These preliminary results indicated that the Ames assay could
detect mutagenic activity in small sample amounts of ambient air
(as a complex mixture), and that the fractionation and analysis
scheme was capable of identifying broad chemical mutagenic classes
within these mixtures. It was concluded that chemical fractiona-
tion of the crude complex sample was both useful and necessary to
accomplish adequate biotesting and chemical identification of
signature mutagenic components in ambient air particulate.
ACKNOWLEDGMENTS
This study was supported by U.S. Environmental Protection
Agency Contract No. 68-02-2724.
REFERENCES
Ames, B., J. McCann, and E. Yamasaki. 1975. Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian
microsome mutagenicity test. Mutation Res. 31:347-364.
Chrisp, C., G. Fisher, and J. Lammert. 1978. Mutagenicity of
filtrates from respirable coal fly ash. Science 199:73-75.
Daisey, T., I. Hawryleck, T. Kneip, and F. Mukai. 1979. Mutagenic
activity in organic fractions of airborne particulate matter.
In: Conference on Carbonaceous Particles in the Atmosphere.
U.S. Dept. of Commerce: Washington, DC. pp. 187-192.
Dehnen, W. , R. Pitz, and R. Toraingas. 1977 . The mutagenicity of
airborne particulate pollutants. Cancer Lett. 4:5-12.
-------
AMBIENT AIR PARTICULATE
41
DeSerres, F., and M.D. Shelby. 1979. Recommendations on data
production and analysis using the Salmonella/microsome
mutagenicity assay. Mutation Res. 64:159-165.
Duce, R.A. 1978. Speculations on the budget of particulate and
vapor-phase non-methane organic carbon in the global
troposphere. Pure and Appl. Geophys. 116:244-273.
Epler, J.L., B.R. Clark, C.-h. Ho, M.R. Guerin, and T.K. Rao.
1978. Short-term bioassay of complex organic mixtures: Part
II, mutagenicity testing. In: Application of Short-term
Bioassays in the Fractionation and Analysis of Complex
Environmental Mixtures. M.D. Waters, S. Nesnow, J.L.
Huisingh, S.S. Sandhu, and L. Claxton, eds. Plenum Press:
New York. pp. 269-289.
Fishbein, L. 1976. Atmospheric mutagens. In: Chemical Mutagens:
Principles and Methods for their Detection. A. Hollaender,
ed. Plenum Press: New York. pp. 219-319.
Flessel, P. 1977. Mutagenic activity of particulate matter in
California Hi-Vol samples. Presented at the Third Interagency
Symposium on Air Monitoring Quality Assurance, San Francisco,
CA.
Henry, W., and R. Mitchell. 1978. Development of a large sampler
for the collection of respirable matter. EPA 600/4-78-009.
U.S. Environmental Protection Agency: Research Triangle
Park, NC.
Hidy, G., and J. Brock. 1970. An assessment of the global sources
of tropospheric aerosols. In: Proceedings of the Second
International Clean Air Congress. H.M. Englund and W.T.
Beery, eds. Academic Press: New York. pp. 1088-1097.
Hughes, T. , L. Little, L. Claxton, M. Waters, E. Pellizzari, and
C. Sparacino. 1978. Microbial mutagenesis testing of air
pollution samples. In: Application of Short-term Bioassays
in the Fractionation and Analysis of Complex Environmental
Mixtures. M.D. Waters, S. Nesnow, J.L. Huisingh, S.S. Sandhu,
and L. Claxton, eds. EPA 600/9-78-027. U.S. Environmental
Protection Agency: Research Triangle Park, NC. (abstr.)
p. 582.
Hughes, T.J., E. Pellizzari, L. Little, C. Sparacino, and A.
Kolber. 1980. Ambient air pollutants: Collection, chemical
characterization and mutagenicity testing. Mutation Res.
76:51-83.
-------
42
ALAN KOLBER ET AL.
King, L.C., and J. Lewtas Huisingh. 1980. Evaluation of the
release of mutagens from diesel particles in the presence of
serum. Presented at the U.S. Environmental Protection Agency
Second Symposium on the Application of Short-terra Bioassays in
the Fractionation and Analysis of Complex Environmental
Mixtures, Williamsburg, VA.
Lee, M., M. Novotny, and K. Bartle. 1976. GC/MS and nuclear
magnetic resonance determination of polynuclear aromatic
hydrocarbons in airborne particulates. Anal. Chea. 48:1566.
Lofroth, G. 1980. Comparison of the mutagenic activity in diesel
and gasoline engine exhaust and in carbon particulate matter.
Presented at the U.S. EPA Second Symposium on the Application
of Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures, Williamsburg, VA.
McCann, T. , E. Choi, E. Yamasaki, and B.N. Anes. 1975. Detection
of carcinogens as mutagens in the Salmonella/microsome test:
assay of 300 chemicals. Proc. Nat. Acad. Sci. USA
72:5135-5139.
Miller, M., and I. Alfheim. 1980. Mutagenicity and PAH-analysis
of airborne particulate matter. Atmos. Exper. 14:83-88.
National Academy of Sciences. 1972. Particulate Polycyclic
Organic Matter, Biological Effects of Atmospheric Pollutants.
Baltimore Press: Washington, DC.
Pellizzari, E., L. Little, C. Sparacino, T. Hughes, L. Claxton, and
M. Waters. 1978. Integrating microbiological and chemical
testing into the screening of air samples for potential
mutagenicity. In: Application of Short-term Bioassays in the
Fractionation and Analysis of Complex Environmental Mixtures.
M.D. Waters, S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L.
Claxton, eds. Plenum Press: New York. pp. 331-351.
Pelroy, R., and M. Peterson. 1978. Mutagenicity of shale oil
components. In: Application of Short-term Bioassays in the
Analysis of Complex Environmental Mixtures. M.D. Waters, S.
Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
Plenum Press: New York. pp. 463-475.
Pitts, J., D. Grosjean, T. Mischke, V. Simmon, and D. Poole. 1977.
Mutagenic activity of airborne particulate organic pollutants.
Toxicol. Lett. 1:65-70.
-------
AMBIENT AIR PARTICULATE
43
Rao, T.K., J.L. Epler, M.R. Guerin, B.R. Clark, and C.-h. Ho.
1980. Mutagenicity of nitrogen compounds from synthetic crude
oils: collection, separation, and biological testing.
Presented at the U.S. EPA Second Symposium on the Application
of Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures, Williamsburg, VA.
Shubik, P., and J. Hartwell. 1969. Survey of compounds which have
been tested for carcinogenic activity. Supp. 2. PHS 149-2.
U.S. Public Health Service: Washington, DC.
Talcott, R., and W. Harger. 1979. Mutagenic Activity of Aerosol-
size Fractions. EPA-600/3-79-032. U.S. Environmental
Protection Agency: Research Triangle Park, NC.
Talcott, R., and E. Wei. 1977. Airborne mutagens bioassayed in
Salmonella typhimurium. J. Natl. Cancer Inst. 58:449-451.
Tokiwa, H., K. Morita, H. Tokeyoshi, K. Takahashi, and Y. Ohnishi.
1977. Detection of mutagenic activity in particulate air
pollutants. Mutation Res. 48:237-248.
U.S. Department of Health, Education, and Welfare. 1977. Registry
of Toxic Effects of Chemical Substances, Vol. 1 and 2. U.S.
Depart, of Health, Education, and Welfare: Washington, DC.
Van Duuren, B. 1976. Tumor promoting and co-carcinogenic agents
in chemical carcinogenesis. In: Chemical Carcinogenesis.
ACS Monograph 173. C. Searle, ed. American Chemical Society:
Washington, DC. pp. 24-51.
-------
Intentionally Blank Page
-------
EVALUATION OF COLLECTION AND EXTRACTION METHODS FOR
MUTAGENESIS STUDIES ON AMBIENT AIR PARTICULATE
R. Jungers, R. Burton, L. Claxton, and J. Lewtas Huisingh
Environmental Monitoring Systems Laboratory and
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION
The extractable organics associated with air particles have
been shown to be carcinogenic (Hueper et al. , 1962; Leiter et al.,
1942) and mutagenic in a short-term microbial bioassay (Lewtas
Huisingh, 1980; Teranishi et al., 1977). The identification and
characterization of the potentially hazardous chemical components
associated with ambient air particles requires that efficient
collection and extraction methods be developed and validated. It
is important that the original chemical composition of the
particle-bound organics be maintained through collection and
extraction to establish that the bioassay results are directly
relatable to the chemistry of the original organics as thev exist
on particles in the atmosphere. The particles of major interest
are those that are in the inhalable (< 15 urn) or respirable
(< 5 urn) size range and therefore can be trapped in the human
respiratory tract. Several studies have shown that the organics
associated with ambient air particles are more mutagenic in the
smaller (respirable) particle size range (Huisingh, 1980; Teranashi
et al., 1977 ) .
In the past, most conventional ambient air samplers were not
designed to collect particles in size-separated stages, nor could
they collect the quantities of particulate matter desired for
integrated biological and chemical studies. A new sample
collection instrument, the Massive Air Volume Sampler (MAVS), has
been designed to separate collected particles into three size
ranges (0 to 1.7 ym, 1.7 to 3.5 ym, and 3.5 to 20 gm) and to
collect large amounts of particles. The objective of this study is
45
-------
46
R. JUNGERS ET AL.
to evaluate current ambient air particle sampling and extraction
methods for short-term mutagenesis bioassay applications.
COMPARISON OF AVAILABLE COLLECTION METHODS
A great number of field air particle samplers with varied
applications have been designed and used. Filtration and impaction
are the two techniques most commonly employed to collect particles
for the determination of particle concentration and chemical
composition. The conventional air particle collection instrument,
recommended in the Federal Register for determining total suspended
particulate (TSP) , is the standard high-volume sampler. Recently,
the size selective inlet (SSI) high-volume samoler has been
introduced to collect size-selected inhalable particles less than
15 ym (McFarland and Rodes, 1979). In addition, the dichotomous
sampler (virtual impactor) is being employed when elemental
chemical analysis is desired (McFarland and Rodes, 1979). All
three of these instruments employ filtration as the primary
collection method. Preliminary studies have shown that these
samplers do not collect sufficient amounts of particles for
complete bioassay studies. Henry and Mitchell (1978) designed and
constructed a massive air volume sampler (MAVS) for the U.S.
Environmental Protection Agency (EPA) that would collect large
quantities of size-fractionated air particles. The principles and
operation of this sampler are described in the next section. The
new sampler, MAVS, capable of collecting gram amounts of sized
particles in a reasonable sampling period, is currently being
evaluated for collecting ambient air particles to be used in
bioassay studies.
A comparison of particle mass that can theoretically be
collected by the TSP high-volume sampler, the SSI high-volume
sampler, the dichotomous sampler, and the new MAVS is shown in
Table 1 for different air pollution TSP concentrations. The
comparison was made for 24-h sampling periods in ambient air with
TSP particle concentrations of 60, 100, and 200 yg/m3. As noted in
the table, the mass of particles collected depends on the ambient
particle concentration and, more importantly, on the flow rate of
sample air through the given sampler. The flow rate of the MAVS
(18.5 m3/min) enables it to collect a greater amount of particulate
matter than the other samplers.
The Massive Air Volume Sampler
Since the MAVS is a relatively new particle sampler, a brief
description of its operation is given here. The sampler, shown in
Figures 1, 2, and 3, has an inlet which serves as the scalping
stage that permits only particles of 20 vim aerodynamic diameter or
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
47
Table 1. Comparison of Amounts of Particulate Matter
Theoretically Collected by Four Tvpes of Samplers
Calculated Particulate
Matter Collected at:
TSP Concentrations3

Vol. of Air

Type of
Flow Rate
Sampled3
60
Ug/m3
100 pg/m3
200 ug/m3
Sampler
(m3/min)
(m3)

(g)
(s>

D ichotomous
0.017
24.0
0,
.0015
0.0024
0.0048
SSI high-volume
1.13
1 ,627 .0
0.
.097
0.16
0.33
TSP high-volume
1.40
2,016.0
0,
.12
0.20
0.40
MAVS
18.5
26,640.0
1.
. 6
2.7
5.3
aDuring 24-h sampling period.
less to enter the sampler. The sample air then enters an impaction
plate assembly containing four stainless steel Teflon-coated dates
with slots, which serve as impactor jets and collection surfaces
for the two-stage impactor. The first stage of the impactor
collects particles between 20 and 3.5 um, and the second stage
collects particles between 3.5 um and 1.7 vim. Immediately below
the impaction plates, the remaining particles less than 1.7 um
pass through a particle charging field and are collected on the
third stage, which consists of vertically charged electrostatic
precipitator (ESP) plates. Both the impaction plates and the
electrostatic collector plates are Teflon-coated to minimize
substrate contamination and to facilitate removal of the particles.
The particle collector plate assemblies, after sampling, are
shipped in sealed containers to the laboratory, where the particles
are mechanically scraped from the Teflon plates. Figure 4 shows
the particles being removed inside a sealed glove box to minimize
contamination of the sample and to insure the safety of personnel.
Ozone Generation in the MAVS
The MAVS does not have the potential artifact-formation
problems associated with substantial flow of reactive gases (e.g.,
ozone and nitrogen oxides) over particles collected on reactive
filtration media, since the MAVS employs impaction and
-------
48
R. JUNGERS ET AL.
SCALPING STAGE
CUT OFF ¦> 20 p
fc
3 5^ STAGE
' >- 1.7y STAGE
ELECTROSTATIC
PRECIPITATOR
CLEAN
AIR
Figure 1. Massive Air Volume Sampler schematic.
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
Figure 2. Massive Air Volume Sampler

Figure 3. Impactor plate and electrostatic precipitator.
-------
50 R. JUNGERS ET AL.
Figure 4. MAVS plate being scraped inside glove box.
electrostatic precipitation onto Teflon-coated surfaces. The ESP
particle-charging section of the sampler, however, generates low
levels of ozone that passes through the fine particle (< 1.7 um)
collection stage. Since ozone may react with organic compounds
(NSF, 1977), including polynuclear aromatic hydrocarbons (PAH), it
is important to determine whether the generated ozone reacts with
the organic compounds associated with the collected particles and
whether such reactivity would bias mutagenicity bioassay results.
Mitchell et al. (1978) reported the ozone concentrations
present in the blower outlet of the first MAVS under different
operating conditions. After subtracting the ambient background
ozone (0.05 ppm), they reported no ozone emitted when a positive
corona was employed and 0.025 to 0.050 ppm ozone emitted when a
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
51
negative corona was employed. As the corona sign was changed from
positive to negative, the particle collection efficiency increased
from 66% to 80% at a corona voltage of 7,800. As the negative
corona voltage was increased, both collection efficiency and ozone
concentration increased.
As the MAVS has become viable as the most acceptable method
for collecting large samples of size-separated ambient particles
for bioassay screening studies, artifact formation due to reaction
of the collected material with ozone has become an important
consideration. Extractable organics from ambient air particles
collected at the third stage of the MAVS, the ESP, where an ozone
reaction would be most suspect, were found to be more mutagenic
than the larger particles collected at the first and second stages
(Huisingh, 1980). A second, more detailed measurement of ozone
within the sampler, as well as at the inlet and outlet, was
performed; the results are given in Table 2. The standard
operating particle charging voltage of 10,000 volts was used along
with negative polarity and corona. As noted in Table 2, the
highest ozone concentration in this study was found at the blower
outlet of the sampler, which is downstream of the third collection
stage. The highest ozone concentration observed in any section of
the sampler where particles are actually collected was 0.08 ppm.
When the voltage was reversed to give a positive corona, no
measurable amounts of ozone were generated in the sampler.
Table 2. MAVS Ozone Generation Test Results
Charging

Ozone (O3)
Wire
Plate

Generated
Voltage3
Voltage3
Locat i on
(ppm)
10 ,000
7 ,800
Sampler inlet
b
10,000
7,800
Precipitator inlet
0.07
10,000
7 ,800
Between plates
0.08
10 ,000
7 ,800
Precipitator outlet
0.09
10,000
7,800
31ower outlet
0.11
10,000
5,000
Between plates
0.08
aAll voltages are negative polarity.
'-'Room ambient concentration was 0.02 ppm O3- All other values were
corrected for this amount to determine the concentration of 0-j
generated.
-------
52
R. JUNGERS ET AL.
MAVS Ozone Field Study
A field study was developed to determine whether the
concentration of ozone generated in the electrostatic precipitator
collector section of the MAVS causes reaction between the ozone and
the organic constituents of the collected particles and
consequently biases the results of the bioassay. Three different
sampling sites were selected on the basis of possible PAH presence
in the air: Durham, NC (EPA Cameo Building in downtown Durham);
Birmingham, AL (industrial site in north Birmingham); and Gadsden,
AL (steel mill coke oven).
Since the MAVS can be operated with either positive particle
charging corona (without ozone generation) or with negative
particle charging corona (with ozone generation), field samples
were collected by co-located samplers (one with positive charging
corona and one with negative charging corona). All samples were
extracted, and split blind samples were both chemically
characterized and tested in the Ames Salmonella typhimurium
microbial mutagenesis bioassay. Chemical characterization and
bioassay results from particles charged with positive corona were
compared with results from particles charged with negative corona.
Samplers were operated in a normal manner with particle charging
voltage of 10,000 volts and collector plate voltage of
approximately 7,800 volts. As shown in Table 2, 0.08 ppm of ozone
was generated in the negative charging mode and no measureable
ozone in the positive charging mode.
At the Birmingham site, a sample was collected simultaneously
by each of the two co-located samplers (one with positive and one
with negative particle charging corona). The same sampling scheme
was used in Gadsden, where the two co-located samplers were
operated for two sequential sampling periods. In Durham, the same
sampling scheme was used, with the sampler corona reversed in the
second of the two sequential sampling periods to eliminate any
possible sampler bias. The sampling scheme for all three tests is
shown in Table 3.
The particles collected at the Durham field site were
mechanically extracted from the collector plates, and the
Birmingham and Gadsden particles were rinsed from the collector
plates with dichloromethane (DCM; methylene chloride). These
particle samples and solvent extract samples were sealed and
blind-coded for laboratory analysis (e.g., organic extraction,
chemical analysis, and bioassay preparation).
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
Table 3. Ozone Field Study Sampling Scheme
53
Particle Charging Corona
Location
Sampler
Experiment 1
Experiment 2
Durham, NC
A
B
+
+
Birmingham, AL
A
3
+
Gadsden, AL
A
B
+
+
Analysis Methodology
The air particle samples from Durham were Soxhlet-extracted
with DCM after mechanical removal from the plates, and those from
Birmingham and Gadsden, which were originally received in DCM,
were filtered. The DCM extract from both types of samples was
dried in a stream of dry nitrogen. One aliquot was used for
chemical analysis, and one aliquot was prepared for bioassay by
adding dimethylsulfoxide (DMSO) to obtain a concentration of
2 mg/ml.
The DCM extracts were prepared for gas chromatographic (GC)
and GC/MS analysis of polycyclic aromatic hydrocarbon (PAH) using
procedures developed by Bj^rseth (1977). The GC analysis was
performed on these extracts with a flame ionization detector (FID)
and a glass capillary column with hydrogen as the carrier gas.
Splitless injections were used while the column was at ambient
temperature, and peak area integration was performed with an
automatic digital integrator.
The GC/MS analysis was performed using a Finnigan Model 3200
MS and a Model 9500 GC with a glass capillary column. Splitless
injections were made while the column remained at ambient
temperatures. Electron impact (EI) spectra were obtained and
processed with a Digital System 150 and INCOS data system.
-------
54
R. JUNGERS ET AL.
Bioassay Methodologies
The S. typhimurium plate incorporation assay was performed as
described by Ames et al. (1975), with minor modifications. The
modifications consisted of adding the standard minimal
concentration of histidine directly to the plate media instead of
to the overlay and then incubating the plates for 72 rather than
48 h. The comparative samples were bioassayed simultaneously in
the same experiment for each tester strain.
The studies were performed with triplicate plates at five
doses with and without metabolic activation. Activation was
provided by a 9000 x g supernatant of liver from Aroclor-1254-
induced CD rats (Ames et al. , 1975). The linear portion of the
dose-response curves were used to calculate a linear regression
line. The slope of the linear regression analysis is reported as
revertants per microgram (rev/pg) of organics tested.
Discussion of Analytical Results
GC chromatographs were obtained under identical conditions and
show a similar relative concentration of PAH components in the
pairs of samples for all three sites. This similarity was shown
regardless of collection date, precipitator ion potential, or
whether the precipitator plates were washed or scraped.
The similarity noted in the chromatograms was the basis of a
reduced number of samples being analyzed by GC/MS. With the
identification of the major components determined by GC/MS, the
total ion chromatogram was compared to the GC chromatograph to
aid in the identification of the various components. The
quantification for various PAH species and the concentration ratios
for two pairs of isomeric PAH are shown in Table 4.
Some PAHs are oxidized readily, while others are relatively
insensitive to oxidation (Committee on Biological Effects, 1972).
Pyrene and anthracene are readily oxidized, while fluoranthene and
phenanthrene are relatively inert to oxidation. Assuming that
reactions with ozone proceed in the same manner in the MAVS with
negative corona as in the laboratory studies, the ratio between
the inert and reactive compounds should reflect any potential
reactions of ozone with the reactive PAHs. Comparison of these
ratios (Table 4) in the presence and absence of ozone (negative
and positive corona) show no significant or consistent changes,
suggesting no significant reaction of the PAHs with ozone. In
fact, the values obtained were well within the range reported in
other studies, including ambient air and aluminum and coke plants
that were conventionally sampled (Bj^rseth et al., 1978; Hoffman
and Wynder, 1976).
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE 55
Table 4. Comparisons of PAH Levels at Three Sampling Locations3
PAH (ug/g)

Durham
Birmingham
Gadsden
Chemical
Compound
+
-
+
-
+
-
Phenanthrene
1470
872
116
64
776
896
616
940
Anthracene
I960
1150
92
72
288
256
180
524
Ratio P/A
0.75
0.76
1.23
0.90
2.69
3.46
3.46
1.79
Fluoranthene
2190
648
208
196
924
1840
1370
2220
Pyrene
1980
696
224
184
1250
1900
1240
2900
Ratio F/P
1. II
0.93
0.94
1.08
0.74
0.97
1. 10
0.77
aPositive corona = +; negative corona =
The data provide interesting information about the MAVS.
Since the Birmingham samples (#1 and it2) were collected in the same
manner, they should exhibit similar patterns. In order to study
this relationship, parent PAH profiles (PPP) were constructed. PPP
depict relative distribution of the key PAH compounds in the
sample, thus providing a convenient method of comparing samples
(see Figure 5). The profile remained relatively constant over the
sampling period, indicating that operation of the electrostatic
precipitator in the positive corona (//I) or negative corona ("2)
mode did not affect the composition of the samples.
The data indicate that no significant degradation of PAH
attributable to ozonation occurred, compared with the inert
compounds. The stability of the profiles indicates that variation
of sampling parameters had no influence on the collected organics.
Therefore, it is unlikely that the MAVS electrostatic precipitation
-------
56
R. JUNGERS ET AL.
Phenathrene
Anthracene
Fluoranthene
Pyrene
Benz(a)anthracene
Chrysene
Benzfluoranthene
Benz(e)pyrene
Benz(a)pyrene
Indenopyrene
Benzperylene
positive corona
negative corona
i 1 1 1 1 1 1 1 r — -! 1 1—
0123456789 10 11
mg PAH/sample
Figure 5. Parent PAH profile of ambient air.
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
57
plates in the positive or negative corona mode caused any artifact
formation due to ozonation.
Discussion of Bioaasay Results
The mutagenic activity in S. typhlmurlum tester strain
TA98 is shown in Figure 6 for each of the three sites as a function
of ionization potential. At the Durham site, the slope of the
mutagenic activity curve ranged from 0.37 to 0.42 rev/ug in the
absence of S-9 activation and from 0.52 to 0.75 rev/yg in the
presence of activation. No significant difference was observed in
mutagenicity as a function either of ionization potential or of
sampler.
2 8-
2.6-
2 4-
2.2-
a 2.0-
53 1.8-
£
1.6-
< 1.4-
y
z
LLl
1.2-
<
2 10-
0 8-
0.6-
• TA 98 - S3
A TA 98 + S9
~
I
T
A ~r
1 i
A
T 4"
T
A
1
T
. 1
1
0 2-
ATibicnt Air
Durham. NC
Anb'Snt A •
Birrn ngham. AL
Coke Oven
Gadsden, AL
Figure 6. Corona potential.
-------
58
R. JUNGERS ET AL.
Although the mutagenic activity of the Birmingham and Gadsden
sam pies was greater than the Durham samples, no significant
difference was observed between the two samples collected during
the same time periods with different coronas.
Evaluation of Extraction Techniques
To determine which extraction technique was the most effective
in removing mutagens from ambient air particles, two extraction
methods and seven solvent systems were evaluated. The two
extraction methods were Soxhlet extraction for 24 h and sonication
for both 30 min and 2 h. The four basic single solvents, selected
for increasing polarity index, were cyclohexane (CH),
dichloromethane (DCM), acetone (Ac), and methanol (MeOH), with
polarity indices of 0.0, 3.4, 5.4, and 6.6, respectively. Three
additional comparative solvent systems were sequential sonication
of cyclohexane and methanol (CH/MeOH), 1:1:1 toluene:
dichloromethane:methanol (T:DCM:MeOH), and direct suspension in
dimethylsulfoxide (DMSO). These solvent systems were selected for
comparison with previous work (Pitts et al., 1977; Pellizzari
et al. , 1979). Only the four single solvents were used in the
Soxhlet extraction method, while two of the comparative solvent
systems were also used in sonication.
The data in Table 5 indicate no significant difference for
percent extractables in the length of time used for sonication, but
acetone and methanol extracted approximately three times as much
mass as cyclohexane and twice as much as DCM. The Soxhlet
extraction method indicated a significant increase of percent
extractables using acetone as the solvent.
Table 5. Ambient Air Particles Percent Extractables
Soxhlet Sonication Sonication
Solvent 24 h 30 min 2 h
CH
DCM
Ac
MeOH
CH/MeOH
T:DCM:MeOH
4.0
6.4
21.0
8.9
2.9
4.3
9.0
11.4
11.6
8.7
3.0
4.7
8.2
11.5
11.6
8.9
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
59
An aliquot of each extract was dried under nitrogen and
diluted with a 1:1 mixture of DCM and cyclohexane. Fifty
microliters were spotted on a TLC plate and developed in a tank
containing a 1:2 mixture of DCM and ethanol. Analysis was done
with a Perkin Elmer MPF-44B Fluorescence Spectrometer. The results
are shown in Table 6. Regardless of method of extraction or
solvent used, the benzo(a)pyrene (B[a]P) analysis is relatively
constant on a weight-by-weight basis.
Table 6. Ambient Air—Glass Fiber Filters B(a)P Analysis
B(a)P (gg/g)
Soxhlet Sonication Sonication
Solvent 24 h 30 min 2 h
CH
11.8
8.2
8.4
DCM
12.0
10.9
9.5
Ac
11.4
11.7
10. 1
MeOH
11.8
11.1
10.4
CH/MeOH

13.2
12.8
T: DCM:MeOH

13.9
11.7
An aliquot of 10/S of the extract was dried and prepared in
aqueous solution for analysis by a Dionex Model 10 ion
chromatograph (IC). The IC was equipped with a standard column and
0.003 M sodium bicarbonate and 0.0024 M sodium carbonate eluent.
The sample was analyzed for fluoride, chloride, nitrate, and
sulfate ions. Total anion concentration was calculated and
compared for the four single solvents used in Soxhlet extraction
and the two additional comparative solvent systems used in the
sonication extraction techniques. Table 7 shows that the
increasing polarity of the solvent is comparable to the increasing
quantity of total anions extracted, regardless of extraction
technique, although the increased time of sonication did extract
increased amounts of total anions. Table 7 also shows that the
methanol in both the Soxhlet and 30-min sonication extraction
techniques and the comparative sorbent system CH/MeOH in the 30-min
sonication extraction technique extracted approximately the same
quantities of total anions. It appears that methanol is the basic
extracting solvent of total anions in both systems.
-------
60
R. JUNGERS ET AL.
Table 7. Comparison of Total Anion Concentrations
Total Anions (jjg/g)
Solvent
Soxhlet
24 h
Sonication
30 min
Sonication
2 h
CH
DCM
Ac
29.12
19.32
583.71
992.67
17.09
21.77
319.66
878.54
949.44
443.69
13.80
81.89
528.49
MeOH
CH/MeOH
T: DCM:MeOH
1124.87
99.749
500.17
Table 8 summarizes anion analysis of the four single solvents
in the Soxhlet extraction technique. The major quantity of anion
extracted for acetone is nitrate ion and for methanol is sulfate
ion. Neither of these ions appreciably affect the microbial
mutagenicity bioassay results. The results of further studies
incorporating sonication and the comparative solvent systems are
shown in Table 9 for nitrate ion and Table 10 for sulfate ion.
It should be noted that while the comparative solvent systems
do not extract significantly more nitrate ion than does the single
solvent acetone, they extract far less of the sulfate ion than does
the single solvent methanol. Therefore, it appears that the two
comparative solvent systems (CH/MeOH and T:DCM:Me0H) are not more
effective than the Soxhlet-extracted single solvents (acetone and
methanol).
Bioassay Results and Discussion
The S. typhimurlum mutagenesis data for tester strain TA98
for each solvent system and extraction technique are shown in
Table 11. DCM extraction, either by Soxhlet or sonication,
resulted in an extractable material that was more mutagenic than
that resulting from any of the other solvents. The least polar
solvent, cyclohexane, was the least effective in extracting
mutagens by either method. Acetone and methanol solvent extraction
generally resulted in less-mutagenic samples than DCM extraction,
except when acetone was used in sonication. Acetone removed
considerably more extractable mass from air particles, including
more inorganics. Therefore, when the mutagenic activity was
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
61
Table 8. Ambient Air Anion Analysis
Anion Concentrations (vjg/g)a
Solvent F1 CI NO3 SO4
CH 0.02 2.6 7.0 19.5
DCM 0.02 2.2 2.8 14.3
Ac 2.41 83.7 477.5 20.1
MeOH 0.77 18.7 56.1 917.1
aSoxhlet extraction for 24 h.
Table 9. Ambient Air—Glass Fiber Filters Nitrate Analysis
Nitrate (yig/g)
Soxhlet Sonication Sonication
Solvent 24 h 30 min 2 h
CH 7.01 6.73 1.22
DCM 2.81 3.93 3.04
Ac 477.48 265.27 290.62
MeOH 56.11 42.92 498.57
CH/MeOH 505.48 468.43
T:DCM:MeOH 318.32 397.02
-------
62
R. JUNGERS ET AL.
Table 10. Ambient Air—Glass Fiber Filters Sulfate Analysis
Sulfate (ug/g)
Sonication
30 tnin
Sonication
2 h
7.61
15.48
12.61
825.41
331.AO
51.54
8.82
71.74
4.26
516.80
425.61
6.88
Solvent
Soxhlet
24 h
CH
DCM
Ac
MeOH
CH/MeOH
T:DCM:MeOH
19.48
14.33
20.06
917.12
Table 11. Mutagenic Activity of Different Solvent Extractable
Material from Air Particulate in TA98 (without S-9 activation)
Soxhlet Extraction
Sonication
Solvent
95% Confidence
Rev/yg Limits Rev/ug
95% Confidence
Limits
CH
DCM
Ac
MeOH
CH/MeOH
T:DCM:MeOH
DMSO
0.11
0. 52
0.28
0.26
0.05
0.43
0.23
0.23
0.16
0.62
0.33
0.30
0.09
0.48
0.45
0.35
0.38
0.37
0.07
0.06
0.37
0.38
0.28
0.30
0.26
0.03
0.13
0.58
0.53
0.41
0.45
0.48
0.10
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
63
calculated per milligram of particle, as shown in Table 12, acetone
appeared to result in greater mutagenic activity. Consequently,
the use of acetone may be advantageous for certain studies.
Analytical problems associated with the use of acetone plus the
presence of higher concentrations of inorganic salts have resulted
in our selection of a solvent other than acetone for most studies.
Table 12. Ambient Air Particle Mutagenicity in TA98
(without S-9 activation)
Solvent
% Extractablea
Revertants/ug
Extractable
Revertants/mg
Particle
CH
DCM
AC
MeOH
4.0
6.4
21.0
8.9
0.11
0.52
0.28
0.26
4.4
33.3
58.8
23. 1
aSoxhlet extraction.
While DMSO can be used directly both to suspend the particles
and administer to the bioassay without evaporation or solvent
exchange, it was the least effective method for detecting the
mutagenic activity present. Thus, DCM is the preferable solvent,
particularly for studies in which the amount of nonnutagenic mass
should be minimized.
REFERENCES
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian microsome mutagenicity test. Mutation Res.
31:347-364.
Bjifirseth, A. 1977. Analysis of polycyclic aromatic hydrocarbons
in particulate matter by glass capillary gas chromatography.
Anal. Chim. Acta 94:21-27.
Bjifirseth, A., O. Bjrseth, and P.E. Fjeldstad. 1978. Polycyclic
aromatic hydrocarbons (PAH) in industrial atmospheres. I.
Analysis of the PAH content in an aluminum reduction plant.
Scand. J. Work Environ, Hlth 4:212-216.
-------
64
R. JUNGERS ET AL.
Code of Federal Regulations. 1980. Title 40, part 50, appendix B.
Reference Method for Determination of Suspended Particulates
in the Atmosphere (High Volume Method). General Service
Administration: Washington, DC. pp. 531-535.
Committee on Biological Effects of Atmospheric Pollutants. 1972.
Particulate polycyclic organic matter. National Academy of
Science: Washington, DC.
Henry, W.M., and R.I. Mitchell. 1978. Development of a large
sample collector of respirable particulate matter.
EPA-600/4-78-009. U.S. Environmental Protection Agency:
Research Triangle Park, NC.
Hoffman, D., and E.L. Wynder. 1976. In: Chemical Carcinogens.
C.E. Searl, ed. ACS Monograph, 173, American Chemical
Society: Washington, DC. pp. 324-365.
Hueper, W.C., P. Kobin, E.C. Tabor, W.W. Payne, H. Falk, and E.
Sawicki. 1962. Carcinogenesis bioassays on air pollutants.
Arch. Pathol. 74:89-116.
Huisingh, J. Lewtas. 1980. Bioassay of particulate organic matter
from ambient air. Presented at the U.S. Environmental
Protection Agency Second Symposium on the Application of
Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures, Williamsburg, VA.
Leiter, J., M.B. Shimkin, and M.J. Shear. 1942. Production of
subcutaneous sarcomas in mice with tars extracted from
atmospheric dusts. J. Natl. Cancer Inst. 3:155-165.
McFarland, A., and C.E. Rodes. 1979. Characteristics of aerosol
samplers used in ambient air monitoring. Presented at the
86th National Meeting of Chemical Engineers, Houston, TX.
Mitchell, R.I., W.M. Henry, and N.C. Henderson. 1978.
Fabrication, optimization, and evaluation of a massive air
volume sampler of sized respirable particulate matter.
EPA-600/4-78-031. U.S. Environmental Protection Agency:
Research Triangle Park, NC.
National Science Foundation, Subcommittee on Ozone and Other
Photochemical Oxidants. 1977. Ozone and other photochemical
oxidants. Printing and Publishing Office, National Academy
of Sciences: Washington, DC.
-------
COLLECTION AND EXTRACTION OF AIR PARTICULATE
65
Pellizzari, E.D., L.W. Little, C. Sparacino, T.J. Hughes,
L. Claxton, and M.D. Waters. 1979. Integrating
microbiological and chemical testing into the screening of
air samples for potential mutagenicity. In: Application of
Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures. M.D. Waters, S. Nesnow,
J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds. Plenum
Press: New York. pp. 382-418.
Pitts, J.N. , D. Grosjean, J.M. Mischke, V.F. Simmon, and D. Poole.
1977. Mutagenic activity of airborne particulate organic
pollutants. Toxicol. Lett. 1:65-70.
Teranashi, K. , K. Haraada, N. Tekeda, and H. Watanabe. 1977.
Mutagenicity of the tar in air pollutants. Proceedings of the
4th International Clean Air Congress, Tokyo, Japan,
pp. 33-36.
-------
Intentionally Blank Page
-------
INTEGRATION OF THE AMES BIOASSAY AND CHEMICAL ANALYSES IN
AN EPIDEMIOLOGICAL CANCER INCIDENCE STUDY
C. Peter Flessel, Jerome J. Wesolowski, SuzAnne Twiss,
James Cheng, Joel Ondo, Nadine Monto, and Raymond Chan
Air and Industrial Hygiene Laboratory
California Department of Health Services
Berkeley, California
INTRODUCTION
The development of the Ames bioassay as an instrument for
assessing public health problems involving mutagenicity and
potential carcinogenicity has resembled the development of other
quantitative techniques for assessing public health problems. The
Ames test has developed from a qualitative assay of samples in
simple matrices into a quantitative determination of complicated
environmental mixtures, and its results are now being integrated
with human epidemiological studies. Originally, the Ames test was
used primarily to determine whether or not a compound was mutagenic
(Ames, 1971). Soon quantitative methods were introduced (Ames et
al . , 1975), and a significant correlation between mutagenicity in
the Ames test and carcinogenicity in animal bioassays emerged
(McCann et al., 1975). Shortly thereafter, the test was applied to
environmental mixtures in the analysis of air particulate material
(Pitts et al. , 1977 ; Talcott and Wei, 1977). Currently, the Ames
test is used to detect mutagens in a variety of media and sample
types. The rapidly expanding list of applications includes
drinking water, cigarette smoke, auto exhaust, foods, drugs, and
urine (Holstein et al. , 1979). Although not a quantitative test in
the sense of having well-established precision and accuracy, the
Ames bioassay yields results that indicate relative mutagenicity.
Thus, it is appropriate to consider its use in epidemiological
cancer studies.
The application of the Ames mutagenicity test to
epidemiological cancer studies is logical because mutagenicity is
in some measure a composite index of total potential
carcinogenicity (McCann et al., 1975). It can be argued that
67
-------
68
C. PETER FLESSEL ET AL.
determining the concentrations of all carcinogens in the medium
also should give, at least in theory, the information needed to
determine the carcinogenic exposure to humans. However, chemical
methods are not available to detect all possible carcinogens in a
given medium. Furthermore, even an exhaustive compilation of
carcinogens would neglect synergistic or antagonistic effects.
Thus, application of both chemical and bioassay techniques more
completely characterizes environmental samples (Bj^rseth et al.,
1980).
Contra Costa County, CA, was suspected of having high rates of
respiratory-tract cancer, based on the findings of the national
survey of cancer mortality by county, 1950 through 1969 (Mason and
McKay, 1973). The northern section of Contra Costa County is
heavily industrialized, with five major petroleum refineries and
many petrochemical plants. These facts prompted the U.S.
Environmental Protection Agency (EPA) to fund an epidemiological
study of cancer incidence as related to airborne emissions in
Contra Costa County. This study is part of a larger study funded
jointly by the State of California and the Occupational Safety and
Health Administration (OSHA). The larger study includes four other
counties in the San Francisco Bay Area and is being carried out by
the California Resource for Cancer Epidemiology (RCE). The present
discussion is restricted to the Contra Costa County portion of the
project. The major objectives of the study are 1) to identify
environmental factors which may contribute significantly to cancer
incidence in the county, 2) to determine whether various groups of
workers are more likely to develop cancer than others, and 3) to
evaluate whether air pollutants have affected the observed
incidence of respiratory-tract cancer in the county.
An important component of the study is environmental
monitoring, consisting of ambient air particulate matter sampling
and subsequent chemical and biological analyses. A major goal is
to determine whether or not mutagenic activity, as measured by the
Ames assay, can be accounted for by the chemical characterization
of the samples. Environmental monitoring will provide a means of
correlating the geographic distribution of current cancer incidence
with current ambient air pollutants and to develop baseline air
pollution data for comparison with future measurements and use in
future epidemiological cancer studies. Because of the latency of
cancer onset, current incidence levels cannot be causally related
to current air quality. Analysis of past and present emission data
and air quality patterns could, however, reveal historical
pollution trends useful in determining the association, if any,
between air pollution and current cancer incidence.
This paper will discuss the design of the environmental
monitoring program and preliminary data from the chemical and
-------
AMES BIO ASS AY IN EPIDEMIOLOGICAL CANCER STUDY
69
biological assays. The planned integration of this environmental
data base with the epidemiological cancer study will be described.
PROCEDURE
The Sampling Program
Fifteen high-volume particulate samplers were placed at
thirteen locations in Contra Costa County and two locations in
adjacent counties (shown in Figure 1), in order to characterize
air quality variations over the entire county. Five of the
locations are permanent stations of the Bay Area Air Quality
Management District (BAAQMD). Although these five stations monitor
for several pollutant gases as well as for particulate matter, the
present paper will discuss only particulate matter results.
VallCjO
P ttiburq
WjMinPi Concord
Concord T'tfdC
Sao H,
Te"->por?-v SfBl'.on
Figure 1. Locations of sampling stations in Contra Costa County,
CA.
-------
70
C. PETER FLESSEL ET AL.
Air particulate material was collected on 8- x 10-in. (20- x
25-cid) glass-fiber filters (EPA Grade Whatman) in standard high-
volurae samplers, which drew approximately 1,200 of air through a
filter during each 24-h run. Filters were collected every sixth
day at each of the 15 sampling stations, from November 3, 1978,
through October 31, 1979. In this period, nearly 900 air-filter
samples were collected and analyzed.
Particulate matter was analyzed for benzene-soluble organics
(BSO), lead (Pb) , total suspended particulate matter (TSP) ,
nitrates (NO3-), sulfates (SC>4=), specific polycyclic aromatic
hydrocarbons (PAH), and mutagenic activity, using the Ames test.
The BAAQMD collected the air samples and analyzed them for TSP,
NO3-, and SO4-, and the Air and Industrial Hygiene Laboratory
(AIHL) carried out the other analyses.
Logistics of Sample Analysis and Data Management
The plan for distributing filter samples for analysis and
reporting results is shown in Figure 2. After weekly sample
collection, the filters were weighed to determine the amounts of
total suspended particulate material and were delivered to AIHL.
There, the filters were logged in and cut, and the pieces were
distributed for further analysis.
The crux of the air-monitoring program is the analysis of
composite air samples from each of the 15 stations for PAH content
and mutagenic activity. For each station, composite samples were
prepared by combining samples collected over each of the following
four-month periods: November, 1978, through February, 1979
(winter); March through June, 1979 (spring); and July through
October, 1979 (summer). These three periods correspond to the
three meteorological seasons of the San Francisco Bay Air Basin.
Filter disks for PAH analysis and mutagenicity testing collected
between November 1, 1978, and May 1, 1979, were stored in the dark
at room temperature. Filters collected between May 1 and October
31, 1979, were stored in the dark at -20°C. Disks (47-nnn diam.)
were cut from each filter and individually extracted ultrasonically
with organic solvents. For each location, aliquots taken during a
given four-month period were combined to provide a composite sample
for PAH and mutagenic analysis.
Analysis Methods for TSP, BSO, Pb, N0^~, and S04~
Standard methods were used to analyze for the following five
pollutants: TSP was determined gravimetrically (BAAQMD, 1977);
NO3"" colorimetrically (BAAQMD Method N-7 , 1976): SO43
turbidiraetrically (BAAQMD Method S-42, 1976); BSO by Soxhlet
-------
AMES BIO ASS AY IN EPIDEMIOLOGICAL CANCER STUDY
71
FILTERS
1 n
2- Deposit area msajtred
3 Cut and disf'tbui*d 'w analysis
IAIHL-
AnJlyied (or
NO3 Caicnmeiricafly
SO4, Turhidi-ncTical y
IBAAQMD)
To BAAQMD
Analyzed PAHi
GC-MS. HPLC
(AlHl)
Analyzed for Pb
Analy(9d for
MUTAGENIC ACTIVITY
in Iht
An« Assay (AlHLl
Figure 2. Logistical plan for analysis of high-volume air filters
collected in Contra Costa County, CA, November, 1978,
through October, 1979.
extraction (AIHL, 1975); and Pb by wavelength dispersive X-ray
fluorescence (Moore, 1976).
Analysis Methods for PAH
Analysis methods for PAH were modified from Bjifrseth et al.
(1980). Individual (47-mm) filter disks were placed in screw-cap
test tubes and extracted twice at 40 to 45°C for 20 minutes, first
with 8 ml and then with 6 ml of cyclohexane (MCB, OmniSolv) in an
ultrasonic bath (Bransonic Models 220 or 32). Each cyclohexane
extract was filtered through a 0.5-ym Fluoropore filter
(Millipore). Extracts comprising each composite were combined in a
round-bottom evaporating flask and concentrated to 6 ml in a rotary
evaporator at 45 to 50°C. To separate PAH from interfering
material, the 6 ml of concentrated cyclohexane extract was combined
with 2 ral of toluene (MCB, OmniSolv) and extracted ultrasonically
-------
72
C. PETER FLESSEL ET AL.
with 12 ml of a 10:1 mixture of N,N-diraethylforraamide (DMF) (MCB
Manufacturing Chemists, Inc., OmniSolv) and water in a screw-cap
test tube for 15 min. The bottom layer, containing the PAH, was
transferred by pipette to a 60-ral separatory funnel. The remaining
cyclohexane phase was re-extracted ultrasonically twice more with
6 ral of the DMF-H2O mixture, and the phases containing the PAH were
combined in the separatory funnel. Following the addition of 24 ml
of distilled water to the separatory funnel, the 2 ml of toluene,
containing the PAH, separated. The toluene phase was transferred
to a centrifuge tube and evaporated to dryness in a heating block
at 45°C under a stream of nitrogen. The residue was then dissolved
in 200 to 400 yl of acetonitrile for analysis by high-pressure
liquid chromatography (HPLC).
A Varian Model 5000 high-pressure liquid chromatograph and
Microbondapak C18 (Waters Associates) column were used to separate
PAH. Column effluents were monitored using ultraviolet (UV)
absorption (at 254 nm) and fluorescence (excitation, 263 ran;
emission, 407 nm). Fluorescence measurements were used to resolve
and quantitate three carcinogenic PAH: benz(a)anthracene ,
benzo(a)pyrene , and chrysene. Fluorescence measurements were made
with a Perkin-Elmer Model MPF-44A spectrofluorometer. HPLC was
performed in a linear gradient from 70% acetonitrile in water to
100% acetonitrile, in 50 min. The flow rate was 0.8 ml/min, the
temperature was 30°C, and the chart speed was 1 cm/min. The
injection was made with a sample loop operated by a rotary valve,
using a 10-yl injection volume.
The efficiency of extraction of PAH from high-volume filters
has been studied. Fluorescence measurements show that more than
95% of PAH can be recovered from a spiked filter.
Peaks observed in the HPLC chroraatograms were identified by
three methods. First, peaks were tentatively identified by
comparing their retention times to those of standards. Second,
the peak height ratios of samples and standards were compared at
the wavelengths used to measure absorbance and fluorescence.
Identifications of the three PAH were confirmed by stopping the
flow during HPLC analysis and scanning the fluorescence spectra
using the optimum excitation wavelength for each compound. Peak
heights were used for quantitation of PAH.
The selectivity of the fluorescence detection method is
illustrated in Figures 3 and 4. Using fluorescence excitation and
emission wavelengths of 263 nm and 407 nm respectively, most of the
UV-absorbing PAH peaks were suppressed, but benzo(a)pyrene,
benz(a)anthracene, and chrysene were enhanced. Such specificity is
critical, because many poorly resolved peaks, including those
containing the 10 PAH standards, are visible in the chromatograms
of air samples using UV detection (Figure 4). The major peak
-------
AMES BIOASSAY IN EPIDEMIOLOGICAL CANCER STUDY
73
eluting just before phenanthrene was due to an organic contaminant
in the cyclohexane used in early experiments. This artifact
disappeared when the brand of cyclohexane specified above was
substituted.
The Ames Test for Mutagenic Activity
Methods for extracting air particulate material from high-
volume glass-fiber filters and making composite samples were
adapted from Pitts et al. (1979). The solvent, a 1:1:1 mixture of
methanol, dichlororaethane, and toluene (MCB, OmniSolv), was
prepared fresh daily and saturated with nitrogen. Extractions
were carried out in an ultrasonic bath under low light. Each
individual 47-mm filter disk cut from a filter was placed in a
16- x 125-mm screw-cap tube with a Teflon liner, and 4 ml of
solvent was added. Tubes were sonicated for 20 rain at maximum
power in the ultrasonic bath containing water at 45°C. Extracts
were filtered through 0.5- m Fluoropore filters (Millipore); filter
disks were re-extracted using 3 ml solvent, and the filtrates were
combined. The volume of each extract was adjusted to exactly 10 ml
with the solvent, and the extracts were stored at -20°C.
Composite samples for mutagenic testing were prepared by
combining aliquots of stored extracts in a vacuum flask, saturating
the extracts with nitrogen, and reducing the volume of solvent in
a rotary evaporator, under reduced pressure and at a temperature of
45°C. The composite samples were then transferred to preweighed
tubes, which were placed in a heat block at 45°C; the remaining
extraction solvent was removed under a stream of nitrogen. After
weighing, residues were redissolved in dimethylsulfoxide for
mutagenic analysis.
The method for detecting mutagens with the Salmonella/
mammalian microsome test was as described by Ames et al. Tl975),
with the following changes: rat liver homogenate (S-9) was
prepared from rats fed commercial rodent foods; rats were
anesthesized with carbon dioxide before surgery; and plates were
incubated for 72 instead of 48 h. S-9 protein concentrations were
determined by the method of Lowry et al. (1951).
Negative solvent (dimethylsulfoxide) and S-9 sterility
controls and positive controls for each strain used were run with
each experiment. The control mutagens for the five tester strains
were sodium azide in TA1535; 9-aminoacridine in TA1537;
2-aminofluorene in TA1538 and TA98; and methyl methanesulfonate in
TA100. The Ames assay was applied according to a two-part protocol
(Pitts et al . , 1979). The first step involved screening the sample
in the five standard Ames tester strains both with and without
metabolic activation. These data gave a qualitative estimate of
-------
74
C. PETER FLESSEL ET AL.
0.03
|
S 0.02
U
o
3
8
9
o.oi
24
M BONDAPAK Cl« COLUMN
bO't CHj CN/4 0" H20 TO
iOCi CHjCtf LINEAR
GRADIENT IN" 50 HIN.
20
16
12
RETENTION TIME (MINUTES)
Figure 3. High-pressure liquid chroraatogram of a 10-PAH standard
detected by fluorescence and UV absorbance.
0.04
S
s
N
0.01
16 12
RETENTION TIME (XINUTES)
Figure 4.
High-pressure liquid chromatogram of the PAH fraction
from air particulate material collected in Antioch, CA,
November, 1978, through February, 1979.
-------
AMES BIOASSAY IN EPIDEMIOLOGICAL CANCER STUDY
75
the mutagenic activity and indicated the most sensitive strain and
the optimum conditions of metabolic activation for subsequent
quantitative analysis. For initial screening, each composite
sample was assayed at one dose in the range 20 to 1000 ug/plate.
Determinations were made with and without added rat liver S-9 at
both low (~ 0.6 mg/plate) and high (~ 3 mg/plate) protein
concentrat ions.
All composite samples exhibited mutagenic activity in the
initial screening and were reanalyzed in the strain showing the
greatest response. The dose range and conditions of metabolic
activation also were chosen to maximize activity. Duplicate
determinations were made at each of several doses and the result
expressed as revertants per cubic meter.
An interlaboratory comparison of the mutagenicity of ambient
particulates was carried out between AIHL and the Statewide Air
Pollution Research Center, University of California, Riverside.
High-volume filter samples collected in southern and northern
California were split and analyzed for mutagenic activity in the
Ames test. The determinations made by the two laboratories agreed
within a factor of two.
RESULTS
As analyses are still in progress, only partial results are
presented .
Table 1 gives the median and maximum values for TSP, BSO, and
Pb for winter and spring. The median and maximum levels of the
three pollutants were the highest in winter, when meterological
inversions frequently occur. The extreme values for individual
24-h runs differed by a factor of 50. For example, the highest
24-h level of BSO was 41.3 mg/m^, in Concord in early December,
1978, and the lowest was less than 0.8 mg/m^ (the detection limit)
at several sites in January and February, 1979. The highest Pb
concentration was found in Antioch, also in early December, while
the highest TSP concentration was in Brentwood in November.
To describe variations in levels of community air pollution
throughout the county, maps showing the geographical distribution
of the seven measured pollutants are being constructed using a
computer program called SYMAP (developed by the Laboratory for
Computer Graphics, Harvard University, Cambridge, MA). In this
program, sampling station coordinates and associated pollutant
levels are used to construct a matrix containing the pollution
levels at each station. Contours are constructed by interpolation.
These distributions will be used to estimate community exposure
levels and will be compared with the patterns of cancer discovered
-------
76
C. PETER FLESSEL ET AL.
Table 1. Analysis of Air Particulate Material Collected
in Contra Costa County, CA
24-Hour Median Value 24-Hour Maximum Value
Nov. 1978- March- Nov. 1978- March-
Pollutant Feb. 1979 June 1979 Feb. 1979 June 1979
Total suspended
particulate material 60 42 229 126
Benzene-soluble
organics 4.8 1.4 41.3 32.1
Lead 0.7 0.2 2.5 0.6
in epidemiological studies. The distributions may also provide
clues to pollution sources.
Thus far, computer-drawn contour maps of TSP mass, BSO, and
Pb levels for winter and spring have been prepared; these are shown
in Figure 5. They were constructed using average values obtained
at each sampling station during Che first two seasons for which
composite samples were analyzed. Panels A, B, and C show the
distributions of these pollutants during winter; panels D, E and F
show Che discribucions in the spring. Concentracions of the
pollutants were generally higher in winter than in spring, and
seasonal variations were most pronounced for Pb and BSO levels,
which changed by more than a factor of three (see Table 1). The
geographical distributions of Pb and BSO were similar. For both
pollutants, the highest levels were found in winter in a north-
south band located in central Contra Costa County. This region
corresponds roughly to Che Diablo Valley, a natural pollution sink
through which runs a major freeway.
At present, three PAH have been quantitated: benzCa)anthracene
benzo(a)pyrene, and chrysene. Concentrations from the Antioch and
Brentwood winter composite samples are listed in Table 2. Antioch
is an urban-industrial site; BrenCwood is a rural locacion. The
range of concentrations of benz(a)pyrene found in these samples
(0.1 to 0.8 ng/m ) was comparable to, although somewhat lower than,
that found in particulate samples collected in San Francisco 20
years ago (Sawicki eC al., 1960). The present values were also
-------
AMES BIOASSAY IN EPIDEMIOLOGICAL CANCER STUDY
CONTRA COSTA
CONTRA COSTA

NOV T9-FE3 79 *
MASS. lltj-T.' . .. .. . .,
°:<0 A ^00
BU «0< SO *
>0< 70
70< 100

:.
^v> •¦>•
NOV. 7B-FC0.
OftGANICS /4Q'rr.
0 (^2 0
2 0*4 0
4.0* 8 0
ao«iso
NOV 7S-fEB 7S
if ad ^/mJ
f>£>* 0 2
(L2«Q3
oj« oa
Ofl « 1 2
MAR. - JUN 79
MASS ^jm*
0< 40
OR *o< so
M 50< 70
?oCS Mjlm
20< 2.0
rv3 2 o« 4 o
E9 4> « .o
¦I 6 CK15 0
MAR - JUN 79
LLAO ^/'T? 1
D.0< 0 2
0 2< 0 3
BB 0 3*0®
O X I 2

»¦-- aS
^«s ""
Figure 5. Computer-drawn contour mapa of the geographical
distribution of levels of total mass, benzene-soluble
organics, and lead in air particulate material
collected in Contra Costa County, CA. A, B, and C
ahow distributions for November, 1978 through
February, 1979, and D, E, and F show March through
June, 1979, distributions.
comparable to those measured more recently in New York City (Dais
et al. , 1979) and Germany (Gusten and Heinrich, 1978).
Extracts from samples collected at the 15 stations have been
analyzed qualitatively for mutagenicity, and all showed activity
-------
78
C. PETER FLESSEL ET AL.
Table 2. Polycyclic Aromatic Hydrocarbon Content and Mutagenic
Activity in Contra Costa County Air Particulate Material
Collected November, 1978, through February, 1979
Sampling Location
Type of Measurement
Ant ioch
Brentwood
PAH (ng/m3)
Benz(a)anthrene
Benzo(a)pyrene
Chrysene
0.81
0.81
0.90
0.10
0.10
0.16
Sum of three PAH 2.52 0.36
Mutagenic activity
(revertants/m3)
Without S-9 6.3 3.9
With S-9 25.4 5.9
at least one strain. The most activity was seen in strains TA98
and TA1538. Adding the S-9 fraction generally enhanced activity;
some samples were most active at the high S-9 protein concentration
(3 mg/plate), while the majority were most active at the lower
concentration (0.6 mg/plate). Dose-response curves for the Antioch
and Brentwood composites (with activities given as revertants/
plate) are shown in Figures 6 and 7. Over Che dose range used (up
to 20 ra3 of air), the mutagenic responses appear linear. The
dose-response curves obtained with the Antioch winter composite
(Figure 6) were somewhat atypical, in that most samples did not
show a fourfold increase in activity in the presence of S-9. The
response to S-9 in samples analyzed to date was more typically that
shown by the Brentwood sample (Figure 7), although stations in more
heavily polluted areas generally showed greater S-9 enhancement.
Seasonal variations in mutagenic activity were also observed.
Activities (expressed as revertants per cubic meter) were generally
higher for samples collected in winter than for samples collected
in spring or summer. Table 2 summarizes results from the Antioch
and Brentwood Stations. Samples from Antioch, a more urban setting
than Brentwood, showed both higher concentrations of PAH and
-------
AMES BIOASSAY IN EPIDEMIOLOGICAL CANCER STUDY
600-.
+S9
pe 400-
200-
-S9
25
20
VOLUME OF AIR («3)
Figure 6. Ames test dose-response curves for a composite sample
from Antioch, CA, November, 1978, through February,
1979, assayed in strain TA98, with and without 3 mg S
protein/plate.
200
W
3 150 -
a,
as
Ui
eu

-S9
S
i 50
5
15
20
25
10
VOLUME OF AIR (m3)
Figure 7. Ames test dose-response curves for a composite sample
from Brentwood, CA, November, 1978, through February,
1979, assayed in strain TA98, with and without 0.6 mg
S-9 protein/plate.
-------
80
C. PETER FLESSEL ET AL.
increased mutagenic activity. The activities and the urban-rural
differences found in this study were similar to those found
previously in California (Flessel, 1977; Pitts et al . , 1977 , 1979).
One can also compare chemical composition and biological
activity in a given sample. For example, the Antioch sample
analyzed above had a benzo(a)pyrene concentration of 0.81 ng/m^.
As the molecular weight of benzo( a)pyrene is 252, this corresponds
to approximately 0.003 nmol/ra^. The specific mutagenic activity of
benzo(a)pyrene is approximately 121 revertants/nmol (Ames et al. ,
1975). Therefore, the amount of benzo(a)pyrene measured accounts
for about 0.36 revertants/m^, or less than 2% of the observed
activity. Clearly, chemicals other than benzo(a)pyrene account for
most of the mutagenicity in this and other air samples (Bj^rseth et
al., 1980).
DISCUSSION
The cancer epidemiology project plan consists of a series of
studies, both prospective and retrospective, to investigate
environmental factors relevant to cancer in Contra Costa County.
The study will include a census tract analysis of cancer incidence
at specific sites and is designed to ultimately attempt to
distinguish between the contributions of occupational and community
exposures to the incidence of cancer. From the mutagencity and
PAH data, community exposures in Contra Costa County will be
estimated. These exposures will be examined for correlations with
present and future cancer rates in various geographical areas.
Although this is fundamentally a prospective study, attempts will
be made to interpret current cancer incidence rates in terms of
current exposures and historical pollution trends. Definitive
studies will require following current Contra Costa residents
through the next several decades and monitoring a larger number and
variety of carcinogens and mutagens in community air. It might be
worthwhile to expand monitoring activities to include data from
community water, soil, food, and workplace samples.
ACKNOWLEDGMENTS
This research was supported in part by the U.S. EPA grant
no. R 806396010. The authors thank W. Riggan, EPA Project
Officer, for his support and interest. The X-ray fluorescence
analyses for lead were performed by H. Moore and A. Alcocer (AIHL).
Benzene-soluble organics were determined by F. Boo (AIHL).
Nitrates, sulfates, and total suspended particulate matter were
measured by S. Balestrieri (BAAQMD). We also wish to acknowledge
the cooperation of M. Imada, E. Jeung, and R. Stanley (AIHL); D.
-------
AMES BIOASSAY IN EPIDEMIOLOGICAL CANCER STUDY
81
Levaggi, W. Siu, D. England, and N. Balberan (BAAQMD); and D.
Austin, W. Mandel, and S. Lam (RCE, California Department of Health
Services) .
REFERENCES
AIHL Method 67. 1975- Determination of Total Organic Materials
in Atmospheric Particulate Matter. Air and Industrial Hygiene
Laboratory, California Department of Health Services,
Berkeley, CA.
Ames, B. 1971. The detection of chemical mutagens with enteric
bacteria. In: Chemical Mutagens: Principles and Methods
for their Detection, Vol. 1. A. Hollaender, ed. Plenum
Press: New York. pp. 267-282.
Ames, B. J. McCann, and E. Yamasaki. 1975. Method for detecting
carcinogens and mutagens with the Salmone11 a/mammalian-
microsome mutagenicity test. Mutation Res. 31:347-364.
BAAQMD Method. 1977. Total Suspended Particulate Gravimetric
Analysis Procedure. Bay Area Air Quality Management District,
San Francisco, CA.
BAAQMD Method N-7. 1976. Determination of Nitrate in Glass Fiber
Hi-Vol Filters. Bay Area Air Quality Management District,
San Francisco, CA.
BAAQMD Method S-42. 1976. Determination of Sulfate in Glass Fiber
Hi-Vol Filters. Bay Area Air Quality Management District,
San Francisco, CA.
Bjrseth, 0., P. Flessel, N. Monto, J. Wesolowski, T. Parker, and
P. Ouchida. (1980). Monitoring for polycyclic aromatic
hydrocarbon (PAH) content and mutagenic activity in products
and emissions from a gasifier demonstration project. In:
Safe Handling of Chemical Carcinogens, Mutagens, and
Teratogens—A Chemist's Viewpoint, Vol 2. D. Waters, ed.
Ann Arbor Press: Ann Arbor, MI. pp. 635-651.
Daisey, J., M. Leyko, and T. Kniep. 1979. Source identification
and allocation of polynuclear aromatic hydrocarbon compounds
in the New York City aerosol: methods and applications. In:
Polynuclear Aromatic Hydrocarbons. P. Jones and P. Leber,
eds. Ann Arbor Science: Ann Arbor, MI. pp. 201-215.
-------
82
C. PETER FLESSEL ET AL.
Flessel, P. 1977. Mutagenicity of Particulate Matter. Presented
at the Third Interagency Symposium on Air Monitoring Quality
Assurance, Air, and Industrial Hygiene Laboratory, California
Department of Health Services, Berkeley, CA.
Gusten, H., and G. Heinrich. 1979. Polycyclic aromatic
hydrocarbons in the lower atmosphere of Karlsruhe. In:
Polynuclear Aromatic Hydrocarbons. P. Jones and P. Leber,
eds. Ann Arbor Science: Ann Arbor, MI. pp. 357-370.
Holstein, M., J. McCann, F. Angelosanto, and W. Nichols. 1979.
Short-term tests for carcinogens and mutagens. Mutation Res.
65:133-226.
Lowry, 0., N. Rosebrough, A. Farr, and R. Randall. 1951. Protein
measurement with the folin phenol reagent. J. Biol. Chem.
193:265-271.
Mason, P., and F. McKay. 1973. U.S. Cancer Mortality by County,
1950-1969. DHEW Publication No. (NIH)-74-615. U.S.
Government Printing Office: Washington, D.C.
McCann, J., E. Choi, E. Yamasaki, and B. Ames. 1975. Detection
of carcinogens as mutagens in the Salmonella/microsome test:
Assay of 300 chemicals. Proc. Natl. Acad. Sci. USA
72:5135-5139.
Moore, H. 1976. Application of Wavelength Dispersive X-ray
Fluorescence Spectrometry to the Determination of Lead in
Atmospheric Particulate Matter Collected on High-Volume Glass
Fiber Filters. AIHL Report 183, Air and Industrial Hygiene
Laboratory, California Department of Health Services,
Berkeley, CA.
Pitts, J., D. Grosjean, and T. Mischke. 1977. Mutagenic activity
of airborne particulate organic pollutants. Toxicol. Lett.
1:65-70.
Pitts, J., K. Van Cauwenberghe, D. Grosjean, J. Schmid, D. Fitz,
W. Belser, G. Knudson, and P. Hynds. 1979. Chemical and
microbiological studies of mutagenic pollutants in real and
simulated atmospheres. In: Application of Short-term
Bioassays in the Fractionation and Analysis of Complex
Environmental Mixtures. M. Waters, S. Nesnow, J. Huisingh, S.
Sandhu, and L. Claxton, eds. Plenum Press: New York. pp.
355-378 .
Sawicki, E., W. Elbert, T. Hauzer, F. Fox, and T. Stanley. 1960.
Benzo(a)pyrene content of the air of American communities.
Am. Ind. Hyg. J. 21:443-451.
-------
AMES BIO ASSAY LN EPIDEMIOLOGICAL CANCER STUDY
Talcott, R., and E. Wei. 1977. Airborne mutagens bioasaayed i
Salmonella typhimurium. J. Natl. Cancer Inst. 58:449-451
-------
Intentionally Blank Page
-------
MUTAGENICITY OF AIRBORNE PARTICULATE MATTER IN RELATION TO
TRAFFIC AND METEOROLOGICAL CONDITIONS
Ingrid Alfheim and Mona Miller
Central Institute for Industrial Research
Oslo, Norway
INTRODUCTION
It is now well established that environmental factors are
major causes of cancer in man (Higginson and Muir, 1973).
Epidemiological studies have shown that the incidence of lung
cancer is higher in urban than in rural areas (Henderson et al. ,
1975). Urban air contains large amounts of particulate pollutants,
which are believed to contribute to this higher lung-cancer
incidence (Menck et al., 1974). Furthermore, the carcinogenic
potential of organic extracts from airborne particles has been
demonstrated in animal experiments (Hueper et al., 1962). These
observations have made it increasingly important to identify
carcinogenic compounds in ambient air. Since animal studies are
expensive and time-consuming, short-term tests for mutagenicity
in microbial systems are currently used to identify possible
care inogens.
The presence of mutagenic compounds in airborne particulate
matter from urban and industrial areas has been documented by us
and others using the Ames Salmonella test system (Dehnen et al.,
1977; Lofroth, 1978; Miller and Alfheim, 1980; Pitts et al., 1977;
Talcott and Wei, 1977; Teranishi et al., 1978; Tokiwa et al.,
1977). Mutagenic activity of airborne particulates from a rural
site was investigated in only one of these studies; all of the
rural samples were reported to be inactive (Pitts, 1977). Part of
the observed mutagenic activity in samples from urban air is
probably due to polycyclic aromatic hydrocarbons (PAH). Several
PAH compounds cause cancer in animals and are also suspected of
causing cancer in man. However, these compounds require metabolic
activation before they can act as mutagens in the Salmonella test
85
-------
86 IN GRID ALFHEIM AND MONA M0LLER
system (Ames et al. , 1973). In the studies cited, all urban
samples were mutagenic both with and without metabolic activation,
indicating the presence of mutagens other than PAH.
Mutagens in ambient air originate from various combustion
sources, including residential heating and motor vehicle exhausts
(Miller and Alfheim, 1980; Lofroth, 1979: Wang, 1978), and may be
transported over long distances (Alfheim and Miller, 1979). Many
parameters will influence the mutagenicity of airborne particulate
matter from a given source, including the distance from the source,
meteorological conditions, the presence of other pollutants, and
the location of the source. In this work, we compared the
mutagenicity of urban air particulate matter at street level to
that at roof level, to determine the contribution of traffic to
the mutagenicity of urban air. We also compared the mutagenicity
of urban and rural airborne particulate matter, and we related
mutagenic activity to meteorological conditions and to the
composition of the particulate matter.
MATERIALS AND METHODS
Sampling
The urban sampling sites were all located in the center of
Oslo, Norway. Two sampling sites were in a narrow street with
heavy traffic (averaging 2000 cars/h during the day and 500 cars/h
at night). Site A, street level, was 2 m above the ground, and
site A, roof level, was 25 ra above the ground. Sampling site B
was in a park with much lower traffic frequency than at site A.
Roof samples were taken at two other locations in Oslo: sampling
site C was at a junction with heavy traffic, and site D was located
in a more industrial area with less traffic. Sites A and C were
considered to be more polluted than sites B and D, based on SO2
and soot measurements. Rural samples were collected in southern
Norway near the coast at Birkenes (site E) and in central Norway in
a mountainous area at Huramelfjell (site F).
Airborne particulate matter was collected during the winter
and spring of 1978 and 1979. Samples were collected on glass fiber
filters (Gelman type A-E) with high-volume samplers. At sites A
and B, the air was also passed through plugs of polyurethane (PUR)
to adsorb the more volatile compounds (especially volatile PAH)
(Alfheim et al., 1977). Separate day and night samples were
collected, each during two L2-h periods (~400 m3 air). At sites C
and D, approximately 700 m3 air passed through each filter during a
24-h period. About 2000 m3 were sampled at the rural sites. For
particle fractionation, a Sierra High Volume Cascade Impactor
Sampler with split filters (Gelman type C-230) was connected to the
s ampler.
-------
MUTAGENICITY OF AIRBORNE PARTICULATES
Extractions and Mutagenicity Testing
87
The organic compounds were extracted from the filters with
50 ml of acetone or 50 ml of cyclohexane, for nonpolar compounds,
in a Soxhlet apparatus for 16 h. For mutagenicity testing, the
samples were either concentrated by evaporation and tested directly
or were evaporated to near dryness and dissolved in
dimethylsulfoxide. Samples from locations A and B were extracted
with acetone, and samples from location C, D, E, and F were
extracted with cyclohexane. Acetone or cyclohexane extracts from
unused filters were tested as controls; positive controls were
2-aminoanthracene and benzo(a)pyrene (B[a]P).
Salmonella typhimurium strains TA98 and TA100 were kindly
supplied by Dr. B.N. Ames, University of California at Berkeley.
Liver homogenate fractions were prepared from male Wistar rats
injected with Aroclor 12 5A, 500 mg/kg i.p., five days prior to
preparation. The assay was carried out as described bv Ames et al.
(1975).
Analysis of PAH
PAH were analyzed by the procedure of Grimmer and Bohnke
(1972) as modified by Bjdrseth (1977). Internal standards were
added to the cyclohexane before extraction of the filters. The
extracts were shaken once with 50 ml and once with 25 ml
dimethylformamide (DMF):water (9:1). The DMF:water phases were
separated, water was added to give a DMF:water ratio 1:1, and the
samples were re-extracted with cyclohexane. The final cyclohexane
extracts were concentrated, first with a modified Vigreux column
under nitrogen and reduced pressure and then with a stream of
highly purified nitrogen at 30°C. The analysis was performed on a
Carlo-Erba Fractovap 2101 AC gas chromatograph with a glass
capillary column and flame ionization detector. The cyclohexane
extract, 2 yl, was injected splitless (Grob and Grob, 1969).
Chemical Fractionation
Acetone extracts from three samples collected at site A,
street level, were pooled, evaporated to near dryness, transferred
to diethylether, and fractionated into acidic, basic, and neutral
fractions. The acidic and basic fractions contained compounds that
were extractable from ether by aqueous sulfuric acid and sodium
hydroxide, respectively, and re-extractable into ether after
neutralization. The neutral fraction was further separated on a
silica column by elution with cyclohexane, benzene, and ether. The
cyclohexane fraction was subdivided into three fractions, on a
-------
88
INGRID ALFHEIM AND MONA M0LLER
column of silica gel with a layer of aluminum oxide on top, by
elution with pentane containing increasing amounts of ether.
One sample from each of these seven fractions was evaporated
to dryness and tested for mutagenic activity. The most mutagenic
fractions were analyzed by glass capillary gas chromatography, as
described above, or by combined gas chromatography/mass
spectrometry (GC/MS) (Bj<£rseth et al . , 1977).
RESULTS
Mutagenicity Testing
Preliminary results revealed very little or no mutagenic
activity in Salmonella strain TA100; therefore, strain TA98 was
used for further studies, including all those reported here. This
strain had 35 to 45 spontaneous revertants per plate. Twice the
number of spontaneous mutants was considered a significant
mutagenic response.
Mutagenic substances either act directly or require metabolic
conversion to mutagenic products. Unsubstituted PAH compounds
require activation with mammalian enzymes to be mutagenic in the
Ames test (Ames et al., 1973). The relative contributions of these
two groups of substances were estimated by testing the extracts in
both the presence and absence of liver microsomal preparations.
All extracts of airborne particulate matter collected on glass
fiber filters were mutagenic both with and without metabolic
activation. The dose response, expressed as number of mutants
per plate, was linear both with and without S-9. The standard
amount of S-9, 50 yl per plate, produced maximum activity for most
samples, but more S-9 was needed for a few extracts. Extracts from
the polyurethane filters showed no or very weak mutagenic
responses. Some samples taken at street level were fractionated
according to particle size. The results showed that only extracts
from particles less than 2.7 ym were mutagenic in both the presence
and absence of S-9.
Figure 1 shows the mutagenicity results for samples from site
A, street and roof levels, and from site B, expressed as
revertants per cubic meter of air. Daytime samples taken at street
level were about twice as mutagenic with microsomal activation as
without. (Mean values in February were 69 and 38 revert ants/m3
with and without S-9, respectively.) The activity of daytime
samples collected at site A, roof level, was approximately the same
with and without S-9; the same was true for site B daytime samples.
These samples were only 5 to 25% as mutagenic as the site A
street-level samples. For site A street-level samples, the
-------
MUTAGENICITY OF AIRBORNE PARTICULATES
SITE A, STREET LEVEL
89
QC
<
100
90
80
70
60
50
40
30
r>
J
to
l-
? 20
<
I—
£E
UJ 10
>
LU
ir
l-
III
60
50
40
30
201-
10
DAY
¦ With S 9
~ Without S-9
SITE A,
ROOF
LEVEL SITE B
25.1 30.1 7 2
31.1 8.2
NIGHT
SITE A, STREET LEVEL
tl
SITE A,
ROOF LEVEL
SITE B
L
-I
-EVEL
1 J
30-31.1
31.1 1.2
13-14,2
14-15.2
20-21.2
21-22 2
30-31.1
31.1-1.2
13-14.2 10-21.2
15 2 21-22.2
DATE
Figure 1. Mutagenicity of samples from site A (street-level and
roof-level samples) and site B, expressed as
revertants per cubic meter of air.
-------
90
INGRID ALFHEIM AND MONA MILLER
mutagenicity at night was only 20 to 25% of that during the day.
The activity of extracts from site A, roof level, and from site B
was relatively constant from one day to the following night.
The daily variation in the mutagenicity of nonpolar extracts
sampled at sites C and D is shown in Figures 2 and 3. The
mutagenicity of the samples from both locations was of the same
magnitude (expressed as revertants per cubic meter of air in
Figure 2). However, when expressed as revertants per milligram of
particulate matter, the mutagenicity was more than twice as high
at site D than at site C (Figure 3). The site C extracts were
more mutagenic in the winter (February) than in the spring (April),
both with and without metabolic activation.
The mutagenicity of urban air samples was compared to that of
samples collected at site E, on the southern coast of Norway, far
from any source of pollution. Samples from this location were
mutagenic in both the absence and presence of metabolic activation
(S-9) (see Figure 4). The highest mutagenic activity at site E was
obtained for samples collected during the winter. Furthermore, the
activity was significantly lower in samples representing air masses
coming from the north than in samples representing air masses
coming from the south. The mutagenic activity of these samples
(expressed as revertants per cubic meter of air) was 5 to 10% of
that for urban samples collected during the same period (sites C
and D). The samples from site F, a rural inland site in the
mountains, far from any pollution source, did not show any
mutagenic activity. These nine samples were all made during the
winter.
Chemical Fractionation of Street Samples
To characterize the mutagenic compounds in urban air, a few
Oslo samples were fractionated, and the fractions were tested for
mutagenicity. The results are given in Table 1. Only 48% of the
mutagenicity assayed in the presence of S-9 was recovered after
fractionation. Without S-9, the recovery was 30%. The mutagenic
activity of the combined fractions was approximately equal to the
sum of the activity of the individual fractions, indicating that
there were no synergistic effects between substances found in
different fractions.
The mutagenicity of the fractionated street-level samples was
found mainly in the neutral aromatic fraction (N-2). Some activity
was also associated with the acidic and the neutral fractions
(N-4) . The GC analysis showed that the N-2 fraction contained all
the common PAH except for the most volatile compounds. In
addition, this fraction contained unidentified polynuclear aromatic
-------
CO
£
<
rr
UJ
>
H
Lil
10 f
9'
s
7
E
5
4
3
2
-8" -10° -n" ~K° -9
° O _
• • . 9 *
8' -P°
9
HI
8 2 9 2 102 11 212 2 13 2 14 2
-12° -9° -n*
• • •
-2° .2"
9 t
1* I*
f 9
o
o
SITE C
l
17 2 IB 2 19.2
26 2
(
8 3 93
SITE D
9 2 10 2 11 2 12 2 13.2 14.2 15.2
17.210 219 2
25.2 26 2
DATE
J I
3.3 A3 6 3 73 8.3 9 3
5°
o
4° ^ 3* 1®
C3 9 &
7.4
1°
en
f 11.1 I
I
94104 114124 134
¦ Willi S-9
Q Without S-9
ft Sun
^ Rain or Snow
o Cloudy
UC
Ci
17.4
'"igure 2. Mutagenicity of roof-level samples from sites C and D, expressed as revertants
per cubic meter of air, and weather conditions on dates of sampling.
a
H
>
O
w
Z
O
H
<
O
n
>
S3
CO
O
53
'Z
F1
TJ
>
ta
H
O
a
>
H
M
Cfl
CD
-------
INGRID ALFHEIM AND MONA MQLLER
»-
<
no i-
100 -
so-
so}-
™f
50*
SO -
i
CO -
30
20 -
SO
• If
o o o
WEATHER CONDITIONS
9 9 ^ o o 3 ° ^ o 3 n o
SITE C
I
i
1 t_L
a< IM 112 12 2KJ *5 2 182 '92 25 2 26?
6 3 '33 74 io& 11.412113t
D
O
X
<
<
X
JJ
>
JJ
DC
cooi-
j*oj-
3 sop
340 |-
320-
300-
2 SO-
260
no
22 0
200
ibc r
150 j-
U3r
120 ¦
too ¦
£0 ¦
50 h
0|.
20'
Jl
SITE D
~ Win S-9
~ Witiojt S-9
~ Sun
Rain or Snow
O Clouny
ICZ <12 12 2 Ifc7 15 2 192192 25.2262 1)1] 6J 238.3
DATE
Figure 3. Mutagenicity of roof-level samples from sites C and D,
expressed as revertants per milligram particulate
matter, and weather conditions on dates of sampling.
-------
Location E
N W E-SF N S-SW W S-SE W NW 5 E N 5 N-NWW NE SW W N St
June| July |Sept | December | January | February | March | April
Figure 4.
Mutagenicity of samples from site K, expressed as revertants per cubic meter of
air. The main directions of the air masses during sampling are indicated as
N = north, W = west, E = east, and S = south.
-------
94
INGRID ALFHEIM AND MONA MQLLER
Table 1. Distribution of Mutagenicity Among Fraction3 of
Extracts of Airborne Particulate Matter3
Net Revertants/m^ of Air
Fract ion
With S-9
Without S-9
Unfractionated
59
35
Ac idic
5
2
Basic
1
1
Aliphatic, N-l
0
0
Aromatic, N-2
18
6
Aromatic, N-3
1
0
Oxygenated, N-4
3
2
Oxygenated, N-5
< 1
< 1
Sum
28 (48%)
11 (30%)
Combined fractions
28 (48%)
14 (39%)
aDay samples collected at site
A, street level ,
in February, 1979.
compounds (PNA). The compound
s responsible for
the mutagenicity in
the acidic and the N-4 fractions have not been
ident i fied.
Analysis of PAH
The gas chromatogr ams of samples from sites C and D allowed
quantification of 33 different PAH compounds. Table 2 shows the
total concentration of PAH, together with the concentrations of
pyrene and B(a)P. The concentration of PAH in particulate matter
from the city air was higher in the winter than in the spring (as
was its mutagenicity). The PAH concentrations in samples from
sites C and D were similar.
DISCUSSION
The contribution from traffic to the mutagenicity of the air
at street level appeared to be substantial. The mutagenicity of
daytime samples at street level in the presence of S-9 was 4 to 20
times higher than for the corresponding samples from roof level or
-------
MUTAGENICITY OF AIRBORNE PARTICULATES 95
Table 2. Concentration of PAH in Airborne Particulate Matter

Concentrat ion
of PAH (ng/m^)
Sampling Site
Date
of Sampling
Pyrene
B(a)P
J>AH (33 comp.)
C
1978
February
20
66.0
12.0
414

M
24
16.0
4.1
127

tf
28
3.9
2.1
72

March
05
1.2
0.6
45

April
04
2.1
0.9
17

M
18
1.5
1.0
14
D
1978
February
20
40.0
7.9
319

March
05
1.9
1.0
51

April
04
3.5
1.4
29
C
1979
February
14
71.0
11.0
355

ii
20
5.5
3.7
60

March
04
1.5
0.1
7

II
10
1.1
0.6
14
from site B.
Furthermore, the mutagenicity
at street
level varied
with traffic
frequency (i.e. ,
day '
vs. night)
, whereas
activity of
samples from
roof level and site B
showed no
such variations.
The mutagenicity of street samples was enhanced by the presence
of liver microsomes, indicating a greater contribution from gasoline
engine exhaust than from diesel exhaust. Exhaust from diesel
engines is more mutagenic in the Ames test than is exhaust from
gasoline engines. However, the activity of diesel exhaust is lower
in the presence of S-9, while the opposite ig true for gasoline-
engine exhaust (Lofroth, 1979). It has been suggested that nitro-
substituted PAH contributes to the mutagenic activity of exhaust
from traffic (Wang, 1978). Most of these compounds are direct-
acting mutagens in the Ames test.
All mutagenic activity at street level was associated with
particles less than 2.7 ym in diameter. Furthermore, only traces
of PAH were found in particles greater than 2.7 ym (A1fheim et al.,
1977). This agrees with cascade impactor measurements made in
Belgium of the size distribution of 60 organic pollutants. Van
Vaeck and Van Cauwenberghe (1978) showed that 95 to 98% of the PAH,
-------
96
INGRID ALFHEIM AND MONA M0LLER
as well as some heterocyclic polyaromatics (PNA), are found in
particles less than 3 urn in diameter. Most carbocyclic acids and
aliphatic hydrocarbons (90%) are also associated with particles
less than 3 ym.
Fractionation of extracts from particulate matter at street
level showed that most of the mutagenic activity recovered was due
to PNA. A similar investigation based on roof sampling in Japan
showed a somewhat different result (Teranishi et al., 1978). In
this study, the mutagenicity was evenly distributed among the
acidic, aromatic, and oxygenated fractions 15% in each fraction),
with slight activity in the basic fraction. Of the original
activity, 76% was recovered after the fractions were combined.
Fractionation of particulate matter of roof samples in Stockholm
(Lofroth, 1979) gave a distribution of mutagenicity similar to that
reported from Japan. The recovery in our experiments was somewhat
lower, possibly because the components of particulate matter
collected at street level were more labile. Earlier
investigations, in which particulate matter from a heavy traffic
area was fractionated and tested for carcinogenicity through
skin-painting experiments on mice (Wynder and Hoffman, 1965),
showed that the aromatic fraction was associated with carcinogenic
properties, while the acidic and oxygenated fractions were
associated with promoter effects.
The mutagenicity of roof-level samples (A, C, and D) and
samples from site B probably has main sources other than traffic.
Mutagenicity and PAH concentrations were both higher in the
February samples than in the spring samples, which is most likely
explained by the higher use of residential-heating fuels during
the winter months in Oslo. The average temperatures during the
sampling period were -8.3°C for February, -0.9°C for March, and
+3°C for April. Using measurements at street level, we found that
the amount of B(a)P could only account for ~1% of the mutagenic
activity in day samples and up to 4% in night samples. Data for
extracts of roof-level samples from other cities show that the
mutagenicity of urban air (revertants per cubic meter) was of the
same magnitude in Oslo as in Stockholm (Lofroth, 1979) and Los
Angeles (Pitts et al., 1977).
Meteorological conditions may also contribute to seasonal
differences in mutagenicity levels and cause daily variations in
the mutagenicity of winter samples. The mutagenicity (revertants
per cubic meter of air) was highest on cold, clear weekdays with
little wind and stagnant air.
Daily measurements of SO2 and soot in the air have been made
at these same sampling sites by the Oslo City Department of Health.
The most mutagenic samples from each sampling site also had the
highest concentrations of SO2, soot, and PAH at that site.
-------
MUTAGENICITY OF AIRBORNE PARTICULATES
97
However, while mutagenicity (revertants per cubic meter of air) and
PAH concentrations were approximately equal for the two sampling
sites, the values for SO2 and soot were nearly twice as high at
site C as at site D. Contributions from different pollution
sources may explain why SO2 and soot concentrations were higher at
site C than at site D, while mutagenicity and PAH concentrations
were equal.
Mutagenic activity in strain TA98 (revertants per cubic
milligram of particulate matter) was high on days with a low total
concentration of particles in the air—mainly days with rain and
snow. The mutagenicity was especially high at location D on such
days. Variations in the amount of nonmutagenic particles will
greatly influence the mutagenicity (revertants per milligram of
particles). Currently, we consider revertants per cubic meter of
air to better express the mutagenic potential of the air samples.
However, this subject requires further investigation.
For most samples from the rural site in southern Norway,
mutagenicity either was the same with and without metabolic
activation or was higher with it. The activity with metabolic
activation might be explained by the presence of PAH compounds in
the samples. Such compounds have previously been demonstrated in
long-range-transported aerosols collected at this site (Lunde and
Bj4>rseth, 1977), and their amount have been shown to vary with the
origin of the air masses in the same way as the mutagenic effect
does. Like the mutagenicity, the concentration of PAH at site E
also was 5 to 10% of that found in samples from Oslo. At site F,
the PAH concentrations were up to 1% of the corresponding Oslo
values.
REFERENCES
Alfheim, I., and M. Miller. 1979. Mutagenicity of long-range
transported atmospheric aerosols. Sci. Tot. Environ.
13:275-278.
Alfheim, I., M. Miller, S. Larssen, and A. Mikalsen. 1977.
Undersifikelse av PAH og mutagene stoffer i Oslo-luft — relasjon
til trafikk. Report to Norwegian State Pollution Control
Authority. 42 pp.
Ames, B.N., W.E. Durston, E. Yamasaki, and F.D. Lee. 1973.
Carcinogens are mutagens: a simple test system combining
liver homogenate for activation and bacteria for detection.
Proc. Natl. Acad. Sci. USA 70:2281-2285.
-------
98
INGRID ALFHEIM AND MONA M0LLER
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian-microsome mutagenicity test. Mutation Res.
31:347-364.
Bjitrseth, A. 1977. Analysis of polycyclic aromatic hydrocarbons
in particulate matter by glass capillary gas chromatography.
Anal. Chim. Acta 94:21-27.
Bjrseth, A., G. Lunde, and N. Gjs. 1977. Analysis of
organochlorine compounds in effluents from bleacheries by
neutron activation analysis and gas chromatography/mass
spectrometry. Acta Chem. Scand . B 31 :797-801.
Dehnen, W. , N. Pitz, and R. Tomingas, 1977. The mutagenicity of
airborne particulate pollutants. Cancer Lett. 4:5-12.
Grimmer, G., and H. Bohnke. 1972. Bestimmung des Gesamtgehaltes
aller Polycyclischen Aroraatischen Kohlenwasserstoffe in
Luftstaub und Kraftfahrzeugabgas mit der Capillar-gas-
Chromatographie. Z. Anal. Chem. 261:310-314.
Grob, K. , and G. Grob. 1969. Splitless injection on capillary
columns (Part I and Part II). J. Chromatogr. Sci. 7:584-591.
Henderson, B.E., R.J. Gordon, H. Menck, J. Soohoo, S.P. Martin, and
M.C. Pike. 1975. Lung cancer and air pollution in
Southcentral Los Angeles County. Am. J. Epidemiol.
101:477-488.
Higginson, J., and C.S. Muir. 1973. In: Cancer Medicine. J.F.
Holland and R. Frei, eds. Lea and Febiger: Philadelphia.
Hueper, W.C., P. Kotin, E.C. Tabor, W.W. Payne, H. Falk, and E.
Sawicki. 1962. Carcinogenic bioassays on air pollutants.
Arch. Pathol. 74:89-116.
Lofroth, G. 1978. Mutagenicity assay of combustion emissions.
Chemosphere 7:791-798.
Lofroth, G. 1979. Salmonella/microsorae assays of exhaust from
diesel and gasoline powered motor vehicles. Presented at
International Symposium on Health Effects of Diesel Engine
Emissions, Cincinnati, OH.
Lunde, G., and A. Bj^rseth. 1977. Polycyclic aromatic
hydrocarbons in long-range transported aerosols. Nature
268:518-519.
-------
MUTAGENICITY OF AIRBORNE PARTICULATES
99
Menck, H.R., J.T. Casagrande, and B.E. Henderson. 1974.
Industrial air pollution: possible effect on lung cancer.
Science 183:210-212.
Miller, M. , and I. Alfheim. 1980. Mutagenicity and PAH analysis
of airborne particulate matter. Atmos. Environ. 14:83-88.
Pitts, J.N., Jr., D. Grosjean, and T.M. Mischke. 1977. Mutagenic
activity of airborne particulate organic pollutants. Toxicol.
Lett. 1:65-70.
Talcott, R., and E. Wei. 1977. Brief communication: airborne
mutagens bioassayed in Salmonella typhimurium. J. Natl.
Cancer Inst. 58:449-451.
Teranishi, K., K. Hamada, and H. Watanabe. 1978. Mutagenicity in
Salmonella typhimurium mutants of the benzene-soluble organic
matter derived from airborne particulate matter and its five
fractions. Mutation Res. 56:273-280.
Tokiwa, H., K. Morita, H. Takeyoshi, D. Takahashi, and Y. Ohnishi.
1977. Detection of mutagenic activity of particulate air
pollutants. Mutation Res. 48:237-248.
Van Vaeck, L., and K. Van Cauwenberghe. 1978. Cascade impactor
measurements of the size distribution of the major classes of
organic pollutants in atmospheric particulate matter. Atmos.
Environ. 12:229-239.
Wang, Y.Y., S.M. Rappaport, R.F. Sawyer, R.E. Talcott, and E.T.
Wei. 1978. Direct-acting mutagens in automobile exhaust.
Cancer Lett. 5:39-47.
Wynder, E.L., and D. Hoffman. 1965. Some laboratory and
epidemiological aspects of air pollution carcinogenesis. J.
Air Pollut. Contr. Assoc. 15 : 155-159.
-------
Intentionally Blank Page
-------
DETECTION OF GENETICALLY TOXIC METALS BY A MICROTITER
MICROBIAL DNA REPAIR ASSAY
Guylyn R. Warren
Chemistry Department
Montana State University
Bozeman, Montana
INTRODUCTION
For some years, our laboratory has been involved in assessing
the genetically toxic effects of inorganic chemicals found in the
environment near raining and smelting operations (Tindall et al.,
1978; Warren et al., 1979) and of metal-containing pesticides used
in Montana (Warren et al., 1976). Although many inorganic species
are known or suspected carcinogens or are genetically active in
many test systems (Flessel et al. , 1979; Sunderman, 1978), most
existing short-term biological screening methods are unsuitable for
use with this class of suspected carcinogen. Only two systems, the
Bacillus subtilus rec assay (Nishioka, 1975; Kaneraatsu et al.,
1980) and an in vitro DNA synthesis fidelity assay (Loeb et al.,
1979), have been useful for a large number of inorganic chemicals.
Mutagenicity of some inorganic chemicals has been demonstrated in a
CHO/HGPRT assay (Hsie et al., 1979) with considerable technical
difficulty due to the general toxicity of inorganic chemicals.
Green and Muriel (1976) have used a repair-deficient series of
Escherichia coli B strains to detect mutagenicity of some chromate
salts; also by this method, they found that nickel chloride salts
do not cause differential lethality.
Recent discoveries about the relationship of DNA repair
functions to mutagenesis in bacteria (Witkin, 1976; Kimball, 1978)
and of error-prone repair mechanisms such as inducible SOS repair
and another excision repair mechanism (Hanawalt et al., 1979),
provide reasons for using batteries of repair-deficient organisms
in a screening system, rather than using a single repair-deficient
strain, such as pol A (Hyman et al., 1980). The pleiotropy of rec"
mutants is a major problem in using only one double-rec-deficient
101
-------
102
GUYLYN R. WARREN
strain, such as Bacillus subtilis M45, for the assay. Such
rec" mutants exhibit permeability changes, possibly due to cell-
wall and membrane-surface defects, as in rec A (Tomizawa and Ogawa,
1968). Differences in inhibitory effects of test chemicals might
be due to differences in their rates of penetration into the
repair-defective strain as compared with wild type rather than to
differences in DNA repair capacity.
To rapidly screen large numbers of samples, we have developed
a microbial repair assay using a series of singly- and raultiply-
mutant DNA-repair-deficient strains in an E. coli K12 background.
The mutational repair defects of the K12 series have been studied
in much more detail, both genetically and biochemically, than have
those of E. coli B or J3. subtilis (Witkin, 1976; Kimball, 1978;
Hanawalt et al., 1979). The strains are stable in culture and easy
to grow, and they carry several biochemical markers that permit
genetic manipulation for strain construction, and possibly
mutagenesis assays, in the same repair strains. A number of metal
salts and metal salt compounds have been tested in the K12 system
(Warren et al., in press a, b; MS). Our results correlate well
with mutagenesis as assayed by the Ames test. This paper describes
in detail the repair assay and its use to assay metal-containing
samples .
METHODS
Construction and Testing of Strains
The derivation of the repair-deficient E. coli bacterial
strains is diagrammed in Figure I. In each case, isogenicity of
the repair defectives with the wild type AB1157 was optimized, to
minimize mutational effects other than on DNA repair. For
discussions of repair defects and phenotypes see Kimball (1978),
Hanawalt et al. (1979), and Witkin (1976).
The bacteria are stored at room temperature in Lederberg stabs
of nutrient broth (Difco) and have remained viable for at least
three years. Each month, working slants are made from single
colonies and stored at room temperature. Bacteria from these
single colonies are streaked onto minimal medium plus the amino
acids required by AB1157 and onto minimal medium alone, to verify
the strains' nutritional requirements. Every two weeks, single
colonies are isolated from the working slants to check the repair
characteristics of the strains. Two colonies of each strain are
isolated and inoculated into Mueller-Hinton (MH) broth (5 ml), and
cultures are grown overnight at 37°C. Ultraviolet-light
sensitivity of each culture is checked by the rapid method of
Greenberg (1967) as follows: A streak of each strain is placed by
capillary pipette on each of seven MH agar plates. Plates are
-------
MICROTITER DNA REPAIR ASSAY FOR METALS
103
AB1157 (a Y10 E_. coli K12, F",X") wild-type for repair
(a!
f-thr-1
leu-6
tni-1
lacY 1
galK2
ara-14
xyi-5
mt',-1
proA2
his-4
argE3
str-31
tsx-33
(MA
GW950
(a)
lex B31
JC5519
(a)
rec B21
rec C22
GW901
(a)
rec F143
( naiR R \ GW802
aricsr P1.JC5Q88 alB } ^ (a)
(at
urv A6
reversion
P1.JC5088(na,B
reversion
¦*- GW801
(a)
rec A56
PI.
GY3565
NG
J1§-*AB2494 P1.JC5088(nalB
(a) reversion
lex A1
P1.PAM26
PAM10Q
(a)
lex C1
UV
* AB1899
la)
lon-1
TRIM
P1.AG61
P1.P3478
-*~GW19t
P1.AB1886
uvr A6
rec A56
GW803
(a)
lex A1
rec A56
po) A-
met A
met
sel'n
¦»~GW991
la)
pol A-
uvr
Figure 1. Construction of a series of E, coli K12 repair-deficient
strains from strain AB1157 (wild-type for repair).
-------
104
GUYLYN R. WARREN
immediately irradiated with known fluences of 254-nm light (for
example, 0, 20, 50, 100, 200, 300, and 400 ergs/mm2) and then
incubated in the dark at 37°C. Results such as those shown in
Table 1 can be determined after incubation for 18 h. Any culture
not giving the expected result is rejected, and a new colony of
that strain is picked and tested. Acceptable cultures are labelled
for use and stored. Broth cultures may be used for testing for up
to two weeks if stored at 2°C. Several of the repair-deficient
strains are still being evaluated as testers, and those with the
most useful traits will be chosen. An example is the lex C mutant
strain, PAM100, which is uniquely sensitive to nickel compounds.
We are constructing a second identical isogenic series of
strains containing an rfa (deep rough) mutant locus. This series
will be tested later this year. A series of repair-deficient
strains containing the ochre mutant trp~ locus from WP2 and one
strain containing two of the Yanofsky trp~ series (Yanofsky, 1963)
for detecting frameshifts and base pair substitutions are under
c onstrue t ion.
ASSAYS
Initially, we developed a sensitivity disc assay in which
zones of inhibition caused by inorganics were compared between each
repair-deficient strain and the wild-type (repair-proficient)
strain. Such a method allowed differential inhibition (in many
instances, lethality) to be grossly quantitated, DNA damage of
some sort should have been the cause, because the only differences
between strains were in repair capacities.
The sensitivity disc assay was done by spreading 0.1 ml of
each broth culture directly on a complete-medium agar plate.
Sterile discs of paper were impregnated with a known concentration
of test agent . Three of these discs were placed on each freshly
spread agar surface, and plates were run in triplicate. Several
different media were tried initially, in an effort to maxmimize
zone size differences. For a group of salts of known metal
carcinogens (chromium, cadmium, mercury, cobalt, and arsenic), the
largest differences were obtained on solid medium containing 10 g/1
agar and MH broth (Difco) . When 0.2% glucose was added, chromate
compounds gave a much clearer and larger differential response,
indicating a glucose requirement for rapid bacterial growth.
Addition of 0.52 sodium chloride increased the inhibitory effects
of cadmium and cobalt. The standard medium now used is MH plus
glucose and salt. The effects of these additions to MH medium are
shown in Table 2 as differences in zones of inhibition (mm).
Differences greater than 5 mm were significant at the 95% level
(Tindall, 1977).
-------
Table 1, Verification of Repair Deficiencies hy Ultraviolet Irradiation
Growth in Streaks oti Mil Plates After Irradiation with 254 nm Light''

— —

— — — — — —
— — —
— — —

Fluentes (
ergs/ram^)

Repair

—— ¦—

— ——
Strain
Defect3

0
20
50
100
200
300
400
ABl157
wild type

++ f
4+4-
H- +
++ +
+++
++*
+++
AB1886
uvr A6

+++
44
- (I)
- (20)
- (0)
- (1)
- (2)
AB1899
Ion-1

f f +
fff
+++
i-i-f
+ + +
M- f
- (8)
AB2494
lex Al

+++
444
+
+-
- (15)
- (0)
- (0)
GW801
rec A56

++
+ -
- (>200)
- (35)
- (0)
- (0)
- (0)
PAM100
lex CI

++
4 «- *
f -
- ¥
- (20)
- (0)
- (0)
GW802
uvr A6, rcc
A56
+ + +
- (3)
- (0)
- (0)
- (0)
- (0)
- (0)
GW803
lex Al, rec
A56
+ hl-
- (>200)
- (45)
- (8)
- (3)
- (0)
- (0)
GWJ 91
pol Aj

+++
+++
++ ~
+++
+~

- (36)
aAclual kill curves can be determined if necessary.
^ + + 1- = maximal growth; ++ = noticeably less dense streak; + - thin streak; +- "
lethaLity; ~ + = spotty but >500 colonies; - = kill (no. of colonies counted).
O
pd
O
H
m
m
a
•?>
>
Cfl
J*
I—*
PS
>
CO
c»
>
<<
*n
o
58
2
H
H
>
r
m
v is ible
o
-------
o
Carcinogenic Metal Salts
K2Cr04a
No2Te03f
BeS04c
KSbO^(C4H4O5)•l/2H2Od
Med iume: MH MHO MHS MHGS MH MHG MHS MHGS MH MUG MHS MHGS MH MHG MHS MHGS
Strain
AB1886

0
-10
3
-4
5
5

3
4
0
0
0
0
2
5
4
2
GW801

2
3
3
6
-4
1

-2
0
0
0
0
0
-1
7
6
4
AB2494

-4
-14
-2
2
12
16

16
8
0
1
0
0
-2
6
I
1
GW802

1
10
3
12
-2
3

0
2
10
10
8
0
-1
11
4
it
GW803

1
4
3
5
16
6

3
8
0
0
0
0
-7
0
2
1
GW19 I

0
3
2
9
10
14

15
12
0
0
1
0
-6
3
-3
-1
a0. 10 M
in
H'} 0,
10
Vii/disc

*MH =
MueLler-Hinton agar
(Di fco)
.

b0.01 M
in
II2O ,
15
[i 1 / disc .

MHG
=
MH
agar
+ 0
. IZ glucose..

c0. 10 M
1 n
H?0,
20
pL/disc

MHS
=
MH
agar
+ 0
.05% NaCl.

d0.10 M
in
h2o,
20
MI / d l s c

MHGS
=
MH
[ agar
+
glllCOSR
+ NaCl.

O
c
t-1
So
>
S3
PO
M
-------
MICROTITER DNA REPAIR ASSAY FOR METALS
107
Kaneraatsu et al. (1980) used a cold incubation period after
spreading and disc placement to enhance the inhibition by test
metals in the B. subtilis rec assay (this allows the test substance
to diffuse before bacterial growth starts). We have observed this
effect with most chemicals, although some (mostly organic
compounds) have shown weaker responses after cold incubation than
without it.
Because migration through an agar medium is affected by the
depth and viscosity of the agar, temperature, charge, and pH, and
because all of these parameters are subject to laboratory
variation, we searched for a more reproducible and sensitive
method. The disc repair assay required high levels of test agent
and was not quantitative, due to solubility problems. McCarroll et
al. (1979) have developed a microtiter assay system to study
effects of organic toxicants using repair-defective strains of E.
coll B obtained from Green and Muriel (see Green and Muriel, 1976;
McCarroll et al . , 1980a, b). We have adapted this system to the
repair assay with E. coli K12. Microtiter techniques (Cooke
Engineering, 1972) can be used to quantitatively measure inhibition
of growth (as in antibiotic testing) and also to quantitatively
measure lethality, since colonies can be counted in each well at
the end of a treatment period (McCarroll et al., 1980b).
Mutagenesis testing also can be done from the same microtiter plate
wells, thus generating actual mutation frequencies (mutants per
survivor). Several loci could be monitored for mutagenesis at the
same time.
A microtiter plate set up for an eight-strain test is shown in
Figure 2. Initially, 0.05 ml of MH broth is added to each well of
the microtiter plate, using a Tridak Stepper, 3-ml syringe, or
Gilson Repetman. Rows 1 and 12 receive a double dose. Rows A
through H will each receive one of the strains of bacteria, with A
receiving the wild type. Row 1 will contain only bacteria and
medium, to serve as the cell control. Row 2 receives 0.05 ml of
the test agent in solution. For most inorganic toxicants, a
convenient initial concentration is 2 mg/ml in water, if possible
(or dimethylsulfoxide if necessary). The test agent is added to
row 2 with a 1-ml Tridak stepper. Serial dilutions are then
performed from row 3 through row 11, using microtiter diluters
(Cooke), which hold 0.05 ml and do automatic serial dilutions. The
final dilutions of mutagen are shown in Figure 2 and range from 1:4
(row 2) to 1:2048 (row 11). Row 12 is left uninoculated as a
sterility control for medium, mutagen, and solvent. Having been
checked and stored as described above, each of the bacterial
cultures is diluted in MH to a final viable count of ~ 1 x 106
colony forming units (CFU)/ml, for optimal sensitivity in
microtiter. Because several of the repair-deficient strains
(especially the multi-mutants; see Table 1) do not reach the same
cell density overnight as does the wild type, the appropriate
-------
108
GUYLYN R. WARREN
>
CD
NJ
<£>
a
«J /
o
"0
>
>
>
>
£
$
$
UJ
CD
DC
30
00
cc

—*

o
o
o

CD
CO
i
CJ

§
CO
CD
00
CT>

•nj
UNTREATED
V1ETAL COMPOUNDS 1:4
WELL
DILUTION
FACTORS

1:1024
1:2048
STERILITY CONTROL
K H ^
X ) !' >( )( )i<
—y"'\> —'V
¦, -X X

i\ )' 1 (. j I ! ( ,
I ) ( ) ' ! ¦ ¦{ '
/.. K. ... y
- --V--
JO
'J1
X
>
2
CO
Figure 2. Sample raicrotiter plate.
dilution must be established for each strain. Bacterial culture
dilutions should be discarded daily; they may be used all day if
kept on ice. With a 1-ml Tridak stepper, 0.05 ml of bacterial
culture dilution is added to each well of the appropriate row for
that strain.
Initial concentrations of test agents can be varied, within
limits of solubility, to test for effects throughout a very large
concentration range. Plates are sealed with tape to retard
evaporation. Sterile lids can be used for short-term treatments,
but incubation for over four hours results in significant fluid
loss by evaporation. The standard incubation time is 24 h. Some
tested compounds initially inhibit growth, but then either
-------
MICROTITER
DNA REPAIR ASSAY FOR METALS
109
resistant organisms overgrow the wells or the test agent is
metabolized and becomes ineffective. For such agents, a shorter
incubation time can be used.
Mammalian microsomal fractions can be used for activation in
the repair assay. Positive controls such as 2-aminoanthracene show
good responses with either of two methods of addition of rat liver
homogenate (S—9). The S-9 mix is prepared according to Ames et al.
(1975), and 0.05 ml is added with the mutagen to wells in row 2, in
place of the initial 0.05 ml of MH broth. Both substances are
simultaneously serially diluted through the wells. Alternatively,
one may add 0.05 ml of S-9 mix in place of MH broth to each well in
the plate and then dilute only the test mutagen. Throughout either
procedure, all plates are kept on ice. For all positive controls
we have studied, the former method results in more differential
lethality than the latter. After incubation, the plates are read
visually, using a microtiter mirror. For each of the lettered
rows, the totally inhibited well with the lowest test-agent
concentration is determined visually and recorded. From the
dilution factor for that well and the known concentration of test
agent, a minimal inhibitory concentration (MIC) can be determined
for each strain. Differential inhibition is given as the factor of
increase in MIC over that of the wild-type strain, in this case
AB1157 (i.e., MIC for repair-deficient strain/MIC for repair-
proficient strain).
RESULTS
Table 3 compares the results of the microtiter repair assay
and the disk assay, using eight strains of E. coli and ten rhodium
complexes recently synthesized by Abbott (Warren et al., in press
a) .
In the optimal test, an inhibitory test-agent concentration
for the wild-type bacteria would be determined. In practice, this
is not possible for all agents, because the wild type often is
resistant to all concentrations within the range of solubility of
the test agent. In such cases, the highest concentration that
allows determination of an MIC for the most sensitive strain has
been used (less than 11 wells inhibited). The factor of increase
in MIC is therefore underestimated.
A series of substitutionally inert rhodium complexes was used
to demonstrate the utility of the method described here, not only
for detecting DNA damage but also as a screening test preceding the
Ames reversion assay (Warren et al., in press a). Dose
responsiveness was predictable, and as the dose was doubled, the
MIC also doubled, until the entire plate was killed. The factor of
increase in MIC for any one test agent did not change once a lethal
-------

Table 3. Comparison
0 f Mi crnt i
tir and
Di sr. Repfli r As
a,iyi , Us i tip.
Rhnd i urn
Cnmpl oki'S

Inhibition of
Stra ins
Tested hy the
ni.H«-. (d) 3
and Mi c rot i ter
(M)b Methods

GWB02
CW80 3

ABI8B6

GWflOl

AB2494
uvr
Aft
lex
Al
AHI 899

uvr A6

rec A56

lex A1
rec
ASA
rec
A56
1 on- 1

Complex
D
M
D
M
D M
n
M
I)
M
n
M
I Rli( Fy r )/,Bi 2 | Br
18.0
16
23.4
32
12.7 0
2 t*.e*
512
22.0
8
13./)
U
IKh(CH3CN)jCl 3 ]
9.3
0
7.3
4
2.0 0
12.0
12
fl.7
4
2 . 7
0
lRh(Bipy)2Cl2]CI
7.5
(1
0
0
0 0
i 6.5
16
n
f)
0
0
IRh(3Fic)4Cl7]Cl
15.3
16
23.0
2
12.7 0
28.3
I2H
20 .0
0
l'». 7
Ih
I Rli( Fy r ) /,Cl 2 ] Cl
23.0
8
22 .6
0
7.5 0
27.0
256
235 .0
0
123.0
0
I Rli( Phen) 2^1 2 )C1
9.0
0
2.0
0
0 0
7.0
64
1.5
0
0
0
1Rh(en)2)Cl2
0
0
0
0
0 0
7.6
16
0
0
0
0
HliCl3(H2r))j
6.6
0
3.3
0
O
C
8.6
8
4.3
0
2.0
0
lHli(NH1)/iCl2|r.l
0
0
0
0
0 0
12.0
16
1 . 3
0
0
0
(Rh(Trien)Cl2ICl
1 .1
0
17.3
0
0.5 0
23.3
64
16.3
0
0.6
2
""For test agents at 20 ing/ml H20, 20 ill/disc; means of results train f1 discu; expros.ifid as difference
in diameter from wild type (mm).
''For test agents at 2 mp,/ral without S 9; means of results from duplicate plates; expressed .is
factor of increase in MIC over that of wild type.
o
c:
M*
r
~<
so
5=
>
sa
pa
M
-------
MICROTITER DMA REPAIR ASSAY FOR METALS
111
2.9
2.7
2
z
(A
<
UJ
(X
o
2 5 _
^ 2.3 -
a 2.1
<
X)
LO
<
o
<
1.9
1.7
3 i
1.5
1.1
0.9
1
3 _i
• ¦ • ¦
TA92 •
TA100 ¦
• • i
/
X
_L
8.00
9.00 10 00
Log Hisr Revertants/nmol (Salmonella)
11.00
Figure 3.
Correlation between sensitivity of the microtiter
repair assay (using strain GW802, uvr A6, rec A56) and
mutagenicity (strains TA92 and TAl00) in the Ames
test for a group of rhodium complexes.
concentration was reached in the wild-type strain. A comparison
between the results of the microtiter repair assay and the Ames
test for these compounds is given in Figure 3. The coefficient of
correlation between the results of the two tests was 0.92,
indicating agreement. No Ames-positive test agents in a series of
23 rhodium complexes were missed by the repair assay, while one
repair-positive agent (very weak) was Ames negative.
An S-9 dose response in the microtiter repair assay with an
organic herbicide, diallate, is shown in Figure 4. Diallate, an
indirect-acting mutagen, was used because this compound is small
-------
O)
<1>
.a
E
D
C
C
o
sz
c
Aroclor 12b4-indur.ed
(24.8 my/ml protein)
Strain:
AB1107
AB1886
GW801
AB2494
GW802
GW803
Hhenobarbital induced
(13 3 mfj/ml protein)

Percent S-9 Fraction in the S 9 mix
(ratio of other constituents constant)
Figure 4. Enzyme? activation of diallato In a microtiter repair assay-
to
O
c:
<
r
r°
>
pa
50
w
2;
-------
MICROTITER DNA REPAIR ASSAY FOR METALS
113
and should have no difficulty penetrating cell walls. The results
were related to the enzyme dose. Also, the type of inducer used
affected the response just as it does in a mutagenesis assay. We
have previously noted that a higher concentration of protein is
required of phenobarbital-induced enzyme than of Aroclor-1254-
induced S-9 for diallate mutagenicity in the Ames test. While the
inhibitory activity of several metal salts was reduced by S-9
(zinc, palladium, mercury, antimony, arsenic, selenium, nickel,
manganese, and cadmium), we have seen only one in which S-9
increases genetic toxicity (telluriura)(Rogers and Warren,
unpublished data).
This system is designed for testing complex mixtures. It has
been used to document the differential lethality of a sample of
smelter rafter dust, air filter particulate extracts, and urine
concentrates (Warren et al., 1979).
DISCUSSION
The microtiter repair assay techniques described here employ
standard microtiter equipment and can be automated. A scanning
device is available for reading optical density directly from the
plates for growth determinations. Various dyes can be used to
indicate growth (e.g., bromophenol blue)(Green and Muriel, 1976).
In cases of agents which kill cells but cause filamentation of
some strains, plate counting may be required, because filamentation
increases optical density without an increase in the number of
viable cells. This is particularly likely with strains AB1899 and
PAM100, both of which filament easily and in response to any
chemical insult .
Strains of E. coli have been constructed with mutations to
block incision (uvr A, B, C) and to partially block excision (pol
Ai , rec B, C), resynthesis and conditioned responses (rec A, lex
A), and recombination (rec A, rec F). Mutagenesis might be caused
by less-well-known repair mechanisms that could be unmasked if the
more rapid error-free repair were removed by mutation and post-
replicative repair were incapacitated as well. The strain GW802
has these defects and is easily the most sensitive strain in the
battery at this time. There is some evidence that an SOS-like
conditioned response system exists in mammals and may be
responsible for tumor induction by some mutagens (Radman, 1977).
Many of the metal mutagens and all of the substitutionally inert
transition metal complexes we have tested require the pKMlOl
plasmid in Salmonel1 a for mutagenesis and do not require a
functional uvr system. Venturini and Monti-Bragadin (1978) found
that no mutation is caused by platinum without the lex function in
E. coli, but that lex* E. coli can be mutated without carrying
pKMlOl. Therefore, these strains are suitable for metal-ion
-------
114
GUYLYN R. WARREN
mutagenesis without the pKMlOl repair system. Because many
mutagens can function as curing agents, a system that does not
require a plasmid would seem Co be advantageous.
If an agent's lethal and mutagenic effects on each bacterial
strain are known, then the general type of DNA damage caused by
that agent can be predicted. For instance, uvr~ strains cannot
recognize DNA helix distortions; agents differentially lethal to
uvr~ strains but not mutagenic are most likely to be strand cross-
linkers (Murray, 1979). Green and Muriel (1976) give five classes
of damaging agents and note that some mutagens fit in more than
one category.
It has also been shown that repair assays correlate
better than do mutagenesis or prophage induction with anti-tumor
activity of alkylating agents (Tamaro et al., 1977), indicating
another useful screening capacity for the microtiter repair system.
CONCLUSIONS
The microtiter technique is rapid, cost effective, and
sensitive and requires very little sample. The bacterial strains
are easily grown and checked and provide an accurate screening
test for both organic and inorganic genetically toxic chemicals
and complex mixtures, regardless of their activity in mutagenesis
systems. This test is a very useful prescreen for mutagenesis
assays.
REFERENCES
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmone11a/
mammalian-microsome mutagenicity test. Mutation Res.
31:347-363.
Cooke Engineering. 1972. Handbook of Microtiter Procedures.
T.B. Conrath, ed. Dynatech Corp.: Cambridge, MA. 475 pp.
Flessel, C.P., A. Furst, and S.B. Radding. 1979. A comparison of
carcinogenic metals. In: Metal Ions in Biological Systems,
Volume 10, Chapter 2, Carcinogenicity and Metal Ions. H.
Sigel, ed. Marcel Decker: New York.
Green, M.H.L., and W.J. Muriel. 1976. Mutagen testing using Trp+
reversion in Escherichia coli. Mutation Res. 38:3-32.
Greenberg, J. 1967. Loci for radiation sensitivity in Escherichia
coli strain Bg_i. Genetics 55:193-201.
-------
MICROTITER DNA REPAIR ASSAY FOR METALS
115
Hanawalt, P.C., P.K. Cooper, A.K. Ganesan, and G.A. Smith. 1979.
DNA repair in bacteria and mammalian cells. Ann. Rev.
Biochem. 48:783-836.
Hsie, A.W. , N.P. Johnson, D.B. Couch, J.R. San Sebastian, J.P.
O'Neill, J.D. Hoeschele, R.O. Rahn, and N. Forbes. 1979.
Quantitative mammalian cell mutagenesis and a preliminary
study of the mutagenic potential of metallic compounds. In:
Trace Elements in Health and Disease. N. Kharasch, ed. Raven
Press: New York. pp. 55-69.
Hyraan, J., Z. Leifer, and H.S. Rosenkranz, 1980. The E. coli
polAj ~ assay: a quantitative procedure for diffusible and
non-diffusible chemicals. Mutation Res. 74:107-111.
Kaneraatsu, 0., M. Hara, and T. Kada. 1980. Rec assay and
mutagenicity studies on metal compounds. Mutation Res.
77: 109-116.
Kimball, R.F. 1978. The relation of repair phenomena to mutation
induction in bacteria. Mutation Res. 55:85-120.
McCarroll, N.E., B.H. Keech, and C.E. Piper. 1980a. A comparative
evaluation of microsuspension microbial DNA repair systems.
Environ. Mutagen. 2:270. (abstr.)
McCarroll, N.E., C.E. Piper, G.M. Fukin, B.H. Keech, and G.
Gridley. 1979. A serial dilution multiwell suspension assay
for DNA damage in E. coli. Environ. Mutagen. 1:123. (abstr.)
McCarroll, N.E., C.E. Piper, and B.H. Keech. 1980b. Bacterial
microsuspension assays with benzene and other organic
solvents. Environ. Mutagen. 2:281. (abstr.)
Murray, M.M. 1979. Substrate specificity of uvr excision repair.
Environ. Mutagen. 1:347-352.
Nishioka, H. 1975. Mutagenic activities of metal compounds in
bacteria. Mutation Res. 31:185-189.
Radman, M. 1977. Inducible pathways in deoxyribonucleic acid
repair, mutagenesis and carcinogenesis. Biochem. Soc. Trans.
5:1194-1199.
Sunderman, F.W.,Jr. 1978. Carcinogenic effects of metals. Fed.
Proc. 37:40-46.
Tamaro, M., S. Venturini, C. Eftimiadi, and C. Monti-Bragadin.
1977. Interaction of platinum compounds with bacterial DNA.
Experientia 33:317-319.
-------
116
GUYLYN R. WARREN
Tindall, K.R. 1977. The mutagenicity of inorganic ions in
microbial systems. Unpublished master's thesis, Montana
State University. University Microfilm: Bozeraan, MT.
Tindall, K.R., G.R. Warren, and P.D. Skaar. 1978. Metal ion
effects in microbial screening systems. Mutation Res. 53:9091.
(abstr.)
Toraizawa, J.-i., and H. Ogawa. 1968. Breakage of DNA in rec*
and rec~ bacteria by disintegration of radiophosphorous atom
in DNA and possible cause of pleiotropic effects of recA
mutation. Cold Spring Harbor Syrap. Quant. Biol. 33:243-251.
Venturini, S., and C. Monti-Bragadin. 1978. R plasmid-mediated
enhancement of mutagenesis in strains of Escherichia co1i
deficient in known repair functions. Mutation Res. 50:1-8.
Warren, G., E. Abbott, P. Schultz, K. Bennett, and S. Rogers.
(in press a). Mutagenicity of a series of octahedral rhodium
111 compounds. Mutation Res.
Warren, G.R., S.J. Rogers, and E.H. Abbott. (in press b). The
genetic toxicology of substitutionally inert transition metal
complexes. In: Inorganic Chemistry in Biology and Medicine,
ACS Advances in Chemistry Series.
Warren, G., S. Rogers, S.G. Mevec, and M.D. Roach. 1979. Mutagen
screening in an isolated high lung cancer mortality area of
Montana. Montana Air Pollution Study, Air Quality Bureau,
Dept. of Health and Environmental Sciences, Helena, MT.
30 pp.
Warren, G., P. Schultz, D. Bancroft, K. Bennett, E.H. Abbott, and
S. Rogers. (MS). Mutagenicity of a series of octahedral
chromium III compounds.
Warren, G.R., P.D. Skaar, and S.J. Rogers. 1976. Genetic activity
of dithiocarbamate and thiocarbamoyl disulfide fungicides in
Saccharomyces cereviaiae, Salmonella typhimurium, and
Escherichia coli. Mutation Res. 38:391-392 (abstr.).
Witkin, E.M. 1976. Ultraviolet mutagenesis and inducible DNA
repair in Escherichia coli. Bacteriol. Rev. 40:869-907.
Yanofsky, C. 1963. Amino acid replacements associated with
mutation and recombination in the A gene and their
relationship to in vitro coding data. Cold Spring Harbor
Symp. Quant. Biol. 28:581-588.
-------
MICROTITER DMA REPAIR ASSAY FOR METALS
117
Zakour, R.A., L.A. Loeb, T.A. Kunkel, and R.M. Koplitz. 1979.
Metals, DNA polymerization, and genetic miscoding. In:
Trace Elements in Health and Disease. N. Kharasch, ed.
Raven Press: New York. pp. 135-153.
-------
Intentionally Blank Page
-------
A CULTURE SYSTEM FOR THE DIRECT EXPOSURE OF MAMMALIAN CELLS
TO AIRBORNE POLLUTANTS
Ronald E. Rasmussen and T. Timothy Crocker
Department of Community and Environmental Medicine
College of Medicine
University of California
Irvine, California
INTRODUCTION
Most airborne pollutants first enter the body through the
respiratory tract. In mammals and most other air-breathers,
mechanisms have evolved to deal with these pollutants, especially
those of a particulate nature. In recent times, however, air
pollutant gases have been introduced into the environment at higher
concentrations than before. Well-known examples are ozone (O3) and
oxides of nitrogen (NO2, NO). Other gaseous pollutants whose
effects are not so well known include short-lived, highly reactive
species produced photocheraically in the urban air mixture called
smog. Peroxyacetyl nitrate (PAN) is one example (Stephens, 1969).
There has not been sufficient time for animals to evolve
resistance to atmospheric oxidant gases. This lack of inherent
resistance can be shown dramatically in rats exposed to 0.8 ppm of
O3 in air. At rest, the rats can tolerate this concentration for
several hours, but if they are forced to exercise (i.e. on a
treadmill) they quickly succumb to lung edema and hemorrhage (R.F.
Phalen, University of California, Irvine, personal communication,
1980). Rats can be acclimatized to O3 by repeated exposure at
rest, so that they can resist the effects of O3 when forced to
exercise. Thus, substantial adaptational changes must have
occurred in the lung. Studies of various enzymes in lungs of rats
exposed to O3 and NO2 also have shown alterations (Chow et al.,
1976). It can be supposed that similar effects occur in the human
lung, since oxidant gas concentrations in some areas frequently
approach, and even exceed, 1 ppm.
119
-------
120
RONALD E. RASMUSSEN AND T. TIMOTHY CROCKER
The cell culture and exposure system described in this report
represents an in vitro approach to the study of the initial
interactions between oxidant gases and respiratory cells. The
goals of the present studies were to document some of the
cytotoxic, biochemical, and cytogenetic effects on mammalian cells
of exposure to low concent rat ions of O3 and NO2. Certain features
of this system allow the exposure of cell cultures to ambient
polluted atmospheres in a manner resembling exposure at the
surface of the respiratory epithelium. Living cells are separated
from the test atmosphere only by a thin layer of nutrient medium
held over the cells by capillary attraction. As this situation
allows close contact between cells and the ambient atmosphere, it
may be especially useful in studying the primary interactions
between cells and airborne pollutants.
MATERIALS AND METHODS
Exposure System
The following features are required in a cell system for
testing exposure to gaseous pollutants:
1) accurate generation and monitoring of the pollutants to be
s tudied,
2) maintenance of the cells in close contact with the test
atmosphere without allowing them to dry out,
3) exposure under biologically sterile conditions, and
4) provision for recovery of the cells for further culturing
and analysis.
The polluted atmospheres were generated by the measured
addition of NO2 or O3 to a stream of clean air. The initial
exposure system consisted of one control and one experimental
chamber. Results with this early system have been reported
elsewhere (Samuelsen et al., 1978; Crocker et al., 1979). The
system was later expanded to a total of six chambers that could
be arranged as desired for control or text exposures. Figure 1 is
a diagram of the system, showing interconnections for using three
different gases, either individually or in combination.
Atmospheres containing NO2 or sulfur dioxide (SO2) were generated
by diluting gas from stock cylinders in a stream of clean air.
Clean air was provided by treating the building air supply
sequentially with Purafil, activated charcoal, and filtration
through a 2-ym bacteriological filter. This treatment reduced the
level of NO2 to < 10 ppb and the O3 level to below detectable
-------
G

o
fns-Mi
^ dPhH- -r^r
(H-lfl-'EHj-o ;
o — 1 o J-
n
-EH To
lo
1°
lo
j
Fi gure
CAI IBRITlnN rj>s
WWirNT AIR
0. WIS
NO-.—»-N0
VENT
- '(.$£>®
' c#j- ® ^
—|-c5^a
-
f~jj IrdtamfTrp Q«essurc
1 M J -T" REGULATOR
tXl
FtOW
3 Kt^iernp
CHECK VAIVF
> THERMOCOUPLE
J r^flEHCAT COll
manual
On-OFF
/ \ WATER
/ TRAP
fOI C70NF
Gfstwroft
0 CRITICAL FLOW (P) niRAF!L
o*rjcr
®HUHimitK © CMAFTflAL
KCi mnuai
H5sJ HFTERING
rXZi i-«*
P*vJ smrhnm
18
CELL
CULTURE
DPOSURE
TH^HprR
ISJ
© FllltR
>
Ui
~<
H
ra
•n
O
so
O
i—^
fa
M
n
H
M
X
"d
o
c»
c
ffl
M
o
>
O
R
r
f
cn
The exposure system.
-------
122
RONALD E. RASMUSSEN AND T. TIMOTHY CROCKER
amounts (i.e., < 1 ppb). Because the cell culture medium was
bicarbonate-buffered, carbon dioxide (CO2) was added to the air
stream to raise its concentration to 5%.
Each exposure chamber was provided with flow controls for the
clean air and pollutant gases. The gases were mixed with the air
stream immediately before they entered the exposure chambers.
Ozone was provided by an individual generator for each exposure
chamber. The chambers were enclosed in 37°C incubators. The
chambers were rectangular with internal dimensions of approximately
10 x 10 x 35 cm and volumes of approximately 3.5 1. Gas flow was
along the long axis of the chamber and could be adjusted up to a
flow rate of 4 l/min. All tubing and fittings were either Teflon
or stainless steel. The exposure chambers were lined with
stainless steel foil; chambers of Lexan plastic are also available.
From the exposure chambers, the gas was carried via heated
sampling lines to a bank of solenoid valves controlled by a
microprocessor. The outflow from each chamber was sequentially
directed to the monitoring instruments, which (except for the N0X
monitor) were enclosed in a 37°C chamber to prevent moisture
condensation in the instruments. The instruments were selected to
provide measurements of specific gases without interference from
other gases. These instruments were a Dasibi model 1003 AH O3
monitor and a Beckman model 952A N0X analyzer. (A monitor for SO2
had yet to be installed.) Output from the monitors was fed to a
multichannel recorder.
Cell Culture Method
Cells of strain V-79 Chinese hamster lung fibroblasts
(obtained from Dr. E.H.Y. Chu, Ann Arbor, MI) were routinely grown
in Eagle's minimum essential medium supplemented with 10% fetal
bovine serum. All cell culture media were from Grand Island
Biological Co.
For exposure to gases, the cells were planted, either as
dispersed cells or as confluent cultures, on Millipore filters
(HAWP, 0.47-ym pore size) that had been thoroughly washed and
autoclaved. The cell-bearing filters were then assembled into the
specially designed holder shown in Figures 2 and 3. These holders
were fabricated from either Lexan plastic (General Electric Co.) or
stainless steel. The filters were positioned cell-side up and
sealed in place with the O-rings and threaded cap. Growth medium
was provided to the cells with a syringe pump connected to the
fitting in the base of the. holder. Medium perfusing through the
filter was drawn off through a small tube in the membrane holder,
shown in Figure 2 as a diagonal line at the right-hand side of the
membrane holder. With this arrangement the cells were kept moist
-------
A SYSTEM FOR DIRECT EXPOSURE OF MAMMALIAN CELLS
123
, 110 CAP
/

MEMBRANE HOLDER
MEMBRANE WITH CELLS
O-RING
Liliilli
1.25 BASE
6.3 CM
Figure 2. Diagram of a filter holder.
Figure 3. Photograph of a disassembled filter holder.
-------
124
RONALD E. RASMUSSEN AND T. TIMOTHY CROCKER
and continuously provided with fresh nutrient, while in nearly
direct contact with the ambient atmosphere. In clean air, nearly
100% viability could be maintained for at least several hours.
Experimental Procedure
Figure 4 shows a filter with cells being placed in a holder.
Before assembly, the base was filled with medium, and after
assembly, the top well of the holder was filled with medium to
protect the cells before placement in the exposure chambers. The
holders were then placed in the exposure chambers (four per
chamber). The bases of the holders were connected to the syringe
pump, while the tubes for withdrawal of medium from the upper wells
were connected to a peristaltic pump. The chamber door was sealed,
and gas flow was initiated. When the desired pollutant
concentration within the chamber had been reached and was stable,
the medium overlaying the cells was drawn off with the peristaltic
pump, and cell exposure was begun. The syringe pump was turned on
at the same time to slowly perfuse nutrient medium through the
filter. To conclude the exposure, the exposure chambers were
opened, the holders removed, and the top well immediately filled
with nutrient medium. Because the gas exposure levels used were of
the same order of magnitude as in the ambient atmosphere (0.15 ppm
NO2; 0.05 ppm O3), no special precautions were necessary.
Subsequent procedures depended on the nature of the
experiment. To estimate colony formation by surviving cells, the
filters, previously seeded with an appropriate number of dispersed
single cells, were transferred directly to petri dishes containing
nutrient medium and incubated for 7 to 10 days to permit colony
development. The cells could then be harvested from the filters
with trypsin and subcultured for further study, such as for
mutagenesis or chromosomal effects.
Measurement of Effects of 0^ and NO? on DNA Replication
The procedures for this study were adapted from those
described by Painter (1977). Cultures of V-79 cells were grown for
48 h with carbon-14-labeled thymidine (^C-TdR) at 0.01 uCi/ml (50
mCi/mmol). The cells were then planted on filters and exposed to
O3, NO2, or clean air at the concentrations and for the times
indicated in the Results section, below. After the gas exposure,
the filters were returned to nutrient medium; at intervals, sample
cultures were labeled for 10 min with tritiated thymidine (3H-TdR;
5 uCi/ml, 60 Ci/mmol). After this labeling, the filters were fixed
in ice-cold 5% trichloroacetic acid, washed several times with 70%
ethanol, and air dried. The radioactivity associated with the
filters was measured by scintillation counting, and the ratio of
-------
A SYSTEM FOR DIRECT EXPOSURE OF MAMMALIAN CELLS
125
Figure 4. The filter with cells being placed in the holder.
dpm:il+C dpm was calculated to give an index of the rate of DNA
synthesis at the time of labeling with ^H-TdR.
Measurement of Cytotoxic Effects of O3 and
Strain V-79 cells were seeded into filters as dispersed
single cells and allowed three to four hours for attachment. At
that time, the filters were placed in holders, which were then put
in the exposure chambers and exposed to O3 or NO2 at the
concentration and for the times indicated below. After exposure,
the filters were removed from the holders, transferred to petri
dishes containing nutrient medium, and incubated for 7 to 10 days
to permit colony development by survivors. Colonies were
visualized by staining the filters with hematoxylin.
Measurement of Direct Effects of NO2 on V-79 Cells
Growing cultures of V-79 cells were double-labeled with
amino acids (0.1 yCi/ml) and ^H-TdR (1.0 pCi/ml) for 24 h and then
seeded onto filters, which were then placed into Lexan filter
holders. After another 24 h, the cells were exposed for 2 h to NO2
-------
126
RONALD E. RASMUSSEN AND T. TIMOTHY CROCKER
at 5 ppm. The filters were then removed from the holders, washed
gently with 0.9% sodium chloride, and air dried. The radioactivity
remaining with the filters was determined by scintillation
count ing.
EXPERIMENTAL RESULTS
Cytotoxic Effects of NO2 and O3
It was recognized early in these studies that V-79 cells were
very sensitive to the effects of NO2 when the gas was in direct
contact with the cells (Samuelsen et al., 1978). Exposure of cells
to 0.15 ppm of NO2 inactivated their colony-forming ability in a
dose-dependent manner; after 6 h, fewer than 90% of the originally
exposed cells could form macroscopic colonies during subsequent
incubation of the filters in immersed culture.
Ozone was somewhat more effective than NO2 in activating
colony formation by V-79 cells. Figure 5 shows the results of a
series of studies with O3 at 0.05 ppm in air. Colony-forming
ability was always compared with that of cells exposed to clean air
in separate chambers. In the clean-air controls, very little loss
of colony-forming ability was seen (less than 10X) for exposures as
long as 6 h.
Effect of NO2 and O3 on DNA Replication
In mammalian cells, many chemical and physical mutagens induce
DNA damage that interferes with DNA replication (Painter, 1978).
This can be shown by measuring, at intervals, the rate of DNA
synthesis in cell cultures after a single treatment with the test
agent. With most mutagens, the rate of DNA synthesis declines with
time after treatment. Chemicals that inhibit DNA synthesis, but do
not damage DNA, do not produce such an effect, and DNA synthesis
returns to its normal rate when the inhibitors are removed.
To test for the effect of NO2 and O3 on DNA replication, V-79
cells were labeled with 14C-TdR as described above and exposed to
either O3 (0.03 ppm, 1 h) , NO? (0.15 ppm, 1 h), or 254-nm
ultraviolet (UV) light (5 J/m^). Immediately after treatment and
one and two hours later, sample cultures from each group were
labeled for 10 rain with ^H-TdR and immediately fixed with cold 5%
trichloroacetic acid. The results are shown in Figure 6. The data
are presented as percentages of the rates of DNA synthesis in
appropriate controls (cultures exposed to clean air as controls for
NO2 and O3 and cultures sham-exposed to UV light).
-------
A SYSTEM FOR DIRECT EXPOSURE OF MAMMALIAN CELLS
127
z
o
—I
0
(J
5
1
o
o
z
>
>
oc
D

I
A
\
^ t
\
\
j i i i i i \
2 3 4 5 6
HOURS OF EXPOSURE
Figure 5. Inactivation of V-79 cells by 0.05 ppm ozone
Exposure to NO2 produced a slight decrease in the rate of DNA
synthesis, but the rate returned to normal when the cultures were
placed in fresh culture medium. Both O3 and UV light also produced
a slight initial decrease in the rate of DNA synthesis: but in
contrast to the results with NC^-exposed cells, the rate of DNA
synthesis continued to fall during subsequent incubation in fresh
medium. This suggests that O3 may have damaged the cellular DNA.
Mechanism of Action of O3 and NO? on Directly-exposed Cells
The oxidizing effects of NO2 and O3 on membrane lipids are
well known (Goldstein and Balchum, 1967: Thomas et al., 1968): the
toxicity seen in the present exposure system may be due in part to
such oxidations of the cell surface membranes. Exposure of thin
films of peanut oil to 0.15 ppm of NO2 in the present system
produced detectable peroxidation (Goldstein et al. , 1969).
-------
128
RONALD E. RASMUSSEN AND T. TIMOTHY CROCKER
m C
T CO
h-
Z 2
> o
to cj
< LJ-
2 o
o H
< Q.
100 _
OZONE
1 2
HOURS AFTER EXPOSURE
Figure 6. The effect of NO2, ozone, and UV-light on the rate of
DNA synthesis in V-79 cells.
Microscopic examination of cells after various periods of
exposure to NO2 tended to support this hypothesis. The first
observable changes were that the cells seemed to become more
rounded and firmly attached to the filter substrate. With longer
times of exposure and higher NO2 levels, the nuclei were distorted.
Finally, 5 ppm of NO2 produced loss of cells from the filters. A
possible trivial explanation could be that NO2 somehow affected the
Millipore membrane so that it could no longer support cell growth
and attachment. This was shown not to be the case through studies
in which filters were exposed to 5 ppm of NO2 for 6 h. This
treatment did not alter cell attachment or the ability of the
filters to support cell growth.
To obtain some quantitative information on cell destruction
by NO2, cultures of V-79 cells were labeled simultaneously with
lLtC-amino acids and ^H-TdR as described above. The doubly-labeled
cells were then planted on filters so as to provide a nearly
confluent cell sheet at the time of exposure to NO2.
-------
A SYSTEM FOR DIRECT EXPOSURE OF MAMMALIAN CELLS
129
Table 1 gives the results of several experiments in which
exposure to NO2 at 5 ppm for 2 h produced an average loss of cells
from the filters of about 50%. Both and were lost from the
filters. In every case, however, proportionately more tritium was
lost from the N02-exposed filters; the ratio of remaining on
the filters was always lower for the N02~exposed filters. Exposure
to clean air did not produce any loss of label, as shown in
experiments 18, 21, and 25.
Table 1. Results of Mammalian Cell Exposure to Nitrogen Dioxide
Experiment 3H DPM ll*C DPM Ratio
Number Exposure x 10-5 ± 1 SD x 10~3 ± 1 SD ^H:^C
15

NO 2
1.72
+
0.14
1.18
+
0.05
14.6

Air
7.58
+
0.16
3.73
+
0.10
20.3
17

NO 2
0.982
+
0.074
3.78
+
0.27
26.0

Air
1.66
+
0.19
5.88
+
0.68
28.2
18

NO 2
0.850
+
0.20
2.75
+
0.57
30.9

Air
2.50
+
0.33
5.34
+
0.82
46.8

Not
Exposed
2.31
+
0.40
5.07
+
0.91
45.6
21

NO 2
4.60
+
0.82
25.00
+
0.47
18.4

Air
11.2
+
0.79
44.50
+
0.21
25 .2

Not
Exposed
10.6
+
0.40
44.10
+
0.20
24.0
25

no2
3.62
+
0.31
14.0
+
1.3
25 .9

Air
5.33
+
0.26
17.7
+
1.3
30.1

Not
Exposed
5.55
+
0.40
19.1
±
1.4
29.1
DISCUSSION
This research project was begun with the goal of developing an
in vitro analog of the respiratory epithelium. Cell culture
methods were to be established for maintaining living cells in
nearly direct contact with test atmospheres, while at the same time
preventing the cells from drying out. In addition, these methods
would allow recovery of the cells for further study. Previous
systems for exposing cells to gaseous materials have relied on
1) solution of the pollutant in the medium bathing the cells or
-------
130 RONALD E. RASMUSSEN AND T. TIMOTHY CROCKER
2) periodic short exposures of the cell layer followed by immersion
of the cells in liquid media.
A method for exposing cells for relatively long periods to
almost any atmosphere has been developed: however, cells exposed in
this system were extremely sensitive to very low levels of
pollutant gases, which may not realistically indicate the effects
of pollutants on experimental animals and humans. For example, O3
and NO2 concentrations of less than 1 ppm caused rapid destruction
of cells in this test system. This great sensitivity may be
related to the relatively large area of the cell surface that was
exposed to the gases, since the cells were spread thinly on the
membrane surface. In contrast, the cells of the respiratory tract
in vivo are closely packed, as in the columnar epithelial areas,
and only a small percentage of the total cell surface area is
exposed to airflow in the airways. Also, in the normal lung the
cells are covered by a mucous layer that shields them from exposure
to oxidants and particulates. Simulation of this mucous layer in
the present exposure system was not attempted.
The exposure system is not limited to the use of the cell
cultures on membrane filters. The exposure chambers will
accommodate conventional organ culture dishes, which are being used
in current studies of the role of oxidant gases in neoplastic
transformation. It is known that O3 and NO2 damage lung cells and
induce hyperplasia (Hackett, 1979; Dungworth et al., 1975), but the
role (if any) of this response in neoplasia is not clear. Using
the present exposure system, we hope to provide evidence on this
subject.
ACKNOWLEDGMENTS
This project has received support from the U.S. Environmental
Protection Agency (contract no. 68-02-1204), the U.S. Air Force
Office of Scientific Research (contract no. 77-3343), and the U.S.
National Institute for Environmental Health Sciences (grant no.
ES-01835). We thank Dr. G. Scott Samuelsen and John T. Taylor, of
the School of Engineering of the University of California, Irvine,
for invaluable aid and counsel. We also thank M.E. Witte and D.L.
Swedberg for able technical assistance.
REFERENCES
Chow, C.K., M.Z. Hussain, C.E. Cross, D.L. Dungworth, and M.G.
Mustafa. 1976. Effect of low levels of ozone on rat lungs.
I. Biochemical responses during recovery and reexposure.
Exper. Molec. Pathol. 25:182-188.
-------
A SYSTEM FOR DIRECT EXPOSURE OF MAMMALIAN CELLS
131
Crocker, T.T., R.E. Rasraussen, M.E, Witte, G.S. Samuelsen, and J.T.
Taylor. 1979. A unique culture system for in vitro exposure
of respiratory cells to pollutant gases. In: Nitrogenous Air
Pollutants. Chemical and Biological Implications. D.
Grosjean, ed. Ann Arbor Science Publishers: Ann Arbor, MI.
pp. 179-188.
Dungworth, D.L., C.E. Cross, J.R. Gillespie, and C.G. Plopper.
1975. The effects of ozone on animals. In: Ozone Chemistry
and Technology. J.S. Murphy and J.R. Orr, eds. The Franklin
Institute Press: Philadelphia. pp. 29-54.
Goldstein, B.D. and O.J. Balchum. 1967. Effect of ozone on lipid
peroxidation in the red blood cell. Proc. Soc. Exper. Biol.
Med. 126:356-358.
Goldstein, B.D., C. Lodi, C. Collinson, and O.J. Balchum. 1969.
Ozone and lipid peroxidation. Arch. Environ. Hlth.
18:631-635.
Hackett, N.A. 1979. Proliferation of lung and airway cells
induced by nitrogen dioxide. J. Toxicol. Environ. Hlth.
5:917-928.
Painter, R.B. 1977. Rapid test to detect agents that damage
human DNA. Nature 265:650-651.
Painter, R.B. 1978. DNA synthesis inhibition in HeLa cells as a
simple test for agents that damage human DNA. J. Environ.
Pathol. Toxicol. 2:65-78.
Samuelsen, G.S., R.E. Rasmussen, B.K. Nair, and T.T. Crocker.
1978. Novel culture and exposure system for measurement of
effects of airborne pollutants on mammalian cells. Environ.
Sci. Technol. 12:426-430.
Stephens, E.R. 1969. The formation, reactions, and properties of
peroxy acyl nitrates (PANs) in photochemical air pollution.
In: Advances in Environmental Sciences and Technology, Vol.
1. J.N. Pitts and R.L. Metcalf, eds. Wiley Interscience:
New York. pp. 119-146.
Thomas, H.V., P.K. Mueller, and R.L. Lyman. 1969.
Lipoperoxidation of lung lipids in rats exposed to nitrogen
dioxide. Science 159:532-534.
-------
Intentionally Blank Page
-------
SESSION 2
DRINKING WATER AND
AQUEOUS EFFLUENTS
-------
Intentionally Blank Page
-------
IS DRINKING WATER A SIGNIFICANT SOURCE OF HUMAN EXPOSURE TO
CHEMICAL CARCINOGENS AND MUTAGENS?
Richard J. Bull
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
Drinking water has long been suspected as a medium through
which chemical carcinogens and mutagens reach man. The first
inquiries into this possibility were made by Heuper and Ruchhoft
(1954), who tested carbon-chloroform extracts from the following
samples: a gravity oil separator effluent from a petroleum
refinery; raw water from a canal polluted by wastes from a
petroleum refinery; raw water from Nitro, WV (the Kanawha River);
and finished water from Cincinnati, OH (the Ohio River). These
samples were applied topically to black male C57 mice in whole-
animal carcinogenicity tests lasting one year. Samples from the
refinery and the ship canal were carcinogenic; the raw water from
Nitro and the finished water from Cincinnati gave negative results.
While this study did not give direct evidence of carcinogens in
drinking waters, it brought attention to the presence of such
chemicals in surface waters used as sources for drinking water.
The development and general application of very sophisticated
analytical tools in the past ten years have greatly altered our
views on pollution of drinking waters. Coupled gas chromatographic
and mass spectral analysis of drinking water has rapidly increased
the numbers of chemicals known to occur in drinking water. Where
previously the identification of a single chemical might take
months or years, the general availability of mass spectrometry and
the development of associated computer systems now allow almost
instantaneous identification of hundreds of chemicals (if reference
fragmentation patterns are available). To date, more than 1152
different chemicals have been identified in extracts of U.S.
drinking waters (Melton, 1979); several of these chemicals are
known to produce tumors in humans or experimental animals.
135
-------
136
RICHARD J. BULL
The most significant finding in recent years is that the
organic chemicals usually found at the highest concentrations in
drinking waters arise not from industrial pollution, but from
disinfection of drinking water with chlorine (Rook, 1974; Bellar et
al., 1974). The first such chemicals identified were the
trihaloraethanes (THMs), primarily bromo- and chloro-substituted
methanes, with occasional traces of iodomethanes. The significance
of these observations was increased by the National Cancer
Institute Carcinogenesis Bioassay Program's finding that chloroform
is carcinogenic in rats and mice. Results from animal studies were
subsequently supported by a number of studies indicating a
correlation between drinking water chlorination and
gastrointestinal- and urinary-tract cancer mortality (see Wilkins
et al. , 1979, for a review). In addition, several chemicals found
in drinking water have been identified as mutagens in the Ames test
(Simmons et al., 1977).
As the above findings became known, workers began to
investigate the biological properties of complex mixtures of
chemicals in drinking water. A number of laboratories have found
organic concentrates of drinking water to be mutagenic in the Ames
test (Loper et al., 1978; Glatz et al., 1978). These studies have
been followed up with demonstrations that such concentrates can
transform BALB/3T3 cells (Lang et al., in press). When chemicals
from water are sufficiently concentrated and fractionated, it is
doubtful that any surface water supply will be found to be without
any mutagenic activity.
Several investigators have shown that the THMs represent only
a fraction of the products of chlorination. Many substances
produced by chlorination of huraic and fulvic acids and isolated
from water remain to be identified; many of these are not available
from commercial sources. Data recently reported by Symons (in
press), of the Drinking Water Research Division of the Municipal
Environmental Research Laboratory, in Cincinnati, indicate that
more than 50% of the organic chlorine produced through chlorination
of drinking water can be in products other than THMs. Although
chloramination, an alternate disinfection method, suppresses THM
formation, it does not suppress the production of non-THM organic
chlorine to the same extent.
Mouse skin initiation/promotion studies with various
disinfectants indicate that these non-THM products cannot readily
be dismissed. In one such experiment (see Table 1), tumor-
initiating activity was found only for drinking water disinfected
with chlorine, chloramines, or ozone (Bull, 1980). Since these
concentrates were prepared by reverse osmosis, they did not contain
THMs .
-------
CHEMICAL CARCINOGENS AND MUTAGENS IN DRINKING WATER 137
Table 1. Tumor Initiation by Reaction Products of Various
Disinfectants3 After 20 Weeks of Promotion with PMA*5
Concentrat ion
No. Animals
Total
% Animals
Samplec
Factorc
with Tumors
Tumors
with Tumors
Non-disinfected
102
0/25
0
0
Chlorine
106
4/25
5
20
Chloramine
142
5/25
8
32
Chlorine dioxide
168
0/25
0
0
Ozone
186
7/25
9
36
Saline
-
1/25
1
4
7,12-dimethyl-

benzo(a)anthracene
~
16/25
35
140
aSubstrate was settled, coagulated, and filtered Ohio River water.
The total dose was 1.5 ml (6 x 0.25 ml) given subcutaneously at
the concentration factor indicated.
kpMA =¦ Phorbol myristate acetate applied at a dose of 2.5 ug, three
times weekly.
cAfter treatment, water was subjected to reverse osmosis with
cellulose acetate to the level indicated by the concentration
factor (initial volume/final volume).
With this brief background, we can summarize the major
categories of hazardous chemicals in drinking water:
1) Trace quantities of a wide variety of synthetic
organic chemicals, usually at < 1 yg/1 in surface
waters.
2) Sporadic occurrence of high concentrations of
individual industrial chemicals, typically involving
groundwater contamination with bulk solvents.
3) Natural (or background) organic chemicals, such as
humic and fulvic acids.
4) Products of the reaction of disinfectants with
background chemicals (including THMs, the major
class of organic compounds in drinking water).
5) Chemicals leached from distribution systems, such as
lead, asbestos, and polycyclic aromatic hydrocarbons.
-------
138
RICHARD J. BULL
6) Water-treatment chemicals (including polyelectrolytes,
coagulants, and corrosion-control chemicals).
The third and fourth of these items are peculiar to and ubiquitous
in drinking water; they pose the greatest problems in assessing the
risks associated with drinking water.
The overwhelming numbers of chemicals that appear in drinking
waters, their low individual concentrations, the complications
associated with drinking-water disinfection, and the need to
demonstrate actual reduction in carcinogenic risk through various
treatment options dictate a bioassay approach to drinking water
risk assessment. The presence of several hundred to several
thousand compounds in a drinking water at a fraction of a microgram
per liter is the rule rather than the exception. Individually,
most of these chemicals might contribute little to human disease
(with the possible exception of the THMs). However, they add up
to concentrations of from one to several milligrams of total
organic carbon per liter of drinking water and could account for
some of the hazards suggested by epidemiological data (Wilkins et
al., 1979).
REFERENCES
Bellar, T.A., J.J. Lichtenberg, and A.D. Kroner. 1974. The
occurrence of organohalides in chlorinated drinking water.
J. Amer. Water Works Assoc. 66:703-706.
Bull, R.J. 1980. Health effects of alternate disinfectants and
their reaction products. J. Amer. Water Works Assoc.
72:299-303.
Glatz, B.A., C.D. Chriswell, M.D. Arguello, H.J. Svec, J.S. Fritz,
S.M. Grimm, and M.A. Thomson. 1978. Examination of drinking
water for mutagenic activity. J. Amer. Water Works Assoc.
70:465-468.
Heuper, W.C. and C.C. Ruchhoft. 1954. Carcinogenic studies on
absorbates of industrially polluted raw and finished water
supplies. Arch. Ind. Hyg. Occup. Med. 9:488-495.
Lang, D.R., H. Kurzepa, M.S. Cole, and J.C. Loper. (in press).
Malignant transformation of BALB/3T3 cells by residue organic
mixtures from drinking water. J. Environ. Pathol. Toxicol.
Loper, J.C., D.R. Lang, R.S. Schoeny, B.B. Richmond, P.M.
Gallagher, and C.C. Smith. 1978. Residue organic mixtures
from drinking water show in vitro mutagenic and transforming
activity. J. Toxicol. Environ. Hlth. 4:919-938.
-------
CHEMICAL CARCINOGENS AND MUTAGENS IN DRINKING WATER 139
Melton, R.G. 1979. GC/MS Analysis of Organics in Drinking Water
Concentrates and Advanced Waste Treatment Concentrates.
Preliminary Report—Combined Results on Five Drinking Water
Supplies. Prepared under contract no. 68-03-2458 for the U.S.
Environmental Protection Agency, Cincinnati, OH.
Rook, J.J. 1974. Formation of haloforms during chlorination of
natural waters. J. Water Treat. Exam. 22:234-243,
Simmons, V.F., K. Kauhanen, and R.G. Tardiff. 1977. Mutagenic
activity of chemicals identified in drinking water. Presented
at the Second International Conference on Environmental
Mutagens, Edinburgh,
Symons, J.M. (in press). Utilization of various treatment unit
processes and treatment modifications for trihaloraethane
control. In: Proceedings, Control of Organic Chemical
Contaminants in Drinking Water. U.S. Environmental Protection
Agency: Washington, DC.
Wilkins, J.R. Ill, N.A. Reiches, and C.W. Kruse. 1979. Organic
chemical contaminants in drinking water and cancer. J.
Epidemiol. 110:420-448.
-------
Intentionally Blank Page
-------
ALTERNATIVE STRATEGIES AND METHODS FOR CONCENTRATING
CHEMICALS FROM WATER
Frederick C. Kopfler
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
INTRODUCTION
The concentration of organic matter in water can range from
several hundred micrograms per liter in groundwater to many
milligrams per liter in industrial or sewage effluents. Generally,
the biologically active materials will be present in concentrations
too low to be detected by testing the small amount of aqueous
sample that can be incorporated directly into a biological test
system. The organic matter in drinking water and wastewater is a
complex mixture and defies complete characterization by current
technology. Consequently, these materials cannot be purchased or
synthesized for biological testing, but must be obtained from the
water to be evaluated. A paradoxical situation results, since the
evaluation of methods for concentrating an organic substance
depends on the existence of reliable analytical methods for the
quantitative analysis of the substance. Therefore most of the data
available concerning organic concentration techniques are based on
performance with a few specific compounds or on general parameters
such as total organic carbon (TOC) or total organic halogen.
Methods for producing samples of organics from water for
biological testing can be divided into two types: concentration
and isolation. In the former, water is removed, leaving the
dissolved substances behind; in the latter, the organic substances
are removed from the water. Combinations of methods have also been
used to prepare samples for biological testing. This paper does
not deal with the theoretical aspects of the methods, but rather
reviews the advantages and disadvantages of representative examples
of both approaches. This paper is intended as a guide to the
141
-------
142
FREDERICK C. KOPFLER
interpretation of data obtained through biological testing of
organic concentrates produced through methods currently in use.
The choice of approach to preparing samples for biological
testing is determined by the biological test system to be used.
The final sample must provide a sufficient quantity of material in
a volume of solvent that is compatible with and can be accommodated
by the test system. Generally, more sample is required for
biological testing than for chemical analysis. Most of the methods
included in this paper were developed and evaluated for preparing
samples for chemical analysis and, in most cases, will have to be
scaled up to prepare samples for anything other than small-scale in
vitro tests .
CONCENTRATION TECHNIQUES
Some volume-reduction methods for preparing organic
concentrates for biological testing include freeze concentration,
freeze-drying, vacuum evaporation, and membrane processes (reverse
osmosis and ultrafiltration). If aqueous concentrates are
required, the degree to which the samples can be reduced in volume
is Limited by the concentration of inorganic substances in the
sample and the aqueous solubility of the organic substances. If
the final concentration required for testing is greater than these
parameters will allow, then the volume-reduction method can be used
as a first step in a combination method, as will be described
1ater .
Freeze concentration is the process whereby a water sample is
frozen into a shell of pure ice, leaving in the center the unfrozen
water containing the dissolved substances. Shapiro (1961) proposed
using this method for concentrating environmental water samples; it
was subsequently evaluated by Baker (1969). The method allows
effective recovery of all components tested, including volatile and
ionized organic species. The components are equally but not
totally recovered.
Freeze concentration works well in distilled water solutions,
but recovery of organic solutes decreases with increasing salt
content, due to alterations of the forming ice surface that result
in the incorporation of solute-rich liquid into the ice (Baker,
1970). Baker demonstrated that initial concentrations of up to 310
rag/l of total dissolved solids have little effect on the recovery
of the test substance m-cresol, with approximately 80% being
recovered in the liquid after a 20-fold reduction in volume (Baker,
1969) . Baker used a rotary evaporator with the flask immersed in a
freezing bath to concentrate samples of a few hundred milliliters
spiked with milligram quantities of test compounds. At one time, a
freeze concentrator capable of concentrating 5-1 samples up to
-------
ALTERNATIVE STRATEGIES FOR CONCENTRATING CHEMICALS
143
30-fold was available, but it is no longer manufactured. If large
samples are to be freeze concentrated, the equipment will have to
be custom-made and evaluated.
Freeze-drying is the process of removing water vapor directly
from the frozen sample by sublimation under vacuum. Large-scale
equipment is available, but it is cumbersome and expensive, and it
requires large amounts of energy. Freeze-drying is a slow process.
In our laboratory, we use an apparatus that allows up to 40 1 of
water to be processed at one time; about 72 h are required to
remove all of the water. The residue remaining is composed largely
of inorganic salts and is hygroscopic, so that a solvent is
required to recover it from the large stainless steel pans used as
sample containers.
Vacuum evaporation (or vacuum distillation) is the boiling of
the aqueous sample at reduced pressure at or near ambient
temperature. This method has been used to concentrate water
samples for chemical analysis (Jolley et al., 1975) and for
biological testing (Johnston and Herron, 1979). The required
apparatus can be assembled from commercially available components.
Freeze-drying and vacuum evaporation can achieve high degrees
of concentration with little contamination, and only substances
volatile at the temperature and pressure used will be lost. For
most environmental water samples, large percentages of TOC can be
recovered. The major drawback in both cases is the difficulty of
recovering the organic substances from inorganic precipitates.
This problem will be addressed at the end of this section.
In reverse osmosis, water is preferentially forced through a
membrane by applying an external pressure that exceeds the osmotic
pressure across the membrane. Reverse osmosis systems are
commercially available in many sizes and generally contain either
cellulose acetate or polyamide (nylon) membranes. The polyamide
membranes are more stable at extremes of pH but are highly
sensitive to chlorine. They cannot, therefore, be used to process
chlorinated waters unless the residual is chemically reduced. One
disadvantage of the commercial systems is that they contain plastic
components and synthetic adhesives that could adsorb sample
components or release contaminants into the concentrate.
A schematic of a reverse osmosis system used to concentrate
water samples is shown in Figure 1. The water to be concentrated
is circulated past the membrane under pressure; a fraction of the
volume is removed, and the remainder is recirculated to the feed
tank. Cellulose acetate membranes commonly used are rated to
reject organic substances of molecular weight greater than 200 and
90 to 97% of the inorganic ions. Generally, this degree of
inorganic rejection is achieved. However, while molecular weight
-------
144
FREDERICK C. KOPFLER
Reject or Concentrate Stream
/ ¦ '
Sample
f

TTT'
Pressure
Regulator
^"-Permeate
t
Pump Pressure Vessel
Containing Membrance
Figure 1. Schematic of reverse osmosis concentrator-.
influences Che rejection of organics by the membrane, polar or
ionized species are rejected more effectively than hydrophobic
nonionized substances. In contrast to the other volume-reduction
methods described, reverse osmotic concentration vor.ks through ah
exponential decay process in the recirculating system. Figure 2
illustrates the percent recoveries obtained for compounds with
rejections of 70, 80, and 90% at volume reductions between 10- and
1000-fold. Table 1 shows the actual concentration, factors obtained
when sample volume is reduced 10-, 100-, and 1000-fold. While
reverse osmosis does not retain all of the substances equally, it
does retain much of the organic carbon in the drinking water
samples.
Each of these concentration methods has its own advantages and
disadvantages, but the one disadvantage shared bv all of these
methods is that inorganic species are concentrated along with the
organic substances of interest. The degree to which samples can be
concentrated by these methods before precipitation of inorganics
occurs varies with the types and concentrations of inorganics
originally present, but is generally 20- to 50-fold. If the
bioassay system is sensitive enough and will tolerate aqueous
samples, and the inorganic salts do not interfere, these
concentrates can be tested directly. Freeze concentration and
reverse osmosis produce only such aqueous concentrates; if more
concentrated samples are required, the samples must be processed
-------
ALTERNATIVE STRATEGIES FOR CONCENTRATING CHEMICALS
145
100.

90-
>
70-
3)
>
60*
0
u
50C
<0
oc
40-

c
3?
30-
o

a.
20-
10
O Compounds with 90% rejection
A Compounds with 80% rejection
~ Compounds with 70% rejection
100
(-80
50
b40
0.1 0.02 0.01 0.0025
Fraction of Original Volume Remaining
0.001
ID
>
O
o
V
oc
c
a
25 u
0)
Q.
¦10
Figure 2. Recovery of compounds by reverse osmosis at various
stages of concentration.
Table 1. Effect of Percent Rejection on the Actual Concentration
of Constituents Obtained by Reverse Osmosis
Actual Concentration Factor
Fraction of Original Volume Remaining
Compound
Percent
Rejection 0.1 0.01 0.001
100
90
80
70
10
8
5.2
5
100
60
40
25
1000
500
250
125
-------
146
FREDERICK C. KOPFLER
further by other means. Freeze-drying and vacuum evaporation can
be used to produce either aqueous concentrates or a solid residue.
If inorganic constituents precipitate, recovery of the organics
from the residue is difficult.
Pitt and Scott C1973) concentrated effluent from sewage-
treatment plants by vacuum evaporation followed by freeze-drying.
Many of the salts were carbonates and could be solubilized in
acetic acid; samples were concentrated 2000-fold, with more than
95% recovery of TOC. River and lake waters were concentrated by
the same methods. Because the starting levels of organic carbon
were much lower than in sewage, a concentration factor of 101* was
required. The inorganics were not predominantly carbonates and
couLd not be redissolved; consequently, about 75% of the organic
carbon originally present could not be recovered. Extraction with
methanol gave only slightly better yields. Recovery studies were
conducted with 13 compounds representing the classes of compounds
expected in natural water samples. Table 2 shows the percentages
of these compounds recovered from a spiked river water sample.
These workers improved the recovery of organic carbon by passing
the water sample through a weak cation exchange column prior to the
evaporative steps. Losses of organic carbon from four samples thus
treated still ranged from < 10 to 57%.
Crathorne et al. (1979) studied the recovery of
5-chlorouracil, 5-chlorouridine, 4-chlororesorcinol, and
5-chlorosalicylic acid from residues obtained from spiked samples
of finished drinking water. The 14 solvents listed in Table 3 were
investigated for efficiency in recovering the compounds from the
residue. Only water and methanol yielded greater than 90%
recovery, but the mean recovery with methanol was about 60% (B.
Crathorne, Water Research Centre, Medmenham Laboratory, Manlaw,
Buckinghamshire, England, personal communication, 1980).
Volume-reduction methods are most successful if precipitation
is prevented during concentration. This has been accomplished by
Pitt as described above, by Kopfler et al. (1977), using Donnan
dialysis to exchange sodium ions for calcium ions during reverse
osmosis, and by Johnston and Herron (1979), through a method to be
described below.
ISOLATION TECHNIQUES
Isolation methods remove organics from water by concentrating
them in organic solvents. One method is to use immiscible organic
solvents to extract the water; another is to adsorb the organics
onto a solid medium and elute them with an organic solvent. Many
variations of these methods have been investigated, because they
have been used extensively to prepare samples for chemical analysis
-------
ALTERNATIVE STRATEGIES FOR CONCENTRATING CHEMICALS
147
Table 2. Recovery of Individual Compounds in Aqueous
Solution after Evaporation-Lyophi1izationa

Ant icipated
Final
Percent
Compound
Coneentration
(yg/ml)
Recovered
Sucrose
10

>50b
Urac il
10

80
Guanosine
10

5
Xanthine
8

10
Uric Acid
25

<5
Hippuric acid
20

5
p-Cresol
13

0
p-Hydroxyphenylacetic acid
15

5
Syringic acid
15

15
p-Hydroxybenzoic acid
18

10
o-Chlorobenzoic acid
40

20
p-Chlorobenzoic acid
40

0
o-Chlorophenol
33

0
aFrom Pitt and Scott ( 1973) .
^Quantification uncertain due to interference of naturally
occurring compound.
Table 3. Solvents Tested for Extracting Concentration Residues3
Solvent
Concentration Residue
Methanol
Diethyl ether
Water
Dimethylsulfoxide
Acetone
Dioxane
Acetonitrile
Ethyl acetate
Carbon tetrachloride
Isopropanol
Chloro form
Pyridine
Dichloromethane
Tetrahydrofuran
aB. Crathorne, Water Research Center, Medraenham Laboratory, Manlaw,
Buckinghamshire, England, personal communication, 1980.
-------
148
FREDERICK C. KOPFLER
Direct liquid-liquid extraction is suitable for the recovery
of organics from water samples of several liters. Large samples,
however, require continuous extractions using large volumes of
solvent or refluxing a smaller volume of solvent to provide pure
extractant.
As with any process employing solvents, impurities can be
concentrated along with sample components. Most highly purified
organic solvents contain preservatives—often an antioxidant—that
can react to add organic contaminants to samples. For example,
cyclohexene is an impurity present in the best grades of methylene
chloride. When methylene chloride is used to extract samples that
contain a chlorine residual (as do most drinking waters and
wastewaters), the cyclohexene produces mono-, tri-, and tetra-
chlorocyclohexenes and cyclohexanes by reacting with the chlorine
residual (Logsdon et al., 1977). Also, peroxides may contaminate
extracts prepared with ether or may react with sample components to
produce new substances not originally present in the sample or
solvent. Another critical area for investigation is whether
changes in the organic residues could occur in concentrates oE
organics during storage between preparation and analysis or
biological testing.
The adsorption-elution methods require the least-complex
apparatus for isolating organics from water. The water sample is
passed through a column of the solid adsorbant, and the organics
are subsequently eluted with a smaller volume of suitable solvent.
The most common adsorption-elution methods used are activated
carbon, ion exchange resins, macroreticular resins, and reverse
phase high-performance liquid chromatography columns.
Activated carbon has been used to remove organics from aqueous
solutions (Buelow et al., 1973). However, recovery from carbon is
not as good as can be obtained with other media (Chriswell et al.,
1977). Formerly, organics were recovered from the carbon through
air-drying followed by prolonged Soxhlet extraction with chloroform
and ethanol. This method has been abandoned because of the air-
drying step and the continued boiling of the extract in the
extraction apparatus. A procedure using supercritical liquid
carbon dioxide as a solvent has recently been developed and is
being evaluated as a means of regenerating granular activated
carbon used in water treatment (Modell et al., 1978). The liquid
carbon dioxide is miscible with water, allowing the drying step to
be eliminated. The extraction takes place at about 30°C in a
closed system, eliminating the high temperature encountered with
organic solvents. The liquid carbon dioxide has also been shown to
be a good solvent for several classes of organics. This method has
not yet been evaluated for producing extracts for bioassay.
-------
ALTERNATIVE STRATEGIES FOR CONCENTRATING CHEMICALS
149
The XAD resins produced by Rhom and Haas have been used
extensively Co recover organic substances from water. Two types
are available: a styrene-divinyl benzene copolymer and a
methacrylate-based copolymer (Dressier, 1979; Gustafson and Paleos,
1971). Resins of both types are available in various pore sizes,
giving different unit surface areas. These resins are produced for
industrial use and contain many lower-molecular-weight
contaminants. The resins must be prepared for laboratory use by
serial extraction in a Soxhlet extractor with methanol, diethyl
ether, and acetonitrile (Junk et a!., 1974). Before use, the
resins should be evaluated to insure that the extraction has, in
fact, removed contaminants from the resin to the degree required.
If resin beads are allowed to dry out, they can crack, exposing
newly contaminated surfaces; thus, clean resin should be stored
under methanol until used.
XAD resins have been used to isolate synthetic organic
chemical contaminants from water for chemical analysis. They have
a great affinity for hydrophobic substances and retain virtually
all of these materials, even when the aqueous sample is passed
through the resin column at high flow rates (Junk et al., 1974).
Much of the organic matter in water is hydrophilic, however;
Thurman and his co-workers (1978) have demonstrated that the
capacity of the resins for these compounds is not great and that
flow rates during the adsorption step must be in the range of 15 to
20 bed voluraes/h for good recovery. Organic acids and bases are
effectively adsorbed from water only after ionization has been
suppressed by pH adjustment. Lowering the pH to protonate the
organic acids generally presents no problem, but attempting to
recover organic bases at a high pH can result in clogging of the
column by inorganic hydroxides after only a small amount of water
has passed through the column. It has been estimated that about
50% of the TOC in the average water sample can be concentrated onto
a column of XAD-8 resin (Malcolm et al., 1977).
Ionizable organic substances can be recovered by elution with
aqueous solutions of inorganic acids or bases. This will give an
aqueous solution more concentrated than the original sample, but
further concentration may still be required for biological testing.
Elution with organic solvents is effective for recovering neutral
organic substances. A variety of solvents have been used to elute
the adsorbed organics from XAD columns; the most commonly used are
diethyl ether, methanol, acetone, methylene chloride, and mixtures
of these solvents. As discussed for liquid-liquid extraction,
precautions must be taken to prevent impurities in the organic
solvents from producing artifacts in the sample.
Workers from the U.S. Geological Survey have also found chat
huraic and fulvic acids are adsorbed onto both types of XAD resins
at pH 2 (Aikin et al., 1979). However, about 20% of the material
-------
150
FREDERICK C. KOPFLER
binds irreversibly to XAD-2 resin and cannot be eluted with alkali
or organic solvents. Much of the organic matter in surface waters
is composed of these naturally occurring acids or derivatives
produced during water disinfection. Because they can bind many
lower-molecular-weight organic substances, these acids should be
recovered to insure recovery of these bound materials. Owing to
its simplicity, it is tempting to use the resin technique for
producing samples for biological testing; however, it must be
remembered that not all of the organic substances are adsorbed and
those that are may not be fully recovered.
COMBINATION METHODS
Methods for isolating organics from water for biological
testing include combinations of several techniques, for convenience
or in attempts to approach 100% recovery of organics from the
water.
Kopfler et al. (1977) use reverse osmosis to reduce the volume
of water samples from thousands of liters to about forty liters.
To prevent precipitation of inorganic salts, sodium ions are
exchanged for calcium and magnesium ions in the concentrate through
a Nafion tubular membrane (Dupont) concurrently with the reverse
osmosis process. The concentrated aqueous sample is then
transported to the laboratory, where it is extracted with pentane
and methylene chloride and passed through a column of XAD-2 resin,
which is eluted with echanol. Recovery of organics by chis method
is estimated to be 35 to 40%. Residues have been tested for
mutagenicity and cellular transformation in vitro and
teratogenicity and carcinogenicity in vivo.
Johnston and Herron (1979) first pass the water through a
"parfait" column containing layers of silica gel, cation exchange
resin, and anion exchange resin. This step results in the
adsorption of some neutral hydrophobic organics as well as ionized
species including inorganic ions. The effluent from the column is
evaporated under vacuum to concentrate the hydrophilic nonvolatile
organic substances. Substances adsorbed on the column are
recovered by separating the layers in the column and eluting each
with 2 M triethylammonium carbonate buffer followed by acetone.
Samples prepared by this method have been tested for mutagenicity
in vitro.
Baird and co-workers (1980) use a series of stainless-steel
columns packed with microparticulate-sized weak ion exchange resins
and XAD resins. They report 85 to 90% removal of TOC from highly
treated wastewater passed through this series of columns. The
columns are eluted with acetonitrile and a 4.5-M sodium chloride
solution containing acetonitrile. Since the presence of
-------
ALTERNATIVE STRATEGIES FOR CONCENTRATING CHEMICALS 151
acetonitrile interferes with the TOC analysis, the actual recovery
of organics cannot be determined. The saline eluates of the
columns are concentrated further by extraction with acetonitrile at
pH 7 and again at pH 1. After these extractions, the saline
solution is still colored, indicating that organics have not been
completely recovered. These extracts have been tested for
mutagenicity in vitro.
Probably the most elaborate device for recovering organics
from water for toxicity testing is that developed in France by
Carbridenc and Sidka (1979). The apparatus is designed to extract
1000 liters of water with 100 liters of chloroform under an inert
gas. The water is extracted first at pH 7, then at pH 2, and
finally at pH 10. The aqueous sample is then neutralized and
passed through small columns containing anion exchange resin
(eluCed with butanol) , XAD-2 (eluted with a mixture of ethanol and
methylene chloride), and activated carbon (washed with ethanol and
extracted with chloroform). The recovery of 51 substances was
determined using a 100-1 version of this apparatus. Forty-five of
the compounds were detected in one or more of the fractions, and
total recovery was calculated to be 88% by weight. Extracts of
drinking water prepared by this method have been tested for
cytotoxicity in vitro and for promotion in vivo.
CONCLUSIONS
The results from biological tests of organic concentrates in
water can be used to estimate the hazards associated with the
water, but only to the degree that the concentrate represents the
organic materials actually present in the water. The concentrate
should contain representative amounts of all the organic materials
originally present, or at least a predetermined fraction of them.
Also, the integrity of the chemicals must be maintained, with no
contaminants present, or at least none that interfere with the
biological tests. Until such methods or combinations are developed
and validated, the information in this paper should serve as a
guide to the representativeness of concentrates produced by various
methods and should allow proper reservations to be made in the
interpretation of biological test results.
REFERENCES
Aikin, G.R., E.M. Thurman, R.L. Malcolm, and H.F. Walton. 1979.
Comparison of XAD macroporous resins for the concentration of
fulvic acid from aqueous solution. Anal. Chem. 51:1799-1803.
-------
152
FREDERICK C. KOPFLER
Baird, R., J. Gute, C. Jacks, R. Jenkins, L. Neisess, B.
Scheybeler, R. Van Sluis, and W. Yanko. (1980). Health
effects of water reuse: a combination of toxicological and
chemical methods for assessment. In: Water Chlorination ,
Environmental Impact and Health Effects, Vol. 3. R.L. Jolley,
W.A. Brungs, and R.B. Cumraing, eds. Ann Arbor Press: Ann
Arbor, MI.
Baker, R.A. 1969. Trace organic contaminant concentration by
freezing—III. ice washing. Water Res. 3:717-730.
Baker, R.A. 1970. Trace organic contaminant concentration by
freezing—IV. ionic effects. Water Res. 4:559-573.
Buelow, R.W., J.K. Carswell, and J.M. Symons. 1973. An improved
method for determining organics by activated carbon adsorption
and solvent extraction. J. Am. Water Works Assoc. 65:57-72.
Carbridenc, R. and A. Sidka. 1979. Extraction des micropollutants
organiques des eaux en vue de la realisation d'essais
biologiques. Presented at the European Symposium on the
Analysis of Organic Micropollutants in Water, Berlin, Federal
Republic of Germany.
Chriswell, C.D., R.L. Ericson, G.A. Junk, K.W. Lee, J.S. Fritz, and
H.J. Svec. 1977. Comparison of aacroreticular resin and
activated carbon as sorbents. J. Am. Water Works Assoc.
69:669-674.
Crathorne, B., C.B. Watts, and M. Fielding. 1979. The analysis of
non-volatile organic compounds in water by high-performance
liquid chromatography. J. Chromatogr. 185:671-690.
Dressier, M. 1979. Extraction of trace amounts of organic
compounds from water with porous organic polymers. J.
Chromatogr. 165:167-206.
Gustafson, R.L., and J. Paleos. 1971. Interactions responsible
for the selective adsorption of organics on organic surfaces.
In: Organic Compounds in Aquatic Environments. S.J. Faust
and J.V. Hunter, eds. Marcel Dekker, Inc.: New York. pp.
213-237.
Heraon, D., P. Lazor, R. Cabridenc, A. Sidka, B. Festy, C.
Gerinroze, and I. Chouroulinkov. 1978. I: Micropollution
organique des eaux destinees a la consummation huraaine. Rev.
Epidem. et Sante Publ. 26:441-450.
-------
ALTERNATIVE STRATEGIES FOR CONCENTRATING CHEMICALS
153
Johnston, J.B., and J.N. Herron. 1979. A Routine Water Monitoring
Test for Mutagenic Compounds. UIUC-WRC-79-0141. University
of Illinois: Urbana, IL. 87 pp.
Jolley, R.L. , S. Katz, J.E. Morchek, W.W. Pitt, and W.T. Rainey.
1975. Analyzing organics in dilute aqueous solution. Chem.
Tech. 5:312-318.
Junk, G.A., J.J. Richard, M.D. Grieser, D. Witiak, J.D. Witiak,
M.D. Arguello, R. Vick, H.J. Svec, J.S. Fritz, and G.V.
Calder. 1974. Use of macroreticular resins in the analysis
of water for trace organic contaminants. J. Chroraatogr.
99:745-762.
Kopfler, F.C., E.W. Coleman, R.C. Melton, R.C. Tardiff, S.C. Lynch,
and J.K. Smith. 1977. Extraction and identification of
organic micropollutants: reverse osmosis method. Ann. N. Y.
Acad. Sci. 298:20-30.
Logsdon, O.J., K.. E. Nottingham, and T.O. Meiggs. 1977 . Formation
of nitrosamines and chlorocycloalkanes during analytical
procedures. Presented at the 91st meeting of the Association
of Official Analytical Chemists, Washington, DC.
Malcolm, R.L., E.M. Thurman, and G.R. Aiken. 1977. The
concentration and fractionation of trace organic solutes from
natural and polluted water using XAD-8, methylmethacrylate
resin. In: Trace Substances in Environmental Health, Volume
XI. D.D. Hemphill, ed. University of Missouri: Columbia,
MO. pp. 307-314.
Modell, M. , R.P. deFilippi, and V. Krukonis. 1978. Regeneration
of activated carbon with supercritical carbon dioxide.
Presented before the Division of Environmental Chemistry,
American Chemical Society, Miami, FL.
Pitt, W.W., and C.D. Scott. 1973. Measurement of molecular
organic contaminants in polluted water. In: Ecology and
Analysis of Trace Contaminants. ORNL-NSF-EATC-1. Oak Ridge
National Laboratory: Oak Ridge, TN. pp. 309-331.
Shapiro, J. 1961. Freezing out, a safe technique for
concentration of dilute solutions. Science 133:2063-2064.
Thurman, E.M., R.L. Malcolm, and G.R. Aiken, 1978. Prediction of
capacity factors for aqueous organic solutes adsorbed on a
porous acrylic resin. Anal. Chem. 50:775-779.
-------
Intentionally Blank Page
-------
DETECTION OF ORGANIC MUTAGENS IN WATER RESIDUES
John C. Loper and M. Wilson Tabor
Departments of Microbiology and Environmental Health
University of Cincinnati College of Medicine
Cincinnati, Ohio
INTRODUCTION
In previous studies (Loper and Lang, 1978; Loper et al. , 1978;
Lang et al., in press; Kurzepa et al., in press), we used short-
terra bioassays to demonstrate the mutagenicity, carcinogenicity,
and toxicity of residues prepared from samples of drinking water
from six U. S. cities. The samples were processed by Gulf South
Research Institute (New Orleans, LA), using reverse osmosis plus
XAD resin sorption-desorption as described by Kopfler et al.
(1977). Using the Ames test, we found city-specific patterns of
dose-dependent mutagenesis that were essentially independent of the
microsomal activation system. One or more samples from each city
showed reproducible transformation frequencies at least three times
the spontaneous frequency. Focus formation induced by these
samples was equivalent to malignant transformation as verified in
nude mice. In these studies, quantitation of mutagenic and
transformation responses were complicated by the toxicity and
heterogeneity of the complex residue mixtures.
These findings justify further efforts at compound
identification, and for that purpose, we have proposed the
development of a coupled bioassay/chemical fractionation procedure
(Loper and Lang, 1978). Such a method would be patterned after
successful analyses of other complex environmental mixtures such as
synthetic fuels (Guerin et al., 1978). Mutagenicity would be
assayed using the Ames test. Initial partitioning of the sample by
1iquid/1iquid extraction would be followed by high-performance
liquid chromatography (HPLC) for separation into smaller
subtractions. Active fractions sufficiently free of inactive and
155
-------
156
JOHN C. LOPER AND M. WILSON TABOR
bi.oci.dal components would be analyzed by gas chroraatography/mass
spectrometry (GC/MS) for identification of peaks.
Drinking water residues are not usually generated in
sufficient quantity for developing such methods. However, from Mr.
Francis Middleton we obtained residue still on hand from previous
use of the U. S. Public Health Service carbon-chloroform extractor.
This mega sampler was used in the 1950's to early 1960's for the
processing of 100,000-gal samples of drinking water and source
water. The sample is a liquid solution in chloroform (CHCI3) of
125 g of residue obtained about 1960 from 50,000 gal of drinking
water (Middleton et al., 1962). When the CHCI3 is removed under a
stream of nitrogen, the residue has the sticky consistency of water
residues recovered from XAD resins. We have termed this material
the "carbon-chloroform extracted organics" (CCEO). In precise
composition, it may or may not closely resemble residues from
drinking water today, but for our work, it is invaluable as an
abundant supply of a complex mixture of water residuals.
Using this CCEO, we have attempted to develop a general method
of coupled bioassay/chemical fractionation for separating mixtures
from current drinking water. With such a method we could test
immediately whether the mutagenicity of such mixtures in the Ames
test is due to summation of the low activity of many mutagens or to
the effects of a few relatively active components. The method
should allow isolation of active subtractions in yields suitable
for compound identification. Some significant biohazardous
properties of residue components may not be detected by our
bioassay procedure of bacterial mutagenesis plating. For example,
co-carcinogens, carcinogen promoters, and certain procarcinogens
and teratological agents would not be recognized. So that other
subtractions could be tested using short-term in vitro and in vivo
mammalian assay systems, the method should permit maximum recovery
of the total sample, distributed into multiple subtractions.
METHODS AND RESULTS
Preliminary Characterization and Primary Partitioning of CCEO
Aliquots of CCEO were shown to be reproducibly stripped of all
toxic traces af CHCI3 by streaming with dry nitrogen for 60 min at
60°C. Mutagenesis testing of the residue using Salmone11a strains
TA98 and TA100, as described elsewhere (Loper et al., 1978),
revealed dose-related, microsomal-activation-dependent TA100
mutagenesis, with some toxicity at higher doses. Following
experimentation with aqueous extraction of the residue from the
original CHCI3 solutions and from solutions in methylene chloride
(CH2CI2) (Tabor and Loper, 1980) , a procedure was adopted for the
semisolid/liquid extraction of neutrals, acids, and bases into
-------
DETECTION" OF ORGANIC MUTAGENS IN WATER RESIDUES
157
hexane. The resulting distributions of weight and mutagenic
activity for TA98 and TA100 with and without microsomal activation
have been detailed elsewhere (Tabor et al., 1980). The neutrals
are one third of the total sample by weight but contain nearly all
of the mutagenicity for TA100. All subsequent procedures were
conducted on portions of this CCEO neutral sample.
Method Development
Quantities of partitioned samples taken for mutagen isolation
had to be sufficient for repeated cycles of separation and
bioassay. To obtain useful data on both mutagenicity and toxicity
with minimum loss of material, tests were limited to single plates
of four dose levels chosen to induce colony counts of two to three
times those seen spontaneously (Loper, 1980). Reverse-phase HPLC,
employing mixtures and gradients of water/acetonitrile, was used
for chemical separation. The instrument was a Waters Associates
ALC/GPC 204 equipped with two 6000A pumps, UK6 injector, solvent
programmer, and 254-nm absorbance detector. A guard column (3.9 mm
x 2.5 cm) packed with pellicular particles bonded with
octadecylsilane (BONDAPAK—Cjg/CORASIL; Waters Associates) was
followed by a radial compression module (RCM; Waters Associates)
containing a 8-mm x 10-cra column packed with 10-ym silica particles
bonded with a high load of octadecylsilane. To achieve adequate
separation levels, various gradient elution conditions were
investigated using the analytical column, as sample quantities were
increased from microgram to milligram levels. One milligram of the
CCEO neutral sample induced approximately 2000 TA100 colonies in
our assay, and for RCM chromatography, 20-mg samples were routinely
loaded in 200-pl volumes of acetonitrile.
A flow diagram of our procedure for partitioned samples is
given in Figure 1. Activity losses accompanying removal of the
bactericidal acetonitrile were minimized by solvent exchange into
CH2CI2 using SEP-PAKS (Waters Associates) packed with pBONDAPAK-C^g
(Waters Associates) followed by evaporation of the CH2CI2 at 40°C
under dry nitrogen in the presence of a small volume of
dimethylsulfoxide. Replicate separations reproduced the HPLC
fingerprint at the top of Figure 2, and with this added material,
we assayed fractions 1 through 6 using both TA98 and TA100 with and
without microsomal activation. As before, all the activity was
detected with TA100 and appeared in fraction 5B. Rechroraatography
of this region gave the isolated major peaks at the bottom of
Figure 2. Subtractions 5B/5 and 5B/6 contained all the mutagenic
activity, and each of these was rechromatographed (Figure 3). Such
subtractions are currently undergoing further study; some
preliminary data are presented here. In one series of four dose-
level determinations, the summed mutagenesis from the active
subtractions (5B/5/2, 5B/6/2, 5B/6/3) was approximately half that
-------
158
JOHN C. LOPER AND M. WILSON TABOR
20 mg Sample
I
rcm/hplc
Separation 1
I
-~Dilute Active Fractions
3X with HoO
4
RCM/HPLC
Separation 2 3
l
Active Fractions
r
GC Analysis 1
on OV 1 or Silar 5CP
J
GC/MS/DS
I
Dilute 3Xwith H20
Load on SEP-PAK C
18
I—
Effluent
I
i
Elute with CH2CI2
r
•HPLC Microanalysis GC Analysis 2 Partially Evaporate
Ir
Add DMSO
I
Evaporate
I
Broassay
Discard
Figure 1. Flow diagram for coupled bioassay/chemical fractionation
of partitioned complex mixtures.
-------
DETECTION OF ORGANIC MUTAGENS IN WATER RESIDUES
15
Figure 2. RCM/HPLC reverse-phase separation of CCEO neutrals. A
20-mg sample was fractionated, using gradient elutions
as shown, and rechromatographed by the procedures given
in Figure 1. Subfractions mutagenic to TA100 in the
presence of microsomal activation are indicated by a
check (/) .
-------
160
JOHN C. LOPER AND M. WILSON TABOR
fBflCTlON SB/5
50% CH-CN
3||5| |
FRACTION Number
Figure 3. Fractions 5B/5 and 5B/6, generated as shown in Figure
2, were rechromatographed separately. Mutagenic
subtractions are indicated by a check (/).
from the initial 20—mg sample. The relative purity of residue in
each subtraction was estimated by GC analysis. Tracings of GC
chromatograms obtained using a column containing 10% SE30 on
Chromosorb WHP 80/100 mesh (Applied Science) as a stationary phase
are shown in Figure 4.
Subtractions 5B/5/2 and 5B/6/2 each contained two different
components, and one other component constituted subfraction 5B/6/3.
To date, three 20-rag aliquots of CCE0 neutrals have been carried
through the procedure as far as peak separation and GC analysis,
using both polar (102 Silar 5CP on Chromosorb Q 80/100 mesh;
Applied Science) and nonpolar stationary phases (10% SE30 on
Chromosorb WHP 80/100 mesh or 3% OVl on Chromosorb W 80/100 mesh:
Applied Science): the patterns shown in Figures 3 and 4 are
representative of the purity of all three aliquots.
-------
DETECTION OF ORGANIC MUTAGENS IN WATER RESIDUES
161
MINUTES
Figure 4. Flame-ionization gas chroraatograms of RCM/HPLC
subfractions . Seven and one-half microliters of a
water/acetonitrile solution (about 50:50 by volume) of
each subfraction was slowly injected into a Perkin-
Elmer Model 900 GC fitted with a 2-mm x 1-ra stainless
steel column containing 10% SE30 on Chromosorb WHP
80/100 mesh. The nitrogen carrier gas flow was at 18
ml/rain, and the temperatures of injector and detector
were 280°C and 350°C respectively. A linear temperature
program (120°C to 220°C at 4a/min) was initiated at the
time of injection. Data were collected continuously and
analyzed using a Spectra Physics Autolab System I
Computing Integrator, and chromatograms were displayed
on a 10-mV recorder at an attenuation of 160.
-------
162
JOHN C. LOPER AND M. WILSON TABOR
Weight values for a sample constituents were calculated based
on peak areas compared with peak, areas obtained from chromatography
of 1 pi of a 1 rag/ml solution in chloroform of American Oil
Chemists Society Reference Mixture No. 6, run under conditions
identical to those for the experimental samples. These weights
have been used in estimating specific activity, in net revertant
colonies per milligram, shown in Table 1 (see Figure 4). Thus for
5B/5/2 and for 5B/6/2, the weights of their two GC components have
been combined, so that results are expressed as net revertant
colonies per total weight of each subfraction. Preliminary GC/MS
analyses indicate that some of these compounds are polyhalogenated:
this is indicated in Table 1 by an X with the number of components
for each subfraction.
TabLe 1. Response of TA100 to Mutagens in the CCEO
Neutrals Mixture
Net Revertant Colonies (+S-9)
RCM/HPLC Subtractions3 Total Per mg
5B/5/2 (IX + 1)
6000
3 X 10-
53/6/1 (?)
?
—
5B/6/2 (IX + I)
4600
10 5
5B/6/3 (IX)
13000
4 X 10 5
aNumber of major peaks in parentheses: X indicates the presence of
halogen(s), based on preliminary MS data.
DISCUSSION AND CONCLUSIONS
We have demonstrated that this coupled bioassay/chemical
fractionation procedure is reproducible for this complex residue
mixture, and we feel it will serve as a general method. For the
present, we assume that concentration methods yield residues
representative of the non-volatile organics in drinking water.
Based on our observations in the six-city study and our results
from CCEO neutral sample, drinking water residues appear to contain
a vast number of non-rautagenic compounds, possibly some of low
mutagenicity, and a few highly mutagenic compounds. Evidence for
-------
DETECTION OF ORGANIC MUTAGENS IN WATER RESIDUES
163
this is summarized in Table 2. Of Che RCM/HPLC subtractions of the
CCEO neutrals, only three were mutagenic. Peak 5B/6/3, containing
a single component by GC, showed a specific activity comparable to
those of the carcinogens S-naphthylamine or 3-methylcholanthrene
(McCann et al., 1975). The other two subfractions contained two GC
components each, and their mutagens must have been comparably
potent (see Table 1). Of course, these subfractions were derived
from the old CCEO sample. For more recently isolated residues, the
presence of highly active compounds was suggested by our data for
the XAD eluate of Seattle sample 1. This fraction showed direct-
acting mutagenesis, which was increased 24-fold in specific
activity (revertant colonies per milligram of sample material) by
extraction into hexane (Loper et al., 1978: see Table 2). The
specific activity of the active compound(s) might be considerably
higher, based on our observation that this hexane-extracted
subtraction yielded 20 major 254-nm absorption peaks through HPLC
(Tabor et al., 1980).
Table 2. Mutagenesis of Drinking Water Residue Fractions
and Known Carcinogens in the Ames Test (TA100)a
Net Revertant
Test Substance Colonies/Plate/mg
XAD eluate of Seattle drinking water
(sample l)
4
X
102
Hexane extract of XAD eluate of Seattle
drinking water (sample 1)

10"
CCEO neutrals fraction 5B/6/3
4
X
10 5
3-propiolactone
5
X
10u
6-naphthy1 amine
6
X
104
3-methylcholanthrene
2
X
10 5
aData for XAD eluate and subfraction is from Loper et al., 1978:
that for known compounds is from McCann et al., 1975.
Final identification of the mutagenic components in
may be of some direct benefit, even though the sample is
20 years old, and some of the halogen content may be due
the CCEO
now nearly
to storage
-------
164
JOHN C. LOPER AND M. WILSON TABOR
in CHCI3 . Identification by GC/MS is in progress. Should
previously unidentified mutagens be detected, their
characterization by MS would permit data searches, by peak
recognition, among MS profiles of residues of current drinking
water samples.
We propose to apply our procedure of coupled bioassay/chemical
fractionation to the identification of mutagens in recently derived
drinking water residues, obtained by desorption from XAD. Should
major mutagens in these samples be identified, a number of
questions concerning water quality could be investigated. These
topics include the reduction or avoidance of mutagens by alternate
disinfection procedures; seasonal changes in the types and amount
of mutagenic activity; and the mutagenic potential of discharges of
diverse industrial processes into surface or ground water destined
for human use. Knowledge of specific mutagens in a water sample
would provide an important criterion for evaluating alternate
residue-isolation procedures and could lead to simplification of
analytical chemical detection methods. Identification of
significant non-volatile bacterial mutagens in water would be a
step toward toxicological assessment of their risk to man.
ACKNOWLEDGMENTS
We are grateful to Debbie Spector for technical assistance;
to our colleagues Dr. Carl C. Smith, for advice and encouragement
in this work, and Dr. Joseph MacGee, for analytical contributions;
and particularly to Mr. Francis Middleton for providing us the
CCEO sample. This research was supported by a grant from the U.S.
Environmental Protection Agency.
REFERENCES
Guerin, M.R., B.R. Clark, C.-h. Ho, J.L. Epler, and T.K. Rao.
1978. Short-term bioassay of complex organic mixtures: part
I, chemistry. In: Application of Short-term Bioassays in the
Fractionation and Analysis of Complex Environmental Mixtures.
M.D. Waters, S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L.
Claxton, eds. Plenum Press: New York. pp. 247-268.
Kopfler, F.C. , W.E. Coleman, R.G. Melton, R.G. Tardiff, S.C. Lynch,
and J.K. Smith. 1977. Extraction and identification of
organic raicropollutants: reverse osmosis method. Ann. N. Y.
Acad. Sci. 298:20-30.
-------
DETECTION" OF ORGANIC MUTAGENS IN WATER RESIDUES
165
Kurzepa, H., A.P. Kyriazis, and D.R. Lang. (in press). Growth
characteristics of tumors induced by tranplantation into
athyraic mice of BALB/3T3 cells transformed in vitro by residue
organics from drinking water. J. Environ. Pathol. Toxicol.
Lang, D.R., H. Kurzepa, M.S. Cole, and J.C. Loper. (in press).
Malignant transformation of BALB/3T3 cells by residue organic
mixtures from drinking water. J. Environ. Pathol. Toxicol.
Loper, J.C. 1980. Overview of the use of short-terra biological
tests in the assessment of the health effects of water
chlorination. In: Water Chlorination: Environmental Impact
and Health Effects, Vol. 3. R.L. Jolley, W.A. Brungs, and R.B.
Cunnning, eds. Ann Arbor Science: Ann Arbor, MI. pp. 937-945
Loper, J.C., and D.R. Lang. 1978. Mutagenic, carcinogenic, and
toxic effects of residual organics in drinking water. In:
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures. M.D. Waters, S.
Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
Plenum Press: New York. pp. 513-528.
Loper, J.C., D.R. Lang, R.S. Schoeny, B.B. Richmond, P.M.
Gallagher, and C.C. Smith. 1978. Residue organic mixtures
from drinking water show in vitro mutagenic and transforming
activity. J. Toxicol. Environ. Hlth. 4:919-938.
McCann, J., E. Choi, E. Yamasaki, and B.N. Ames. 1975. Detection
of carcinogens as mutagens in the Salmonella/microsome test:
assay of 300 chemicals. Proc. Natl. Acad. Sci. USA.
72:5135-5139.
Middleton, F.M., H.H. Pettit, and A.A. Rosen. 1962. The mega
sampler for extensive investigation of organic pollutants in
water. In: Proceedings of Che 17th Industrial Waste
Conference. Engineering Ext. Ser. 112:454-460. Purdue
University: Lafayette, IN.
Tabor, M.W., and J.C. Loper. 1980. Separation of mutagens from
drinking water using coupled bioassay/analytical fractionation
Int. J. Environ. Anal. Chera. 8:1-19.
Tabor, M.W., J.C. Loper, and K. Barone. 1980. Analytical
procedures for fractionating non-volatile mutagenic components
from drinking water concentrates. In: Water Chlorination:
Environmental Impact and Health Effects, Vol. 3. R.L. Jolley,
W.A. Brungs, and R.B. Cumming. eds. Ann Arbor Science: Ann
Arbor, MI. pp. 899-912.
-------
Intentionally Blank Page
-------
SHORT-TERM METHODS FOR ASSESSING IN VIVO CARCINOGENIC
ACTIVITY OF COMPLEX MIXTURES
Michael A. Pereira and Richard J. Bull
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
INTRODUCTION
The carcinogenic activity of a chemical or a complex mixture
derived from an environmental sample is assessed most efficiently
through a three-tier decision scheme (Bridges, 1973; Weisburger and
Williams, 1977; Bull and Pereira, in press). In tier 1, the
samples are screened for evidence of carcinogenic and mutagenic
activity. The Ames Salmonella mutation bioassay and _in vitro and
in vivo cytogenetic assays appear to sucessfully identify most
chemicals and samples with carcinogenic activity. These two types
of assays detect the two major classes of genotoxic agents,
mutagens and clastogens, and would therefore form the backbone of
tier 1. Other possible assays for tier 1 include mammalian cell
mutation, sister-chromatid exchange, raicronuclei, and unscheduled
DNA synthesis. The nature of tier 1 bioassays—especially their
lack of correlation to carcinogenic potency, the absence of a
direct demonstration of cancer or neoplasia, and the number of
false positives—requires that carcinogenic activity be confirmed
in tier 2.
Tier 2 is the level in the decision tree where false positives
are eliminated, insuring that Che time-consuming and expensive tier
3 bioassay is used only for chemical and environmental samples that
are virtually certain to contain carcinogenic activity. The
function of the tier 3 bioassay is to provide data for assessing
Che quantitative risk associated with the carcinogenic activity of
chemicals and samples. To be effective, tier 2 must confirm
carcinogenic activity wich a minimum of false positives, while
maintaining an acceptably low level of false negatives.
167
-------
168 MICHAEL A. PEREIRA AND RICHARD J. BULL
Quantitation of carcinogenic activity in environmental samples
of complex mixtures, and thus ranking of mixtures according to
carcinogenic activity, poses a unique problem. Lifetime feeding or
exposure studies with rodents in a bioassay following the National
Cancer Institute (NCI) protocol is presently the only acceptable
procedure for obtaining data on carcinogenic potency that may be
extrapolated to humans (IRLG, 1979). This type of bioassay is not
feasible with most complex mixtures derived from drinking water and
other environmental samples. For example, the expense of obtaining
a single drinking-water sample for an NCI bioassay greatly exceeds
that of the bioassay itself, potentially costing millions of
dollars. Furthermore, the uniqueness of each drinking-water sample
(containing thousands of unknown chemicals) prevents the
generalization of the bioassay results to other drinking waters
(Bull et al., in press). The composition of drinking waters varies
depending on treatment (especially disinfectant), source (ground or
surface), seasons, and the effects of industrial use and municipal
waste disposal. Therefore, each drinking water would have to be
considered a unique test substance, and each would require a
separate bioassay, costing millions of dollars, at least until
enough assays were performed to determine whether generalizations
among drinking waters were possible.
CARCINOGENESIS TESTING MATRIX
The lack of a tier 3 bioassay for complex mixtures derived
from environmental samples means that any ranking or quantitation
of carcinogenic activity must be obtained at tier 2. To accomplish
this, tier 2 bioassay results would have to relate quantitatively
Co carcinogenic potency. Also, since no short-term bioassay
appears to be sensitive to all chemical carcinogens, a
Carcinogenesis Testing Matrix (CTM)(Bull and Pereira, in press)
has been proposed for tier 2. It includes mouse lung adenoma
(Shimkin and Stoner, 1975; Stoner and Shimkin, in press), mouse
skin initiation/promotion (Slaga et al., in press), rat liver foci
(Pereira, in press), in vitro cell transformation (DiPaolo, 1979;
Styles, 1979), and in vivo sister-chromatid exchange (Latt et al.,
1979). These bioassays were chosen because evidence, or some
reasonable rationale, indicates that their results will relate to
carcinogenic potency, at least when taken together. The in vivo
bioassays appear most attractive, because their results should be
influenced by the same pharmacokinetic and metabolic factors as in
lifetime-exposure carcinogenicity bioassays. Where possible,
bioassays that directly measure the acquisition of neoplastic
properties, such as benign tumors, preneoplastic lesions, and cell
transformation, are included in the CTM.
The increased sensitivity of the CTM bioassays (compared with
lifetime exposure bioassays) greatly reduces the sample requirement.
-------
ASSESSING IN VIVO CARCINOGENICITY
169
These bioassays require a few applications of sample, at most.
This increased sensitivity can be illustrated with the rat liver
foci bioassay (Pereira, in press). Briefly, the bioassay protocol
is as follows:
1) the sample is administered by any convenient route in
single or multiple doses;
2) the rats are given 500 ppm sodium phenobarbital in their
drinking wacer, starting four to seven days after the
sample and continuing for seven weeks; and
3) the rats are then sacrificed, and their livers excised
and examined histochemically for foci of gamma glutamyl
transpeptidase.
A two-thirds hepatectoray can be performed either 18 to 24 h before
or 14 days afcer administering Che sample. Performing the partial
hepatectomy 18 to 24 h before exposure increases the spectrum of
chemical classes of carcinogens to which the assay is sensitive and
increases the response. Performing it 14 days after exposure also
increases the response and keeps the initiation step distinctly
separated from the promotion.
Table 1 shows the results for diethylnitrosamine (DENA) in the
rat liver foci bioassay. A single dose of DENA (30 rag/kg) gave
positive results when administered either 24 h after or 14 days
prior to partial hepatectoray. When the partial hepatectomy was
performed 14 days after DENA exposure, a single dose as low as 300
pg/kg was detected. Since partial hepatectomy 24 h prior to the
DENA increases sensitivity to DENA (Scherer and Eramelot, 1976), one
would expect an even lower single dose to be detected. Ten low
daily doses of 300 ug/kg, for a total dose of 3 mg/kg, appear to be
additive in the bioassay (Table l). The additivity of multiple low
doses can also result in increased sensitivity (Ford and Pereira,
in press). The use of tumor promoters, preneoplastic lesions, and
benign tumors in short-term bioassays greatly decreases the amount
of sample required so that acquiring sufficient amounts of sample
becomes feasible.
DIFFICULTIES IN RANKING POTENCY FROM SINGLE 3I0ASSAYS
The use of two or more bioassys to rank the carcinogenic
activities of complex mixtures could result in different rankings
in the different bioassays, or in some mixtures giving a positive
result in one bioassay while other mixtures were positive only in
another bioassay. Table 2 outlines a hypothetical situation where
four environmental samples of complex mixtures (1, 2, 3, and 4)
were tested in bioassays A and B. The possible results include
-------
170 MICHAEL A. PEREIRA AND RICHARD J. BULL
Table 1. Rat Liver Foci Bioassay of Diethvlnitrosamine (DENA)
GGTase-Positive
DENAa No. of Partial Foci/cm2
(mg/kg) Animals Hepatectoray (mean ± std. error)
Experiment l*3
30
10
-24
h
18.8
+
4.6
30
10
+ 14
days
3.92
>
0.91
0
10
-24
h
0.40
±
0.14
:periment 2C

0.3
10
+ 14
days
1.02
+
0.17
0.3 x 10
9
+ 14
days
2.16
+
1 .02
3
52
+ 14
days
1.51
+
0.25
0 x 10
23
+ 14
days
0.59
+
0.20
aRats wereadministered DENA in 0.3 mldistilled water by gastric
intubation. One week later the rats received 500 ppra
phenobarbital in their drinking water for one week. After partial
hepatectoray on day 14, the phenobarbital was decreased to 250 ppm.
The rats were maintained on this concentration of phenobarbital
for four to five weeks. Cyrostatic sections of liver were stained
for GGTase activity and at least 2 cm2 examined for foci,
bpereira, in press.
cFord and Pereira, in press.
Table 2. Possible Outcomes of a Two-Bioassay Matrix Testing
of Drinking Water
Bioassay
Drinking Water Sample A
1 +
2 +
3 +
4 -
-------
ASSESSING IN VIVO CARCINOGENICITY
171
samples 1 and 2 positive in bioassay A, and 1 and 3 positive in
bioassay 3. Of the three environmental samples possessing
carcinogenic activity, the question then is, which sample is the
most hazardous? To rank samples 2 and 3, which are positive in
different bioassays, one must critically compare the results from
the two bioassays. That sample A is negative in both bioassays
does not mean it is not a potent carcinogen, unless it is
demonstrated that the two bioassays as a set can detect all the
carcinogens conceivably present in the environmental samples.
Since most bioassays appear at least somewhat selective in their
responses to well-established chemical carcinogens, it seems
inappropriate to judge relative hazard on the basis of how many
tests are positive.
The relative hazard of the environmental samples could be
derived from the relative responses in the bioassays of the CTM, if
the responses in the various bioassays were calibrated to the same
standardized estimate of carcinogenic potency. A test substance
would have a standardized estimate of carcinogenic potency derived
from each bioassay of the CTM. A decision would have to be made on
a procedure for using these individual estimates to arrive at a
single estimate. This might be accomplished by averaging the
individual estimates or by accepting the estimate of highest
carcinogenic potency.
The following examples more explicitly illustrate the problem:
1) The Ames Salmonella mutagenicity bioassay has been
proposed as correlating with carcinogenic potency (Meselson and
Russell, 1977). This correlation requires the exclusion of
nitrosamines, since their Ames test response is very weak compared
with their carcinogenicity. When environmental samples of complex
mixtures of unknown chemical composition were assayed, the Ames
test would predict an erroneously low carcinogenic activity if
highly active nitrosamines were present. When a liquid
preincubation is used with the Ames test, nitrosamines can be
detected, though still not at a level reflecting their potency.
The use of liquid preincubation could also change the ranking of
the other carcinogens. The results of the two bioassays (standard
and preincubation Ames tests) would have to be compared as if they
were two separate tests.
2) As another example, when drinking water concentrates from
five cities were assayed in the Ames Salmonella mutagenicity and
mouse skin initiation/promotion tests, the results of these two
tests were not correlated (Loper et al., 1979; Robinson et al. ,
1980). The mouse skin initiation/proraotion assay is very sensitive
to polycyclic aromatic hydrocarbons (PAH) and nitrosamines
(Pereira, in press). Comparing the responses of PAH and
nitrosamines in the mouse skin and Salmonella mutagenicity tests
-------
172
MICHAEL A. PEREIRA AND RICHARD J. BULL
reveals many discrepancies (Andrews et a!., 1978a, b). Among 25
polycylic aroraati.cs, mutagenicity and carcinogenicity were
positively correlated for 58% and negatively correlated for 41%
(Andrews et al., 1978a). Since the two assays rank these chemicals
differently, the responses of the two assays to drinking water
samples are not expected to be correlated if either PAH or
nitrosamines are present. The critical question is how to rank the
carcinogenic activity of environmental samples based on results
from more than one assay.
VALIDATION OF THE CARCINOGENESIS TESTING MATRIX
Results from two or more bioassays can be compared by
calibrating each bioassay with respect to the carcinogenic potency
of the chemicals to which each bioassay is sensitive. Calibration
curves similar to the one used by Meselson and Russell (1977) can
be determined; these investigators used the reciprocal of the log
response in the short-terra bioassay versus the reciprocal of the
log carcinogenic potency. The reciprocal of the carcinogenic
potency of each chemical can be calculated from lifetime exposure
bioassay results as the dose (Dose 1/2) required to produce tumors
in 50% of the animals (rats and mice) in two years. The highest
estimate of carcinogenic potency is to be used in the calibration
curves, since this is the value employed for extrapolation to man
by the Carcinogen Asessment Group of the U.S. Environmental
Protection Agency, Office of Health and Environmental Assessment.
For certain chemicals or chemical classes, individual
bioassays of the CTM will respond poorly relative to the
carcinogenic potencies of the chemical classes. As long as other
bioassays of the CTM correctly indicate the potencies of such
compounds, the responses for these compounds will be dropped from
the calibration curve of any bioassay that responds poorly. For
the matrix, it is the overall correlation with carcinogenic
activity that is important. This method of selecting chemicals to
be incorporated into the calibration curve of a bioassay results in
a nonrandom grouping of chemicals, making further testing necessary
for validation of the CTM.
If a chemical is extremely potent in a particular bioassay,
compared with its carcinogenic potency derived from lifetime
exposure, it may be necessary to delete that bioassay from the CTM.
The CTM is being proposed for use with complex mixtures of
undefined chemical composition and where a bioassay is not
feasible. Therefore, such a response in a bioassay would tend to
result in consistent overestimates of carcinogenic risk. Before a
decision on keeping the bioassay in the CTM is made, the data for
the chemical in the long-terra carcinogenesis bioassay and the
short-term bioassays will be carefully reviewed. Additional
-------
ASSESSING IN VIVO CARCINOGENICITY
173
chemicals of Che same general class will be tested in the bioassay
to determine whether this is a general problem with the bioassay or
whether it is confined to a single chemical.
Validating the CTM to rank the carcinogenic hazard of
chemicals and environmental samples will involve three rounds of
testing the ability of each component bioassay and of the CTM as a
whole to predict carcinogenic potency. Archetypal chemicals
representing the various chemical classes of carcinogens will be
tested in round one to determine calibration curves. In round two,
carcinogenic and noncarcinogenic (or weakly carcinogenic) analogues
will be tested to determine the ability of the CTM and its
component bioassay to predict relative carcinogenic potency. At
this time, it should be possible to decide whether the CTM is valid
and practicable and what each bioassay contributes to the CTM. The
third round will involve testing additional carcinogenic and
noncarcinogenic analogues, surrogate mixtures containing two or
more carcinogens, and complex mixtures spiked with known
carcinogens. The resulting CTM will then be ready for use in
determining the relative carcinogenic hazard associated with
drinking water and other environmental samples of complex mixtures.
REFERENCES
Andrews, A.W., L.H. Thibault, and W. Lijinsky. 1978a. The
relationship between carcinogenicity and mutagenicity of some
poLynuclear hydrocarbons. Mutation Res. 51:311-318.
Andrews, A.W., L.H. Thibault, and W. Lijinsky. 1978b. The
relationship between mutagnicity and carcinogenicity of some
nitrosamines. Mutation Res. 51:319-326.
Bridges, B.A. 1973. Some general principles of mutagenicity
screening and possible framework for testing procedures.
Environ. Hlth. Perspect. 6:221-227.
Bull, R.J., and M.A. Pereira. (in press). Development of a
short-term testing matrix for estimating relative carcinogenic
risk. J. Environ. Pathol. Toxicol.
Bull, R.J., M.A. Pereira, and K.L. Blackburn. (in press).
Bioassay techniques for evaluating the possible
carcinogenicity of absorber effluents. In: Conference on
Practical Application of Adsorption Techniques.
DiPaolo, J.A. 1979. Quantitative transformation by carcinogens
of cells in early passage. In: Environmental Carcinogenesis.
P. Emmelot and E. Kriek, eds. Elsevier Press: Amsterdam,
pp. 365-380.
-------
174
MICHAEL A. PEREIRA AND RICHARD J. BULL
Druckrey, H., A. Schildback, D. Schmahl, R. Preusraann, and S.
Ivankovic. 1963. Quantitative Analyse der carcinogenen
Wirkung von Diathynitrosamin. Arzneimittel-Forsch. 13:841-851.
Ford, J.O., and M.A. Pereira. (in press). Short-term in vivo
initiation/promotion bioassay for hepatocarcinogens. J.
Environ. Pathol. Toxicol.
IRLG, Interagency Regulatory Liaison Group, Work Group on Risk
Assessment. 1979. Scientific basis for identification of
potential carcinogens and estimation of risks. J. Natl.
Cancer Insc. 63:241-248.
Latt, S.A., R.R. Schreck, K.S. Loveday, and C.R. Shuler. 1979.
In vitro and in vivo analysis of sister chromatid exchange.
Pharmacol. Rev. 30:501-535.
Loper, J.C., D.R. Lang, R.S. Schoeny, B.B. Richmond, P.M. Gallagher,
and C.C. Smith. 1978. Residue organic mixtures from drinking
water show in vitro mutagenic and transforming activity. J.
Toxicol. Environ. Hlth. 4:919-938.
Meselson, M., and K. Russell. 1977. Comparisons of carcinogenic
and mutagenic potency. In: Origins of Human Cancer, Book C.
H.H. Hiatt, J.D. Watson, and J.A. Winsten, eds. Cold Spring
Harbor Laboratory: Cold Spring Harbor, NY. pp. 604-628.
Pereira, M.A. (in press). Rat liver foci bioassay. J. Environ.
Pathol . Toxicol.
Robinson, M. , J.W. Glass, D. Cmehil, R.J. Bull, and J.G. Orthoefer.
1980. Initiating and promoting activity of chemicals isolated
from drinking waters in the SENCAR mouse: a five-city survey.
Presented at the U.S. Environmental Protection Agency Second
Symposium on the Application of Short-term Bioassays in the
Fractionation and Analysis of Complex Environmental Mixtures,
Williamsburg, VA.
Scherer, E., and P. Emmelot. 1976. Kinetics of induction and
growth of enzyme-deficient islands involved in
hepatocarcinogenesis . Cancer Res. 36:2544-2554.
Shimkin, M.B., and G.D. Stoner. 1975. Lung tumors in mice:
Application to carcinogenesis bioassay. Adv. Cancer Res.
21:1-58.
Slaga, T.J., S.M. Fischer, L.L. Triplett, and S. Nesnow. (in
press). Comparison of complete carcinogenesis and tumor
initiation in mouse skin: Tumor initiation-promotion, a
reliable short-term assay. J. Environ. Pathol. Toxicol.
-------
ASSESSING IN VIVO CARCINOGENICITY
175
Stoner, G.D., and M.B. Shirakin. (in press). Strain A mouse lung
tumor bioassay. J. Environ, Pathol. Toxicol.
Syles, J.A. 1979. Cell trans format ion assays. In: Mutagenesis
in Sub-mammalian Systems. G.E. Paget, ed. Baltimore
University Press: Baltimore. pp. 53-71.
Weisburger, J.H., and G.M. Williams. 1977. Decision point
approach to carcinogen testing. In: Structural Correlates
of Carcinogenesis and Mutagenesis. HEW Publication No, (FDA)
78-1046. Rockville, MD. pp. 45-52.
-------
Intentionally Blank Page
-------
THE INITIATING AND PROMOTING ACTIVITY OF CHEMICALS ISOLATED
FROM DRINKING WATERS IN THE SEN CAR MOUSE: A FIVE-CITY
SURVEY
Merrel Robinson, John W. Glass, David Cmehil,
Richard J. Bull, and John G. Orthoefor
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
INTRODUCTION
Means of properly evaluating the carcinogenic risk posed by
organics in drinking water are of utmost concern to the U.S.
Environmental Protection Agency (SPA). While rodent lifetime
exposure and human epidemiological studies serve as the only
generally accepted means, the cost and time involved are highly
prohibitive. In addition, formulation of human epidemiological
data requires human exposure of sufficient magnitude to allow
separation of confounding factors from the relationship in
question. Since a major part of EPA's regulatory activities is
directed cowards preventing significant increases in the
carcinogenic risk to the population, quick and reliable
investigative methodology is a necessity. Short-terra bioassays can
be used to identify most potential problems and to provide an
initial risk assessment.
Loper et al. (1978) showed that all the organic material
isolated from the drinking water of six cities contained measurable
mutagenic activity in Salmonella tester strains. These cities were
selected to represent the most common types of drinking water
sources. Since a large number of chemicals that are known
carcinogens react positively in the Ames test, while noncarcinogens
do not (McCann et al. , 1975), these results suggest that chemical
carcinogens are to be found in most drinking water.
Mouse skin initiation/proraotion studies offer one means of
confirming Che presence of chemical carcinogens in complex
mixtures. A positive response in this assay system may be
classified as a true carcinogenic response, on the basis of
177
-------
178
MERREL ROBINSON ET AL.
evidence that strongly indicates a quantitative association of
benign papillomas with malignant tumors (3outwell, 1974: Burns et
al., 1976; Shubik et al . , 1953; Van Duuren et al. , 1973). The
system may therefore be used as a short-term _in vivo method of
assessing carcinogenic activity. Through the use of appropriate
experimental designs, the system allows the activity of tumor
initiators and tumor promoters to be clearly differentiated
(Barenblum, 1941; Hennings and Boutwell, 1970: Mufson et al., 1977:
Sivak and Van Duuren, 1971; Van Duuren, 1969).
The mouse skin bioassay was applied in our study to test both
the tumor-initiating and tumor-promoting potential of a complex
mixture of organic chemicals concentrated by reverse osmosis (RO)
from drinking water of five cities. These samples were obtained
from the same cities and processed by the same methods as employed
for Ames testing by Loper ec al. (1978).
The test samples for this study were concentrated from
drinking water supplies of Miami, Seattle, Philadelphia, Ottumwa
(IA), and New Orleans. These cities (Table 1) were selected to
represent both surface and ground types of water supply; sources
potentially contaminated by agricultural runoff, industrial wastes,
or municipal wastes; and uncontaminated sources (Tardiff and
Denizer, 1973) .
METHODS
Table 1. Cities Selected for Extraction and Bioassay
of Residual Organics in Drinking Water
City
Origin of
Water Supply
Type of
Water Supply
Miami, FL
New Orleans, LA
Otturawa, IA
Philadelphia, PA
Seattle, WA
Ground
Surface
Surface
Sur face
Surface
Uncontaminated3
Industrial wastes
Agricultural runoff
Municipal wastes
Uncontaminated3
aNo known contamination from municipal, agricultural, or industrial
wastes; however, contamination from decomposition products of
natural origin is possible.
-------
INITIATING/PROMOTING ACTIVITY OF DRINKING WATER CHEMICALS 179
The concentrated organics were prepared for EPA by Gulf South
Research Institute using the procedure described by Kopfler et al.
(1977). The samples were concentrated from multiple 200-1
quantities of tap water, which received sufficient concentrated
hydrogen chlorine (HC1) to maintain a pH of 5.5, by RO using a
cellulose acetate (CA) membrane. The reject from the CA membrane
was passed through a heat exchanger to maintain the water
temperature at < 15°C. Part of the reject stream was diverted
through a Donnan softening unit to avoid salt precipitation. The
CA permeate was adjusted to pH 10 and then concentrated by RO using
a nylon membrane, and the nylon permeate was discarded. Both the
CA and the nylon concentrates were then adjusted to neutral pH and
extracted with pentane and methylene chloride. The aqueous phases
were adjusted to pH < 2 with HC1 and again extracted with methylene
chloride. For the purposes of this study, these fractions were
combined and the solvent removed to produce the reverse osmosis
extract (ROE) sample. The residual aqueous concentrate was passed
through a column of purified XAD-2 resin. After removing metallic
oxides and other inorganic agents by elution with 1 M HC1 and
distilled water, the organics were then eluted from the column with
957> ethanol. The ethanol was removed from the eluate by vacuum
distillation, and the eluates from both columns were combined to
produce the XAD sample.
Once the samples were concentrated, they were administered to
mice subcutaneously (into the back). The following studies were
conducted using the mouse skin bioassay.
Drinking Water Concentrates as Initiators
Male SENCAR mice (sensitive to carcinogens) were obtained from
Dr. T.J. Slaga, of Oak Ridge National Laboratories, Oak Ridge, TN.
The mice were 8 to 10 weeks old when the study began. The ROE and
the XAD samples were administered over a two-week period in six
injections of 0.1 ml of a 10% Eraulphor (a polyoxyetheylated
vegetable oil), for a total dose of 4.5 mg/mouse (total dose = 150
mg/kg body weight). To maintain control over dosage and to allow
comparison of results with prior studies (Bull, 1980), the
subcutaneous route of administration was chosen. The 7,12-
dimethyl benz(a)anthracene (DMBA-positive control) was given in six
injections of 0.1 ml in 10% Emulphor, for a 25 ug/'nouse total dose.
The DMBA was obtained from Eastman Kodak Company (Rochester, NY)
and was purified by thin-layer chromatography by Dr. F. Bernard
Daniel of the EPA Health Effects Research Laboratory (Cincinnati,
OH). There were 60 animals in each exposure group.
Two weeks after the last initiating dose, the promoting phase
was begun. Forty mice of each group received 1.0 ug phorbol
myristate acetate (PMA) in 0.1 ml acetone applied to the shaved
-------
180
MERREL ROBINSON ET AL.
back three days a week for 20 weeks. The remaining 20 animals in
each group received only acetone. The ?MA was obtained from Dr.
Peter Borchert (University of Minnesota) and required no further
purification. The animals were weighed weekly and observed for
tumor incidence. The incidence of both papillomas and carcinomas
was charted weekly. Any of these that persisted for three weeks or
more were included in the cumulative count.
Following completion of the promotion period, the animals were
held for a total of one year for study and then were sacrificed.
Moribund animals were sacrificed as needed. Major organs and all
macroscopically evident lesions were sectioned and fixed in 10%
buffered formaldehyde solution for subsequent nistopathological
evaluation.
Drinking Water Concentrates as Promoters
The tumor-promoting potentials of the ROE and XAD samples were
also tested in the SENCAR mouse. Groups of 20 mice (for ROE) and
30 mice (for XAD) received an initiating dose of 2.56 yg DK3A in
0.1 acetone topically to the shaved area of the back. Two weeks
later, the promoting schedule with water concentrate samples was
begun. The ROE from each city was applied at a dose of 100 pg per
mouse per application in 0.1 ml acetone, three times a week for 18
weeks. The XAD dose was 500 yg/mouse in 0.1 ml acetone, three
times a week for 18 weeks. The only reason for the differing doses
was the availability of sample, which was much more limited for the
ROE. A positive control group received 1 yg PMA in 0.1 ml acetone
per application, following the same initiating dose of DMBA. After
completing the treatment, surviving animals were held for
observation of tumor incidence until they were one year old and
then sacrificed. Histological evaluation was done in the same
manner as in the tumor-initiating-potential study.
Drinking Water Concentrates as Complete Carcinogens
A third study was done with the concentrate samples to test
their potential as complete carcinogens. In groups the same sizes
as above, the ROE and XAD samples were administered topically at
dose levels of 100 yg/mouse and 500 yg/mouse, respectively, in 0.1
acetone, three times a week for 20 weeks. Thereafter the same
protocol was followed as in the other two studies.
-------
INITIATING/PROMOTING ACTIVITY OF DRINKING
WATER CHEMICALS 1S1
RESULTS
Initiating Activity
Table 2 presents the results at the end of 50 weeks of the
study testing the initiating activity of the water concentrate
samples. Positive results are apparent with several of the
samples. The data indicate significantly greater numbers of
papillomas per mouse in the animals treated with Ottumwa ROE (0.40)
and New Orleans XAD (0.33). Marginal responses occurred in Ottumwa
XAD (0.28) and Philadelphia XAD (0.25). Smaller response rates
were observed with Miami XAD (0.23), New Orleans ROE (0.18), and
Seattle XAD (0.20), compared with the vehicle control (0.10).
Essentially negative results were seen in Miami ROE (0.15),
Philadelphia ROE (0.10), and Seattle ROE (0.13). Using the
increase over the control response and applying a normalization
factor based on the amount of water processed to obtain samples,
the cities were ranked according to relative activity per unit
volume. The ranking was as follows: Miami (48), Ottumwa (34), New
Orleans (28), Philadelphia (26), and Seattle (7).
The time course of tumor development is presented in Figures 1
and 2. In the groups receiving ROE, no persistent papillomas
occurred after week 25, except in the Ottumwa group, where the
cumulative count continued to rise throughout the remainder of the
50-week period. More variation was seen in the groups receiving
the XAD; in all of these groups, tumor incidence tended to increase
with time relative to the control group. The New Orleans group
attained the highest level of tumors per animal within the last
five weeks of the study, and this change accounted for its
statistically significant difference.
All lesions that were observed grossly at the time of necropsy
were histologically examined; Table 3 gives the final distribution
of skin tumors, but not the total count, since some lesions meeting
the criteria later regressed or coalesced. The skin lesions were
predominantly papillomas. As their incidence was low, the low
incidence of squamous cell carcinomas observed in the experimental
group was not surprising. Interestingly, the group giving rise to
the most carcinomas (Seattle XAD) was negative by papilloma count.
However, due to limited numbers, the incidence of these tumors did
not correlate with papilloma incidence. This fact illustrates the
types of problems encountered when testing relatively small
quantities of complex mixtures where it is not possible to test at
levels approaching a maximally tolerated dose because of sample
preparation expense. Another problem was that a few fibrosarcomas
were observed that could have been injection-site related.
-------
Table 2. Cumulative Tumor Count at 50 Weeksa
Mice with Total Number Tumors Normalization Normalized
Sample Tumorof Tumors Per Mouse Factorc Activity^
M i ain i

ROE (
6

6
0.15
23
1 . 2
XAD
a

9
0.23
360
46.8
48.0
ROE
11

16
0.40
12
3.6
XAD
10

11
0.28
170
30.6
34.2
ROE
4

5
0. 13
14
0.4
XAD
«

10
0.25
1 70
25 .5
25.9
ROE
5

5
0.13
1.3
0.0
XAD
8

8
0.20
66
6.6
e'.f>
ROE
5

7
0. 18
10
0.8
XAD
8

13
0. 33
120
27.6
28.4
DMBA (25 pg)
38

286
7.15
-
-
Kmulphor
4

4
0. 10

—
aToLal dose of
150 tng/kg applied
suhc ut aneousIy
to each aniinal ,
foI lowed
by 20 weeks oi
promotion with
1 .0 pg
PMA three
times weekly.

^Out of a total
of 40 mice.

cAdjustment tor
amount
of water
processed to obtain concentrates
=

normalization factor = fraction wt x 10^
total recovered x volume processed.
•^Obtained by multiplying tumors per mouse minus conLrol incidence by the normalization factor
-------
INITIATING/PROMOTING ACTIVITY OF DRINKING WATER CHEMICALS 183
LSGENC

X X -
flO-fllA
-PITA
~ 0 "
TO-CTT
-•tia
© O -
30—PHI
-^HA
»—# -
RQ-SEA
-PT1A
A—& ~
RQ— NO
-fTIA
k A ~
ETIULPHCR
-PITA

m
P fff y' & ,.»•
gooa
k x x ii tie*
>¦¦¦««Utlt¥
y

V7
2^
T—
10
¥—
15
T-
20
-»—
33
35
~r~
40
30
TIME (WEEKS)
45
Figure 1. Tumor incidence through week 50 after animals received
an initiating dose of 150 mg/kg ROE sample s.c. and PMA
as a tumor promoter three times a week from week 0 to 20,
LESEHD

> * -
XAD-nlA
-PTA
O—0 -
xao-qtt
-ptia
iy—o -
XAD-PHI
-priA
~—* -
XAD-SEA
-PMA
A—A -
XAD-NO
-*MA
k—4 "
3ULPHQH
-PHA
—T—
19
30
~r—
40
—w—
40
10
S0
TIME (WEEKS)
Figure 2. Tumor incidence through week 50 after animals received
an initiating dose of 150 mg/kg XAD sample s.c. and PMA
as a tumor promoter three times a week from week 0 to 20.
-------
184

MERREL ROBINSON ET AL.
Table
3.
Summary of the Macroscopically Observed Lesions

Skin Tumors
Systemic Tumors
Sample Pap.a
Car.k Fib.Sa .c
Pul.Ad.Ca.^ Hem.(li.)e Hepat.^
Mia-ROE
4

Ott-ROE
8

3
Phi-ROE
2

Sea-ROE
2

3
N.0.-ROE
2
1
3
Mia-XAD
5

Ott-XAD
2
1 1
2
Phi-XAD
2
1
2
Sea-XAD

2
1 3
N.O.-XAD
8
1
1 1 4
aPapi1 loma.
^Carcinoma.
"-Fibrosarcoma.
^Pulmonary adenocarcinomas.
eHeraangioma (liver).
^Hepatoma.
The systemic tumors observed were distributed somewhat
unevenly among the groups. The most frequent kind were hepatomas.
Four were observed with New Orleans XAD, three each with Seattle
ROE and XAD, and two with Ottumwa and Philadelphia XAD. Animals
treated with the other four samples gave no evidence of hepatomas,
and only one animal in the control group gave evidence of
hepatomas. Pulmonary adenocarcinomas were also somewhat elevated
in New Orleans and Ottumwa ROE samples. Due to the relatively low
incidence of these tumors, the differences could not be considered
statistically significant. However, incidences in experimental
groups generally exceeded those observed in control animals.
Promoting Activity
The study to determine the promoting potential of the water
concentrate samples is continuing, and Table 4 shows the results
through week 38. The groups receiving 500 yg XAD sample per
application from Miami, New Orleans, and Ottumwa have yielded one
papilloma each. No papillomas have occurred in groups treated with
-------
INITIATING/PROMOTING ACTIVITY OF DRINKING WATER CHEMICALS 185
Table 4. Drinking Water Concentrates as Promoters: Number of
Papillomas Per Number of Micea
City
Sample Vehicle PMA Mia. N.O. Ott . Phil. Sea.
Controls 0/20 319/20
ROE (100 yg) 0/20 0/20 0/20 0/20 0/20
XAD (500 ug) 1/20 1/30 1/30 0/30 1/30
aInitiator equals DMBA (2.56 pg/mouse applied topically) . Drinking
water concentrate or PMA (1.0 yg in 0.1 ral acetone) applied
topically three times a week for 18 weeks. Results at 38 weeks.
Philadelphia and Seattle XAD, nor have papillomas occurred in any
of the ROE samples. In the positive control group, using 1 ug PMA
three times weekly, 19 of 20 mice had tumors, with a total of 319
papillomas. At this point, it appears that drinking water samples
at the doses applied do not promote DMBA tumorigenesis.
Complete Carcinogenic Activity
Table 5 gives the results after 38 weeks of studying the
potential of the ROE and XAD samples as complete carcinogens. Only
one papilloma has been observed, and that is in the New Orleans ROE
group. To this point in time, no evidence of complete
carcinogenesis of organic chemicals from drinking water has been
demonstrable. Again, the chemicals isolated from drinking waters
do not seem to be complete carcinogens at the doses applied.
DISCUSSION
Within the limits of possible dosage in the present work, it
appears that tumorigenic and/or carcinogenic substances were
present in the drinking waters. These chemicals were primarily
initiators in the mouse skin rather than promoters or complete
carcinogens. However, this conclusion must be clearly couched in
terms of the doses that were actually administered. Compared with
PMA, organic chemicals present in the ROE and XAD fractions were
less than 1/100 and 1/500 as potent as promoters, respectively.
-------
186
MERREL ROBINSON ET AL.
Table 5. Drinking Water Concentrates as Complete Carcinogens:
Number of Papillomas Per Number of Mice3
City
Sample Miami N.O. Ottum. Phila. Seattle
ROE (100 mg) 0/20 1/20 0/20 0/20 0/20
XAD (500 tag) 0/30 0/30 0/30 0/30 0/30
aDrinking water concentrate applied topically three times a week
for 20 weeks. Results at 38 weeks.
In view of the extreme potency of PMA, this was not altogether a
satisfying result. A further reservation is that no evidence
exists to indicate that mouse skin is a universal target tissue for
tumor promoters.
In the case of the Ottumwa ROE sample and all of the XAD
samples, tumor development was late. This contrast with the time
course of tumor development for the positive control DMBA (Figure
3) suggests that the chemicals in drinking water responsible for
initiating tumors may differ from DMBA with respect to underlying
raechanism(s). In terms of initiating activity, individual samples
produced significant increases in the number of tumors. At equal
doses of organic material, however, there was little to distinguish
positive from negative responses in the different samples. On the
other hand, if the data were adjusted to the amount of water
processed, the total units of activity present could vary among
water samples by a factor of seven. Although this calculation was
based on somewhat nonsignificant data, it does suggest that the
total risk observed might parallel the level of organic material
present. Undoubtedly, a wide variety of variables underlie this
parallel. For example, previous work has shown that disinfecting
drinking water can increase the levels of carcinogens isolated from
drinking water (Bull, 1980). However, the observation argues that
a prudent course of action in drinking water treatment might
involve reducing the total organic carbon present in the finished
drinking water. Although the data obtained in the present study
cannot be used to estimate risks to populations consuming these
drinking waters, it certainly justifies further research into
carcinogenic risks associated with drinking water.
-------
INITIATING/PROMOTING ACTIVITY OF DRINKING WATER CHEMICALS 187
B.&-
n
5 b-®-
2 3.0-
CL
tn
g 4-®-
^ 3.0-
2.0-
0.0"
»
40
-f-
43
—* »—
10 IS
—? T T
20 as 3[
TIME (WEEKS)
Figure 3. Tumor incidence through week 50 after animals received
an initiating dose of 25 yg DMBA s.c. and ?MA as a tumor
promoter three times a week from weeks 0 to 20.
REFERENCES
Barenblum, I. 1941. The cocarcinogenic actions of croton resin.
Cancer Res. 1:44-48.
Boutwell, R.K. 1974. The function and mechanism of promoters of
carcinogenesis. Critical Rev. Toxicol. 2:419-443.
Bull, R.J. 1980. Health effects of alternate disinfectants and
reaction products. J. Am. Water Works Assoc. 72:299-303.
Burns, F.J., M. Vanderlaan, A. Sivak, and R.E. Albert. 1976.
Regression kinetics of mouse skin papillomas. Cancer Res.
36:1422-1427 .
Hennings, H-, and R.K. Boutwell. 1970. Studies on the mechanism
of skin tumor promotion. Cancer Res. 30:312-320.
Kopfler, F.C., W.E. Coleman, R.G. Melton, R.G. Tardiff, S.C. Lynch,
and J.K. Smith. 1977. Extraction and identification of
organic micropollutants: Reverse osmosis method. Ann. N.Y.
Acad. Sci. 298:20-30.
-------
188
MERREL ROBINSON ET AL.
Loper, J.C., D.R. Lang, R.J. Schoeny, B.B. Richmond, P.M.
Gallagher, and C.C. Smith. 1978. Residue organic mixtures
from drinking water show in vitro mutagenic and transforming
activity. J. Toxicol. Environ. HIth. 4:919-938.
McCann, J., E. Choi, E. Yamasaki, and B. Ames. 1975. Detection of
carcinogens as mutagens in the Salmonella/microsome test:
assay of 300 chemicals. Proc. Natl. Acad. Sci. USA
72:5135-5139.
Mufson, R.A., R.C. Simsiman, and R.K. BouCwell. 1977. The effect
of the phorbol ester turaor promoters on the basal and
catecholamine-stimulated levels of cyclic adenosine 3 ' : 5 1 —
monophosphate in mouse skin and epidermis in vivo. Cancer
Res.*37:665-669.
Shubik, P., R. Baserga, and A.C. Ritchie. 1953. The life and
progression of induced skin tumors in mice. Brit. J. Cancer
7:342-351.
Sivak, A., and 3.L. Van Duuren. 1971. Cellular interactions of
phorbol myristate acetate in tumor promotion. Chem.-Biol.
Interact. 3:401-411.
Slaga, T.J., G.T. Bowden, and R.K. Boutwell. 1975. Acetic acid, a
potent stimulator of mouse epidermal macromolecular synthesis
and hyperplasia but with weak tumor-promoting ability. J.
Nat. Cancer Inst. Vol. 55. 4:983-987.
Tardiff, R.G., and M. Denizer. 1973. Toxicity of organic
compounds in drinking water. Water Quality Conference,
University of Illinois, Urbana-Champaign. J.S. Government
Printing Office 1974-657-053/1084. pp. 23-37.
Van Duuren, B.L., A. Sivak, A. Segal, I. Seidman, and C. Katz.
1973. Dose-response studies with a pure tumor-promoting
agent, phorbol myristate acetate. Cancer Res. 33:2166-2172.
Van Duuren, 3.L. 1969. Tumor-promoting agents in two-stage
carcinogenesis. Progress Exp. Tumor Res. 11:31-68.
-------
AQUEOUS EFFLUENT CONCENTRATION FOR APPLICATION TO BIOTEST
SYSTEMS
William D. Ross, William J. Hillan, Mark T. Wininger,
JoAnne Gridley, Lan Fong Lee, and Richard J. Hare
Monsanto Research Corporation
Dayton, Ohio
Shahbeg S. Sandhu
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION
Potential chemical mutagens in industrial effluents may be
present at concentrations below the detection limits of biotests
such as the Ames mutagenicity test. These chemicals may accumulate
in biological food chains. Many insecticides and other chemicals
are known to accumulate in living organisms where tissues act as
effective storage depots for toxic compounds (Loomis, 1978). This
effect is especially significant for human health when dilute
toxicants enter the human food chain, such as through seafoods.
Mollusks such as the oyster tend to accumulate toxicants, because
they filter-feed, which concentrates and magnifies the effects of
toxic materials. Because of this potential for bioaccumulation,
methods are needed to determine the bioactivity of low
concentrations of potential toxicants in industrial effluents.
The objective of the research program discussed in this paper
was to evaluate and compare three methodologies for concentrating
potential chemical mutagens in typical industrial effluents for
application to in vi tro biotest systems. For this study, the Ames
Salmone11a mutagenicity assay (Ames et al., 1975) was used. The
optimum concentration methodology would ideally meet the following
criteria:
1) concentration of relatively large quantities (> 3 1) of
aqueous sample;
2) concentration factors of > 200 times;
189
-------
190
WILLIAM D. ROSS ET AL.
3) little or no loss of volatile compounds;
4) efficient concentration and extraction of potential
t oxi cant s;
5) maintenance of high integrity of chemicals by preventing
artifact formation;
6) maintenance of the relative concentrations of all
compounds;
7) use of methods and reagents that are compatible with the
biotest system; and
8) maintenance of microbial sterility of the resulting
sample.
This paper describes the experimental approach and resulting
data for three concentration methods: adsorption using
macroreticular resins (XAD), freeze-drving (1vophi1ization), and
reverse osmosis (ultrafiltration). These methods were used to
concentrate added standard chemical compounds (i.e., potential
toxicants) in "typical" aqueous effluents for apolication to the
Ames mutagenicity test.
METHODOLOGY
Three methods were used to concentrate aqueous effluents for
application to in v i tro biotest systems: sorbent extraction,
1yophi1ization, and reverse osmosis. A schematic is presented in
F igure 1 .
Five-gallon samples of raw wastewater obtained from an
industrial plant served as typical standard effluent samples and
were used to evaluate each of the concentration methodologies.
To check for microbial contamination, aliquots of the neat
effluent samples were streaked with a sterile applicator onto both
Difco Bacto nutrient agar and Ames histidine-free bottom agar. The
plates were highly contaminated, indicating the need for filter
sterilization. The samples were sterilized bv drawing them through
a series of Millipore filters of 1.3-, 0.45-, and 0.2-um pore size.
The Ames Salmonella mutagenicity assay (Ames et al. , 1975) was used
to test the neat filtrate for mutagenicitv, using two histidine-
requiring strains, TA98 and TA100. The specific procedure for
analyzing the neat effluents used five concentrations in
triplicate: 0.01, 0.1, 0.5, 0.75, and 1.0 ml/plate. Each
concentration was tested with and without rat liver S-9 fractions
in the plate-incorporation test. Spot and toxicity tests were also
-------
AQUEOUS EFFLUENT CONCENTRATION
191
MITER
SltAlUZATION
siokst
Of NlAt
SAMPlf
*£S1 1OR
CONTAMtMNTS
HLH
AQUOuS
INDUSTRIAL '
tfFvLtNT
SAMPU
CONCENTRATION
MfTHODOLOCr
CHEMlCAt
'>AflACT(«IZ AflCH
PRIORITY
POLLUTANTS
TOTAL ORGANIC
CAfldOfc
flEVEBSf
OSMOSiS
TtaCN HLUR
ACSOiFTlON
OtSORrTICN
3I0IEST CF
CONCEKTRME
AM£S
-irOPHiliiATXN
PARTICULATE
DtSOR*HOs
WITH
SOLVtNT
9 lOTf ST
OF OCSORBED
COMPOUNDS
Figure 1. Schematic of concentration and biotesting of textile-
industry effluents.
performed. The positive controls used were 2-nitrofluorene (2NF),
sodium nitrite, 9-amino-acridine, benzo(a)uyrene (B[a]P), and
2-amino-anthracene, while tap water served as a negative control.
The concentrates from sorbent extraction and 1yophi1ization
were tested using six concentrations in which the samples were not
toxic: 0.1, 0.4, 2.0, 10, 30, and 100 gl/plate. If the samples
were found to be toxic at a designated concentration, a lower,
nontoxic concentration was used as the highest concentration.
Many effluent samples contain large amounts of particulate
material that impede flow through XAD resins, reverse osmosis
membranes, and sterilizing filters. The large particles (> 5 um)
in the effluent used in this study were removed by filtration.
Potential toxic and mutagenic compounds adsorbed to particles were
also removed. These compounds were assayed for bioactivitv. The
industrial effluent was filtered under vacuum through polyester
drain discs placed in series with 5-gm Teflon type LS Millipore
filters. Twenty-four-hour Soxhlet extractions were carried out on
filter blanks with four fresh polyester drain discs and six 5-um
Teflon type LS Millipore filters, using 200 ml of methvlene
chloride. A cellulose extraction thimble retained the filters.
The methylene chloride was reduced to 5 ml with a Kuderna-Danish
-------
192
WILLIAM D. ROSS ET AL.
concentrator over a steam bath. Five milliliters of dimethvl-
sulfoxide (DMSO) was added, and attempts were made to remove the
rest of the methylene chloride; however, approximately 6 ml of
organic extract remained, indicating that approximately 1 ml of
methylene chloride would not evaporate. The material was
transferred to a micro—Snyder apparatus for more efficient removal
of methylene chloride; however, the volume still was not reduced.
The presence of residual methylene chloride was confirmed by
infrared spectrophotometry.
Ames mutagenicity testing of the Soxhlet filter extract of
recovered particulates indicated no mutagenicity, but a slight
toxic response was found. A Soxhlet reagent blank was also tested
and gave a slight mutagenic response in the spot test and a toxic
response. A methylene chloride control gave both a mutagenic and
a toxic response. Although there was no clear mutagenic response
for the Soxhlet filter particulate extract, methylene chloride
should not have been used in processing samples for biotest systems
because of its potential for causing a false mutagenic response.
Acetone or other nonbioactive solvents are recommended for future
Soxhlet extractions.
Three known mutagenic materials were used in this evaluation:
acridine orange (AO), B(a)P, and 2NF. All are positive mutagens in
the Ames microbial test system. AO, a orecursor to some dyes which
might be found in textile effluents, is a highly colored compound.
AO was used first in all of the experiments because it could be
easily traced by visual methods. B(a)P is a chemical mutagen
commonly found in environmental samples. This compound requires
S-9 activation and also represents less-stable compounds. It is
light sensitive and presents some concentration problems. A
commonly used positive standard in the Ames test is 2NF, which does
not require activation. The selection of nonmutagenic
concentrations of standard compounds for spiking the neat effluent
sample was based on previous experimental data (unpublished). All
of the concentrations of positive mutagens added to the neat
effluent theoretically should be nutagenic when concentrated by at
least a factor of 200 times. The spiked concentrations of mutagens
prior to concentration are listed in Table 1.
The standard EPA method (EPA, 1977) was used for chemical
characterization of these selected priority organic compounds.
This procedure involves solvent extraction and gas chromatography/
mass spectrometric (GC/MS) analysis of the effluent. The GC/MS
system used was a Hewlett-Packard 5985A GC/MS/Data System. Total
organic carbon (TOC) analyses were performed on the neat standard
effluent samples and on the concentrates to determine efficiencies
of recovery of organic compounds. The TOC analysis was performed
on a Technicon II Autoanalyzer using the TOC cartridge at 550 nm;
the recorder read full scale at 200 ppm.
-------
AQUEOUS EFFLUENT CONCENTRATION
193
Table 1. Standard Mutagenic Compounds and Concentrations
Added to Effluent Samples
Compound
Concentrat ion
(mg/1)
Benzo(a)pyrene
Acridine orange
2-Nitrofluorene
0.17
7.60
0.17
Sorbent Extraction
Sorbent materials have been used primarily to remove organic
compounds from potable water whose matrix is relatively clean (Junk
et al., 1974; Loper et al. , 1978). More recent investigations
(Rappaport et al., 1979) have used macroreticular resins such as
Amberlite (Rohm and Haas) to remove potential mutagens from
effluents and wastewaters. Compared with drinking water,
industrial effluent samples are usually higher in particulates,
contain larger numbers of much higher concentrations of organics,
and have more extreme pH values. Processing waste effluents with
XAD resins presents additional problems not encountered in the
treatment of drinking water. Considerably more research is
required with the various sorbent materials to determine the optima
of parameters such as depth of bed, flow rates, desorption methods,
and solvents, as well as breakthrough limits. Such studies would
be highly complex because of the different interactions of each
chemical compound with the sorbent materials.
The adsorbent used in this study was Rohm and Haas Amberlite
XAD-2, a low polarity styrene-divinylbenzene copolymer possessing
the macroreticular characteristics necessary for high sorptive
capacity. Recovery efficiencies of about 80 organic compounds in
water have been reported (junk et al., 1974), The efficiencies
vary from 35 to 100%; on average, however, recovery efficiencies
are above 78%.
The initial XAD-2 (Applied Science) column was prepared by
washing the resin with a 50:50 mixture of methanol and water and
pouring off the excess solvent to leave a slurry of XAD-2. The
system consisted of a 300-ml burette with a plug of silanized glass
wool placed in the bottom to retain the particulate XAD resin. The
XAD-2 slurry was added to the column, and a glass wool plug was
placed on top of the slurry to prevent disruption of the XAD
particulates forming a column of XAD-2 resin 2 cm long by 1 cm in
-------
194
WILLIAM D. ROSS ET AL.
diameter. The resin bed was washed with 3- Co 30-ml aliauocs of
deionized water and maintained wet at all times.
Three liters of raw wastewater effluent was processed by
filtering the sample through a 5~im type LS Millipore filter to
remove coarse particulates. This filter became plugged after 200
ml of sample was processed, and a pre-filter consisting of a Uni-
pore polyester drain disc (Bio-Rad Laboratories) was olaced ahead
of the 5-pm Millipore filter. This modification allowed filtration
of 600 to 700 ml of sample before the system became plugged with
particulates. The 3-1 sample was subsequently processed in 600-to
700-ml aliquots. Filters were replaced whenever plugging occurred.
The filters with particulates were retained for solvent extraction
of organics. After about half of the effluent had been processed,
Che flow rate slowed considerably, and highly purified nitrogen
under pressure was applied to facilitate the filtering process.
The average flow rate was about 1.2 1/h.
Direct extraction with DMSO was evaluated as a means of
reducing experimental time by eliminating desorption with one
solvent followed by exchange with DMSO. Three milliliters of DMSO
were added to the XAD column, saturating the XAD resin. The
solvent was permitted to stand for 30 min. Then the DMSO solution
was drained into a sterile test tube.
Unconcentrated raw wastewater was filter-sterilized by passage
through a series of Millipore filters, as described earlier. Ames
mutagenicity and toxicity tests were also performed. The samole
was found Co be nonmutagenic and nontoxic. The DMSO-extracted
XAD-2 concentrate solution was then Cested Eor mutagenicity. The
concentrated effluent processed through the methanol/water-washed
XAD-2 columns indicated some Ames mutagenicity bioactivitv, but no
dose response. The XAD-2-processed control Cau water gave a
similar response.
A new XAD-2 column was prepared by the method of Junk et al.
(1974), whereby three solvents (methanol, acetonitrile, and
diethyl ether) are used to wash the resin in a Soxhlet extraction
apparatus. The XAD resin was refluxed with each solvent for 8 h
and then stored in methanol. A 3-1 tap water blank was Drocessed
with washed XAD-2 and then tested for mutagenicity. No mutagenicity
was found. Three liters of effluent were concentrated on the
washed XAD-2; again, no mutagenicity was found.
We concluded from this study that l) the neat effluent samole
and the 600-fold concentrate were nonmutagenic; 2) DMSO extracts
were marginally mutagenic if the XAD-2 was washed onlv with
mechanol/water; and 3) DMSO could be used directly as a desorbent
solvent if Che XAD-2 were washed properly (i.e., by the Junk
method).
-------
AQUEOUS EFFLUENT CONCENTRATION
195
Standard mutagens were added to the standard effluent in order
to determine the efficiency of recovery and to evaluate the
mutagenicity of the concentrate using the Ames test system. The
three standard mutagens were AO, B(a)P, and 2NF.
The zinc chloride salt of AO was made up by adding 22.8 mg AO
to 3 1 of particulate—free raw wastewater. This solution,
containing 170 ppb AO, was processed (concentrated bv adsorption)
by the procedure described previously. The flow rate through the
XAD averaged 18 ml/min. The AO content of the effluent was
measured by a colorimetric technique with an Aminco DW-2 dual
wavelength UV-Vis spectrophotometer scanning the range of 400 to
650 ran. The wavelength monitored was 433 nm. The adsorption
efficiency for AO was determined by comparing the concentration
prior to processing with that of the XAD-2 filtrate. The starting
material (neat effluent) contained 7.6 pg/ml, and the filtrate
contained 4.5 pg/ml, indicating a collection efficiency of 40.8%.
Desorption was achieved by adding 5 ml of DMSO for a residence time
of 30 min. A comparison of the amount of AO collected (9.3 mg)
with the desorbed amount in the DMSO (7.5 mg) indicated a
desorption efficiency of 80.6%. Mutagenicity was tested by
applying the unconcentrated effluent containing 7.6 mg/1 (7.6 Dpm)
AO to the Ames plate-incorporation test. No mutagenicity was
found. The final XAD-2/A0 extract, concentrated from 3 1 of
starting sample to 5 ml of DMSO (with a combined 32.9% recovery and
desorption efficiency), contained 1.5 mg/ml (1,500 pom). The total
concentration factor was 196. A dose response was found using TA98
with S-9.
Five hundred micrograms of 2NF was added to 3 1 of particulate
filtered sample effluent, making a concentration of 167 pe/1 (167
ppb). The XAD-2 was prepared, as described previously, with three
solvents in a Soxhlet system. Three liters of sample effluent
containing the 167 ppb 2NF was processed. The recovered 2NF was
desorbed with 5 ml of DMSO. Mutagenicity testing indicated no
response in the Ames test to the blank or to the neat,
unconcentrated sample, A positive dose response with TA98 and S-9
and a response to TA100 (no dose response) was found for the XAD-2
concent rat e.
Addition of 500 ug of B(a)P to 3 1 of neat effluent gave a
concentration of 167 ug/1 (167 ppb). The 3 1 of neat effluent was
concentrated by passing it through the XAD-2 column. Five
milliliters of DMSO was used to desorb the recovered B(a)P. The
blank sample and the neat sample containing 167 ppb B(a)P gave
negative responses in the Ames test. The concentrated 3(a)P sample
indicated bioactivity in tests performed on two different dates:
on 7/13 at 30 pl/plate with TA98 and S-9 (with no dose response)
and on 7/19 at 30 and 40 pl/plate (with no consistent dose
response). Recovery experiments were performed to determine
-------
196
WILLIAM D. ROSS ET AL.
extraction efficiency of B(a)P by XAD-2 from spiked effluent with
subsequent desorption with DMSO. Much care was taken to isolate
all samples containing B(a)P from light in order to eliminate
potential light-degradation problems. Five milligrams of B(a)P
was added to 3 1 of industrial effluent, and the sample was
processed through the XAD-2 column. Recovery efficiencies were
determined by measuring the B(a)P solutions with an Aminco DW-2
dual wavelength UV-Vis spectrophotometer. The aqueous permeate was
analyzed at 270 ntn by scanning the range of 230 to 410 nm. The UV
analysis showed that no B(a)P passed through the XAD-2, indicating
a possible 100% recovery of B(a)P. The column was desorbed with
5 ml of DMSO. The UV analysis indicated a desorption efficiencv of
45°= of the 3(a)P. This low desorption efficiencv may indicate a
permanent bonding of the B(a)P to the XAD-2 resin. The UV data
gave no evidence of degradation of the 3(a)P.
Lyophi1i zation
This freeze-drying approach to concentration is best applied
to effluents that contain water-soluble, nonvolatile, heat-labile
pollutants. Inorganic salts and biological compounds of large
molecular weight are retained by this method. Bieri et al. ( 1979)
have reported successful use of freeze-drving to remove water from
Chesapeake Bay samples for application to chemical characterization
tests. These researchers reported the potential loss of volatile
compounds below Cj^ hydrocarbons and problems with chemical
contamination from vacuum-pump oils. Van De Meent et al. (1977)
suggested that freeze-drving led to catalytic conversion of
alcohols to olefins. However, Bieri found no evidence of this
problem (Bieri et al., 1979). In the present study, contamination
in lyophilized samples was demonstrated. Consequently, all samples
had to be filter-sterilized before treatment or after concentration
prior to adding them to the Ames test system.
In this experiment, the lyophi1ization system was built around
a stainless steel drum manifold (Virtis Model 10-MR-ST). The other
components were a vacuum pump, a backup trap, a vacuum gauge, and
the sample-containing filter-seal flask. A vacuum of 0.133 mbar
was maintained over processing times of 48 to 50 h. The system
could run unattended during much of this time, including overnieht.
The backup trap retained much of the volatile material and could be
analyzed for volatile compounds.
Three liters of neat effluent sample was processed to drvness,
leaving 0.8 g of a dry white residue. An inorganic chemical
characterization analysis (EDAX) indicated the presence of sulfur,
silicon, potassium, calcium, and iron. TOC analysis of the residue
using the Technicon II Autoanalvzer, was 6.3%. This drv residue
was dissolved in 10 ml of 50:50 DMSO:water solution. A
-------
AQUEOUS EFFLUENT CONCENTRATION
197
concentration factor of 300 times was achieved. Problems were
encountered in completely dissolving the residue. A "dissolved"
sample was applied to the Ames mutagenicity assay using the plate-
incorporation test and indicated no mut agen i c i t y in the Ivophilizeci
neat effluent. Microbial contamination was encountered, but not
enough to prevent a valid test. Filter sterilization is
recommended prior to lyophi1ization.
AO was added at a concentration of 0,0076 nig/ml to 3 1 of
effluent sample. The sample was lyophilized in approximately 55 h.
Using the colorimetric method, 96,3% of the AO was retained in the
lyophilized sample. The freeze-dried powder was dissolved in a
50:50 mixture of sterile distilled water and DMSO and applied to
the Ames mutagenicity assay. A two-point dose response was
obtained at 10 ul and 2 ul with TA98 and S-9 activation. Microbial
contamination was so high that the plates had to be hand counted;
lower concentrations could not be counted because of the
c ont aminat ion.
Reverse Osmosis
Reverse osmosis (RO) has become a versatile separation and
purification method. This process physically separates contaminant
from water by circulating the aqueous solution at high pressures
over the surface of a semipermeable membrane. Two factors
influence the concentration of contaminants: the physical and
chemical properties of the contaminants and the properties of the
membrane. Recent RO technology has improved rapidly as membrane
technology has advanced. Much of the development and applied
research has been directed toward the purification of aqueous
effluents, whereas this study is concerned with the concentrate.
The RO system used in this investigation was manufactured by
Abcor, Inc. (Cambridge, MA). The system is a bench-type static RO
and ultrafiltration test cell designed specifically for studying
membrane selectivity in a laboratory situation. The RO unit has a
200-ml capacity, maximum operating pressure of 105 kg/cm (eauge),
and a membrane diameter of 3 in. (7.6 cm). It is 6.25 in. high by
3,75 in. in diameter (15.9 x 9.5 cm) and weighs 6 lbs (2.7 kg).
The stainless steel unit has a Teflon-coated magnetic stirring bar
for constant agitation at the membrane surface and uses a SEPA-97
(Osmonics) cellulose acetate membrane of nominal 5-& (0.5 nm) pore
size.
Three liters of raw wastewater was filtered to remove
particulates to prevent clogging of the RO membrane. A I'nipore
polyester drain disc (Bio-Rad Cat. No. 334-0659) was placed over a
type LS Millipore filter with a 5.0-ym pore size. The filtered
effluent was spiked with AO to obtain a concentration of
-------
198
WILLIAM D. ROSS ET AL.
0.0076 mg/ml. The spiked effluent was processed bv RO. A
processing time of about 48 h was required to reduce the samole to
168 ml. Problems were encountered with a much-reduced flow rate
toward the end of the processing experiment. The RO unit was
dismantled and a large buildup of what appeared to be inorganic
materials highly colored with AO was found on the membrane,
impeding flow through the membrane. The retained solution
apparently became saturated with both dissolved salts and AO as the
concentration increased; this phenomenon will limit the
concentrat ion factor attainable by RO. The efficiency of
concentration was tested colorimetricallv: the concentration of AO
in the starting unprocessed spiked sample was 7.6 mg/ml and in the
processed concentrate 7.1 mg/ml, A conductivity measurement of the
filtrate (EP Meter, Myron L Co.), reduced from 3 1 to 200 ml,
indicated that 77% of the conductivity was removed by the process.
This result is an obvious limitation of the RO system.
CONCLUSIONS
This investigation has demonstrated that an industrial
effluent spiked with subtoxic amounts of standard mutagens (AO,
2NF, and B[a]P) can be concentrated by XAD-2 resin. The work also
showed that AO could be concentrated by lyophi1ization and RO. The
other two mutagens were not tested with the latter two techniques.
The concentrated spiked effluent exhibited positive dose responses
in the Ames Salmone1 la biotest. The RO techniaue was found to
retain 84% of the TOC. However, only a tenfold concentration was
achieved, because of clogging of the membrane by precipitated
compounds .
This evaluation also showed that the adsorbent concentration
procedure has the following advantages over lvophi1ization and RO
techniques: low cost of equipment, short processing time, ease of
portability, and potential for selectivity for specific chemical
classes by resins. A disadvantage of this approach is the need for
a desorption solvent, which increases the chances of sample
alteration, chemical contamination, loss of sample, and possible
lack of compatibility with biotest systems.
The lyophi1ization procedure has the advantage of good
recovery of inorganic components and relatively high organic
compound recovery. However, it requires much processing time, and
the equipment costs are high.
Reverse osmosis also concentrates inorganic components. It is
portable and has the potential advantage of recovering specific
chemical classes of compounds by use of selective membranes.
-------
AQUEOUS EFFLUENT CONCENTRATION
199
ACKNOWLEDGMENTS
This report was supported by the U.S. Environmental Protection
Agency, contract no. 68-02-1874.
REFERENCES
Ames, B., J. McCann, and E. Yamasaki. 1975. Methods for detecting
carcinogens and mutagens with the Salmonella/mamma 1i an-
microsome mutagenicity test. Mutation Res. 31:347-364.
Bieri, R.N., M.K. Cueman, R.J. Huggett, W. Maclntyre, P. Shou, C.W.
Su, and G. Ho. 1979. Investigation of organic pollutants in
the Chesapeake 3ay: Report #1, Grant R806012010. U.S.
Environmental Protection Agency: Annapolis.
EPA, U.S. Environmental Protection Agency. 1977. Sampling and
analysis procedures for the screening of industrial effluents
for priority pollutants. U.S. Environmental Protection
Agency: Cincinnati, OH.
Junk, G.A., J.J. Richard, M.D. Grieser, D. Witiak, J.L. Witiak,
M.D. Arguello, R. Vick, H.J. Svec , J.S. Fritz, and G.V.
Calder. 1974. Use of macroreticular resins in the analysis
of water for trace organic contaminants. J. Chromatogr.
99:745-762.
Loomis , T.A. 1978 . Essentials of Toxicology. Lea and Febiges:
Philadelphia.
Loper, J.C., and D.R. Lang. 1978. Mutagenic, carcinogenic, and
toxic effects of residual organics in drinking water. In:
Application of Short-Term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures. M.D. Waters, S.
Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
Plenum Press: New York. pp. 513-528.
Rappaport , S.M., M.G. Richard, M.C. Hollstein, and R.E. Talcott.
1979. Mutagenic activity in organic wastewater concentrates.
Environ. Sci. Technol. 13:957-961.
Van De Meent, D., W.L. Maters, J.W. Peheew, and P.A. Schenck.
1977. Formation of artifacts in sediments upon freeze-
drying. Org. Geochem. 1:7-9.
-------
Intentionally Blank Page
-------
SESSION 3
TERRESTRIAL SYSTEMS
-------
Intentionally Blank Page
-------
POTENTIAL UTILITY OF PLANT TEST SYSTEMS FOR ENVIRONMENTAL
MONITORING: AN OVERVIEW
Shahbeg Sandhu
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
Research over the past decade has shown a significant
proportion of genetic diseases in man to be caused by natural and
man-made chemical mutagens. Human cancer is one of the diseases
for which direct associations with certain environmental factors
have been established (such as the association between cigarette
smoking and lung carcinoma). The recently published "Atlas of
Cancer Mortality" (Mason et al., 1975) provides further evidence
for the association between human cancer and environmental factors.
In this study, certain specific types of cancer appear to be
associated with certain industrial activities. Chemical mutagens
are also believed to contribute to birth defects, behavioral
abnormalities, and aging. It has been suggested that environmental
chemicals play a role in causing atherosclerosis (Benditt, 1973)
and, most damaging of all, in deteriorating the human gene pool.
Awareness of the role of environmental chemicals as human
health hazards has led to the development and use of various
techniques for identifying potentially toxic chemicals and, if
possible, eliminating exposure to them. These methodologies
include several nontraditional types of short-term bioassays. A
variety of test systems, ranging from the use of viruses to that of
circulating human lymphocytes and sperm cells, have been used to
identify potentially harmful chemicals or mixtures of chemicals.
Over 100 different assays have been developed for detecting toxic
chemicals (Hollstein et al., 1980).
Despite their historical role in the formulation of principles
of genetics and genetic toxicology, plant test systems have not
been employed in the recent rapid advances in the development of
genotoxin-detection technology. This general neglect may perhaps
203
-------
204
SHAHBEG 5ANDHU
be attributed to the apparently great phylogenetic distance of
plants from animals. However, the use of plants as test organisms
for detecting the genetic effects of individual chemicals or as
monitors of environmental quality offers the following advantages:
1) Plants, like animals, are eukaryotes; thus, their
organization and cell structure are similar.
2) Plants undergo mitosis and meiosis, thus making it
possible to evaluate somatic and germ cell mutations and
their transmission to future generations. This caDacitv
is especially important when we consider that most of the
short-term bioassays use bacteria or mammalian cells in
culture, which do not undergo meiotic cell division.
3) Plants can easily be propagated from vegetative tissue or
even from a single cell. Thus any variant line can be
genetically characterized.
4) Plants show a wide array of genetic endpoints, including
gene mutation, DNA repair, primary DNA damage, and
chromosomal aberrations. In certain plant species (e.g.,
barley and Arabidopsis) , multiple-locus forward-mutation
test systems have been developed that may be particularly
relevant to human genetic systems. Using Arabidopsis, it
is already feasible to monitor one hundred loci at once.
This attribute is significant when we consider that
different chemicals elicit qualitatively and
quantitatively different responses at different loci in
the same genome and that most of the commonly used
short-term bioassays monitor genetic alterations at only
one or two loci.
5) Plants are relatively easy and inexpensive to work with.
Furthermore, mutational events are very easily scored by
technicians.
6) Perhaps the most significant attribute of plants as test
systems is their suitability for in situ monitoring.
For the last three years, the U.S. Environmental Protection
Agency (EPA) and the National Institute of Environmental Health
Sciences (NIEHS) have made concerted efforts to develop and use
plant bioassays for environmental monitoring. There are two areas
in which plant systems show promise for immediate use: 1) as part
of a short-term first- or second-level laboratory test batterv for
evaluating the mutagenicity of specific environmental chemicals or
chemical mixtures, and 2) in field monitoring studies, as
indicators of the mutagenicity of the total environment.
-------
PLANT TEST SYSTEMS FOR ENVIRONMENTAL MONITORING 205
Possible Role of Plants for Mutagenicity Evaluation in the
Laboratory

A number of articles have been published emphasizing the need
for short-term bioassays (see Hollstein ec al. , 1979, for a review)
and for their integrated use with in vivo animal bioassavs in
identifying environmental chemicals and estimating risk from
exposure to thera (Waters et al., 1980). An extensive data base is
not yet available for comparing the genetic responses to chemicals
of plant test systems with those of other in vitro and in vivo
test systems. However, preliminary comparisons baser! on
of the literature show a fairly good concordance. The
potencies of eight chemicals in various in vitro and in
st systems were compiled by Clive and Soector (1978). The
Table 1 show that mutation responses in plants correlate
th those in mammalian systems.
Comparative Mutational Potencies of Eight Chemicals in
Bacteria, Plants, Insects, and Mammals3
Mammals
Chemical
Bacteria
Pi ant s
Insects
in vivo
in vitro
Trenimon
2
1
4
1
4
Mitomycin C
4
2
1
1
2
MNNG
5
3
7
5
2
Triethylene-

melamine
8
4
8
3
5
E t hylenemel amine
3
5
3
4
8
Ethylmethane

sulfonate
7
6
6
7
3
Methylmethane

sulfonate
1
7
2
8
7
Dimethyl-

nitrosamine
6
8
5
6
6
aData from Clive and Spector (1978).
Redei et al. (1980) have compiled data from the literature on
the mutagenic response of the multilocus Arabidopsis test system to
a number of known animal carcinogens. The list includes several
compounds that require metabolic activation to produce genetic
animal
review
genetic
vivo te
data in
well wi
Table 1
-------
206
SHAHBEG SANDHU
effects. This report shows an 84% correlation of genetic responses
in Arabidopsis with those in animal tests. It also shows that
Arabidopsis can provide the enzymes needed to transform prorautagens
into reactive metabolites. Because it is a multiple locus system
and because Arabidopsis has a short generation time and can grow on
a variety of inedia, this test seems especially well suited for
further development for screening environmental chemicals in the
laboratory. This is the only existing short-term bioassay
for analyzing the genetic effects in progeny of environmental
exposure of the parents. The EPA, in a cooperative effort with the
University of Missouri, is currently validating this assay.
Nearly every biology student first visually encounters
chromosomes in onion or broad bean root tips—the chromosomes from
these materials are large and easy to manipulate. A few chemical
pesticides have been tested for their ability to induce chromosomal
aberrations in plant root tips; Table 2 compares the clastogenic
response to these pesticides in plant root tips with that in
mammalian cells in culture.
Table 2. A Comparison of Responses by Plant Root Tip Cells and
Mammalian Cells in Culture to Pesticides3
Chromosome Aberrations
Compound
Plant Root Tips
Mammalian Cells
in Culture
Apholate
Atrazine
2 ,4-D
DDT
Dichlorvos
Di eldrin
Ethylene dibroraide
Griseofulvin
Hempa
Heptachlor
Maleic hydrazide
Mercury compounds
Phosphamidon
2,4,5-T
Tepa
+
+
aData from W. ¥~. Grant (1978) .
-------
PLANT TEST SYSTEMS FOR ENVIRONMENTAL MONITORING 207
In general, the positive or negative responses to these
pesticides are very similar in plant root tips and mammalian cells
in culture. The only exception seems to be raaleic hydrazide. It
has recently been observed by Dr. Michael Plewa (University of
Illinois, personal communication, 1980) that although animals
cannot transform this compound to genetically reactive metabolites,
plants (at least Zea mays) do convert it into biologically active
forms.
The Health Effects Research Laboratory of aPA at Research
Triangle Park, NC, is in the process of evaluating the Vicia faba
root tip assay for possible inclusion in the level one test
battery. Several pesticides for which we have fairly extensive
data will be tested for chromosomal effects in this assay.
The presentation by Dr. Constantin in this symposium
(Constantin et al., 1980) further illustrates the utility of plant
cytogenetics assays in concert with microbial mutagenicity assays
for evaluating the potential health effects of complex
environmental mixtures. With few exceptions, plant bioassays do
not have as well-defined gene markers or genetically engineered
tester strains as are found in microbial bioassays. On the other
hand, very few microbial assays can be used to evaluate the
chromosomal effects of exposure to environmental chemicals.
The point of this discussion
profitably used for toxicological
chemicals. These assays will not
systems in the foreseeable future
complementary information.
is Chat plant bioassays could be
evaluation of environmental
be able to replace microbial test
, but will be useful in furnishing
The Role of Plants for In Situ Environmental Monitoring
Perhaps the most useful testing application for plants in the
future will be in monitoring the mutagenicity of the total
environment. Environmental chemicals exert their effects not in
isolation but in concert with other chemicals and environmental
factors. This environmental milieu is impossible to reproduce in
the laboratory. By growing experimental plants in the ground at
the site to be tested, one can evaluate the multimedia exposure
effects. Several plant systems have shown a great deal of promise
for in situ environmental monitoring. The Tradescantia stamen hair
assay (Schairer et al., 1978) has been used to monitor ambient air
quality at several industrial sites in the U.S. The waxy pollen
maize assay (Plewa, 1978) has been developed and applied to detect
the mutagenicity of agricultural chemicals. Klekowski (1978) has
developed a very useful bioassay for monitoring mutagens in
effluent streams, rivers, and lakes. An excellent review of these
test systems has been edited by de Serres (1978).
-------
208
SHAHBEG SANDHU
In Che present symposium, Dr. Ma describes the potential use
of a newly developed bioassay for monitoring complex environmental
mixtures (Ma et al., 1980). The Tradescantia micronucleus assay
appears to be even more sensitive than the Tradescantia stamen hair
assay. This assay is under further validation with financial
support provided by EPA.
Another assay, developed by Dr. Vig, of the University of
Nevada (Vig, 1980), uses a soybean system for measuring point
mutations and chromosomal aberrations in somatic cells. By using a
series of promutagens (compounds that require mammalian metabolic
activation enzymes for biotransformation to express their genetic
potential), Dr. Vig has shown that plants have the ability to
activate these compounds to mutagenic levels.
In the past, the lack of concurrent controls has caused some
difficulties in interpreting data from in situ bioassays. Data
from historical controls cannot be used as a substitute for on-site
controls. In situ monitoring with plant test systems is not
intended to take the place of more rigorous testing to evaluate the
health hazards of exposure to a particular environment. None of
the in situ plant bioassays have reached a stage of development
where they could be used to identify specific mutagenic compounds
from the environment. Their main utility so far appears to be in
raising a "red flag," so that priorities can be set for applying
more specific bioassays and chemical analysis to track down the
sources of toxic chemicals.
REFERENCES
Benditt, E., and F. Benditt. 1973. Evidence for a monoclonal
origin of human atherosclerotic plaques. Proc. Nat. Acad.
Sci. US 70:1753.
Constantin, M.J., K. Lowe, T.K. Rao, F.W. Larimer, and J.L. Epler.
1980. The detection of potential genetic hazards in complex
environmental mixtures using plant cytogenetics and microbial
mutagenesis assays. Presented at the U.S. Environmental
Protection Agency Second Symposium on the Application of
Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures, Williamsburg, VA.
Clive, D., and J.F.S. Spector. 1978. Comparative chemical
mutagenesis: an overview. In: Proceedings Comparative
Chemical Mutagenesis Workshop. F.J. de Serres, ed. National
Institute of Environmental Health Sciences: Research
Triangle Park, NC.
-------
PLANT TEST SYSTEMS FOR ENVIRONMENTAL MONITORING
209
de Serres, F.J. 1978. Introduccion: utilization of higher plant
systems as monitors of environmental mutagens. Environ.
Hlt'n. Perspect. 27:3-6.
Grant, W.F. 1978. Chromosome aberrations in plants as a
monitoring system. Environ. Hlth. Perspect. 27:37-43.
Hollstein, M., J. McCann, F. Angelosanto, and W. Nichols. 1979.
Short-term tests for carcinogens and mutagens. Mutation Res.
65:133-226.
Klekowski, E. 1978. Screening aquatic ecosystems for mutagens
with fern bioassays. Environ. Hlth. Perspect. 27:99-102.
Ma, T.-H., V. Anderson, and S. Sandhu. 1980. A preliminary study
of the clastogenic effects of diesel exhaust fumes using the
Tradescantia micronucleus assay. Presented at the U.S.
Environmental Protection Agency Second Symposium on the
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, Williamsburg, VA.
Mason, T.J., F.W. McKay, R. Hoover, W.J. Blot, and J.F. Fraumeni,
Jr. 1975. Atlas of Cancer Mortality for U.S. Counties:
1950-1969. Dept. of Health, Education, and Welfare
publication no. 75-780. National Institutes of Health:
Washington, DC.
Plewa, M.J. 1978. Activation of chemicals into mutagens by green
plants: a preliminary discussion. Environ. Hlth. Perspect.
2 7:45-50 .
Redei, G.P., M.M. Redei, W.R. Lower, and S.S. Sandhu. 1980.
Idendification of carcinogens by mutagenicity for Arabi dopsi s.
Mutation Res. 74:469-475.
Schairer, L.A., J. Van't Hof, C.G. Hayes, R.M. Burton, and F.J.
de Serres. 1978. Measurements of biological activity of
ambient air mixtures using a mobile laboratory for in situ
exposures: preliminary results from the Tradescantia plant
test system. Application of Short-term Bioassays in the
Fractionation and Analysis of Complex Environmental Mixtures.
M.D. Waters, S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L.
Claxton, eds. Plenum Press: New York. pp. 419-440.
Vig, B.K. 1980. Soybean system for testing the genetic effects of
industrial emissions and liquid effluents. Presented at the
U.S. Environmental Protection Agency Second Symposium on the
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, Williamsburg, VA.
-------
210
SHAHBEG SANDHU
Waters, M.D., V. Simmon, A.D. Mitchell, T.A. Jorgenson, and R.
Valencia. 1980. An overview of short-term tests for
Che mutagenic and carcinogenic potential of Desticides.
J. Environ. Sci. Hlth. B.Pesticid. Food Contam. Agric.
Wastes B15:867-906.
-------
ARABIDOPSIS ASSAY OF ENVIRONMENTAL MUTAGENS
G.P. Redei
Department of Agronomy
University of Missouri
Columbia, Missouri
INTRODUCTION
In the past, Arabidopsis assays have been employed for testing
the mutagenic effects of a variety of chemicals. Although it is
potentially useful for determining mutagenic hazards of complex
mixtures, this species has not been much used for this purpose.
An Arabidopsis assay system was initiated In the early 1960's
by Andreas Mviller ( 1963, 1965b). The suggested procedure was based
on the principles first exploited by Gregor Mendel in his famous
pea experiments. In the autogamous species of pea, within
individual flowers, segregation of the alleles at raeiosis and the
random combination of the gametes at fertilization resulted in the
reappearance of both dominant and recessive phenocypes of the
parents among the embryos developed on the Fj plants. Therefore
when Mendel crossed two different varieties of peas (Yellow vs.
green and Smooth vs. wrinkled cotyledons), a study of only the F|
plants, bearing the F2 seeds, was frequently satisfactory for his
genetic analyses. Thus, he saved considerable time and labor,
enabling him to make more comprehensive studies.
But heterozygosity within the nucleus of a cell can also arise
by mutation. Mutation from a dominant to a recessive allele in the
diploid cells is concealed. If the mutation takes place early in
the diploid germline, and the heterozygous cell gives rise to a
sector encompassing both the pollen-producing (androeciura) and the
egg-producing (gynoecium) lineages, segregation may become evident
among the embryos of the same individual. Since the germllne is
commonly multicellular, the Fj plant may be chimeric after a
211
-------
212
G. P. REDEI
mutational event. This is more of an advantage than a disadvantage
because it permits the screening of a large population at a low
cost.
Such an analysis is impractical, however, in monoecious
plants, such as maize, because the silks and the tassel may not
differentiate concordantly (from the same cell lineage).
Obviously, dioecious plants and the majority of bisexual animals
are not amenable to such an analysis.
Though a large number of assay systems are available for
mutagens, none of them suits entirely the needs for testing all
hazardous compounds. The most efficient and the most widely used
microbial assays involve the detection of revertants at a specific
locus. The mutability of individual loci may vary by one or more
orders of magnitude (Table 1). Furthermore, the reversion assay
uses special alleles, one or another at a time. The mutability of
these special sites depends to a great extent on the nature of the
inducing agent (Table 2). Therefore it is not easy to draw firm
conclusions as to how complex mixtures of mutagens or even pure
compounds would affect other genes. Even worse, the mutability of
many important human genes cannot be directly measured, and the
inferences based on indirect methods are also somewhat tenuous.
Approximations based on mammalian assays, such as the mouse
specific locus assay (which is probably the best), are not much
better either, because we do not know for sure whether the six or
seven loci used represent accurately all the loci of mice (or those
of man).
PURPOSE
Table 1. Rates of Spontaneous Mutation3
Organism
Mutant Phenotype
Rate
E. coli
Streptomycin resistance
Histidine auxotrophy
4 x 10"1*
2 x 10-6
Neurospora
Inositol independence
Adenine independence
2 x 10~5
4 x 10"6
Drosophila
Yellow body
White eye
1 x 10"4
3 x 10"5
aRedei, 1980, p. 576.
-------
ARABIDOPSIS ASSAY OF ENVIRONMENTAL MUTAGENS
213
Table 2.
Site
Specificity of Induced Mutations
in Yeast3

Revertants per 10^ Cells

Sites
EMSb
NMGC UVd
Y-rays
c vl-131
1226
1775 41
18
cyl-133
4
1 83
8
cyl-9
8
8 2430
19
aAbridged from Prakash and Sherman, 1973.
bEthylraethane sulfonate.
cNi t rosomethylguan id ine.
^Ultraviolet.
There is obviously a need for imorovement in the testing
systems; we must search for approaches capable of detecting a
variety of genetic alterations (gene mutations and chromosomal
effects) at as large a number of positions in the genome as pos-
sible. We must also strive to learn much more about the capabili-
ties of metabolism in activating or detoxifying the potentially
hazardous chemical compounds present in the human environment.
The tests must be relatively fast, reliable, reproducible, and
inexpensive, and the information obtained should be applicable for
predicting human hazards. The Arabidopsis assay outlined here
appears quite valuable for meeting many of these goals.
THE ARABIDOPSIS ASSAY
Culture of the Plants
Arabidopsis thaliana (L.) Heynh. is a plant of the family
Cruc i ferae (mustards), with a haploid chromosome number of five.
The early genotypes may produce eight generations a year in the
laboratory, where the life cycle can be hastened by continuous
illumination (long-day plant). Under short daily light regimes,
the vegetative period is quite long, and only a few generations can
be grown in a year. Under such conditions, the plants grow much
larger and may produce 50,000 seeds each.
The plants can be grown in pots on any good soil or other
media. In the greenhouse, we culture then on soil in 5-in
(12.7-cra) pots, each with several plants. In growth chambers, we
have successfully raised 200 or more plants on Pro-Mix medium
(Premier Brands) to maturity in petri plates 10 cm in diameter.
-------
214
G. P. REDEI
Before planting, the medium was moistened with a nutrient solution
of the following composition:
After germination, the lid was removed, and the lost moisture was
replaced by distilled water as needed (at least two times a day).
In soil culture, supplemental nutrients are not needed. In any
case, the seeds should not be covered after planting, because
Arabidopsis requires light for germination.
When large quantities of seeds need to be planted, time may be
saved by suspending the seeds in a fluid yet viscous agar solution
(ca. 0.15%) and spreading thera dropwise with the aid of a
separatory funnel or a pipette. After planting, the seeds must not
be allowed to dry for any length of time. To avoid washing them to
the rim of the container, watering is best done initially with a
fine mist. Later, after germination, the pots can be placed in a
tray containing distilled water. In the growth chamber, the
intensity of the light should not exceed 800 foot-candles (8608
lux), and a constant temperature of about 24°C is satisfactory.
The seeds of Arabidopsis loose their germination ability
within two to three years at higher temperatures and humidity.
Therefore, it is advisable to store the seeds under a regime where
the sum of the temperature in degrees Fahrenheit arid the relative
humidity is below 100; the lower this figure, the longer the
viability.
Mutagen Assay
Generally, for Laboratory assays, seeds are exposed to
mutagens. Though the mature embryo contains about 6000 to 7000
cells, only two of the cells represent the diploid germline at this
stage. As the seeds germinate and the seedlings develop, the
number of cells in the germline increases. The consequences of
mutation at two different stages of the growth of the germline are
diagrammed in Figure 1.
Either the mutants can be detected at the embryo stage (Figure
2) or the M2 generation can be planted and classified for seedling
and plant traits. The most convenient method of assessing the
mutagenic effects is to determine the frequency of mutational
events expressed at the embryo stage. For this purpose the fruits
ammonium nitrate (NH3)
magnesium sulfate (MgS04'7H20)
calcium phosphate, monobasic (CaH^(PO4)2 *H2O)
potassium phosphate, monobasic (KH2PO4)
potassium phosphate, dibasic (K2HPO4)
ferric i trate
200 mg/1
100 mg/1
100 mg/1
100 mg/1
50 mg/1
2.5 mg/1
-------
ARABIDOPSIS ASSAY OF ENVIRONMENTAL MUTAGENS
21
«1
It
THE FRUITS REPRESENT THE CELL LINEAGES 0C
EIGHT CELLS IN THE GERMLINE AT THE 5TAC.E OF
the mutagenic exposure
eRUIT5 BORN BY HETEROZYGOUS
SEC'CR SEGREGATE FOR CCTYL"
EDCN COLOR VISIBLE THROUGH
TRANSPARENT TESTA WHEN IM-
MATURE FRUITS ARE DISSECTED
ONE OF THE EIGHT CELLS OF THF
~ germline is heterozygous
because of a mutation
plant
IN THE GE.RmlINE the
GENETICALLY effective
CELL NLMBER IS TWO; /—
ONE IS HETEROZYGOUS (•
BECAUSE OF MUTATION
CONTAIN I Nf, THE
GERMLINE

AT THE SEEDLI'IC. STAr,
ber of the 'hnetical
MUTAGENIC EXPOSURE
INCREASED the num-
THE DRY SEED STAGE /-«
MATURE EMBRYO IN DRY SEED
IS EXPOSED TO MUTAGEN
Figure 1. Consequences of mutagenic treatment at stages when the
germline consists of two diploid cells (left) and when
it has increased to eight (right). Explanations are on
the diagram.
Figure 2. An Arab i dops i s fruit, showing segregation for the
color of the embryos.
-------
216
G. P. REDEI
are opened with sharp forceps before the seed coat turns brown and
opaque. Through the immature seed coat, the green cotyledons of
the normal embryos can be seen. Many types of mutants have white,
yellow, or pale green cotyledons, which can also be classified at
this stage (Figure 2). Persons with very good vision may not need
a magnifier: if necessary, an enlarger should be used which does
not interfere with the use of the hands. With a little experience,
approximately 70 fruits can be screened within an hour.
Organization of the Germline
The fruits appear on the stem in a rather well-organized
pattern (Figure 3). The arrangements (phyllotaxis) follows a
spiral, and after some turns, two or more fruits appear on the same
vertical line. It is expected that the vertically aligned fruits
belong to the same cell lineage. In chimerical plants, the genetic
constitution of the same lineage is expected to be the same, while
other cell lineages may be different (Figure 1) indicating that the
germline is composed of two cells in the mature seed. (This
statement will be supported later.) If one of these cells contains
a mutation, the plant is expected to have two sectors, one of the
normal constitution (homozygous wild type) and another that is
heterozygous for the mutation. The fruits situated on the stem
may represent one or the other sector. Though the periodicity of
the fruit arrangement on the stem may be influenced to some extent
by developmental factors or by the external conditions, the
experimenter must be familiar with the phyllotaxis in order to make
the work efficient. As a rule of thumb, it is advisable to scan
fruits which sit on opposite sides of the stem, as these will most
likely represent different sectors of the chimera. It is possible,
however, for phenotypically identical mutants to occur in opposite
fruits; these, most likely, originated from independent mutational
events (Figure 4).
Calculation of the Mutation Frequency
Within single fruits, the segregation of mutant and wild-type
embryos is expected to correspond to monogenic Mendelian ratios.
Dominant and recessive mutations may be distinguished if the number
of embryos is sufficiently large. Single fruits may contain up to
30 or 40 embryos, though frequently their number is much reduced
when the chemicals are toxic. Occasionally, recessive mutations
may mimic ratios expected on the basis of dominant inheritance,
because of the small numbers and/or reduced transmission of the
chromosomes (gametes) involved. The simultaneous induction of two
phenotypically similar mutations may result in a theoretical
proportion of 9 wild-type:7 mutant. Poor transmission of the
chromosome carrying the wild-type allele (because of a large
-------
ARABIDOPSIS ASSAY OF ENVIRONMENTAL MUTAGENS 217
Figure 3. Arab idopsi s main stem showing the phyllotaxis of the
frui ts (in a right-hand spiral, as illustrated at
right) .
deficiency or other gametophyte factors) may give a 2:1 ratio.
It is sometimes not possible to distinguish clearly among these
proportions on the basis of the few embryos in a fruit.
In addition, we may observe a deficiency of the mutant class:
this situation is actually the most common. The reduction of the
mutant class within a fruit can be caused by the poor transmission
of the chromosome carrying the mutant allele or by the early elimi-
nation of some of the zygotes homozygous for defective alleles.
When there is normal inheritance of the mutants and not too
high a frequency of mutation (i.e., the majority of plants would
incur only a single mutation), we can determine rather accurately
the number of sectors in the mutant plants. Alternatively (or
additionally), we may harvest all the seeds from individual
plants and determine segregation ratios in the planted M2
generation. In the case of normal transmission and viability, the
proportion of the wild-type and mutant individuals in the progeny
is expected to reflect the number of cells in the germline that
gave rise to the fruits. (This is what we call the genetically
effective number of cells, or GECN.) Where the germline is
-------
218
G. P. REDEI
Figure 4, The time course of Che development of Che germline.
In Che mature embryo. the germline consists of abouC
two cells (left): in the maturing plant, the number
of cells in the germline increases (right). If one
of the two cells of the germline in the mature seed
incurs a mutation, the growing plant becomes a
chimera containing two identical-size sectors. The
genetic constitutions of Che fruits (embryos) are Chen
determined by their locations on the stem.
represented by a single cell at the time of the mutation,
segregation for wild-type and mutant should show a proportion of
3:1 (GECN 3 1). Where the germline contains two cells (GECN = 2)
at the time of the mutagenesis and only one of the two cells
suffers a mutation, the expected segregation ratio is 7:1: that is,
one of the two cells segregates 4:0 while the other displays 3:1
proportions, and pooling the data, we obtain the 7:1 ratio. Simi-
larly, if the germline were eight cells, seven of the fruits would
be expected to yield only wild-type embryos, and one would display
3:1 segregation progeny, a proportion of 31:1 ([7 x 4] + 3 : l).
Thus, on the basis of genetic data, we can infer the number of
genetically effective cells at the time of the mutagenic exposure
if it is of relatively short duration, that is, lasting for a few
-------
ARA BID OPS IS ASSAY OF ENVIRONMENTAL MUTAGENS
219
hours rather Chan for many days. The number of genetically
effective cells is reasonably consistent at a particular
developmental stage. Poor sampling of the seed output of the
plants and reduced transmission of certain mutants may, however,
cause variation around the most frequent class (Figure 5.)
Figure 5. Variations in the apparent number of cells in the
gerraline. The mode (the most frequent class) indicates
that the number of genetically effective cells (GECN)
is generally two. The classes shown were established
by grouping the segregation data of all mutants
classified and counted.
When the number of genetically effective cells is known, the
frequency of mutation can be easily calculated by the following
formula:
-------
220
G. P. REDEI
R » no. of independent mutations observed
no. of progenies classified x GECN x ploidy x dose of mutagen.
This formula expresses mutation frequency on a genome basis,
because we take into consideration the diploid nature of all the
cells in the germline by substituting 2 for ploidy in the
denominator. By this procedure, mutation frequencies can be
directly compared with those of other higher organisms or
microoganisms. The dose of the mutagen used may be omitted from
the formula.
In an experiment, we treated mature seeds with 0.3%
ethylmethane sulfonate, and the embryos were classified before
maturity as outlined (Table 3). The fruits not used for embryo
analysis were harvested, and in the M2 generation, phenotypically
distinguishable mutant classes were counted in 308 families. Some
of the families displayed no mutants at all, and others showed one
or more types. The frequency distribution of these families is
shown in Figure 6, with the theoretical expectation based on the
Poission distribution.
Table 3. Frequency of Embryo Mutants after Mutagenic Treatment
with Q.2% Ethylmethane Sulfonate for 15 Hours

No. of
No. of
No. of Fruits
Mutation
Treatment
Plants
Genomes
Analyzed
Frequency
Untreated
192
768
599
0.0013
Treated
205
818
485
0.4707
This analysis considered all the mutants that germinated and
expressed themselves during early or later stages of development.
Many mutations that could be identified in the immature fruits
could not be detected after planting the seeds. Apparently, these
involved lethality that prevented germination of the individuals
affected. A number of additional types of mutants could, however,
be identified during the later stages of development.
A closer examination of the classes of families with various
or no mutational events revealed some similarities with and some
discrepancies from the Poisson distribution. Curiously, the
frequency of families with one or more mutations was higher than
-------
ARABIDOPSIS ASSAY OF ENVIRONMENTAL MUTAGENS
221
. 36
.30'
.28
m
poisson dis*ribit:on
•10
-06-
•C«
0 1 2 3 4 5 6 7 8 9 10 11 1} 13 14 15 16 17 1« 19 20 I
308 Fa.nili'es with Number of Phenotypical1y Distinguishable Mutations
0 12 3 4 5 6
Number Observes 22 115 113 45 6 2 0
Frecuency 0.071 0.373 C.383 0.146 0.020 0.C07 0
Figure 6. Comparison of Che distribution pattern of families with
multiple mutations with Poisson distributions for
averages of one to five independent events (m). The
abscissa (i) represents the numbers of events expected
to occur at the frequencies given by the ordinate.
expected on the basis of the curves shown. These curves show
Poisson distributions with 1, 2, 3, 4, and 5 average mutational
events expected per family. On the other hand, the frequencies of
classes 1 and 2 were nearly equal, a feature characteristic of the
Poisson distributions of integer numbers (Figure 6, top).
In the 308 families, 520 mutations were identified on a
phenotypic basis. This classification was obviously loaded with a
systematic error, because in the presence of two independent
mutations per cell, we would also expect, besides the two single
mutants, the double-mutant class. In the case of three mutational
events, three single mutants, three different types of double
-------
222
G. P. REDEI
mutants, and one triple—mutant phenotype may be exhibited in a
family. Since progeny analysis could not be performed with the
lethal or subvital types, the genetic constitution of these
phenotypic classes could not be determined.
The average number of mutational events per family can be
determined, however, if we consider the zero-mutation class. The
frequency of this group is not influenced by the complications just
mentioned. The theoretically expected frequency of the
zero-mutation class (based on the Poisson distribution) was
determined for 2.0 to 3.0 mutational events per progeny, and the
frequencies are shown in Table 4.
Table 4. Theoretically Expected Frequency
of the Zero-Mutation Class
Average no. 2.0 2.5 2.6 2.7 3.0
of mutations
Expected 0.135335 0.082085 0.074274 0.067206 0.049787
frequency of
zero-nutation
class
Figure 6 (bottom) shows the result of an experiment where the
frequency of the zero-mutation class was 0.071, which indicates an
average of 2.6 to 2.7 mutations per family (Table 4). Since at the
time of the treatment with EMS, the gerraline of each mature embryo
contained four genomes, the average frequency of mutations in this
material can be computed as (approximately) 2.6/4 = 0.65. This is
considerably higher than the observed 0.47 for mutations expressed
at the embryo stage in the immature fruits; it is also higher than
the empirically found value of 0.59 for all the mutations observed.
We may also conclude that under the conditions of this experiment,
approximately 10% of the mutations were missed, either because of
misclassification or due to some other accidental causes. The
difference between the experimentally observed value of 0.59 and
the frequency of 0.65 predicted on the basis of Poisson's
exponential binomial limits is not so large as to give basis for
serious reservations concerning validity.
-------
ARABIDOPSIS ASSAY OF ENVIRONMENTAL MUTAGENS
223
Calculation of the Number of Target Loci
From the viewpoint of effectiveness of a mutagen assay, it is
important to know how many potential targets can be hit by a
mutagenic chemical. It seems that the mutability of the gene loci
is not uniform across the entire genome (Table I). This may be due
to differences among the various loci in number of nucleotide pairs
(the size of the genes). The genes may be more mutable during
periods of replication or transcription, and these differences in
state may be reflected in the observed mutabilities. Also, certain
sites within a gene may appear as "hot spots" when exposed to one
agent but not when another mutagen is applied (Table 2). The
nature of particular base pairs and/or the conformation of the DNA
may also affect mutability. The metabolic machinery, the genetic
repair systems, etc., may also affect the various loci differently.
Assays that are capable of monitoring mutational responses at
a large number of loci, presumably representing all of the loci in
a fair manner, are particularly attractive. The number of genes
cannot be directly counted in higher organisms. Since nucleotide
sequencing became practical, the number of genes of a few viruses
could be determined (Fiers, 1975; Sanger et al., 1977). Such an
approach is still impractical for mammals or higher plants, which
may have six to eight orders of magnitude more DNA per cell than
the smallest viruses.
The number of genes (cistrons) in Drosophila is believed to be
about 5,000 (Garcia-Bellido and Ripoll, 1978), a figure that is
close to the number of bands detectable by the most revealing
counts on the salivary gland chromosomes. Belling (1928) assumed
that the number of chromomeres observed in the lily chromosomes,
2193, indicates the number of genes in this plant. Because of the
small size of the chromosomes of Arabidopsis, chromomeric organiza-
tion cannot be determined (Figure 7), but even if this plant were
quite favorable for cytological studies, an estimate of the number
of genes on such a basis would not be sufficiently realistic.
The number of gene loci can be calculated more precisely on
the basis of overall mutation frequencies if the average rate of
mutation per locus is known. In Arabidopsis, estimates of mutation
rate are available for specific loci involved in the synthesis of
thiamine (vitamin B^). For the calculations, mutations induced
with EMS at the three loci and in the following numbers were used:
py 35, tz 5, £h 11. The average induced mutation rate at these
three loci is 3.2 x 10~5, which is comparable to the data for mice
from the literature (Table 5).
These loci of Arabidopsis were chosen for studies not on the
basis of their mutability, but because of our interest in the
genetic control of a biosynthetic path. They may represent to a
Q,
-------
224
G. P. REDEI

Figure 7. Salivary gland chromosomes of Drosophila compared on
an equal scale with the chromosomes of Arabidopsis,
shown within the box. (Courtesy of Dr. H.K. Mitchell
and Dr. Lotti Sears, respectively.)
Table 5. Induced Mutation Rates in Arabidopsls and Mouse
Organism
Loci
Mutagenic
Agent
Rates
Arabidopsis
21
tz
th
(EMS)
(EMS)
(EMS)
2.0 x lO"1*
2.5 x 10"5
7.0 x 10"5
Mouse
Isozyme loci
7 Specific loci
X-raysa
EMS°
1.7 x lO-4*
7.8 x 10"5
aMalling and Valcovic, 1978.
bEhling, 1978.
-------
ARAB I DO PS IS ASSAY OF ENVIRONMENTAL MUTAGENS
225
fair extent the rest of the genes of this plant. This assumption
cannot be proven, however. In the absence of better information in
this or other systems of higher eukaryotes, we have no other
choice.
When the estimated overall frequency of induced mutations is
divided by the average frequency of mutations per gene, we obtain
an estimate of the number of loci capable of mutation or of the
expression of a mutant phenotype at a particular developmental
stage. For example, the average frequency of mutations detectable
at the embryo stage was found to be approximately 0.47 (Table 3),
and the average rate of mutation per locus was estimated to be 3.2
x 1Q~^. Hence, the minimal number of loci responding with
mutations expressed at the stage of immature embryos is 0.47/3.2 x
10"5 = 14,687. Similarly, on the basis of mutations expressed at
other developmental stages, we can calculate the number of genes
expressed in mutant states at early, late, or all developmental
stages (Table 6).
Table 6. Calculation of the Number of Loci in Arabidopsis
Frequencies
Embryo mutations per genome 0.47
Early seedling mutations per genome 0.32
Late seedling mutations per genome 0.12
Frequency of all mutations observed 0.47 +0.12 ¦ 0.59
Average mutation frequency of 3 loci: 3.2 x 10-^
Estimate of the number of loci with embryo mutations:
0.47
3.2 x 10*5 = 14,687
Estimate of all the potentially mutable loci (based on Table 4
data):
0.65
3.2 x 10_i x 20,313
Estimate of the number of all loci with mutations observed:
0.59
3.2 x 10-:> - 18,438
-------
226
G. P. REDEI
Some objections may be raised to this procedure of calculating
gene numbers. For example, these average mutation frequencies may
include repeated mutations at some loci and none at others. The
repeated occurrences do not seriously bias the data, however,
because the probability of including twice a locus with a high, say
10-2 , frequency of mutation is only 10"1*. Such cases do not much
affect the average frequencies, which are in the 10_1 range. Thus
the expected error in connection with each case is in the 0.0001
range. Similarly, the very stable loci barely bias the figures
because the use of approximation shown in Table 5 affords the
proper correction.
Since this mutagen test assesses forward mutations, the actual
number of sensitive sites substantially exceeds that of the number
of genes. If we assume that an average locus consists of one
thousand nucleotide pairs, the number of potentially mutable sites
per plant may reach several millions.
Critical Population Size
The potencies of various mutagens are very unequal, yet we
must evaluate their effectiveness in a reliable manner. Therefore
some guidelines are needed as to the size of the populations to be
tested in order to find some mutants even if their expected
frequency is very low. Also, we must establish some criteria of
clearance for an apparently innocuous compound. We need to know
how many plants to screen to find at least one mutant or what
number of mutations represents a significant increase over the
spontaneous rate.
The critical size of a population can be defined as the number
of plants to be tested, or better, the number of genomes, which may
yield at least one mutation at a chosen level of probability (P).
The rationale of the procedure is that we rule out the chance of
finding no mutational events at all more frequently than specified
by 1-P. Let us use a very simple example: When a plant is
heterozygous for a single allelic pair, according to the Mendelian
rule, we expect 3/4 of the progeny to have the dominant phenotype
and 1/4 of the individuals to display the recessive genotype. How
many individuals (n) do we then need in the M2 generation in order
to find at least one recessive plant with a probability (P) of
0.99? Since P = l-(3/4)n,
n = log (1-P) -2 16.008 = 17.
log (3/4) -0.1249387
-------
ARAB ID OPS IS ASSAY OF ENVIRONMENTAL MUTAGENS
227
Similarlyif we expect an induced frequency of nutations of
1/200 genomes, the number of genomes to be tested (n) at P = 0.99
should be
n » log (1—P) = -2 = 918.7
log (199/200) -0.002177
to find at least one genome within the germline with a recessive
mutation at the 0.99 level of probability. When we treat mature
seeds, where the germline is represented by four genomes (as we
discussed earlier), we need to test (918.7)/4 = 230 surviving
plants.
The question remains whether an "induced" frequency of
mutations is higher than the spontaneous rate. Certainly we must
use a negative control. The frequencies between the two groups
should be compared by determining the appropriate confidence
limits. We must remember that even if no mutation is observed in
100 genomes, this may not negate the possibility that we missed up
to four according to the 95% confidence limit. Similarly, we may
want to know whether the 10% mutations observed among 200 genomes
is significantly different from the 15% observed in 100 genomes.
Reference to the 0.95 confidence belt convinces us readily that
they are not. Were we to have 25% mutations in the latter group,
the difference would be significant at the 0.95 level.
CORRELATION BETWEEN MUTAGENIC EFFECTS IN ARABIDOPSIS AND
CARCINOGENICITY IN AMIMALS
The embryologist Theodore Boveri attributed cancer to the
presence of an abnormal complex of chromatin as early as 1914.
Evidence for and against a genetic cause of cancer has been
entertained and occasionally negated ever since. In the widely
used Ames Salmonella assay system, of more than 200 carcinogens
over 90% showed mutagenic effects (Hollstein et ai. 1979). A
survey (Rinkus and Legator, 1979) of 465 known or suspected
carcinogens examined by the Salmonella S-9 method indicated a lower
correlation (77%) with mutagenicity. It seems that not all groups
of carcinogens are equally efficient mutagens for all organisms.
With Arabidopsis so far, approximately 110 chemical compounds
had been tested for mutagenicity, according to a survey of about
four dozen publications. Interestingly, this compares favorably
with Drosophila, for which about 1000 publications are listed by
the Environmental Mutagen Information Center as being concerned
with tests of about the same number of compounds (Hollstein et al.,
1979).
-------
228
G. P. REDEI
We have found information concerning carcinogenicity and
neoplastic effects for 52 compounds tested for mutagenicity in
Arabidopsis (Redei et al., in press). The correlation between
carcinogenicity in animals and mutagenicity in Arabidopsis seems
comparable to or better than that for Salmonella (Table 7). It is
particularly noteworthy that only one compound, di-2-
chloroethylamine phosphamide ester (endoxan, Cytoxan), was negative
in the mutagenicity assay among those that are listed as category I
carcinogens for animals (proven in at least two animal systems)
according to the Occupational Safety and Health Administration
(OSHA)• For this particular carcinogen, only a short negative note
is available concerning mutagenicity in Arabidopsis (Miiller,
1965a), which may require revision on repetition. This compound
(synonym cyclophosphamide), when administered to male mice, showed
dominant lethal effects for spermatozoa, but no dominant lethals
were induced when spermatocytes or spermatogonia were tested
(Rohrborn, 1970). According to Heddle and Bruce (1977), this
compound is carcLnogenic and shows positive responses in the mouse
sperm abnormality, bone marrow micronucleus, and Salmonella tests.
Triethylenemelamine is another false negative in Arabidopsis; it
was found to be mutagenic in several other assays, including the
sex-linked recessive lethal test in Drosophila, in the mouse
specific locus test, and in a Salmonella assay (Hollstein et al.,
1979). Again, the negative result in Arabidopsis is found in the
same undetailed note (Miiller, 1965b).
Table 7. Number of Compounds Tested for Mutagenicity in
Arabidopsis and for Carcinogenicity in Animal Assays
Carcinogens Neoplastic
Mutagenic in Arabidopsis
Nonmutagenic in Arabidopsis
Total
40 (87%) 4 (66.6%)
_! A
46 6
OSHA Categories of Compounds That Are Nonmutagenic in Arabidopsis
Di-2-chloroethylamine phosphamide ester I
Ethyl alcohol II
Chloramphenicol II
Maleic acid hydrazide III
Sulfathiazole III
Triethylenemelamine ?
-------
ARABIDOPSIS ASSAY OF ENVIRONMENTAL MUTAGENS
229
The remaining four nonmutagenic suspected carcinogens are only
category II and III compounds, indicating that the evidence for
carcinogenicity is fragmentary or insufficient. I was unable to
find positive mutagenic information for them in the commonly used
mutagenic assay systems.
CONCLUSIONS
The Arabidopsis assay as outlined provides direct evidence
concerning the genetic nature of most of the mutagenic alterations.
The majority of the mutants detected can be subjected to formal
genetic analysis. This system permits the simultaneous study of
forward mutation at thousands of loci; therefore, the information
obtained must be relevant for practically all the genes of this
organism and presumably for other eukaryotes. Because of the very
small size of the chromosomes of Arabidopsis, cytological proof for
chromosomal aberrations is hard to obtain (except for some very
gross ones). The sterility of the fruits can be, however, easily
determined by counting the missing embryos in the linear array. A
considerable number of these defects, perhaps most, are presumably
due to large deficiencies and two- or multiple-hit aberrations of
the chromosomes.
The mutation frequencies in Arabidopsis can easily be
expressed on the genome basis, and the test results can be compared
with those of any prokaryote or eukaryote. Mutagenicity
information in Arabidopsis correlates very well with the
carcinogenicty data for animal test systems. There is direct
evidence (Redei et al., 1980) that the metabolic system of of
Arabidopsis can activate several promutagens into genetically
effective compounds.
A mutagen assay with Arabidopsis can be completed within four
to five weeks. Because of the small size of the plants, up to 200
or more individuals can be raised per 10-cm-diameter petri plate
in Inexpensive growth chambers. Therefore, the assays are not only
fast and revealing, but they are also very inexpensive- The plants
can be grown year round, both in the laboratory and in the
environment (this species is winter-hardy even in the North).
Though sufficient information is lacking on Its utility for in situ
testing of environmental pollutants, there is no apparent reason
why it could not be employed successfully for this purpose too.
ACKNOWLEDGMENTS
This study was supported by EPA contract D4541 NAST, and it is
contribution no. 8497 from the Missouri Agricultural Experiment
Station. I appreciate the valuable comments and assistance of
-------
230
G. P. REDEI
Dr. Shahbeg Sandhu and Dr. William Lower during the course of the
investigations. Some of the data were collected with the technical
assistance of Patti Reichert and Magdi Redei; their conscientious
efforts and enthusiasm are gratefully acknowledged.
REFERENCES
Belling, J. 1928. The ultimate chromomeres of Lilium and Aloe
with regard to the number of genes. Univ. California Publ.
Bot. 14 :307-318.
Ehling, U.H. 1978. Specific-locus mutations in mice. In:
Chemical Mutagens: Principles and Methods for Their
Detection, Vol. 5. A. Hollaender and F.J. de Serres, eds.
Plenum Press: New York. pp. 233-256.
Fiers, W. 1975. Chemical structure and biological activity of
bacteriophage MS2 RNA. In: RNA Phages. N.D. Zinder, ed.
Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY.
pp. 353-396.
Garcia-Bellido, A., and P. Ripoll. 1978. The number of genes in
Drosophila melanogaster. Nature 273:399-400.
Heddle, J.A., and W.R. Bruce. 1977. Comparison of tests for
mutagenicity or carcinogenicity using assays for sperm
abnormalities, formation of micronuclei, and mutations in
Salmonella. In: Origins of Human Cancer, Book C, Human Risk
Assessment. H.H. Hiatt, J.D. Watson, and J.A. Winsten, eds.
Cold Spring Harbor Laboratory Press: Cold Spring Harbor, NY.
pp. 1549-1557.
Hollstein, M. , J. McCann, F.A. Angelosanto, and W.W. Nichols.
1979. Short-term tests for carcinogens and mutagens.
Mutation Res. 65:133-226.
Mailing, H.V., and L.R. Valcovic. 1978. New approaches to
detecting gene mutations in mammals. In: Mutagenesis,
Advances in Toxicology, Vol. 5. W.G. Flamm and M. Mehlman,
eds. Hemisphere: Washington, DC. pp. 149-171.
Miiller, A.J. 1963. Embryonentest zum Nachweis rezessiver
Letalfaktoren bei Arabidopsis thaliana. Biol. Zbl.
83:133-163.
Miiller, A.J. 1965a. A survey on agents tested with regard to
their ability to induce recessive lethals in Arabidopsis.
Arabidopsis Inf. Serv. 2:22-24.
-------
A RABID OPS IS ASSAY OF ENVIRONMENTAL MUTAGENS
231
Muller, A.J. 1965b. In: Induction of Mutations and the Mutation
Process. J. Veleminsky and T. Gichner, eds. Publishing House
of the Czech. Acad. Sci. : Praha. pp. 46-52.
Prakash, L., and F. Sherman. 1973. Mutagenic specificity:
reversion of iso-l-cytochrorae c mutants of yeast. J. Molec.
Biol. 79:65-82.
Redei, G.P. 1980. Basic Plant Genetics. University of Missouri:
Columbia, MO. p. 567.
Redei, G.P., M.M. Redei, W.R. Lower, and S.S. Sandhu. 1980.
Idendification of carcinogens by mutagenicity for ArabidoDsis.
Mutation Res. 74:469-475.
Rinkus, S.J., and M.S. Legator. 1979. Chemical characterization
of 465 known suspected carcinogens and their correlation with
mutagenic activity in the Salmonella typhimurium system.
Cancer Res. 39:3289-3318.
Rohrborn, G. 1970. The activity of alkylating agents. I.
Sensitive mutable stages in spermatogenesis and cogenesis.
In: Chemical Mutagenesis in Mammals and Man. P. Vogel and G.
Rohrborn, eds. Springer-Verlag: Berlin, pp. 294-316.
Sanger, F., G.M. Air, B.G. Barrell, N.L. Brown, A.R. Coulson, J.C.
Fiddes, C.A. Hutchison III, P.M. Slocorabe, and M. Smith.
1977. Nucleotide sequence of bacteriophage X174 DNA. Nature
265:687-695.
-------
Intentionally Blank Page
-------
SOYBEAN SYSTEM FOR TESTING THE GENETIC EFFECTS OF
INDUSTRIAL EMISSIONS AND LIQUID EFFLUENTS
Baldev K. Vig
Department of Biology
University of Nevada
Reno, Nevada
INTRODUCTION
The testing of complex mixtures found in today's air, water,
and soils for harmful effects on man i3 a gigantic and important
task. This paper introduces a eukaryotic test system that yields
at least qualitative estimates of genetic damage of the type
observable in man. The soybean spot test is based on inducing
various types of genetic damage by treating seeds or seedlings with
the chemical or mixture of chemicals in question. This review
briefly describes the nature of the test system, the kinds of data
obtained for various chemicals, their postulated genetic effects,
and the potential usefulness of this material for environmental
mutagen testing. This summary is only a guideline to understanding
the system; the details of previous work can be found in references
(for reviews, see Nilan and Vig, 1976; Vig, 1975, 1978).
THE TEST SYSTEM
The Origin of Spots
In 1956, Weber and Weiss described a light-green colored plant
of soybean (Glycine max [L.] Merrill) whose progency segregated in
a ratio of 1 dark green: 2 light green: 1 golden yellow. The
yellow color is due to lack of chlorophyll. The alleles
controlling this trait are symbolized by and y^ : Y^Y]^
plants are dark green, Y^y^i plants are light green, and yj^yjj
plants are golden yellow. In heterozygotes, the embryonic leaves
(which develop into two simple, opposite leaves) and the first
compound leaf occasionally show a few dark green spots and about an
233
-------
234
BALDEV K. VIG
equal number of yellow spots. Although these spots resemble the
leaves of the two homozygotes in color, the intensity of their
colors depends on the number of layers of palisade cells involved
in the mutational event. An altered embryonic leaf cell usually
develops its colony of cells during division and expansion without
being outcompeted by the surrounding tissue. The yellow spots
survive because the light green background tissue provides the
necessary carbohydrates for their growth.
Besides these single spots, the Y^y^ leaves also produce
twin spots composed of adjacent equal-sized, mirror-image spots,
one dark green and one yellow. The cells of such twin spots appear
to have complementary genotypes, i.e., YjjYh and y^yii.
The origin of twin spots is best attributed to somatic
crossing over in Y^yjj cells. The failure of one of these cells
to develop into a visible sector results in a single dark green or
yellow spot. The frequency of occurrence of these twin spots
coincides with the frequency of somatic crossing over in several
other organisms (see Vig, 1978) and can be dramatically increased
by applying substances known to cause mitotic recombination, such
as mitomycin C (German and LaRock, 1969; Holiday, 1954; Vig and
Paddock, 1968) .
The single yellow spots may also develop by the multiplication
of a cell (or cells) that has lost the chromosome segment carrying
the Yjj allele. Duplication of Y^ allele or deletion of the yij
allele, followed by cell multiplication, may give single dark green
spots (Vig, 1969b). These situations may result from nonhomologous
translocations involving the chromosome carrying this gene.
Consequently, Y^y^ leaves treated with a given mutagen or complex
mixtures of mutagens may be analyzed to distinguish between the
modes of action of various mutagens. Furthermore, specific locus
mutations may be induced (yjj to Y^j, in ynyn cells) to give
light green spots on a yellow background.
The Protocol
A 10- to 15-g sample of seed of variety T219, L65-1237, or
L72-1937 is treated for the desired length of time with a solution
of the chemical or mixture of chemicals to be tested. The seed is
then thoroughly washed in running tap water, sown 6 mm (0.25 in.)
deep in washed, coarse sand of no nutritive value in galvanized
metal flats, and watered as required, depending on the temperature
and humidity of the greenhouse. The protocol can be altered so
that the seed is watered with a solution of the suspected mutagen.
Agents like mitomycin C, caffeine, or nitrosourea may be used as
positive controls.
-------
SOYBEAN SYSTEM FOR TESTING GENETIC EFFECTS
235
The plants are ready for analysis in four Co five weeks, when
Che second compound leaf unfolds. For both heterozygous and
horaogyzous plants, spots are counted on the two simple leaves and
Che first compound leaf, which is considered equivalent in area to
three simple leaves. Usually, the simple leaves have more spots,
due to a larger initial number of cells than in the compound leaf.
Because the spontaneous background frequency of the control varies
with the age of the seed, a statistical procedure such as the
t-test is used to confirm the induction of damage by the mutagen.
The doses and number of replicates required are determined by Che
chemical's mutagenic effectiveness.
STUDIES WITH SOME TEST AGENTS
The Induction of Twin Spots
Chemicals whose primary geneCic effect is to cause somatic
recombination produce a preponderance of twin spots on the
heterozygous leaves. Two such chemicals, caffeine and mitomycin C,
deserve special mention. Four-hour treatments with caffeine in
concentrations ranging from 0.0625% to 0.5% increase the
frequencies of all types of spots, especially doubles (Vig, 1973b).
In one experiment, the ratio of total spots to twin spots was 11:1
in the control; for seeds treated for 4 h with 0.0625%, 0.125%,
0.25%, and 0.5% caffeine, the ratios were 3.6, 4.0, 3.4, and 2.7,
respectively. Several other experiments have confirmed that about
one third of the spots induced by caffeine are twin spots.
When seeds are Created with mitomycin C at a wide range of
concentrations and at various stages of seed germination, usually
one third or more of the spots are twins. The same frequency of
twin spots occurs following treatment of seeds with mitomycin C
solution and following application of the chemical in lanolin paste
to the growing tip of the seedling, including the unexpanded third
through fifth compound leaves (Vig and Paddock, 1968). In these
studies, concentrations of mitomycin C were as low as 0.00325% (for
24 h) for seed treatments and 0.005% in lanolin paste. Twin spots
are also preferentially induced by applying the chemical to the
seed at various physiological ages for 4-h periods (i.e., from 0 Co
4, 4 to 8, 8 to 12, 12 to 16,..., 32 to 36 hours after germination
[Vig, 1973a]) .
The alkylating agents diepoxybutane, trenimon, and
methylmethane sulfonate affect the frequency of spots similarly.
These three agents are potent inducers of spots of all types, about
one third of which are twin spots. In the case of diepoxybutane,
concentrations as low as 0.5 ppm applied to seed for 24 h are
effective, and concentrations of trenimon as low as 0.25 ppm
applied under similar conditions increase the total spot frequency
-------
236
BALDEV K. VIG
Co about twice Chat of the control (Vig and Zimmermann, 1977).
Methylmethane sulfonate induces similar types of spots when applied
at various phases of germination (Vig et al., 1976) and at
concentrations as low as 6 ppm. At high concentrations, however,
these alkylating agents affect leaf expansion drastically,
resulting in a reduced spot frequency calculated on per leaf basis.
Preponderant Production of Yellow Spots
While ethyl methanesulfonate slightly increases the frequency
of twin spots over that of the control, it is far more effective in
producing yellow spots than the other two types of spots. In one
study (Vig et al., 1976), treatment with a 0.25% solution of this
chemical produced about 50% more single yellow spots than single
dark green spots and about 2.5 times as many single yellow spots as
twin spots. Although ethylmethane sulfonate and methylmethane
sulfonate are closely related chemicals, they produce different
frequencies of the various spot types at any given concentration
(Vig et al. , 1976) .
Treatment of the seed with cobalt-60-eraitted gamma rays gives
results similar to those for ethylmethane sulfonate, although the
relative frequency of yellow spots is even higher. This has been
found over a range of 10 to 750 R for both dry and pre-soaked seed
(Vig, 1974); the total frequency of spots is generally higher in
leaves from pre-soaked seed. Exposure of seed to beta particles
from tritiated water results in equal frequencies of the three
types of spots (Vig, 1974; Vig and McFarlane, 1975). Even tritium
concentrations as low as 0.01 uC/ml for 96 h (equivalent to 4.5 R)
are effective in producing spots. These differences could be due
to internal availability of beta particles to the DNA from within
the organic fraction of the embryo (Vig and McFarlane, 1975).
A Lack of Production of Twin Spots
Treatment of seeds with sodium azide (NaNj) greatly increases
the frequencies of single dark green and yellow spots (Vig and
McFarlane, 1975). However, the frequency of twin spots is barely
above that of the control (Vig, 1973c), and the chemical generally
does not produce light green spots on yellow leaves. Thus, these
spots are probably not due to somatic crossing over, point
mutations, or deletions of chromosome segments. We tentatively
conclude that NaN3 causes nondisjunction, giving Y^Y^y^ sectors
that are dark green and ^nYiiyn sectors that are nearly yellow.
The monosomic cell lines are presumably inviable or outcorapeted by
the normal and trisoraic lines, and are thus lost. We have not
found similar results for any other mutagen that we have tested.
-------
SOYBEAN SYSTEM FOR TESTING GENETIC EFFECTS
237
The Induction of Point Mutations from y; i to Yp
Light green spots on yellow (ynyjl) leaves are produced by
specific locus mutations, to Y^. Not all chemicals that
induce spots in the heterozygotes also produce light green spots in
yellow horaozygotes. Caffeine produces such spots at concentrations
of 0.05% or higher (Vig, 1973b). This effect distinguishes
caffeine from chemicals like mitomycin C or nitrosoamines, which
induce both twin spots and single spots on Y^y^ leaves, but have
no effect on y^y^ leaves.
The alkylating agents methylmethane sulfonate, ethylraethane
sulfonate, raethylethane sulfonate, and methylbutane sulfonate cause
mutation of yjj to Y^ at concentrations as low as 0.02% when
applied to seed for 20 h (Vig et al., 1976). Diepoxybutane and
trenimon also induce light green spots on yellow leaves. For these
chemicals, a concentration of 0.25 ppm applied to seed for 24 h is
mutagenic (Vig and Ziramerraann, 1977); thus, these chemicals induce
not only somatic recombination and chromosome deletions (as
indicated by the induction of spots on Yj^yu leaves), but also
point mutations of the allele yy to Y^.
No other chemical has been found to induce the mutation of y^
to Y^. However, as in other such test systems, beta particles and
gamma rays induce this point mutation at about the same frequency
as they do spots on Y^y^ leaves (Vig, 1974, 1978; Vig and
McFarlane, 1975) .
MUTAGENS REQUIRING METABOLIC ACTIVATION
In recent years, increasing attention has been focussed on the
mutagenic action of chemicals requiring metabolic activation. The
S-9 fraction of rat liver homogenate is commonly used to activate
promutagens. Recent studies indicate that liver is not the only
system with the enzymatic machinery needed for such activation. In
1963, Veleminsky and Gichner demonstrated the mutagenic activity of
some promutagens in plant systems without raamraal ian metabolic
activation (see Arenaz and Vig, 1978; Klekowski and Levin, 1979).
We have treated soybean seeds with aqueous solutions of
dimethylnitrosoamine at concentrations as low as 1.25 ppm for 24 h
(Arenaz and Vig, 1978). At this dose, we found a 2.8-fold increase
in the frequency of twin spots, a 2.6-fold increase in dark green
spots, and 1.7-fold increase in yellow spots, demonstrating that
this plant system can activate this chemical. Treatments with
concentrations from 60 to 500 ppm appeared to cause maximal
conversion of the chemical into true mutagen. Methylnitrosourea, a
related nitrosoamide that does not require metabolic activation,
showed no such saturation effect. However, methylnitrosourea is
-------
238 BALDEV K. VIG
much more toxic, in that interference with leaf expansion (yielding
an artificially low spot frequency) occurs at 125 pptn. In
comparison, dimethylnitrosoamine is tolerated by the plant at doses
as high as 500 ppra (Arenaz and Vig, 1974).
Despite data demonstrating metabolic activation in the plant,
it remains to be shown that metabolites produced through the
mediation of plant extracts (e.g., from nitrosoamines, atrazine,
and other such chemicals) can cause mutations or chromosome changes
in any mammalian system.
CONCLUSIONS
The soybean spot test is well suited for assaying both pure
chemicals and complex mixtures. The results for several agents
tested in this system are summarized in Tables 1 and 2. The system
allows not only quick and inexpensive assessment of the genetic
damage in a eukaryotic system, but also discrimination among
different genetic end points and thus among modes of action of the
agents. Thus, a chemical like methylraethane sulfonate, which can
cause DNA crosslinks, induces twin spots and single spots on Y^y^
leaves and light green spocs on ynyn leaves, unlike NaN3, which
produces only single spots on leaves and apparently has no
other effect. This discriminatory ability of the system is
advantageous in preliminary screening.
The soybean test system is suitable for use with complex
mixtures. Klekowsky and Levin (1979) recently showed that effluent
from paper mills induces spots on Y^iy^| leaves. The system should
be adaptable to testing of liquid, solid, or gaseous effluents, as
long as the seed or Che seedlings (as in Vig and Paddock, 1968) can
be treated at appropriate stages of development.
REFERENCES
Arenaz, P. 1977. Ineffectiveness of hycanthone raethanesulfonate
in inducing somacic crossing over and mutations in Glvcine
max. Mutation Res. 48:187-190.
Arenaz, P. 1978. Inability of aminoazotoluene to induce somatic
crossing over and mutation in Glycine max (L.) Merrill.
Mutation Res. 50:295-598.
Arenaz, P., and B.K. Vig. 1976. Induction of somatic mosaicism in
the soybean by some carcinogens. Genetics 83:93.
-------
Table I. Summary of Qualitative Genetic Effects of Agents Tested for
Induction of Spots oil Lhe Leaves of the Soybean (Glycine max)
Agent
Y11y11 Plants
Y] |y) | Plants
Twin Spots Single Spots Single Spots
Post ulated
Primary
Median ism
Re ferenee
Hitoinycin C +
Caffeine +
Actinomycin +
Daunomycin -
5-Fluorodeoxy uridine
N-methyl, N-nitro, +
N-n i t"rosoguan id ine
t'thylmethane sul fonate
Het.hylethane sulfonate
Mcthylmethane sul fonate
Methylbutane sulfonate
Colchicine +
Puromycin -
Sod Lum az ide t
N-mcthyI-N-nitrosurea
Dimethyl nitrosamine +
Trenimone
0iepoxybutane >
Cam fur
( l-[5-nitru-2 - furylJ-
2-[6-amino~3~pyrida7.yl | -
el lianehydrochlor ide)
t +
(not tested)
+ t
(not tested)
+
+ +
+
i
+
''Somatic crossing over usually with accompanying chromosome breakage
''Chi oiuosowe breakage as primary cause.
cPoint mut at i on.
''Nondisjunction (?).
a ,h V i g and Paddock, 1968
a,b,c, Vig, 19 7 3b
ri.h Vig, 1973b
b Vig and Paddock, 1968
b Vig, 1973b
h Arenaz and Vig, 1976
b ,c Vig fit al . , 1976
c V i g et al., 1976
a,b ,c Vig et al., 1976
c Vig eta I. ,1976
a Ashley, 1978;
Vig, 1969b
b Vig, 1973b
d Vig, 197 1c
b Arenaz and Vig, I97B
b Arena?, and Vig, 1978
a,c Vig and Ziimiiermann , 197 7
a,c Vig and Zimmertnanii, 197 7
a Vig and Z iinmermann , 197 7
to
o
CO
tn
>
•z
to
<
to
H
tn
o
?o
H
w
to
H
t—*
V,
o
o
M
H
H
n
tn
~*1
Tl
M
to
to
CO
-------
240
BALDEV K. VIG
Table 2. Agents That Did Not Induce Spots on the Leaves of the
Soybean (Give ine max)
Agent
Re ference
Aminoazo toluene
Nucleosides, d-A
d-C
d-G
d-T
Aluminum potassium
n itrate
Copper sulfate
Ferrous sulfate
Hycanthone
Ammonium thiocyanide
1-nitroso-2-
naphthyl-2,
6-disulfonic acid
Cytosine arabinoside
2-Chloroethanol
Hydroxylamine
hydrochloride
Urea
N, N-
dinitrosopiperazine
N-nitroso-
methyl urea
Arenaz, 1978
Vig, 1972
Vig and Mandeville, 1972
Vig and Mandeville, 1972
Vig and Mandeville, 1972
Arenaz, 1977
Vig, 1975
Vig, 1975
Vig, 1975
Vig, 1975
Vig, 1975
Vig, 1975
Vig, 1975
Vig, 1975
Arenaz, P., and B.K. Vig. 1978. Somatic crossing over in Glycine
max (L.) Merrill: activation of dimethyl nitrosoamine by
plant seed and comparison with methyl nitrosourea in inducing
somatic mosaicism. Mutation Res. 52:367-380.
Ashley, T. 1978. Effect of colchicine on somatic crossing over
induced by mitomycin C in soybean (Glycine max). Genetica
49:87-96.
German, J., and J. LaRock. 1969. Chromosomal effects of
mitomycin, a potential recombinogen in mammalian cell
genetics. Texas Rep. Biol. Med. 27:409-418.
Holiday, R. 1954. The induction of mitotic recombination by
mitomycin C in Ustilago and Saccharomvces. Genetics
50:323-335.
-------
SOYBEAN SYSTEM FOR TESTING GENETIC EFFECTS
241
Klekowski, E., and D.E. Levin. 1979. Mutagens in a river heavily
polluted with paper recycling waste: results of field and lab
mutagen assays. Environ. Mutagen. 1:209-219.
Nilan, R.A., and B.K. Vig. 1976. Plant test systems for detection
of chemical mutagens. In: Chemical Mutagens, Their
Principles and Methods of Detection, Vol. 4. A. Hollaender,
ed. Plenum Press: NY. pp. 143-170.
Vig, B.K. 1969a. Increase induced by colchicine in the incidence
of somatic crossing over in Glycine max. Theoret. Appl.
Genet. 41:145-149.
Vig, B.K. 1969b. Relationship between mitotic events and leaf
spotting in Glycine max. Canadian J. Cytol. 11:147-152.
Vig, B.K. 1972. Suppression of somatic crossing over in Glycine
max (L.) Merrill by deoxyribose cytidine. Molec. Gen.
Genet. 116:158-165.
Vig, B.K. 1973a. Mitomycin C induced mosaicism in Glycine max
(L.) Merrill in relation to postgermination age of the seed.
Theoret. Appl. Genet. 43:27-30.
Vig, B.K. 1973b. Somatic crossing over in Glycine max (L.)
Merrill: effect of some inhibitors of DNA synthesis on the
induction of somatic crossing over and point mutations.
Genetics 73:583-596.
Vig, B.K. 1973c. Somatic crossing over in Glycine max (L.)
Merrill: mutagenicity of sodium azide and lack of
synergistic effect with caffeine and mitomycin C. Genetics
75:265-277.
Vig, B.K. 1974. Somatic crossing over in Giveine max (L.)
Merrill: differential response to 3H-emitted 0-particles and
®^Co-emitted y-rays. Radiat. Bot. 14:127-137.
Vig, B.K. 1975. Soybean (Glycine max): a new test system for
study of genetic parameters as affected by environmental
mutagens. Mutation Res. 31:49-56.
Vig, B.K. 1978. Somatic mosaicism in plants with special
reference to somatic crossing over. Environ. Hlth. Perspect.
27:27-36.
Vig, B.K., and W.F. Mandeville. 1972. Ineffectivity of metallic
salts in induction of somatic crossing over and mutation in
Glycine max (L.) Merrill. Mutation Res. 16:151-155.
-------
242
BALDEV K. VIG
Vig, B.K., and J.C. McFarlane. 1975. Somatic crossing over in
Glycine max (L.) Merrill: Sensitivity to and saturation of
the system at low levels of tritium emitted 0-radiation.
Theoret . Appl. Genetics 46:331-337.
Vig, B.K., R.A. Nilan, and P. Arenaz. 1976. Somatic crossing
over in Glycine max (L.) Merrill: induction of somatic
crossing over and specific locus mutations by methyl
methanesulfonate. Environ. Exp. Bot. 16:223-234.
Vig, B.K., and E.F. Paddock. 1968. Alteration by mitomycin C of
spot frequencies in soybean leaves. J. Hered. 59:225-229.
Vig, B.K., and E.F. Paddock. 1970. Studies on the expression of
somatic crossing over in Glycine max L. Theoret. Appl.
Genetics 40:316-321.
Vig, B.K., and F.K. Zimmerraann. 1977. Somatic crossing over in
Glycine max. An induction of the phenomena by carofur,
diepoxyoutane and trenimon. Environ. Exp. Bot. 17:113-120.
-------
MUTAGENICITY OF NITROGEN COMPOUNDS FROM SYNTHETIC CRUDE
OILS: COLLECTION, SEPARATION, AND BIOLOGICAL TESTING
T.K. Rao, J.L. Epler, M.R. Guerin, B.R. Clark, and
C.-h. Ho
Biology and Analytical Chemistry Divisions
Oak Ridge National Laboratory
Oak Ridge, Tennessee
INTRODUCTION
Short-term mutagenesis assays have been used to test complex
environmental mixtures, in order to 1) serve as predictors of
long-range health effects, 2) guide chemical separation procedures
for the isolation and concentration of biologically active
materials, 3) identify chemical agents responsible for biological
activity, and 4) determine priorities for further, extensive
testing. Organic extraction coupled with chemical-class
fractionation is a prerequisite for most of these assays.
Our emphasis has been on evaluating various test materials
from the newly emerging synfuel technologies. We have previously
used the class chemical separation procedure developed by Swain et
al . (1969) to fractionate certain coal-derived (Epler et al., 1978)
and shale-derived (Epler et al., 1979b) oils for mutagenicity
testing. To minimize chemical reactivity during the fractionation
procedure, several synfuels were separated with Sephadex LH-20 gel
chromatography (Rao et al., in press). The bacterial mutagenicity
assay developed by Ames (Ames et al., 1975) employs certain
well-characterized histidine-auxotrophic mutants of Salmonella
typhimurium. Mutagenicity results indicate that the alkaline
fractions containing azaarenes, primary amines, and nitro
polyaromatics are mutagenic, along with the neutral-fraction-
containing polycyclic aromatic hydrocarbons (PAH) and nitrogen-
containing PAH (Ho et al., in press). The biological activity of
these organic compounds is characterized by the ability to revert
the frameshift strains TA1538 and TA98 (Rao et al., 1978) and
dependence on specific activation systems.
243
-------
244
T.K. RAO ET AL.
Source of Samples
Samples used in this study were supplied by U.S. Environmental
Protection Agency/U.S. Department of Energy Synfuel Research
Materials Facility (Coffin et al. , 1979). Process operating
conditions at the time of sampling, sampling conditions, and sample
histories are not sufficiently defined to allow process-specific
conclusions. Samples (with repository numbers) and their sources
are listed in Table 1. A detailed process description has been
presented elsewhere (Guerin et al., in press).
Table 1. Synfuel Samples and Their Sources
Sample3
Source
Petroleum
Wilmington crude oil (5301)
Recluse crude oil (5305)
Shale-derived oils
Shale oil (in situ) (4101)
Paraho shale oil (4601)
HDT-Paraho shale oil (4602)
Coal-derived oils
SRC II fuel oil (1701)
H-coal dist.—raw (1601)
H-coal dist.—HDT
low severity (1602)
H-coal dist .—HDT
medium severity (1603)
H-coal dist.—HDT
high severity (1604)
ZnCl2 dist. (1801)
Bartlesville Energy Technology Center
Bartlesville Energy Technology Center
Laramie Energy Technology Center
US Navy/Standard Oil Co. of Ohio
US Navy/Standard Oil Co. of Ohio
Pittsburgh and Midway Mining Co
Mobil Research/EPRI
Mobil Research/EPRI
Mobil Research/EPRI
Mobil Research/EPRI
Conoco Coal Development Co.
aRe spository numbers in parentheses
Purpose
The objective of this study was to identify mutagenic activity
in fractionated synfuel 3anples and to isolate and identify the
mutagenic chemical agents.
-------
MUTAGENICITY OF COMPOUNDS FROM SYNTHETIC CRUDE OILS
245
MATERIALS AND METHODS
Chemical Fractionation
The chemical fractionation procedure (Figure 1) has been
described by Guerin et al. (in press). Acidic and basic fractions
were separated by 1iquid/1iquid partitioning into ether-soluble
and -insoluble acids and bases (water-soluble fractions were
generally inactive in the mutagenicity assays). The neutral
fraction was separated using Sephadex LH-20 into aliphatic,
aromatic, polyaroraatic, and polar fractions by isopropanol elution.
Solvents were removed by rotary evaporation; the residue was
dissolved in d imethyl sul foxide for mutagenicity assays. The basic
fraction was subfractionated by use of basic alumina and Sephadex
LH-20. The ether-soluble base fraction was loaded onto the column
and eluted with 500 ml benzene (benzene subtraction) followed by
700 ml ethanol . Ethanol was removed by rotary evaporation, and the
residue was separated further on a Sephadex LH-20 gel column. The
column was eluted sequentially with 250 ml of isopropanol
(isopropanol subfraction) and 600 ml of acetone (acetone
subfraction), and the eluting solvent was removed by rotary
evaporat ion.
Mutagenicity Assay
Histidine-auxotrophic strain TA98 of S. typhimurium, obtained
through the courtesy of Dr. B.N. Ames (University of California at
Berkeley), was used for these studies. The procedure (based on the
work of Ames et al. , 1975) was to overlay minimal-medium agar
plates with soft agar containing the fraction being tested,
bacterial cells (2 x 10®), and liver horaogenate (S-9 mix) from
Aroclor-1254-induced rats (for metabolic activation). Activity
in revertants per milligram of test substance was derived from the
slope of the induction curve. Total mutagenic activity of a
starting material was computed from the activities of its
subtractions corrected for the percentages by weight contributed
by the subtractions to the starting material.
RESULTS AND DISCUSSION
The crude oils could not be tested for mutagenicity because of
their toxicity; when tested, they yielded questionable results. To
overcome this problem and also to obtain a more homogeneous
distribution of sample in the test medium, the samples were
fractionated by the general procedure described above (see Figure
1). Recovery and reproducibility have been tested by using
triplicate samples of shale oil and crude oil. Chemical recovery
-------
SI AH I ING MATH HIAI I
T-
HOIAHY kVAHOHAIION
E1 HER/ACID
VOLATILES
NONVOLAHLFS
qagls (Anurous)
ACIDS AND NEUTHAIS iEIHEH
pH/ETHER
pH/FTHfcH
NEUTRAL S IE 1 HIH)
ETMCn SOLUOLE
BASES IFTHf R)
IIASI S"
SEPHADCX LM 20/ISOPHOHANOL
INSOI DHL E
ACIDS
POLAR
AROMA1 ICS
ACIDS (AQUEOUS!
'01 YAROMAT ICS
INSOLUBLE
BASES
L 1 HER SOLUHIE
ACIDS (I fHEH)
waiiii r.oi uni f
BASCS IWAI LR)
waiih sui imii
ACILIS (AQUEOUS)
1
"NEU1HALS"
—Y
ACIDS'
F igure 1 .
Chemical fractionation procedure (from Guerin et al . , in press). Materials
are shown in boxes, their phases are indicated in parentheses, and the step
of Lhe procedure are given in boldface type.
-------
MUTAGENICITY OF COMPOUNDS FROM SYNTHETIC CRUDE OILS
247
and reproducibility are generally adequate for biological testing
purposes, even though recoveries are unacceptably 1ow with certain
samples. The losses are caused by volatile matter, which is
difficult to use in bioassays.
Table 2 gives the distribution of mutagenic activity in
petroleum oils, shale oils, and coal-derived oils. The acidic
fractions were inactive in the mutagenicity assays. All of the
activity was found in the basic and neutral fractions: the activity
levels varied from sample to sample. The petroleum samples showed
activity in the neutral fraction, which was weakly mutagenic. In
the shale oil samples, high in nitrogen, both the neutral and basic
fractions were mutagenic. Significant mutagenic activity was also
observed in both the neutral and basic fractions of the coal-
derived oils. The shale oils and coal-derived oils were more
mutagenic than were the petroleum samples.
Hydrotreatment (HDT) seemed to reduce mutagenic activity, as
seen in the results obtained with HDT-hydrogenated coal (H-Coal)
distillates. The high severity hydrotreatment completely
eliminated mutagenic activity from the H-Coal sample, while the
medium and low severity treatments were less effective. Zinc
chloride- (ZnC^) catalyzed distillation apparently eliminated
alkaline mutagens from the coal-derived oil.
Mutagenic activity in the basic and neutral fractions led us
to examine these fractions to isolate and identify chemical agents
responsible for the biological activity. When the basic fraction
was subfractionated, biological activity was found in the ultimate
acetone fraction that comprised approximately 10" of the starting
ether-soluble basic (ESB) fraction. Results for subfractionated
SRC-II ESB and shale oil ESB are given in Table 3. The acetone
fractions with high specific activities can be used for biological
assays in other genetic systems as well as for chemical analysis
(Epler et al. , 1979a). Azaarenes and primary aromatic amines are
the organic constituents of this fraction suspected of causing the
mutagenic activity.
Chemical separation and analysis of the neutral fraction
suggested PAH, alkylated PAH (Griest et al., 1979), and nitrogen-
PAH (Ho et al., in press) as the possible mutagenic agents.
Mutagenic activity of PAH and alkylated PAH appeared to increase
with the ring size (see Figure 2). The aliphatic constituents of
coal-derived materials were not mutagenic, while the nitrogen-PAH
fraction obtained from coal oil (Ho et al., in press) had a
specific activity more than twice that of the PAH fraction (Table
4), The nitrogen-PAH fraction from a shale-derived oil was not
mut agenic.
-------
Table 2. Distribution of Mutagenic Activity in Synfuels3
10
00
Distribution of Activity (%)
Sample
Total Mutagenic Activity
( revert an t s/ing ) ^
Neut ral
Ac id s
Bases
Other
Pet roleum
Wilmington crude oil
Recluse crude oil
5
6
100
100
0
0
0
0
0
0
Shale-derived oils
"Tin situ)
Shale oi1
Paraho shale oi1
HDT-Paraho shale
oil
I 78
390
0
54
31
0
2
0
0
42
69
0
2
0
0
Coal-derived oils
SKC;- 1 I fuel oi f
H-coal dist . — raw
H-coal dist.—IJDT
low severity
H-coal d i st . - HDL'
medium severity
H-coal tlist.—HUT
li igh sever i r y
ZnCl^ dist.
1000
350
540
210
0
530
65
63
100
0
LOO
0
0
0
0
0
aFroin Guerin eL al . (1980, in press).
^Determined from the linear portion of a dose-response curve with strain TA98,
35
37
0
0
0
0
0
0
0
0
H
*
to
>
O
M
H
>
-------
MUTAGENICITY OF COMPOUNDS FROM SYNTHETIC CRUDE OILS 249
Table 3. Distribution of Mutagenic Activicy in the
Ether-soluble Basic Fraction (ESB)
Test Relative Weight Specific Activity Weighted Activity
Substance {%) (rev/mg) (rev/nig)
SRC-II ESB - 0 14,000
Benzene 74 0 0
Isopropanol 6 400 24
Acetone 15 68,000 10,200
Total 95 10,224
Shale oil ESB - - 2,500
Benzene 77 0 0
Isopropanol 12 0 0
Acetone _9 20,000 1,800
Total 98 1,800
200
'60-
CL
X
od
CD
<
>
OJ
c:
A
NEUTRAL
SUBTRACTIONS
80"
40
'PAH
Alk-PAHs
O
J Alk-PAH SUBTRACTION
8(o!A
Apheron.
Napri
—T
0 20
-1—
60
100 O 20
/ DLATE
50
100
140
Figure 2.
Mutagenicity of the polycyclic aromatic hydrocarbon
(PAH) and alkylated (PAH) subfractions, including
benz(a)anthracene, phenanthrene, and naphthalene.
-------
250
T.K. RAO ET AL.
Table 4. Mutagenic. Activities of Neutral Sub fract ions of a
Coal-derived Oil and a Shale Oil3
Specific Activity (rev/mg)
Subfrac t ion
Coal-derived Oil
Sh a1e Oil
Aliphatic (AL)
0
0
PAH (I)
1390
120
Neutral N-PAH (II)
3250
0
Polar (III)
3380
1100
aFrom Ho et al. (in press).
CONCLUSIONS
Our conclusions are as follows: 1) Short-term bioassays such
as the Ames Salmonella histidine reversion assay can be effectively
applied to complex environmental samples. 2) Proper chemical
extraction and fractionation methods should be coupled to the
assay. 3) The shale- and coal-derived oils were relatively more
mutagenic than was petroleum crude oil. 4) Mutagenic activity was
mainly associated with the basic and neutral fractions.
5) Azaarenes, aromatic amines, PAH, alkylated PAH. and nitrogen-PAH
were the organic constituents of these fractions suspected of
causing the mutagenic activity.
REFERENCES
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian-microsome mutagenicity test. Mutation Res.
31:347-364.
Coffin, D.L., M.R. Guerin, and W.H. Griest . 1979. The interagency
program in health effects of synthetic fossil fuels
technologies: operation of a materials repository. In:
Proceedings of the Symposium on Potential Health and
Environmental Effects of Synthetic Fossil Fuel Technologies,
CONF-780903. U.S. Department of Energy: Oak Ridge, TN.
-------
MUTAGENICITY OF COMPOUNDS FROM SYNTHETIC CRUDE OILS
251
Epler, J.L., B.R. Clark, C.-h Ho, M.R. Guerin, and T.K. Rao.
1979a. Short-terra bioassay of complex organic mixtures:
Part II. Mutagenicity testing. Environ. Sci. Res.
15:269-290.
Epler, J.L., T.K. Rao, and M.R. Guerin. 1979b. Evaluation of
feasibility of mutagenic testing of shale oil products and
effluents. Environ. Hlth. Perspect. 30:179-184.
Epler, J.L., J.A. Young, A.A. Hardigree, T.K. Rao, M.R. Guerin,
I.B. Rubin, C.-h. Ho, and B.R. Clark. 1978. Analytical and
biological analyses of test materials from synthetic fuel
technologies. I. Mutagenicity of crude oils determined
by the Salmonella tvphimurium/microsomal activation system.
Mutation Res. 57:265-276.
Griest , W.H., B.A. Tomkins , J.L. Epler, and T.K. Rao. 1979.
Characterization of multialkylated polycyclic aromatic
hydrocarbons in energy-related materials. In: Polynuclear
Aromatic Hydrocarbons. P.W. Jones and P. Leber, eds. Ann
Arbor Science Publishers, Inc: Ann Arbor, MI. pp. 395-409.
Guerin, M.R., C.-h. Ho, T.K. Rao, B.R. Clark, and J.L. Epler.
1980. Polycyclic aromatic primary amines as determinant
chemical mutagens in petroleum substitutes. Environ. Res.
23:42-53.
Guerin, M.R., I.B. Rubin, T.K. Rao, B.R. Clark, and J.L. Epler.
(in press). Distribution of mutagenic activity in petroleum
and petroleum substitutes. Fuel.
Ho, C.-h, M.R. Guerin, B.R. Clark, T.K. Rao, and J.L. Epler. (in
press). Preparative-scale isolation of alkaline mutagens from
complex mixtures. Environ. Sci. Technol.
Rao, T.K., B.E. Allen, D.W. Ramey, J.L. Epler, I.B. Rubin, M.R.
Guerin, and B.R. Clark. (in press). Analytical and
biological analyses of test materials from the synthetic
fuel technologies. III. Fractionation with LH-20 gel for
the bioassay of crude synthetic fuels. Mutation Res.
Rao, T.K., J.A. Young, A.A. Hardigree, W. Winton, and J.L. Epler.
1978. Analytical and biological analyses of test materials
from the synthetic fuel technologies. II. Mutagenicity of
organic constituents from the fractionated synthetic fuels.
Mutation Res. 54:185-191.
Swain, A.P., J.E. Cooper, and R.L. Stedman. 1969. Large scale
fractionation of cigarette smoke condensate for chemical and
biologic investigations. Cancer Res. 29:579-583.
-------
Intentionally Blank Page
-------
THE DETECTION OF POTENTIAL GENETIC HAZARDS USING PLANT
CYTOGENETICS AND MICROBIAL MUTAGENESIS ASSAYS
Milton J. Constant in and Karen Lowe
Comparative Animal Research Laboratory
University of Tennessee
Oak Ridge, Tennessee
T.K. Rao, Frank W. Larimer, and James L. Epler
Biology Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee
INTRODUCTION
The recent realization that industrial effluents and wastes
could pose health hazards to future generations has led to
considerable research on the health effects of these substances and
to efforts to regulate their release. The Research Conservation
and Recovery Act (RCRA) specifically addresses the potential health
and environmental hazards of solid wastes. Evaluation of long-term
consequences of exposure to the huge number of potentially
hazardous compounds and complex mixtures would tax the scientific
community's whole-aniraal testing capabilities. Therefore, many
short-term bioassays have been developed to rapidly screen
compounds for toxicity, mutagenicity, teratogenicity, and
carcinogenicity. It is hoped that these short-terra tests will
predict health hazards early in the development of new
technologies, allowing technological changes to be made at early
stages, and resulting in timely whole-aniraal testing where
screening indicates potential hazards. Time and money should thus
be saved in both technological development and biological testing.
In the approach that we have taken with an array of complex
mixtures from the technological world, chemical characterization
and preparation are coupled with short-tern bioassays. Other such
investigations have usually involved chemical analyses and
preparation for biological testing (e.g., cytotoxicity,
mutagenicity, and carcinogenicity assays). The genetic test for
Salmonella histidine-mutant reversion (Ames et al., 1975) has been
widely used to screen potential mutagens or carcinogens.
Subsequent fractionation procedures are carried out to isolate and
identify the mutagens in the material; the bioassay is used to
253
-------
254
MILTON J. CONSTANTIN ET AL.
trace the biological activity and guide the separations. The
biological tests function as 1) predictors of long-range health
effects such as mutagenesis, teratogenesis, or carcinogenesis: 2)
predictors of toxicity to man and his environment: 3) a mechanism
to rapidly isolate and identify hazardous agents in complex
material; and 4) indicators of relative biological activity,
through the correlation of control data with changes in
environmental or process conditions.
Plant systems may offer an additional means of evaluating
complex mixtures. According to Kihlman (1971), root tips are the
ideal plant tissue in which to study the effects of chemicals on
chromosomes because 1) they are readily available throughout the
year; 2) they are inexpensive; 3) they are easy to handle; 4)
they provide data within a few days; 5) they are directly exposed
to the chemical in aqueous solutions of known concentrations; 6)
they provide numerous dividing cells for analysis; and 7) they
have few, relatively large chromosomes.
Barley (Hordeum vulgare [L] emend Lam.) has 2n = 14
chromosomes, ranging from 6 to 8 ym in length. The barley embryo
has from five to seven seminal roots that yield numerous dividing
cells for analysis within 24 to 48 h after germination starts.
Following seed treatments, the effects of chemical and physical
agents can be assessed in terras of cytogenetic and genetic end
points in the same population of plants. This unique advantage has
made barley a useful mutagenesis test organism.
The most frequently used method for studying the effects of
chemicals on the chromosomes of barley root tip cells is the
analysis of anaphases for detectable aberrations (Nicoloff and
Gecheff, 1976). The assay has been used to test chemicals either
suspected or known to be mutagens; generally, the results have
agreed with those from other organisms tested with the same
compounds. The assay has not been used widely to test complex
mixtures.
In this report, we present data from a study comparing
microbial mutagenicity assays (Salmonella and yeast) with a plant
cytogenetic assay. The materials used were aqueous extracts from a
fly ash sample and an arsenic- (As) contaminated groundwater sample
(provided by the U.S. Environmental Protection Agency). The
purpose of this research was to assess the potential of complex
environmental mixtures (extracts of solid wastes) to produce
cytogenetic effects in a higher plant and gene mutations in
microbes. A complex environmental mixture capable of inducing both
end points warrants close scrutiny as a potential health hazard to
humans.
-------
PLANT CYTOGENETICS AND MICROBIAL MUTAGENESIS ASSAYS 255
MATERIALS AND METHODS
Preparation of Liver Hoaogenate (S-9)
The microsomal preparation was made according to Ames et al.
(1975). The livers of Sprague-Dawley rats (induced with either
Aroclor 1254 [Ar S-9] or phenobarbital [i|iB S-9]) were washed in an
equal volume of 0.15 M NaCl, minced with sterile scissors, and
homogenized with a Potter-Elvehjem apparatus in 3 vol of 0.15 M
KC1. The homogenate was centrifuged at 4°C for 10 tnin at 9000 x g.
The supernatant was collected and stored at -80°C. The activation
system (S-9) mix contained, per milliliter, 0.3 ml of liver
homogenate, 8 ymol magnesium chloride, 33 umoi potassium chloride,
5 pmol glucose-6-phosphate, and 4 umol nicotinamide adenine
dinucleotide phosphate in 100 vmol of sodium phosphate buffer (pH
7.4).
Bacterial Mutagenicity Assay
Among bacterial mutagenicity test systems, the assay developed
by Ames using the histidine auxotrophic strains of Salmonella
typhimurium is widely used as a screening test to detect potential
genetic and carcinogenic hazards. Of the four standard tester
strains generally used in the assays, the missense strains TA100
and TA1535 detect base-pair substitutions, while strains TA1537 and
TA98 detect frameshift mutations. The general procedure was
described by Ames et al. (1975). A bacterial suspension (2 x
i08/ml) was added to 2 ml of molten top agar (45°C) containing test
substance. S-9 mix (0.5 ml/plate) was added, when required, to
provide metabolic activation. The top agar was overlayed on
Vogel-Bonner (1956) minimal medium, and revertant colonies were
counted after two days of incubation at 37°C. Known chemical
mutagens as positive controls and solvent (negative) controls were
routinely run.
Yeast Assays
The Saccharomyces assay measures both forward and reverse
mutation. Forward mutation is detected by the inactivation of the
arginine permease gene (CAN1), leading to resistance (canr) to the
toxic antimetabolite canavanine. Reverse mutation is monitored by
use of a histidine auxotroph (his 1-7) chat reverts by base-pair
subs titution.
The test strain used in these experiments was constructed from
stocks of Saccharomyces cerevisiae maintained at Oak. Ridge National
Laboratory (QRNL) and from stocks obtained from the Berkeley
-------
256
MILTON J. CONSTANTIN ET AL.
collection. Strain XL7-10B has the genotype a p+ CAN1 hisl-7
lysl-l ura1.
Details of the preparation of media, rat-liver homogenates,
and the mutagenicity assay have been described by Larimer et al.
(1978, 1980). The test material was mixed with yeast cells in
buffer or rat-liver horaogenate (S—9) in buffer and incubated for 3
or 24 h at 30°C with shaking before it was plated on selective
media. Mutant clones were scored on selective plates after
incubation for five days at 30°C. To determine survival, dilutions
were plated on yeast-extract-peptone-dextrose plates and scored
after two days at 30°C.
Plant Cytogenetic Assay
Seeds of Himalaya barley from R.A. Nilan (Washington State
University, Pullman, WA) were stored under refrigeration and
hand-picked for quality just prior to each experiment.
Approximately 25 seeds were sown erabryo-side-up on Whatman #1
filter paper in glass petri plates. The filter paper was
thoroughly saturated by adding 7.5 ml of either double-distilled
water or a test solution at pH 7.0. Each treatment was done in
triplicate; plates were placed in sealed polyethylene bags and
cultured for 42 to 46 h at 25°C under 10 to 15 uE m-2sec-1
fluorescent light.
Germinating seeds were killed and fixed in 3:1 ethanol-
glacial acetic acid and stored in the refrigerator until the roots
were processed. Excised roots were hydrolyzed for 9 min in 1 N HC1
at 60°C, reacted with Schiff's Reagent for at least 15 min, treated
with pectinase (0.42 in water, pH 4.0) for at least 30 min, and
squashed in aceto-carmine; Deckglaskitt was used to seal around the
coverslip. Cells were observed at 1000X magnification for the
presence of bridges and fragments at anaphase.
Data were expressed as aberrations per hundred anaphases and
as percentage of aberrant anaphases. Statistical inferences were
drawn on the bases of analysis of variance and chi-square analysis
of 2 x 2 contingency tables for aberrant vs. normal anaphases in
control vs. chemical treatments.
Preparation of Samples
Aqueous extracts of the fly ash samples were prepared
according to the extraction procedure (EP) given in the Federal
Register (1978). The As-contaminated sample was used as provided
by the U.S. Environmental Protection Agency. Resin concentration
(XAD-2; Isolab, Inc.) procedures were described by Epler ec al.
-------
PLANT CYTOGENETICS AND MICROBIAL MUTAGENESIS ASSAYS
257
(1980). A 500-nl aqueous extract was passed through A g XAD-2 at a
flow rate of 1.2 ml/rain. The column was rinsed with deionized
water and then eluted with acetone. The acetone eiuate was
evaporated to dryness, and the sample was dissolved in 2 ml
dimethylsulfoxide.
RESULTS AND DISCUSSION
Salmonella Assay
Aqueous samples from the As-contaminated groundwater and its
XAD-2 concentrate (12.5-fold) were tested for mutagenicity over a
nontoxic dose range. The results are given in Figure 1. The
groundwater sample was not mutagenic (Figure 1A), even with
metabolic activation, for strains TA98 and TA100. However, XAD-2
concentrate of groundwater (Figure IB) exhibited a clear dose
response with the fratneshift strains TA1537 and TA98 and the
highly sensitive TA100 strain. The missense strain TA1535 was not
reverted. These results suggest the presence of a mutagenic
constituents) in the aqueous sample whose activity was evident
only when the sample was concentrated by an appropriate method.
The EP extract and its XAD-2 concentrate from fly ash sample
were not mutagenic when tested with the basic set of tester strains
(see Table 1). Addition of the metabolic activation system did not
influence the result. Lack of mutagenic activity of fly ash from
power plants conflicts with earlier reports of Chrisp et al. (1978)
and Kubitschek and Venta (1979), who used serum extraction, and
Hobbs et al. (1979), who used dimethylsulfoxide for extraction.
The EP extraction procedure is probably not adequate to extract
mutagenic agents from fly ash. When benzene extraction was used
with a similar fly ash sample from a different power plant, a
dose-dependent increase in the induction of histidine revertants
was observed (unpublished data). Results obtained with known
chemical mutagens (positive controls) are given in Table IB. The
alkylating agents ethylmethane sulfonate (EMS) and methylraethane
sulfonate (MMS) were very specific in reverting TA1535 and TA100,
respectively. The frameshift strains were reverted by 8-amino-
quinoline (8-AmQ), specific to TA1537, and benzo(a)pyrene (B[a]P),
which is mutagenic only when activated with Aroclor—1254-induced
rat-liver horaogenate (Ar S-9 mix).
Saccharomyces Mutation Assay
The As-contaminated groundwater sample was not mutagenic (see
Table 2). The XAD-2 concentrate of this sample was mutagenic
without metabolic activation for a 24-h exposure, giving a dose-
dependent response. Metabolic activation appeared to reduce the
-------
258
MILTON J. CONSTANTIN ET AL.
8. As-gro jndwcter and XAD-2
Corcentrate
•TA10C
A. As-groLrdwater
-1000
200-
- 30C
160-1
+ S-9
TA100
c TA98
L 600
f 12C-
+ S-9
U
(-400 >
w 80-
LlJ a
TA98
- 200
40-
oo
50
0.5 5.0 0
•0
25
0.C5
CONCENTRATION (ul/PLATE)
Figure 1. Mutagenicity of As-contaminated groundwater and its
XAD-2 concentrate in the Salmonella assay. Each point
represents an average from at least three independent
experiments.
mutagenic potential of the XAD-2 concentrate. Neither of these
test materials was toxic. More induced forward mutations to canr
were seen than hisl-7 base-pair substitutions. This is the typical
response of this system to a frameshifting agent. The fly ash EP
extract and its XAD-2 concentrate were not toxic or mutagenic to
yeast (Table 3).
Plant Cytogenetic Assay
The cytogenetic effects observed in barley root tip cells are
presented as percentage of aberrant anaphases in Table 4, which
shows the distribution of anaphases into normal and aberrant
classes for negative, positive, and solvent controls and complex
environmental mixtures. The probability values were determined
through chi-square tests for independence of the treatments and the
proportions of aberrant anaphases, in 2 x 2 contingency tables.
-------
PLANT CYTOGENETICS AND MICROBIAL MUTAGENESIS ASSAYS
259
Table 1. Salmonella Mutation Assay: Fly Ash—Aqueous
Extract (EP) and Its Concentrate (XAD-2)
his+ Revertants/Plate
TAi 535a TA15373 TA98 TA100
Treatment and
Concentration
(gin/plate) EP EP EP XAD-2 EP XAD-2
No activation
Solvent control
21
14
32
57
136
231
50
18
15
27
30
141
180
• S-9 activation

Solvent control
15
16
46
33
192
209
10
15
14
35
42
108
203
25
8
15
34
28
141
206
50
16
15
53
26
123
206
75
16
18
42
38
172
208
isitive controls*5

50 ul (5%) EMS
704
11
48
-
386
-
50 yl (5%) MMS
42
13
48
-
1495
-
20 yg 8-AmQ
6
102
56
-
270
-
20 ug B(a)P + S-9
7
122
625

1700

aXAD-2 not tested with these strains.
^Underscoring represents positive response to mutagen.
The solvent controls (phosphate buffer and acetic acid extraction
solution) yielded aberrant anaphases at frequencies not
significantly different frora those of the negative control,
distilled water. In contrast, the frequency of aberrant anaphases
was increased in the seeds treated with EMS, a positive control,
and in those treated with the fly ash extract and the As-
contaminated groundwater. The fly ash extract was used as we
received it, except for the adjustment to pH 7.0, whereas the As-
contaminated groundwater was diluted 1:8 and 1:16 with distilled
water (1.25 and 0.625 ml extract diluted to 10 ml of solution),
and these results were pooled. The more concentrated solutions
were toxic to the germinating seed; i.e., germination was delayed
and root growth was inhibited.
-------
bo
cn
o
Z Survival
canr/IO' Survivors
hla'/io' Survivors
3 h
24 h
3 h
24 h
3 h
24 h
Concent rat Ion
(nD
Simple XAD-2
Scrapie XAD-2
Sample XAD-2 Sample XAD-2
Sample XAD-2 Sample XAD-2
No activation
Control
0. I
1.0
10
20
50
100
inn
71
84
75
89
100
I 16
127
138
156
100
94
81
85
99
100
95
98
93
86
27
23
20
13
16
24
19
15
17
17
19
16
19
19
13
15
17
14
69
176
6
16
8
13
13
II
8
8
9
I I
9
Ik
17
55
VB S-9 activation
Contrul
0. I
1.0
10
20
50
100
100
107
59
82
91
100
105
98
103
1 12
100
101
87
103
119
100
94
86
79
71
14
21
27
21
18
17
18
20
16
21
17
14
21
14
18
15
19
31
B9
107
12
10
10
14
12
10
II
13
8
7
12
9
10
9
12
6
29
24
r
H
O
•7,
Ar S-9 activation
Control 100
0. I 9/
1.0 92
10 87
20
50
100 88
EMS, U v/v 22
100
107
1 I I
122
90
100
94
96
91
103
100
108
122
91
82
17
18
13
21
24
448
15
12
25
13
13
21
20
27
22
22
14
13
41
74
118
9
11
8
12
I 1
1270
10
11
5
8
I I
9
7
10
13
8
9
17
38
n
o
7,
CD
H
>
d
x
w
H
>
f
-------
PLANT CYTOGENETICS AND MICROBIAL MUTAGENESIS ASSAYS
261
±
<
X
so i i so n s: o
c* 8N rN O
a.
rt O i/* vC I
9 CB >7 4 1 1 00
Q 00 >9 U*i I
O
"¦N.
+
a'
mm1
Cl 1^489
a* J I o AO «
4 1 I 3) 09 4

— C* ff* J I
^ ® (N I I —
3
V3
5N
A
5
a.
v
I n »r n i"
N

o
r*»

a
ij

a
u

O
>
ffl

CJ
•H

>
o

a
c

CB
0

mm
o

u
MM
o

u
M
©

CJ
Ul
w
¦
•

¦
¦

*J
•
«

•
v
O
e
o
M
©
©
©
©
i
c
©
_
©
©
©
©
i
e
©
t—•
©
©
©
©
8m
1
<0
o

-------
262
MILTON J. CONSTANTIN ET AL.
Table 4. Chromosome Aberrations in Barley Embryos
Treated with Extracts of Complex Environmental Mixtures:
Numbers of Aberrant Anaphases
Aberrant Anaphases
Normal
Treatments
Anaphases
N
N
7.
pa
Distilled water
2182
48
2. 15

Phosphate buffer*5
2197
75
3.30
0.05 to 0.10
EMSC
2241
184
7.59
< 0.001
Acetic acid
3784
88
2.27
0.30 to 0.50
extraction solution^

Fly ash extract
2874
437
13.20
< 0.001
As-contaminated
1221
190
13.47
< 0.001
groundwater

aP =¦ probability that difference was due to
chi-square test for independence.
^Monobasic and dibasic phosphate buffer at
chance,
pH 7.0.
according to
CEMS at 0.025 M in phosphate buffer at pH 7.0; seeds soaked
aerobically in water at ~ 1°C for 16 h and in EMS for 2 h at ~ 1°C
plus 6 h at 24°C, rinsed, and cultured on a water-saturated
Whatman //1 filter.
^Weak acetic acid extraction solution as described in Federal
Register (1978).
Table 5 shows the same experimental data as in Table 4, but
expressed as aberrations per hundred cells (the mean for each
treatment). An arcsine transformation of the data (arcsin p,
where p is a proportion) was done prior to the analysis of
variance, to reduce the heterogeneity of the treatment variances.
A Student-Newman-Keuls comparison of treatment means (Steel and
Torrie, 1960) showed no difference between the negative control
(distilled water) and either of the solvent controls (phosphate
buffer and acetic acid extraction solution), whereas the positive
control (EMS) and the two complex environmental mixtures (fly ash
extract and the As-contaminated groundwater) induced a
significantly greater number of aberrations per hundred anaphases.
Although the two methods of data analysis address different
aspects of the cell population's response to seed treatment, the
conclusion is the same: the barley root tip system responded to an
unknown mutagenic substance(s) in the two complex environmental
-------
PLANT CYTOGENETICS AND MICROBIAL MUTAGENESIS ASSAYS
263
Table 5. Chromosome Aberrations in Barley Embryos
Treated with Extracts of Complex Environmental Mixtures:
Aberrations per Hundred Cells
Mean Number of
Aberrations/ Standard
Treatments3 Embryos 100 Cells*3 Deviation
Distilled water
60
2.75
3. OA
Acetic acid
60
2. 57
2.85
extraction solution

Fly ash extract
60
16,30
9. 10
Phosphate buffer
59
4.37
4.23
EMS (0.025 M)
63
11.01
9.73
As-contaminated
63
16. 19
9. 10
groundwater
aSee footnotes to Table 4 for details.
''Not arcsine-transfonned.
mixtures much as it did to EMS, a known mutagen. This response was
observed as increases in the percentage of aberrant anaphases and
in the number of aberrations per hundred anaphases.
Our results with the barley root tip cytogenetic aberration
assay agree with expectations. According to Brewen and Preston
(1978), structural changes in chromosomes constitute a significant
proportion of mutagenic events. Chemicals that induce mutations in
eukaryotes invariably also induce chromosomal structural changes
(Evans, 1976; Kunzel, 1971). Fly ash from coal combustion is
mutagenic in bacteria; known mutagens have been isolated and
identified. Arsenic and heavy metals are known to induce both
mutations and cytogenetic effects.
Chrisp et al. (1978) reported that horse serum, phosphate-
buffered saline, and cyclohexane filtrates of fly ash of 2.2-um
mass median diameter (the finest particle size fraction tested)
induced histidine revertants in _S. typhlmurlua strains TA98 and
TA1538. The order of activity was horse serum !>> cyclohexane >
saline. More recently, Fisher et al. (1979) reported that serum
filtrates of the most respirable stack-collected fly ash are
mutagenic in _S. typhimurium strains TA98, TA100, and TA1538.
However, after heating to 350°, these serum filtrates are not
mutagenic in the Salmonella assays. The authors hypothesized that
the mutagenic activity of fly ash is associated with organic
compounds. Lee et al. (1980) have found diraethylsulfate and its
-------
264
MILTON J. CONSTANTIN ET AL.
hydrolysis product monoethylsulfate at concentrations as high as
830 ppm in fly ash and airborne particulate matter from coal
combustion. These compounds are known mutagens.
In the case of the As-contaminated groundwater, numerous
metals were present (e.g., in gg/1: cadmium, 485; nickel, 935,
lead, 117; antimony, 297; thallium, 7720, and zinc, 251; Epler et
al., 1980). Arsenic, especially in the arsenite state, is known to
induce mutations and chromosome aberrations (Rossner, 1977). Some
of the heavy metals are known to be mutagenic and/or carcinogenic
(Freese, 1971; Miller and Miller, 1971).
CONCLUSIONS
The following conclusions were reached: 1) The Salmonella and
Saccharomyces assays indicated the presence of mutagenic activity
in the XAD-2 concentrate of the As-contaminated groundwater but not
in the aqueous extract of the fly ash sample. 2) Both assays
implicated frameshift mutagenesis as the mechanism involved. 3)
The Hordeum root tip assay indicated mutagenic activity in both
complex mixtures tested. 4) Chemical analyses of both complex
mixtures showed the presence of heavy metals, implicating them as
the possible cause of chromosomal aberrations.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the staff of the Analytical
Chemistry and the Environmental Sciences Divisions of ORNL who
contributed to the research that led to this comparative study.
REFERENCES
Ames, B.N. , J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian-microsorae mutagenicity test. Mutation Res.
31:347-364.
Brewen, J.G., and R.J. Preston. 1978. Analysis of chromosome
aberrations in mammalian germ cells. In: Chemical Mutagens:
Principles and Methods For Their Detection. Vol. 5. A.
Hollaender, ed. Plenum Press: New York. pp. 127-150.
Chrisp, C.E., G.L. Fisher, and J.E. Lammert. 1978. Mutagenicity
of filtrates from respirable coal fly ash. Science 199:73-75.
-------
PLANT CYTOGENETICS AND MICROBIAL MUTAGENESIS ASSAYS
265
Epler, J.L., F.W. Larimer, T.K. Rao, E.M. Burnett, W.H. Griest,
M.R. Guerin, M.P. Maskarinec, D.A. Brown, N.T. Edwards, C.W.
Gehrs, R.E. Milleraan, B.R. Parkhurst, B.M. Ross-Todd, D.S.
Shriner, and H.W. Wilson, Jr. 1980- Toxicity of Leachates.
Final Report for Office of Research and Development, U.S.
Environmental Protection Agency, Cincinnati, OH. Oak Ridge
National Laboratory: Oak Ridge, TN.
Evans, H.J, 1976. Cytological methods for detecting chemical
mutagens. In: Chemical Mutagens: Principles and Methods For
Their Detection, Vol. 4. A. Hollaender, ed. Plenum Press:
New York. pp. 1-29.
Federal Register. 1978. A3 (Dec.): FR58946.
Fisher, G.L., C.E. Chrisp, and O.G. Raabe. 1979. Physical factors
affecting the mutagenicity of fly ash from a coal-fired power
plant. Science 204:879-881.
Freese, E. 1971. Molecular mechanisms of nutations. In:
Chemical Mutagens: Principles and Methods For Their
Detection, Vol. 1. A. Hollaender, ed. Plenum Press: New
York. pp. 1-56.
Hobbs, C.H., C.R. Clark, L.C. Griffis, R.O. McClellan, R.F.
Henderson, J.O. Hill, and R.E. Royer. 1979. Inhalation
toxicology of primary effluents from fossil fuel conversion
and use. ORNL Publication Conf-780903. Oak Ridge National
Laboratory: Oak Ridge, TN.
Kihlman, B.A. 1971. Root tips for studying the effects of
chemicals on chromosomes. In: Chemical Mutagens: Principles
and Methods For Their Detection, Vol. 2. A. Hollaender, ed.
Plenum Press: New York. pp. 489-514.
Kubitschek, H.E., and L. Venta. 1979. Mutagenicity of coal fly
ash from electric power plant precipitators. Environ.
Mutagen. 1:79-83.
Kunzel, G. 1971. The ratio of chemically induced chromosome
aberrations to gene mutations in barley. Mutation Res.
12:397-409.
Larimer, F.W., A.A. Hardigree, W. Lijinsky, and J.L. Epler. 1980.
Mutagenicity of N-nitrosopiperazine derivatives in
Saccharomyces cerevisiae. Mutation Res. 77:143-148.
Larimer, F.W., D.W. Ramey, W. Lijinsky, and J.L. Epler. 1978.
Mutagenicity of methylated N-nitrosopiperidines in
Saccharomyces cerevisiae. Mutation Res. 57:155-161.
-------
266
MILTON J. CONSTANTIN ET AL.
Lee, M.L., D.W. Later, D.K. Rollins, D.J. Eatough, and L.D. Hansen.
L980. Dimethyl and raonocnethyl sulfate: presence in coal fly
ash and airborne particulate matter. Science 207:186-188.
Miller, E.C., and J.A. Miller. 1971. The mutagenicity of chemical
carcinogens: Correlations, problems, and interpretations.
In: Chemical Mutagens: Principles and Methods For Their
Detection, Vol. 1. A. Hollaender, ed. Plenum Press: New
York. pp. 83-119.
Nicoloff, H., and K. Gecheff. L976. Methods of scoring induced
chromosome structural changes in barley. Mutation Res.
34 :233-244.
Rossner, P. L977. Mutagenic effect of sodium arsenite in Chinese
hamster cell line Dede. Mutation Res. 46 (3):234-235.
Steel, R.G.D., and J.H. Torrie. 1960. Principles and Procedures
of Statistics. McGraw-Hill: New York. p. 481.
Vogel, H.J., and D.M. Bonner. 1956. Acetylornithase of E. coli:
partial purification and some properties. J. Biol. Chem.
218:97-106.
-------
SESSION 4
MOBILE SOURCES
-------
Intentionally Blank Page
-------
SHORT-TERM CARCINOGENESIS AND MUTAGENESIS BIOASSAYS OF
MOBILE-SOURCE EMISSIONS
JoeLlen Lewtas Huisingh
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION
The combustion emissions from mobile sources, including both
gases and particles, are very complex and may have thousands of
separate components. Qualitative and quantitative identification
of all of these individual components is a tremendous task. The
analytical challenge is facilitated if the number of compounds
requiring identification can be reduced.
Short-terra bioassays can be used to narrow the compounds
requiring identification to those potentially responsible for
adverse health effects. Initial screening of complex mixtures is
useful to
1) indicate particular emissions or portions of an emission
that are potentially toxic, mutagenic, or carcinogenic and
that should be evaluated in confirmatory and, possibly,
long-term bioassays;
2) biologically direct the fractionation and identification
of hazardous components and specific chemicals in complex
mixtures; and
3) compare the relative biological activity of similar
emissions that result from different sources, fuels,
control technologies, or operating conditions.
The introduction of increasing numbers of light-duty diesel
automobiles has stimulated environmental concern over the health
effects of diesel particulate emissions. Currently, diesel
269
-------
270
JOELLEN LEWTAS HUISINGH
automobiles erait over one hundred times the particles (grams per
mile) emitted by gasoline-powered, catalyst-equipped (gasoline-
catalyst) automobiles. Diesel particles, emitted as carbonaceous
soot, serve as condensation nuclei for higher-molecular-weight
organic combustion vapors, which condense onto the soot particles
as the exhaust is diluted and cooled to ambient'temperature. The
diesel particles emitted into the ambient air contain 10 to 50%
extractable organic constituents.
The gaseous organics that do not adsorb onto particles are
currently regulated only as total hydrocarbon emissions.
Components of this general class of emissions that are of potential
concern, such as aldehydes, nitrosamines, phenols, and cyanides,
are not specifically regulated. Research to apply short-term
bioassays to these gaseous emissions is being initiated; much of
the research completed to date, however, has focussed on the
organic compounds extracted from diluted particulate emissions.
APPLICATION OF MICROBIAL ASSAYS TO MOBILE SOURCE EMISSIONS
Mutagenic activity resulting from organics extracted from
diesel particulate emissions was first detected using microbial
mutagenesis assays. Particles collected from two heavy-duty diesel
engines were subjected to extraction and fractionation techniques
(Huisingh et al., 1979). The resulting organic fractions (acidic,
basic, and neutral) were then screened using bioassays that
employed bacteria (Salmonella typhimurium) to detect gene mutations
and mammalian cells to detect cellular toxicity. None of the
diesel organic fractions was found to be highly cytotoxic in the
mammalian cell assays. All but one of the fractions showed some
mutagenicity in the S. typhimurium plate-incorporation assay for
gene mutations.
The neutral components of the die3el extract accounted for 84%
of the mass and were fractionated into four subfractions
(paraffins, aromatics, and transitional and oxygenated polar
neutrals). The paraffinic fraction (39% by weight) was not
mutagenic, and the aromatic fraction (13% by weight) accounted for
only 1.5% of the mutagenic activity in the TA98 strain of S.
typhimurium. The two polar neutral fractions, transitional and
oxygenated, were the most mutagenic. These two fractions accounted
for one third of the mass of the extractable organics and over 90%
of the mutagenic activity in both TA98 and TA1538 strains of S^.
typhimurium.
These results suggest that there is more than one mutagen
present in the polar neutral fractions of organics bound to diesel
particles. These mutagens are not artifacts of the extraction or
-------
SHORT-TERM BIOASSAYS OF MOBILE-SOURCE EMISSIONS 271
fractionation processes (Huisingh et al., 1979), but appear Co be
products of the combustion process, since fractions of uncombusted
fuel were not mutagenic.
Various fuels appear to differ in the mutagenicity of their
particle-bound combustion organics. Studies comparing the
mutagenic activity of combustion emission organics from two
passenger cars operated with five different fuels show that the
poorest quality fuel (No. 2 diesel fuel) generated the largest
quantity of mutagenic particle-bound organics (Huisingh et al.,
1979). This minimum-quality fuel had the lowest Cetane index
(41.8), highest aromatic content, and highest nitrogen and sulfur
contents.
The effects of engine, fuel, and operating conditions on the
mutagenicity of automotive emissions were studied using short-term
bioassays. These conditions are variable and may affect not only
the mutagenic activity of the organic fractions but also the amount
of extractable organics present on the diesel particles and the
particulate emission rate. These factors can be accommodated by
calculations that determine the mutagenic activity on a per-mile-
or per-kilogram-fuel-consumed basis.
In comparing different diesel automobiles, Claxton and Kohan
(1980) found as much as a three-fold difference in their mutagenic
emission rate. Although the extractable organics from the
gasoline-catalyst automobile emissions were more mutagenic than
many of the diesel organics, the amount of extractable organics and
the particle emission rate were so low for the gasoline-catalyst
automobile that the net mutagenic activity per mile was two orders
of magnitude less than that from a comparable diesel automobile.
The total mutagenic activity resulting from automotive
emissions depends on the release of the mutagenic organics from the
particles. The ability of physiological fluids (serum, lung-cell
cytosol, and lung-lavage fluid) to release mutagens from diesel
particles has been compared with the extraction capability of
solvents. Serum and lung cytosol were found to remove 80 to 85% of
the solvent-extractab1e mutagenic activity from the diesel
particles (King et al., in press). The serum- and cytosol-
associated mutagens were essentially undetectable when the serum
itself was tested in the j>. typhimurium mutagenesis bioassay. This
effect is possibly due to binding of the mutagens by the serum.
Other studies have shown that whole diesel particles are engulfed
by mammalian cells in vitro and are capable of causing gene
mutations (Chescheir et al., 1980).
-------
272
JOELLEN LEWTAS HUISINGH
APPLICATION OF MAMMALIAN CELL BIOASSAYS TO MOBILE-SOURCE EMISSIONS
The extractable organics from diesel particles, although
showing a low cellular toxicity in the microbial bioassays, were
mutagenic in a microbial (S. typhimurium) assay and positive in a
yeast (Saccharomyces cerevisciae) assay for DNA damage (mitotic
recombination). These results indicated the presence of
potentially mutagenic or carcinogenic chemicals in diesel emission
organics (Huisingh et al., in press b).
Mammalian cell bioassays were initiated to verify the
microbial screening results; mammalian cells are much more similar
to human cells in cellular and chromosomal organization than are
microbes. Diesel organics gave positive results in two forward
mutational assays using mammalian cells. Two assays for DNA
damage—unscheduled DNA synthesis (UDS) and sister chromatid
exchange (SCE) assays—were also used. The UDS assay was negative
and the SCE assay positive with the diesel organics tested. The
carcinogenesis assay for morphological oncogenic transformation in
the mammalian (BALB/c 3T3) cells was positive.
Additional research is needed to determine which bioassays are
most useful in evaluating automotive emissions and to develop new
methods to expose these test systems to "difficult" samples, such
as gases and insoluble organics.
COMPARATIVE BIOASSAYS OF MOBILE SOURCE EMISSIONS
A matrix of in vitro and in vivo bioassays is currently being
used to quantitatively compare the effects of a series of mobile-
source emissions (extractable organics from particulate emissions:
Huisingh et al., in press a). The normalized rankings for four
bioassays are compared in Table 1. The quantitative results from
the mobile-source samples show a general overall consistency
(Nesnow and Huisingh, in press). The Cat sample was very weak in
all of the assays. The Nissan sample showed the highest activity
in these assays, while the other three mobile sources showed
intermediate activity.
In theory, gene mutation and skin tumor initiation arise from
similar mechanisms and, thus, should give similar results (assuming
equal toxicity and mutagen or carcinogen transport and activation
by the various cell types). A comparison of the results of the
microbial and mammalian cell mutation assays with the results of
the rodent skin tumor initiation assay seems to support this
hypothes is.
-------
SHORT-TERM BIOASSAYS OF MOBILE-SOURCE EMISSIONS 273
Table 1. Activity Rankings for Mobile-source Emissions3,^
Light-duty Diesel
Heavy-duty Gasoline-
Diesel catalyst
Activity Cat Nissan Olds VW Rab Mustang
Microb i al
4.3
100
23
22

25
mutat ionc

Sister chromatid
0
100
0
50

1
exchange^

Mammalian cell
1
100
64
50

36
mutat ione

Rodent skin tumor
0
100
45
1

35
initiat ion^

aAll data are exp
ressed as a
percentage
of the
N i s s an
d iesel

activity, which
was assigned
a value of
100.

"Cat is the Caterpillar 3208,
4-stroke cycle engine:
Olds is

Oldsmobile; VW Rab is Volkswagen Rabbit.
CS. typhimurium histidine reversion assay; TA98 with S-9 activation
(Aroclor-induced).
^Chinese hamster ovary cell assay with Aroclor-induced S-9
act ivat ion.
eL5178 mouse lymphoma forward mutation asssay at the thymidine
kinase locus with Aroclor-induced S-9 activation.
^SENCAR mouse assay using TPA (12-0-tetradecanoylphorbol~13-acetate
as the tumor promoter.
Other comparative source samples (roofing tar, coke oven
emissions, and cigarette smoke condensate) were also evaluated in
this study (Huisingh et al., in press a; Nesnow and Huisingh, in
press). The quantitative results for these samples, which required
metabolic activation, showed less agreement between these
bioassays. Thus, it may not be possible to quantitatively
extrapolate from _in vitro to _in. vivo results for all types of
complex mixtures.
CONCLUSIONS
Short-term carcinogenesis and mutagenesis bioassays, now being
widely applied to the evaluation and characterization of mobile
source emissions, show that the organics associated with both
diesel and gasoline-catalyst particulate emissions exhibit
-------
274
JOELLEN LEWTAS HUISINGH
mutagenic and carcinogenic activity. The relative potency of
different mobile sources varies significantly.
Current research is focussing on the following areas: 1)
comparative potency of the emissions from a variety of mobile
sources, 2) comparative evaluation of a battery of bioassays for
mobile-source applications, 3) identification of the hazardous
components in diesel emissions, and 4) determination of the
effective dose and target for those hazardous components.
REFERENCES
Chescheir, G.M., N.E. Garrett, J. Lewtas Huisingh, M.D. Waters, and
J.D. Shelburne. 1980. Mutagenic effects of environmental
particulates in the CHO/HGPRT system. Presented at the U.S.
Environmental Protection Agency Second Symposium on the
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, Williamsburg, VA.
Claxton, L., and M. Kohan. 1980. Bacterial mutagenesis and the
evaluation of mobile source emissions. Presented at Che U.S.
Environmental Protection Agency Second Symposium on the
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, Williamsburg, VA.
Huisingh, J.L., R. Bradow, R. Jungers, L. Claxton, R. Zweidinger,
S. Tejada, J. Bumgarner, F. Duffield, V.F. Simmon, C. Hare, C.
Rodriguez, L. Snow, and M. Waters. 1979. Application of
bioassay Co the characterization of diesel particle emissions.
Part I. Characterization of heavy duty diesel particle
emissions. Part II. Application of a mutagenicity bioassay
to monitoring light duty diesel particle emissions. In:
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures. Plenum Press:
New York. pp. 382-418.
Huisingh, J. Lewtas, R. Bradow, R. Jungers, B. Harris, R.
Zweidinger, K. Cushing, B. Gill, and R. Albert. (in press a).
Mutagenic and carcinogenic potency of extracts of diesel and
related environmental emissions: study design, sample
generation, collection, and preparation. In: Proceedings of
the International Symposium on Health Effects of Diesel Engine
Emissions, December, 1979. U.S. Environmental Protection
Agency: Cincinnati, OH.
-------
SHORT-TERM BIOASSAYS OF MOBILE-SOURCE EMISSIONS
275
Huisingh, J. Lewtas, S. Nesnow, R. Bradow, and M. Waters. (in
press b). Application of a battery of short-term mutagenesis
and carcinogenesis bioassays to the evaluation of soluble
organics from diesel particles. In: Proceedings of the
International Symposium on Health Effects of Diesel Engine
Emissions, December, 1979. U.S. Environmental Protection
Agency: Cincinnati, OH.
King, L., M. Kohan, A. Austin, L. Claxton, and J. Huisingh. (in
press). Evaluation of the release of mutagens from diesel
particles in the presence of physiological fluids. Environ.
Mut agen.
Nesnow, S., and J. Lewtas Huisingh. (in press). Mutagenic and
carcinogenic potency of extracts of diesel and related
environmental emissions: summary and analysis of the results.
In: Proceedings of the International Symposium on Health
Effects of Diesel Engine Emissions, December, 1979. U.S.
Environmental Protection Agency: Cincinnati, OH.
-------
Intentionally Blank Page
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
TUMORIGENESIS OF DIESEL EXHAUST, GASOLINE EXHAUST, AND
RELATED EMISSION EXTRACTS ON SENCAR MOUSE SKIN
Stephen Nesnow
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
Lairy L. Triplett and Thomas J. Slaga
Biology Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee
INTRODUCTION
Recent advances in the study of particulate emissions have
brought to light several facts concerning their health effects.
Many emission sources produce respirable particles with associated
organic substances (Waters et al., 1979). These organic substances
may be unburned fuel or they may result from pyrosynthetic
reactions at or near the combustion source and photosynthetic and
oxidative processes that occur after their initial formation
(Crittenden and Long, 1976). Some of these organic materials
contain known carcinogens and are mutagenic in short-term bioassays
(Huisingh et al., 1979). Previous work by Kotin et al. (1966) and
by Mittler and Nicholson (1957) gave conflicting results on the
mouse skin tumorigenicity of diesel exhaust components. Similar
studies with gasoline exhaust revealed a positive tumorigenic
response from multiple application of condensates and extracts to
mouse skin (Kotin et al. , 1964; Mittler and Nicholson, 1957;
Hoffmann and Wynder, 1963; Hoffmann et al., 1965). The present
study was performed to examine the tumorigenicity of the organics
associated with diesel exhaust particulate emissions using a
sensitive mouse skin tumorigenesis model (SENCAR) and to compare
the tumorigenic potency of the organics from particulate emissions
of diesel, gasoline, and related emission sources.
The SENCAR mouse is a relatively new stock of carcinogen-
sensitive animals, which up to this time has not been used
extensively in bioassay programs. A description of the SENCAR
system and of mouse skin tumorigenesis in general follows, to
explain the strengths and weaknesses of this short-term in vivo
carcinogenesis bioassay.
277
-------
278
STEPHEN NESNOW ET AL.
The SENCAR mouse stock has been selected for its increased
sensitivity to two-stage carcinogenesis using 7 ,12-dinethylbenz(a)-
anthracene (DMBA) as the initiator and 12-0-tetradecanoylphorbol-
13-acetate (TPA) as the promotor. This system is also more
sensitive to other polycyclic aromatic hydrocarbons (PAH) such as
benzo(a)pyrene (B[a]P) (Slaga et al., in press a). In addition to
its well-docuniented response to PAH (Slaga et al . , 1978b), the
mouse skin tumorigenesis bioassay system has identified many
chemicals other than PAH as potential carcinogens (Table 1). These
chemicals represent a wide variety of structural classes, including
aldehyde, carbamate, epoxide, haloalkyether, haloaromatic,
haloalkylcarbonyl, hydroxylamine, lactone, nitrosamide, sulfonate,
sultone, and urea. This list of 32 chemicals includes such well-
known chemical carcinogens as aflatoxin B1, bis(chloromethyl)ether,
chloromethyl methyl ether, urethane, N-acetoxy-2-acetamidofluorene,
3_propiolac tone , N-methyl-N'-nitro-N-nitrosoguanidine ,
1,3-propanesultone, N-nitrosomethyl urea, triethvlenemelamine, and
4-nitroquinoline-N-oxide. The mouse skin tumorigenesis bioassay
can also detect chemicals that cause tumors in the respiratory
tract of animals (Table 2). Of 11 known animal respiratory
carcinogens, the mouse skin tumorigenesis system has to date
detected PAH, quinolines, and carbamates. Of 11 highly suspect
occupational respiratory carcinogens, the mouse skin tumorigenesis
system has to date detected chloromethyl ethers and coke oven
emissions. These results indicate that the mouse skin
tumorigenesis bioassay can detect both dermal and nondermal
carcinogens.
The two basic protocols that can be employed to detect
chemical carcinogens in the mouse skin tumorigenesis assay are
illustrated in Figure 1. Multiple application of the test agent
for up to 60 weeks will give rise primarily to malignant carcinomas
of the skin. This protocol for complete carcinogens is a test for
agents exhibiting both tumor-initiating and tumor-pronoting
activities. The bioassay protocol for tumor initiators is a single
application of test agent followed one week later by multiple
applications of a potent tumor promoter. Tumor initiation is one
step in the multistep carcinogenic process and involves the
conversion of a normal cell into a preneoplastic one. In the case
of chemical carcinogens, it involves the interaction of chemicals
or their activated forms with cellular DNA. These initiated cells
remain dormant for periods of up to one year, or until they are
stimulated to progress into hyperplastic or neoplastic lesions.
This stimulation is called tumor promotion and is accomplished by
applying croton oil or its most active component, TPA. An
initiated cell is, therefore, an irreversibly formed preneoplastic
lesion that can be stimulated to express the transformed phenotype.
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
Table 1. Chemicals Other Than PAH Detected by Mouae Skin Bloassav
279
ClasB
Chemical
Reference
Aldehyde
Carbamate
Epoxide,
dlepoxide
Haloalkylether
Halaaromatic
ttaloalkylcarbony!
Hydroxylamine
Lactone
Multifunctional
Nacural products
Ultrosamide
Sulfonate
Suitone
Urea
Malonaldehyde
Urethane
Vinyl carbamate
Ethyl N-phenylcarbamate
Glyeidaldehyde
1,2,3,4-Olepoxybutane
1 ,2,4 ,5-DlepoicypeDtaae
I,2,6,7-Dleporyheptane
Chloroethylene oxide
Bls(chloronethyl)ether
Chloromethyl methyl ether
2,3,4 ,5-Tetrachlorooitrobenzene
2,3,4,6-Tetrachloronitrobenzeue
2,3,5,6-Tetrachlorocltrobenzene
Pentachloronltrobenzene
Chloroacetone
3-Bromoproplonic acid
N-Acetoxy-4-acetamldoblphenyl
N-AceCoxy—2-acetamldofluorene
N-Hydroxy-2-aminonaphthalene
N-Acetoxy-2-acetoamidophenanthrene
N-( 4-!lechory )benzoyloxyplperldine
H-(4-Nitro)benzoyloxyplperidlne
N-Acetoxy-4-acetamidoetllbene
g-Proplolactone
Trlechylanenelamlne
4-Nltroquinoliae—N—oxide
Aflatoxln B1
N-Mechyl-K' -nitro-N-
nltrosoguanidine
Allyl aethylsulfonate
1,3-Propanesultone
tt-Nltroaomethylurea
Shamberger et al.
1974
Salaaan and Roe, 1933
Slaga et al., 1973
Dahl et al., 1978
Roe and Salaman, 19 55
Shamberger et al., 1974
Vac Duuren et al., 1965
Van Duures et al., 1965
Van Duuren et al., 1965
Van Duuren et al., 1965
Zajdela et al,, 1980
Van Duuren ec al., 1969
Zajdela ec al., 1980
Slaga et al., 1973
Slaga et al., 1973
Van Duuren et al., 1969
Searle, 1966
5earle, 1966
Searle, 1966
Searle, 1966
Searle, 1966
Searle, 1966
Scribner and Slaga, 1975
Scrlbner and Slaga, 1975
Slaga et al., 1978b
Clayson and Gamer, 1976
Scribner and Slaga, 197 5
Scrlbner and Slaga, 1975
Scribner and Slaga, 1975
Scrlbner and Slaga, 1975
Roe and Salaman, 1955
Slaga ec al., 1973
Henniage and Boutvell, 1969
Boe and Salaman, 1955
Hennlngs and Boutwell, 1969
Lindenfelser ec al., 1974
HennlogB et al. , 1978
Fujli, 1976
Roe, 1957
Slaga et al., 1973
Craffl and Hoffman, 1966
-------
280
Table 2.
STEPHEN NESNOVV ET AL.
Response of Carcinogens in Humans, Animals, and Mouse Skin
Occupational Animal Mouse
Respiratory Respiratory Skin
Sample Carcinogen3 Carcinogen3 Tumorigen'-'
Arsenic +

Asbestos +
+

Beryllium +
+

Carbamates
+¦
+
Chlororaethyl ethers +
+
+
Chromium +

Coke oven +

+
Isopropyl oil +

MOCAc +
+

Mustard gas +
+

Nickel +
+

Nitrosamines
+

PAH
+
+
Quinolines
+
+
Vinyl chloride +
+

aFrank, 1978.

^Slaga et al., 1978b, in press: Van Duuren,
1976.

cMethylene bis(ortho-chloroaniline).

Relationships between tumor initiators and complete
carcinogens have been previously described. Various structurally
diverse chemicals (Table 3) are both complete carcinogens and tumor
initiators in mouse skin from CD-I and the genetically related
SENCAR mouse. Some agents, however, appear to have only tumor-
initiating activities in mouse skin (Table 4). The correlation
between potencies as complete carcinogens and as tumor initiators
is excellent for the 12 chemicals that show both kinds of activity
(Table 5). The relationship between the production of papillomas
and the production of carcinomas in the same animals treated with
the skin tumor initiator DMBA or B(a)P is shown in Table 6. These
results indicate that the number of papillomas per mouse at 15 to
20 weeks correlates well with the number of malignant carcinomas
formed at 50 weeks, for animals treated with these two strong skin
tumor initiators.
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
281
PROTOCOLS
TEST
AGENT
TUMOR INITIATION
TUMOR PROMOTION
COCA RCINOGEN ESI S
COMPLETE CARCINOGENESIS
B-A-P
t
T^T
TEST
AGENT
b-a-p r
~ L
j * i
J !_
TPA, 2 x WEEKLY
t r~
_L
r
2 !
'TEST AGENT, WEEKLY'
T
T
_L
1
J TPA, 2* WEEKLY
T
T
1 ! 1
i
1
i
i i
{TEST AGENT, WEEKLY'
1
I
T
_! 2 f !
i
i
WEEK OF EXPERIMENT
I
I
SCORE
FOR
PAPILLOMAS
SCORE
FOR
CARCINOMAS
Figure 1. Protocols for bioassays of test agents as tumor
initiators, tumor promoters, cocarcinogens, and complete
c arc inogens.
Table 3. Compounds That Are Both Complete Carcinogens and
Tumor Initiators in CD-I and SENCAR Mouse Skina
7,12-Dimethylbenz(a)anthracene
3-Methylcholanthrene
BenzoC a)pyrene
7-Methylbenz(a)anthracene
Dibenz(a,h)anthracene
5-MethyIchrysene
B-Propiolac tone
Bis(chloromethyl)ether
2-Hydroxybenzo(a)pyrene
Benzo(a)pyrene-7 ,8-oxide
BenzoC a)pyrene-7,8-d iol
7 ,12-Dimethylbenz(a)anthracene-
3,4-diol
aHecht et al., 1979; SIaga et al., 1978b , in press b.
-------
282 STEPHEN NESNOVV ET AL.
Table 4. Agents That May Be Pure Tumor Initiators in Mouse Skin3
Benzo(a)pyrene-7, 8-d iol-9,10-epoxide Dibenz(a,c)anthracene
N-Methyl-N'-ni tro-N-nitrosoguanidine Chrysene
Benz(a)anthracene-3 ,4-diol-l,2-epoxide Urethane
Benz(a)anthracene Triethylenemelamine
aScribner, 1973; Scribner and Slaga, 1975; Slaga st al., 1973,
1978a, 1979; Van Duuren, 1976.
Table 5. Comparison of Complete Carcinogenesis and
Tumor Initiation in Mouse Skin
Relative Potency3
Complete Carcinogenesis Tumor Initiation
Compound (carcinomas) (papillomas)
7 ,12-Dimethylbenz(a)anthracene
100
100
3-Methylcholanthrene
50
50
Benzo(a)pyrene
30
30
2-Hydroxybenzo(a)pyrene
30
30
7-Bromomethyl-12-

methylbenz(a)anthracene
20
20
Benzo(a)pyrene-7,8-oxide
20
20
Dibenz(a,h)anthracene
20
20
Benz(a)anthracene
5 ± 5
5
Dibenz(a,c)anthracene
0
3
Pyrene
0
0
Benzo(a)pyrene-4,5-oxide
0
0
Anthracene
0
0
aRelative potency was determined from dose-response data. DMBA was
given a maximum value of 100 (Slaga et al., in press b) .
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
283
Table 6. Dose-response Studies on the Ability of DMBA and B(a)P
Co Initiate Skin Tumors in SENCAR Mice3
No. of Papillomas 7. of Mice % of Mice
Dose per Mouse With Papillomas With Carcinomas
Initiator (nmol) at 15 Weeks*5 at 15 Weeks at 50 Weeks^5
DMBA
100.0
22.0
(100)
100
100

DMBA
10.0
6.8
(32)
100
40

DMBA
1.0
3.2
(15)
93
22

DMBA
0.1
0.5
(2)
20
5

B(a)P
200.0
7.5
(100)
100
55
(100)
CP
100.0
3.2
(43)
78
30
(55)
B(a)P
50.0
1 .4
(19)
60
18
(33)
aMice were treated one week after initiation with twice weekly
applications of 5 ug TPA.
^Values in parentheses represent percent normalized to the highest
dose tested of each agent (Slaga et al., in press b).
MATERIALS AND METHODS
Sample Generation and Isolation
The details of sample generation and isolation have been
reported elsewhere (Huisingh et al., in press). Briefly, the
mobile-source samples consisted of particulate emissions from two
diesel-fueled vehicles, one gasoline-fueled vehicle, and one diesel
engine (Table 7); a heavy-duty Caterpillar 3304 engine mounted on
an engine dynamometer at 2200 rpm steady state with an 85-lb load;
a Datsun-Nissan 220-C; an Oldsraobile 350; and a 1978 Mustang 11-302
V-8 catalyst engine (with emission controls and using unleaded
gasoline) mounted on a chassis dynamometer with a repeated highway
fuel economy cycle of 10.24 mi, an average speed of 48 raph, and a
running time of 12.75 min. The Caterpillar, Datsun-Nissan, and
Oldsraobile engines were fueled with the same batch of No. 2 diesel
fuel. Particulate samples were collected using a dilution tunnel
in which the hot exhaust was diluted, cooled, and filtered through
Pallflex Teflon-coated fiberglas filters.
The comparative sources employed were cigarette smoke
condensate, coke oven samples, and roofing tar emissions.
Cigarette smoke condensate was obtained by condensing smoke from an
85-nnn nonfilter Kentucky reference cigarette 2R1. Condensate was
collected in acetone and refrigerated in Dry-Ice-isopropanol bath.
Cigarette smoke condensate acetone suspension was adjusted with
-------
284
STEPHEN NESNOW ET AL.
Table 7. Mobile Source Sample Generation
Dr iving
Sample Description Fuel Cycle
Diesel
Cat
Caterpillar 3304
Diesel
No.
2
Mode IIa
Nissan
Nissan Datsun 220C
Diesel
No .
2
HWFETb
Olds
Oldsraobile 350
Diesel
No .
2
HWFET
Gasoline
Mustang 1978 Mustang, II- Unleaded HWFET
302, V-8 catalyst gasoline
and EGR
aMode II cycle was conducted at 2200 rpm steady state with an 85-lb
1 o ad.
^Highway fuel economy cycle (HWFET) was a 10.24-mi cycle averaging
48 raph and caking 12.75 min.
appropriate amounts of acetone and water. Coke oven samples were
collected from the top of a coke oven battery ac Republic Steel,
Gadscon, AL, using Che Massive Air Volume Sampler. Due to local
wind conditions, various types of aerosols were sampled; thus, an
unknown but significant portion of the emission sample may have
been from the urban environment. The roofing tar emission sample
was collected using a conventional Car pot with external propane
burner. Pitch-based tar was heated to 360 to 380°F (182 to 193'C),
and emissions were collected using a 6-ft (1.8-m) stack extension
and Teflon socks in a baghouse.
The mobile source, coke oven, and roofing tar emission samples
were Soxhlet-extracted with dichloromethane. The dichloromethane
was removed by evaporation under dry nitrogen, and the samples were
shipped in coded form in dry ice to Oak Ridge National Laboratories
where the animal experiments were conducted. Table 8 shows the
amount of organic material extracted from the particles with
dichloromethane and the amount of B(a)P per milligram extract or
per milligram particle in each sample. The B(a)P analysis was
performed according to the method of Snook et al. (1976) or Swanson
et al. ( 1978). Percent extractable of organic material from the
particles varied from 8% of the Nissan sample to a maximum of 99%
for the roofing tar sample. Since cigarette smoke condensate was
not a particulate sample per se, the complete sample was used in
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN 285
Table 8. Benzo(a)pyrene Analysis3
Sample
Extractable (%)
B(a)P
(ng/ng
extract)
B( a)P
(ng/mg
particle)
Diesel:

Cat
27
2
0.5
N i s s an
8
1173
96 .2
Olds
17
2
0.4
Gaso 1 ine:

Mustang
43
103
44.1
Coraparat ive

Sources:

Cigarette
—
<1
—
Coke
7
478
31.5
Roofing tar
>99
889
889
aB(a)P analysis was performed according Co Swanson ec al. ( 1978) ,
except for analysis of cigarette smoke condensate, which was
performed according to Snook et al. (1976).
the biological analysis. B(a)P in the extracts varied from less
than L ng/tng extract for the cigarette sample to a high of 1173
ng/mg extract for the Nissan sample.
Animals
SENCAR mouse stock, selected for its increased sensitivity to
carcinogenesis (Boutwell, 1964) was used in this study. These mice
were derived by breeding Charles River CD-I mice with male STS
(skin-tumor-sensitive) mice that were originally derived from
Rockland mice. Mice were selected for sensitivity to the DMBA-TPA
two-stage system of tumorigenesis for eight generations. These
mice were initially obtained from Dr. R. Boutwell (McArdle
Laboratory for Cancer Research, University of Wisconsin, Madison,
Wl) and now are being raised at the Oak Ridge National Laboratory,
Oak Ridge, TN.
-------
286
STEPHEN NESNOW ET AL.
Chemicals
TPA was obtained from Dr. P. Borchert (University of Minnesota
Minneapolis, MN) and B(a)P from Aldrich Chemical Co. All the
agents were prepared under yellow light immediately before use and
applied topically in 0.2 ml of spectral-quality acetone.
Tumor Experiments
These studies employed 80 mice per treatment group (40 of each
sex). All the mice were shaved with surgical clippers two days
before the initial treatment, and only those mice in the resting
phase of the hair cycle were used. Five dose levels were used for
the tumor-initiating activities of the various samples, except for
Che Mustang sample, which was tested at four dose levels. B(a)P
was used as the standard for the tumor-initiation studies, using
four'dose levels. One week after application, the tumor promoter
TPA was administered twice weekly. All samples at all doses were
applied as a single treatment, except for the 10-mg dose, which was
administered in five daily doses of 2 rag. Skin tumor formation was
recorded weekly, and papillomas greater than 2 ram in diameter were
included in the cumulative total if they persisted for one week or
longer. Both the number of mice with tumors and the number of
tumors per mouse were determined and recorded weekly. Papillomas
and carcinomas were removed randomly for histological verification.
RESULTS AND DISCUSSION
The organic extracts from particulate emissions described
previously were applied to the backs of SENCAR mice according to
Che protocols cited in Materials and Methods. The production of
benign papillomas on a weekly basis is depicted in Figure 2 for
both the reference standard B(a)P and the Nissan sample. In both
cases, after a 7-to 8-week latency period, the percenC of animals
bearing tumors rose dramatically between weeks 8 to 14, with a 95
to 100% tumor incidence observed in both of these dose groups.
Mean number of papillomas per mouse began to rise from control
between weeks 6 to 8, increasing much more slowly than did the
number of animals with tumors. A plateau was reached during weeks
22 to 25. In both cases, the numbers of papillomas per animal
ranged from five to six.
3(a)P exhibited a linear dose response between 2.52 and 100.92
ug (10 to 400 nmol) in both male and female SENCAR mouse skin
(Figure 3). The males seemed to be more sensitive than the females
to this carcinogen, although this sex difference was not evident
for the complex mixture samples evaluated. The most active sample
tested in this series was the coke oven extract. The response to
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN 287
:-o

i»i<
C 1 2 3 a 5 6 1 B J '0 'J '3 M '5 16 '< -9 19 20 2^ 22 23
WEE <5
¦°i I 1 : i I i l i ; i : i i m ( • ; i i i j i )
j
L
a
3
e
S 5
2
C
1
2 \3
} i i * s y s !4 •*. ><¦ 'inw ?' j: j* ?:• js
WEEKS
I
| [ ]
1 >i- 1
! (•
< —
2
r
=t"
: I I I I : ; I I I ; 111''1 i_i
' 1 Z 4 5 a : B 5 «:t •? D '< ^ 'I It 'I H K II JI 73 74 75
WESKS
Figure 2. SENCAR mouse skin-tumor initiation. Male SENCAR
(40) were initiated with either a single dose of B(a)P
(50.4 ug) or five daily treatments with Nissan extract
(2 rag). Animals were then treated biweekly with TPA (2
pg). Left: B(a)P. Right: Nissan extract.
this sample in both male and female animals was biphasic. An
initial linear dose response was observed between 0.1 and 2 rag
extract, with animals carrying an average of five to six
papillomas. The roofing tar extract and Nissan extract (Figure 4)
also produced a large tumor response in both male and female
animals.
The Oldsraobile sample exhibited a linear dose response up to 1
tog and a subsequent loss of activity at 10 mg. The response to the
Oldsraobile sample was one tenth that for the Nissan, coke oven, and
roofing tar samples. The gasoline-fueled Mustang II sample also
produced a weak response in both male and female animals. The
"goodness of fit" (R2) to the linear regression analysis for the
female animals was extremely low, 0.686, indicating a lack of
-------
288
STEPHEN NESNOW ET AL.
. » "« !—
*0C —
BENZC U) PVR6N5, yg
3ENZC (a) PVR5NE,
OK= OVEN :X RAC ,
QKE OVEN =X^RACr rig
J 00
:e
M
3 00i—
:o

9 00
3 oo
ROO= NG TAR EXTRACT, my
ROOs'NG "AR EX'RA;
mg
Figure 3. SENCAR mouse skin tumor initiation dose-response plots
(mean number of papillomas per mouse, after subtracting
background level). Graphs on left are for males and
those on right are for females. There were 40 animals
per dose group. The numbers of surviving animals at
scoring were as follows: 3(a)P—males, 156: females,
156; coke oven extract—males, 195; females 197; roofing
extract—males, 197; females, 196.
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
289
= « 00 -
• 3.00 -
0 w
:oo
o 00
3 CC
200
NISSAN extract, -n,q
NISSAN EXTRAC. ng
3 50
20
C 10
•: :c
200
0-05W03ILE cXTRAC"", r-ig CLDSf/OBILE EXTRACT, -ng
=V
C
o
a.
Z
<
jj
c jc Qu
MUSTANG EXTRACT, mg MLSTANG EXTRACT, nc
Figure 4. SENCAR mouse skin tumor initiation dose-response plots
(mean number of papillomas per mouse, after subtracting
background level). See Figure 3 caption for
explanation. Nissan extract—males, 190; females, 198:
Oldsmobile extract—males, 156: females, 157: Mustang
extract—males, 188; females, 195.
-------
290
STEPHEN NESNOW ET AL.
linear dose response. The Caterpillar sample and cigarette smoke
condensate produced two to three times the numbers of tumors found
in the controls (Figure 5). However, there was no observable dose
response for the doses tested (0.1 to 10 rag). The lack of activity
of the cigarette smoke condensate was disappointing, although not
unexpected. Cigarette smoke condensate when applied to female ICR
Swiss mice twice weekly produced tumors only at relatively high
doses (Gori et al. , 1977; Wynder and Hoffmann, 1967). It was
expected that the increased sensitivity of SENCAR mice to
carcinogens would allow tumors to be observed after treatment with
10 mg whole-smoke condensate. However, this was not the case.
Cigarette smoke condensate is not an extract of isolated
particulates but a suspension of organics, particles, and
volatiles. Therefore, it has not been concentrated to the same
extent as the other samples. The detectability limit of the SENCAR
mouse skin tumorigenesis assay is above the doses and
concentrations tested of the cigarette smoke condensate.
The formation of spontaneous tumors in animals treated with
acetone and promoted twice weekly with TPA was 0.08 and 0.05
papillomas/mouse in male and female animals, respectively, at 22
weeks after initiation; 7 to 8% of the animals had tumors. Animals
initiated with up to 100.92 yg of B(a)P, followed by promotion with
acetone alone, did not produce tumors.
A preliminary analysis of the results obcained was performed
using a linear regression statistical analysis to produce potencies
in terras of papillomas per animal per milligram agent. The results
of these calculations are found in Table 9. The R2 (goodness of
fit) of the data to the linear response was greater than 0.920 for
8 out of 12 of the test groups and greater than 0.84 for 11 out of
12. Potency values ranged from 0 to 101 papilloraas/mouse/rag
agent. The higher of these values was obtained from the B(a)P
treatment groups and was an extrapolation from the microgram dose
range, where the data was obtained, to the milligram range.
Obviously, this number is theoretical and based on strict linearity
throughout a 1000-fold dose range, an assumption not yet proven.
Also, it is a physical impossibility to have 100 papillomas on the
back of a mouse. However, for comparative purposes, these values
give a good approximation of the true values. A relative ranking
of each of the test groups to each other after normalizing to the
Nissan sample is also found in Table 9. The ranking indicates
that the potency of B(a)P was greater than that of the coke oven
sample, which was in turn greater than those of the roofing tar and
Nissan samples. The potencies of these samples were greater than
those of the Oldsmobile and Mustang samples. All of these samples
were greater in potency than were cigarette smoke condensate and
the Caterpillar sample, whose potencies were not significantly
different from zero.
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
291
AR6TTC SMOKE CONDENSATE
Figure 5. SENCAR mouse skin tumor initiation dose-response plots
(mean number of papillomas per mouse, after subtracting
background level). See Figure 3 caption for
explanation. Caterpillar extract—males, 196: females
191; cigarette smoke condensate—males, 187; females,
194.
The results presented here confirm and expand the earlier
observations by Kotin et al. (1966) on the turaorigenesis of diesel
exhaust components and clearly indicate the tumorigenic potential
of these materials. The results also indicate a range of response
of diesel engines, presumably due to differences in engine
technology.
Comparison of the tumor data in Table 9 with the B(a)P
content per milligram extract in Table 8 indicates a lack of
correlation between the two parameters. This result suggests that
B(a)P and possibly other associated PAH are not reliable markers
for tumorigenic activity in these complex mixtures, and that other
non-PAH chemicals in the mixtures make major contributions to their
overall potency.
-------
292

STEPHEN NESNOW ET AL.
Table 9. SENCAR
Mouse Skin Tumor
Initiation:
Sample Rankings3
Sample
Papillomas/
Mouse/mg
R2
Relative Ranking
BenzoC a)pyrene
101 (M)
71.1 (F)
0.999
0.979
20000
13000
Coke oven
2.00 (M)
1.65 (F)
0.960
0.922
400
310
Roofing tar
0.640 (M)
0.571 (F)
0.975
0.977
130
110
Ni ssan
0.532 (F)
0.507 (M)
0.991
0.998
100
100
Olds
0.148 (F)
0.135 (M)
0.896
0.844
28
27
Mustang
0.097 (F)
0.073 (M)
0.686
0.842
18
14
Cigaret te
0
—
0
Caterpillar
0
—
0
aA linear regression model was applied to Che individual daea
points to obtain both slope potency and R2. M and F refer to
results from male and female animals, respectively.
In conclusion, the SENCAR mouse skin tumorigenesis bioassay
for tumor initiation is a quantitative short-term in vivo rodent
carcinogenesis system Chat detects a variety of structurally
diverse chemical carcinogens. This bioassay system has also
shown its utility in evaluating complex environmental mixtures for
tumorigenic potential. It gives excellent dose responses with both
pure substances and complex mixtures and has shown utility for
comparative potency analysis. Additional statistical models are
being evaluated to analyze this data, and the results will be
reported elsewhere.
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
293
ACKNOWLEDGMENTS
The authors wish to thank R.L. Bradow, R.H. Jungers , B.D.
Harris, T.O. Vaughan, R.B. Zweidinger, K.M. Cushing, J. Bumgarner,
and B.E. Gill for Che sample isolation and characterization, and
Carol Evans for the ADP programming. The research was sponsored by
the U.S. Environmental Protection Agency, contract no. 79D-X0526,
under the Interagency Agreement, U.S. Department of Energy no.
40-728-78, and the Office of Health and Environmental Research,
U.S. Department of Energy, under contract no. 7405 eng-26 with the
Union Carbide Corporation.
REFERENCES
Boutwell, R.K. 1964. Some biological aspects of skin
carcinogenesis. Progr. Exp. Tumor Res. 4:207-250.
Clayson, D.B., and R.C. Garner. 1976. Carcinogenic aromatic
amines and related compounds. In: Chemical Carcinogens.
C.E. Searle, ed. American Chemical Society: Washington, DC.
pp. 366-461 .
Crittenden, B.D., and R. Long. 1976. The mechanisms of formation
of polynuclear aromatic compounds in combustion systems. In:
Polynuclear Aromatic Hydrocarbons. R. Freudenchal and P.W.
Jones, eds. Raven Press: New York. pp. 209-223.
Dahl, G.A., J.A. Miller, and E.C. Miller. 1978. Vinyl carbamate
as a proraucagen and a more carcinogenic analog of ethyl
carbamate. Cancer Res. 38:3793-3804.
Frank, A.L. 1978. Occupational lung cancer. In: Pathogenesis
and Therapy of Lung Cancer. C.C. Harris, ed. Marcel Dekker,
Inc.: New York. pp. 25-51.
Fujii, M. 1976. Carcinogenic effect of N-butyl-N-nitrosourechane
on CDF-1 mica. Gann 67:231-236.
Gori, G.B., ed. 1977. Toward Less Hazardous Cigarettes: The
First Set of Experimental Cigarettes. Department of Health,
Education, and Welfare publication no. 76-905. National
Institutes of Health: Washington, DC.
Graffi, A., and F. Hoffman. 1966. Strong carcinogenic effect of
methl nitrosourea on mouse skin in the drop test. Acta 3iol.
Med. Ger. 16:Kl-K3.
-------
294
STEPHEN KESNOW ET AL.
Heche, S.S., E. LaVoie, R. Mazzarese, N. Hirota, T. Ohmori, and D.
Hoffmann. 1979. Comparative mutagenicity, tumor-initiating
activity, carcinogenicity and in vitro metabolism of
fluorinated 5-methylchrysenes . J. Natl. Cancer Inst.
63:855-861 .
Hennings, H., and R.K. Boutwell. 1969. Inhibition of DNA
synthesis by initiators of aouse skin turaorigenesis. Cancer
Res. 29:510-514.
Hennings, H., B. Michael, and E. Patterson. 1978. Croton oil
enhancement of skin tumor initiation by MNNG. Proc. Soc. Exp.
Biol. Med. 158:1-4.
Hoffmann, D. , E. Theisz, and E.L. Wynder. 1965. Studies on the
carcinogenicity of gasoline exhaust. J. Air Pollut. Contr.
Assoc. 15:162-165.
Hoffmann, D., and E.L. Wynder. 1963. Studies on gasoline engine
exhaust. J. Air Pollut. Contr. Assoc. 13:322-327.
Huisingh, J. Lewtas. (in press). Mutagenic and carcinogenic
potency of extracts of diesel and related environmental
emissions: preparation and characterization of the samples.
Proceedings of the International Symposium of Health Effects
of Diesel Engine Emissions.
Huisingh, J., R. 3radow, R. Jungers, L. Claxton, R. Zweidinger, S,
Tejada, J. Bumgarner, F. Duffield, M. Waters, V. Simmon, C.
Hare, C. Rodriguez, and L. Snow. 1979. Application of
bioassay to the characterization of diesel particle emissions.
In: Application of Short-term Bioassays in the Fractionation
and Analysis of Complex Environmental Mixtures. M.D. Waters,
S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
Plenum Press: New York. pp. 383-418.
Kotin, P., H.L. Falk, and M. Thomas. 1964. Aromatic hydrocarbons,
II. Presence in the particulate phase of gasoline-engine
exhausts and the carcinogenicity of exhaust extracts. AHA
Arch. Ind. Hyg. Occup. Med. 9:164-177.
Kotin, P., H.L. Falk, and M. Thomas. 1966. Aromatic hydrocarbons,
III. Pr esence in the particulate phase of diesel-engine
exhausts and the carcinogenicity of exhaust extracts. AMA
Arch. Ind. Hyg. Occup. Med. 11:113-120.
Lindenfelser, L.A., E.B. Lillehoj, and H.R. Burmeister. 1974.
Aflatoxin and trichothecene toxins: skin tumor induction and
synergistic acute toxicity in white mice. J. Natl. Cancer
Inst. 52:113-116.
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
295
Mittler, S., and S. Nicholson. 1957. Carcinogenicity of
atmospheric pollutants. Ind. Med. Surg. 26:135-138.
Roe, F.J.C. 1957. Tumor initiation in mouse skin by certain
esters of methane sulfonic acid. Cancer Res. 17:64-67.
Roe, F.J.C., and M.H. Salaman. 1955. Further studies on
incomplete carcinogenesis: triethylene raelamine (TEM),
1,2-benzanthracene, and g-propiolactone as initiators of skin
tumor formation in the mouse. Br. J. Cancer 9:177-203.
Salaman, H.H., and F.J.C. Roe. 1953. Incomplete carcinogens:
Ethyl carbamate (urethane) as an initiator of skin tumor
formation in the mouse. Br. J. Cancer 7:472-481.
Scribner, J.D. 1973. Tumor initiation by apparently
noncarcinogenic polycyclic aromatic hydrocarbons. J. Natl.
Cancer Inst. 50:1717-1719.
Scribner, J.D., and T.J. Slaga. 1975. Tumor intiation by
N-acetoxy derivatives of piperidine and N-arylacetamides. J.
Natl. Cancer Inst. 54:491-493.
Searle, C.S. 1966. Tumor initiatory activity of some
chloromononitrobenzenes and other compounds. Cancer Res.
26:12-17.
ShaTT.berger, R.J., T.L. Andreone, and C.E. Willis. 1974.
Antioxidants and cancer IV. Initiating activity of
malonaldehyde as a carcinogen. J. Natl. Cancer Inst.
53:1771-1773.
Slaga, T.J., G.T. Bowden, B.G. Shapas, and R.K. Boutwell. 1973.
Macromolecular synthesis following a single application of
alkylating agents used as initiators of mouse skin
tumorigenesis. Cancer Res. 33:769-776.
Slaga, T.J., W.M. Bracken, A. Viaje, D.L. Berry, S.M. Fischer, D.R.
Miller, W. Levin, A.H. Conney, H. Yagi, and D.M. Jerina.
1978a. Tumor initiating and promoting activities of various
benzo(a)pyrene metabolites in mouse skin. In: Polynuclsar
Aromatic Hydrocarbons. R.I. Freudenthal and W. Jones, eds.
Raven Press: New York. pp. 371-382.
Slaga, T.J., S.M. Fischer, L. Triplett, and S. Nesnow. (in
press a). Comparison of complete carcinogenesis and tumor
initiation in mouse skin: Turaor initiation promotion, a
reliable short term assay. In: Proceedings of the
Symposium on Short-term In Vivo Carcinogenesis Bioassays.

-------
296
STEPHEN NESNOW ET AL.
Slaga, T.J., G.L. Gleason, J. DiGiovanni, D.L. Berry, M.R. Juchau,
and R.G. Harvey. 1979. Tumor initiating activities of
various derivatives of benz(a)anthracene and
7,12-dimethylbenz(a)anthracene in mouse skin. In:
Proceedings of the Third International Battelle Conference on
Polynuclear Aromatic Hydrocarbons. P.W. Jones and P. Leber,
eds. Ann Arbor Press: Ann Arbor, MI. pp. 753-764.
Slaga, T.J., A. Sivak, and R.K. Boutwell, eds. 1978b.
Carcinogenesis: A Comprehensive Survey, Vol. 2. Mechanisms
of Tumor Promotion and Cocarcinogenesis. Raven Press: New
York.
Slaga, T.J., L.L. Triplett, and S. Nesnow. (in press b) .
Mutagenic and carcinogenic potency of extracts of diesel and
related environmental emissions: two-stage carcinogenesis in
skin tumor sensitive mice (SENCAR). Proceedings of the
International Symposium on Health Effects of Diesel Emissions.
Snook, M.E., R.F. Severson, H.C, Higraan, R.F. Arrendale, and O.T.
Chortyk. 1976. Polynuclear aromatic hydrocarbons of tobacco
smoke: Isolation and identification. Zeit. zur Tabak. pp.
250-276.
Swanson, D., C. Morris, P. Hedgecoke, R. Jungers, R. Thompson, and
J. Bumgarner. 1978. A rapid analytical procedure for the
analysis of benzo(a)pyrene in environmental samples. In:
Trends in Fluorescence, Vol. 1, No. 2. Perkin-Elmer Corp.:
Norwalk, CT. pp. 22-27.
Van Duuren, B.L. 1976. Tumor-promoting and co-carcinogenic agents
in chemical carcinogenesis. In: Chemical Carcinogens. C.E.
Searle, ed. American Chemical Society: Washington, DC. op.
24-51.
Van Duuren, B.L., L. Orvis, and N. Nelson. 1965. Carcinogenicity
of epoxides, lactones and peroxy compounds, Part II. J.
Natl. Cancer Inst. 35:707-717.
Van Duuren, B.L., A. Sivak, B.M. Goldschmidt, C. Katz, and S.
Melchionne. 1969. Carcinogenicity of haloethers. J. Natl.
Cancer Inst. 43:481-486.
Waters, M.D., S. Nesnow, J.L, Huisingh, S.S. Sandu, and L. Claxton,
eds. 1978. Application of Short-term Bioassavs in the
Fractionation and Analysis of Complex Environmental Mixtures.
Plenum Press: New York.
-------
TUMORIGENESIS OF EMISSION EXTRACTS ON MOUSE SKIN
297
Wynder, E.L., and D. Hoffmann, eds. 1967. Tobacco and Tobacco
Smoke. Academic Press: New York. pp. 133—315.
Zajdela, F., A. Croisy, C. Malaveille, L. Toraatis, and H. BarCsch.
1980. Carcinogenicity of chloroethylene oxide, an ultimate
reactive metabolite of vinyl chloride, and
bis(chlororaethyl)ether after subcutaneous administration and
in initiation-promotion experiments in mice. Cancer Res.
40:352-356.
-------
Intentionally Blank Page
-------
BACTERIAL MUTAGENESIS AND THE EVALUATION OF MOBILE-SOURCE
EMISSIONS
Larry Claxton and Mike Kohaxi
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION'
Interest In developing a rapid, inexpensive means of detecting
and evaluating the potential health hazards of mobile-source
emissions is increasing. Faced with the staggering numbers of
chemicals created through combustion processes that have never been
assayed for mutagenicity and/or carcinogenicity, the chemist faces
a futile task of identifying and controlling all potential health
hazards. This study will demonstrate how bioassay techniques, and
particularly the Salmonella assay, can be coupled with the
fractionation of chemically complex emissions to identify
components requiring more extensive analysis and control.
Emission source organics vary with factors such as time, fuel,
and environmental conditions. In this study, several variables
influencing mobile source studies are examined: day-to-day
variation from a single source; variation among several vehicles of
the same make, model, and configuration; and variation between
different light-duty mobile sources. Also explored are the effects
of storage and the creation of artifacts during the initial
collection of sample. By understanding certain characteristics and
uses of the available bacterial strains, the investigator can use
microbial bioassays to aid in identification of mutagens in complex
mixtures, characterize and compare the types of mutagenic
components within complex mixtures, and screen various mobile
sources for levels of mutagenic compounds.
299
-------
300
LARRY CLAXTON AND MIKE KOHAN
MATERIALS AND METHODS
Bioassays
The primary mutation assay used was the Salmonella typhimurium
plate-incorporation assay as described by Ames et al. Cl975).
However, the test protocol had the following minor modifications:
1) minimal histidine was added to the base agar in the petri dish
rather than to the soft-agar overlay; 2) plates were counted at 48
and 72 h to provide an additional check for toxicity factors:
3) colony counting was performed with an Artek automatic colony
counter; and 4) when adequate sample was available, each dose was
done in triplicate. Microsomal activation was provided using a
9000 x g supernatant of Aroclor-1254-induced Charles River CD-I
rats, as described by Ames et al. (1975).
Six indicator strains of S. typhimurium were used: TA98,
TA100, TA1537, TA1538, TA1535,~and TA98-FR1 The TA98-FR1 strain
is a nitroreductase-deficient strain (Rosenkranz and Speck, 1975),
which was provided by Dr. Herbert Rosenkranz (New York Medical
College, Valhalla, NY). TA98-FR1 is deficient in only one of
several nitroreductase enzymes (H.S. Rosenkranz, personal
communication, 1979). All other strains were provided by Dr. Bruce
Ames (University of California at Berkeley). The test results for
all six indicator strains are presented in this summary paper:
however, the data is given for only one strain, due to the large
volume of data collected.
A second bioassay was also conducted: the. 8-azaguanine
forward mutation assay. Two strains of S. typhimurium,
TM677 and TM35, were used, as described by Skopek et al. (1978a, b).
Samples
All samples were organic extracts from automotive exhaust
particles. In each case, the vehicle was operated on a chassis
dynamometer, the exhaust was diluted and cooled in a stainless
steel dilution tunnel, and the particles were collected on Pallflex
T60A20 glass fiber filters. The entrapped particles were then
extracted with dichloromethane (Huisingh et al., 1978). The
remaining organics were solvent exchanged to dimethylsulfoxide
(DMSO) to a final concentration of 2 mg exhaust organics/ml DMSO.
The DMSO solution was used in the bioassay. Each sample was given
a unique identificaiton number to prevent bias in testing and to
allow computerization of all data. These identification numbers
are used for sample identification in this paper.
Sample CMBX-79-0092 is a Nissan 220C diesel vehicle, and the
results for this sample are reported to demonstrate the typical
-------
BACTERIAL MUTAGENESIS AND EVALUATION OF EMISSIONS
301
response in the five major tester strains. Samples CMBX-79-0001 to
CMBX-79-0012 were 12 aliquots of the sane sample. Each aliquot was
tested during consecutive months to ascertain any effects of
storage. Multiple samples were collected on three consecutive days
from the same Oldsmobile 350 vehicle and supplied as samples MSER-
78-0122 to MSER-78-0135. A collection of exhaust organics from the
Oldsmobile 350 diesel was chemically fractionated by the Research
Triangle Institute (RTI), Research Triangle Park, NC, and was
assigned numbers MSER-79-0039 to MSER-79-0046• The method of
chemical fractionation is reported elsewhere (Little, 1978: Lee et
al. , 1976). MSER-79-0032 to MSER—79-0036 were samples from five
separate gasoline automobiles of the same make, model, and
configuration (Ford LTD, catalyst equipped). The variation among
different makes and models of diesel automobiles was demonstrated
with the TA£B samples. The nitroreductase-deficient strains were
used with the following samples: an Oldsmobile 350 diesel (MSER-
79-0074), a 1978 Datsun 810 gasoline (MSER-79-0064), an Oldsmobile
260 diesel (TAEB-79-0029), and a VW Dasher diesel (TAEB-79-0034)
vehicle. The comparison of forward and reverse mutation systems
using a preincubation protocol used an Oldsmobile 260 sample (TAE3-
79-0030) and a VW Dasher sample (TAEB-79-0036). Table 1 summarizes
the samples .
RESULTS
The qualitative response of the five Ames tester strains to
exhaust extracts from various internal combustion engines was very
consistent. TA1535 generally gave either a very low or a negative
response. Each of the frameshift tester strains has given positive
results. Since TA100 responds to frameshift mutagens as well as to
other mutagens, it also gave a positive response. Figure 1 shows
this typical response of the five tester strains to the organics
from the exhaust of a Nissan 220C automobile.
Tables 2 and 3 summarize the results of bioassay data from
multiple Highway Fuel Economy Test (HWFET) cycles run on three
consecutive days with the same Oldsmobile 350 diesel vehicle. The
bioassay data is expressed for the organics as revertants per
microgram organic (slope of the linear regression line). By using
the percent extractable mass and the particulate emission rate
(PER) for each automobile, the revertants per gram particulate and
the revertants per mile were calculated from the given slope. The
revertants per microgram ranged from 2.86 to i.33, and the
revertants per mile ranged from 187,000 to 284,000. Even with the
variation of test runs and bioassays, there is less than a twofold
difference in the bioassay values for different filter extracts
from the same vehicle operated on the HWFET cycle. The overall
coefficient of variation is approximately 11%.
-------
302
LARRY CLAXTON AND MIKE KOHAN
Table I. Sample Information
Sample
Sample Number
Oldsmobile 350 diesela
Total particle extract
Acid I fraction
Acid II fraction
Base I fraction
Base II fraction
Insoluble tars
Polar neutral fraction
Polynuclear aroraatics
Nonpolar neutral fraction
Ford LTD gasoline automobiles3
Total particle extract
Nissan 220C diesel3
Baseline study for strains
Comparative vehicles3,*5
Oldsmobile 350 diesel3,*3
Oldsmobile 260 diesel*1
Mercedes Benz 300D diesel'5
Open Record E diesel^
Chevrolet truck 350 diesel*5
VW Dasher wagon diesel^
1978 Datsun 810 gasoline3
VW Dasher diesel°
CMBX-79-0001 to 0012
MSER-78-0122 to 0135
MSER-79-0039
MSER-79-0040
MSER-79-0041
MSER-79-0042
MSER-79-0043
MSER-79-0044
MSER-79—0045
MSER-79-0046
MSER-79-0032 to 0036
CMBX-79-0092
TAEB-78-
MSER-79-
TAEB-78-
TAEB-79-
TAEB-78-
TAEB-78-
TAEB-78-
TAEB-78-
MSER-79-
TAEB-79-
0501-0507
0074
•0502.
¦0029:
•0504.
¦0506
•05 11 i
¦0513.
¦0064
•0034,
0503
0030
0505
0510
0512
0514
0036
aSupplied by Ron Bradow and Roy Zweidinger, U.S. Environmental
Protection Agency (EPA), Research Triangle Park (RTP), NC.
Fractionation by Edo Pellizzari, RTI, RTP, NC.
''Supplied by Tom Baines, Emission Control Technology Division,
EPA, Ann Arbor, MI.
The same diesel automobile (Oldsmobile 350) was used to
generate multiple samples that were pooled subsequently for a large
single sample. Twelve of the aliquots (CMBX-79-0001 to CMBX-79
-0012) from this sample were stored in glass vials at -80°C in the
dark. Each month for 12 months, one aliquot was tested in the
microbial assay, and the results are recorded in Table 4. For
-------
WITH ACTIVATION
WITHOUT ACTIVATION
TA9B
1200
TA9B
TA100
TAI5JB
TAlb38
-O TAI537
MICROGRAMS PER PLATE
Figure 1. Exhaust organics from a Nissan 220C vehicle tested in the five tester strains
of S. typhimurium. q
CO
-------
304
LARRY CLAXTON AND MIKE KOHAN
Table
2. Comparative Results
: of E
xhaust Organics from an
Oldsmob ile
350
Diesel Tested with S.
tvphimurium TA98
Without Acti
vat ion
Sample
No. Slope^
%
Rev x 10
5 PERd
Rev x
(MSER-
78-)a (Rev/plate/ug)
Ext .c
/g Partic
(g/mile)
10 5/mile
Day 1

0122
4.33
10.8
4.68
0.502
2.35
0123
3.99
10.8
4.31
0.484
2.09
0124
3.82
12.4
4 .74
0 .495
2.34
Day 2

0126
3.99
13.8
5.51
0.516
2.84
0129
3.86
11.4
4.40
0.517
2.28
0130
3.41
12.8
4.36
0.487
2.13
Day 3

0132
3.72
12.6
4.69
0.518
2.43
0133
2.86
11.2
3.20
0.584
1.87
0134
3.35
11.6
3.89
0.569
2.21
0135
3.46
10.8
3.74
0.568
2.12
aNurabers assigned to each automobile exhaust sample collected by
EPA, Environmental Sciences Research Laboratory (ESRL), RTP, NC.
*>S1 ope of linear regression line.
cPercent dic'nlororaethane-extractable mass.
"^Particle emission rate (courtesy of Roy Zweidinger, EPA, ESRL, RTP,
NC).
these storage samples, the revertants per microgram ranged from
4.05 to 7.23, with a mean of 5.52. The coefficient of variation
for slope was 18.1%. Although obvious variance was seen from month
to month, none of the differences were significant.
To test variation among vehicles of the same make, model, and
configuration, five Ford LTD gasoline vehicles of the same
configuration were tested. Table 5 provides the bioassay data for
these vehicles. The data show chat the variation among different
vehicles was much greater than the variation when the same vehicle
was retested. In the bioassay, the revertant per microgram organic
levels varied from 1.02 to 9.61. While the various parameters of a
single vehicle could vary by a factor of two, the same parameters
for different individual vehicles could show a 5- to 10-fold
difference. It can be seen from Table 5 that a single sample
(i.e., sample MSER-79-0036) could be responsible for most of the
variation, when small numbers of samples were used. When the
-------
BACTERIAL MUTAGENESIS AND EVALUATION OF EMISSIONS 305
Table 3. Daily and Total Statistics of Results for Exhaust Organics
from an Oldsmobile 350 Diesel Tested with S. typhimurium TA98
Without Activation
Sample No. Slope'5
(MSER-78-)a (Rev/plate/yg)
%
Ext.c
Rev x 10 -
/g Partic.
PERd
(g/mile)
Rev x
10 '/mile
Day 1

Mean
4.05
11.3
4.58
0.494
2.26
Standard dev.
0.26
0.9
0.23
0.009
0.15
Coeff. of var.
0.06
0.08
0.05
0.02
0.07
Day 2

Mean
3.75
12.7
4.76
0.507
2.42
Standard dev.
0.30
1.2
0.65
0.017
0.37
Coeff. of var.
0.08
0.08
0.14
0.03
0.15
Day 3

Mean
3.35
11.6
3.88
0.560
2.16
Standard dev.
0.36
0.8
0.62
0.029
0.23
Coeff. of var.
0.11
0.07
0.16
0.05
0.11
All days combined

Mean
3.68
11.8
4.35
0.524
2.27
Standard dev.
0.42
1.0
0.64
0.037
0.26
Coeff. of var.
0.11
0.09
0.15
0.07
0.11
aNumbers assigned to each automobile exhaust sample collected by
EPA, Environmental Sciences Research Laboratory (ESRL), RTP, NC.
^Slope of linear regression line.
cPercent dichloromethane-extractable mass.
^Particle emission rate (courtesy of Roy Zweidinger, EPA, ESRL, RTP,
NC) .
revertants per microgram organic were normalized to revertants per
mile, the variation was reduced from a 10-fold to a 5-fold
di f ference.
The TAEB samples (Table 6) demonstrated the degree of
variation in mutagenic response found between diesel vehicles of
different makes and models. Among the diesel vehicles tested, the
revertants per microgram organic ranged from 1.15 to 3.23, while
the revertants per mile ranged from 250,000 to 878,000.
Results from bioassay techniques can be used to guide the
chemical fractionation of complex exhaust organics. This
-------
CO
Table 4. Mutagenicity of l)infu?l Kxhad^t Organic Samples Hfnied Over Varying Periods of Time ^
Tested with S. t yph imuri inn TA100
Mean Revertants per Plate for Monthly Samples
Compound
Act"
Dogi* (pg)
He an
snh
Hcnn
sn
Mean
sr>
Henri
SD
Mean
SI)

JAN

FEB

MAR

HAY

J UN

Positive control
•f

7 34 . 50
62 .91
741 .00
4 3.02
7/8.67
16.1)2
669.11
10.12
993 6/
4 3.13

291.00
2.81
305.00
2.83
84.6/
7 .51
466.66
43 13
510 6/
25 70
Negative control

Ukmetliy 1 .nil lox i de
~
100.00
ul
95 .00
7.07
90.00
7.07
81 .67
3.21
94 .00
18.08
85 67
4.73
U i mt*r It y I sill (oxide
_
100.00
Ml
89 .00
8 .49
101.50
6.36
88 .00
17.35
90.67
4.51
85.00
9.54
<)lri»mohilc J50 dieacl
*
60.00

227.67
11 .37
250.67
8. 50
242 33
4.51
248.00
9.64
310.00
19.47

~
100.00

318.53
40,50
393.33
28.02
362.50
12.02
305.00
34.39
421.33
19.22

~
200.00

504.67
87 .05
-
-
564.31
4 6.61
481.67
4 .73
660 13
6.51

~
600.00

-
-
1 117 OU
14 .42
91 7.00
12.12
1185.0(J
21 .0/

-
60.00

473.00
10.05
453.67
47.88
471 33
4.93
494.6 7
18.61
532.00
56. 15

-
100.00

613.33
29.14
591.00
78.48
628.33
2 7.39
605.31
12.22
7 39 33
20 .03

-
200.00

-
-

-
107/.67
20.03
934.00
8. 72
1117.33
16.92

-
600.00

-
-
-
-
1176.33
44 .00
1125.no
48.12
1298.33
82 . 50

JUL

Aim

SF.P

OCT

NOV

Positive control
~

6 74.00
49.43
118.67
21 .83
338 33
20.50
1065 67
74 . 10
1168.00
3/ 11

477.67
22 50
365.67
18.50
662.67
21 .03
591.67
3 . 79
6/5.67
1/21
Negat ive cont rul

Uinto t It y I tiiil 1 <»k »d2 .6 /
12 42
528.67
26.54
580.67
3.06
494 31
20.40
685 .00
44 84

-
100 .00

366 .00
11.14
701.00
33.96
735.33
59.80
/6/ . 33
28.5 7
104 1.00
20 66

-
2 00.00

685.67
44.38
926.33
44 . 74
1 182.33
75 . 39
1 2 30. 13
16.17
1565.67
2 5 74

600.00

1119.00
25 . 36
961.33
16.62
1227.33
146.17
1436.67
19.30
1920.00
59.21
flMetaholic activat ion.
'•St
po
SO
o
f
>
X
H
O
•z
>
'z
~
£
*
M
?!
O
sc
>
-------
BACTERIAL MUTAGENESIS AND EVALUATION OF EMISSIONS
307
Table 5. Comparison of Exhaust Organics from Gasoline Vehicles of
the Same Make, Model, and Configuration (Ford LTD Automobiles)
in S. typhimuriura TA98 Without Activation2
Sampl
e No.
Slope'5
%
Rev x 10s
PERd
Rev/
(MSER-79-)
(Rev/plat e/ug)
Ext.c
/g Partic.
(g/mile)
Mile
0032
Lt. blue
7.54
5.8
4.37
0.0079
3455
0033
Mid. blue 7.51
3.4
2.55
0.0100
2553
0034
Gray
9.49
3.2
3.04
0.0180
5466
0035
Silver
9.61
3.8
3.65
0.0100
3652
0036
Brown
1.02
21 .4
2.18
0.0050
1091
Mean

7.03
7.52
3.16
0.0102
3243
Standard dev.
3.51
7.83
0.87
0.0048
1602
Coeff.
of var .
0.50
1.04
0.28
0.47
0.49
aNumbers assigned to each automobile exhaust sample collected by
EPA, ESRL, RTP, NC.
t>Sl ope of linear regression line.
cPercent dichloromethane-extractable mass.
^Particle emission rate.
possibility was demonstrated with exhaust organic samples from an
Oldsmobile 350 diesel vehicle chemically fractionated at RTI under
the direction of Edo Pellizzari. When adequate sample was
available, each fraction was bioassayed with all five tester
strains. Results (Table 7) were similar to previous results with
exhaust organics from heavy-duty diesel engines (Little, 1978). (A
more complete summary is in preparation.) TA1535 gave negative
results with all fractions except for the polynuclear aromatic (PNA)
fraction when exogenous activation was used. The basic and nonpolar
neutral fractions were negative; however, the basic fraction could
not be adequately tested due to a lack of sample. The acid I, acid
II, polar neutral, and PNA fractions gave positive results with each
strain that responds to frameshift mutagens. This activity was
demonstrated both with and without activation.
When sample CMBX-79-0047 was tested with TA98 and TA98-FR1
(Table 8), the nitroreductase-deficient strain demonstrated
approximately one half the activity of TA98. With cigarette smoke
condensate, both strains provided equal responses. When the results
of other mobile-source organics simultaneously tested with these two
strains were compared (Table 9), this relationship was not
maint ained.
-------
CO
o
OQ
Table 6. Comparison of Organic. Exhaustfl from Diesel Vehicles U9ing
S. typliimurium TA98 Without Activation
Sample No.
(TAEB-78-)
Vuli ic 1 e
HC
( ppm)
N0X
( pptn)
CO
( ppm)
S1 ope''
(Rev/plate
/M«)
Mean
%
Ext .<
Rev x 10
/g
Part ic .
PERd
(g/mile)
Rev x 10
/mile
501 + 0507
502 f 503
511 f 512
504 + 505
506 + 510
513 + 514
Olds 350
Olds 760
Chev . truck
350 dienel
M.B. 3000
Opel Record E
VW Dasher
He nn
Standard deviation
Coefficient of variation
0.593
0.584
0. 784
0.251
0.454
0.503
0.528
0. 176
0.33
1 .49
1 .59
1.53
I .38
2.08
0.97
1.51
0.36
0.24
1 .508
I . 504
1 .590
I . 377
I .540
1 . 1 70
1 .448
0. 153
0. 11
1 .85
1.51
2.64
3.2 3
1.15
I .45
1 .97
0.80
0.41
21.1
16.5
53.4
2 3.3
51.6
53. 7
36.6
18.0
0.49
3.90
2.49
14.10
53
,93
79
6.96
4.06
0.58
0.H46
I .005
0.623
0.840
0.483
0. 322
0.687
0.256
0.37
aProvided by Tom Baines, Emission Control Technology Division, EPA, Ann Arbor, Ml.
''Slope of linear regression line.
cPercenr d ich lorome.r hane-ext rac t ab le mass.
''Particle emission rate.
3 . 30
2 . 50
8.78
6. 32
2.87
2.51
4 . 38
2 . 59
0.59
r>
>
ja
~<1
o
f
>
X
H
O
5:
5
X
M
W
O
33
>
-------
Table 7.
Activity of Organic Fractions from an Oldsinohile 350 Diesel Exhaust
Sample Tested in S. typli imur iuni"
Sample
TAI00
+ S-9
-S-9
Acid I fraction
Acid 11 fraction
Base I fractionc
Baae II fraction0
Insoluble Lars
Polar neutrals
I'NA
Non-polar neutrals
Specific Activity for Strain
TA98
fS -9
-S-9
TA153 7
TAI538
+ S -9
-S-9
+ S 9
-S-9
NT
NT
NT
NT
TA1535
fS-9
S-9
NT
NT
NT
NT
aPosiLive activity indicates a done-dependent response with at leant one dose pivinf* a two-fold
increase over spontaneous activity.
''NT = not tested; * = positive; - = negative.
clncomplete Lesting.
Od
>
n
H
w
ta
t—t
>
t-1
5?
3
>
o
w
X
M
CO
C/5
>
Z
a
cn
<
>
tr
G
>
H
I—I
O
YA
O
•n
M
£
m
rj)
~—H
o
CM
Go
O
<£>
-------
310 LARRY CLAXTON AND MIKE KOHAN
>
il
a
CC sC
^ n ^ c — ^ ® ec
>C«C — IT
-9" sS ^ ^

o^-ircc^e
« &
T~ ^
< >

I
c c
r— »c
f»
rx
C
C
^ r*. O O
I—' (¦" c o *c
r- o
^QCnJ^l^CC^P^U-.
NNNesm^ff'O®
r«»r*r*«.f*r*CCC:—':""
^^"'^r-'r-CCCr-*:-'
O-a^^r^i — — <— i-* rv. ~
p-.r*f*"»
a <
at f-
e u.
^ CO
._. o>
r <
3 H
ac
= .!:
t =i
III
w ?s
Q£ 9
iu —
i a.
oo
ON >
<
H »-
o;
Z
r* C
•m r-
sC
o
o o
o o
PS
CN *3-
B n r ^
C C r* M
^
^ — r-
«C C ® -¦
N -s. r-» r-» c Q
COOe^sO^>£CC
r-rJMiA-CC-
\CeNr»»0**s''*^Cs0 o
c

u
u ¦+. ^

c
—

C
c
3
k*
C 3 3
E

Q
»—
c
O fl 1
c

c
C
u
a.
4
(J —

i
&

C
C
>k >k

B
• «-
E
41
u
C
« x r

o
>
kd
• •*
> jj
4J
a>

u
¦ d
-------
BACTERIAL MUTAGENESIS AND EVALUATION OF EMISSIONS
311
Table 9. Comparison of Exhaust Organics from Different
Light-Duty Automobiles with Nitroreductase-deficient (TA98-FR1)
and -competent (TA98) Strains of S. typhimurium
Specific Activity
(rev/100 ug)
Metabolic
Sample Activation TA98 TA98FR1
MSEX-79-0064
(Datsun 810 gasoline auto.)
MSER-79-0074
(Olds 350 diesel auto.)
TAEB-79-0029
(VW Dasher diesel wagon)
TAEB-79-0034
(VW Dasher diesel auto.)
154 210
f 341 490
79 221
+ 81 207
46 104
+ 65 106
108 330
+ 156 257
The preliminary study comparing the response of complex
mixtures in a forward and reverse mutation system was completed
using four strains of S.typhimurium: TM677, TM35, TA100, and
TA1535, Each of the four strains were used for forward mutation to
8-azaguanine resistance. Strains TA100 and TA1535 were used for
reverse mutation to histidine prototrophy. Throughout this study,
the preincubation assay as described by Sfcopek et al. (1978a, b)
was used. Two samples, from an Oldsmobile 260 and a VW Dasher,
were used for comparison. Table 10 gives a summary of the results.
The mutagenic activity of the two samples was detected by the
forward mutation system using strains TM677 and TA100: however,
TA100 was not as sensitive as strain TM677. The preincubation
protocol also was used for the reverse mutation assay in which
strain TA100 provided a positive response. TA1535 was negative for
the reversion assay.
DISCUSSION
Microbial mutagenesis assays can be used for rapid initial
evaluation of combustion organics and emissions. Previously
published reports (Huisingn et al., 1978) stated that most of the
mutagenic activity associated with diesel samples from heavy-duty
-------
TabLe 10. Comparison of Korward and Keverse Mutation Systems With the
Preincubation Protocol Using Two Diesel Exhaust Samples
8-Azaguanine Resistance Histidine Reversion
(mutants/10 ^ survivors) (mutants/10® survivors)
Sampie
Dose
TM677
TA100
TM35
TA1535
TA100
T AI 5 3'
S pont aneou9

4.4
1 .2
1.2
0.6
4.1
0.2
Cont rol: MNNG
2. OA
|jM
37 .9
95.0
42 .0
27.0
7.9
1.9
Olds 260
12.5
pg/ml
5.4
2.7
1 . 1
0.9
4.6
0.3

2 5.0
(jg/inl
6.1
4.2
1 . 5
1.3
6. 1
0.2

50.0
lig/ml
7.9
4.6
I .3
1.7
5.4
0.4

1 00.0
Mg/ml
14.6
9.3
I .2
2.0
I0.fi
0.3
VW Dasher
12.5
Mg/inl
7.0
9.6
1 .6
1.2
4.4
0. 1

25.0
lig/inl
10.7
19.1
1 .6
1.8
5.9
0.1

50.0
|lg/inl
18.4
34.3
1 .4
1.8
5.8
0.3

1 00.0
llg/inl
14.2
43.4
I .6
1 . 5
9 . 5
0.3
-------
BACTERIAL MUTAGENESIS AND EVALUATION OF EMISSIONS
313
engines was detected by the indie
frameshift mutagens. This paper
passenger automobiles. Each dies
which testing has been completed
produced a negligible response in
TA98, TA100, TA1537, and TA1538.
the five tester strains to Nissan
organics. Although the magnitude
qualitative response was similar
date.
ator strains that respond to
provides data on light-duty diesel
el and gasoline organic sample for
in all five strains to date
TA1535, but positive responses in
Figure 1 shows the response of
220C diesel automotive exhaust
of response varied, the overall
for all automobiles tested to
A great concern in testing complex environmental samples is
the variability expected from multiple mixtures of chemicals.
Therefore, the difficulty in sorting out the factors that cause
this variability was explored. Tables 2, 3, 4, and 6 show the
variability of bioassay and chemical data with different types of
comparisons. Tables 2 and 3 compare the results for samples
collected from one automobile at various times. Both the bioassay
data and chemical data showed less than a twofold difference in all
comparisons and a coefficient of variation less than 20%. A
comparison of samples from vehicles of the same make, model, and
configuration (Table 5) demonstrates an increased variability over
multiple samples taken from one vehicle. There was nearly a
tenfold difference in revertants per microgram of organic material
between two automobiles. In this comparison, one automobile (the
brown Ford LTD, sample MSER-79-0036) introduced significantly
different values for most of the parameters measured: however, the
variability for the bioassay was normalized to a large extent with
the calculation of revertants per mile. Clearly, differences in
emission characteristics can result in widely different mutagenic
activities in emission organics. The brown Ford LTD, for example,
may have emitted an unusual amount of unburned fuel or oil that
would have diluted the concentration of mutagens and increased the
percent extractable. However, since the amount of fuel burned for
a predetermined distance would not be altered, the calculation of
revertants per mile would normalize the data. Table 6 is the
averaged data from replicate experiments for six different light-
duty diesel vehicles. The variation among different diesel
vehicles is similar to the variation between the different Ford LTD
gasoline automobiles. There are, however, two dramatic differences
First, the organics of the gasoline vehicles demonstrate more
mutagenicity on a per weight of particulate basis. Second, when
normalized to revertants per mile, there is an approximately 50- to
100-fold difference between the diesel vehicles and the Ford LTDs.
The mean for the different diesel vehicles is 438,000 rev/mils, and
the mean for the Ford LTDs is 3,200 rev/mile. This difference is
attributable mainly to the difference in PER and demonstrates the
need for adjusting the data for specific needs and comparisons.
Table 11 summarizes the data from different comparisons.
-------
Oj
J—1
Table 11. Comparison of Summary Data Deinonstrating Hie Effect of Differing Sampling
Parameters (Derived from Tables 4, 7, ami 9)
HC
( ppm)
N0X
( ppm)
CO Slope 7, Rev x 10 s PERb
(ppm) (rev/pi ate/pg) Ext. /g Partic. (g/mile)
Rev x
10 '/mi I e
Different runs within same auLomobile (diesel)
Mean 0.249 1.550 0.899 3.68 11.8 4.35 0.524 2.27
Standard deviation 0.032 0.10 0.070 0.42 1.0 0.64 0.037 0.26
Coeff. of var. 0.13 0.07 0.08 0.11 0.09 0.15 0.07 0.11
Vehicles of same make, model, and configuration (gasoline)
Mean 0.23 1.74
Standard deviation 0.20 0.66
Coeff. of var. 0.89 0.38
Different diesel vehicles
Mean 0.528 1.51
Standard deviation 0.176 0.36
Coeff. ofvar. 0.33 0.24
aSlopf of linear regression line.
''Particle, emission rate.
0.3/
0. 10
0.27
1 .448
0. 153
0. 11
7 .03
.3.51
0. 50
1 .98
0.80
0.41
7. 52
7.83
1 .04
36.6
18.0
0.49
3. 16
0.87
0.28
6 .96
4.06
0. 58
0.0102
0.0048
0.47
0.687
0.256
0.37
0.032
0.016
0.49
4 . 38
2.59
0. 59
>
&
fo
<
o
X
H
O
2
a
V
hv*
I—<
*
w
w
o
X
>
5*
-------
BACTERIAL MUTAGENESIS AND EVALUATION OF EMISSIONS
315
Decisions on pollution control devices, alterations in engine
design, criteria for fuel characteristics, and some environmental
regulations rely upon an understanding of which specific organics
are likely to have a detrimental health effect. Bioassay-guided
chemical fractionation has the potential to speed the identification
of biologically active compounds in this category. Table 7
indicates that the chemist needs to place a higher priority on
chemical identification of the polar neutral fraction than the
nonpolar neutral fraction. Also, by examining the specific
activity of the different fractions, one notices that the PNA
fraction gives a positive response without activation. This
response may be due to spillover of polar neutral chemicals into
the PNA fraction with this particular fractionation scheme.
Nitroreductase-deficient strains of bacteria cannot metabolize
the nitrogen components of a chemical; therefore, a decreased
response in the nitroreductase-deficient strain supports the
conclusion that active nitro compounds are present in the organic
mixture. If activity of the compound does not depend on the
reduction of the nitrogen group, activity in a nitroreductase-
deficient strain should not differ from that of its parental
strain. This decreased response occurred with some samples, though
not all (Tables 8 and 9). This demonstrates that nitroaromatics
may be one class of mutagens prominent in mobile source emissions.
Whether these nitro compounds are true artifacts or are also
created under normal environmental conditions is yet to be
demonstrated.
The results of the 8-azaguanine forward mutation bioassay are
consistent with earlier results in the Ames reversion assay.
Although the Ames tester strains TA100 and TA1535 can be used for
this forward mutation assay, the strains developed by Skopek et al.
(1978a, b) were more sensitive and gave fewer technical problems in
the performance of the assay. Although the assay detects a range
of mutagens equivalent to the range detected by all five of the
routinely used Salmonella strains, these forward mutation strains
cannot be used in the same diagnostic manner. In other words, the
forward mutation system detects a variety of mutagenic insults,
whereas a specific type of DNA damage must occur to be detected by
reverse mutation systems. As the forward mutation assay is more
thoroughly validated, it becomes increasingly feasible to use the
8-azaguanine system for general screening and to resort to Ames
tester strains for better characterization of the substance(s)
being tested.
These samples demonstrate the uses of microbial mutagenesis
assays for both qualitative and semi-qualitative assessment of
mobile source emissions. This paper has demonstrated a means of
comparing and evaluating the potential health hazard of polluting
soot material from various mobile sources and from new technological
-------
316
LARRY CLAXTON AND MIKE KOHAM
developments for each of these sources. The classical chemical
examination of exhaust components cannot be used effectively to
quantitate and evaluate Che thousands of organic chemicals derived
from each combustion source. Whole-animal evaluation of multiple
mobile sources and technological alterations would involve
extremely slow evaluation and high costs. Although still
controversial in many aspects, the Salmonella assay for mutagenicity
provides a rapid, inexpensive means to assess genetically toxic
effects. Recent results (Nesnow and Huisingh, in press) indicate
that the plate-incorporation test correlates well with other tests
for genotoxicity of combustion organics. By assuming that
increased response in the plate-incorporation test represents
greater potential health hazard, the Salmone1la assay can be used
to compare various sources, evaluate technological developments,
and guide chemical characterization. Since mutation assays have
some power to predict heritable effects, carcinogenesis, and
teratogenesis (Hollaender and de Serres, eds., 1978), this
assumption does not rely on direct correlation of Salmonella assay
results with any one end point.
REFERENCES
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian microsome mutagenicity test. Mutation Res.
31:347-364.
Claxton, L.D., and C.C. Evans. (MS). The microbial mutagenicity
of various environmental substances. U.S. Environmental
Protection Agency: Research Triangle Park, NC.
Hollaender, A., and F.J. de Serres, eds. 1978. Chemical Mutagens:
Principles and Methods for Their Detection. Vol. 5. Plenum
Press: New York.
Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R. Zweidinger, S.
Tejada, J. Buingarner , F. Duffield, M. Waters, V.F. Simmon, C.
Hare, C. Rodriguez, and L. Snow. 1978. Application of
bioassay to the characterization of diesel particle emissions.
M. Waters, S. Nesnow, J. Huisingh, S. Sandhu, and L. Claxton,
eds. In: Application of Short-term Bioassays in the
Fractionation and Analysis of Complex Environmental Mixtures.
Plenum Press: New York. pp. 381-418.
Lee, M.L., M. Novotny, and K.D. Bartle. 1976. Gas chromatography/
mass spectrometric and nuclear magnetic resonance
determination of polynuclear aromatic hydrocarbons in airborne
particulates. Anal. Chem. 48:1566.
-------
BACTERIAL MUTAGENESIS AND EVALUATION OF EMISSIONS
317
Little, L. 1978.' Microbiological and chemical testing of air
samples for potential mutagenicity: First annual report.
EPA/68-02-2724. U.S. Environmental Protection Agency:
Research Triangle Park, NC.
Nesnow, S., and J. Lewtas Huisingh. (in press). Mutagenic and
carcinogenic potency of extracts of diesel and related
environmental emissions: Summary and discussion of results.
In: International Symposium on the Health Effects of Diesel
Engine Emissions. U.S. Environmental Protection Agency:
Cincinnati, OH.
Rosenkranz, H.S., and W.T. Speck. 1975. Mutagenicity of
metranidazole: Activation by mammalian liver microsomes.
Biochera. Biophys. Res. Commun. 66:520-525.
Skopek, T.R., H.L. Liber, D.A. Kaden, and W.G. Thilly. 1978a.
Proc. Natl. Acad. Scie. USA 7 5:4465-4469.
Skopek, T.R., H.L. Liber, J.J. Krolewski, and W.G. Thilly. 1978b.
Proc. Natl. Acad. Sci. USA 75:410-414.
-------
Intentionally Blank Page
-------
COMPARISON OF THE MUTAGENIC ACTIVITY IN CARBON PARTICULATE
MATTER AND IN DIESEL AND GASOLINE ENGINE EXHAUST
Goran Lofroth
Radiobiology Department
University of Stockholm
Stockholm, Sweden
INTRODUCTION
Airborne carbon particulate matter is a variable and complex
mixture of components, including a variety of organic compounds.
Its origin in urbanized and industrialized areas is primarily
through various combustion processes. Motor vehicles are often
viewed as a major source.
Talcott and Wei (1977) and Pitts et al. (1977) first showed
that the Salmonella/microsome mutagenicity test, as developed by
Ames et al. (1975), can be used to detect mutagenic—and thus
potentially carcinogenic—activity in airborne particulate matter.
Exploratory studies of motor vehicle exhaust (Tokiwa et al., 1978)
and of emissions from a stationary combustion plant (Lofroth, 1978)
also demonstrated the feasibility of the Salmonella assay for this
purpose.
The present report summarizes some of the results obtained in
an ongoing study of the mutagenicity, as detected by the Salmonella
assay, of airborne particulate matter collected in the Stockholm
area. The results are compared with mutagenicity data for motor
vehicle exhaust samples obtained in a previous study (Lofroth,
1979) and by additional analyses of these samples.
319
-------
320
GORAN LOFROTH
MATERIALS AND METHODS
Motor Vehicle Exhaust Samples
Exhaust samples from gasoline and indirect infection (IDI)
diesel passenger cars were the same as those described by Lofroth
(1979). They were obtained daring U.S. Federal Test Procedure 1973
cycles by sampling the undiluted exhaust and by collecting the
particulate matter on the filter at 25 to 50°C and also the formed
aqueous condensate containing compounds that had not adsorbed onto
particulate matter at the point of sampling. Particulate matter
was Soxhlet-extracted overnight with acetone, and after removal of
the major part of the acetone, the residue was dissolved in
dimethylsulfoxide (DMSO). The aqueous condensates were treated in
various ways, including extraction with n-pentane followed by
removal of the pentane and dissolution of the residue in DMSO.
These pentane-extracted samples, shown to contain the maior part of
the mutagenicity of the condensates, are the condensate samples
that have been further analyzed in the present study, togecher with
the samples of particulate matter. All samples dissolved in DMSO
were stored frozen at -20°C. No detectable change in their
mutagenic activity was noted over a period of about one year.
Airborne Particulate Matter
Airborne particulate macter was collected wich Sierra 305 High
Volume Samplers or similar equipment on 20.3- x 25.4-cra (8- x
10-in) glass fiber filters (Stora Kopparberg Special-produkter,
Sweden) wich a flow rate of 68 m^/h (40 fVmin) . Sampling sites
were located on the roof of a ten-story building in the northern
part of the inner city of Stockholm and on the roof of a two-story
house in a suburban community 22 km NNW from the center of
Stockholm. Sampling was usually performed for 24 h, starting and
ending at 6 to 7 AM. Night samples were collected between 10 PM
and 6 AM.
To study the mutagenic activity of size-fractionated
particles, samples were taken in the inner city of Stockholm.
Particulate matter was collected with a Sierra 305 High Volume
Sampler equipped with five stages of cascade impactors with slotted
glass fiber collection substrates and the regular glass fiber
filter as back-up filter. Simultaneous sampling at the same flow
rate (68 mVh) was done using a second 305 Sampler without
impactors. Sampling was done over four days: substrates and
filters were changed every 24 h. Filters were extracted
individually; the samples were first assayed separately and then
combined and assayed as one sample. The four substrates from each
impactor stage were extracted and assayed as one sample.
-------
MUTAGENIC ACTIVITY IN CARBON PARTICULATES AND EXHAUST
321
Collected filters were wrapped in aluminum foil and stored at
-20°C until they were extracted (between one and eight days). The
filters were Soxhlet-extracted for 16 h with 250 ml acetone. The
acetone was first evaporated under vacuum to about 10 ml and then
under a stream of nitrogen on a heating block at < 40°C to 0.3 to
0.5 ml. This residue was diluted with DMSO to a known volume,
usually about 4 ml/filter. All samples were stored frozen at -20°C
prior to and between mutagenicity assays.
Meteorological data were acquired from three official
stations. One is located in the inner city of Stockholm, and
records are made from three daily observations. Two are located at
airports, in the western part of the city and north of Stockholm,
and observations are made every 30 min.
Mutagenicity Assay
Mutagenicity was determined by the Salmonella plate-
incorporation method with bacterial cultures fully grown overnight,
as described by Ames et al. (1975). Assays generally were
performed with strains TA98 and TA100, obtained from Dr. B.N. Ames
(University of California, Berkeley, CA), and with the
nitroreductase-deficient strains TA98 NR and TA100 NR, obtained
from Dr. H.S. Rosenkranz (New York Medical College, Valhalla, NY).
The raicrosome-containing rat-liver supernatant (S-9) was prepared
from Aroclor-1254-induced male Sprague-Dawley rats and was used
with the necessary cofactors.
All assays included tests with positive control compounds.
The bacterial strains were routinely checked for the presence of
known characteristics, including spontaneous reversion frequency,
sensitivity to ultraviolet light and crystal violet, and
sensitivity or resistance to ampicillin. 3enzo(a)pyrene was used
as a positive control for the S-9; in TA98, a 5-yg dose of this
compound yielded between 300 and 600 revertants/plate with S-9 at
20 and 50 ul/plate.
The sources of the chemicals used were as follows:
2-nitropropane was purchased from Merck-Scnuchardt: FRG and
2-nitronaphthalene from EGA-Chemie: and FRG and 1-nitropyrene
(labeled 3-nitropyrene) from Koch-Light, England. Di- and
tetranitropyrenes were obtained from Dr. R. Mermelstein, Xerox
Corporation, USA.
During the course of this investigation, it was found that the
strain TA100 could vary in its response to large molecules and that
the change in characteristics was not readily detected with crystal
violet. The quantitative monitor presently used is 1-nitropyrene
(Koch-Light); 1 ug of the commercial product gives between 1000 and
-------
322
GORAN LOFROTH
L500 revertants/plate in TA100. Other samples of 1-nitropyrene may
give different responses, due to the presence of variable amounts
of more mutagenic impurities.
Each sample was assayed with one plate per dose level at two
or more dose levels in at least three independent tests. The
mutagenic response, expressed as revertants per cubic meter of air
for airborne particulate matter and as revertants per gram of
consumed fuel for motor vehicle exhaust, was calculated from the
linear part of the dose-response curve.
RESULTS AND DISCUSSION
Mutagenicity Pattern of Motor Vehicle Exhaust
Results presented in a previous report (Lofroth, 1979) showed
both similarities and differences in the mutagenic responses of
Salmonella to gasoline exhaust and diesel exhaust from normal
passenger cars. All samples were mutagenic in the absence of
mammalian metabolic activation, showing that the mutagenic
compounds present were either directly acting mutagens or were
converted to ultimate mutagens by bacterial metabolism. Both
gasoline and diesel exhausts were more mutagenic in TA100 than in
TA98.
Enhancement of the mutagenic activity by mammalian metabolic
activation was observed for only one type of sample: particulate
matter from gasoline exhaust. All other samples were less
mutagenic in the presence of S-9 than in its absence. The decrease
in mutagenic activity depended on the amount of S-9: it is thus not
appropriate to report the mutagenic activity in the presence of S-9
for such samples. For this reason, these results cannot be
compared with those of Ohnishi et al. (1980), who reported
Salmonella mutagenicity data for several gasoline and diesel
exhaust samples assayed in the presence of mammalian metabolic
activation.
Diesel exhaust was far more mutagenic than was gasoline
exhaust. Irrespective of the manner of comparison—revertants per
volume of exhaust, revertants per test cycle, or revertants per
amount of fuel consumed—diesel exhaust was more than ten times as
mutagenic as gasoline exhaust.
Assays involving anaerobic incubation during the first 16 h of
the 48-h incubation indicated that the exhaust samples contained no
or undetectable amounts of certain nitro compounds that become more
mutagenic when assayed under anaerobic conditions.
-------
MUTAGENIC ACTIVITY Cs CARBON PARTICULATES AND EXHAUST 323
Mutagenicity Pattern of Urban Particulate Matter
Sampling of urban particulate matter for mutagenicity studies
was attempted in the early part of 1975, buc the low volumes
sampled did not yield conclusive results. High-volume sampling was
started at the end of 1977 to test the feasibility of the assay,
and regular sampling was begun during 1978.
Several different Soxhlet-extraction solvents—acetone,
benzene, cyclohexane, dichloromethane, and methanol—were tested in
the early phase of the study. Acetone extraction gave the highest
yield of mutagenic activity. Extraction with benzene, cyclohexane,
dichloromethane, and methanol followed by further extraction with
acetone resulted in samples containing additional mutagenic
activity; however, extraction with acetone followed by further
extraction with any of the solvents resulted in samples with no
additional detectable mutagenicity.
The samples collected above the rooftops were generally tested
in the strains TA98 and TA100 and occasionally in the other tester
strains, TA1535, TA1537, and TA1538. No detectable mutagenicity
was observed with TA1535. Of the plasmid-containing strains TA98
and TA100, the former was usually the most responsive, although
TA100 sometimes showed equal or higher responses. Addition of S-9
usually, but not always, decreased the mutagenicity.
It was previously reported (Lofroth, 1979) that many urban
particulate samples showed a higher mutagenic activity in the assay
involving anaerobic incubation than in the regular aerobic assay.
This has been further confirmed, indicating that in contrast to the
motor vehicle exhaust samples, urban particulate samples may
contain nitro compounds that give this type of mutagenic response.
Seasonal variation. The mutagenic activity of the collected
urban particulate matter varied seasonally. Although there were
differences between consecutive days (see below), an average
mutagenicity could be calculated for periods during which 24-h
samples were collected. A number of such periods, consisting of
two to four weekdays, were studied. Results to date are presented
in Figure 1, which gives the average mutagenic activity in TA98 in
the absence of S-9 as a function of the time of year when the
samples were collected. The mutagenicity was higher during the
winter than during the summer. Many factors—several
meteorological parameters as well as emissions of various
compounds—may have contributed to the level of mutagenic activity,
Thus, it is not certain that the higher mutagenic activity during
the winter months was due to heating of buildings.
-------
324
GORAN LOFROTH
10 11 12 1 2
MONTH OF SAMPLING
Figure 1. The mutagenicity of extracts of particulate matter
collected during weekdays above the rooftops in the
inner city of Stockholm. Each point represents the
average of two to four consecutive 24-h samples.
Inner city and suburban sites. Mutagenic activity was
somewhat higher in particulate matter from the inner city of
Stockholm than in that collected simultaneously at the suburban
site (Table 1 reports mutagenicity in the absence of S-9). The
seasonal variation observed for the inner city sampling site
(Figure 1) was also found for the suburban site (Table 1). The
suburban site was located in an area where residential heating is
generally electric.
Daily variations. The mutagenic activity varied within the
same day and between consecutive days. Mutagenic activities in the
nighttime hours, when motor vehicle traffic was low, were compared
with activities in other periods, in order to assess the
contribution from stationary sources such as residential heating.
The mutagenicity of particulate matter collected between 10 PM and
6 AM was investigated for three different periods, during which
-------
MUTAGENIC ACTIVITY IN CARBON PARTICULATES AND EXHAUST
325
Table 1. Mutagenic Activity of Extracts of Particulate Matter
Collected Above the Rooftops in the Inner City of Stockholm and
Simultaneously in a Suburban Area 22 kia NNW of Stockholm.
Sampling Period
TA98
Revertant s/m3
of Aira
Inner
City
Suburban
Dec. 20-22, 1978
36

27
Feb. 19-23, 1979
55

30
Apr. 9-12, 1979
18

12
May 28-June 1, 1979
12

7
July 2-6, 1979
2.
.2
0.9
Sept. 3-7, 1979
13

4.4
Oct. 15-19, 1979
25

13
Dec. 17-21, 1979
35

22
Feb. 4-8, 1980
48

28
aAverage mutagenicity for two to four consecutive days.
time complete 24-h samples were also collected and assayed. The
results from one of these periods are given in Table 2. Nighttime
samples were appreciably less mutagenic than the 24-h samples.
Similar results were obtained for the other periods: the average
nighttime and 24-'n samples, were respectively, 5 and 25
revertants/m3 for Oct. 15 through 19, 1979, and 17 and 48
revertants/ni3 for Feb. 4 through 8, 1980.
These results indicate that emissions during the night, such
as those associated with residential heating, were not a ma^or
direct source of the mutagenicity present in 24-h samples.
However, it cannot be ruled out that some type of enhanced
formation of mutagenic components occurred where motor vehicle
exhaust was mixed and interacted with emissions from stationary
sources.
The variations between consecutive days shown in Table 2 were
typical for most of the investigated periods. Changes in
meteorological parameters were probably a major reason for these
variations. Although the acquired meteorological data have not yet
been fully evaluated and compared with the mutagenicity results, a
gross comparison revealed that wind speed may have been an
important factor in the day-to-day variations in mutagenic
act ivity.
-------
326
GORAN LOFROTH
Size distribution. Organic components associated with urban
particulate matter for the most part are adsorbed to small,
respirable particles (Van Vaeck et al., 1979). Talcott and Harger
(1979) have reported that mutagenic activity was associated
primarily with particles less than about 2 urn in samples collected
in southern California during long-term sampling.
The results for two complete sampling periods of the present
project are given in Table 3. It is evident that the smaller Che
particles, the greater was the mutagenic activity. Also, the
mutagenic activity of most fractions decreased after addition of
the mammalian metabolic activation system; the exceptions were
impactor stage 5 and, in one case, impactor stage 4. Particulate
matter collected on these two stages was intensely black, in
contrast to the greyish black color of the other fractions.
Samples collected simultaneously without size fractionation
had a mutagenicity greater than the sum of the mutagenic activities
of the fractionated samples. Reconstitution of the fractionated
samples gave about the sane mutagenicity as that expected from the
sum of the fractions. The lesser mutagenic activity in size-
fractionated samples, compared with unfractionated samples, cannot
be explained. However, it may have been due to loss of particulate
matter to other surfaces of the impactors (Chenp; and Yeh, 1979).
It is also conceivable that, if mutagens form as artifacts during
sampling (Pitts et al . , 1978), such reactions may be decreased by
Table 2. Mutagenicity
of Extracts of Particulate
Mat ter
Collected at
Night and Over 24 Hours

TA98 Revertants/m^
of Air

Inner City
Suburban
Sampling Date

(December 1979)
8 h Night3 24 h
24 h
17-18
13 48
33
18-19
9 23
12
19-20
12 31
17
20-21
12 36
24
Average
11 35
22
a10 PM to 6 AM.
-------
MUTAGENIC ACTIVITY IN CARBON PARTICULATES AND EXHAUST 327
Table 3. Mutagenic Activity of Extracts of Size-fractionated
Particulate Matter Collected in the Inner City of Stockholm
Compared with the Mutagenic Activity of Samples Collected
Simultaneously Without Size Fractionation3
TA98 Revertants/m3 of Air

Oct.
22-26, 1979
Nov.
19-23, 1979

Particle

Size
-S-9
+ S-9
-S-9
+ S-9
Sample
(um)

(50 ul/plate)

(50 ul/plate)
Impactor stage 1
7.2
0.3
0.1
0.2
0.2
2
3.0
0.5
0.4
0.3
0.2
3
1.5
0.5
0.5
0.4
0.3
4
1.0
0.7
1.2
0.6
0.5
5
0.5
1 .4
2.5
1.0
1.4
Back-up filters

6.2
4. 5b
5.4
2. 7b
Sum of impactor stages

1-5 and back-up
filters
10
9
8
5
Reconstituted sample:

impactor stages
1-5

and back-up filters
-
-
7
7b
Filters without impactors
15
9b
16
9b
a?arCicle sizes are those given by the manufacturer at 50% collec-
tion efficiency for spherical particles with unit mass density.
^Approximate figures; assays in the presence of S-9 often give
nonlinear dose-response curves, making it difficult to arrive at a
single response.
collecting precursors on the impactor substrates instead of on the
filter through which polluted air is flowing during the entire
sampling period.
Chromatographic separation. Extracts of urban airborne
particulate matter having sufficiently high mutagenic activity were
separated by high performance liquid chromatography (HPLC) , and the
eluates tested for mutagenicity. The system separates standard
-------
328
GORAN LOFROTH
polycyclic aromatic hydrocarbons and ni troarenes in narrow peaks
(1 ml, 1 or 2 fractions). The mutagenic activity in extracts oF
particulate matter eluted over a wide range, with no major single
peaks (as shown in Figure 2). Most of the mutagenic activity was
found in the range where di- to tetracyclic compounds elute. The
results suggest that the mutagenicity of extracts of urban airborne
particulate matter is caused by many compounds and that it may be
unprofitable to look for single compounds as major airborne
mutagens .
Mutagenic Response in Nitroreductase-deficient Tester Strains
Salmonella strains deficient in nitroreductase activity were
developed in conjunction with mutagenicity studies of nitrofuran
derivatives (Rosenkranz and Speck, 1976). Nitrofuran derivatives
that are mutagenic in the original strains are less mutagenic in
the nitroreductase-deficient strains.
"= 200
£ ioo
spontaneous background
0
0
1.0
2.0
RETENTION TIME (as a factor of the retention time of pyrene}
Figure 2. Mutagenic response of fractions from a reverse-phase
HPLC separation of an extract of airborne particulate
matter. The sample (50 yl, containing activity
corresponding to about 1600 revertants) was applied to
a Spherisorb S5-0DS column (220 mm x 4.6 mm i.d.) with
methanol-water (8:2) as elutant. Fractions of 0.67 ml
(20 s) were collected and assayed in TA98 in the absence
of S-9 , using 0.2 and 0.4 ml.
-------
MUTAGENIC ACTIVITY IN" CARBON PARTICULATES AND EXHAUST 329
In the present study, nitroreductase-deficient strains TA98 NR
and TA100 NR were used to investigate the behavior of some
nitroarenes, motor vehicle exhaust samples, and extracts of
airborne urban particulate matter. Typical mutagenic nirroarenes
were less mutagenic in the TA98 NR strain than in the original
strain (Figure 3): but whereas the mutagenic resoonse of the
dicyclic 2-nitronaphthalene almost disappeared, the response
decreased only slightly for the tetracyclic 1-nitropyrene. Strains
TA100 and TA100 NR behaved similarly.
The mutagenic activity of the motor vehicle exhaust samples
was more or less the same in TA98 and TA98 NR (Figures 4 and 5 and
Table 4). In contrast, several extracts of urban airborne
particulate matter were less mutagenic in the nitroreductase-
deficient strain (Figure 6 and Table 4). Preliminary studies
comparing the mutagenic activity of these samples in TA100 and
TA100 NR gave similar results: the motor vehicle samples produced
about the same response in both strains, whereas airborne
particulate matter was less mutagenic in TA100 NR than TA100.
These results indicate that extracts of urban airborne particulate
matter contained certain types of nitro compounds (specifically,
those that were metabolized to ultimate mutagens by the bacterial
nitroreductase system that had been abolished in the deficient
strains). The motor vehicle exhaust samples did not seem to
contain detectable amounts of these compounds.
Thus, comparisons between the regular assay and the assay
involving anaerobic incubation and between the regular and the
nitroreductase-deficient strains both imply that extracts of
filter-collected urban airborne particulate matter contain nitro
compounds. Due to the uncertainty about artifactual formation of
nitro compounds during sampling (Pitts et al . , 1978), it is,
however, premature to state that airborne particulate matter
contains nitro compounds prior to collection. Wang et al. (1980),
after comparing the mutagenic response in TA98 and a
nitroreductase-deficient strain, suggested that n:troaromatic
compounds are present in airborne particulate matter.
Environmental Nitro Compounds
The simultaneous presence of organic compounds and nitrogen
oxides in combustion emissions and polluted air requires extensive
studies on the possible presence and formation of mutagenic or
carcinogenic nitro and nitroso compounds (Pitts et al., 1978;
Ehrenberg et al., 1980) .
Aliphatic nitro compounds should not be discounted as
mutagens. In a series of tested nitroalkanes, the common
industrial solvent 2-nitropropane was found to be mutagenic in the
-------
330
GORAN LOFROTH
TA98 TA98NR
2-mtronaphthalene
1-nitropyrene
LLI
l-
<
_i
500-
CC
LU
<
oc
LU
>
LU
IX
2-nitronaohthalene (fig/alatej
0
100
0 1-nitropyrene (rig/plate) 200
Figure 3. Mutagenicity of 2-nitronaphthaiene and 1-nitropyrene
in TA98 and TA98 NR in the absence of S-9. The
spontaneous mutation rate has been subtracted.
TA98 TA98NR
OJ
l-
<
_i
£T
LLI
to 200-
H
Z
<
H
e
L_
>
UJ
c
0
0.5
EXHAUST, expressed as fuel consumed (g/platek
Figure 4. Mutagenicity of diesel exhaust in TA98 and TA98 NR in
the absence of S-9.
-------
MUTAGENIC ACTIVITY IN CARBON PARTICULATES AND EXHAUST 331
TA 98 TA 98 N R
Particulate
£? 100-
LU
2
0
4
EXHAUST, expressed as fuel consumed (g/plate)
Figure 5. Mutagenicity of gasoline exhaust in TA98 and TA98 NR
in Che absence of S-9.
• TA 98
~ TA 98 NR
300
LU
<
-J
to
2
<
£ 100-
>
LU
tr
0
10
20
. SAMPLE (m3 of air/platel
Figure 6. Mutagenicity of urban airborne particulate matter in
TA98 and TA98 NR in the absence of S-9,
-------
332
GORAN LOFROTH
Table 4. Absolute and Relative Mutagenic Response of Motor Vehicle
Exhaust Samples and Extracts of Urban Airborne Particulate
Matter in TA98, TA98 NR., and TA100 in the Absence of S-9
TA98 Percent of
Response in TA98
Revertants/ Revertants/
Sample g Fuel Used of Air TA98 NR TA100
Diesel Exhaust
Particulate matter 800
Condensate 210
Gasoline Exhaust
Particulate matter 24
Condensate 39
Urban Airborne
Particulate Matter
4 days: March, 1979
(composite sample)
4 days: Nov., 1979
(composite sample)
4 days: Dec., 1979
(average)
90 >100
110 • >100
125 >100
115 >100
19 70 <100
18 60 <100
35 60 <100
Salmonella assay (unpublished data). The mutagenicity of
2-nitropropane was higher in TA100 than in TA98 and was not related
to the mutagenicity of nitrite. The mutagenic behavior of
2-nitropropane suggests that higher homologs of aliphatic nitro
compounds may have contributed to the mutagenicity of motor vehicle
exhaust samples.
Recent studies show that some photocopies and toners contain
mutagenic compounds, and it has been suggested that tri- to
pentacyclic nitroarenes were responsible for the mutagenicity in
some cases (Lbfroth ec al., 1980). In one case, the problem was
traced to the presence of small amounts of dinitropyrenes present
as impurities in a particular carbon black product (Rosenkranz
et al . , 1980) .
Hughes, et al . ( 1979) reported that pyrene adsorbed to various
particulate substrates is nitrated in the presence of nitrogen
dioxide and nitric acid and that mononitropyrene can be further
-------
MUTAGENIC ACTIVITY IN CARBON PARTICULATES AND EXHAUST 333
nitrated to dinitropyrer.es. Nitropyrenes are strongly mutagenic in
the Salmonella system and gave the following responses in TA98:
1-nitropyrene 4 revertants/ng
1,3-dinitropyrene 300
1,6-dinicropyrene 450
1,8-dinitropyrene 750
1,3,6,8-tetranitropyrene 60
Except for I-nitropyrene, none of the nitropyrenes showed an
increased mutagenic response with anaerobic incubation. In
addition, 1,6- and I,8-dinitropyrene gave about the same mutagenic
response in the nitroreductase-deficient strain TA98 NR as in TA98.
Nitropyrenes and particularly dinitropyrenes do not act in the sane
manner as several other nitroaroraatic compounds. Thus,
dinitropyrenes or similarly behaving mutagenic nitroarenes can be
present in noCor vehicle exhaust samples and extracts of airborne
particulate matter without being detected by the currently used
Qodifications of the Salmonella test.
Model calculation
The investigated urban area—Stockholm and neighboring
communities—has an area of approximately 3650 km^ , of which the
inner city of Stockholm occupies about 50 km2. The average fuel
consumption for transportation is estimated at 0.7 metric ton/year
for the whole area; 0.1 metric ton/year is used in the inner city.
Approximately 10X of the fuel is diesel. If the mutagenic activity
of vehicle emissions is assigned the values 2C0 revertants/g
gasoline consumed and 2C00 revertants/g diesel fuel consumed
(Lofroth, 1979), the mixed use then corresponds to an emission
equivalent to 400 revertants/g transportation fuel consumed. The
average emission per unit area and unit tine can then be calculated
to be about 100 revertants/ra2/h in the inner city and 8
revertants/m2/h in the outlying areas.
If the specific emission from an area is known, the resulting
concentration in the space above the area can be approximately
calculated by a box model:
E • X 1 V r * v
C = = ' I En xn
L * v L * v n
where C = concentration; E = emission/(area)(time); x = distance
along the emission area; L = mixing height; and v = wind speed,
* 0.
If it is assumed that concentrations are monitored in the
center of circular areas, the distance over the inner city is about
-------
334
GORAN LOFROTH
4 km and over the outlying areas, about 30 km. The mixing height
and the wind speed nay be assigned values of 200 n and 3 m/s,
respectively. Using these figures, the average mutagenic activity
can be calculated as 0.3 revertants/m3 of air. Actual measurements
of mutagenic activity reported in the present study were
appreciably higher, e.g. , 20 to 50 revertants/m^ during the winter
months.
The difference between calculated and measured average
mutagenic activities seemed to be too large to be due solely to the
use of a simple model. Several other explanations are more or less
probable:
1) The actual mutagenic activity of motor vehicle emissions
was 10 to 100 times higher than that from the tested
passenger cars. Such levels would not conform with
reported data (Lofroth, 1979; Ohnishi et al., 1980).
2) The emissions from DI heavy-duty diesel engines were not
considered separately. However, preliminary studies of
exhaust samples from such an engine indicate that the
mutagenic activity is of the same order of magnitude as
that from IDI diesel engines (Rehnberg and Lofroth,
unpublished data).
3) There are other major sources of mutagenic components.
Analyses of nighttime-collected urban particulate matter
indicate that residential heating, etc., was not a major
direct source of the mutagenic activity. Currently, most
of the stationary energy production in the Stockholm area
is from oil combustion. Residential heating is to a large
extent provided by larger district plants, hopefully
having low emissions of organic compounds.
4) Mutagenic components in motor vehicle exhaust samples are
not the same as those in extracts of urban particulate
matter. This explanation is supported by the reported
differences in the mutagenic characteristics of these
samples. These differences and the increased mutagenic
activity could be caused by transformations during the
residence time in the atmosphere or by artifactual
transformations during sampling.
CONCLUSIONS
Several differences in the mutagenic characteristics and in
the level of mutagenic activity have been found between motor
vehicle exhaust samples and urban particulate matter collected
above the rooftops. If major sampling artifacts are absent, the
-------
MUTAGENIC ACTIVITY IN CARBON PARTICULATES AND EXHAUST 335
differences imply that transforraations occur after emission,
changing Che composition of mutagenic compounds. It nay be that
either emissions or actual ambient samples are best suited for
analyses.
Much information is still lacking, including evaluations of
the adequacy of various sampling methods, mutagenicity studies of
samples collected at street level, and mutagenicity studies in
conjunction with measurements of other pollutants, including
nitrogen oxides. Mutagenicity studies in the Salmonella system
can only suggest human health implications, which will ultimately
require further evaluation.
ACKNOWLEDGMENTS
The present project is supported by grants from the National
Swedish Environment Protection Board and the Swedish Natural
Science Research Council.
REFERENCES
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian-microsome mutagenicity test. Mutation Res.
31:347-364.
Cheng, Y-S. , and H-C. Yeh. 1979. Particle bounce in cascade
impactors. Environ. Sci. Technol. 13:139 2-1396.
Ehrenberg, L., S. Hussain, M. Noor Saleh, and U. Lundqvist. 1980.
Nitrous esters - A genetic hazard from nitrogen oxides (N0X)?
Hereditas 92:127-130.
Hughes, M.M., D.F.S. Natusch, D.R. Taylor, and M.V. Zeller. 1979.
Chemical transformations of particulate polycyclic organic
matter. Presented at the 4th International Symposium on
Polynuclear Aromatic Hydrocarbons, Columbus, Ohio.
Lofroth, G. 1978. Mutagenicity assay of combustion emissions.
Cheniosphere 7:791-798.
Lofroth, G. 1979. Salmonella/microsome mutagenicity assays of
exhaust from diesel and gasoline powered motor vehicles.
Presented at the International Symposium on Health Effects
of Diesel Engine Emissions, Cincinnati, Ohio.
Lofroth, G., E. Hefner, I. Alfheim, and M. Miller. 1980.
Mutagenic activity in photo copies. Science 209:1037-1039.
-------
336
GORAN LOFROTH
Ohnishi, Y., K. Kachi, K. Sato, I. Tahara, H. Takeyoshi, and H.
Tokiwa. 1980. Detection of mutagenic activity in automobile
exhaust. Mutation Res. 77:229-240.
Pitts Jr., J.N., D. Grosjean, T.M. Mischke, V.F. Simmon, and D.
Poole. 1977. Mutagenic activity of airborne particulate
organic pollutants. Toxicol. Lett. 1:65-70.
Pitts Jr., J.N. , K.A. Van Cauwenberghe, D. Grosjean, J.P. Schniid,
D.R. Fitz, W.L. Belser Jr., G.B. Knudson, and P.M. Hynds.
1978. Atniospheric reactions of polycyclic aromatic
hydrocarbons: Facile formation of mutagenic nitro
derivatives. Science 202:515-519.
Rosenkranz, H.S., and W.T. Speck. 1976. Activation of
nitrofurantoin to a mutagen by rat liver nitroreductase.
3iochem. Pharmacol. 25:1555-1556.
Rosenkranz, H.S., E.C. McCoy, D.R. Sanders, M. Butler, D.K.
Kiriazides, and R. Merraelstein. 1980. Nitropyrenes:
isolation, identification and reduction of mutagenic
impurities in a carbon black in toners. Science
209:1039-1043.
Talcott, R., and E. Wei. 1977. Airborne mutagens bioassayed in
Salmonella typhinurium. J. Natl. Cancer Inst. 58:449-451.
Talcott, R., and W. Harger. 1979. Mutagenic activity of aerosol
size fractions. EPA 600/3-79-032. U.S. Environmental
Protection Agency: Research Triangle Park, NC.
Tokiwa, H. , H. Takeyoshi, K. Takhashi, K. Kachi, and Y. Ohnishi.
1978. Detection of mutagenic activity in automobile exhaust
emissions. Mutation Res. 54:259-260. (abstr.)
Van Vaeck, L., G. Broddin, and K. Van Cauwenberghe. 1979.
Differences in particle size distributions of major organic
pollutants in ambient aerosols in urban, rural, and seashore
areas. Environ. Sci. Technol. 13:1494-1502.
Wang, C.Y., M-S. Lee, C.M. King, and P.O. Warner. 1980. Evidence
for nitroaromatics as direct-acting mutagens of airborne
particulates. Chemosphere 9:83-87.
-------
MUTAGENIC EFFECTS OF ENVIRONMENTAL PARTICULATES IN THE
CHO/HGPRT SYSTEM
G.M. Chescheir III and Neil E. Garrett
Health Effects Research Program
Northrop Services, Inc.
Research Triangle Park, North Carolina
John D. Shelburne
Department of Pathology
Duke University Medical Center and
Veterans Administration Hospital
Durham, North Carolina
Joellen Lewtas Huisingh and Michael D. Waters
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION
The emission of particulate matter into the atmosphere from
stationary fuel combustion and transportation-related sources is a
serious environmental problem. Natusch (1978) has shown that trace
elements and chemicals associated with particulate matter from coal
combustion nay constitute a health hazard. In an earlier study,
Natusch and Wallace (1974) reported that nany known or potential
carcinogens are preferentially concentrated on the surface of
respirable coal fly ash. According to Davison et al. (1974),
greater quantities of trace elements are associated with particles
of fly ash too small to be trapped effectively by conventional
particulate-control devices. The potential hazard of respirable
particles was brought into sharper focus through studies in which
organic extracts of fly ash particles (Chrisp et al., 1978; Fisher
et al., 1979) and diesel exhaust particulates (Huisingh et al.,
1978) were shown to be mutagenic in Salmonella typhlmurium.
Studies with mammalian cells are necessary to confirm the
mutagenic effects of environmental particulate matter. The Chinese
hamster ovary (CKO) cell is being evaluated as a test system for
such particles. These cells, which form discrete cell colonies in
culture, were shown by Wininger et al. (1978) to be a convenient
system for testing environmental chemicals. In our laboratory we
have shown that the CHO cell is capable of phagocytizing
particulate matter. This characteristic has been exploited to
evaluate the toxicity of a variety of particles (Garrett et al.,
1979, 1980), including samples from coal gasification, fluidized
bed combustion, and conventional coal combustion. Furthermore, we
have shown that the CHO system is useful in determining the
337
-------
338
G.M. CHESHEIR III ET AL.
toxicity of such diverse environmental agents as liquid effluents
from textile mills (Campbell et al., 1979), polychlorinated
biphenyls (PCB)(Garre11 and Stack, 1980) , and organic
condensates from a refuse energy recovery system. Because the CKO
system is useful in evaluating the toxicity of chemical and
particulate matter, the studies have been extended to detect
possible mutagenic effects of atmospheric particles. The CHO cell
assay, which measures mutation at the hypoxanthine-guanine
phosphoribosyl transferase (HGPRT) locus, detects 95% of the
chemicals known to have carcinogenic activity in vivo (Hsie et al.,
1978).
This report examines the use of the CHO/HGPRT system to assay
toxic and mutagenic effects of environmental particles. Since CHO
cells readily phagocytize whole particles, the system provides a
straightforward method for testing mutagenesis of particulate
matter without the complexities of extraction and fractionation.
METHODS
The Chinese hamster ovary cell line (CH0-K1) was obtained from
the American Type Culture Collection and maintained in medium
supplemented with fetal calf serum (screened for virus and
mycoplasma). Exponentially growing stock cultures were harvested
by washing the flasks once with phosphate-buffered saline and then
with 0.257, trypsin. Cells were plated in Ham's F-12 media
containing 107. serum and no antibiotics. The cells were usually
treated with a three-day exposure to 1 uM aminopterin to prevent
the appearance of spontaneous mutants and were subcultured twice
before use in a mutation assay.
Mutation induction at the HGPRT locus was measured by the
method of O'Neill (1977). Ham's F-12 media containing 5% dialyzed
fetal calf serum and antibiotics was used. After trypsinization,
0.5 x 106 cells were transferred to Corning 75-cm2 flasks and
incubated at 37°C. After a 24-h attachment and growth period, the
test sample was added to the medium, and the cultures were
incubated for 20 h. The exposed cells were then washed three times
with saline, and the cells were harvested and counted.
The initial cell survival was determined in experiments in
which aliquots of the cell suspension were added to media in 25-cm2
flasks (300 cells./flask). The flasks were incubated for seven days,
and the colonies were fixed and stained with 0.04% crystal violet.
Mutation induction was determined in cells that were
subcultured every 48 h. The flasks were washed and trypsinized,
and 1 x 106 cells were added to 75-cm2 flasks. After eight days of
culture, cells were plated for selection and post-expression colony
-------
MUTAGENIC EFFECTS OF PARTICULATES IN CHO/HGPRT SYSTEM 339
survival. Cloning efficiency was determined in experiments in
which 200 cells were seeded in hypoxanthine-free media in 25-cra2
flasks. Colonies of mutant cells were obtained in 75-cm2 flasks
after seeding 2 x I05 cells in media without hypoxanthine and with
10 pM 6-thioguanine. The flasks for cloning efficiency and
selection were incubated at 37°C for seven days, and the colonies
were then fixed and stained. The number of phenotypic mutants was
monitored by changing the media in the flasks after five days of
incubation to media without thioguanine but containing hypoxanthine
and 1 pM aminopterin. The flasks were then incubated an additional
five days to develop the colonies.
In each experiment, two replicate cultures were formed for the
negative and positive controls and each concentration of the test
substance. From each replicate, three flasks were derived for
measuring initial cell survival, five for selection, three for the
post-expression survival, and three for determining mutants
resistant to aminopterin.
Phagocytic activity was determined by adding particulate
samples to CHO cells cultured in Lab-Tek microslides. Before
particles were added, the cells were incubated for a 24-h
attachment period at 3 7 0 C in a humidified atmosphere of 5% carbon
dioxide in air. After a 24-h incubation with the particles, the
cells were washed three times with saline, fixed with methanol, and
stained with May Grunwald and then Giemsa solutions. The cells
were washed with water and then with acetone-xylene solutions.
Phagocytosis was confirmed by preparing electron micrographs
of cells exposed to particulate samples in 25-cm2 flasks. The
cells were preincubated 24 h at 37"C. After treatment with the
particles for 20 h, the cells were fixed overnight with
glutaraldehyde in Millonig's phosphate buffer. After post-fixation
with osmium tetroxide, the monolayers were stained en bloc with
uranyl acetate in water, dehydrated with ethanol, and embedded.
Sections were examined with a transmission electron microscope.
Size analysis was performed using a Coulter Counter TA II
after counting 10,000 to 20,000 particles. The instrument was
calibrated using a 100-um aperture and standard 10.05-yn
polystyrene latex sphere. Filtered (0.45 urn) physiological saline
was used as the electrolyte. Samples of 1 to 2 mg were added to
polystyrene test tubes and vortexed for 2 min. Saline (5 ml) was
then added, and the tubes were vortexed for 30 sec. Immediately
before counting, each tube was vortexed for 10 sec, and 0.1 to 1.0
ml of the sample was added to 100 ml saline. The particles were
sized, and population mean diameters for particles > 2 ua were
determined from cumulative plots.
-------
340
G.M. CHESHEIR III ET AL.
Data reduction and statistical analysis was performed using
modified versions of programs written for the Texas Instruments
TI-59 calculator (Garrett and Stack, 1980).
RESULTS AND DISCUSSION
The response of the CHO cells to known mutagenic agents was
checked using N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) and
cis-dichlorodiamine platinura-II (Pt[NH3]2CI2)• The platinum
complex presumably causes base substitution in DNA by miscoding.
MNNG, an alkylating agent, is a more potent mutagen in the CHO
system. Both compounds increased the mutation frequency with
increasing dose. The effect of platinum on mutation frequency in
the CHO cells is shown in Figure 1A. The dose response was linear
with a correlation coefficient of 0.99. A concentration of 1.5
Mg/ml platinum yielded a mean value of 96 mutant colonies/10 flasks
in 12 independent experiments. This dose of platinum gave a
mutation frequency of 1.16 ± 0.38 x L0~** and was selected as a
positive reference against which to measure the mutagenic effects
of environmental particles. Spontaneous mutation in the untreated
(negative control) cells gave rise to a mean of 5 mutant colonies/
10 flasks (N=23) and a mutation frequency of 5.3 ± 3.8 x 10~6.
The range of mutagenic responses obtained for the negative and
positive control is shown in Figure LB. A logarithmic plot
normalizes the standard deviation of the data and illustrates the
structure of the distributions. Other investigators have
transformed CHO mutation data to correct for the nonhoraogeneity of
variances and non-normal distributions in test cultures and in
untreated controls (Irr and Snee, 1979). Figure IB shows that the
mean mutation frequency for the positive reference was
approximately 10 times the maximum frequency for the untreated
control, facilitating a comparison with weakly mutagenic substances.
The size distribution of particles tested in the mutation
assay is shown in Figure 2. Exhaust particles from two diesel
engines had a similar particle size distribution (Figure 2, A and
B). Brief sonication of one diesel sample apparently dissociated
the particulate, so that a larger number of small particles was
observed (Figure 2C). Fly ash from oil combustion (Figure 2D) was
also a heterogeneous mixture with a size distribution similar to
that of the diesel particles. The fly ash from coal combustion
tested in these experiments was < 3 ym (Figure 2r) and 2 to 5 tin
(Figure 2H). Iron oxide particles (Figure 2E) were similar in size
distribution to the coal fly ashes and were used in these
experiments as a reference particulate. Silica particles, obtained
from a commercial source, were 4 to 9 ym (Figure 2G).
-------
MUTAGENIC EFFECTS OF PARTICULATES IN CHO/HGPRT SYSTEM
341
T
o
u.
5
1 -
I N= 12
N=23
1.0 2.0
DOSE (/ag/'ml)

>
LU
8 -
6 -
O 4
O
^ CONTROLS
I 1 1.5 jig/ml Pt(NH3)2 Cl2

10 100
MUTATION FREQUENCY (x 10"6}
Figure I. Mutation frequencies (MF) obtained in uncreated controls
and CHO cells treated with Pt(NH3)2Cl2« A. The effect
of ?t(NH3)2Cl2 on mutation with increased dose (0 to 3
Ug/cil). B. Histogram of the mutagenic responses
obtained in control CHO cells and cells treated with 1.5
ug/ml of Pt(NH3)2Cl2- There were three events with
mutation frequencies equal to zero.
-------
Intentionally Blank Page
-------
MUTAGENIC EFFECTS OF PARTICULATES IN CHO/HGPRT SYSTEM
343
The relative size of particles ingested by CHO cells is shown
in light and electron micrographs (Figures 3 through 5). Figure 3
shows control cells with very prominent nucleoli and high ratios of
nuclear material to cytoplasm; no phagocytic material is seen.
Treated cells examined by phase contrast microscopy (not shown) and
fixed and stained cells examined by light microscopy revealed a
large number of particles in the cytoplasm of CHO cells. These
particles were frequently arranged closely around the nucleus.
Electron microscopy of sections of fixed and embedded cells
confirmed that the particles were In the cells, often closely
apposed to the nucleus. The treated cells differed from the
control cells in that a membrane could often be seen around each
particle or group of particles, forming a very large membrane-
limited phagosome. Figure 4 shows CKO cells in which particles of
fly ash were present in phagosomes. Frequently the particles
either sectioned poorly or appeared to have fallen out of the
section. The phagosomes of cells treated with diesel particulates
contained small amorphous particles of the diesel exhaust material
and empty regions suggesting fluid accumulation (Figure 5).
These experiments provided evidence that a variety of
particulates of environmental concern are trapped close to the cell
nucleus. The effect of these particles on mutation was
investigated by studying both the particles alone and particles In
combination with MNNG and Pt(NH3)2Cl2« The latter experiments
tested the possibility that the particles could facilitate passage
of another mutagen to the cell nucleus, with subsequent damage to
DMA. Iron oxide, silica, and fly ash particles from coal
combustion did not produce a statistically significant difference
in mutation frequency in cells treated simultaneously with MNNG and
platinum (Figure 6, A and C). Fly ash and silica did contribute to
cell toxicity (Figure 6, B and D). The toxic effect of iron oxide
with mutagens was not different from that of the mutagen alone.
Because of the high baseline mutation of MNNG and Pt(NH3)2Cl2,
these experiments would detect relatively large influences on
mutation, but not small changes due to the particles alone.
Additional experiments tested the effects of the particles without
an exogenous mutagen. As shown in Figure 7, A and B, six particles
and one organic extract produced mutation in excess of the control
values. These substances were toxic to the cell cultures to
varying degrees (Figure 6, C and D). The model particles of coal
fly ash (2 to 5 un) produced a small increase in mutation frequency
(17 x 10~6 versus 8.4 x I0~6 for controls, at a dose of 125 ug/ml).
Another sample of fly ash (0 to 3 um) from conventional coal
combustion caused a four-fold increase in mutation (9.8 x 10"^ at
250 ug/ml versus 2.5 x 10-^ for controls. Similarly, fly ash from
oil combustion increased the mutation frequency (29.1 x I0-6 at
75 ug/ml versus 11.8 x 10~6; p » 0.15). Two samples of exhaust
-------
344
G.M. CHESHEIR III ET AL.

Figure 3. Light micrograph (A) and electron micrograph (B) of
control CHO cells. The cells were magnified 200 times
in the light micrograph and 5000 times in the electron
micrograph. The control cells exhibit prominent
nucleoli and high nuclear-to—cytoplasm ratios. No
phagocytized material is seen.
-------
MUTAGENIC EFFECTS OF PARTICULATES IN CHO/HGPRT SYSTEM 345
Figure 4. Light micrograph (A) and electron micrograph (B) of CHO
cells after a 20-h exposure to coal fly ash. In A,
cells were exposed to 200 ;jg/ml of fly ash, < 3 iim in
diameter, and magnified 1000 times; in B, cells were
exposed to 250 yg/tnl of 2 to 5 gm coal fly ash and
magnified 7 100 times. Many fly ash particles are
visible in the cytoplasm, immediately adjacent to the
cell nucleus.
-------
346
G.M. CHESHEIR III ET AL.
Figure 5. Light micrograph (A) and electron micrograph (B) of CKO
cells after a 20-h treatment with sonicated diesel
(No. 1) particles at 100 ug/ml. A, cells were
magnified 1000 times, and in B, 5000 tines. Phagosomes
of the treated cells contain small amorphous particles
of the diesel exhaust material. Particles are closely
associated with the cell nucleus.
-------
MUTAGENIC EFFECTS OF PARTICULATES IN CHO/HGPRT SYSTEM
347
CH CONTROLS
o.ooh
z
o
c
5® 60 ¦
20 -
I'oc r
u.
o
! 60h
20 ¦
_L
MNNG
a
FLV ASH
VNSG
a
FSOs
MNNG
9
SILICA
P»
a
*IY fiSH
?*
a
Ft,0,
P!
a
SILICA
Figure 6. Mutagenicity (A, C) and toxicity (3, D) of particles in
combination with MS3JG and Pt(NH3)2Cl2« Mutagenicity is
expressed as a percent of the mutation frequency of MNNG
or Pt(NH3)2Cl2* CHO cells were exposed to 0 to 3 um
fly ash at 200 pg/ml, ferric oxide particles at 100
Ug/ml, and silica particles at 1000 ug/ml. Toxicity was
measured as the initial cell survival in the CHO cloning
assay. Toxicity data are expressed as a percent of the
untreated control.
particles from diesel engines produced about a three-fold increase
in mutation over the control values. One sample of particles
increased mutation from 2.2 x 10~^ for the control to a value of 5
x 10-^ at 750 ug/ml. A. methylene chloride extract of this diesel
sample increased mutation to five times the control frequency
-------
348
G.M. CHESHEIR III ET AL.
100-
3 60-
20 ¦
a
:on~«xs
ioh
r
n
COAL fur ASH COALfiraSH
c-Sg 0-3u
CiESEL". CIESlL-i
jONICA'E
OiESEL 2 3'CSEL 2
•X7RACT
Fieure 7. The effect of environmental particulates on mutation
frequency (A, B) and toxicity (C, D) in CHO cells. CHO
cells were exposed to oil fly ash at 75 ug/ml, 2 to
5 pin coal fly ash at 125 ug/ml, 0 to 3 gm coal fly ash
at 250 ug/ml, diesel exhaust particles (No. 1) at 500
ug/ml, sonicated No. 1 diesel particles at 100 ug/ml,
diesel exhaust (No. 2) at 750 ug/ml, and a methylene
chloride extract of No. 2 diesel particles at 50 ug/ml.
Toxicity or cells per milliliter is expressed as percent
of the untreated control.
(10.4 x 10 ^ at 50 ug/ml versus 2.2 x 10-^ for the control;
p = 0.078). Higher concentrations of the extract were severely
toxic. Another diesel particle was evaluated with and without
sonication and mutation was also increased (13.4 x 10-f> at
-------
MUTAGENIC EFFECTS OF PARTICULATES IN CHO/HGPRT SYSTEM 349
500 ng/ml for untreated particles (p = 0.12); 14.9 x 10-^ at
100 Mg/ml for sonicated particles (p » 0.059); the control rate was
was 4.9 x LO-6).
These data are, to our knowledge, the first to demonstrate
that whole particles of fly ash from coal and oil combustion and
exhaust particles from diesel engines can cause mutation in a
mammalian cell culture system. For these experiments, it was not
necessary to extract the particles with solvents. These results
suggest that this test system might be used to assay environmental
particulates that have not been previously treated with solvents.
Such a test would more accurately reflect natural exposure to
particulate matter in the atmosphere, since extraction with organic
solvents and subsequent concentration of the extract have no
counterparts in natural exposure by inhalation.
REFERENCES
Campbell, J.A., N.E. Garrett, J.L. Huisingh, and M.D. Waters.
1979. Cellular toxicity of liquid effluents from textile
mills. In: Symposium Proceedings: Textile Industry'
Technology. EPA-6C0/2-79-104. U.S. Environmental Protection
Agency: Research Triangle Park, NC. pp. 239-248.
Chrisp, C.E., G.L. Fisher, and J.E. Lammert. 1978. Mutagenicity
of filtrates from respirable coal fly ash. Science 199:73-75.
Davison, R.L., D.F.S. Natusch, J.R. Wallace, and C.A. Evans. 1974.
Trace elements in fly ash. Dependence of concentration on
particle size. Environ. Sci. Technol. 8:1107-1113.
Fisher, G.L., C.E. Chrisp, and O.G. Raabe. 1979. Physical factors
affecting the mutagenicity of fly ash from a coal-fired power
plant. Science 204:879-881.
Garrett, N.E., J.A. Campbell, J.L. Huisingh, and M.D. Waters. 1979.
The use of short term bioassay systems in the evaluation of
environmental particulates. In: Proceedings of the Symposium
on the Transfer and Utilization of Particulate Control
Technology, Vol. 4. EPA-600/7-79-044d. U.S. Environmental
Protection Agency, Research Triangle Park, NC. pp. 175-186.
Garrett, N.E., G.M. Chescheir III, N.A. Custer, J.D. Shelburne,
J.L. Huisingh, and M.D. Waters. 1980. An evaluation of the
cytotoxicity and mutagenicity of environmental particulates in
the CHO/HGPRT system. In: Proceedings of the 2nd Symposium
on the Transfer and Utilization of Particulate Control
Technology, Vol. 4. E?A-600/9-80-039d, U.S. Environmental
Protection Agency: Research Triangle Park, NC- pp. 524-535.
-------
350
G.M. CHESHEIR III ET AL.
Garrett, N.E., and H.F. Stack. 1980. Cellular toxicology data
analysis using a programmable calculator. Comput. 3iol.
Med. 10:153-167.
Hsie, A.W. , J.P. O'Neill, J.R. San Sebastian, D.8. Couch, P.A.
Brimer, W.N.C. Sun, J.C. Fuscoe, N.L. Forbes, R. Machanoff,
J.C. Riddle, and M.H. Hsie. 1978. Quantitative mammalian
cell genetic toxicology: Study of the cytotoxicity and
mutagenicity of seventy individual environmental agents
related to energy technologies and three subtractions of a
crude synthetic oil in the CHO/HGPRT system. In: Application
of Short-term Bioassays in the Fractionation and Analysis of
Complex Mixtures. M.D. Waters, S. Nesnow, J.L. Huisingh, S.S.
Sandhu, and L. Claxton, eds. Plenum Press: New York,
pp. 292-315.
Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R. Zweidinger,
S. Tejada, J. Bumgarner, F. Duffield, M. Waters, V.F. Simmon,
C. Hare, C. Rodriguez, and L. Snow. 1978. Application of
bioassay to the characterization of diesel particle emissions.
In: Application of Short-term Bioassays in the Fractionation
and Analysis of Complex Environmental Mixtures. M.D. Waters,
S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
Plenum Press: New York. pp. 381-M8.
Irr, J.D., and R.D. Snee. 1979. Statistical evaluation of
mutagenicity in the CHO/HGPRT system. In: Banbury Report 2.
Mammalian Cell Mutagenesis'. The Maturation of Test Systems.
A.W. Hsie, J.P. O'Neill, and V.K. McElheny, eds. Cold Spring
Harbor Laboratory: New York. pp. 263-275.
Natusch, D.F.S. 1978. Potentially carcinogenic species emitted to
the atmosphere by fossil-fueled power plants. Environ. Hlth.
Perspect. 22:79-90.
Natusch, D.F.S., and J.R. Wallace. 1974. Urban aerosol toxicity:
The influence of particle size. Science 186:695-699.
O'Neill, J.P. 1977. A quantitative assay of mutation induction
at the hypoxanthine-guanine phosphoribosyl transferase locus
in Chinese hamster ovary cells (CHO/HGPRT system):
Development and definition of the system. Mutation Res.
45:91-101.
Wininger, M.T., F.A. Kulik, and W.D. Ross. 1978. In vitro clonal
cytotoxicity assay using chinese hamster ovary cells (CH0-K1)
for testing environmental chemicals. In Vitro 14:381.
-------
A PRELIMINARY STUDY OF THE CLASTOGENIC EFFECTS OF DIESEL
EXHAUST FUMES USING THE TRADESCANTIA MICRONUCLEUS BIOASSAY
Te-Hsiu Ma and Van A. Anderson
Department of Biological Sciences and
Institute for Environmental Management
Western Illinois University
Macomb, Illinois
Shahbeg S.Sandhu
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION
Automobile engine emissions are one of the major urban air
pollutants in the industrialized nations. The direct and indirect
effects of these complex agents on human health are crucial
problems for modern society. Epidemiological studies carried out
in Europe (Barth and Blacker, 1978; Blumer et al., 1977a, b) and
Japan (Shimizu et al. , 1977) found that cancer mortality rates were
higher in populations near heavily travelled highways than in those
away from them. The increasing popularity of diesel-engine-powered
vehicles, especially passenger cars, demands a better understanding
of the health effects of diesel emissions.
Current information on the health effects of diesel engine
exhaust fumes is limited to the results of laboratory tests with
experimental animals. In an early review, Goldsmith (1964)
concluded that low concentrations of diesel exhaust fumes stimulate
the growth of cultured mammalian cells. Studies using mammals such
as cats, rats, mice, and Guinea pigs have shown that aryl
hydrocarbon hydroxylase increases in the lungs, liver, and
prostate; however, no pathological change is seen in the lungs of
exposed animals (see Barth and Blacker, 197S, for a review).
Campbell et al. (1979), in their studies of the effects of light-
duty engine diesel exhaust fumes on mice, found increased
susceptibility to infection and mortality rates. Guerrero et al .
(1979) found that the sister-chromatid exchange rate in golden
hamsters was not altered at a dose of 20 mg/aninal for 24 h, while
in another study (Pereira et al., 1979) a six-month treatment gave
negative results in the bone marrow micronuclei bioassay, but
positive results in the sperm morphology test. Other studies have
351
-------
352
TE-HSIU MA ET AL.
shown biochemical changes in the lungs of rats and Guinea pigs
exposed Co diesel exhaust fumes. At doses of 250 to 1,500 ,
20 h/day, 5.5 days/week, for 12 to 24 weeks, prostaglandin
dehydrogenase activity was reduced (Chaudart and Dutta, 1979). A
36-week exposure caused increases in lung weight and in lipids and
collagen in lung tissue (Misioroski et al., 1979). In strain A
mice, inhalation of diesel exhaust fumes for seven months showed no
effect on lung structure (Orthoefer, 1979), and inhalation for 28
weeks did not cause changes in sperm morphology (Pereira et al.,
1979b). Negative results were also obtained in the Drosophila
recessive mutation test, after an 8-h exposure to diesel exhaust
fumes containing 11.6 ppm of hydrocarbons (Schuler and Niemeier,
1979) .
The whole-animal bioassays for diesel exhaust fume effects are
time-consuming and costly. An alternative approach to this urgent
problem would be using a battery of short-term bioassays similar to
those proposed by Huisingh et al. (1979) for diesel exhaust
particulates. The Tradescantia (spiderwort) micronucleus (Trad-
MCN) is a very sensitive and quick bioassay (Ma, 1980) that is
especially suitable for testing gaseous agents in ambient air (Ma
et al., 1980), as well as in chambers (Ma et al., 1978; Ma, 1979).
The reliability and efficiency of this bioassay were verified by
tests of well-known mutagens (Ahmed and Ma, 1980; Ma, 1979; Ma and
Anderson, 1979; Ma et al., 1978, 1980a) and dose-response curves
established using 1,2-dibromoethane (Ma et al., 1978) and X-ray
treatments (Ma et al., 1980b). Therefore, the Trad-MCN bioassay
was used to determine the clastogenic effects of total diesel
exhaust fumes in the present study.
MATERIALS AND METHODS
The Tradescantia paludosa clone 03 was used for all
experiments. A population of this clone was cultivated in the Duke
University Phytotron, Durham, NC, under optimal growing conditions
in clean air. The plant cuttings bearing young inflorescences
(about 5 cm long) were maintained in tap water in a plastic cup
(200 ml) before, during, and after treatment. Generally, 15 to 20
cuttings constituted an experimental group, and each experiment
involved three groups treated with different doses, one baseline
control group (in a clean-air room), and one field control group
(in the experimentation area, outside of the treatment chamber).
One group exposed to 40 R of X-rays served as the positive control
for all experiments.
Two pilot studies were conducted to establish an adequate dose
rate and total dose for both the concentration of fumes and the
duration of exposure. The final series of experiments was carried
out at three different concentrations (1/200, 1/45, and 1/23
-------
CLASTOGENIC EFFECTS OF DIESEL EXHAUST FUMES
353
dilutions) of Che total exhaust fumes. The fumes were generated by
a Nissan (1970 Datsun) diesel automobile at the Emissions
Measurements Characterization Division, Environmental Sciences
Research Laboratory, U.S. Environmental Protection Agency, Research
Triangle Park, NC. The automobile ran on a chassis dynamometer
under a simulated city driving cycle (23 min/cycle). The diluted
exhaust fumes were introduced into a specially designed Plexiglas
chamber (31-1 capacity) at a flow rate of 32 1/min. Three groups
of plant cuttings were each treated with a different dose by
exposing them to a given concentration of fumes in the treatment
chamber for one, two, or three driving cycles. Gas flow was
interrupted for 5 min between the driving cycles. The temperature
in the chamber was about 28°C, the relative humidity was in the
range of 50 to 80%, and the atmospheric pressure was 0.99 to 1.0
bar (745 to 750 inmHg). Nitrogen oxides, hydrocarbons, carbon
monoxide, and carbon dioxide in each of the dilutions of fumes were
sampled and analyzed during the treatment. The concentrations of
these gases are given in Table 1. The doses of each treatment
group were derived from a combination of dilution factor and number
of driving cycles.
There was a 30-h recovery time after the end of the treatment,
to allow the raeiotic prophase I pollen mother cells to arrive at
the early tetrad stage, where the broken chromosomes appear as
micronuclei (MCN) outside of the nucleus proper. The young
inflorescences of treated and control groups were fixed in aceto-
alcohol (1:3) after the recovery period; the fixed samples were
transferred into 70% ethanol after 24 to 48 h of fixation.
Microslides of the tetrad stage of pollen mother cells were
prepared by the aceto-carmine squash method. The frequency of MCN
on each slide was expressed as the number of MCN per 100 tetrads.
An average of 350 tetrads were observed on each slide. The mean
frequency and the standard error of the mean for each experimental
group was derived from an examination of five slides. The treated
and control groups were statistically compared using the standard
error of the difference in means, with a significance level of
0.01 .
RESULTS AND DISCUSSION
Since the end point of this bioassay is the frequency of MCN
in tetrads, and MCN represent the broken pieces of chromosomes, the
diesel exhaust fumes effect demonstrated in this study is
appropriately referred to as a clastogenic rather than a mutagenic
effect. The preliminary data from three series of experiments are
shown in Table 1. The only significantly positive results were for
DEF-2, T-2, and DEF-3, T-3. All lower doses (fewer driving cycles
or lower concentration of fumes or both) gave negative results.
The negative result for DEF-2, T-3 was due to an overdose, a common
-------
Co
CJI
Concentration of Gases3
ExperLmenldl
Groups
(Treatment = No.
of Driving Cycles)
DKK-l
Baseline control
Kield control
Treatment 1
Treatment 2
Treatment .3
DEF-2 and -3
Baseline control
Field control
DEK-2
Trfiat.mont 1
Treatment 2
Treatment 3
DEF-3
Treatment I
TreaLment 2
Treatment 3
HCb
( ppm)
3.04
3.04
3.04
46. 95
47.36
4 3.84
18.04
20.65
2 1 .00

MCN per

NO
CO
C02
100 Tetrads
Sl arnlard
S igni f ic ance
( ppm)
(ppm)
(X)
(mean)
Error
(p < 0.01)

8.32
1 .63

5.67
0.61

1.5
0.9
0.04
5.45
0.68
-
1.5
0.9
0 .04
5.97
0.67
-
1 .5
0.9
0.04
7.75
0. 78
—

4. 76
0,90

5.04
0. 70

18.0
18.0
0.49
8. 56
2.07
_
17.0
16.0
0.49
20.81
3.13
+
18.0
I 7.0
0.49
10 .90
1 .68
-
8.8
8.0
0.23
7. 35
0. 32
-
10.0
9.5
0.25
7. 54
I .62
-
9.5
8.5
0.25
00
o
0.44
+

62 .06
3.10
+
Positive! control (40 R X-rays)
a,I1icsp. concent rat ions were obtained from the constant volume sampler before a tenfold dilution.
L
°HC = Hydrocarbons.
H
M
X
w
I—I
C
s
>
M
>
f
-------
CLASTOGENIC EFFECTS OF DIESEL EXHAUST FUMES
355
phenomenon in this test system. Such overdose samples usually have
relatively low frequencies of MCN accompanied by abortive cells in
tetrads and/or microspores.
The concentrations of fumes that induced MCN were probably
comparable to some concentrations measured in a separate in situ
monitoring project. Samples obtained through 1- to 6-'n monitoring
at public parking garages, bus stops, and truck stops (Ma et al.,
1980a) sometimes gave positive results in the Trad-MCN bioassay.
According to Zdrazil and Picha (1978), the particulate exhaust
concentration is around 8.21 ug/m^ at the orifice of the exhaust
pipe, and around 0.205 ug/m^ in an ordinary working garage. These
figures may help to indicate the actual magnitude of exhaust
pollution.
In another study (Ahmed and Ma, 1980), the Trad-MCN bioassay
was used concurrently with a human lymphocyte chromosome aberration
bioassay to establish the relative effectiveness of these two
tests. Results of such comparative studies may be used to
extrapolate to human cell systems the mutagenicity of a given agent
as determined in the Trad-MCN assay.
The Tradescantia stamen hair mutation test (Schairer et al.,
1978) is also capable of detecting the mutagenic effect of diesel
exhaust fumes. Although no report on direct testing of diesel
exhaust fumes in the laboratory is yet available, Tradescantia
stamen hair in situ tests were used at interstate highway junctions
and in downtown areas of several large American cities. This kind
of in situ monitoring is comparable to the present laboratory
study. The stamen hair mutation test, which measures the rate of
somatic mutation at a particular locus, and the Trad-MCN test,
which measures the frequency of chromosome damage, are an ideal
combination of bioassays for better assessment of the effect of
diesel exhaust fumes on living systems.
ACKNOWLEDGMENTS
The authors wish to express their deepest appreciation to Dr.
Roy Zweidinger and the staff of the Emissions Measurements
Characterization Division, Environmental Sciences Research
Laboratory, U.S. Environmental Protection Agency, Research Triangle
Park, NC, for providing diesel exhaust fumes and their dilution,
characterization, and measurements; and to Dr. Henry Hellmers,
Director of the Duke University Phytotron, Durham, NC, for raising
the Tradescantia plants.
-------
356
TE-HSIU MA ET AL.
REFERENCES
Ahmed, I., and T.H. Ma. 1980. Chromosome breakage induced by
maleic hydrazide in cultured human lymphocytes and
Tradescantia pollen mother cells. Environ. Mutagen. 2:287.
(abstr.).
Barth, D.S., and S.M. Blacker. 1978. The EPA program to assess
the public health: Significance of diesel emissions. J. Air
Pollut. Control Assoc. 28:269-771.
Bluraer, M., W. Bluraer, and T. Reich. 1977a. Polycyclic aromatic
hydrocarbons in soil of a mountain valley: Correlation with
highway traffic and cancer incidence. Environ. Sci. Technol.
1 1: 1082-1084.
Blumer, W., M. Blumer, and T. Reich. 1977b. Carcinogenic
hydrocarbons and the incidence of cancer mortality among
residents near an automobile highway. Fortschro. Med.
95:1497-1498 and 1551-1552.
Campbell, K.I., E.L. George, and I.S, Washington, Jr. 1979.
Enhanced susceptibility to infection in mice after exposure to
diluted exhaust from light duty diesel engines. Presented at
the International Symposium on Health Effects of Diesel Engine
Emissions, Cincinnati, OH.
Chaudhart, A., and S, Dutta. 1979. Effect of exposure to diesel
exhaust on pulmonary prostaglandin dehydrogenase (?GDH)
activity. Presented at the International Symposium on Health
Effects of Diesel Engine Emissions, Cincinnati, OH.
Goldsmith, G. 1964. Air pollution and health. Science
145:184-186.
Guerrero, R.R., D.E. Rounds, and J. Ort'noefer. 1979. Sister
chromatid exchange analysis of Syrian hamster lung cells
treated in vivo with diesel exhaust particulates. Presented
at the International Symposium on Health Effects of Diesel
Engine Emissions, Cincinnati, OH.
Huisingh, J., S. Nesnow, R. Bradow, and M. Waters. 1979.
Application of a battery of short term mutagenesis and
carcinogenesis bioassays to the evaluation of soluble organics
from diesel particulates. Presented at the International
Symposium on Health Effects of Diesel Engine Emissions,
Cincinnati, OH.
-------
CLASTOGEKIC EFFECTS OF DIESEL EXHAUST FUMES
357
Ma, T.H. 1979. Micronuclei induced by X-rays and chemical
mutagens in meiotic pollen mother cells of Tradescantia—a
promising mutagen test system. Mutation Res. 64:307-313.
Ma, T.H., and V.A. Anderson. 1979. Micronuclei induced by
ultraviolet light and chemical mutagens in meiotic pollen
mother cells of Tradescantia. 10th Annual Meeting of the
Environmental Mutagen Society, New Orleans, LA.
Ma, T.H., V.A. Anderson, and I. Ahmed. 1980a. In situ monitoring
of air pollutants and screening of chemical mutagens using
Tradescantia Micronucleus Bioassay. Environ. Mutagen. 20:287.
(abstr .).
Ma, T.H., G.J. Kontos, Jr., and V.A. Anderson. (1980b). Stage
sensitivity and dose response of meiotic chromosomes of pollen
mother cells of Tradescantia to X-rays. Environ. Exp. Bot.
20:169-174.
Ma, T.H., A.H. Sparrow, L.A. Schairer, and A.F. Nauman. 1978.
Effect of 1,2-dibromoethane (DBE) on meiotic chromosomes of
Tradescantia. Mutation Res. 58:251-158.
Misioroski, R.L., K. Strora, and M. Schvapil. 1979. Lung
biochemistry of rats chronically exposed to diesel
particulate. Presented at the International Symposium on
Health Effects of Diesel Engine Emissions, Cincinatti, OH.
Orthoefer, J.G. 1979. The strain A mouse as an inhalation
carcinogenesis model in diesel exhaust research. Presented at
the International Symposium on Health Effects of Diesel Engine
Emissions, Cincinnati, OH.
Pereira, M.A., P.S. Sabharval, C. Ross, P. Kaur, A. Choi, and T.
Dixon. 1979a. In vivo studies on the mutagenic effects of
inhaled diesel exhaust. Presented at the International
Symposium on Health Effects of Diesel Engine Emissions,
Cincinnati, OH.
Pereira, M.A., P.S. Sabharwal, and A.J. Wyrobek. 1979b. Sperm
abnormality bioassay of mice exposed to diesel exhaust.
Presented at the International Symposium on Health Effects of
Diesel Engine Emissions, Cincinnati, OH.
Schairer, L.A., J. Van1t Hof, C.C. Hayes, R.M. Burton, and F.J. de
Serres. 1978. Exploratory monitoring of air pollutants for
mutagenicity activity with the Tradescantia stamen hair
system. Environ. Hlth. Perspect. 27:51-60.
-------
358
TE-HSIU MA ET AL.
Schuler, R.L., and R.W. Niemeier. 1979. A study of diesel
emission on Drosophi1 a. Presented at the International
Symposium on Health Effects of Diesel Engine Emissions,
Cincinnati, OH.
Shimizu, H., K. Aoki, and T. Kuroishi. 1977. An epidemiological
study of lung cancer in relation to exhaust gas from cars.
Lung Cancer 17:103-112.
Zdrazil, J., and F. Picha. 1978. Carcinogenic hydrocarbons from
exhaust fumes in the working atmosphere. Cesk. Hvg.
8:344-348.
-------
ABILITY OF LIVER HOMOGENATES AND PROTEINS TO REDUCE THE
MUTAGENIC EFFECT OF DIESEL EXHAUST PARTICULATES
Yi Y. Wang and Eddie T. Wei
School of Public Health
University of California
Berkeley, California
INTRODUCTION
An estimate by the U.S. Environmental Protection Agency (EPA)
indicates that by 1985, 25% of the automobiles produced in the
United States may be diesel powered, because diesel engines provide
greater fuel economy Chan do spark-ignition gasoline engines
(Santodonato et al . , 1978). Automotive diesel engines may emit
from 30 to 80 times more particulates Chan a comparable gasoline
engine (Springer and Baines, 1977). Concern has been expressed
about the potential health hazards of these particulates, since
extracts of diesel particulates contain chemicals that are
mutagenic in the Ames Salmonella typhimurium bioassay (Huisingh et
al., 1978; Wei et al., 1980), a short-term test for determining the
mutagenic and carcinogenic potential of chemicals (Ames and McCann,
1976).
The majority of the mutagenic activity in extracts of diesel
exhausts does not require mammalian liver enzymes for activation.
In fact, the addition of rodent liver horaogenates (5-9) in the Ames
assay mixture decreases mutagenic activity (Huisingh et al., 1978;
Wei et al., 1980; Clark and Salraeen, 1980). The mutageniciCy-
reduction effects of S-9 are also observed when S-9 is added to
mutagenic extracts of airborne particulates (Talcott and Wei, 1977)
and gasoline-engine exhausts (Wang et al., 1978). Furthermore, S-9
reduces the mutagenic activity of chemicals such as sodium azide,
sodium dichromate, sodium nitrite, and 5-nitro-2-furoic acid
(deFlora, 1978).
359
-------
360
YI Y. WANG AND EDDIE T. WEI
Some investigators have suggested that the mutagenicity-
reduction effects of S-9 are due to enzymatic degradation of the
mutagen (Clark and Salmeen, 1980; deFlora, 1978); however, no
experimental evidence has been brought forth to substantiate this
hypothesis. We have examined here the properties and components of
S-9 that account for its mutagenicity-reduction effect on diesel
exhaust samples. The results indicate that the mutagenicity-
reduction effect of S-9 is due to non-enzymatic binding of mutagens
to liver proteins.
MATERIALS AND METHODS
The defined diesel exhaust sample used in this study was
obtained from General Motors Research Laboratories (GM), Warren,
MI, Details on the collection and characterization of the sample
have been described by Schreck et al. (1978). The defined sample
was collected on a mini baghouse filter attached to a 2.1-1
Peugeot diesel engine. The engine was operated on a water-brake
dynamometer at a 69-km/tn cruise condition and 9 kW (12 hp). Diesel
fuel #2 was used. Particulates were weighed, mixed with an
appropriate volume of dimethylsulfoxide (DMSO), sonicated for
several minutes, and then vortexed immediately before bioassay.
The Ames test procedures were followed without modification (Ames
et al. , 1975). The Salmonella strain employed was TA98, which has
been shown to be the most sensitive strain for detecting diesel
exhaust mutagens (Huisingh et al., 1978; Wei et al., 1980). The
post-mitochondrial supernatant fraction of rat liver homogenates
(S-9) was prepared from Aroclor-1254-treated rats, according to the
standard methods of Ames et al. (1975). The cofactors were added
to the S-9 prior to bioassay. One half milliliter of S-9 mix,
containing 50 ul of S-9, was applied on each plate. All
experiments have been repeated at least once, and all samples have
been tested in duplicate in each experiment.
In some experiments, S-9 was separated according to the
procedure of Frantz and Mailing (1975). The enzymatic activity in
S-9 was inactivated by heat or by omission of the NADPH-generating
system. S-9 mix was heated in a boiling water bath (100'C) for
five minutes. The boiled S-9 mix was cooled at room temperature
and vortexed before being applied to the assay. Part of the heat-
treated S-9 was filtered through a 0.45-Mm Millipore filter unit to
remove coagulated proteins. The filtered or unfiltered heat-
treated S-9 mix was then used in the bioassay. The NADPH-
generating system was omitted from the S-9 mix by replacement of
the cofactor mix with an equivalent volume of sodium phosphate
buffer.
-------
ABILITY OF LIVER HOMOGENATES TO REDUCE MUTAGENICITY
361
Albumin was removed from 'he cytosol fraction by gel column
chromatography. In this experiment, uninduced perfused rat liver
noraogenates were used. The column was prepared according to the
method of Cuatrecasas ( 1970). Agarose gel (Sepharose 4B-CL:
Pharmacia, Sweden) was activated with cyanogen bromide
(approximately 250 mg/ral of settled gel) and the cyanogen-bromide-
agarose was stirred overnight ac 8 to 10°C in 2 vol of 0.1 M sodium
carbonate, pH 9, containing 30 mg/ml JJ-aminocaproic acid. The
£-aminocaproic-acid-agarose produced was washed, resuspended in two
volumes of methanol containing 20 mg/ral L-tryptophan methyl ester
and 20 mg/ml dicvclohexylcarbodiimide, and stirred for 4 h at room
temperature. The gel was washed with several volumes of methanol
prior to resuspension in 50 mM monobasic potassium phosphate.
Cytosol fractions were passed through the column, and the filtrates
collected were termed the "low-albumin cytosol."
The protein content of S-9, cytosol, and low-albumin cytosol
was determined according to a method based on Bradford's Coomassie
Brilliant Blue G250 dye-binding assay (Bradford, 1976) (BioRad
Protein Assay Kit with a BioRad bovine plasma albumin standard) .
The albumin content of S-9, cytosol, and low-albumin cytosol was
measured with a colorimetric method based on the formation of an
intense blue albumin-bromocresol green complex (Doumas and Biggs,
1972). Glutathione was measured according to Che method of Ellman
(1959), which is specific for thiol groups. Reduced glutathione
(Sigma) and bovine serum albumin (Metrix) were obtained from
commercial sources. Positive controls in the Ames test were
2-nitrofluorene and 2-aminofluorene (Aldrich).
RESULTS AND DISCUSSION
The addition of S-9 to the incubation mixture of the Ames
bioassay reduced the mutagenic activity of the GM defined diesel
exhaust sample (see Table 1). The decrease in mutagenic activity
was proportional to the amount of S-9 added to the incubation
mixture (see Table 2). As shown in Table 3, the mutagenicity-
reduction effect was still observed with heat treatment of the S-9
mix or without the NADPH-generating system. The mutagenicity-
reduction effect disappeared when the heated S-9 mix was filtered
to remove coagulated materials. The above results indicate Chat
enzymatic accivity was not responsible for the mutagenicity-
reducing ability of S-9. It is evident thac proceins, the
principal denatured materials in heaC-Created S-9, produced Che
mutagenicity-reduction effect.
-------
362
YI Y. WANG AND EDDIE T. WEI
Table 1. Mutagenic Activity of the GM Defined Diesel Exhaust
Particulate Sample in the Presence or Absence of S-9 Mix3
Particulates (mg/plate)
TA98 Mean Net
Revert ant s/Pi ate*3
Without S-9 Mix
With S-9 Mixc
Spontaneous
34
43
0.125
584
446
0.25
890
500
0.5
1138
888
1
1567
1090
aFor details on the collection of the sample, see Schreck et al.,
1978.
^The mean number of spontaneous revertants was subtracted.
cThe concentration of protein in the S-9 mix used was 2.9 mg/plate.
Table 2. Relationship 3etveen the Amount of S-9 Added in the
Assay Mixture and the Number of Revertants Induced by the
Diesel Exhaust Mutagens3

Spontaneous
S-9
TA98 Mean Net
Prote in
Revertant s/
( u1/0.5 ml S-9 Mix)
Revert ants/pi ate
(rag/plate)
Plate
0
1385
0
23
5
1175
0.3
25
10
1052
0.6
25
20
882
1.2
31
50
770
2.9
33
100
702
5.8
34
150
649
8.7
35
a0ne half milligram GM defined diesel particulates was used in each
plate. The protein concentration in the S-9 was 58 rng/ml. One
half milliliter S-9 mix was applied on each plate.
-------
ABILITY OF LIVER HOMOGENATES TO REDUCE MUTAGENICITY 363
Table 3. Effects of S-9 Mix Subjected to Various Treatments on
the GM Defined Diesel Exhaust Mutagens3
TA98 Mean Net Revertants/Plate

With Boiled
S-9 Mix
Particulates
Wi thout
With Boiled
and Filtered
Without
(rag/plate)
S-9 Mix
S-9 Mix
S-9 Mix
Cofactors
Spontaneous
29
40
44
44
0.0625
185
126
231
15
0.125
380
197
330
103
0 .25
641
392
697
395
0.5
1075
855
1130
854
NADPH-generacing cofactor system. HeaC-denatured materials,
mainly coagulated proteins, were removed from the heat-treated S-9
mix by filtration. The filtrate was also applied on the assay.
The protein content of the S-9 mix was 2.9 rag/plate.
Albumin is the principal protein formed in the liver
(Rothschild et al., 1972). When albumin is removed from the
cytosol fraction by gel filtration, the detoxifying activity is
reduced by nearly 90% (see Table 4). Exogenous bovine serum
albumin, when added to the Ames bioassay, simulates the
rautagenicity-reducing activity of S-9 on diesel exhaust mutagens
(see Table 5). These results provide strong evidence that albumin
in S-9 or from exogenous sources acts as a nonspecific
mutagenicity-reduction agent for diesel exhaust mutagens.
Albumin contains thiol groups that may interact with
electrophilic mutagens (Peters and Reed, 1978; Ross, 1962). As
glutathione is present in S-9, we investigated the activity of
glutathione on the diesel exhaust mutagens. Glutathione and other
sulfhyaryl compounds have been shown to reduce the mutagenicity of
certain mutagens (Rosin and Stich, 1979; Hollstein et al., 1978;
Srinivasan and Fugimori, 1979). Table 6 shows that glutathione
significantly reduced the mutagenic activity of diesel exhaust at
doses of 25 umol/plate or more, but was ineffective at 0.8 pmol
/plate. The amount of endogenous glutathione in S-9, 0.05 ymol
/plate (in 50 yl of S-9), was too small to account for the overall
mutagenicity-reduction activity of S-9.
-------
364
YI Y. WANG AND EDDIE T. WEI
TabLe 4. The Effects of S-9 Mix, Cytosol, and Low-Albumin
Cytosol on the Mutagenic Activity of the GM Defined
Diesel Exhaust Particulate Sample3
TA98 Mean Net Revertants/Plate

Wi thout
Wi th
Wi th
With Low-Albumin

S-9 Mix
S-9 Mixb
Cytosolc
Cytosol
Particulates

(mg/plate):

Spontaneous
27
42
33
30
0.125
461
362
153
480
0.25
781
656
309
744
0.5
1197
945
760
1061
Protein

(rag/plate)
0
1.9
7.5
3.7
Albumin

(mg/plate)
0
1.4
5.3
0.3
aOne half milliliter S-9 aiix, containing 50 ul S-9, 0.5 ral cytosol,
or 0.5 ml low-albumin cytosol, was applied on each place. The
amount of tocal protein, and more specifically, the amount of
albumin used on each plate are given in table.
^S-9 was prepared from perfused livers of uninduced male Sprague-
Dawley rats.
cCytosol fraction was obtained from the S-9 according to the method
of Frantz and Mailing (1975).
These results clearly show that the ability of liver
homogenates to reduce the mutagenic activity of diesel exhaust
samples is due not to enzymatic activity but to an interaction
between mutagens and liver albumin. The nature of the interaction
is not known. It may be adsorption of mutagens onto proteins by
van der Waal forces, a process frequently observed in the binding
of drugs to protein. Binding of mutagens to protein may reduce the
effective dose of mutagens to bacterial DNA, so that the observed
mutagenicity is decreased. Another possible explanation of the
decrease would be that protein might block the transport of the
mutagens to the DNA.
-------
ABILITY OF LIVER HOMOGENATES TO REDUCE MUTAGENICITY
365
Table 5. Mutagenic Activity of the GM Defined Diesel Exhaust
Particulate Sample in the Absence or Presence of Various
Amounts of Bovine Serum Albumin
TA98 Mean Net Revertants/Plate
Without With Bovine Serum Albumin
Bovine
Particulates Serum
(rug/Plate) Albumin 2.5 mg/plate 12.5 mg/plate 20 rag/plate
Spontaneous
24
23 27 20
0. 125
318
207 198 148
0 .25
579
362 255 170
0.5
1024
736 534 392
1
1278
1146 859 569
Table 6. Mutagenic Activit
y of the GM Defined Diesel Exhaust
Sample in the
Absence or Presence of Exogenous Glutathione

TA98
Mean Net Revertants/Plate

With Glutathione
Part iculates
Wi t'nout

(rag/place)
GlutaChione
0.8 ymol/plate 25 umol/plate
Spontaneous
21
20 20
0 .125
456
435 262
0.25
712
673 472
0.375
907
835 670
0.5
1062
1055 836
-------
366
YI Y. WANG AND EDDIE T. WEI
The chemical identification of the direct-acting mutagens is
an important objective in safety evaluation of diesel exhausts.
When this objective is obtained, the molecular interactions between
the mutagens, glutathione, and albumin may be revealed.
ACKNOWLEDGMENTS
This research has been supported by the Northern California
Occupational Health Center. Y. Y. Wang is the recipient of the
Grossman Endowment for the School of Public Health, University of
California, Berkeley.
REFERENCES
Ames, B.N., and J. McCann. 1976. Carcinogens are mutagens: a
simple test system. In: Screening Tests in Chemical
Carcinogenesis. R. Montesano, H. Bartsch, and L. Tomatis,
eds. IARC Publications V. 12, IARC, Lyon, France. pp.
493-504 .
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian microsome mutagenicity cesc. Mutation Res.
31:347-364.
Bradford, M. 1976. A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal. Biochem. 72:248-254.
Clark, C.K., and I.T. Salmeen. 1980. Influence of sampling filter
media on the mutagenicity of diesel soot. In: Lovelace
Institute's 1978-1979 Annual Report to the Department of
Energy. Lovelace Inhalation Toxicology Research Institute:
Albuquerue, NM. pp. 211-216.
Cuatrecasas, P. 1970. Protein purification by affinity
chromatography. J. 3iol. Chera. 245:3059-3065.
deFlora, S. 1978. Metabolic deactivation of mutagens in the
Salmonel1a-microsome test. Nature 271:455-456.
Doumas, 3.T., and H.G. Biggs. 1972. Determination of serum
albumin. In: Standard Methods of Clinical Chemistry, Vol. 7.
G.R. Cooper, ed. Academic Press: New York. pp. 175-188.
Ellraan, G.L. 1959. Tissue sulfhydryl groups. Arch. Biochem.
B iophys. 82:70-77 .
-------
ABILITY OF LIVER HOMOGENATES TO REDUCE MUTAGENICITY
367
FranCz, C.N., and H.V. Mailing. L 9 7 5 . The quantitative microsomal
mutagenesis assay method. Mutation Res. 31:165-380.
Hollstein, M. , R. Talcoct, and E. Wei. 1978. Quinoline:
conversion to a mutagen by human and rodent liver. J. Natl.
Cancer Inst. 60:405-410.
Huisingh, J., R. Bradow, R. Jungers, L. Claxton, R. Zweidinger, S.
Tejada, J. Bumgarner, F. Duffield, M. Waters, V.F. Simmon, C.
Hare, C. Rodriguez, and L. Snow. 1978. Application of
bioassay to the characterization of diesel particle emissions.
In: Application of Short-term Bioassays in the Fractionation
and Analysis of Complex Environmental Mixtures. M.D. Waters,
S.L. Nesnow, J.L. Huisingh, S.S. Sandhu, and L. Claxton, eds.
Plenum Press: New York. pp. 381-418.
Peters, T., Jr., and R.G. Reed. 1978. Serum albumin: conformation
and active sites. In: Albumin, Structure, Biosynthesis,
Function. T. Peters and I. Sjoholra, eds. Pergamon Press:
Oxford, England. pp. 11-20.
Rosin, M.P., and H.F. Stich. 1979. Assessment of the use of the
Salmonel1 a mutagenesis assay to determine the influence of
antioxidants on carcinogen-induced mutagenesis. Int. J. Cancer
23:722-727 .
Ross, W.C.J. 1962. Biological Alkylating Agents. Butterworths:
London, England.
Rothschild, M.A., M. Oratz, and S.S. Schreiber. 1972. Albumin
synthesis. New Eng. J. Med. 286:748-757.
Santodonato, J., D. Basu, and P. Howard. 1978. Health Effects
Associated with Diesel Exhaust Emissions: Literature Review
and Evaluation. EPA-600/1-78-063. U.S. Environmental
Protection Agency: Research Triangle Park, NC. p. 9.
Schreck, R.M., J.J. McGrath, S.J. Swarin, W.E. Hering, P.J.
Groblicki, and J.S. MacDonald. 1978. Characterization of
Diesel Exhaust Particulate for Mutagenic Testing. GMR-2755.
General Motors Research Laboratories: Warren, MI.
Springer, K.J., and T.M. Baines. 1977. Emission from Diesel
Version of Production Passenger Cars. SAE paper no. 770818.
Society of Automotive Engineers: Detroit, MI.
Srinivasan, B.N., and E. Fugimori. 1979. Benzo(a)pyrene-serum:
albumin/cysteine interactions: fluorescence and electron spin
resonance studies. Chem.-3iol. Interact. 28:1-15.
-------
368
Y1 Y. WANG AND EDDIE T. WEI
Talcott, R.E., and E,T. Wei. 1977. Airborne mutagens bioassayed
in Salmonella typhimurium. J. Natl. Cancer Inst. 58:449-451.
Wang, Y.Y., S.M. Rappaport, R.F. Sawyer, R.E. TalcoCC, and E.T. Wei.
1978. Direct-acting mutagens in automobile exhaust. Cancer
Lett. 5:39-47.
Wei, E.T., Y.Y. Wang, and S.M. Rappaport. 1980. Diesel emissions
and the Ames test: a commentary. J. Air Pollut, Control
Assoc. 30:267-271.
-------
Intentionally Blank Page
-------
SESSION 5
STATIONARY SOURCES
-------
BIO ASS AYS OF EFFLUENTS FROM STATIONARY SOURCES: AN
OVERVIEW
R.G. Merrill, Jr., W.W. McFee, and N.A. Jaworski
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
Major point sources of pollutants will be with us for a long
time. Ever since western civilization began to industrialize and
to require high inputs of energy from combustion, pollutants have
been emitted in increasing amounts. In 1977, stationary electric
power sources were emitting 19.3 million tons of sulfur oxides, 7
million tons of nitrogen oxides, 3.1 million tons of suspended
particles, and 0.1 million tons of volatile organic compounds into
the atmosphere of the United States (U.S. EPA, 1978). These
figures represent about 70% of the sulfur oxides and of the
nitrogen oxides emitted each year. Clearly, stationary sources of
air pollutants deserve close attention on the basis of the mass of
emissions alone. The upward trend in emissions from stationary
sources is expected to level off or decrease in some regions of the
United States after 1990 but to continue upward in others, in spite
of advances in control technologies. New systems of energy
conversion and fuel processing, for example, fluidized bed
combustion and coal gasification, will introduce new pollution
control problems. The industrial growth anticipated in synthetic
fuels will increase the need for stringent analysis of the
potential hazardous impact of pollutants in the coming years.
In the next few years, we, as a nation, are faced with
important energy-related decisions that will have long-term effects
on our quality of life. A certain trade-off between abundant,
cheap energy and a pleasant, healthy environment seems inevitable.
We should make these hard decisions based on the best information
available.
To secure this information is the task of the Environmental
Assessment Program of the U.S. Environmental Protection Agency
371
-------
372
R.G. MERRILL, JR., ET AL.
(EPA). We have learned a lot about the threats of pollution from
industrial air, water, and residual waste streams in the last few
years. However, we want to do all we can to avoid being surprised
by the cumulative effects of effluents from present processes or
the introduction of new materials by a shift in technology. The
nation's economic growth depends, in large measure, on industrial
expansion. To aid industry and regulating agencies in protecting
our environment, we must know the chemistry and the potential
effects of products, by-products, and wastes. We can and must
protect the environment, even while we allow industry to grow.
However, environmental assessment is no easy task. For
several years we have been systematically measuring, as precisely
as possible, the emissions of many industrial technologies,
especially energy-processing technologies. The major components of
the waste streams are generally well-characterized, but with the
increasing concern over trace-level contaminants, complex mixtures,
and their subtle, cumulative effects on the environment, we have
been faced with increasingly difficult assessment problems.
At a conference two years ago, Stephen Gage (1978) said, "The
emergence of the Environmental Assessment Program as a distinct and
important part of EPA's environmental research efforts over the
past several years is an excellent example of how the Agency's
efforts have turned from primarily research in reaction to known
environmental problems to include research which anticipates and
tries to avoid future environmental problems. This change of
emphasis also predated the rechartering of EPA's position toward
toxics."
Some of EPA's best efforts in environmental assessment are in
the area of coal utilization. It is obvious that the United States
is going to have to use much more coal if our economy is to grow.
Continued electricity generation is important, and part of the coal
will probably have to be converted to gases and liquids to help
meet our varied fuel needs. These processes introduce new
compounds and mixtures into the stationary source emission picture.
Some of thera may present a serious health risk. EPA's Environ-
mental Assessment Program activities have made considerable
progress identifying the hazards in the processes, but much remains
to be lone.
It is extremely difficult to obtain representative samples
from process streams that are at high temperature and pressure and
that often include highly reactive chemical species. Chemically
analyzing complex fixtures of chemicals is another great challenge,
even with today's powerful analytical chemistry techniques. Once
the emissions are chemically characterized, determination of the
relative degree of health or environmental hazard is the next
difficult step. This activity is particularly troublesome when
-------
BIOASSAYS OF EFFLUENTS FROM STATIONARY SOURCES
373
cumulative effects, such as those which might result from exposure
to carcinogenic, mutagenic, or teratogenic agents, are considered.
In this third area—determination of potential health or ecological
hazards—bioassays add an important dimension to the process of
environmental assessment.
Since 1975, EPA's Industrial Environmental Research Laboratory
(IERL), in Research Triangle Park, NC, has been conducting a series
of environmental assessment programs designed to
1) systematically characterize the physical, chemical, and
biological characteristics of all effluent streams from an
energy conversion or industrial process;
2) rank those streams according to hazard potential;
3) identify control technology programs needed to reduce the
hazard of those streams; and
4) predict the effects of those streams on the environment,
in conjunction with the health and ecological research
laboratories of EPA's Office of Research and Development.
Examples of EPA environmental assessment programs currently
underway include assessments of coal gasification, fluidized bed
combustion, stationary conventional combustion, and coal cleaning
and liquefication processes. In performing these environmental
assessments, IERL is supporting the regulatory and enforcement
offices of EPA by anticipating future control technology needs and
developing the data bases needed to support development of
s t andard s.
The phased approach to source sampling, chemical analysis, and
bioassay has been designed to provide Che data needed to evaluate
potential environmental impact (Dorsey et al., 1978). This phased
approach incorporates the three levels of sampling and analysis
shown in Figure 1. This scheme, which relies heavily on bioassays,
is successfully meeting the environmental assessment goals listed
above.
The phased approach was adopted because it offered potential
cost savings over a direct approach in which all streams would be
carefully sampled and completely analyzed in one pass. In each
level of this approach (see Figure 2), both chemical and biological
characterizations of an effluent stream are performed. The
chemical characterization provides a quantitative numerical rank
of a stream's potential hazard based on an engineering model for
potential discharge severity. The bioassay characterization
provides a direct measure of a biological response. The dual
-------
374
E.G. MERRILL, JR.. ET AL,
COMPLEX SOURCE SAMPLES
CHEMICAL
—
—
+
+
+
—
—
—
—
+
BIOLOGICAL

—
-h
—
_
—
—
—
—
+
1
LEVEL
SCREENING ALL
STREAMS FOR
POTENTIAL
HAZARD
LEVEL
VERIFICATION AND
quantitation
of hazardous
COMPONENTS IN
INDICATED STREAMS
CHEMICAL
—
—
—
_
+
BIOLOGICAL
+
+
—

LEVEL
LONG TERM
MONITORING
FOR VARIATION
IN OUTPUT AND/OR
EVALUATION OF NEW OR
ADDITIONAL CONTROL
TECHNOLOGY ON
INDICATED STREAMS
CHEMICAL
BIOLOGICAL
S POSITIVE RESPONSE
INDICATES CONTINUED
CONCERN AND POTENTIAL
FOR HAZARD
Figure 1. The phased approach to environmental assessment.
chemical and biological characterizations are designed to
complement each other.
Numerous advantages can be gained by including bioassavs in
the assessment of stationary sources of oollutants. For example,
it may be possible to use organisras the same as or similiar to
those likely to be affected by the source. In bioassays, one can
frequently use the whole sample without separation or modification,
-------
BIOASSAYS OF EFFLUENTS FROM STATIONARY SOURCES
375
REPORT
INPUT TO
IMPACT
ANA LVSIS
FIELD
SAMPLES
INORGANIC
ELEMENTAL ANALYSIS
ISPARK SOURCE MASS
AND ATOMIC ABSORPTION
SPECTROMETRY)
ORGAN!C
LIQUID CHROMATOGRAPHY
INFRARED AND
LOW RESOLUTION
MASS SPECTROMETRY
PHYSICAL
SOLIDS MORPHOLOGY
3IOASSAY
in vitro CYTOTOXICITY
BACTERIAL MUTAGENICITY
ECOLOGICAL TESTING
in vivo TOXICITY
Figure 2. Flow chart of Level 1 scheme.
thus allowing the combined synergistic or antagonistic effects to
be expressed, which might be either lost in separation or
unobserved in chemical analyses alone. We are constantly faced
with the dilemma of whether we should separate, extract, and
isolate, in an attempt to understand what is present and how it
behaves, or work with the whole sample, so that it can express the
combined properties. The use of bioassavs allows both approaches.
Another advantage is that bioassay results have some built-in
interpretation that is absent in chemical analysis. A chemical
result indicating the presence of substance X is meaninzful only
insofar as the potency of X in causing health or environmental
damage is known. In contrast, when all the minnows die in an
aquatic test, we don't necessarily know why, but it is obvious that
this material contains one or more substances likelv to be
dangerous in aquatic systems.
-------
376
R.G. MERRILL, JR., ET AL.
One disadvantage is apparent in the example above. 3ioassays
do not always satisfy our desire to know the specific components.
Another disadvantage is that of complexity and costs. It is more
difficult to set up a laboratory to do routine bioassays at low
costs. Bioassays often cannot be started on short notice and thus
may require several weeks or months to complete, creatine delavs
and higher costs.
Even though we have argued that bioassays have some built-in
interpretative value, they are also sometimes difficult to
evaluate. Some often-asked questions are: "Do the results suggest
a similar response in other organisms? Can the results be applied
to humans? Does mutagenicity relate to carcinogenicity? So what
if fruit flies die or don't reproduce?" The bioassays used in the
first phase don't provide unequivocal answers. However, thev are
effective screening devices, providing a basis for subseauent, more
definitive tests. All tests must eventually be related to biologic
effects. Bioassays, like chemical results, are easier to interpret
when supported by other bioassays or analyses.
The best situation we can hope for in environmental assessment
is to have a combination of a battery of bioassays and the most
complete chemical and physical analyses we can afford. The results
of chemistry and bioassays support each other and provide a much
more complete picture than either alone. Bioassays should
complement chemical assays in such a way that the task of assessing
potential environmental impact is possible and practical.
The presentations in this session deal with advances in the
application of short-term bioassays to complex mixtures from
stationary energy-related sources. In particular, the application
of in vitro and in vivo bioassays to fly ash, fluidized bed
combustion effluents, and coal-1iauefication material are
d iscussed.
It is important that we look not only at Dresent-day
processes, but also at technologies of the future, so that when
they are commercially applied, we are not surprised bv their
effects. It is a distorted perspective to consider only today's
problems while allowing tomorrow's to get ahead of us and oossiblv
out of hand. It should be our goal to help avoid any ootential
environmental problems that nay be associated with new combustion
processes, synthetic fuel production, or energy-conversion
processes.
Bioassays of complex mixtures have already contributed to our
understanding of processes and hazards. Engineers, chemists, and
biologists have worked together in developing new techniaues and in
interpreting the results. A series of pilot studies on selected
stationary sources have taught us a lot about the combined use of
-------
BIOASSAYS OF EFFLUENTS FROM STATIONARY SOURCES
377
chenical and biological tests in environmental assessment.
Experience in these research and development programs has allowed
refinements in the bioassavs to provide cost-efficient, reliable
assessment of environmental impact from a wide variety of source
and sample types. It should be reiterated that health, ecological,
and industrial laboratories are working together on the collection
and interpretation of results. The people involved have a right Co
be proud of their accomplishments, but we all realize the
inadequacies of our techniques and the need to improve them. The
work reported at this conference is contributing to our
margin of safety and flexibility in future energy production,
REFERENCES
Dorsey, J.A., L.D. Johnson, and R.G. Merrill. 1978. A phased
approach for characterization of multimedia discharges from
processes. In: Monitoring Toxic Substances. D. Schuetzle,
ed. ACS Symposium Series, No. 94. American Chemical Society:
Washington, DC. pp. 29-48.
Gage, S.J. 1978. Keynote address. In: Symposium Proceedings:
Process Measurements for Environmental Assessment, Atlanta,
GA. E.A. Burns, ed. EPA 600/7-78-168. U.S. Environmental
Protection Agency: Research Triangle Park, N.C. Dp. 1-3.
U.S. Environmental Protection Agency. 1978. National Air Oualitv,
Monitoring and Emissions Trends Report, 1977.
EPA-450/2-78-052. U.S. Environmental Protection Agency:
Research Triangle Park, NC.
-------
Intentionally Blank Page
-------
COAL FLY ASH AS A MODEL COMPLEX MIXTURE FOR SHORT-TERM
BIOASSAY
Gerald L. Fisher, Clarence E. Chrisp, and Floyd D. Wilson
Battelle Columbus Laboratories
Columbus, Ohio
INTRODUCTION
Combustion of coal for electric power generation has increased
markedly throughout the past decade and is expected to continue to
increase throughout the remainder of this century. Coal combustion
produces a variety of biologically active inorganic and organic
compounds. Of major concern is the release of oxides of carbon,
nitrogen, and sulfur; biologically active trace elements: siliceous
primary particulate matter; and organic compounds. Most of the
primary particulate matter produced during coal combustion is coal
fly ash. Interaction of fly ash particles with inorganic and
organic compounds formed during the combustion process produces a
unique complex mixture. This mixture may serve as a model for
other mixtures resulting from interaction of relatively inert
carrier particles with biologically active metals and organic
compounds; the bulk of atmospheric particulate matter may serve as
a matrix for subsequent interaction of airborne vapors and
condensable gases. In this report, we review our studies that
indicate the complexity of the physical and chemical properties of
coal fly ash and attempt to relate such properties to the
biological activity of coal fly ash.
SAMPLE COLLECTION
The fly ash samples described below were collected from a
western U.S. power plant burning low-sulfur (0.5%), high-ash (20%)
coal (McFarland et al., 1977a). Unless otherwise indicated, the
samples were collected and size-fractioned in situ at 95°C
downstream of the plant's electrostatic precipitator (ESP). On
379
-------
380
GERALD L. FISHER ET AL.
occasion, ESP-collected ash was also studied. Four fractions of
stack-collected material with volume median diameters (VMD) of 20,
6.3, 3.2, and 2.2 pm and geometric standard deviations of
approximately 1.8 were obtained. Material was size-classified
using a specially thermostatically controlled (95°C) system
containing two cyclones in series followed by a centripeter with 25
parallel jets. Cyclone-separated material was deposited in a
collection hopper, while centripeter particles were collected on
fabric filters and removed for hopper deposition by cleaning with
reverse air jets operating at one-minute intervals. Thus, in
contrast to standard filter collection techniques, collected fly
ash sanples were not continually exposed to the reactive gases in
the flue stream. Such an exposure may lead to changes in the
chemical composition and biological activity of collected
polynuclear aromatic hydrocarbons (Chrisp and Fisher, in press).
Our samples were collected continuously over a 30—day period.
Thus, variations in coal composition, combustion conditions, and
other parameters of plant operation that may affect chemical
composition should be reflected in these samples. In this regard,
Kubitschek and Kirchner (1980) have demonstrated the iraporrant
effects on mutagenic activity of combustion conditions during
start-up and shut-down of a bench-model fluidized-bed combustion
s y s t em.
PHYSICAL AND CHEMICAL CHARACTERIZATION
Microscopic Studies
In previous studies (Fisher, 1979: Fisher et al., 1976:
1978b), we have described the physical and morphological properties
of coal fly ash and have generally found it to be an extremely
heterogeneous, complex mixture with a variety of morphological
forms. Morphology generally depends both on matrix composition and
on exposure conditions during combustion. Upon heating,
aluminosilicate inclusions in the coal initially become rounded and
then through degassing, become vesicular (Fisher, 1979). Further
heating results in the formation of solid spheres, hollow spheres
(cenosphere), or sphere-within-sphere structures (plerosphere) .
Crystals form somewhat later in morphogenesis, with internal
(quench) crystals forming rapidly during the transition from liquid
to solid phase (Fisher et al., 1976). Such quench crystals have
been identified as mullite by Gibbon (1979). While internal
crystal formation occurs in milliseconds, surface crystal formation
appears to be a much slower process, talcing days or months.
Surface crystals apparently form through sulfuric acid interaction
with metals found on fly ash surfaces.
-------
COAL FLY ASH AS MODEL MIXTURE FOR SHORT-TERM BIOASSAY 381
Previous analysis of individual fly ash particles demonsCrat ad
extreme matrix heterogeneity among morphologically similar
particles (Pawley and Fisher, 1977). Mora recently, we have
completed a detailed comparison of 1ight-microscopic morphology
and individual-particle elemental composition using scanning
electron microscopic (SEM) X-ray analysis. Comparative light and
electron microscopy is a powerful tool for assessing physical and
chemical properties of coal fly ash (Fisher et al. , in press). We
have demonstrated the presence of relatively pure mineral phases,
although the bulk of the ash is amorphous and appears to have the
composition of the clay minerals generally associated with coal.
Pure quartz, calcium phosphate, titanium dioxide, calcium oxide,
iron oxide, and alumina have been observed (T.L. Hayes, E.E. Lai,
B.A. Prentice, and G.L. Fisher; University of California at
Berkeley; unpublished data; 1979). The degree of pigmentation of
spherical particles depends on their iron content. Statistical
cluster analysis has confirmed the usefulness of the light-
microscopic morphological classifications (Fisher et al., in
press). In particular, within each of the specific morphological
classes defined by light microscopy, the individual particles
generally have very similar elemental compositions. However, the
most morphologically amorphous particles tend to be most diverse in
elemental composition.
We have also recently completed X-ray diffraction analysis of
the crystalline phases within coal fly ash (Hansen et al., MS).
These studies show the predominant crystalline species in coal fly
ash to be quartz, mullite, and magnetic iron oxides. The highest
concentration of crystalline material is found in the coarsest fly
ash fractions; we found quartz concentrations of 4%, mullite 8%,
and iron oxide 0.6%. The crystalline mineral concentrations
decrease with decreasing particle size; in the finest fly ash
fraction, quartz was 1.3%, mullite 47*, and magnetic oxides 0.03%.
We hypothesize that quartz intrusions within the coal itself
produce the silica, whereas mullite and magnetite are formed during
the combustion and cooling processes. Mullite is usually
associated with a quench crystal phase that occurs during rapid
cooling and hence is generally encapsulated within the
aluminosilicate matrix of the clay mineral. This does not appear
to be the case for quartz or magnetic iron oxides.
As described in our earlier work, surface crystal may form
(Fisher et al. , 1976; 1978b) as the result of chemical interaction
or formation of sulfuric acid on fly ash surfaces, possibly through
leaching of minerals and heavy metals from within the fly ash by
the surface-associated sulfuric acid. Electron microprobe analysis
of the larger crystals in fly ash has identified only calcium and
sulfur, thus the majority of crystals appear to be calcium sulfite,
present as either gypsum or anhydrite. We postulate, however, that
sulfate crystal formation may also increase the biological
-------
382
GERALD L. FISHER ET AL.
availability or refractory metal oxides through conversion to the
more soluble metal sulfates.
We found (as have many other investigators; see Coles et al . ,
1979; Ondov et al., 1977) chat the concentration of the volatile
(at coal-combustion temperatures) trace elements or their oxides
are highest in the finest fly ash particles. The elements most
enriched in the finest fly ash particles are cadmium, zinc,
selenium, arsenic, antimony, tungsten, molybdenum, gallium, lead,
vanadium, fluorine, and sulfur. However, the relatively refractory
elements uranium, chromium, barium, copper, beryllium, and
manganese also are enhanced in the finest fly ash fractions. While
it appears that vapor-phase condensation enhances volatile trace
elements in fine fly ash particles (Natusch et al., 1974), other
mechanisms also may be important in this phenomenon. Filtration
studies with neutron-activated coal fly ash indicate that some
elements are highly enriched in particles of the size range from
0.2 to 0.4 yra VMD (Fisher et al., 1979). In particular, the
elements antimony, arsenic, tungsten, uranium, and chromium have a
disproportionately high concentration in particles less than
0.4 ym. It has been hypothesized that homogeneous nucleation and
subsequent coagulation of primary particles or the condensation of
reaction products enhance submicron particles. Relatively high
concentrations of biologically active trace elements in submicron
particles may present special technological problems in particulate
abat ement.
We have compared the distribution of elements in fly ash
associated with the alurainosilicate matrix with that of elements
either in separate mineral phases or associated with the
particulate surface (Hansen and Fisher, 1980). Preferential
dissolution with either hydrochloric or hydrofluoric acid indicated
that more than 70% of the titanium, sodium, potassium, magnesium,
hafnium, thorium, and iron is associated with the aluminosilicate
matrix. On the other hand, more than 70% of the volatile elements
arsenic, selenium, molybdenum, zinc, cadmium, tungsten, vanadium,
uranium, and antimony are associated with the particulate surface.
It also appears that the larger portion of the calcium, scandium,
strontium, lanthanum, and the rare earth elements is associated
with a separate mineral phase, possibly an apatite phase, which has
a particle size distribution similar to that of the aluminosilicate
phase. These findings, then, allow computation of the probable
biological availability of trace elements in coal fly ash. In
agreement with these observations, we have found relatively high
solubilities for molybdenum, calcium, selenium, barium, arsenic,
tungsten, zinc, and antimony in coal fly ash treated with a
tris-hydrochloric acid buffer at pH 7.4 (Fisher et al., 1979b).
-------
COAL FLY ASH AS MODEL MIXTURE FOR SHORT-TERM BIOASSAY
383
IMMUNOTOXICOLOGY STUDIES
Although many sensitive in vitro bioassays exist for
mutagenesis, few are available for screening environmental
toxicants for their potential effects on host cellular defense
factors. The development of assays for monitoring immunity effects
is hampered by the complexity of the cellular and humoral factors
involved in the host reaction to neoplasia. We have begun to
develop a variety of methods for the study of environmental factors
affecting cellular immunities. A major effort has been made toward
developing assays that reflect inhibition of macrophage function.
In Vitro Studies
Using SEM X-ray analysis to investigate particles contained in
macrophages, we find that the matrix composition of particles
contained within phagocytes may vary dramatically (Hayes et al.,
1978; Pawley and Fisher, 1977). Furthermore, we hypothesize that
variation in the chemical composition of phagocytized particles may
also represent variation in the toxicological potential of such
particles (Hayes et al. , 1978; 1980). We have proposed a model
for the exposure of individual lung cells to the foreign elements
in fly ash. Segregation of elements in specific particles of fly
ash allows much higher exposure levels within individual cells than
are predicted by a model based on uniform distribution of elemental
concentrations among all particles. Furthermore, we have
demonstrated that comparative microscopic techniques can be used to
correlate the viability of individual cells with the elemental
composition of phagocytized particles. Light-microscopic analysis
of macrophages deposited on SEM finder grids and treated with
trypan blue dye indicates the viability of the cells: subsequent
electron-microscopic analysis of the same cells indicates the
elemental composition of the phagocytized particles. Studies are
now under way to compare the toxicities of the various fly ash
coraposit ions.
To determine whether trace elements in fly ash can alter the
function or viability of macrophages, we have calculated the
elemental content of macrophages that have phagocytized a few fly
ash particles (Hayes et al. , 1980). Our calculations indicate
that the concentrations of many biologically active trace elements
generally are increased by one, two, or even three orders of
magnitude. These calculations, then, indicate that the trace
elements in fly ash potentially could produce cellular damage.
Further studies are necessary to evaluate the biological
availability of trace elements in the fly ash. We have found that
many trace elements in fly ash inhibit either lectin-induced or
mixed-lymphocyte-induced lymphocyte blastogenesis (Shifrine et al. ,
in press). Chromium, lead, vanadium, and copper appear to be
-------
384
GERALD L. FISHER ET AL.
effective inhibitors in the lymphocyte-sEimulaEion assays, at
concentrations similar Eo those found in coal fly ash.
Interestingly, both lymphocytes and macrophages appear to be
extremely sensitive to vanadium.
In vitro exposure of macrophages to particulate matter of
similar size distributions allows us to compare the toxicity of fly
ash, silica, and glass beads (Fisher et al., 1977: Whaley et al.,
1977). Silica was chosen as a positive control because it is a
well-documented macrophage toxicant. Glass beads (aluminosi1icate
particles) were chosen as a negative control because they are
apparently inert. The effects of in vivo exposure to these
particles have been studied using both rat and mouse pulmonary
macrophages (Fisher et al . , 1977). Particulate exposure was
performed at a 40:1 particle:eel 1 ratio, comparable to test
particle combinations used in the phagocytic assay (Fisher et al.,
1978a). We find that the phagocytic activity of macrophages
increases with time in incubation media (Fisher et al., 1977). The
degree and rate of enhancement of control phagocytosis depends on
the species derivation of macrophages: rat macrophages show the
effect after two hours in culture, while murine macrophages take
two days in culture. Exposure to fly ash delays the increase in
phagocytic rates. The lag can be seen after two and four hours in
culture for rat macrophages and after one and two days for murine
macrophages. Interestingly, although silica exposure did not delay
the increase in phagocytosis, the final phagocytic capability
(after seven days) was markedly below that of controls. Further
studies are necessary to determine the nature of the enhanced
phagocytosis associated with in vitro culture and to evaluate the
significance of the lag in this effect, caused by fly ash.
We have also developed techniques for evaluating the
proliferative capacity of lavaged cells from the lung (Boorraan et
al. , 1979a, b) . Clonogenic technique was adapted from previously
described techniques for quantitation of bone marrow granulocyte-
monocyte progenitors (Wilson et al., 1974). The basic culture
system involves plating lung-lavaged macrophages into a semisolid
methyl cellulose medium. Colonies tend to be smaller and slower to
grow than those of hematopoietic cells. Preliminary studies
indicate that all particle types (i.e., fly ash, silica, or glass
beads) may affect the ability of these lavaged cells to divide
(Whaley et al., 1977). Agents that tend to stimulate phagocytosis
appear to decrease the cells' proliferative capacity. Similarly,
agents that inhibit phagocytosis tend to stimulate proliferation.
These observations suggest that particle-induced phagocytic
functions may preclude differentiation and subsequent division of
progenitors.
We have also demonstrated that cloning techniques are useful
in studying the in vitro dose-response characteristics of
-------
COAL FLY ASH AS MODEL MIXTURE FOR SHORT-TERM BIOASSAY
385
lymphohematopoietic progenitors exposed to trace elements in
semisolid culture systems (Wilson et al., 1980). Because of the
enhanced concentrations of zinc and selenium in fine fly ash
particles and because of the known biological activity of these
elements, we evaluated the responses of murine spleen B-lymphocyte
progenitors and bone marrow granulocyte-monocyte progenitors to
selenite and zinc exposures. At physiological concentrations, both
elements significantly suppressed cell proliferation from splenic
B-lymphocytes, but not from granulocyte-monocyte progenitors. The
results demonstrate the feasibility of using lymphohematopoietic
cloning techniques as sensitive short-term bioassays to determine
the effects of fossil fuel combustion products on cellular pathways
involved in hemaCopoiesis and immunological processes. We are
presently studying the effects of in vivo exposure on the
progenitor cell function.
In Vivo Studies
We have also performed in vivo inhalation studies with mice
acutely exposed to fly ash and silica aerosols and with rats
chemically exposed to fly ash alone. As part of this effort, we
have developed techniques for generating well-dispersed fly ash
aerosols. We use a Wright dustfeed mechanism for fly ash
deagglomeration and aerosolization, a cyclone for separating larger
particles, and a krypton-85 discharger for reducing particulate
charge to Boltzraann equilibrium (Raabe et al., 1979). In an
attempt to improve aerosolization procedures, we compared the
efficacy of a fluidized-bed generator to that of the Wright
dustfeed system (McFarland et al., 1977b). The results of the
study indicate that aerosols produced by the fluidized-bed
generator are relatively unstable over time and that
deagglomeration is markedly less effective than with the Wright
dustfeed mechanism. Aerosols produced by the Wright dustfeeder
had smaller aerodynamic size and broader size distributions than
those produced by the fluidized bed, with or without use of the
Wright dustfeed as a feed mechanism. For these reasons, we have
continued to use the Wright dustfeed rather than the fluidized bed
for generation of stable deagglomerated aerosols of fly ash, as
well as other nonhygroscopic particles.
For acute inhalation studies, we have used exposure via the
nose only for up to two hours in small chambers. Chronic
inhalation studies have been performed in immersion chambers for
periods of up to 20 h per day for 180 days (Raabe et al. , 1979).
Particle size distributions are continuously monitored using a
light-scattering particle counter. We also obtain size-
distribution and aerodynamic data using SEM analysis of point-to-
plane ESP samples or cascade impactor samples. Total mass is
measured periodically in filter samples. For acute inhalation
-------
386
GERALD L. FISHER ET AL.
studies, we have used the finest fly ash fractions of the size-
classified stack-collected fly ash. However, because chronic
inhalation studies require relatively large masses of material, we
have employed size-classified material collected from the hopper of
the power plant's ESP.
In acute inhalation studies, using a stack-collected fly ash,
mice were sacrificed 2, 6, and 15 days after exposure, and
macrophage function, pulmonary pathology, and progenitor cell
kinetics were evaluated (Fisher and Wilson, 1980). Macrophage
functional assays indicated a depression in phagocytic capacity of
fly-ash-exposed mice (compared with controls) at 6 and 15 days
after exposure. Similar depressed phagocytic activity was observed
in silica-exposed animals. Progenitor-cell assays indicated an
initial depression in pulmonary alveolar macrophage colonies at 2
days after exposure and a marked elevation at 15 days after
exposure. In contrast to the changes observed in pulmonary
macrophage progenitors, macrophage precursors in bone marrow and
spleen were not significantly affecced. These data suggest that
the elevated progenitor cell activity is due to recruitment of
progenitors from the lung itself (i.e., local production). The
increased proliferation of progenitor cells two weeks after acute
exposure in vivo contrasts with the continued depression observed
in in vitro studies. Similarly, preliminary studies with
intraperitoneal exposure to zinc indicate that lymphohematopoietic
progenitors are less sensitive to in vivo exposure than are cells
exposed in vitro.
Further evidence to support the recruitment hypothesis is
provided by observing changes in the number of particles within
phagocytes (Fisher and Wilson, 1980). For the 2-, 6-, and 15-day
observation periods, the number of particles continually decreased
within the cells, suggesting enhanced production of phagocytic
cells in the lung. On the other hand, we have not observed similar
effects with long-term low—level chronic exposure to fly-ash
aerosols derived from the power plant's ESP. It is not clear
whether the difference in biological response reflects the
difference in concentration (200 vs. 2 mg/m^), the difference in
species (rat vs. mouse), or perhaps, most importantly, the
difference in fly ash source (stack vs. ESP hopper-collected ash).
We have also developed techniques for quantifying fly ash
deposited in rat lungs during the chronic inhalation studies
(Fisher et al . , 1980). Because it is difficult to separate
particulate matter from lung tissue, we chose to evaluate elemental
analysis as a measure of fly ash in the lung. Selection of the
appropriate element was based on the following criteria:
-------
COAL FLY ASH AS MODEL MIXTURE FOR SHORT-TERM BIOASSAY
387
1) the elemental analysis should be specific, and detection
limits should be appropriate for a sensitive indicator of
of lung burden;
2) the levels of the element in the tissue should be iow and
relatively constant;
3) the dissolution of the element should parallel the
particulate mass dissolution; and
4) the element should be uniformly distributed throughout
the size range of the fly ash under study.
Only aluminum and silicon met these criteria; aluminum was chosen
because its analysis is more sensitive and less troublesome than
that of silicon. The lung content of fly ash calculated from the
aluminum analysis was in quantitative agreement with calculations
based on available deposition and clearance data.
We have evaluated the mutagenic properties of coal fly ash
extracts using the Ames Salmonella assay system (Ames et al., 1975:
Chrisp et al., 1978; Fisher et al., 1979a). Our studies indicate
that in keeping with a model of surface deposition, the finest fly
ash fractions are indeed the most mutagenic (Fisher, 1980).
However, the 3.2-um fraction of fly ash is more mutagenic than the
2.2-um fraction. Ac first we assumed that this was due to
antimutagens in the fly ash, and we evaluated the possible role of
selenium, as the selenite, or fluorine, as the fluoride. These
elements were chosen because of their relatively higher
concentrations in the finest fly ash fraction than in the 3.2-iim
fraction, and because selenium, as the selenite, is an antimutagen
for acetylaminofluorene and its derivatives. Fluoride is generally
recognized as an enzyme inhibitor. Adding these elements to
extracts of the 3.2-um fraction, however, did not alter its
mutagenic activity. Thus, we do not have experimental support for
the hypothesis that antimutagens are present in the finest fly ash
frac t ion.
Natusch (Colorado State University, personal communication,
1980) suggests that the difference in mutagenic activity may be due
to differences in chemical absorption of mutagens by the two fly
ash fractions. Further evidence for the chemisorption of mutagens
on fly ash surfaces is the photostability of these compounds. We
have irradiated fly ash samples with ultraviolet light, sunlight,
and X-irradiation, with no decrease in mutagenic activity.
However, heating the fly ash to temperatures of 200 to 250°C
results in the loss of approximately half of the mutagenic
activity, while heating above 300°C results in complete loss of
detectable mutagenic activity (Fisher et al., 1979a). The biphasic
nature of the loss in mutagenic activity with heating indicates the
-------
388
GERALD L. FISHER ET AL.
presence of at least two mutagens or classes of mutagens. Most
recently, we have demonstrated Chat the loss of mutagenicity is due
to decomposition of surface-associated materials, as opposed to
volatilization (Hansen et al., MS), further supporting the
suggestion that mutagens are on fly ash surfaces.
Fly ash collected by the power plant's ESP does not appear to
be mutagenic in the Ames test (Fisher et al., 1979a). Furthermore,
even when classified by size and sampled in a size distribution
equivalent to that of our finest stack-collected fraction, 5SP-
collected materials are still not mutagenic. Our studies indicate
that temperatures of approximately 100'C may be critical for
absorption of mutagens on fly ash surfaces. Indeed, the
calculations of Natusch and Tomkins (1978) indicate that the
deposition of vapor-phase polynuclear aromatic hydrocarbons on fly
ash particles is extremely sensitive to temperature changes around
100'C.
We have compared the efficiencies of mutagen extraction by a
variety of solvents. Normal saline is a very poor extractant of
mutagenic activity, whereas horse serum and serum from other
species are fairly efficient extractants of fly ash mutagens
(Chrisp et al. , 1978). We found that a mutagen-seruin protein
complex forms that can be isolated from the serum, although it is
available to the bacterial genome. Further studies indicated that
albumen alone is nearly as efficient as the total serum in
extracting mutagens from fly ash (C.E. Chrisp and G.L. Fisher,
unpublished data). Direct extraction of fly ash with
dimethylsulfoxide and sonication results in the highest detectable
levels of mutagens.
We have not identified the chemical composition of the fly ash
mutagens, although recent studies using acidic, neutral, and basic
aqueous fly ash extracts further indicate that mutagens in fly ash
are not inorganic. Acidic aqueous fractions were not mutagenic,
whereas basic aqueous fractions contained approximately half of the
mutagens extractable with dimethylsulfoxide. These results
indicated that a significant portion of the mutagenic activity in
coal fly ash could be accounted for by the presence of weak organic
acids (Hansen et al . , 1980).
To evaluate the carcinogenic potential of coal fly ash, we
have modified the tracheal implant system described by Griesemer et
al. (1974). For these studies, we have packaged fly ash in 0,2-um
Nuclepore filters. This method allows for the slow release of
mutagens to the sensitive tracheal epithelial cells and minimizes
the risks of local tissue damage and toxicity. We have used the
Araes bacterial mutagenesis system to monitor the release of
mutagens from fly ash in tracheal implants (Chrisp and Fisher,
-------
COAL FLY ASH AS MODEL MIXTURE FOR SHORT-TERM BIOASSAY 389
1980). Studies are now in progress Co evaluate the carcinogenic
potential of the fly ash using the tracheal implant system.
CONCLUSIONS
In conclusion, oar results demonstrate the extreme complexity
of coal fly ash in terms of matrix composition, morphological
appearance, and surface trace element and organic chemical
composition. Assays are being developed to measure the potential
immunotoxicity of fly ash. Acute inhalation studies have
demonstrated that coal fly ash may be as toxic as quartz to the
pulmonary alveolar macrophage. Further in vivo studies comparing
the cytotoxicity of fly ash and «-quartz are required to
substantiate this hypothesis. The feasibility of applying
sophisticated cloning techniques to the evaluation of potential
lymphohematopoetic effects from complex mixtures has been
demonstrated. Mutagens in coal fly ash appear to be absorbed to
fly ash surfaces and hence may exist in the environment for
relatively long periods of time. Techniques are now being
developed to evaluate the carcinogenic potential of coal fly ash
through a combination of bacterial mutagenesis assays and tracheal
implant carcinogenesis assays. Detailed chemical, morphological,
and toxicological analyses indicate the usefulness of coal fly ash
as a model comolex mixture.
ACKNOWLEDGMENTS
This work was supported by the U.S. Department of Energy
through the Laboratory for Energy-Related Health Research,
University of California, Davis, CA. This manuscript was also
presented at the Second Symposium on Process Measurements for
Environmental Assessment, February 1980, Atlanta, GA.
REFERENCES
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian-microsome mutagenicity test. Mutation Res.
31:347-360.
Boonnan, G.A., L.W. Schwartz, and F.D. Wilson. 1979a. Formation
of macrophage colonies in vitro by free lung cells obtained
from rats. J. Reticuloendothel. Soc. 26:855-866.
Boonnan, G.A., L.W. Schwartz, and F.D. Wilson. 1979b. In vitro
macrophage colony formation by free lung cells during
pulmonary injury. J. Reticuloendothel. Soc. 26:867-872.
-------
390
GERALD L. FISHER ET AL.
Chrisp, C.E., and G.L. Fisher. 1980. Mutagenesis of coal fly ash
linked to tracheal graft assay for carcinogenesis. Presented
at the 11th Annual Meeting of the Environmental Mutagenesis
Soc iet y.
Chrisp, C.E., and G.L. Fisher. (in press). Mutagenicity of
airborne particles. Mutation Res.
Chrisp, C.E., G.L. Fisher, and J. Lammert. 1978. Mutagenicity of
respirable coal fly ash. Science 199:73-75.
Coles, D.G., R.C. Ragaini, J.M. Ondov, G.L. Fisher, D. Silberman,
and B.A. Prentice. 1979. Chemical studies of stack fly ash
from a coal-fired power plant. Environ. Sci. Technol.
13:455-459.
Fisher, G.L. 1980. Size-related chemical and physical properties
of power plant fly ash. In: Aerosol Generation and Exposure
Facilities. K. Willeke, ed. Ann Arbor Science: Ann Arbor,
Ml. pp. 203-214.
Fisher, G.L. 1979. The morphogenesis of coal fly ash. In:
Proceedings of the Symposium on the Transfer and Utilization
of Particulate Control Technology, IV, Denver, CO.
pp. 433-440.
Fisher, G.L., D.P.Y. Chang, and M. Brummer. 1976. Fly ash
collected from electrostatic precipitators: Microcrystalline
structures and the mystery of the spheres. Science
192:553-557 .
Fisher, G.L., C.E. Chrisp, and O.G. Raabe. 1979a. Physical
factors affecting the mutagenicity of fly ash from a
coal-fired power plant. Science 205:879-881.
Fisher, G.L., K.L. McNeill, C.B. Whaley, and J. Fong. 1978a.
Attachment and phagocytosis studies with murine pulmonary
alveolar macrophages. J. Reticuloendothel. Soc. 24:243-252.
Fisher, G.L., K.L. McNeill, C.B. Whaley, and J. Fong. 1977.
Functional studies of lavaged pulmonary alveolar macrophages
exposed to particles in vitro. In: Radiobiology Laboratory
Annual Report. UCD 472-124. University of California:
Davis, CA. p. 50.
Fisher, G.L., and D.F.S. Natusch. 1979. Size-dependence of the
physical and chemical properties of fly ash. In: Analytical
Methods for Coal and Coal Products, III. C. Karr, ed.
Academic Press. pp. 489-541.
-------
COAL FLY ASH AS MODEL MIXTURE FOR SHORT-TERM BIOASSAY
391
Fisher, G.L., B.A. Prentice, T.L. Hayes,' and C.E. Lai. Cin press).
Comparative analysis of coal fly ash by light and electron
microscopy. In: Air. American Institute of Chemical
Engineers.
Fisher, G.L., B.A. Prentice, D. Silberman, J.M. Ondov, A.H.
Bierraann, R.C. Ragaini, and A.R. McFarland. 1978b. Physical
and morphological studies of size-classified coal fly ash.
Environ. Sci. Technol. 12:447-451.
Fisher, G.L., D. Silberman, B.A. Prentice, R.E. Heft, and J.M.
Ondov. 1979b. Filtration studies with neutron-activated coal
fly ash. Environ. Sci. Technol. 13:689-693.
Fisher, G.L., D. Silberman, and O.G. Raabe. (1980). Chemical
characterization of coal fly ash in rodent inhalation studies.
Environ. Res. 22:298-306.
Fisher, G.L., and F.D. Wilson. 1980. The effects of coal fly ash
and silica inhalation on macrophage function and progenitors.
J. Reticuloendothe1. Soc. 27:513-524.
Gibbon, D.L. 1979. Microcharacterization of fly ash and
analogs: the role of SEM and TEM. In: Scanning Electron
Microscopy, Vol. 1. 0. Johari, ed. Scanning Electron
Microscopy, Inc.: Chicago. pp. 501-510.
Griesemer, R.A., J. Kendrick, and P. Nettesheim. 1974. Tracheal
grafts. In: Experimental Lung Cancer. E. Karbe and J.F.
Park, eds. Springer-Verlag: Berlin. pp. 539-547.
Hansen, L.D., and G.L. Fisher. 1980. Elemental distribution
in coal fly ash particles. Environ. Sci. Technol.
14:1111-1117.
Hansen, L.D., G.L. Fisher, C.E. Chrisp, and D.J. Eatoug'n. 1980.
Chemical properties of bacterial mutagens in stack-collected
coal fly ash. Presented at the Fifth International Symposium
on Polynuclear Aromatic Hydrocarbons, Columbus, OH.
Hansen, L.D., D. Silberman, and G.L. Fisher. (MS). Crystalline
components of stack-collected, size-fractionated coal fly ash.
Hayes, T.L., J.B. Pawley, and G.L. Fisher. 1978. The effect of
chemical variability of individual fly ash particles on cell
exposure. In: Scanning Electron Microscopy. Scanning
Electron Microscopy, Inc., AMF: O'Hare, IL. pp. 239-244.
-------
392
GERALD L. FISHER ET AL.
Hayes, T.L., J.B. Pawley, G.L. Fisher, and M. Goldman. 1980.
A toxicologies! model of fly ash exposure to lung cells.
Environ. Res. 22:499-509.
Kubitschek, H.E., and F.R. Kirchner. 1980. Biological monitoring
of fluidized bed combustion effluents. Presented at the
U.S. Environmental Protection Agency Second Symposium on the
Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, Williamsburg, VA.
McFarland, A.R., R.W. Bertch, G.L. Fisher, and B.A. Prentice.
1977a. A fractionator for size-classification of aerosolized
solid particulate matter. Environ. Sci. Technol. 11:781-784.
McFarland, A.R., C. Ortiz, G.L. Fisher, and O.G. Raabe. 1977b.
Aerosolization of bulk, powders utilizing fluidized bed and
Wright dust feed techniques. In: Radiobiology Laboratory
Annual Report. UCD 472-124. University of California:
Davis, CA. pp. 10-13.
Natusch, D.F.S., and B.A. Torakins. 1978. Theoretical
consideration of the adsorption of polynuclear aromatic
hydrocarbon vapor onto fly ash in a coal-fired power plant.
In: Carcinogenesis 3. P.W. Jones and R.I. Freudent'nal , eds.
Raven Press: New York. pp. 145-154.
Natusch, D.F.S., J.R. Wallace, and C.A. Evans. 1974. Toxic trace
elements: preferential concentration in respirable particles.
Science 183:202-204.
Ondov, J.M., R.C. Ragaini, R.E. Heft, G.L. Fisher, D. Silberman,
and B.A. Prentice. 1977. Interlaboratory comparison of
neutron activation and atomic absorption analyses of
size-classified stack fly ash. In: Proceedings of the Eighth
Symposium on Methods and Standards for Environmental
Measurement Materials Research. National Bureau of Standards:
Gaithersburg, MD. pp. 565-572.
Pawley, J.B., and G.L. Fisher. 1977. Using simultaneous three
color x-ray mapping and digital-scan-stop for rapid elemental
characterization of coal combustion by-products. J. Microsc.
110:87-101.
Raabe, O.G., K.D. McFarland, and B.K. Tarkington. 1979.
Generation of respirable aerosols of powerplant fly ash for
inhalation studies with experimental animals. Environ. Sci.
Technol. 13:836-840.
-------
COAL FLY ASH AS MODEL MIXTURE FOR SHORT-TERM BIOASSAY
Shifrine, M., G.L. Fisher, and N.J. Taylor. (in press). Effect
trace elements in coal fly ash on lymphocyte blastogenesis.
J. Environ. Pathol. Toxicol.
Whaley, C.3., F.D. Wilson, G.L. Fisher, M. Shifrine, and K.L.
McNeill. 1977. Growth of canine alveolar macrophage
colonies. In: Radiobiology Laboratory Annual Report. UCD
472-124. University of California: Davis, CA. pp. 47-49.
Wilson, F.D., G.L. Fisher, and B.A. Concoby. 1980. Studies
on in vitro dose-response characteristics of trace elements
(Zn, Se) on lymphohematopoietic progenitors using semisolid
culture systems. In: Pulmonary Toxicology of Respirable
Particles. C.L. Sanders, F.T. Cross, G.E. Dagle, and J.A.
Mahaffey, eds. CONF-791002. U.S. Department of Energy:
Washington, DC,
Wilson, F.D., L. O'Grady, C. McNeill, and S.L. Munn. 1974. The
formation of bone marrow derived fibroblastic plaques in
vitro, preliminary results contrasting these populations to
CFU-C. Exp. Hematol. 3:353-354.
-------
Intentionally Blank Page
-------
POSSIBLE EFFECTS OF COLLECTION METHODS AND SAMPLE
PREPARATION ON LEVEL 1 HEALTH EFFECTS TESTING OF COMPLEX
MIXTURES
D.J. Brusick
Department of Genetics and Cell Biology
Litton Bionetics. Inc.
Kensington, Maryland
INTRODUCTION
The Level 1 Environmental Assessment Program of the U.S.
Environmental Protection Agency (EPA) Industrial Environmental
Research Laboratory (IERL) has two main goals: first, to detect
potentially hazardous emissions from stationary sources, and
second, to develop a data base that will permit a relative ranking
of industrial streams with respect to their potential biohazard.
The level of bioassay applied does not permit either qualitative
or quantitative risk assessment. However, since both chemical and
biological effects are evaluated, the level of data integration and
coordination is increased, reducing the chance chat a potentially
hazardous stream will go undetected.
During the initial phases of the IERL Environmental Assessment
Program, considerable time was given to methods of sample
collection, preparation, and analysis for chemistry assessment
(Lentzen et al., 1978). The recent introduction of bioassays
requires a similar appraisal of the functions of sample collection,
storage, and pretest handling as they relate to the specific health
effects and ecological tests proposed for Level 1 biological
assessment .
Because of the need to rank streams according to potential
biohazard, the initial approach taken by IERL was to evaluate
samples in the specific bioassays in a state as similar as possible
to that found at the time of sampling. However, recent results
reported in the scientific literature on analysis of complex
environmental mixtures have shown that pretest processing of
samples (concentration of liquids, extraction of particulate) often
395
-------
396
D.J. BRUSICK
results in enhanced biological activity (Klekowski and Levin, 1979:
Kub it sc'nek and Venta, 1979; LofroCh, 1978: Pitts et al . , 1977;
Teraniski et al., 1978). In this report, several of the sampling
and pretest procedures are reviewed to illustrate how the
application of these techniques to Level 1 Environmental Assessment
will affect the test responses and ultimately the goals of this
program.
METHODOLOGY
The IERL Level 1 Environmental Assessment Program evaluates
source emissions according to the following five parameters:
1) rate of release into the environment,
2) distribution between physical states,
3) chemical composition,
4) detection of potential specific health effects, and
5) detection of potential impact on the ecosystem.
A schematic of the above parameters is given in Figure 1.
The initial Level 1 methods manual for biological testing
contained protocols for five health effects tests and eight
biological tests (Duke et al . , 1977). The types of data obtained
from these tests were varied and not amenable to interpretation by
anyone other than a biologist. Clearly, a method of developing
uniform data was needed. Out of a review of Level 1 bioassays and
data from several pilot studies, a system of uniform data analysis
and formatting was recommended (Brusick, 1980). The review study
also recommended elimination or substitution of certain test
procedures. Table 1 identifies the current status of
recommendations for Level 1 bioassays. No procedures have yet
been approved for the Soil Microorganism Toxicity Assay; test
substitutions or protocol modifications are being considered for
some of the other tests as well.
The data of three pilot studies have been evaluated following
the recently proposed data-formulating procedures (Brusick, 1980).
The results were recorded as high (H), moderate (M), low (L), or
nondetectable (ND) on a summary sheet such as in Figure 2. Table 2
defines the limits of each category.
The studies on Coal Gasification, Fluidized Bed Combustion,
and Textile Plant Liquid Effluents received very little pretest
sampling or sample history evaluation. Most of the samples were
-------
EFFECTS OF COLLECTION AND PREPARATION ON TESTING
397
TYPICAL STEPS INVOLVED IN ENVIRONMENTAL ASSESSMENT
Source
Release
>
X7
Rate of Release Measured
and Sampling Performed
V
Sample Transported
to Sites of Analysis
V
Preanalysis Handling
• Concentration
• Extraction
Environmental

Chemical Analysis Biological Analysis
Coordinated
"Results
Potential for Environmental
Effects Determined
Figure 1. An overview of the seeps involved in Level I
environmental assessment.
-------
r.
=r i>|
H
o
co
—J

~ 1
.X
5
«-•

a *> —
a
&

r.
w

0
£
t.
s X

Oa
r»

f ¦ ii ¦••
c ^
rr

V •
X

w
X Ifl ^

2
5 i
9
r»
yi
n
rr
» f! ;
— . (s
lc
r.

rr
•<

.•
rt
I u - ¦
c. e

C
f.

•<
w

to 5

zr
r:

>A
¦8
p.

C
o

i— •
—
O
U

;
c.
5
2
r*

ro
1
JT

r.
S

3
u

n
w

tc
«s

X
3
C
3
5
-
&
C
0

2
rys

•—»
MM
"D

*—
"C
rr
*5
~
c.

XT
i-.
—-
w

u
—-
n
3.
*-
"<

—.
<

¦» ¦
s
**«».

F*.
<— !

s
«—•
rs
i
mm
a

»"
.

n
v-
&>

•> .
<-»
n
3
n
o

©•
n

9
•<
-J

.X
w
*
£
o
¦»

a
z

Tfci
rr
— •
3
o
n
o

n
ft

"«
z.
- •

¦- •
•<
n

<
-
(* j

X
X

ft
(A
5

1
~
1
~
¦f

*
bacteria
Mut agenes is
CHO Clonal
Toxic i:v
RAM
In Vivo
Roden:
Toxicity
F i sh
Invertebrate
A1 gae
In sec:
Toxic i t *
Soil
y.ic rocrgan i sns
Toxic '11 y
ant
Tox i: c v
Hoisnaa ra
-------
EFFECTS OF COLLECTION AND PREPARATION ON TESTING 399
9IOASSAY SUMMARY TABLE
Technical Directive or Project No.
Contract No.
Health Effects Tests
Ecological Eltects Tests
Aquatic
Fresh Waler
Notes
Sample Identification
ND m No Detectable Toxicity
L — Low Toiicity
M = Moderate Toxicity
H = High Tonicity
L3I-01«8 R 11 80
Figure 2. Example of bioassay summary sheet.
-------
Table 2. Definition of Effectiveness Categories ^
o
Range of Concentration or Dosage
Assay AcLivity Measured Units MAl)a High Moderate l.ow Not Detectable
Ames test
Hut agenes i s'}
mg/plate
5
<0.05
0.05-0.5
0.5-5
ND
at
>5

|il/pl at e
50
<0.5
0.5-5.0
5-60
ND
at
>50
RAM/Wl-38 and
Lethality EC50c
ing/ml
1
<0.01
0.01-0.1
0. 1-1
ND
at
>1
CHO toxicity

p 1 / in I
600
<6.0
6.0-60
60-600
ND
at
>600

|i I / m 1
20c

ND
at
>20
Kodent

toxicity
LeLhality LD50e
gin/ kg
10
<0.1
0.1-1
1 -10
ND
at
>10
Aquatic Tests

A1 gae
Growth inhibition
gin/1
1
>0.01
0.01-0.1
0. l-l
ND
at
>1

EC50
%
100
<20
20-75
75-100
ND
at
>100
I' i sh
Lethaliry lX50f
gin / 1
1
<0.01
0.01-0.1
0. 1-1
ND
a t
>1

X
100
<20
20-75
75-100
ND
at
>100
Invertebrate
Lethality LC50
gm/1
1
<0.01
0.01-0.1
0. 1-1
ND
at
>1

X
100
<20
20-75
75-100
ND
at
>100
aMaximum applicable dose ("technical I imi tat ions) .
^Negative response at 5 mg/plate or aL level of toxicity is given as ND: positive response requires
calculation of minimum effective concent rat ion (MF.C) to produce a positive inut agen i c response;
II, M, and T. designations are made from MEC values of positive agents.
•-Calculated concent rat ion expected to produce effect in 50% of population.
''Volumes used for solvent exchange samples (this maximum keeps DMSO below level of toxicity).
^Calculated dosage expected to kill 50% of population.
^Calculated concentration expected to kill 50% of population.
P
c_
03
SO
c
C/J
o
-------
EFFECTS OF COLLECTION AND PREPARATION ON TESTING
401
placed directly into Che bioassays: only a few filters or sorbent
collectors were extracted or sonicated with solvents. A severity
assessment was conducted both on chemicals detected and biological
responses; a comparison of the analyses indicated generally
complementary results (Sexton, 1979). In a few instances, toxicity
was not predicted by the chemical analysis. Pretest processing
could have a significant impact on such data. With the increased
use of Level 1 testing procedures, the questions of whether to use
a sample directly or to reduce it to its active constituents or to
concentrate a dilute sample become more relevant.
Important components in the final data analysis and
interpretation of Level 1 bioassays are
1) sampling methods (selecting a representative sample),
2) storage (maintaining proper composition and preventing
degradation) ,
3) shipping (same as for storage), and
4) pretest handling (altering chemical composition, physical
state, or preferential extraction and concentration of
potential toxicants and mutagens).
Sampling methods, storage, and shipping can be grouped into a
single parameter called sample history (i.e., documentation of
sample handling until receipt at the testing laboratory). Pretest
handling encompasses the pretest sample processing techniques
conducted at the testing facilities. Table 3 describes the
possible effects of each technique.
Sample History
Both the final interpretations of the test results and the
ranking of emission sources according to potential hazard can be
affected by sample history. Representative sampling is
particularly decisive.
The environmental fate of the toxic substances in emissions,
for example, profoundly influences the final assessment. An
emission with extremely low volume and release rate may represent a
negligible health or ecological hazard even if it is highly toxic.
However, if the emission were large (e.g., of fly ash from power
plants), significant human and ecosystem exposure would result
(Fisher et al., 1979; Kubitschek and Venta, 1979). If a potential
hazard is identified as a function of its release rate (volume per
unit time), toxicity, and environmental fate, a quantified
assessment may be possible. Thus, if the bioaccumulation of any
-------
Tub 11" 3. Effects of Pretest Handling Hethods
O
to
Pretest Handling Method Employer!
Particulate extract! on (a) Organic solve nt/6on lent ion
(b) Soxli I ot/organ ic solvent
Possible K. f f e c t
Preferential releases tn toxic and mut agen i c. organic
materials from bound state. If these organirs ,ir^
not released under normal environmental or
phy si ill op, i r al conditions, they ran skew ranking
scheme.
Liquid concentration
Flow through XAD-2 resin column
followed by soxhlet extraction
from resin and solvent exchange
Concentrates organics and permits inorganics to pass
through. Some of the inorganics may be important
toxicants. Again, pre Fei cut i al concentration of
organics might increase detection of mutagens and
skew ranking scheme. Concentration of chemicals may
introduce artifacts by sltering chemical: chemical
dynamics not encountered in dilute solutions.
tut rat t ion from
collect ion filter
Sonication in organic solvent
Preferential release of toxic anil mutagenic agenLs
preferentially attached to filter materiat.
Degradation and alteration ti f chemicals is known to
occur when hound to filter. Sonication uf riltei
may release .small particles that would be toxic in
RAN assay.
Solvent exchange and
solubi1izat ion
Addition of sample to methylene
chloride, acetone, benzene,
DMSO, etc.
Preferential release of soluble compounds to the
target organism. Norirepresent at i ve composition of
original sample by extraction of organics.
Particulate sizing
and grinding
(a) Filter collection
(b) Grinding and sizing with
mill and wire scteeu
Uevel opinenL of a physical state more amenable to
cellular phagocytosis or animal absorption. Tli i s
could skew ranking by producing abnormally high
levels of toxicity.
a
pa
C
(/}
O
*
-------
EFFECTS OF COLLECTION AND PREPARATION ON TESTING
403
single component is very large, the potential hazard could be
significant. However, only toxicity is being recorded at present
(see Figure 2). This factor is not being related to the type of
emission release rate or to the volume of emission sample.
Attempts should be made to document Che history of each sample.
Pretest Handling
Several laboratories in the United States (particularly IERL
of EPA in Research Triangle Park, NC, and Oak Ridge National
Laboratory in Oak Ridge, TN) have initiated programs to analyze
complex mixtures for mutagenic activity (Epler, 1979: Huisingh et
al . , 1978). The results of the Ames Salmonella assay serve as a
preliminary indicator for further fractionation tests to determine
the biologically active components (pure substances).
The Ames test procedures specify how to collect the desired
sample (particulate, gas, or liquid) on a filter or solid sorbent;
extract the organics from the filter or sorbent (by sonication or
Soxhlet methods) into an organic solvent: exchange solvents to
dimethylsulfoxide (DMSO) or evaporate to dryness and resuspend in
DMSO; and conduct the bioassay on the concentrated extract .
Chemical fractionation and further testing may eventually lead to
specific associations between chemicals or chemical classes with
biological activity (Huisingh et al., L978). Table 3 describes
some of the pretest sample processing techniques currently used and
indicates their possible effect on the interpretation of Level 1
bioassay data.
The Level 1 toxicity assessments listed in the Bioassay
Summary (Figure 2) include several types of biological systems and
phylogenetic levels. If a significant amount of pretest processing
is anticipated, all the Level 1 tests, and not just the Ames or one
or two selected tests, should be evaluated. Otherwise, the data
balance will be upset and the ranking of the test site might be
biased by an abnormally toxic response from an extract or
concentrate in a single assay, erroneously exaggerating the
potential hazard. This precaution is not meant to preclude pretest
processing. However, as with the sample history factor (release
rate), the pretest sample processing must be factored into the
final assignment of a toxicity value and its contribution to the
potential hazard. A mechanism needs to be developed to normalize
the data obtained from preprocessed samples. The simplest approach
might be to divide the actual test response by a concentration
fac tor.
-------
404
D.J. BRUSICK
RECOMMENDATIONS
The following recommendations would facilitate the
implementation of sample history documentation and pretest sample
processing and enhance Che reliability of the Level 1 bioassays:
1) Documentation of sample collection, storage, shipping, and
pretest processing should be available for all test
samples. Examples of forms for this purpose are shown in
Figures 3 and 4.
2) Pretest processing should be factored into the final
toxicity designations of H, M, L, and ND. Specific
methods need to be developed to normalize data obtained
from samples modified prior to evaluation.
3) Emissions from a given site should be applied uniformly to
the spectrum of Level 1 bioassays so as not to bias the
final interpretation. Data related to environmental fate
should be included.
4) Discharge-severity calculations used to factor chemical
and physical information should be expanded to include
biological response and fate. Specific methods need to be
developed.
CONCLUSIONS
Level 1 environmental assessment bioassays should permit
accurate ranking of emissions from stationary-site sources with
respect to their potential hazard. Pretest processing should be
kept to a minimum and applied uniformly across all Level 1
bioassays. The ranking must ensure that the potential hazard will
be derived from emissions in the state in which they were released
into the environment.
The Level 1 Environmental Assessment Program should include
the environmental fate and emission release rate along with the
chemical and bioassay toxicity determinations. This approach would
result in a second level determination called a severity potential
hazard. A general scheme has been proposed to develop this
assessment and is given in Figure 5. If a potential hazard can be
calculated with reasonable accuracy, the usefulness of Level 1
results will be greatly enhanced.
-------
EFFECTS OF COLLECTION" AND PREPARATION ON TESTING
405
Sample Collection .-orm No.
LEVEL 1 SAMPLE COLLECTION FORM
I. SAMPLE HISTORY
A.
Company Name
B .
Sampling Manager
C.
Contract No.
D .
Sampling Date
E.
Emission Source
F.
Approximate Rate of Emission (Volume/Time)
G.
Proportion of Emission Sampled (Volume Captured)
II. SAMPLE TYPE
A. Sample No.
3. Name
C. Sample Description
III. HANDLING CONDITIONS
Storage
Container
~ Amber Glass Boctle
~ Polyethylene 3ottle
~ Coated Bag or Bottle
Temoerature
~ Ambient
~ Refrigerate (C to 4aC)
~ Freeze (-20"C)
3. Approximate Time of Storage Before Shipping
IV. SHIPPING HISTORY
A. Biological Contractor
B. Address:
Light
~ Keeo in 2ark
C. Carrier
D. Date Shipped 3y
E. Special Packaging
Two copies of this form must accompany each sample.
Figure 3. Example of Level 1 sample collection form.
-------
406
D.J. BRUSICK
Sample Processing Form No.
LEVEL 1 SAMPLE PROCESSING FORM
SAMPLE IDENTIFICATION
A. Sample Collection "ora No.
B. Contract No.
C. Project Officer
D. Sample Mo.
II. SAMPLE TYPE AND PROCESSING REQUIREMENTS
Basic Type

Subtype

C Solid
Q
Solid Granular
~

~
Slurry (>50% Solids)
~

~
Particulates from

Filter
a

~
Filter/Unit Particulates

~
~ Liquids
~
Suspensions (<50% Solids)
~

Q
Effluent
~

a
Leachate
~

~
Extract

a
Condensate

~ Gas
~
Pressure Collection

a
Vacuum Collection

Processing
Organic Solvent
Remove Particula
from Filter
III. BIOASSAYS REQUESTED
~ Ames, Salmonella
C RAM Toxicity
C CKO Clonal Toxicity
~ Rodent Quantal Toxicity
~ Freshwater Fish Toxicity
~ Freshwater Invertebrate
~ Algal Test
~ Insect Toxicity
O Pla.it Test
Q Soil Test
Figure 4. Example of Level 1 sample processing form.
-------
ASSESSMENT OF POTENTIAL HAZARD
ASSESSMENT OF -
ENVIRONMENTAL
FATE
BiOACCUMUL AT ION
MEASUREMENT
MEASUREMENT AND LEVEL I
EMISSION^ DOCUMENTATION SAMPLE COLLECTlON_ ^ CHEMICAL _ DISCHARGE
SITE OF RELEASE RATE [DEFINE CONDITIONS' ANALYSIS' SEVERITY
Wt iGMTEO
- DISCHARGE -
severit yiwds]
/
QUANTITATIVE
» RANKING OF
TOKICITV
PRETEST
PROCESSING*
LEVEL I
BIOASSAYS
/
TOXiClT *
DE F INI I ION
*()eli'i—t
O
5",
>
a
"-0
^3
>
?0
>
H
HH
O
2
o
H
PI

H
HH
O
Figure 5. Proposed scheme for a second sLage evaluation of Level I results (defined in ^
o
-j
Lentzen et al., 1978) °
-------
408
D.J. BRL'SICK
REFERENCES
Srusick, D.J. 1980. Level 1 Biological Testing Assessment and
Data Formatting. EPA-600/7-80-079. U.S. Environmental
Protection Agency: Research Triangle Park, NC.
Duke, K.M. , M.E. Davis, and A.J. Dennis. 1977. IERL-RTP
Procedures Manual: Level 1 Environmental Assessment
Biological Tests for Pilot Studies. EPA-600/7-77-043. U.S.
Environmental Protection Agency: Research Triangle Park, NC.
Epler, J.L. 1979. Mutagenicity testing of energy—related
compounds. In: Energy and Health Proceedings. N.E. Breslow
and A.S. Whitteraore, eds. Siara Institute for Mathematics and
Sociology: Philadelphia. pp. 17-36.
Fisher, G.L., C.E. Chrisp, and O.G. Raabe. 1979. Physical factors
affecting the mutagenicity of fly ash from a coal-fired power
plant. Science 204:879-881.
Huisingh , J., R. Bradow, R. Jungers , L. Claxton, R. Zweidinger, S.
Tejada, J. Bumgarner, F. Duffield, M. Waters, V.F. Simmon, C.
Hare, C. Rodriguez, and L. Snow. 1978. Application of
bioassay to the characterization of diesel particulate
emissions. In: Application of Short-terra Bioassays in the
Fractionation and Analysis of Conplex Environmental Mixtures.
M.D. Waters, S. Nesnow, J.L. Huisingh, S.S. Sandhu, and L.
Claxton, eds. Plenum Press: New York. pp. 381-418.
Klekowski, E., and D.E. Levin. 1979. Mutagens in a river heavily
polluted with paper recycling wastes: Results from field and
laboratory mutagen assays. Environ. Mutagen. 1:209-219.
Kubitschek, H.E., and L. Venta. 1979. Mutagenicity of coal fly
ash from electric power plant precipitators. Environ.
Mutagen. 1:79-82.
Lentzen, D.E., D.E. Wagoner, E.D. Estes, and W.7. Gutknecht . 1978.
IERL-RTP Procedures Manual: Level 1 Environmental Assessment
(Second Edition). EPA-600/7-78-201. U.S. Environmental
Protection Agency: Research Triangle Park, NC.
Lofroth, G. 1978. Mutagenicity assay of combustion emissions.
Chemosphere 10:791-798.
Pitts, J.N., Jr., D. Grosjean, T.M. Mischke, V. Siramon, and D.
Poole. 1977. Mutagenic activity of airborne particulate
organic pollutants. Toxicol. Lett. 1:65-70.
-------
EFFECTS OF COLLECTION AND PREPARATION ON TESTING
409
Sexton, N.G. 1979. Biological Screening of Complex Samples from
Industrial/Energy Processes. EPA-600/8-79-021. U.S.
Environmental Protection Agency: Research Triangle Park, NC.
Teranisiki, K. , K. Haraada, and H. Watanabe. 1978. Mutagenicity in
Salmonella typhimurium mutants of the benzene-soluble organic
matter derived from airborne particulate matter and its five
fractions. Mutation Res. 56:273-280.
-------
Intentionally Blank Page
-------
BIOLOGICAL MONITORING OF FLUIDIZED BED COAL COMBUSTION
OPERATIONS I. INCREASED MUTAGENICITY DURING PERIODS OF
INCOMPLETE COMBUSTION
H.E. Kubitschek, D.M. Williams, and F.R. Kirchner
Division of Biological and Medical Research
Argonne National Laboratory
Argonne, Illinois
INTRODUCTION
The Ames Salmonella microsome assay (Ames and Yamasaki, 1975)
is used extensively to screen environmental pollutants because of
its rapidity, its relatively low cost, and its sensitivity in
detecting carcinogens (McCann et al., 1975). We applied the Ames
assay to the study of mutagenic particulates produced by an
experimental process-development fluidized bed combustor (FBC) at
Argonne National Laboratory. This FBC also was used for mouse
inhalation toxicology experiments (Kirchner et al., 1980; this
study is reported in this volume as Biological Monitoring of
Fluidized Bed Coal Combustion Operations II.).
The original plan was to determine the relative mutagenicities
of the different effluents produced by the FBC and to measure the
mutagenicity of the particulate effluent during the mouse
exposures. However, our earliest measurements (Kubitschek and
Haugen, 1980) indicated that the mutagenicity was much greater than
that for a larger FBC (Clark et al., 1978). This excessive
mutagenicity in the Argonne FBC was traced to operating conditions:
fly ash deposited on the final filter during start-up periods was
up to 60 times as mutagenic as that produced during steady
operation (Kubitschek and Williams, 1980). This finding, and the
variability in the mutagenicity of the fly ash produced during a
1000-h (40-day) run, made it apparent that the Ames assay might be
used to examine the effects of process conditions on levels of
mut agenic i ty.
411
-------
412
H.E. KUB1TSCHEK ET AL.
MATERIALS AND METHOD
The Argonne FBC, operated by the Chemical Engineering Division,
is a process-development, 6-in.- (15.2-cm-) diameter combustor
maintained at atmospheric pressure. This FBC burned high-sulfur
(5.5%) bituminous coal (Sewickly) and used as the sorbent either of
two calcitic limestones (Greer or Grove) treated with sodium
carbonate. The combustor operated at a power level of 21 kW, with
coal feed controlled by the concentration of oxygen in the off-gas
stream. Oxygen was normally maintained at 3%, and sulfur dioxide
at 700 ppm. The operating temperature was 850°C.
Particulates were collected in cyclones and also on a porous
metal filter, located downstream in the off-gas line. This filter
had an efficiency greater than 98% for particles 6 urn in diameter
or larger. Particle cutoff diameters for the first and second
cyclones were approximately 8 and 5 vim, respectively. An important
design feature of the F3C was the availability of twin off-gas
clean-up trains downstream from the combustion chamber, each
containing a primary cyclone, secondary cyclone, and porous metal
filter. The effluent was vented to the exhaust system. With this
equipment, off-gas trains were used alternately to collect filter
particulate samples, and equipment could be cleaned to reduce
cross-contamination between the samples.
Initially, particulate samples (fly ash) were collected after
runs of a single day's duration (8 to 10 h). These samples were
extracted for 45 min at 37°C with dimethylsulfoxide, which yielded
greater observable mutagenicity than did any of the other organic
solvents tested (Kubitschek and Haugen, 1980). Metabolic
activation did not increase observable mutagenicity (Kubitschek and
Haugen, 1980), which agreed with similar observations by Fisher et
al. (1979). In all of the experiments described below, mutagenic
activity was determined with the Ames Salmonella assay using strain
TA98 without microsomal enzyme activation.
RESULTS
Mutagenicity During Start-up and Steady Operation
In a series of tests of fly ash from single daily runs of the
Argonne FBC, we observed mutagenicity levels in excess of 1000
revertants/mg (Kubitschek and Haugen, 1980; Kubitschek and
Williams, 1980). These values were much greater than those
obtained by Clark et al. (1978) for fly ash samples from the
18-in.- (45.7-cm-) diameter F3C at the Morgantown Technology
Center. Clark et al. found that when mutagenicity was detectable,
activities were less than 3 revertants/mg.
-------
BIOLOGICAL MONITORING OF COAL COMBUSTION I
413
We first considered the possibility that these very different
mutagenicities might be due to the different temperatures at which
the samples were collected. Natusch and Tompkins (1978) predicted
that the adsorption of mutagenic polycyclic aromatic hydrocarbons
would increase greatly as tempertures were decreased below 150°C,
and the collection temperature of the Argonne filter (70°C) was
known to be somewhat less than that for the Morgantown samples
(estimated to be in the range of 70° Co 90°C) . However, when the
Argonne FBC was operated for longer periods, somewhat lower sample
activities were observed, indicating that sample activities might
depend on operating conditions.
To distinguish between these possible explanations, fly ash
samples were collected from the first and second cyclones and the
filter both during start-up periods and during steady operation.
The observed mutagenicities (Kubitschek and Williams, 1980) are
shown in Table 1. An inverse relationship between mutagenicity and
particulate collection temperature can be seen, in agreement with
the predictions of Natusch and Tompkins (1978). However,
mutagenicity levels varied much more widely when operating
conditions were changed, and fly ash mutagenicity during start-up
of operations was as much as 60-fold greater than that observed
during later steady operation. Clearly, the great bulk (> 98%) of
Che difference between our earlier determinations and those of
Clark et al. (1978) can be assigned to excessive mutagenicity that
was deposited in our samples during start-up and shut-down of
operac ions.
Table 1
Particulate Effluent Mutagenicity During Start-up
and Steady FBC Operationa>
Operat ion
Site
Temperature
Start-up
St eady
Rat io
Primary cyclone
150°C
3
+
2
0
+
1
Secondary cyclone
95°C
470
+
25
8
+
1 59
Filter
70"C
1400
+
20
22
±
2 64
•From Kubitschek and Williams, 1980, by permission of the
publisher.
^Values are averages ± standard errors for net numbers of TA98 His+
revertants/mg ash extracted.
-------
414
H.E. KUBITSCHEK ET AL.
While Che data shown in Table 1 are those for the experiment
with the greatest difference between start-up and steady operating
conditions, the same qualitative results were obtained in each of
three experiments. These results suggest that excessive
mutagenicity might be produced during periods of incomplete
combustion, as would be expected during start-up of combustion
(Kubitschek and Williams, L980).
Mutagenicity During Departures from Steady Operation
Particulate mutagenicity was monitored for samples collected
daily during a LOOO-h exposure of mice to gaseous and particulate
effluents from the FBC. During this period, the specific
mutagenicity of the fly ash varied widely (Figure 1) with unusually
high levels of mutagenic activity produced during six transient
periods. Later, by examining the record of operations, we found
that each of these peaks of mutagenic activity (Table 2) was
associated with a departure from steady combustor operation, due
either to start-up of operations or to mechanical breakdown
(Kubitschek et al., 1980).
Examination of the gas concentration records indicated that
carbon monoxide (CO) concentrations also increased during the same
periods of high fly ash mutagenicity. Figure 1 shows the sum of
the peak values for CO concentrations in excess of 1000 ppm during
each 12-h period of the run. This level was chosen because
mutagenicity did not appear to be closely correlated with the CO
peaks of smaller magnitude, which became more frequent as CO
concentrations approached the background level of approximately 500
ppm. Good correlation between this peak CO concentration and
sample mutagenicity is evident in Figure 1; the correlation
coefficient is 0.72.
Correlations in time and in intensity for the individual
transients supported the correlation between mutagenicity and peak
CO concentrations. The mean time of occurrence of mutagenicity in
each peak and Che mean time of peak CO production for the
corresponding periods are shown in Figure 2 and Table 2. The
coefficient of correlation between Che mean times of occurrence was
0.9998. This close correlation is especially noteworthy
considering that the average duration of these six transient
periods was more than three days, while the average deviation
between corresponding CO and mutagenicity peaks was less than 5 h.
Good correlation also was observed between the magnitude of CO
peak production and mutagenicity (shown in Figure 3 and Table 2);
the correlation coefficient was 0.81. The average net number of
revertants per milligram produced per CO value at the standard
level of 1000 ppm CO was 16.4 (standard error = 2.1). Thus, both
-------
1 )""" " I —I 1
• Net TA98 reverfants per mg fly ash
V/A Sum of CO concentration peaks above IOOO ppm
I I CO concentration peaks in presence of excess oxygen
Day
Fly ash mutagenicity and peak CO concentrations during a AO-day period of KHC
operation (Kubitschek fit al., 1980). The bar graph shows the sum of the CO
concentration peaks for each half-day period. The letter T (day 10) indicates a
lar^e decrease in operating temperature. The dashed line indicates a period of
about: 10 days during which the FBC was shut down for repairs.
-------
416
H.E. KUBITSCHEK ET AL.
Table
2. Mutagenicity and CO
Produc t ion
During Departures

from Steady FBC Operation
a

Mean
Time (days)*5
<

Period

CO
Mutagenicity

(days)
CO Peaks
Mutagenicity (
ppm x 10-3)
(rev,/ mg)
Rat ioc
0-2
0.55
0.58
21.2
274
12.9
6-8
7.42
7.67
11.7
126
10.8
9 — 13d
11 .60
11.35
18.4
458
24.9
17-19
18.50
18.09
8.4
154
18.3
20-21
21.10
20.97
3.7
56
15.1
32-35
33.80
33.63
14.4
237
16.5

Mean: 16
.4 ± 2.0
aFrora Kubitschek et al., 1980.
bTh e mean time of occurrence for each transient increase in CO
production (CO peaks) and in mutagenicity is given in days
numbered sequentially during the tested periods.
cRevertants per milligram: ppra CO x 10"^.
^Values for the low temperature period on day 10 are excluded.
the time of occurrence and the degree of mutagenicity of these
transient periods of departure from steady operation were correlated
with increased concentration of CO in the combustion chamber.
CONCLUSIONS
The observed correlations strongly support our earlier
suggestion (Kubitschek and Williams, 1980) that excessive effluent
mutagenicity occurs during incomplete coal combustion. The
increase in mutagenicity during periods of incomplete combustion
might have been a result of the evolution of adsorbed hydrocarbons
(especially during start-up) or, alternatively, the result of only
partial oxidation of Che raacromolecules of which coal is composed,
with the release of complex hydrocarbon moieties that are mutagens
or are capable of forming mutagens. Preliminary results for
chemical characterization of mutagens in fly ash samples from the
Argonne FBC are consistent with either hypothesis: several classes
of mutagenic compounds were identified, including 2- to 6-ring
aromatics, phenols, carbonyls, alcohols, and carboxylic acids, and
other very polar compounds (Kubitschek and Haugen, 1980), and the
-------
BIOLOGICAL MONITORING OF COAL COMBUSTION I
417
30
V)
o
Q
c
0)
C7>
o
20 -
-------
418
H.E. KUBITSCHEK ET AL.
500
400
o>
E
300
c
o
w
V
>
V
X.
o 200
E
3
v>
100
0
50
150
100
200
Sum of CO Peak Values
Figure 3. Correlation between specific mutagenicity and the sura of
the CO concentration peak values (redrawn from figure in
Kubitschek et al. , 1980).
operation, then biological risk would be determined mainly by the
frequency and duration of those operational transients.
ACKNOWLEDGMENTS
These studies could not have been carried out without the generous
assistance of K.M. Myles and G.W. Smith, of the Chemical
Engineering Division, during the course of the run. We thank them
for their help and for making the records of operation available to
-------
BIOLOGICAL MONITORING OF COAL COMBUSTION I
419
us. We also thank V.A. Pahnke, of our division, for collecting
samples. This work was supported by the U.S. Department of Energy
under contract No. W-31-109-ENG-38.
REFERENCES
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods for
detecting carcinogens and mutagens with the Salmonella/
mammalian-microsome mutagenicity test. Mutation Res.
31:347-364.
Clark, C.R., R.L. Hanson, and A. Sanchez. 1978. Mutagenicity of
effluents associated with the fluidized bed combustion of
coal. In: Inhalation Toxicology Research Annual Report
1977-1978. Lovelace Biomedical and Environmental Research
Institute.
Fisher, G.L., C.E. Chrisp, and O.G. Raabe. 1979. Physical
factors affecting the mutagenicity of fly ash from a
coal-fired power plant. Science 204:879-881.
Kirchner, F.R., D.M. Buchholz, V.A. Pahnke, and C.A. Reilly, Jr.
1980. Biological monitoring of fluidized bed combustion
operations II. Mammalian responses following exposure to the
gaseous effluents. Presented at the U.S. Environmental
Protection Agency Second Symposium on the Application of
Short-term Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures, Williamsburg, VA.
Kubitschek, H.E., and D.A. Haugen. 1980. Biological activity of
effluents from fluidized bed combustion of high-sulfur coal.
In: Health Implications of New Energy Technologies. W.N. Rom
and W.E. Archer, eds. Ann Arbor Science Publishers: Ann
Arbor, MI. pp. 381-394.
Kubitschek, H.E., and D.M. Williams. 1980. Mutagenicity of fly
ash from a fluidized bed combustor during start-up and steady
operating conditions. Mutation Res. 77:287-291.
Kubitschek, H.E., D.M. Williams, and F.R. Kirchner. 1980.
Correlation between particulate effluent mutagenicity and
increased carbon monoxide concentration in a fluidized bed
coal combustor. Mutation Res. 74:329-333.
McCann, J., E. Choi, E. Yamasaki, and 3.N. Ames. 1975. Detection
of carcinogens as mutagens in the Salmonella/microsome test:
Assay of 300 chemicals. Proc . Nat. Acad. Sci. USA
72:5135-5139.
-------
420
H.E. KUBITSCHEK ET AL.
Natusch, D.F.S., and 3.A. Tomkins . 1978. Theoretical
consideration of the adsorption of polynuclear aromatic
hydrocarbon vapor onto fly ash in a coal-fired power plane.
In: Carcinogenesis, Vol. 3: Polynuclear Aromatic
Hydrocarbons. ?.W. Jones and R.I. Freudenthal, eds . Raven
Press: New York. pp. 145-303.
-------
BIOLOGICAL MONITORING OF FLUIDIZED BED COAL COMBUSTION
OPERATIONS II. MAMMALIAN RESPONSES FOLLOWING EXPOSURE TO
GASEOUS EFFLUENTS
F.R. Kirchner, D.M. Buchholz, V.A. Pahnke, and
C.A. Reilly, Jr.
Division of Biological and Medical Research
Argonne National Laboratory
Argonne, Illinois
INTRODUCTION
Several coal combustion technologies, including fluidized bed
combustion, have potential for meeting government-imposed
environmental pollution guidelines. Fluidized bed combustion
enables the combustion of high sulfur coal while limiting the
emission of sulfur dioxide.
In a previous study (Kirchner et al., 1980), detrimental
biological effects were observed in animals exposed to whole
(gaseous and particulate) effluents derived from the atmospheric
pressure fluidized bed coal combustor (FBC) operated by the
Chemical Engineering Division at Argonne National Laboratory (ANL),
Argonne, IL. In the study presented here, the animals were exposed
only to gaseous FBC components in order to determine which effluent
components are responsible for these observed biological effects.
Three testing systems were applied.
To assess pulmonary alveolar macrophage (PAM) function, one of
the organism's first lines of defense against airborne pollutants,
3rennan at al. (1980) modified the assay of neutrophil function
first described by Tan et al. (1971). The assay differentiates
between defective phagocytosis and impaired intracellular killing.
Effective PAM function is required to eliminate inhaled micro-
organisms, particulate pollutants, and other inhaled particles.
Increased numbers of PAMs were observed in the lungs of
animals exposed only to FBC fly ash (Brennan et al., 1980),
taxing the lung macrophage progenitor stem cell compartment. To
assess the systemic toxic effects following acute exposure to FBC
421
-------
422
F.R. KIRCHNER ET AL.
gaseous effluent, the Till and McCuiloch spleen colony assay (1961)
was used. The lungs and other tissues thought to be at risk, were
also examined histopathologically during and after the exposures.
METHODOLOGY
The process-development-scale atmospheric pressure FBC
described by Kirchner et al. (1980) was used to expose mice and
rats to gaseous effluents of fluidized bed combustion. The
effluent was diluted in two stages by an effluent dilution system.
The effluents were taken fron a point downstream of the final
filter particle cleanup system, to eliminate the particulate
component of the effluents (Figure 1). Following a 20-fold
dilution, the effluents were delivered to the top of an Atmospheric
Effects Simulator (AES)(Figure 2) that exposed the diluted effluent
to simulated sunlighc for 15 min. The animals were thus exposed to
diluted gaseous combustion effluents only. The levels of carbon
monoxide, nitrogen monoxide, sulfur dioxide, nitrogen oxides, and
total vapor-phase hydrocarbons were monitored instruraentally in the
exposure chambers throughout the experiments.
PURPOSE
Two experiments, each with a 500-h (~ 21-day) continuous
exposure period, were conducted. In Experiment I, 154 male
B6CFi/Anl mice, each 120 days old, and 8 male Fisher F-344 rats,
each 90 days old, were exposed in chambers designed and built in
this laboratory (Kirchner et al., 1980; Figure 3). At the same
time, similar control animals (66 mice and 8 rats) were placed in
identical chambers through which only HEPA-filtered room air was
passed. In Experiment II, previously exposed animals (88 mice and
8 rats) were re-exposed for a second 500-h period. In addition,
previously untreated animals (88 mice and 8 rats) were exposed for
500 h. The two control groups in Experiment II were also exposed
to HEPA-filtered room air: one group consisted of 44 control mice
from Experiment I, and the other 44 untreated mice. The animals
were maintained on a cycle of 12 h light and 12 h darkness. They
had food (Wayne Lab Blox, Allied Mills) and water ad libitum. The
animals' condition was checked daily.
The animals were exposed to gaseous effluents obtained only
during steady-state operation conditions. Steady state is defined
by the following conditions: 1) a period at least 12 h after
initiation of coal feed; 2) sulfur dioxide concentrations in the
undiluted off—gas at a constant 700 ± 50 ppm; 3) a bed temperature
of 850 ± 5°C; and 4) carbon monoxide, nitrogen monoxide, and total
vapor-phase hydrocarbon concentrations stabilized.
-------
TO BUILDING EXHAUST
ANIMAL EXPOSURE
ENCLOSURE
TO BUILDING
EXHAUST
FINAL
FILTER
2° DILUTER
DILUTER
Q
FLOW
CONTROL
ORIFICE (I CFM)
A P SENSORS
A CFM HOUSE
COMPRESSED AIR
15 CFM HEPA
FILTERED AIR
OVERPRESSURE DIPTUBE
AES CHAMBER
a
o
r
o
o
o
>
o
'2.
H
O
35
1° a 2° CYCLONES 2
2
o
o
~n
o
o
>
f
n
o
<»
5
FBC §
K-l
o
Figure I.
Schematic diagram of the FBC and the associated effluent delivery system, the
AES (reproduced from Kirchner et al., 1980).
¦p.
to
co
-------
*&MS?'V SU0&&
Figure 2. AES with one of the protective covers holding the
aluminum reflectors removed. Note sample ports on far
side of the chataber used for collecting gaseous and
particulate samples from the A.ES (reproduced from
Kirchner et al., 1980).
-------
Figure. 3. Animal exposure system shown with doors to aluminum enclosure removed
(reproduced from Klrchner ct al., 1980).
-------
426
F.R. KIRCHNER ET AL.
To verify that the animals were exposed only to the gaseous
component of the effluent, the concentration of particulate in the
diluted off-gas was monitored with optical and electrical aerosol
analyzers. For the smallest-diameter particles (0.01 to 1.0 urn)
the Thermo Systems, Inc., electrical aerosol analyzer was used.
For particles in the range of 0.3 to 3.0 pm or greater, the Royco
forward-light-scattering particle counter was used. These analyses
also indicated the efficiency of the particle cleanup system.
After 250, 500, and 1000 h of exposure, two randomly selected
animals from each group were sacrificed by cervical dislocation.
Tissues from the lung, liver, kidney, spleen, and heart were fixed,
stained with hematoxylin and eosin, and examined histologically.
Four to eight days after termination of exposures, the PAM assays
(from mice and rats), femoral bone marrow spleen colony assays
(mice only), and histopathological examinations (mice only) were
performed.
RESULTS
The levels of particulate effluent measured by the electronic
particle counters were less than 10% of those observed during
exposures in which the effluent was taken from a point just before
the final filter of the particle cleanup system. Mo particles
larger than 1.0 pm were detected.
The combustion-gas concentrations during the two experiments,
given in Table 1, are similar to those previously reported by
Kirchner et al. (1980). The only toxicant above the human
threshold linit value (TLV)(ACGIH, 1977) was sulfur dioxide (TLV =
5 ppm)(animals exposed to 22 t 5.1 ppm). The high value for
sulfur dioxide resulted from the 20-fold dilution factor; the
concentration would have been below the TLV following atmospheric
dilution. (Note: TLVs are based on time-weighted average
concentrations to which workers can be exposed for a normal 8-h
workday or 40-h workweek. The mice, however, were exposed
continuously during the 500-h exposures.)
In Experiment I, three experimental mice died (1.5% of the
number entering the experiment); the deaths occurred at days 6, 11,
and 13 of exposure. In Experiment II, four effluent-exposed mice
died, two from the 500-h exposed group (one at day 12 and one at
day 16) and two from the 1000-h exposed group (both on day 16).
These mice represented 2.3% of the animals entering the experiment
from each group. No moribund animals were observed and no rats or
control animals died during either experiment.
-------
BIOLOGICAL MONITORING OF COAL COMBUSTION II
427
Table I. Concentration of Gaseous Effluents ir. the
Atmospheric Effects Simulator During a 500-Hour Exposure3
Gas
Concentration (ppm)b
TLV (ppm)c
Carbon monoxide
17 ±11
50
Nitrogen monoxide
19.0 + 6.3
25
Nitrogen oxides
1.1 ± 0. 1
5
Sulfur dioxide
22.0 ± 5.1
5
Total hydrocarbons

(vapor phase)
2.1 ± 0.75

aAverages of readings taken every 6 h during the 5C0-h exposure
± 1 standard error).
bDet ermined by instrumental gas analyzers.
cThreshold limit values for chemical substances in workroom air
adopted by the American Conference of Governmental Industrial
Hygienists (1977).
Four to eight days after the termination of the experiments,
PAMs were taken from both rats and mice. No impairment was found
in their ability to engulf and kill a challenge dose of
Staphylococcus aureus (Table 2).
The spleen colony assay showed that a 500-h exposure to the
effluent increased spleen colony formation 167% over control
values. The sizes of the spleen colonies from mice exposed 500 h
were generally larger than those from the control mice. In mice
exposed for 1000 h, however, the number of spleen-colony-forming
units (CFl's) was comparable to the control value (Table 3).
Histological changes in the lungs of exposed mice were limited
to a slight accumulation of fly ash in the macrophages in mice
sacrificed after 500 and 1000 h of exposure. Fly ash was not
present in the control mice or the animals exposed for 250 h.
These histological changes were mild compared with all levels and
times of exposure for previous exposures to the whole effluent
(Kirchner et al., 1980). No evidence of epithelial hyperplasia
or obvious proliferation of macrophages was seen. The gross and
histological appearances of all other organs were normal.
Three days after termination of Experiment II, the 500- and
1000-h exposed mice had lost 10% of their initial body weights,
while the control mice had insignificant weight loss (< 1%). The
rats, too, had moderate weight loss (500-h exposed, 9%; 1000-h
-------
428
F.R. KIRCHNER ET AL.
Table 2. Pulmonary Alveolar Macrophage Function
Four to Eight Days After In Vivo Exposures to Effluents3
Staphylococcus aureus
Duration of Phagocytized Phagocytized but
Experiment Treatment Exposure (h) & Killed (%) Not Killed (%)
Mice
cont rol
500
98.7
0.2

exposed
500
99.3
0.0
Rats
control
500
95.1
1.5

exposed
500
93.8
1.8
Mice
control
500
99.9
0.0

exposed
500
98.2
0.1

exposed
1000
99.9
0.0
Rats
control
500
93.8
1.5

exposed
500
97.6
0.3

exposed
1000
98. 1
0.7
aPooled alveolar macrophages from 12 mice/group; results are
expressed as the percent of the total challenge dose of bacteria-
Table 3. Hemopoietic Colony-forming Units (CFUs) in Mice
Four to Eight Days After Termination of Exposures to FBC Effluents3

Duration of
CFUs/10sb
Treatment
Exposure (h)
Nucleated Cells
Control
—
46.5 ± 3.65
Exposed
500
78.0 ± 4.16

1000
41.6 ± 2.50
aFor each determination, pooled femoral bone marrow cells (2.0 x
101*/recipient) from two donors were injected into 15 irradiated
recipient mice.
^Mean number of spleen colonies per mouse from 15 mice (± 1
standard error).
-------
BIOLOGICAL MONITORING OF COAL COMBUSTION II
429
exposed, 13%), with a negligible (< 2%) weight loss in the.
controls.
CONCLUSION'S
In the previous study (Kirchner et al., 1980), exposure of
the rodents to whole (gaseous and particulate) FBC effluent for
500 h significantly impaired the PAMs' ability to phagocytize and
kill a challenge dose of Staphylococcus aureus. However, after an
exposure of 1000 h, the exposed animals' macrophages appeared to
accommodate the exposure conditions and were again able to engulf
and kill the bacteria at the same level as those of the control
animals. In the present studies with FBC effluent gas only, the
bacterial cell killing was unimpaired in either the 500- or 1000-h
exposure groups, Indicating that the temporary reduction in
macrophage function was probably due solely to the high level of
particulate matter.
The spleen colony assay results indicated that exposure to the
gaseous effluent for 500 h acted as a strong proliferation and
differentiation stimulus to the pluripotential stem cells. The
numerous colonies represent large numbers of proliferating
pluripotent stem cells; the large size of the nodules indicates the
rapidity of the proliferative response of these cells to the
stimulus (McCulloch, 1970). Prolonged exposures (1000 h) resulted
in CFUs similar to those of unexposed mice in both colony size and
numbers of colonies. These results differed from those observed in
the previous work with particulate effluents (Kirchner et al.,
1980), in which there appeared to be a cumulative toxic effect
(i.e., decrease in CFUs) in the longer exposures, with little, If
any, recovery or adaptation to the effluent exposures. The
observed decrease in CFUs from 500 to 1000 h of exposure may have
been due to a depletion of the stem cell compartment resulting from
the strong persistent stimulus (500-h exposure) for stem cell
proliferation (Till, 1976). The initial demand for rapid
proliferation may have reduced the pluripotent stem cell pool,
perhaps simultaneously decreasing the committed stem cell pool.
The systemic stress on the hematopoietic stem cell compartment may
ultimately impair one or more of the functional end cells. This
possible alteration of hemopoiesis is currently under
investigation.
ACKNOWLEDGMENTS
The authors would like to thank V. Ann Ludeman for her
technical assistance and Dr. T.E. Fritz for conducting the
histopathologic diagnosis. This work was supported by the U.S.
Department of Energy under contract no. W-31-I09-ENG-38.
-------
430
F.R. KIRCHNER ET AL.
REFERENCES
ACGIH, American Conference of Governmental and Industrial
Hygienists. 1977. Threshold Limit Values for Chemical
Substances in Workroom Air Adopted by ACGIH for 1977.
Brennan, P.C., F.R. Kirchner, J.O. Hutchens, and W.P. Norris.
1980. The effect of reaerosolized fly ash from an atmospheric
fluidized bed corabustor on murine alveolar macrophages. In:
Pulmonary Toxicology of Respirable Particles. C.L. Sanders,
F.T. Cross, G.E. Dagle, and J.T. Mahaffey, eds. CONF-791002.
U.S. Department of Energy: Washington, DC. pp. 279-288.
Kirchner, F.R., J.O. Hutchens, P.C. Brennan, D.A. Haugen, H.E.
Kubitschek, D.M. Buchholz, R. Kumar, K.M. Myles, and W.P.
Norris. 1980. Mammalian responses following exposure
to the total diluted effluent from fluidized bed combustion of
coal. In: Pulmonary Toxicology of Respirable Particles.
C.L. Sanders, F.T. Cross, G.E. Dagle, and J.A. Mahaffey, eds.
CONF-791002. U.S. Department of Energy: Washington, DC.
pp. 29-46.
McCulloch, E.A. 1970. Control of heraatopoiesis at the cellular
level. In: Regulation of Hematopoiesis, Volume 1. A.S.
Gordon, ed. Appleton-Century-Crofts: New York. pp. 133—
159.
Tan, J.S., C. Watanahunakorn, and J.P. Phair. 1971. A modified
assay of neutrophil function: use of lysostaphin to
differentiate defective phagocytosis from impaired
intracellular killing. J. Lab. Clin. Med. 78:316-322.
Till, J.E. 1976. Regulation of hemopoietic stem cells. In: Stem
Cells of Renewing Cell Populations. A.B. Cairnie, P.K. Lala,
and D.G. Osmond, eds. Academic Press: New York. pp. 143—155.
Till, J.E., and E.A. McCulloch. 1961. A direct measurement of the
radiation sensitivity of normal mouse bone marrow cells.
Radiat. Res. 14:213-222.
-------
IN VITRO AND IN VIVO EVALUATION OF POTENTIAL TOXICITY OF
INDUSTRIAL PARTICLES
Catherine Aranyi and Jeannie Bradof
Illinois Institute of Technology Research Institute
Chicago, Illinois
Donald E. Gardner and Joellen Lewtas Huisingh
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
INTRODUCTION
Alveolar macrophages (AM) protect the lungs by phagocytosis
and digestion of inhaled irritant particles and infectious agents.
Reduced activity of the AM system can impair the lung's defensive
capacity and increase susceptibility to respiratory disease. Since
resistance to infection may be lowered by exposure Co an inhalation
hazard, changes in the major functional characteristics of AM can
be used to monitor environmental stresses in the intact animal. In
addition, since these cells can be obtained easily by tracheo-
bronchial lavage and maintained in culture, they are frequently
used in in vitro cellular toxicology to assess the potential
inhalation hazard of various substances.
The advantages of in vitro screening assays in terras of cost
and time efficiency are well known. The rabbit alveolar macrophage
(RAM) test, a rapid, efficient in vitro assay, has been used
extensively by the U.S. Environmental Protection Agency (EPA) and
in our laboratories to compare the cytotoxicity of a variety of
soluble compounds and particulate materials (Aranyi et al., 1977,
1979; Mahar, 1976; Waters et al., 1974a, b; 1975a, b; 1973). This
system is capable of rapid screening and toxicity ranking of test
materials and thereby identifies not only the potentially hazardous
but also the inert agents. Based on the in vitro results, the
number of samples to be studied further in vivo can be reduced
considerably.
The purpose of these studies was to determine whether in vitro
exposure of AM to various complex industrial particles produced
the same relative toxicity ranking as inhalation exposure to
431
-------
432
CATHERINE ARANYI ET AL.
aerosols of these particles in vivo. Our main objective was Co
establish the correlation between Che effects of in vitro and in
vivo exposures and to determine whether inhalation hazards could be
predicted on the basis of in vitro screening assays.
MATERIALS AND METHODS
Particulate Stationary Source Samples
The particles were collected as baghouse samples, from
electrostatic precipitators (ESP) or by cyclone sampling train from
various stationary point sources, including four coal-fired power
plants (three conventional and one fluidized-bed process), a steel
foundry, and an aluminum and a copper smelter. Because of the
large sample requirement for the aerosol inhalation exposure
studies, it was not possible to provide sufficient amounts of
samples collected by special emission source samplers located
beyond the normal in-plant control devices. Thus, the samples
reported here were not true emission or effluent samples and were
not necessarily similar in composition or toxicity to emission
samples.
Foundry, smelter, and selected coal combustion samples (fly
ash nos. I, 2, and 3) were provided and inorganic analysis of these
particulate materials was performed by EPA Industrial Environmental
Research Laboratory (IERL) at Research Triangle Park, N.C. The
conventional power plant coal fly ashes (nos. L, 3, and 4)
originated from high-sulfur-containing Eastern coals, and nos. 1
and 3 were collected as baghouse samples. Fly ash no. 4, a <
3.3-um aerodynamic size fraction of an ESP hopper fly ash, has been
studied in depth by Griest and Guerin (1979) and contains
arsenic, barium, cobalt, chromium, copper, lead, strontium, and
zinc as some of the more prevalent trace metals, as well as traces
of polycyclic aromatic hydrocarbons. The no. 2 fly ash was
collected at 840"C (1550°F) from the second cyclone of a
calcium-oxide-fluidized-bed coal combustion process in an
experimental demonstration plant. Because of the high collection
temperature, residual organic compounds were < 0.1%. Major
inorganic components identified by spark source mass spectroscopy
were calcium, magnesium, iron, silicon, and sulfur. The steel
foundry particles were collected as a baghouse sample. No
extractable organics were found, and the major inorganic
constituents were iron, silicon, magnesium, and zinc. The copper
smelter dust was collected by ESP at 200°C (400°F). No organic
components have been identified, but trace metals such as lead,
arsenic, copper, iron, antimony, and zinc were found in high
concentrations. The aluminum smelter sample, collected as a
baghouse dust, contained 3 mg./g of extractable organic compounds,
-------
TOXICITY OF INDUSTRIAL PARTICLES
433
which were mostly fused aromatics. No information on Che inorganic
constituents was available.
All collected samples were air-classified, and only the
particles in the size-fraction of ( 3 pm were used in the
experiments. Since our previous studies (Aranyi et al., 1979)
demonstrated that the smallest particles had the most deleterious
effects on AM in vitro, we wanted to explore the in vivo
correlation for this size range in particular.
In Vitro Methods
The in vitro experimental procedures have been described
previously in more detail (Aranyi et al., 1979). Briefly, AM
obtained from rabbits by tracheobronchial lavage were centrifuged
and washed in Hanks' Balanced Salt Solution (HBSS), and total and
differential cell counts and viability were determined. AM
suspensions and separate particle suspensions at twice the
projected exposure concentrations were prepared in Medium 199/HBSS
supplemented with serum and antibiotics, and equal volumes of the
two suspensions were mixed. The final concentration of AM in the
test suspensions was maintained constant at 10^ AM/ral: the
concentration of the particles was increased stepwise to 1000 yg/ml
to attain a dose-relaced change in viability from approximately 20
to 90%. The test suspensions were incubated in wells of disposable
plastic cluster dishes placed on a rocker platform for 20 h at 37°C
in a humidified CO2 atmosphere. Immediately after incubation,
percent viability was determined by dye exclusion. The test
suspensions were subsequently washed, centrifuged, and resuspended
in HBSS before total cellular protein and adenosine triphosphate
(ATP) levels were monitored. An aliquot was treated with sodium
deoxycholate, the resulting cell lysate was centrifuged at 10,000 x
g, and the supernatant was used for the Lowry protein assay. A
second aliquot was used for ATP determination; after extraction
from Che cells with dimethylsulfoxide (DMSO), ATP was determined
through the luciferin-luciferase reaction in a Dupont 760
Luminescence 3iorieter.
In Vivo Methods
Inhalation exposures. All inhalation exposure facilities are
located in rooms maintained under negative pressure relative to
outside areas. The animal exposure chambers as well as the aerosol
generation and dilution systems are housed in second chamber
enclosures (safety cabinet-glove box) that permit safe handling of
the animals and maintenance and monitoring of the experimental
environment.
-------
434
CATHERINE ARANYI ET AL.
Animals were exposed to Che experimental environment in
Plexiglas chambers of various sizes (87 to 476 liters) that can
hold up to 240 mice in individual compartments of wire cages. The
compressed air supplied to the exposure chambers for dilution or
dissemination of Che test agents was passed through appropriate
filter systems to dry the air and remove all traces of oil and
particulates. Ambient temperature and humidity were maintained
throughout the exposures by providing an adequate flow of
conditioned air.
Aerosols of the coal fly ash and the copper smelter dust were
generated with a Wright Dust Feeder. Mass concentration was
monitored optically, using a Phoenix JM7000 Aerosol Smoke and Dust
Photometer, and gravimetrically, by weighing particles collected on
membrane filters on an analytical raicrobalance and measuring the
air volumes sampled by a gas meter for the corresponding time
intervals. Aerosol particle size distribution was monitored by an
ASAS-300A Active Scattering Aerosol Spectrometer (Particle
Measuring Systems, Inc.).
The infectious aerosol was generated with a Model 841
DeVilbiss nebulizer using Streptococcus pyogenes (Lancefield Group
C), grown in Todd Hewitt Broth from stock cultures obtained from
colonies isolated from the hearts of infected mice. For the
bactericidal activity assay, aerosols of 35S-labeled K. pneumoniae
were disseminated with a Retec X-70 disposable nebulizer. Aerosol
particle size produced by both nebulizers was between 1 and 5 um
MMD.
Kadiolabeled K. pneumoniae were cultured in modified
Anderson's medium in which the sulfate requirement of the bacteria
was provided by -^S-labeled sodium sulfate. Before aerosolization,
the bacteria were repeatedly washed and centrifuged for removal of
unattached radiolabel. 3acterial counts were determined in a
Petroff-Hauser counting chamber by dark field microscopy and also
by culture plate technique. Radioactive counts were measured with
a Mark III Liquid Scintillation System Model 6880 (Searle Analytic
Inc .) .
Health-effect assays. Groups of 4- to 6-week-old female CD^
mice (Charles River Laboratories) were exposed to either aerosols
of the test particles at 2 mg/m- mass concentration or to filtered
air for 3 h/day, 5 days/week, for one, two, or four weeks. Health-
effect assays followed within one hour of the last exposure.
Pulmonary free cells were obtained from mice by tracheo-
bronchial lavage. Total cell counts were made in a hemocytometer,
and differential counts were made of smears of cells fixed in
methanol and stained with Wright's stain. Viability was determined
-------
TOXICITY OF INDUSTRIAL PARTICLES
435
by dye exclusion. Cellular ATP concentration in the lavaged cells
was monitored, using a Dupont 760 Luminescence Bioineter.
Pulmonary bacterial activity was determined in the lungs of
individual animals through a modification of the method of Green
and Goldstein (1966), whereby mice exposed to particles and control
mice exposed Co filtered air are challenged with radiolabeled live
bacteria. The ratio of the viable bacterial count to the
radioactive count in each animal's lung gives the rate at which
bacteria are destroyed by the lung in a given time after infection.
The streptococcus infectivity model (Ehrlich, 1966: Gardner, 1979)
was used to determine the effect of particle exposure on
susceptibility to respiratory infection. Groups of exposed and
control mice were simultaneously challenged with streptococcus
aerosol. After the challenge, the mice were removed to a clean-air
isolation room, and mortality rate and survival time were recorded
over a 14-day holding period.
RESULTS
In Vitro Tests
The effects of in vitro incubation of the various particulate
samples on rabbit AM were monitored in dose-response experiments,
in which cell viability (percent), total protein (micrograms,
percent of control), and ATP (femtograms ATP per microgram protein,
percent of control) were measured. The means and standard errors
for these parameters were calculated from six or nine replicates at
each concentration, with triplicate assay determinations of each
replicate. When regression analysis was applied to these data,
highly significant negative linear dose-response relationships were
observed for each parameter in all samples (P was generally
< 0.001 and occasionally < 0.01).
The estimated concentrations that were required to reduce the
experimental parameters to 50% of the control responses were
calculated from the linear regressions. From these data, the
samples could be ranked by relative toxicity, as shown in Table 1.
Two of the samples examined, those collected from the copper and
aluminum smelters, were highly cytotoxic; the copper smelter sample
was the more toxic of the two. Three of the power plant coal fly
ashes (nos. 4, 3, and 2) showed intermediate-to-low cytotoxicity,
and the last two samples had very little effect on AM. In fact,
the EC5Q values for fly ash no. 1 and the steel foundry particles
could be obtained only by extrapolation above the tested
concentration range for all three experimental parameters; in the
cases of fly ashes nos. 2 and 3, this was necessary only for
viability.
-------
436
CATHERINE ARANYI ET AL.
Table L. Concentration of Particles Required to Reduce Alveolar
Macrophage Viability, Total. Protein Content, and
ATP Levels to 50% (EC50)a

SC50 (yg/ml)

Sample

Viab i1ity
Total Protein ATP/Protein
Steel foundry

>1000b
>1000b
>1000b
part ic1es

Coal fly ash no.
1
>1000b
>1000b
>1000b
(convent ional)

Coal fly ash no.
2
>1000b
952
537
(fluid ized-bed)

Coal fly ash no.
3
>1000b
930
553
(convent ional)

Coal fly ash no.
4
949
856
445
(convent ional)

Aluminum smelter
dust
114
139
60
Copper smelter dust
11
6
5
aEstimated by lin
ear
regression analys
is for experimental

parameters expre
ssed
as: viability,
%; total protein (yg)
, Z of
control; and ATP
/protein (fg/yg), % o
f control.

^Highest concentr
ation tested.

The samples collected from the copper and the aluminum
smelters were not only much more toxic than the other samples, but
aLso behaved differently. At exposure concentracions between 250
and 1000 yg/ml (used for all other test particles), they produced
initial large decreases in the experimental parameters that
remained fairly constant over this entire range. Only at
concentrations below 250 yg/ml could a monotonic decrease in the
parameters be observed with increasing exposure concentration.
The high cytotoxicity of the copper and aluminum smelter
samples at low concentrations and the absence of a dose—dependent
response at higher concentrations, suggested that soLuble compounds
released continuously into the medium during the incubation period
produced these results, in addition to the particles per se. To
substantiate this hypothesis, the test particles were preincubated
without AM in the maintenance medium at the highest exposure
concentration (1000 yg/ml), under similar conditions (20 h at 37aC)
to those used in the cytotoxicity experiments. After incubation,
the particles were removed from the media by ultracentrifugation
-------
TOXICITY OF INDUSTRIAL PARTICLES
437
and Millipore filtration (0.22-um pore size). These filtered
culture media were subsequently incubated for 20 h at 37°C with AM,
and the viability of Che cells was compared with Chat of the
unexposed control AM. The viability of the cells exposed to the
supernatant fractions from the copper and aluminum smelter dust
samples was substantially Lower than that of the unexposed control
cells, demonstrating that soluble cytotoxic components were
released from these samples during incubation (see Table 2). No
such difference in viability was found for coal fly ash and steel
foundry particles, indicating that if any compounds were
solubilized from Che particles of Chese samples, the amounts were
not toxic to the AM.
Table 2. Examination of Test Particles for Soluble Cytotoxic
Components by Alveolar Macrophage Viability3
Sample
Viability (%)
Control
96.0 to 98.0
Steel foundry particles
97.5
Coal fly ash no, 1 (conventional)
96.3
Coal fly ash no. 2 (fluidized-bed)
97.3
Coal fly ash no. 3 (conventional)
96.7
Coal fly ash no. U (conventional)
97.0
Aluminum smelter dust
46.8
Copper smelter dust
65. 1
aDetermined after exposure at 37 C for 20 h to the particle-free
medium separated from the preincubated particles. For details,
see text .
Spark-source mass spectroscopic analysis by EPA of the copper
smelter particles showed that major trace metal constituents with
concentrations ranging up to 20% by weight were lead, arsenic,
copper, iron, antimony, and zinc, with the arsenic as high as 13%,
present mostly in water-soluble form. Thus, soluble arsenic could
be one of the components responsible for the cytotoxicity of the
copper smelter particles. No information is now available on the
solubility of the other trace metals in this sample, nor on similar
properties of the aluminum smelter dust.
Thus, the in vitro RAM test has enabled us to evaluate the
relative cytotoxicity of seven industrial particulate samples. The
data have demonstrated that five of these—a foundry particulate
-------
438
CATHERINE ARANYI ET AL.
and fly ash samples from three conventional combustion processes
and one fluidized-bed process—had low to intermediate effects on
AM. The samples collected from a copper and an aluminum smelter ,
however, were much more toxic than all others, and the copper
smelter sample was the more toxic of the two. In the case of the
two smelter samples, we found that in addition to the particles per
se, cytotoxic soluble components released during incubation
contributed to the total toxicity to AN.
In Vivo Tests
The next step, and a major objective of these studies, was to
confirm the in vitro evaluation by demonstrating parallel effects
in in vivo assessments of the inhalation hazard of these samples in
the intact animal. Female CD^ mice were exposed, as described, to
aerosols of the copper smelter dust and the fluidized-bed coal fly
ash (fly ash no. 2), particles that had shown very high and low
cytotoxicity, respectively, in vitro. The means and standard
deviations of aerosol mass concentration were 2033 ± 153 ug/m3 for
the copper smelter dust and 2043 ± 308 gg/m^ for the fly ash. 3oth
aerosols had log—normal size distributions, with count median
diameters and a^'s of 0.225 um and 2.9 for the copper smelter dusc
and 0.134 ym and 2.1 for the fly ash.
The effects of inhalation of the particles were evaluated
after 5, 10, and 20 exposures by examining changes in pulmonary
cellular lavage, in susceptibility to respiratory streptococcal
infection, and in pulmonary bactericidal activity to inhaled
radiolabeled K. pneumoniae. Results are sunrraarized in Figures 1,
2 , and 3.
Total cell counts and AT? levels (expressed as percent of the
control responses; see Figure 1) generally did not change
significantly or exhibit any trend related to the number of aerosol
exposures. Similarly, differential cell counts and percent
viability of the lavaged cells (not shown) were not affected by the
exposures. However, as seen in Figure 2, the percent mortality
after streptococcus inhalation challenge was greater in aerosol-
exposed than in control mice, for 5, 10, and 20 daily 3-h exposures
to copper smelter dust aerosols (2 mg/m^). No significant changes
were observed for any of the exposures to the coal fly ash. Mean
survival time (not shown in the figures) was significantly lower
than that of the control groups only following inhalation of the
copper smelter dust (i.e., treatments that also significantly
increased the mortality rate) . The percent of bactericidal
activity in response to inhaled radiolabeled K. pneumoniae was
significantly less for exposed than for control mice (Figure 3),
for 5, 10, or 20 daily 3-h exposures to the copper smelter dust
aerosol. Similar doses of the coal flv ash aerosol had no effect.
-------
TOXICITY OF INDUSTRIAL PARTICLES
439
" P <0 08
:numB«olTiC<
- - i?C
00
2C
2C
S
t.
0
20
:Q
10
s
0
Number cf 3->"»r Expgiur*)
Figure 1. Changes in pulmonary cellular lavage following multiple
daily 3-h aerosol exposures of mice Co 2 mg/m3 of copper
smelter dust (shaded bars) or fiuidized-bed coal fly ash
(unshaded bars).
Thus, these data clearly demonstrate that the copper smelter dust
was not only more cytotoxic in vitro to AM but also more
deleterious in inhalation exposures in vivo than was the
fluidized-bed coal fly ash.
CONCLUSIONS
In assays of the < 3 ym size fraction of a series of
stationary point-source samples by the SAM test, generally
low-to-intermediate cytotoxicity was found for samples collected
from a foundry and from several coal-fired power plants. However,
particulate samples from an aluminum and a copper smelter were
highly toxic to AM, as measured by viability and total cellular
-------
440
CATHERINE ARANYI ET AL.
ic-.
c
u
c
0
1 ?c-
>
t
"i
5
2
I
(L
A\.
• • . p< o 001
( I o^wic#
268'
: 315
sa
10 2C 4 10
Number ol Daily 3-1* iiooau'n
:o
Figure 2. Excess mortality from streptococcus aerosol infection in
exposed mice following multiple daily 3-h aerosol
exposures to 2 mg/ra3 of copper smelter dust (shaded
bars) or fluidized-bed coal flv ash (unshaded bars).
protein and ATP levels. In contrast to all others, the two smelter
samples also contained soluble components that contributed
substantially to their overall in vitro cytotoxicity.
The copper smelter particles and the fluidized-bed coal fly
ash, chosen on the basis of their respectively high and low in
vitro cytotoxicity, were used in aerosol exposures to examine their
effects in vivo on pulmonary free cells, bactericidal activity, and
resistance to respiratory infection in mice. The results obtained
after multiple daily 3-h exposures to 2 rag/m^ of these aerosols
correlated well with the in vitro data; inhalation of the aerosols
-------
TOXICITY OF INDUSTRIAL PARTICLES
441
c
0
(J
V
&
u
£
5
e
^ .10.
g
e
a
• • • P <0.001
( ) number of mica
' 5C-
I
&
-2C - „
20
Number of O aity 3-hr Expoaurm
Figure 3. Percent change in bactericidal activity in response to
inhaled JC. pneumoniae in exposed mice following multiple
daily 3-h aerosol exposures to 2 mg/ra^ of copper smelter
dust (shaded bars) or f1uidized-bed coal fly ash
(unshaded area).
-------
442
CATHERINE ARANYI ET AL.
of the copper smelter dust produced significant differences from
the controls in more of the experimental parameters than did
aerosols of the coal fly ash sample. Thus, the validity of
inhalation hazard prediction on the basis of an in vitro screening
assay has been demonstrated.
ACKNOWLEDGMENTS
These studies were supported by the U.S. Environmental
Protection Agency under grant no. R80514101/'02. The authors are
indebted to Mr. J. Hingeveld and Mr. W. O'Shea for their excellent
technical assistance. The critical review and comments of Dr. R.
Ehrlich and Dr. J. Fenters of IITRI and Dr. R. Merrill from
EPA/IERL are highly appreciated.
REFERENCES
Aranyi, C., S. Andres, R. Ehrlich, J.D. Fenters, D.E. Gardner,
and M.D. Waters. 1977. Cytotoxicity to alveolar macrophages
of metal oxides absorbed on fly ash. In: Pulmonary
Macrophage and Epithelial Cells (Conf-760972), C. Sanders,
R.P. Schneider, G.E. Dagle, and H.A. Ragan, eds. Energy
Research and Development Administration Symposium Series 43,
Technical Information Center, Energy Research and Development
Administration: Springfield, VA. pp. 58-65.
Aranyi, C., F.J. Miller, S. Andres, R. Ehrlich, J. Fenters, D.E.
Gardner, and M.D. Waters. 1979. Cytotoxicity to alveolar
macrophages of trace metals absorbed on fly ash. Environ.
Res. 20:14-23.
Ehrlich, R. 1966. Effect of nitrogen dioxide on resistance to
respiratory infection. Bacteriol. Rev. 30:604-613.
Gardner, D.E. 1979. Alteration in host-bacteria interaction by
environmental chemicals. In: Assessing Toxic Effects of
Environmental Pollutants. S.D. Lee and J. Bryan, eds. Ann
Arbor Science Publishers, Inc.: Ann Arbor, MI. pp. 87-103.
Green, G.M., and E. Goldstein. 1966. A method for quantitating
intrapulmonary bacterial inactivation in individual animals.
J. Lab. Clin. Med. 68:669-677.
Griest , W.H., and M.R. Guerin. 1979. Identification and
quantification of polynuclear organic matter on particulates
from coal-fired power" plant. EPRI EA-1092/DOE No. RTS 77-58
(Interim Report), Dept. of Energy and Electric Power Res.:
Palo Alto, CA.
-------
TOXICITY OF INDUSTRIAL PARTICLES
443
Mahar, H. 1976. Evaluation of Selected Methods for Chemical and
3iologicaL Testing of Industrial Particulate Emissions.
EPA 600/2-76-137. U.S. Environmental Protection Agency:
Research Triangle Park, NC.
Waters, M.D., D.E. Gardner, C. Aranyi, and D.L. Coffin. 1975a.
Metal toxicity for rabbit alveolar macrophages in vitro.
Environ. Res. 9:32-47.
Waters, M.D., D.E. Gardner, and D.L. Coffin. 1974a. Cytotoxic
effects of vanadium on rabbit alveolar macrophages in vitro.
Toxicol. Appl. Pharmacol. 28:253-263.
Waters, M.D., J.L. Huisingh, and N.E. Garrett. 1978. The cellular
toxicity of complex environmental mixtures. In: Application
of Short-terra Bioassays in the Fractionation and Analysis of
Complex Environmental Mixtures. M.D. Waters, S. Nesnow, J.L.
Huisingh, S. Sandhu, and L. Claxton, eds. Plenum Press: New
York. pp. 125-167.
Waters, M.D., T.O. Vaughan, J.A. Campbell, F.J. Miller, and D.L.
Coffin. 1974b. Screening studies on metallic salts using
the rabbit alveolar macrophage. In Vitro 10:342-343.
Waters, M.D., T.O. Vaughan, J.A. Campbell, A.G. Stead, and D.L.
Coffin. 1975b. Adenosine triphosphate concentration and
phagocytic activity in rabbit alveolar macrophages exposed to
divalent cations. J. Reticuoloendothel. Soc. 18:296.
-------
Intentionally Blank Page
-------
MUTAGENICITY AND CARCINOGENICITY OF A RECENTLY
CHARACTERIZED CARBON BLACK ADSORBATE: CYCLOPENTA(CD)
PYRENE
Avram Gold
Department of Environmental Sciences Engineering
University of North Carolina
Chapel Hill, North Carolina
Stephen Nesnow, Martha M. Moore, Helen Garland,
Gavnelle Curtis, Barry Howard, and Deloris Graham
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
Eric Eisenstadt
School of Public Health
Harvard University
Boston, Massachusetts
INTRODUCTION
Polycyclic aromatic hydrocarbons (PAH) are widespread
environmental contaminants that may be metabolically activated to
mutagenic or carcinogenic derivatives (Particulate Polycyclic
Organic Matter, 1972; Gelboin and Ts'o, 1978). Intensive research
indicates that PAH are prorautagens or procarcinogens containing the
bay region geometric feature (Jerina et al., 1978). Studies show
that the biological activity of these compounds results from
metabolism to bay region diol-epoxides, which are capable of
forming covalent adducts at nucleophilic sites within DNA.
CyclopentaCcd)pyrene (CPP, I)(see Figure l), a non-bay region
PAH, was recently characterized and shown to be highly mutagenic in
the Salmonella typhimurium assay (Eisenstadt and Gold, 1978). CPP
was initially identified in extracts of furnace black as the major
contributor to the high mutagenic activity of the extracts.
Because of its unique structure and wide environmental distribution
as a component of soots (Lee et al., 1977; Grimmer, 1977; Kaden et
al., 1979), carbon blacks (Wallcave et al., 1975; Gold, 1975), and
cigarette smoke (Snook et al., 1977), CPP has evoked considerable
interest (Ittah and Jerina, 1978; Konieczny and Harvey, 1979;
Ruehle et al . , 1979). The 3,^-oxide of CPP, predicted as an
ultimate mutagenic metabolite, has been synthesized and shown to be
a powerful direct-acting mutagen to S. typhimuriurn. Confirmation of
the 3,4-oxide as a primary metabolite and ultimate mutagen requires
that the expected enzymatic hydration product, trans CCP
3,4-dihydrodiol , be identified as a metabolic product.
445
-------

446
A VRAM GOLD ET AL.
3
2
4
I
Figure 1. Structure of cyclopenta(cd)pyrene (CPP, I).
Response of different assay systems to treatment with specific
mutagens may vary (Maner et al., 1978), and for this reason, the
mutagenicity of CPP and its 3,4-oxide in mammalian cells as well
as in microbial systems is of interest.
This study reports identification of trans-CPP 3 ,4-dihydrodiol
as the major CPP metabolite of both 3-methylcholanthrene- (3-MC)
and Aroclor-1254-induced rat liver microsomes. CPP and its
3,4-oxide were tested for mutagenicity in the L5178Y mouse lymphoma
system and for cell transformation in the C3H10T1/2CL8 mouse embryo
fibroblast system. The results are discussed in relation to the
proposal that CPP 3,4-oxide may be an ultimate mutagenic or
carcinogenic metabolite of CPP.
MATERIALS AND METHODS
Chemicals
CPP, CPP 3,4-oxide, and 4-oxo-CPP were obtained as previously
described (Gold et al. , 1973, 1979). Tritiaced CPP ([H3]CCP) was
prepared by catalytic exchange labeling (New England Nuclear,
Boston, MA) and was purified prior to use by chromatography over
silica.
Carbon Black Analysis
Semireinforcing furnace black was extracted with methylene
chloride (CH2CI2) and fractionated by standard procedures (Gold,
1975; Rosen and Middleton, 1955). A 120-mg aliquot of the PAH
-------
CYCLOPENTA(CD)PYRENE MUTAGENICITY AND CARCINOGENICITY 447
V
fraction was further separated on a size 3 Lo Bar silica column
(Merck) with an initial eluent of 4% CH2CI2 in hexane changed to
5% CH9CI2 in hexane after 500 ml. The separation was followed by
ultraviolet (UV) detection, and individual peaks were collected.
Ames assays were performed w
Ames et al. (1975). Dose-respons
4, 10, and 20 yg of test mixture
(revertants per microgram) was de
linear portion of the curve. Met
0.5 ml S-9 from Aroclor-treated r
ith strain TA100 as described by
e curves were obtained from 1, 2,
per plate, and mutagenic potency
termined from the slope of the
abolic activation was supplied by
at s .
Metabolism Studies
Rat liver microsomes and S-9 were prepared from Aroclor-
1254-treated male Sprague-Dawley rats weighing 150 to 200 g,
according to the procedures of Ames et al. (1975), except that the
microsomal pellet was resuspended in buffer and centrifuged a
second time at 100,000 x g for 60 min. Microsomes were stored
frozen at -80°C until used.
The concentrations of ingredients in the 1-ral microsomal
metabolism mixture were: 50 mM Tris-hydrochloride, pH 7.4; 3 mM
magnesium chloride (MgC^); 0.8 mM NADP; 5 mM glucose-6-phosphate:
0.4 units glucose-6-phosphate dehydrogenase: 0.89 mM [G_-3H]CPP
(specific activity, 5 x 101* dpra/nraol) and 0.2 mg microsomal
protein. The reaction was started by first adding [^HjCPP to the
microsomal mixture at 4"C and then shaking the mixture at 37°C for
the indicated times. One volume of acetone was added to stop the
reaction. Two volumes of ethyl acetate were added and the mixture
was shaken or vortexed vigorously. The ethyl acetate-acetone phase
was removed, dried with anhydrous sodium sulfate (Na2S0^) and
evaporated to dryness under nitrogen gas. The residue was
dissolved in a small volume of methanol and subjected to high
performance liquid chromatography (HPLC) analysis. Ninety to
ninety-five percent of the input radioactivity was recovered in the
organic extract.
Aroclor-1254- and 3-MC-induced enzymes produced similar
metabolite profiles (Figure 2), and preparative work was done with
3-MC-induced S-9 activation using 50- to 100-ral reaction volumes.
HPLC separations were performed on a Perkin-Elmer Series II
liquid chromatograph with a Perkin-Elmer 4.6-ram x 25-cm ODS-SilX-I
column or 2.6-mm x 25-cm ODS SilX-II column. A UV detector at
254 nm was used to monitor fractions for preparative work.
Fractions were collected (0.5 ml) and counted by liquid
scintillation for quantitative analysis of [3H]CPP metabolites.
-------
448
A
A VRAM GOLD ET AL.
T
0 4 8 12 min
Figure 2. HPLC trace at 254 nm of CPP metabolites from (A)
Aroclor-1254-induced rat liver microsomes (ODS SilX-II
column, 2.6 aim x 25 cm) and (B) 3-MC-induced rat liver
microsomes (ODS SilX-I column, 4.6 mm x 25 cm).
The same linear gradient was used for all separations: a 1%/min
gradient from 35:65 H2O:methanol to 100% methanol, at a flow rate
of 1.5 ml/min.
In Vitro Mammalian Mutagenesis Assay
The TK+//~L5178Y mouse lymphoma mutagen assay, developed by
Clive and co-workers to identify mutagens that induce genetic
damage at the thymidine kinase (TK) locus, was performed according
to published methods (.Clive et al . , 1979: Clive and Spector, 1 975 ),
using Fischer's medium. Trifluorothymidine (I ug/ml) was used to
select for thyraidine-kinase-deficient mutants. CPP, dissolved in
-------
CYCLOPENTA(CD)PYRENE MUTAGENICITY AND CARCINOGENICITY 449
Co published methods (Clive et al . , 1979: Clive and Spector, 1975),
using Fischer's medium. Trirluorothyraidine (' yg/ral) was used Co
selecc for thymidine-kinase-deficient mucants, CPP, dissolved in
dimethylsulfoxide (DMSO), was tested in 3% horse serum at a cell
concencraCion of 0.6 x 10^ cells/ml, wich and wiChouC an Aroclor-
1254-induced rat hepatic S-9 acCivation system. The CPP 3,4-oxide
(dissolved in DMSO) was tesCed in 10% horse serum at a cell
concentration of 0.6 x 10^ cells/ml without metabolic activation.
The positive controls ethyl raethanesulfonate (EMS) and
2-acetylaminofluorene (AAF) were dissolved in saline and DMSO,
respectively. Two days were allowed for the expression of newly
induced mutants.
Oncogenic Transformation Assay
The mouse embryo fibroblast cell line C3H10T1/2CL8 was
derived by Reznikoff et al. (1973; Nesnow and Heidelberger, 1976;
Gehly et al., 1979) and donated by Dr. Charles Heidelberger for use
in chese experimenCs. Cell cultures were incubated at 37°C in
humidified incubators with an atmosphere of 5% carbon dioxide in
air. All cultures were grown in Eagle's Basal Medium with Earle's
salts and L-glutamine supplemented with 10% heat-inactivated fetal
calf serum (Grand Island Biological Co.). The cells were routinely
checked for Mycoplasma contamination and found to be Mvcoplasma-
f ree .
For transformation, cells were seeded onto 60-tnm petri dishes
(1000/dish) in 5 ml of medium (12 replicates/treatment) and 24 h
later treated with the hydrocarbon dissolved in acetone (25 yl) .
After an additional 24 h, the medium was removed, and the cells
received fresh complete medium containing penicillin (100 units/tnl)
and streptomycin (50 yg/ml). Medium was changed weekly until the
cells reached confluency, whereupon the fetal calf serum
concentration was reduced to 5%.
At the end of six weeks, the dishes were washed with 0.9%
sodium chloride solution, fixed with methanol, stained with Giemsa,
and scored for oncogenic transformation. Three different types of
foci have been described after treatment of C3H10T1/2CL8 cells with
polycvclic hydrocarbons. Only Type II and Type III foci were
scored in this assay, since it has been demonstrated that these
foci produce fibrosarcomas upon injection into irradiated C3H mice,
33% and 67% of the time, respectively.
Control cultures received the appropriate solvent and were
treated the same way as the exposed cells. Cytotoxicity assays
were performed concurrently with the transformation assays, and
using the same protocol, except that the dishes were plated with
200 cells (6 replicates/treatment) and stained 10 to 12 days later.
-------
450
AVRAM GOLD ET AL.
RESULTS AND DISCUSSION
A fractionation scheme was applied to the organic extract of
serairein foreing furnace black used in the rubber industry. The
Ames test was used in conduction with the chemical separation Co
identify mutagens. The data shown in Table 1 indicate that the PAH
fraction of the CH2CI2 extract of the semireinforcing furnace black
accounted for essentially all of the mutagenicity in the total
extract. Table 2 shows the mutagenic activity of the fractions
obtained on further chromatographic resolution of the PAH fraction.
Specific activity of the total PAH fraction calculated from the
activities of the fractions resolved in Table 2 appeared to be
about the same as that observed for the total PAH fraction (see
Table 1). A similar observation has been reported for kerosene
soot (Kaden et al . , 1979). It is apparent that benzo(a)pyrene
(B[a]P) contributed only slightly to the mutagenic activity of the
PAH fraction and therefore to the mutagenic activity of the total
organic extract, while CPP was the compound principally responsible
for the activity. An estimate of mutagenicity based on B(a)p
content would have seriously underestimated the mutagenicity of Che
sample. This result underscores the inherent danger in using a
marker compound such as B(a)P as an index of hazard even for
samples with similar pyrogenic origins like soots and carbon
blacks .
Table 1. Mutagenic Activity of Fractionated Carbon Black
Extracts in Salmonella Strain TA100
Weight Activity
Fraction (rag) (rev/pg)
Neutral
saturated hydrocarbon
polycyclic aromatic
pol ar
Ac id ic
Basic
Total Extract*5
aNonlinear dose-response curve.
^From 314 g furnace black.
9 not active
135 38
27 4
49 0.5a
6 not active
346 14
-------
Tabic 2. Distribution of Mutagenic Activity in Aromatic traction
of Furnace Black Extract with Ames Strain TA10D
Fract ion
Numbe r
Compos it ion
Fr ac t ion
Weight
(mg)
naph t ha I ene
ac en a pht h y lene 10.4
phenanthreiie
pyrene 70
f I no r an tliene 30
ben£o(ghi)fLuorantbenea
eye Iopent a(cd)pyrene
benzo(ghi)fIuoranthone 8.4
c yr. I open t a(cd ) pyrenea
eye Iopenta(cd)pyrene 2
isomers of inolee. wt . 252
benzo(a)pyrene
benzo(e)pyrene
benzo(gbi)perylene 4.6
isomer a of inolee. wt . 276 and 300
Ac t i v i t y
(rev/ug)
29
(.'out r i hut i on to Total
Polycyclic Aromatic Activity
(% wt. oE Polycyclic Aromatic
Fraction x Activity)
not active
not active
93
380
90
Calculated activity of total polycyclic aromatic fraction (rev/yg)
aMnjor compononL.
21
24
47
n
k1,
n
r
O
"0
Pi
y
H
>
O
a
T3
~<
Pd
m
2
m
SI
d
H
>
O
M
X
>—I
o
t—t
H
»-<
£
a
n
>
to
n
'7,
O
o
pi
VJ
n
H
a*
-------
452
A VRAM GOLD ET AL.
Structure-reactivity relationships for CPP were of
considerable interest because of its high mutagenicity and unusual
structure—containing a fused cyclopenteno ring and lacking a bay
region. Hence, an investigation of the metabolism, mutagenicity,
and carcinogenicity o£ CPP was undertaken.
To identify the structure of the metabolic products of CPP,
synthetic metabolites were produced by acid-catalyzed decomposition
of the 3,4-oxide. This procedure cleanly yielded three products
(Figure 3). The two most polar peaks were identified as the
3,4-dihydrodiol isomers based on the mass spectrum and TJV spectrum
of the mixture. As a result of the rigid planarity of the fused
five-merabered ring in CPP, the trans diol was expected to be the
less polar isomer, because the dipole moments of the hydroxy
substituents are largely apposing. The rigid geometry also led to
the prediction of greater shielding for the C3 and Cu protons of
the trans isomer in the nuclear magnetic resonance (NMR) spectrum.
The major diol resulting from hydration was the less polar isomer
(peak B, Figure 3) and had the more shielded C3 and C^ protons in
the NMR spectrum (Figure 4a). On this basis, it was assigned the
trans configuration. This assignment was consistant with the
expectation that the major isomer of acid-catalyzed epoxide
hydration would be the trans isomer (Bruice et al., 1976; Keller
and Heidelberger, 1976).
Based on its physical-chemical properties, the third and
major product of the acid catalyzed reaction was identified as
4-oxo CPP. As illustrated in Figure 5, it was readily
distinguished by UV spectroscopy from the known 3-oxo compound
(Gold et al., 1978) .
The HPLC chromatograms (Figure 2) of the metabolic mixtures
produced by 3-MC- and Aroclor-1254-induced rat liver microsomes
indicated a single major metabolite accounting for 507, of the
products. This metabolite was identified as trans-CPP
3,4-dihydrodiol. Its chromatographic retention time and NMR
spectrum (Figure 4b) corresponded to those of the trans isomer from
the hydration of the 3,4-oxide. Also, its UV and mass spectrum
were consistent with the assigned, structure.
The TK+/~ L5178Y mouse lymphoma assay has been used to measure
the mutagenicity of diverse chemical agents (Clive et al., 1979).
Since these cells lack the enzymes necessary to activate
promutagens, CPP was tested both with and without S-9 activation,
to confirm the activation-dependence of mutagenicity typical of
PAH. As expected, CPP was not mutagenic without activation over
the concentrations tested (0.75 to 30 ug/ml)(unpublished data).
On activation by Aroclor-1254-induced hepatic S-9, CPP was
mutagenic, with twice the mutation frequency of the control (Table
3). The 3,4-oxide was tested without activation to determine
-------
CYCLOPENTA(CD)PYRENE MUTAGENICITY AND CARCINOGENICITY
453
OD
254 nr
20
M1N
16
12
Figure 3. HPLC trace at 254 nm of acid catalyzed decomposition
products of CPP 3,4-oxide, peak A, cis-CPP
3,4-dihydrodiol; peak 3, trans-CPP 3,4-dihydrodiol;
peak C, 4-oxo CPP.
whether ic was a direct-acting (and a possible ultimate) mutagen.
It was found to be mutagenic to L5178Y cells over a dose range
similar to that of CPP. Over the 70 to 20% survival range, the
oxide was two- to six-fold more mutagenic than the parent
hydrocarbons.
The C3H10T1/2CL8 transformation assay responds to a variety of
carcinogens, including PAH (Nesnow and Heidelberger, 1976).
C3H10T1/2CL8 mouse embryo cells contain cytochrome P-450 mixed
function oxidase, epoxide hydrase, and conjugating enzymes
necessary to metabolize and activate or detoxify chemical
carcinogens, especially PAH (Gehly et al, 1979; Nesnow and
Heidelberger, 1976). As shown in Table 2, CPP produced a dose-
related response in the formation of both Type II and Type III
transformed foci. At the highest dose used, every plate contained
at least one Type III focus . In accord with the recent report
(Wood et al., 1980) that CPP is a weaker tumor initiator than
B(a)P, the data in Table 4 indicate that CPP was also less active
than B(a)P in the C3H10T1/2CL8 trans format ion assay.
-------
454
AVRAM GOLD ET AL.
68.17
S5 68
65.65
S5.60
£572
Figure 4. In part a, 270 MHz NMR (acetone-d^) of cis-trans-CPP
3 ,4-dihydrodiol mixture. In part b, 270 MHz NMR
(acetone-d^) of major CPP metabolite. Underlined
resonances are consistent with CPP 9.0-dihydrodiol
cochroraatographing with the 3,4-dihydrodiol.
oo
30O
JSO
400
nm
Figure 5. UV-VIS spectrum (CH2CI2) of 4-oxo and 3-oxo CPP.
-------
CYCLOPENTA(CD)PYRENE MUTAGENICITY AND CARCINOGENICITY 455
Table 3. Mutagenesis of L5178Y House Lymphoma Cells
by CPP and CPP 3,4-Oxide
Concentrat ion
( ug/ml)
Total
Vi able
Clones
Total
Mutant
Clones
Total
Survival3
(% of Control)
Mut an t
Frequency
(x 10e)
CPP (with S-9)

i .20
439
263
110
120
1 .30
365
317
74
172
1 .40
436
300
56
138
1 .50
407
284
44
139
1 .60
399
375
20
188
2-AAF

30
270
785
31
581
DMSO (1%)
453
178
100
79
CPP 3,4-oxide

0.60
536
783
71
292
0 .70
472
803
63
340
0 .84
536
912
72
340
0 .96
463.
776
58
335
1 .08
475
1234
52
519
1 .20
476
1065
44
448
1 .32
406
1174
32
578
1 .44
341
1090
22
640
1 .56
360
1423
23
790
1 .68
459
1118
36
487
1 .80
273
1204
10
882
2.04
333b
1214
6
1460
EMS

500
259
1638
25
1267
DMSO (1%)
623
297
100
95
Untreated Control 558
220
100
79
aSurvival calculations described by Clive and Spector (1975)
combine both relative growth in suspension and relative plating
e ffic ienc y.
^Cells cloned at a density of 12 cells/tnl; all other cultures
cloned at 6 cells/tnl to determine viability.
-------
UI
cr>
Concent rat ion Plating No. Type IF Foci No. Typo III Foci Dishos with Type II
(Mg/"il) Efficiency (%) /Total Dishes /Total Dishes and III Foci (%)
CPP
0.01
0.03
10.0
CPP 3,4-oxide
0 .001
0.003
0 .01
0 .03
0.1
0.3
1 .0
3.0
Acetone (0.5%)a
B(a)Pa
(1)
27
28
26
27
29
29
22
32
3 L
30
32
32
32
31
27
30
1 8
0/12
1/12
0/12
3/12
4/1 I
6/12
13/12
2/12
l/ll
0/12
l/ll
1/12
0/12
0/12
1/12
2/24
38/24
2/12
1/12
1/12
2/12
7/H
7/12
12/12
0/12
0/11
0/12
0/11
0/12
2/12
2112
2/12
0/24
59/24
11
17
8
25
73
7b
100
8
9
0
9
8
17
I 7
25
8
100
aKe.sulLsj from two separate experiments combined
>
<
>
V
O
O
1-
D
tn
H
>
1-
-------
CYCLOPENTA(CD)PYRENE MUTAGENICITY AND CARCINOGENICITY 457
CPP 3,4-oxide transformed C3H10T1/2CL8 cells at concentrations
of 0.3 ug/ml. At 3 pg/ral, it produced both Type II and Type III
foci, with virtually no toxicity. The lack of toxicity is
remarkable, because structurally related K-region oxides of B(a)P
and 3-MC were highly cytotoxic (unpublished data). Although the
direct-acting oxide was a more potent mutagen than CPP, it was less
active than the parent hydrocarbon in the C3H10T1/2CL8
transformation assay. This was probably due to the C3H10T1/2CL8
cells' ability to detoxify arene oxides to dihydrodiols (Nesnow and
Heidelberger, 1976; Gehly et al., 1979). L5178Y mouse lymphoma
cells lack this ability (Clive et al., 1979).
The potent direct-acting mutagenicity of the 3,4-oxide in both
bacterial and mammalian assays, its ability to transform
C3H10T1/2CL8 cells, and the identification of CPP 3,4-dihydrodiol
as a major metabolite (Gold et al., 1979) in 3-MC- and Aroclor-
induced metabolism of CPP are strong evidence that CPP 3,4-oxide is
both a primary product of enzymatic oxidation and an ultimate
mutagen or carcinogen. Further support for this conclusion is
found in the report that the addition of epoxide hydrase to a
purified, reconstituted tnonooxygenase activating system drastically
reduces the mutagenicity of CPP (Wood et al., 1980). The effect of
epoxide hydrase can readily be explained if CPP 3,4-oxide is the
ultimate mutagen and, like other arene oxides (Gelboin and Ts'o,
1978), is a good substrate for epoxide hydrase. Since the B(a)P
bay region diol-epoxides are poor substrates for epoxide hydrase
(Gelboin and Ts'o, 1978), CPP would be less mutagenic than B(a)P in
the C3H10T1/2CL8 system due to the more rapid deactivation of
metabolically-generated CPP 3,4-oxide.
An important distinction between B(a)P diol-epoxides and CPP
3,4-oxide is that the latter is an arene oxide: the epoxidized
bond is adjacent to the aromatic nucleus at both termini. The
arene oxides tested, 3-MC 11,12-oxide and B(a)P 4,5-oxide, both
failed to transform C3H10T1/2CL8 cells (unpublished data). CPP is
the first arene oxide reported to transform this cell type.
REFERENCES
Ames, B.N., J. McCann, and E. Yamasaki. 1975. Methods of
detecting carcinogens and mutagens with the Salmonella/
mammalian-microsorae mutagenicity test. Mutation Res.
31:347-364.
-------
458
A VRAM GOLD ET AL.
Bruice, P.Y., T.C. Bruice, P.M. Dansette, H.G. Selander, H. Yagi,
and D.M. Jerina. 1976. Comparison of the mechanisms of
solvolysis and rearrangement of K-region vs. non-K-region
arene oxides of phenanthrene; comparison solvolytic rate
constants of K-region and non-K-region arene oxides. J. Am.
Chem. Soc. 98:2965-2973.
Glive, D., K.O. Johnson, J.F.S, Spector, A.G. Batson, and M.M.M.
Brown, 1979. Validation and characterization of the
L5178/TK+/~ mouse lymphoma mutagen assay system. Mutation
Res . 59:61-108.
Clive, D., and J.F.S. Spector. 1975. Laboratory procedures for
assessing specific locus mutations at the TK locus in cultures
L5178Y mouse lymphoma cells. Mutation Res. 31:17-29.
Eisenstadt, E., and A. Gold. 1978. CyclopentaCcd)pyrene: a
highly mutagenic polycyclic aromatic hydrocarbon. Proc. Nat.
Acad. Sci. USA 75:1667-1669.
Gehly, E.B., W.E. Fahl, C.R. Jefcoate, and C. Heidelberger. 1979.
The metabolism of benzo(a)pyrene by cytochrome P450 in
transformable and non-transforraable C3H mouse fibroblasts.
J. Biol. Chem. 254:5041-5048.
Gelboin, H.V., and P.O.P. Ts'o, eds. Polycyclic Hydrocarbons
and Cancer, Vols. 1 and 2. Academic Press: New York,
Gold, A. 1975. Carbon black adsorbate: separation and
identification of a carcinogen and some oxygenated
polyaromatics, Anal. Chem. 47:1469-1472.
Gold, A., J. Brewster, and E. Eisenstadt. 1979. Synthesis of
cyclopenta( cd)pyrene-3,4-epoxide, the ultimate mutagenic
metabolite of the environmental carcinogen
eyelopenta(cd)pyrene. J. Chem. Soc. Chem. Coramun. 903-904.
Gold, A., J. Schultz, and E. Eisenstadt. 1978. Relative
reactivities of pyrene ring positions: cyclopenta(cd)pyrene
via an intramolecular Friedel-Crafts acylation. Tetrahedron
Lett . 4491-4494.
Gold, A., J. Schultz, and E. Eisenstadt. 1979. Synthesis and
metabolism of eyelopenta(cd)pyrene. In: Polynuclear Aromatic
Hydrocarbons. P.W. Jones and P. Leber, eds. Ann Arbor
Science: Ann Arbor, MI. pp. 695-704.
Grimmer, G. 1977. IARC monographs on the evaluation of
carcinogenic risk of chemicals to man, Vol. 16. International
Agency for Research on Cancer, Lyon, France. p. 29.
-------
CYCLOPENTA(CD)PYRENE MUTAGENICITY AND CARCINOGENICITY 459
Ittah, Y., and D.M. Jerina. 1978. Synthesis of
cyc1 openta(cd)pyrene. Tetrahedron Lett. 4495-4498.
Jerina, D.M., H. Yagi, R.E. Lehr, D.R. Thakker, M. Schaeffer-
Ridder, J.M. Karle, W. Levin, A.W. Wood, R.L. Chang, and A.H.
Conney. 1978. Bay region theory of carcinogenesis by
polycyclic aromatic hydrocarbons. In: Polycyclic
Hydrocarbons and Cancer, Vol. 1. H.V. Gelboin and P.O.P.
Ts'o, eds. Academic Press: New York. pp. 173-188.
Kaden, D.A., R.A. Hites, and W.G. Thilly. 1979. Mutagenicity of
soot and associated polycyclic aromatic hydrocarbons to
Salmonella typhimurium. Cancer Res. 39:4152-4159.
Keller, J.W., and C. Heidelberger. 1976. Polycyclic K-region
arene oxides; products and kinetics of solvolysis. J. Am.
Chem. Soc. 98:2328-2336.
Konieczny, M., and R.G. Harvey. 1979. Synthesis of
cyclopenta(cd)pyrene. J. Org. Chem. 44:2158-2160.
Lee, M.L., P.G. Prado, J.B. Howard, and R.A. Hites. 1977. Source
identification of urban airborne polycyclic aromatic
hydrocarbons by gas chromatographic mass spectrometry and
high resolution mass spectrometry. Biomed. Mass Spectrom.
4:182-186.
Maher, V.M., and J.M. McCormick. 1978. Mammalian cell mutagenesis
by polycyclic aromatic hydrocarbons and their derivatives.
In: Polycyclic Hydrocarbons and Cancer, Vol. 2. H.V. Gelboin
and P.O.P. Ts'o, eds. Academic Press: New York,
pp. 137-160.
Nesnow, S., and C. Heidelberger. 1976. The effect of modifiers
of microsomal enzymes on chemical oncogenesis in cultures of
C3H mouse cell lines. Cancer Res. 36:1801-1080.
Particulate Polycyclic Organic Matter. 1972. National Academy of
Sciences: Washington, DC.
Reznikoff, C.A., J.S. Bertram, D.W. Brankow, and C. Heidelberger.
1973. Quantitative and qualitative studies of chemical
transformation of cloned C3H mouse embryo cells sensitive to
postconfluence inhibition of cell division. Cancer Res.
33:3239-3249.
Rosen, A.A., and F.M. Middleton. 1955. Identification of
petroleum refinery wastes in surface waters. Anal. Chem.
27:790-794.
-------
460
A VRAM GOLD ET AL.
Ruehle, P.H., D.L. Fischer, and J.C. Wiley. 1979. Synthesis of
cyclopenta(cd)pyrene, a ubiquitous environmental carcinogen.
J. Chem. Soc. Chem. Coramun. 302-303.
Snook, M.E., R.E. Severson, R.F. Arrendale, H.C. Highman, and O.T.
Chortyk. 1977. The identification of high molecular weight
polynuclear aromatic hydrocarbons in a biologically active
fraction of cigarette smoke condensate. Beitrage zur
Tabakforsch. 9:97-101.
Wallcave, L. , D.L. Nagel, J.W. Smith, and R.D. Waniska. 1975.
Two pyrene derivatives of widespread environmental
distribution: cyclopenta(cd)pyrene and acepyrene. Environ.
Sci. Technol. 9:143-145.
Wood, A.N., W. Levin, R.L. Chang, M. Huang, D.E. Ryan, P.E. Thomas,
R.E. Lehy, S. Kumer, M. Koreeda, H. Akagi, Y. Ittah, P.M.
Dansette, H. Yagi, D.M. Jerina, and A.H. Conney. 1980.
Mutagenicity and tumor-initiating activity of
cyclopenta(cd)pyrene and structurally related compounds.
Cancer Res. 40:642-649.
-------
MUTAGENICITY OF COAL GASIFICATION AND LIQUEFACTION
PRODUCTS
Rita Schoeny, David Warshawsky, Lois Hollingsworth, and
Mary Hund
Department of Environmental Health
University of Cincinnati College of Medicine
Cincinnati, Ohio
George Moore
Pittsburgh Energy Technology Center
U.S. Department of Energy
Pittsburgh, Pennyslvania
INTRODUCTION
As it becomes evident that the shortage of oil resources is
not a transient phenomenon, emphasis is being directed to the use
of domestic coal. Increased use of coal, however, adds coal
combustion products to the pollution burden. Although various
technologies are being developed to produce cleaner-burning fuels
from coal, such as gaseous fuels, de-ashed low-sulfur boiler fuels,
and synthetic crude oils, problems remain in the production of
these materials.
The production of liquid fuels from coal has been associated
with an increased risk of cancer, and certain of these coal-derived
liquids have been shown to be carcinogenic in experimental animals
(Bingham, 1975; Ketchara et al., 1960; Sexton, 1960a, b; Weil et
al . , 1960). Heavy exposure to coal hydrogenation materials causes
both benign and malignant skin tumors. Composition data
(Swansiger, 1974; Battelle, 1974; ORNL, 1975; Electric Power Res.
Inst., 1975; ERDA, 1976) suggest that various liquefaction products
and by-products are likely to contain polycyclic substances of
considerable carcinogenic potential; these compounds are most
likely to be found in the high-boiling-point aromatic fractions of
the product liquids (TWR, 1976). The total products of
hydrogenation, high-boi1ing-point distillates, centrifuged oils,
char, residues, recycled solvent oil, recycled solvent, and liquid
coal are all potentially hazardous materials (Freudenthal et al . ,
1975) .
461
-------
462
RITA SCHOENY ET AL.
Synthetic natural gas is not expected to pose a carcinogenic
risk, as any trace elements, organic carcinogens, or cocarcinogens
present in the raw product will have been removed during clean-up
and scrubbing operations. Rather than the fuel produced, it is the
coal gasification process itself that should be the primary concern
(Kornreich, 1976). The greatest potential hazard in coal
gasification is in the early stages of the processes, in which coal
goes through a series of structural degradations of complex organic
compounds. During these early stages, leaks and spills should
contain hazardous material (Freudenthal et al., 1975). Potentially
carcinogenic polycyclic organic material is likely to concentrate
in the tars, oils, and char (Kornreich, 1976; TRW, 1976). The
crude gas, if it contains tars of high-boiling-point oils, must
also be considered a potential hazard.
Hazardous chemicals may be synthesized whenever coal is
subjected to severe conditions such of those of pyrolysis,
hydrogenation, or gasification. As use of synthetic fuels will
probably increase, sensitive and rapid in vitro studies should be
carried out on products, intermediate streams, and wastes of coal
conversion processes to determine potential hazards. A variety of
techniques are available for assessing the environmental risks and
potential health effects of coal processing technologies. These
techniques include chemical and physical characterizations,
microbial assays, and tests for acute toxicity and irritation,
subchronic toxicity and teratology, chronic toxicity, and
carcinogenesis, among others. To conduct so many tests on all new
technological developments would not be appropriate, as each change
of experimental condition or operation would lead to new health
study requirements. The cost would be prohibitive, and much of the
resulting data useless, as it would apply to defunct processes.
Delaying biological assessment until the process is ready for
production would likewise be inappropriate. An acceptable position
on testing new fossil energy processes must be found. A suggested
compromise would involve chemical characterizations and rapid
bioassay studies in small-scale developmental programs followed by
detailed characterization and short-term and long-term
toxicological testing programs on pilot processes.
For this project, materials ranging from solid residue to
liquid products and waters, produced through advanced coal-
conversion technologies (including gasification and liquefaction),
have been selected and screened using the Salmonella/microsomal
mutagenesis assay. This assay, developed by 3.N. Ames et al.
(1975), is recognized as one of the most useful short-term assays
for mutagenesis, based on the number of compounds it detects as
mutagens and on its high correlative of positive responses with
known carcinogens (Bridges, 1976; McCann et al., 1975; KcCann and
Ames, 1976; Purchase et al., 1976; Simmon 1979; Committee 17,
1975). It is being used to investigate the health effects of
-------
COAL GASIFICATION AND LIQUEFACTION PRODUCTS
463
compounds already in Che environment and of materials under
development. It is particularly useful in the evaluation of
mixtures of substances, such as cigarette smoke concentrate (Kier
et al . , 1974; Sato et al., L 9 7 7 ) , synthetic crude oil (Epler, 1978;
Epler et al., in press), and organic extracts of drinking water
(Loper et al., 1978) .
MATERIALS AND METHODS
Sample Preparation
Coal-related materials were provided by the Department of
Energy. These samples are not considered to be process discharges
and may not be identical to materials eventually generated from
advanced coal processes developed for commercial use. Although the
samples are not representative of all materials derived from
advanced coal processes, they were selected for study because of
their immediate availability and the desirability of testing
materials of widely differing properties.
All samples were stored at 5°C. Materials ETTM-01, ETTM-02,
ETTM-08, and ETTM-09, which were tars or viscous liquid, were
prepared for testing by weighing a small amount (20 to 70 mg) and
adding dimethylsulfoxide (DMSO) to obtain a presumptive
concentration of 10 rag/ml. In no case did all the material
dissolve. The amount of insoluble sample was subtracted from the
total to give the concentration used in calculating the mutagenic
doses. The sample solutions were filter-sterilized prior to
testing. All samples were applied in 0.1-ml aliquots. The sample
solutions were further diluted in DMSO, so that the following
percentages of the sample solution were assayed: 100%, 50%, and
0.5%, or 100%, 50%, 10%, 5%, and 1%. For ETTM-02, the liquid
fraction was assayed by applying 0.1 ml of the undiluted substance,
as well as the concentrations listed above.
The powdered samples ETTM-03 and ETTM-04 did not dissolve in
DMSO, nor were mutagenic substances in detectable quantities washed
from them into the DMSO. Aqueous leachates were made from these
samples. Five grams of the powder were added to 45 g distilled
water at pH 3.0, 5.0, 6.0, 7.0, and 10.0. These suspensions were
stoppered and shaken at room temperature overnight. Each
suspension was centrifuged, and the supernatant fraction was
aspirated and filter-sterilized for use in the mutagenic assays.
Liquids ETTM-06 and ETTM-07 changed from yellow to turquoise
when mixed with DMSO, and ETTM-06 underwent an exothermic reaction.
These two samples were also diluted in double-distilled water. The
black particulate matter floating on sample ETTM-06 was removed
-------
464
RITA SCHOENY ET AL.
during fi1ter-steri1ization, and Che solvent and sample phases were
mixed vigorously iramediately prior Co assay.
Organic extracts of ETTM-01, ETTM-02, ETTM-08, and ETTM-09
were prepared as follows: Approximately 5 g or less of each sample
was weighed, and a volume (in ml) of solvent equal Co five times
the sample weight (in g) was added. This mixture was agicated
vigorously in the dark at room temperature for two hours. After
centrifugation to settle particulates, the solvent was removed, and
an equal amount of fresh solvent was added. The two extracts were
pooled and evaporated under nitrogen gas. This procedure was
carried out sequentially with hexane, toluene, methylene chloride,
and acetonitrile. This simple organic extraction procedure was
chosen for its ease and appropriateness for these samples. The
type of organic extracts produced, moreover, is suitable for
analysis by high performance liquid chromatography.
Mutagenicity Assays
Salmone11 a/microsomal assays were carried out according to the
methods described by Araes et al. ( 1975). Microsomal extracts for
routine assays (S-9) were made from livers of male Sprague-Dawley
rats (150 to 200 g body weight) that had been administered 500
rag/kg Aroclor 1254 (?CB) on day one and killed on day six. On the
day of assay, S-9 from four animals was pooled. Assays were also
done using S-9 from rats treated with 3-methylcholanthrene (3-MC,
40 mg/kg) and from rats given corn oil. Plates were scored using
an automatic colony counter. Only counts of at least twice the
spontaneous values were considered to indicate mutagenicity.
Colony counts below the range observed for spontaneous revertants,
clearing of bacterial lawns, or the appearance of pinpoint his"
colonies was scored as a toxic response.
RESULTS AND DISCUSSION
Table 1 summarizes the mutagenic activity of the samples. No
sample was direct acting (i.e., mutagenic in the absence of S-9).
The liquefaction products, distillate oils, heavy liquid, and
residue were mutagenic. Of the gasification products, only the
tar was active in these assays. Neither DMSO extracts nor aqueous
leachates of gasification particulate and ash were mutagenic.
Figure 1 is a plot of dose-response data for ETTM-01. This
plot, typical of those for the other active samples, indicates a
linear dose response at the lowest concentrations tested, with
decreases in the slopes of the curves at the higher concentrations.
-------
COAL GASIFICATION AND LIQUEFACTION PRODUCTS
465
Table 1. Mutagenicity of Diraethylsulfoxide-soluble Components
o f Coal-related Materials
Mutagenicity3
Sample TA1535 TA1538 TA98 ' TA100
ETTM-01
Liquefaction product
_
+
4-

ETTM-02
Gasification tar
-
~
+
+
ETTM-03
Gasification particulate
ND
ND
-
-
ETTM-04
Gasification ash
ND
ND
-

ETTM-06
Liquefaction untreated
water
ND
ND
—

ETTM-07
Liquefaction light oils
ND
ND

-
ETTM-08
Liquefaction heavy liquid
(with solids)
ND
ND
~
4-
ETTM-09
Liquefaction product
(filtered)
ND
ND
4-
4-
ETTM-10
Liquefaction distillate
oils
""
+
+
~
ETTM-11
Liquefaction residual

•f

~
a+ = mutagenic; - = not mutagenic; ND = not determined.
As these samples are mixtures, there are several possible
explanations for this pattern:
1) The samples may be toxic to the organisms at the higher
doses. When these concentrations of sample were tested
without S-9, there was apparent toxicity.
2) At the high concentrations, activation enzvraes may be
saturated, or metabolism channelled into detoxification
pathways.
3) Components in the mixture may interact, contributing to
nonlinear kinetics.
Revertant colonies per microgram of sample, derived through
regression analysis of the linear portion of the dose response, are
given in Table 2. Strains TA98 and TA1538 (frameshift mutants with
and without a misrepair-enhancing plasmid) were the most sensitive
to the mutagenic action of the samples. The amount of Aroclor-
induced S-9 routinely used in the assay was 50 ul/plate. Figure 2
is a representative plot of rautagnenicity in response to varying
-------
466
RITA SCHOENY ET AL.
36C0
w 2400
<
Q.
O
o
TA98 •
TA100 a.
20C0
100 200 300 400 500 600 7CC
SAMPLE CONCENTRATION Ug/plate)
SCO
Figure 1
Mutagenicity of ETTM-01 in Salmonella with Aroclor-
induced S-9 (50 gl/plate) . Correlation coefficients (r)
for the linear regression lines up to 39.3 gg of sample
are as follows: strain TA98, 0.982; TA100, 0.971.
For concentrations above 39.3 pg/plate, lines are drawn
through the means of the data points.
amounts of S-9. For all the mutagenic samples, 50 pi of S-9
provided optimal or nearly optimal activation for mutagenesis. The
active samples were also assayed for mutagenicity in the presence
of 50 gl/plate uninduced S-9 and 50 gl/plate 3-MC-induced S-9.
Uninduced S-9 was uniformly poor for sample activation. For all
but one sample (ETTM-01; see Figure 3), Aroclor-induced S-9 was
most effective in generating metabolites mutagenic for TA98 and
TA100.
-------
COAL GASIFICATION AND LIQUEFACTION PRODUCTS 467
Table 2. Relative Mutagenic Activities of Coal-related Materials
A.

Revertant
Colonies/ gga

Sample
TA1535
TA1538
TA98
TA100
ETTM-01

30.12
18.54
6.89
ETTM-02

-
11 . 16
6.75
6 .49
ETTM-08

ND
ND
16 .66
8.38
ETTM-09

ND
ND
11 .55
6.76
ETTM-10

-
7.30
10.88
2.10
ETTM-11

27.92
26 .01
11 .28
B .

Coefficients of
Linear Correlation (R)

TA1358

TA98
TA100
Satnpl e
Nb
R
N
R
N R
ETTM-01
16
0.921
32
0.888
59 0.918
ETTM-02
6
0.999
46
0.918
45 0.857
ETTM-08

-
60
0.960
57 0.843
ETTM-09

-
39
0.948
54 0.887
ETTM-10
18
0.987
64
0.905
78 0.911
ETTM-11
28
0.912
64
0.933
55 0.950
Calculated
from linear
post ions
of dose-
¦response curves: - - no
dose response; ND ¦ not
determined.

bN ¦ number
of data points.

Sequential organic extracts were prepared on two occasions
from each of the six mutagenic samples. Whenever possible,
five-point dose-response assays were done with strains TA98 and
TA100. For comparison, a dose-response assay of the unfractionated
whole sample was run in tandem with these assays. Representative
results are presented in Tables 3 and 4 and Figures 4 and 5. No
extracted material was mutagenic in the absence of S-9. The
samples varied widely in the percentage extractabla by the solvents
and in the mutagenicity of the extracted fractions. For the
majority of samples, the toluene-extractable components were most
mutagenic, although they were not always the largest fractions by
-------
468
RITA SCHOENY ET AL.
1 100
S-9 pl/plate
1000
900
100
800
700
600
500
400
300
200
100
100
200
300
400
SAMPLE CONCENTRATION ( fig/plate)
Figure 2. Effect of varying S-9 concentration on TA98 mutagenesis
by ETTM-10 with Aroclor (PCB)-induced S-9.
weight. Tables 3 and 4 show the mutagenic contribution of each
fraction to the whole sample (i.e., percent of whole sample present
in fraction x revertant colonies per milligram). Comparing the
sums of the fractional mutagenic contributions with the activity of
the unfractionated sample reveals two patterns. In the first,
typified by ETTM-01, the sura of the activities of the fractions was
less than the activity of the whole sample. Synergistic actions
among components of the sample could contribute to the greater
mutagenicity of the unfractionated sample. It is also possible
that material was lost or altered during extraction. ETTM-10
represents the second pattern, in which the sum of fraction
activities was greater than the activity of the parent mixture.
This probably indicates that antagonistic components in the whole
sample inhibit total mutagenicity. It is also conceivable that
-------
COAL GASIFICATION AND LIQUEFACTION PRODUCTS
469
ETTIV-08 ETTiVi-01
2000
• PCB-'nduc e d S 9
1800
A j MC Induced S 9
1600
1400
LLI
I-
<
_J
1200
c.
00
LU
1C00
z
o
_J
O 800
CJ
6C0
4C0
2C0
00 200 3C0 400
10C 200 3C0 4C0
SAMPLE CONCENTRATION (^g/plate)
Figure 3. Effect of induction of S-9 (50 yl/plate) enzymes on
sample mutagenicity for TA98.
mutagenic forms were generated by the extraction process, although
this appears unlikely, as no direcC-acting forms were produced.
While mutagenicity in this assay is not proof of a compound's
carcinogenic potential, ic does indicate an urgent need for further
study. If the types of material represented by these samples are
to be produced in large amounts or are found to be widespread in
the environment, they may pose a significant health problem.
It is likely that production of synthetic fuels will increase,
necessitating the use of in vitro assays to assess the health
hazards of products, intermediate streams, and wastes of coal
processes. Future work should include samples from alternative
processes, feedstocks, and varying process conditions. Bioassay
-------
470
RITA SCHOENY ET AL.
Table 3. Liquefaction Product (ETTM-01)
A.

I x II

I
II
Whole Sample

Percent
TA98
Normalization
Extract
Extracted3
(colonies/mg)^
(colonies/mg)
Hexane
29.09
7 ,640
2,222
To 1uene
55.26
16,490
10,761
Methylene
00
5,220
200
chloride

Acetonitrile
0.16
-
-
Res idue
1.74
—
-
Sum of fractions
100

13,183

(74.5Z)
Whole sample
100

17,690

(100%)
B.
No. of Data Coe
fficient of

Points Linear Regression
Whole
30
0.902
Hexane
34
0.975
Toluene
28
0.923
Methylene chloride !
28
0.905
aInitial weight prior to extraction: ETTM-01A, 5.2948 g;
ETTM-013, 2.9320 g. Sura of organic extract weights: ETTM-01A,
5.4655 g (a gain of 3.22Z); ETTM-01B, 3.2128 g (a gain of 9.60%).
^Calculated from linear portions of dose response curves: - * no
dose response .
data from both complete and fractionated samples will be correlated
with chemical characterizations to identify specific compounds or
functionality effects. Such results will be compared with those
from toxicology testing on larger-scale advanced coal processes.
Assessing hazards early in the process-development cycle will
facilitate development of technology to minimize health risks.
-------
COAL GASIFICATION AND LIQUEFACTION PRODUCTS
Table 4. Liquefaction Distillate Oils (ETTM-10)
471
A.

I x II

I
II
Whole Sample

Percent
TA98
Normali zat ion
Extract
Extracted
a (colonies/mg)b (colonies/mg)
Hexane
97.47
9,500
9,260
Toluene
0.86
271 ,370
2,334
Methylene
1.21
63,890
774
chloride

Acetonitrile
0.05
63,510
32
Res idue
0.41
13,260
54
Sum of fractions
100

12,453

(134%)
Whole sample
100

9,280

(100%)
B.
No
. of Data
Coefficient of

Points
Linear Regression
Whole

24
0.815
Hexane

31
0.937
Toluene

24
0.777
Methylene chloride
24
0.934
Acetonitrile

18
0.958
Res idue

38
0.868
aInitial weight prior to extraction: ETTM-10A, 5.0212 g; ETTM-10B,
4.7879 g. Sum of organic extract weights: ETTM-10A, 4.2488 g (a
loss of 15.4%); ETTM-10B, 4.5672 g (a loss of 4.61%).
^Calculated from linear portions of dose response curves.
ACKNOWLEDGMENT
This work was supported by U.S. Department of Energy contract
No. AS-22-78ET0022.
-------
472
RITA SCHOENY ET AL.
4000 -1
Ace tonitrite
3000 -
LU
I-
<
—I
Tolutne
Q.
CO
2000 -
UJ
Residue
O
o
1000 -
aoo
Hgiaie Extract
rnhjene Extract
600
Methylene CTor-ci*
Ac etonitnle E * 'f 3 c'
PcsiClo
400
20C
M»- "00
SAMPLE CONCENTRATION
-------
COAL GASIFICATION AND LIQUEFACTION PRODUCTS
473
Whole
Hexane Extract
ToUene Extract
Methylene Chloride
Extract
Acetoni'rile Extract
Residue
~ Methylene Chloride
5CC
LlI
Whole
j 400
lexane
o
200
Residue
200
300
400
00
SAMPLE CONCENTRATION ( pig/plate)
Figure 5. Mutagenicity of organic solvent extracts of ETTM-02B for
strain TA98 with Aroclor-induced S-9 (50 yl/plate).
Correlation coefficients (r) for the linear regressions
are as follows: whole sample, 0.924; hexane extract,
0.940; toluene extract, 0.980; methylene chloride
extract, 0.994; acetonitrile extract, 0.989: residue,
0.965.
Bingham, E. 1975. Carcinogenic potency of oil fractions derived
from fossil fuels. Presented at Workshop on Health Effects of
Coal and Oil Shale Mining Conversion Utilization, Department
of Environmental Health, Kettering Laboratory, University of
Cincinnati, Cincinnati, OH.
Bridges, B.A. 1976. Short-term screening tests for carcinogens.
Nature 261:195-200.
-------
474
RITA SCHOENY ET AL.
Committee 17 Report, Environmental Mutagen Society USA. 1975.
Environmental mutagenic hazards. Science 187:503-514.
Electric Power Research Institute. 1975. Status Report of
Wilsonville SRC Pilot Plant. Electric Power Research
Institute. May. Pleasanton, CA. p. 32.
Energy Research and Development Administration, 1976. Fossil
energy research program of the Energy Research and Development
Administration, fiscal year 1977. Washington, D.C.
Epler, J.L. 1978. The use of short-term tests in the isolation
and identification of chemical mutagens in complex mixtures.
In: Chemical Mutagens: Principles and Methods for their
Detection, Vol. VI. A. Hollaender, ed. Plenum Press: New
York. pp . 1-54.
Epler, J.L., F.W. Larimer, C.E. Nix, T. Ho, and T.K. Rao. (in
press). Comparative mutagenesis of test materials from
synthetic fuel technologies. In: Second International
Conference on Environmental Mutagens. D. Scott, F.H. Sobels,
and B.A. Bridges, eds. Elsevier/North Holland Biomedical
Press: Amsterdam, Holland.
Freudenthal, R.I., G.A. Lutz, and R.I. Mitchell. 1975.
Carcinogenic potential of coal and coal conversion products.
Battelle Columbus Laboratories, Columbus, OH.
Ketchara, N., and R.W. Norton. 1960. The hazards to health in the
hydrogenation of coal. III. The industrial hygiene studies.
Arch. Environ. Hlth. 1:194-207.
Kier , L.D., E. Yamasaki , and B.N. Araes. 1974. Detection of
mutagenic activity in cigarette smoke condensates. ?roc.
Natl. Acad. Sci. USA 71:4159-4163.
Kornreich, M.R. 1976. Coal conversion processes: potential
carcinogenic risk. MTR-7155, MITRE Technical Report, 4-14 to
4-16.
Loper, J.C., D.R. Lang, R.S. Schoeny, B.3. Richmond, P.M.
Gallegher, and C.C. Smith. 1978. Residue organic mixtures
from drinking water show in vitro mutagenic and transforming
activity. J. Toxicol. Environ. Hlth. 4:919-938.
McCann, J., and B.N. Ames. 1976. Detection of carcinogens as
mutagens in the Salmonella/microsome test: assay of 300
chemicals: discussion. Proc. Natl. Acad. Sci. USA
73:950-954.
-------
COAL GASIFICATION AND LIQUEFACTION PRODUCTS
475
McCann, J., E. Choi, E. Yamasaki , and B.N. Ames. 1975. Detection
of carcinogens as mutagens i_n the Salmone 11 a/micro soma! test:
assay of 300 chemicals. Proc. Natl. Acad. Sci. USA
72:5135-5139.
Oak Ridge National. Laboratory Coal Technology Program Annual
Report. 1975. ORNL-5069. Oak Ridge, TN. p. 52.
Purchase, I.F.H., E. Longstaff, J. Ashby, J.A. Styles, D. Anderson,
P.A. Lefevre, and F.R. Westwood. 1976. Evaluation of six
short-term tests for detecting organic chemical carcinogens
and recommendation for their use. Nature 264:624.-627 .
Sato, S., Y. Seino, T. Ohka, T. Yahagi, M. Nagao, T. Matsushima,
and T. Sugimura. 1977. Mutagenicity of smoke condensates
from cigarettes, cigars, and pipe tobacco. Cancer Lett.
3:1-8.
Sexton, R.J. 1960a. The hazards to health in the hydrogenation of
coal. I. An introductory statement on general information,
process description, and a definition of the problem. Arch.
Environ. Hlth. 1:181-186.
Sexton, R.J. 1960b. The hazards to health in the hydrogenation of
coal. IV. The control program and its effects. Arch.
Environ. Hlth. 1:208-231*.
Simmon, V.F. 1979. In vitro mutagenicity assays of chemical
carcinogens and related compounds with Salmonella typhimurium.
J. Natl. Cancer Inst. 62:893-899.
Swansiger, J.T. 1974. Liquid coal composition analysis by mass
spectrometry. Anal. Chem. 46:730-734.
TRW Systems and Energy. 1976. Carcinogens relating to coal
conversion processes. ERDA Contract E(49-l8)-2213 .
Washington, DC. pp. 26-28.
Weil, C.S. and N.I. Condra. 1960. The hazards to health in the
hydrogenation of coal. II. Carcinogenic effect of materials
on the skin of mice. Arch. Environ. Hlth. 1:187-193.
-------
Intentionally Blank Page
-------
SESSION 6
HAZARD ASSESSMENT
-------
Intentionally Blank Page
-------
THE ROLE OF SHORT-TERM TESTS IN ASSESSING THE HUMAN HEALTH
HAZARDS OF ENVIRONMENTAL CHEMICALS: AN OVERVIEW
Michael D. Waters
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
The problems Co be addressed in this paper and in this session
are twofoLd: first, to identify and confirm mutagenic and
presumptive carcinogenic cheniicaLs through short-term tests (based
on the use of plant as well as animal materials): and second, to
determine the relationship of demonstrated effects in short-term
tests to actual hazards and risks to human health.
When we consider the complexity of environmental samples, it
becomes clear that removing all chemical hazards is not feasible.
Increased exposure to man-made carcinogens and mutagens in the
environment must be presumed to entail additional human health
risk. Therefore, our research objective and our regulatory
objective must be to identify the potential hazards of such
compounds and to minimize the risk to human health.
Even the task of identifying those chemicals that pose
mutagenic or carcinogenic hazards is complex, because of the
diversity of such agents and the multiplicity of their interactions
with biological systems. Mutagens can induce heritable changes in
the genetic material of either somatic or germinal cells of living
organisms. Alterations can occur at the single gene and the
chromosomal level. Many carcinogens act through mutational
mechanisms, but some may act through other physiological changes
that may alter, for example, the individual sensitivity or pattern
of tumor expression. Eventually, we shall have to develop hazard
assessment procedures that take into account the range of genotoxic
and physiological properties of substances evaluated as
carcinogens.
479
-------
480
MICHAEL D. WATERS
Our understanding of Che mechanistic basis of genetic toxic
effects is undoubtedly best developed in the field of mutagenesis.
Much of this information was obtained from the kinds of short-term
systems that we are discussing. We know much less about the
measurement of human genetic effects. Cytogenetic techniques have
demonstrated that changes in structure and number of chromosomes
are associated with a variety of human diseases. The cri-du-chat
syndrome and one form of Down's syndrome are associated with
structural alterations (a chromosomal deficiency and translocation,
respectively). Kleinfelter1s and Turner's syndromes are
attributable to alterations in chromosome number. It is estimated
that chromosomal abnormalities occur in the human population at a
combined frequency of around 0.5% of all live births. Some 50Z of
all spontaneous abortions involve chromosomal defects. The
relationship between heritable chromosomal alterations in man and
exposure to environmental mutagens remains uncertain. But exposure
of various experimental organisms to such agents definitely
increases the frequency of chromosomal abnormalities.
Similarly, the induction of gene mutations by environmental
chemicals is well known in experimental organisms. Such mutations
give rise to new alleles that can have a variety of effects,
depending on the mode of gene expression. In man, dominant alleles
may be responsible for physical defects, such as dwarfism. It is
estimated that dominant mutations may cause adverse effects in up
to 0.5% of the human population. Perhaps of greater concern are
the recessive mutations. Well over 1000 disease states display
inheritance patterns characteristic of recessive mutations,
including phenylketonuria (PKU), Tay-Sachs, and cystic fibrosis.
The association of such genetic anomalies with chemically-induced
mutation in man has not been established. However, since induced
recessive mutations can remain unexpressed for many generations,
exposures to environmental mutagens could be covertly increasing
the genetic load of the human gene pool. Certainly human exposure
to man-made mutagens has increased, and the genetic risk associated
with such exposure must be evaluated.
Heritable genetic damage is not our only concern. The somatic
mutation theory of carcinogenesis extends the concern to
carcinogenesis as well. Fortunately, microbial mutagenesis tests,
when coupled with mammalian metabolic activation, detect a major
proportion of the chemicals that have been shown to be carcinogenic
in animals. The overall qualitative correlation between the
mutagenicity of chemicals in microbial systems and the
carcinogenicity of the same chemicals in experimental animals is
quite good. Limited data suggests that the quantitative
correlation between mutagenic potency in mammalian cell systems and
carcinogenic potency in animals may be somewhat better than in the
case of the microbial systems, but fewer chemicals have been
tested. Dr. de Serres discusses in his paper (1980) results of
-------
HUMAN HEALTH HAZARDS OF ENVIRONMENTAL CHEMICALS
481
various mutagenicity and related tests conducted on a series of
carcinogens and noncarcinogens.
To the extent that such correlations depend on more accurate
measurement of metabolism, we have made several recent advances.
Perfusion techniques and cell dissociation methods have made it
possible to prepare metabolically active primary cells for
co-cultivation with mutagenesis indicator systems or with cells
that are subject to transformation in vitro. While these technical
innovations may improve our ability to detect genetically toxic
agents missed with microsomal activation methods, there is no
substitute for the intact mammal, when mutagenic or carcinogenic
potency must be quantified directly. In addition to
pharmacokinetic considerations, species, strain, and sex
differences must be considered, as well as organ specificity and
differences in sensitivity of various cell types and cell stages.
Ultimately, potency is a question of dose-response relationships
and the probability of effects at environmental dose levels.
Short-term tests that can be applied directly to man are highly
desirable in this regard. Dr. Wyrobeck (1980) specifically
addresses such tests using sperm. Similar procedures are being
developed with laboratory animals. The ability to identify
mutations in single cells of exposed animals will have significant
advantages over presently available in vitro and in vivo
mutagenesis bioassays.
Finally, 1 would like to discuss two maior approaches to
assessing genetic risk to humans resulting from exposure to
chemical mutagens. One uses experimental data obtained directly
from induced germinal mutations in intact animals, and the other
uses data on chemical dose to the germ cells of animals together
with mutagenesis results obtained with other-than-gerrainal cells in
various short-term tests.
Tests that directly provide information on induced germinal
mutations include the specific locus test, the X-Y chromosome loss
test, and the heritable translocation test. These tests are
complete, in that they measure genetic damage in germ cells which
is expressed in the subsequent generation. This information on
induced mutation frequency may be combined with information on
levels of human exposure. To estimate human risk, the induced
mutation frequency is extrapolated downward to the estimated level
of human exposure. Theory suggests the use of a linear or
no-threshold model for point- or gene—mutational effects.
Chromosomal alterations, on the other hand, are thought to proceed
by multi-hit mechanisms. Thus, linear extrapolation of
translocation data is likely to overestimate risk at lower levels
of exposure. Other models may be more appropriate when supported
by sufficient data. As was mentioned earlier, gene mutations
usually occur at lower doses than do chromosomal mutations;
-------
482
MICHAEL D. WATERS
therefore, gene mutation data would be expected to be more
sensitive for human risk assessment.
The issue of sensitivity tends to favor the second approach to
genetic risk assessment. Because short-term test systems,
particularly the in vitro systems, are usually more powerful, it is
possible to detect more-readily-induced mutants and to relate this
number to the chemical estimation of mutagen-DNA interactions or
binding to DNA. A similar determination can be made of mutagen-DNA
interactions in the germinal cells of the intact animal that has
been exposed via an appropriate route. With knowledge of
mutagen-DNA interactions in the intact animal and in the short-term
mutagenesis test, a relationship can be constructed between
exposure in the one system and induced mutation frequency in the
other. To the extent that binding of the test mutagen to DNA can
be measured at anticipated human exposure levels, it may not be
necessary to perform a high-to-low dose extrapolation. If
extrapolation is required, it may be assumed that DNA binding is
directly proportional to the exposure level, unless there is
evidence to the contrary. It is extremely important to understand
the relationship between exposure, DNA binding, and mutation in all
of our short-term mutagenesis tests in order to make maximum use of
resulting data for relative chemical potency evaluation and for
hazard assessment. These same considerations of exposure,
effective dose, and response apply as well to other genetically
mediated effects.
As far as carcinogenesis is concerned, we are in a better
position to address the issue of human risk assessment because we
have epidemiological evidence of cancer in man. This evidence
provides us with the necessary information on human exposure and
response to validate our long-term whole animal models. On the
strength of correlations with whole animal data, short-term test
results provide evidence of the carinogenic potential of previously
untested pure chemicals and complex mixtures. We have seen in this
symposium that comparative studies of the relative biological
activity of complex mixtures can be used to provide early
information on the carcinogenic potential of such materials.
However, at this stage in the development of short-term tests,
their results are best considered suggestive rather than
conclusive. Given the evidence from these tests and human exposure
considerations, it is reasonable to require the performance of
long-term whole animal tests for carcinogenesis and related effects
to define more precisely the extent of human health risk. Dr.
Albert, who is directly involved in the process of health risk
assessment, elaborates on these points in his paper (1980).
-------
HUMAN HEALTH HAZARDS OF ENVIRONMENTAL CHEMICALS
483
REFERENCES
Albert, Roy E. 1980. Assessing carcinogenic risk resulting from
complex mixtures. Presented at U.S. Environmental Protection
Agency Second Symposium on the Application of Short-term
Bioassays in the Fractionation and Analysis of Complex
Environmental Mixtures, Williamsburg, VA.
de Serres, Frederick J. 1980. International program for the
evaluation of short-term tests for carcinogenicity. Presented
at the U.S. Environmental Protection Agency Second Symposium
on the Application of Short-term Bioassays in the
Fractionation and Analysis of Complex Environmental Mixtures,
Williamsburg, VA.
-------
Intentionally Blank Page
-------
THE INTERNATIONAL PROGRAM FOR THE EVALUATION OF SHORT-
TERM TESTS FOR CARCINOGENICITY (IPESTTC)
Frederick J. de Serres
National Institute of Environmental Health Sciences
Research Triangle Park, North Carolina
INTRODUCTION
The major impetus for developing the International Program for
the Evaluation of Short-term Tests for Carcinogenicity (IPESTTC)
was that our need for rapid identification and control of
carcinogens is not satisfied by traditional rodent bioassays.
Because of resource limitations, rodent studies cannot be carried
out on a large enough scale to identify all carcinogenic chemicals
in the environment within a reasonable period of time. This need
places tremendous pressure on the scientific community to develop
test systems for identifying chemical carcinogens in the
environment at a lower cost and on a shorter time scale.
A major problem in selecting short-term tests for
carcinogenicity has been that the mechanisms of action of chemical
carcinogens have not been well understood. However, during the
past decade, developments in genetic toxicology have allowed
progress towards a unifying theory of the action of many
carcinogens. This theory is based on the hypothesis that chemical
carcinogens induce mutations in somatic cells, and that these
mutations change the behavior of the cells so that cancer develops.
It is a logical step from accepting the somatic mutation theory of
cancer to using short-term mutation tests for detecting chemicals
that potentially produce mutations and cancer. However, other
theories of the induction of cancer exist, and many other highly
accurate tests have been developed to identify carcinogens.
The availability of a large number of short-terra tests has
created a problem for both the scientist and administrator who must
select the most appropriate and accurate test systems for
485
-------
486
FREDERICK J. DE SERRES
carcinogenicity screening, During Che past few years, various
laboratories have tackled this problem using well-known validation
studies. The purpose of these studies is to assess how effectively
the short-term test can distinguish between carcinogens and
noncarcinogens that have been classified as such by their activity
in whole-animal systems. However, the problem remains of how to
compare the performance of different test systems when the data
describing the performance has been developed in different
laboratories using tests with different protocols and different
criteria.
IPESTTC was designed specifically to examine the abilities of
various test systems to distinguish between known chemical
carcinogens and known noncarcinogens. Although many of the test
systems may also detect mutagenic events, this was not the primary
purpose of this program. As the program developed, it became
apparent that a secondary aim could be achieved, namely, providing
a body of data on the mutagenicity of various chemicals. This body
of data would not only describe the effects of the chemicals on
various test systems, but would help to make clear what additional
information was required to describe completely the biological
activity of these chemicals.
HISTORICAL DEVELOPMENT OF IPESTTC
This program emerged from research on short-term tests
supported by the United Kingdom Medical Research Council (MRC)
under commission from the United Kingdom Health and Safety
Executive (HSE). Early discussions on how best to carry out this
study led to a proposal by scientists at the Imperial Chemicals
Industries, Ltd. (ICI) to expand the study to include a larger
number of chemicals, which would be tested blind as coded samples.
ICI scientists took on the responsibility for selecting the
chemicals and preparing them in a high state of purity and in large
enough quantity that all investigators could work with samples
taken from the same batch. As it has evolved, the program is a
unique attempt to gather a large set of test results with which to
objectively evaluate the ability of short-term tests to correctly
distinguish between the carcinogens and noncarcinogens.
The initial selection of test systems included a wide variety
of assays, but none involved inducing cancer in animals. In each
case a correlation with carcinogenic activity in laboratory animals
was sought. It was agreed that the most useful method for
assessing the performance of short-term tests would be to compare
the results for pairs of structually-related chemicals where one
was a known carcinogen and the other a known noncarcinogen. The
investigators would not know the identity of the chemicals tested;
in other words, the test chemical would be evaluated as coded
-------
EVALUATION OF SHORT-TERM TESTS FOR CARCINOGENICITY
487
samples in a blind trial. Scientists of Imperial Chemicals
Industries, Limited, agreed to supply the chemicals for this
testing program and selected the 25 carcinogens and 17
noncarcinogens (including 14 paired compounds). All chemicals were
synthesized and prepared for the study in as pure a state as
possible and in large enough quantity that all investigators could
use samples from the same batch. Since the 50-g quantities
prepared exceeded the requirements of the original HSE/MRC program,
the National Institute of Environmental Health Sciences (NIEHS)
developed a plan to include a larger variety of test systems. This
expansion made it possible to look for interlaboratory variation in
test performance. In other cases, where the time required to test
the set of 42 chemicals exceeded that allowed in the program plan,
samples were divided among laboratories performing the same test,
in an effort to make the final test data as complete as possible.
Many investigators financed their own testing. In other
cases, the work was financed by the HSE/MRC science in the United
Kingdom, NIEHS or the U.S. Environmental Protection Agency in North
America, or various other mechanisms in other parts of the world
such as Japan and the Soviet Union. By the time of the final
meeting at St. Simons Island, GA, in October, 1979, 30 different
assay systems were part of the program and data from over 50
laboratories were considered.
SELECTION OF TEST CHEMICALS
The three main criteria for selecting the test chemicals were
1) to have as large a range of chemical types and chemical classes
as possible in a group of 42; 2) Co ensure the highest possible
purity and to ensure that all samples of each chemical under test
would be taken from the same batch; and 3) to obtain a balance
among different chemical classes, not including too many chemicals
from any particular class. In addition, 11 of the 25 chemical
carcinogens selected were included, because they were known to be
difficult to detect as mutagens in assays for point mutation in
either Salmone11a typhimurium or Escherichia coli.
In the selection of noncarcinogens, an attempt was made to
find structural analogs of the chosen carcinogens. When a test
system gives a positive response to a carcinogen, one cannot
determine whether the assay is responding to some chemical property
other than the carcinogenicity inherent in the chemical structure.
Therefore, the most meaningful assay systems for screening tests
would be those that gave a positive response with a given
carcinogen and a negative result with its noncarcinogenic
structural analog. This is another reason for stressing purity of
the test chemicals: to be sure that the noncarcinogens (for
example) are not contaminated with trace levels of carcinogenic
-------
488
FREDERICK J. DE SERRES
structural analogs. Such contamination would elicit a positive
response for noncarcinogenic chemicals that should give negative
results. All chemicals (except auramine) were more than 99% pure,
and at least six different criteria of purity were evaluated.
The chemicals selected for testing are given in Table 1. This
table lists the 25 carcinogens and the 11 of these 25 that are
difficult to detect in bacteria. Also listed are the 17
noncarcinogens arranged in order of the strength of the data
available in the literature for making this evaluation. Selection
of the noncarcinogens proved exceptionally difficult and somewhat
disappointing, since much attention has been given in the past to
the development of criteria for carcinogenicity, but not
noncarcinogenicity. In Table 2, the list of chemicals are
reorganized to show the 14 pairs of carcinogens and noncarcinogenic
analogs.
SELECTION OF ASSAY SYSTEMS
An effort was made to include representatives of all available
types of short-term assays thought to be potentially useful
carcinogenicity pre-screening tests. Several of the assays
included in the study were part of the original HSE/MRC study in
the United Kingdom. As the program grew, subsequent sponsors
selected other assays to fill in gaps in the program. Many assays
were included because of their sponsors' interest in participation
and willingness to do so without financial support. The
developmental status of the assays ranges from those considered to
be the best-standardized and -validated (such as the Salmonella/
microsome assay) to those in the earliest stages of development
(such as the inductest and the diptheria toxin resistance system in
human fibroblasts).
No effort was made to standardize protocols. Without
knowledge of optimum protocols, standardization would only insure
that each investigator using a given assay would make the same
mistakes. As a result, the program allowed a comparison of results
from different protocols. Among the investigators using
established assays, general agreement on a protocol was reached for
maximum comparability of the results (as with the sex-linked
recessive lethal test in Drosophila and the micronucleus test in
mice) .
The assay systems may be divided into five groups as shown in
Table 3. In group 1 are listed the two inductests used in the
study; one assays the lysis of bacteria (E. coli and B. subtilis),
the other, the expression of genes linked to prophage lambda. The
other two tests are the degranulation test (which assays the
dissociation of ribosoraes or polysomes from rat liver endoplasmic
-------
EVALUATION OF SHORT-TERM TESTS FOR CARCINOGENICITY
489
Table 1. Chemicals Selected for Testing in the International
Program for the Evaluation o£ Short-term Tests for Carcinogenicity
Chemicals Classified as Carcinogens
4-Nitroquinoline-N-oxide
Benzo(a)pyrene
9,10-Dimethy1anthracene3
2-Acety1aminofluorene
N-nitrosoraorpholine
2-Naph thylamine
Hydrazine sulfate
Diethylstilbestrol13
Cyclophosphamide
3-Aminot ri azole*3
4-Dimethylaminoazobenzene
(butter yellow)'5
4,4'-Methvlenebis-
(2-chloroanilin) (MOCA)
Hexamethylphosphoramide (HMPA)b
Methylazoxymethanolacetats
Auramine (technical grade)'3
Benzidine
6-Propiolactone
Chloroform*5
Dimethylcarbamoyl chloride
Urethane^3
DL-Ethionine*5
Ethylenethioureab
Safrole^
Epichlorohydrin
O-Toluidine hydrochloride'5
Chemicals Classified as Noncarcinogens
Pyrene3
Anthracene3
4-Acetylaminofluorenea
1-Naphthy I amine''
Azoxybenzenee
Sugar (sucrose)e
Dinit rosopentamethylene
tetraminee
Isopropyl N(3-chlorophenyl)
carbamate0
3-Methyl-4-ni troquinoline-
N-oxidee
3,3' ,5,5'-Tetramethylbenzidinee
y-Butyrolactone
1,1,1-Trichloroethane^
Dimethylformamide^
. .9
Diphenylmtrosamine
Methionine3
Ascorbic acid3
4-Dimethylaminoazobenzene-4-
sulfonic acid Na salt^
aData not convincing.
''Carcinogen difficult to detect in bacterial assays.
cBest evidence for noncarcinogenicity.
^Intermediate evidence.
ePoorest evidence for noncarcinogenicity.
-------
490 FREDERICK J. DE SERRES
Table 2. Pairs of Structural Analogs Among the Chemicals Tested
Carcinogen Noncarcinogenic Analog
4-Nitroquinoline-N-oxide
Benzidine
4-Dimethylaminoazobenzene
(butter yellow)
Benzo(a)pyrene
3-Propiolactone
9,10-Dimethylanthracene
Chloro form
2-Acety1 amino f1uorene
Dimethylcarbamoyl chloride
2-Naphthylamine
N-Nit rosomorpholine
Urethane
Methylazoxyraethanol acetate
DL-Eth ionine
3-Methyl-4-nit roquinoline-N-oxide
3,3' ,5 ,5'-Tetramethylbenzidine
4-Dimethylaminoazobenzene-4-
sulfonic acid Na salt
Pyrene
v-Butyrolactone
Anthracene
1,1,1-Trichloroethane
4-Acetylaminofluorene
Dimethylformamide
1-Naphthylamine
Dinitrosopentamethylene tetramine
Isopropyl N(3-chlorophenyl)carbamate
Azoxybenzene
Meth ionine
reticulum) and the nuclear enlargement assay (in which a positive
result is indicated by an increase in the size of the nuclei in
both HeLa cells and fibroblasts in culture).
Group II includes assays for the induction of point mutations
in bacteria, including assays for forward and reverse mutation in
both Salmonella and E. coli. The Salmonella/microsome reverse-
mutation assay was conducted in 13 separate laboratories using a
total of seven strains. Several procedures were used for metabolic
activation, and the variations here included 1) source of S-9 ,
2) use of different chemicals as inducers, 3) variation in the
amount of S-9, 4) plate incorporation versus pre-incubation,
5) the use of hepatocytes from rat liver, and 6) the addition of
the comutagen norharman. Data were also obtained for a Salmonella
assay that measured forward mutations to azaguanine resistance. In
E. coli, two systems were included: one in which both forward and
reverse mutation at four loci are screened simultaneously and
another in which tryptophan reversion is measured in different
strains of _E. coli strain WP2.
The tests in group III measure various types of genetic damage
in yeast, including forward mutation in Saccharomyces pombe and the
following endpoints in S. cerevisiae: reverse mutation, mitotic
crossing over in five different strains, induction of aneuploidy in
strain D6, and differential survival of wild type and a multiply
repair-deficient strain.
-------
EVALUATION OF SHORT-TERM TESTS FOR CARCINOGENICITY
491
Table 3. Assay Systems Used in the International Program for Che
Evaluation of Short-term Tests for Carcinogenicity
I. Prokaryotic Repair, Phage Induction, Nongenetic Assays
Inductests:
Baci1lus subt ilis rec assay Escherichia coli pol assay
Escherichia coli Nuclear enlargement
Degranulation test
II. Prokaryotic Mutation Assays
Salmonella/raicrosome fluctuation assay Escherichia coli 343
Salmonella 8-AZA resistance
III. Yeast Assays
Forward mutation—_S. pombe
Reverse mutation—XV185-14C
Mitotic recombination - PG-154, PG-155,
D4, D7, JD1
IV. Mammalian In Vitro Assays
Unscheduled DNA synthesis (WI-38, HeLa) CHO-HGPRT, APRT, TK, OUS
Sister-chromatid exchange (CHO) V79-HGPRT
Chromosome aberrations (CHO, RLj) Human fibroblasts—
BHK21—transformation diphtheria toxin
resistance
V. In Vivo Assays
Aneuploidy - D6
Repair assay - RAD, URA
Sex-linked recessive lethal—
Drosophila
Sister-chromatid exchange—mouse
Micronucleus—mouse
Sperm morphology—mouse
-------
492
FREDERICK J. DE SERRES
The assays in group IV are mammalian in vitro systems,
including assays for gene mutation in Chinese hamster, V79, mouse
lymphoma, and human fibroblast cells. Chromosome effects in vitro
include sister-chromatid exchange, chromosome aberrations, and
unscheduled DNA synthesis. The BHK21 cell transformation assay
based on growth in soft agar is also in this group of assays.
The Group V assays are whole-animal systems, including sex-
linked recessive lethals in Drosophila, the micronucleus test in
mice, and sister-chromatid exchange in liver cells and bone marrow.
The only plant system included in the program is the induction of
gene mutations resulting in color changes in Tradescantia stamen
hairs.
EVALUATION OF THE DATA BASE
The data base that has corae out of this program is broad and
complex. Eventually, it can be used not only to compare the
performances of the individual assay systems, but also to evaluate
the effects of the chemicals both qualitatively and quantitatively.
One of Che difficulties in comparing the performances of the
various assay systems is that due to resource limitations, not all
42 chemicals were tested in each assay. Since the samples were
coded, the chemicals tested should be a random sample of the 42.
However, the chemicals were usually tested in order of their
receipt, so that the gaps in the data base involve the same group
or groups of chemicals. Resource limitations also precluded
interlaboratory comparisons of each assay system.
Each assay system's performance has been evaluated (Table 4)
by its ability to correctly classify as positive the carcinogenic
chemicals (sensitivity) and as negative the noncarcinogenic
chemicals (specificity) and by its overall ability to correctly
classify both carcinogens and noncarcinogens (accuracy). It should
Table 4. Terms Used in Describing Assay Performances
Sensitivity - # positive results with carcinogens
(true positive fraction) 4 carcinogens tested
Specificity = $ negative results with noncarcinogens
(true negative fraction) # noncarcinogens tested
Accuracy = # correct results
# chemicals tested
-------
EVALUATION OF SHORT-TERM TESTS FOR CARCINOGENICITY
493
be remembered that two assay systems can have the same sensitivity,
specificity, and accuracy, and yet differ in the sets of chemicals
that they correctly classify.
In the present study, 11 of the 25 carcinogens were selected
because they are difficult to detect as mutagens in the bacterial
assays for point mutation. Because of this, the sensitivity,
specificity, and accuracy figures will differ markedly from
previous studies. The data from the present program are being
selectively analyzed so that these comparisons can be made for
selected portions of the data base, with groups of chemicals
omitted or with their classifications changed.
The calculations for the 42 chemicals show that no single
assay system is sufficiently sensitive, specific, and accurate to
be used in isolation. The data clearly indicate the need for a
battery of tests.
EVALUATION OF TEST DATA ON THE 42 CHEMICALS
The main problem with this evaluation was that six of the
noncarcinogens may not have been correctly classified. The
following six chemicals gave a high frequency of positive results
over a wide range of assay systems: 3-methyl-4-nitroquinoline-N-
oxide, 4-acetylaminofluorene, 1-naphthylamine, azoxybenzene ,
diphenylnitrosamine, and 4-dimethylaminoazobenzene sulfonic acid
(methyl orange). Those results indicate that they are probably
carcinogens and are thus raise 1 assified for the purpose of the
present program. An advantage of computerization of the IPESTTC
data is that the classification of these six chemicals can be
changed and new estimates of sensitivity, specificity, and accuracy
obtained.
PROGRAM COORDINATION AND COMPLETION
IPESTTC was managed by a coordinating committee consisting of
F.J. de Serres, Chairman, J. Ashby, P. Brookes, B. Bridges, I.
Purchase, M. Shelby, and T. Sugimura. This group was responsible
for the initial selection of assay systems and investigators,
distribution of test samples, collection of data, and organization
of the various workshops held during the course of the study. In
addition, this group was responsible for the follow-up work
required to complete the present data base, as well as its final
analysis .
-------
494
FREDERICK J. DE SERRES
An interim report of this program will be published as a book
(de Serres and Ashby, in press) containing reports from all of the
investigators. It will also contain summary reports from the
various test-system and test-chemical work groups that met to
evaluate the data base after decoding of the test chemicals. The
data, along with more detailed discussion, will be published early
in 1981 .
REFERENCES
de Serres, F.J., and J. Ashby. (in press). International Program
for the Evaluation of Short-Term Tests for Carcinogenicity.
Elsevier/North Holland: Amsterdam.
-------
SPERM ASSAYS IN MAN AND OTHER MAMMALS AS INDICATORS OF
CHEMICALLY INDUCED TESTICULAR DYSFUNCTION
Andrew J. Wyrobek
Lawrence Livermore Laboratory
University of California
Livermore, California
INTRODUCTION
Concern about human exposure to chemical agents has led to the
development of numerous bioassays to detect mutagens and
carcinogens rapidly and inexpensively. Recent attempts to compare
and evaluate the efficacy of these bioassays have shown clearly
that no single assay is sufficient (Coordinating Committee, 1978).
A number of sperm assays have been developed and evaluated
(Wyrobek, in press). Although the mutagenic basis of chemically
induced sperm anomalies is generally not well understood, these
assays play an important role in mutagenesis and carcinogenesis
testing. First, sperm assays can measure chemical damage to the
germ cells occurring during spermatogenesis or transit through the
efferent ducts. Agents that are found to be mutagenic or
carcinogenic in other bioassays can be tested directly for their
spermatotoxic effects. This possibility is of major importance
because the activity of an agent in bacteria or mammalian somatic
cells is often a poor predictor of its activity in the testes after
exposure in vivo (Coordinating Committee, 1978). Second, since
animal sperm assays are as inexpensive as other short-term tests,
many agents can be tested. Third, sperm assays have not only been
developed and applied to mice, other rodents, and a variety of
domestic animals, but they are also applicable to men exposed to
chemicals (Wyrobek and Glendhill, in press).
This paper describes the application and methodology of sperm
assays in men and laboratory animals and discusses the predictive
value of induced sperm changes, correlations of results to
carcinogenicity and mutagenicity, and the relative strengths and
weaknesses of sperm assays (for more detailed accounts with
495
-------
496
ANDREW J. WYROBEK
specific agents, see Topham, 1980b; Wyrobek and Bruce, 1975:
Wyrobek et al., in press b, c).
METHODOLOGY
Human Sperm Assays
Human semen assays have a long history in Che diagnosis of
infertility. Thus, it is not surprising that early attempts to
assess altered testicular function in men exposed to chemicals have
involved measuring changes in the sperm parameters commonly used in
fertility diagnosis, such as sperm density (counts), motility, and
morphology.
Of these parameters, morphology is the most constant in an
unexposed man and is statistically highly sensitive to small
changes (Wyrobek et al., in press c). In the past, this assay has
not been widely used, because the scoring criteria were generally
difficult to standardize, and inteclabocatory discrepancies were
unavoidable. We have improved the human morphology assay by
describing 10 classes of spermhead shapes and categorizing at least
500 sperm per individual into one of these classes. Through the
intermittent use of coded standard slides, we have shown constancy
in scoring the same set of standard slides for up to three years
and have demonstrated the objective nature of the scoring criteria
(Wyrobek, in press). We have adapted this method to men
occupationally exposed to carbaryl (Wyrobek et al., in press e) ,
anesthetic gases (Wyrobek et al., in press a), dibromochloropropane
(DBCP), and mercury (Wyrobek, et al., in preparation). Men exposed
to carbaryl and DBCP showed marked sperm changes when compared with
unexposed men. Men exposed to cancer-chemotherapeutic agents
showed drug-related decreases in sperm counts and increases in
sperm shape abnormalities (Wyrobek et al., 1980).
Among the more recently developed human sperm assays is the
YFF test, which is based on the unique fluorescence of the human Y
chromosome when stained with quinacrine (Kapp, 1979). Men exposed
to adriamycin and DBCP, for example, showed exposure-related
increases in the proportion of sperm having two fluorescent spots,
which are presumably due to the presence of two Y chromosomes in
one sperm because of errors in meiotic disjunction (Kapp, 1979).
A recent survey of the literature (Wyrobek et al., in press c)
showed that human sperm assays have been more widely used than was
generally suspected: more than 80 papers were found on the use of
semen assays in assessing testicular function. About 75% of the
exposures involved experimental or therapeutic drugs; about 15%
were occupational exposures; and about 10% involved personal drug
use. The studies cover 37 single agents, 10 complex mixtures, and
-------
SPERM ASSAYS AS INDICATORS OF DYSFUNCTION
497
12 sets of multiple agents. Tables 1 and 2 list the single agents
and complex mixtures categorized by those agents that 1) were
found to induce significant changes in the sperm parameters of
exposed men, 2) gave suggestive but inconclusive evidence of
change, and 3) had no effect. Of the 37 single agents, 21 were
positive, 9 suggestive, and 6 negative. Almost all of the agents
showing some effect had a reduction in semen quality (e.g.,
reduction in spera counts, decrease in sperm motility, decrease in
the proportion of sperm with normal shapes, or increased
frequencies of YFF sperm). However, five agents (chlomiphene
citrate, coenzyme Q-7, fluoxymestrone, kallikrein, and methadone)
were found to increase sperm counts and/or motility in some of the
infertile patients studied.
Table 1. Single Agents Studied with Human Sperm Assays3
Suggestive but
Positive Effects Inconclusive Effects No Effects Observed
Aspartic acid
Chlorambucil
Chlomiphene citrate
Cyclophosphamide
Cyproterone acetate
Doxorubin hydrochloride
Snovid
Gossypol
6 Medroxyprogesterone
Metandienone
Nilevar
Norlutin
Prednisone
Progesterone
Salicylazosulfapyride
Testosterone enanthate
Testosterone propionate
WIN 13099
WIN 17416
WIN 18446
Centchroman
Colchicine
Methadone
Methotrexate
Metronidazol
Nitrofurantoin
Trimeprimine
CIBA-32644 Ba
Lysine
Methyl testosterone
Ornithine
Testosterone
WIN 59491
aFor data on specific agents and their chemical names, see Wyrobek
et al. (in press c). Table entries are based on studies of sperra
counts, motility, morphology, and YFF. The assignment of
individual agents to columns is based on the data provided in the
papers reviewed by the U.S. Environmental Protection Agency (EPA)
Gene-Tox panel (see Waters, 1979) and may change as more data
becomes available.
-------
498 ANDREW J. WYROBEK
Table 2. Complex Mixtures Studied with Human Sperm Assays3
Positive Effects
Suggestive but
Inconclusive Effects No Effects Observed
Alchoholic beverages Carbaryl^3
Carbon disulfide'5 Diethylstilbestroi
Dibromochloropropane^ Tobacco smoke
Lead"
Epichlorohydrin^
Glycerine workers'3
Polybrominated
biphenylsb
Mari juana
aFor data on specific agents and their chemical names, see Wyrobek
et al. (in press c). Table entries are based on studies of sperm
counts, motility, and YFF. The assignment of individual mixtures
to columns is based on the data provided in the papers reviewed by
the EPA Gene-Tox panel (Waters, 1979) and may change as more data
becomes available.
t>Oc cupational exposures-
Eleven complex mixtures were studied using the sperm assays.
As shown in Table 2, five showed positive effects, three showed
suggestive but inconclusive effects, and three showed no effect.
Most of these studies involved occupational exposures in which
single active agents were implicated (e.g., DBCP or lead).
Sperm assays have also been used in men exposed to at least 12
sets of two or more agents in consort (Wyrobek et al., in press c).
Ten combinations caused reductions in semen quality (e.g.,
cyclophosphamide plus prednisone, Danazol plus testosterone
enanthate, and MVPP cancer therapy).
Of the 60 different human exposures evaluated, including all
single agents, multiple agents, and complex mixtures, 97% used
sperm counts as one of the parameters measured, of which 25% used
counts as the only parameter measured. Furthermore, 58% of all the
exposures evaluated studied motility, 42% studied morphology, and
only 7% used YFF.
Because of our poor understanding of the genetic mechanisms
underlying various induced sperm anomalies, the only information
that can be gained from these assays at present is whether human
spermatogenesis is affected by exposure to a chemical agent or
mixture of agents. These data, together with results of other
short-tern assays for mutagenicity, may indicate which of these
agents are potential germ cell mutagens. Clearly, more research is
needed to develop sensitive sperm assays with defined mutational
-------
SPERM ASSAYS AS INDICATORS OF DYSFUNCTION
499
end points, so that risks of heritable damage may be assessed
directly using sper*n.
Animal Sperm Assays
Sperm assays have also been used to assess chemically induced
changes in testicular function in a variety of animal species. At
least 13 agents have been studied in rabbits, 14 in rats, 2 each in
sheep, cattle, dogs, hamsters, and monkeys, and 1 in pigs (for
review, see Wyrobek et al., in press b). Most of the studies
(covering approximately 159 agents) used the mouse sperm morphology
assay (Wyrobek and Bruce, 1975). Approximately 40 agents produced
dose-dependent increases in sperm shape abnormalities, 105 were
negative, and 5 showed marginal responses. However, of the
negatives, only about 50 were known to have been tested to lethal
doses, and the remainder should be retested at higher doses.
Positive results came from a wide variety of chemical classes,
including antimetabolites, alkylating agents, spindle poisons,
polyaromatic hydrocarbons, aromatic amines, estrogens, and others.
Negative responses will occur with any compound that rapidly kills
the animal or whose active form does not reach the testes either
because of the route of exposure or the metabolism required to
activate the agent. Because only seven of the chemical agents
reviewed were tested in two or more species, meaningful
interspecies comparisons of results of sperm assays are not yet
possible.
DISCUSSION
The Genetic Implications of Chemically Induced Sperm Anomalies
Evidence from human studies. Although it is generally agreed
that major reductions in sperm counts and motility are linked to
reduced fertility, it remains unclear which sperm parameter(s)
indicates embryonic failure or heritable genetic abnormalities.
Human data on this question is very limited. In one study, fathers
of 201 spontaneous abortions showed significantly higher sperm
abnormalities and lower sperm counts than 116 fathers of normal
pregnancies (Furuhjelm et al., 1962), suggesting a link between
poor semen quality and frequency of spontaneous abortions.
Although several studies support this observation (Czeizel et al.,
1976; Joel, 1966; Takala, 1957), some studies found no such
correlation (Kneer, 1957). Clearly, more human studies are needed
to compare exposure of the male parent, induced sperm defects, and
reproductive outcome.
-------
500
ANDREW J. WYROBEK
Evidence from animal studies. Most of the studies on genetic
validation of induced sperm abnormalities have been conducted with
mice. Three lines of evidence link induction of abnormal sperm and
heritable genetic abnormalities. First, it Is clear that sperm
shaping and the production of abnormal sperm is polygenically
controlled by autosomal as well as sex-linked genes (Beatty, 1972;
Brozek, 1970; Hugenholtz and Bruce, 1979; Krzanowska, 1972; Topham,
1980a; Wyrobek and Bruce, 1978; Wyrobek, 1979). Second, in at
least three independent studies with numerous mutagens and
nonmutagens, germ cell mutagens generally induced sperm
abnormalities, while nonmutagens generally had no effect (Bruce and
Heddle, 1979; Topham, 1980b; Wyrobek and Bruce, 1978). Third, in
several studies using agents that induce sperm abnormalities, sperm
abnormalities were transmitted to the male offspring of the exposed
mice (Hugenholtz and Bruce, '979; Sotomayor, 1979; Staub and
Matter, 1976; Topham, 1980a; Wyrobek and Bruce, 1978).
A brief survey of the literature indicates that many of the
compounds that are active in the mouse sperm morphology test are
also active in the heritable specific locus, sperm morphology,
heritable translocation, and/or dominant lethal tests in mice (for
review, see Wyrobek et al., in press b). Therefore, the mouse
sperm morphology test may be a useful screening test for compounds
that constitute a potential genetic hazard for mammals. Spindle
poisons that may cause nondisjunction in germ cells can also be
identified. Further murine studies are needed to understand the
quantitative relationships among dosage regime, appearance of
abnormal sperm shapes in the semen, time between exposure and
conception, fertility of the exposed male, frequency of genetically
abnormal offspring, and fertility of the abnormal offspring.
Correlations of Sperm Abnormality Results with Results of Short-
term Assays for Carcinogenesis
Several attempts have been made to compare results of the
murine sperm morphology assay with other short-term tests for
carcinogenesis. As part of the International Program for the
Evaluation of Short-term Tests for Carcinogenesis, six pairs of
carcinogens and noncarcinogens and five unpaired carcinogens were
surveyed as unknowns, using the mouse sperm abnormality assay
(Wyrobek et al., in press d). No false positive responses were
found, suggesting that the sperm assay has a high specificity for
carcinogens. However, several false negatives were obtained,
indicating that not all carcinogens induce sperm abnormalities in
mice. This data may be very important in assessing which
carcinogens may also be active in the testes. The detailed
comparison of the sperm abnormality assay and the other assays
surveyed in this program is still in progress.
-------
SPERM ASSAYS AS INDICATORS OF DYSFUNCTION
501
A comparison of the potency (dose required to double
background frequencies) of some 30 agents studied in both the
Salmonella/microsome and spent) abnormality assays showed no
apparent correlation, suggesting that the assays are measuring
different biological phenomena (unpublished results). In a
different study of 61 agents (Bruce and Heddle, 1979) with the
mouse sperm morphology and Salmonella/microsome assays, each assay
was found to correctly identify approximately 60% of the
carcinogens and noncarcinognens tested. The assays together
identified nearly 90% of the agents, suggesting that the two assays
measure different end points. These authors recommended that the
number of false negatives, which is relatively high with each of
the two assays, may be reduced by using a battery of both assays
for the identification of potential carcinogens.
CONCLUSIONS
Advantages
The major advantages of sperm assays are that the cells
examined are readily available in both animals and man, and that
sperra carry the paternal genome in the form that will be ultimately
involved in fertilization. Other advantages are the following:
1) Sperm are examined after exposure of a whole mammal. This
helps ensure that artifacts (false positives and false
negatives) due to problems of tissue penetration,
metabolism, pharmacokinetics, and dosage encountered in
non-gonadal, cultured-cell, or nonmaiamalian systems are
minimized.
2) The changes in sperm parameters probably arise from
interference by the test substance with the
differentiation of the sperm cell. Thus, these changes
are intrinsically relevant to safety evaluation and
assessment of potential effects of the agent on male
fertility.
3) The laboratory methods are generally rapid, inexpensive,
and quantitative.
4) Sperm assays have major advantages over other approaches
for assessing induced changes in testicular function.
Testicular biopsies are impractical, traumatic, invasive,
and may themselves affect testicular function.
Epidemiological surveys of reproductive function using
questionnaires exclusively require large sample sizes and
are generally expensive. Analyses of blood levels of
gonadotrophins are expensive and generally insensitive to
-------
502
ANDREW J. WYROBEK
small changes in testicular function. Compared with these
methods, sperm assays are noninvasive, inexpensive,
require small sample sizes for effective analyses, and are
sensitive to small changes.
Disadvantages
The major disadvantages of sperra assays are the following:
1) Heritability of the induced damage is not yet clearly
demonstrated.
2) Limited sperm sampling times and dosage regimens may
reduce the sensitivity of the assay (e.g., agents that
only exert transient effects may be missed by using single
sampling times).
3) Other factors such as ischemia, infection, and starvation
may produce spurious false positive responses.
Applications
The availability of animal and human sperra assays suggests
several applications in the assessment of chemically induced
spermatotoxicity (antifertility effects) and heritable genetic
abnormalities. Animal sperra assays (such as the mouse morphology
assay) may be used to screen large numbers of agents to establish a
ranking that sets priorities for sperm studies in exposed men.
This approach would minimize the use of human studies that
generally have complex requirements for epidemiological and
statistical input and often require lengthy interactions with union
officials, industry representatives, employees, physicians, and
patient-donors. Furthermore, animal sperra studies also may be
useful in evaluating the relative effects of the components of a
complex mixture that is suspected of affecting human sperra (such as
an occupational exposure).
Since little is known of the quantitative relationships
between induced sperm abnormalities and heritable genetic damage,
indirect methods are needed to assess the genetic risk to offspring
of men who show induced sperra anomalies. By combining data from
short-terra mutagen bioassays (e.g., Salmonella/microsome assay,
mammalian somatic cell mutation assays), which may demonstrate
mutagenic potential, with data from animal and human sperra assays,
which may demonstrate activity in the testes, we may be able to
evaluate whether a mutagen (or carcinogen) is active in the testes.
For select agents, a more objective assessment of germ cell
mutagenicity may be required. This could be done using various
-------
SPERM ASSAYS AS INDICATORS OF DYSFUNCTION
503
murine Fi generation bioassays (e.g., heritable chromosomal
translocation, dominant skeletal nutations, and heritable sperm
abnormalities) to quantify the relationships between heritable
consequences and chemically induced sperm anomalies. The combined
use of data from animal and human sperm assays, short-term _in vitro
bioassays, and murine F^ generation mutational bioassays may
provide a feasible approach to genetic risk assessment in men
exposed to agents that cause sperm anomalies.
ACKNOWLEDGMENTS
I wish to thank G. Watchmaker and L. Gordon for their
technical input and M. Mendelsohn and B. Ishida for their incisive
suggestions in the preparation of this manuscript. I am especially
thankful to the members of the EPA Gene-Tox panel on human and
sperm assays for access to some of their data and results presented
in this manuscript. This work was performed under the auspices of
the U.S. Department of Energy, under contract no. W-7405-ENG-48,
and the U.S. Environmental Protection Agency, under a Pass-Through
Agreement and under contract no. 79-D-X0826.
REFERENCES
Beatty, R.A. 1972. The genetics of size and shape of
spermatozoon organelles. In: The Genetics of the
Spermatozoon: Proceedings of an International Symposium.
R.A. Beatty and S. Gleucksohn-Waelsch, eds. University of
Edinburgh: Edinburgh, pp. 97-115.
Brozek, C. 1970. Proportion of morphologically abnormal
spermatozoa in two inbred strains of mice, their reciprocal
and F2 crosses and backcrosses. Acta Biol. Cracov. (Ser.
Zool•) 13:189-198.
Bruce, W.R., and J.A. Heddle. 1979. The mutagenic activity of 61
agents as determined by the micronucleus, Salmonella, and
sperm abnormality assays. Can. J. Genet. Cytol. 21:319-334.
Coordinating Committee. 1978. International program for the
evaluation of short-term tests for carcinogenicity. Mutation
Res. 54:203-206.
Czeizel, E., M. Hancsok, and M. Viczian. 1976. Examination of the
semen of husbands of habitually aborting women. Orvosi
Hetilap 108:1591-1595.
-------
504
ANDREW J. WYROBEK
Furuhjelm, J., B. Jonson, and C.G. Lagergren. 1962. The quality
of human semen in spontaneous abortion. Int. J. Fertil.
7:17-21.
Hugenholtz, A.P., and W.R. Bruce. 1979. Radiation-induced
heritable sperm abnormalities in mice. Environ. Mutagen.
1 :127-128.
Joel, C.A. 1966. New etiologic aspects of habitual abortion and
infertility, with special reference to the male factor.
Fertil. Steril. 3:374-380.
Kapp, R.W. 1979. Detection of aneuploidy in human sperm.
Environ. Hlth. Perspect. 31:27-31.
Kneer, M. 1957. Der habituelle Abort. Dtsch. Med. Wochenschr.
82:1059-1061.
Krzanowska, H. 1972. Influence of Y chromosome on fertility in
mice. In: The Genetics of the Spermatozoon: Proceedings of
an International Symposium. R.A. Beatty and S. Gleucksohn-
Waelsch, eds. University of Edinburgh: Edinburgh. pD.
370-386.
Sotomayor, R.E. 1979. Spermatid head abnormalities in
translocation heterozygotes from EMS- or CPA-treated sires.
Environ. Mutagen. 1:129. (abstr.)
Staub, J.E., and B.E. Matter. 1976. Heritable reciprocal
translocations and sperm abnormalities in the Fj offspring
male mice treated with triethylenemelamine (TEM). Arch,
fuer Genet. 49:29-41.
Takala, M.E. 1957. Studies on the seminal fluid of fathers of
congentially malformed children (199 sperm analyses). Acta
Obst. et Gynec. Scandinav. 36:29-41.
Topham, J.C. 1980a. Chemically-induced transmissible
abnormalities in sperm head shape. Mutation Res. 70:109-114.
Topham, J.C. 1980b. The detection of carcinogen-induced sperm
head abnormalities in mice. Mutation Res. 69:149-155.
Waters, M.D. 1979. The Gene-Tox program. In: Mammalian Cell
Mutagenesis: The Maturation of Test Systems, Banbury Report
2. A.W. Hsle, J.P. O'Neill, and V.K. McElheny, eds.
Cold Spring Harbor Laboratory: Cold Spring Harbor, NY.
pp. 449—457.
-------
SPERM ASSAYS AS INDICATORS OF DYSFUNCTION
505
Wyrobek, A.J. 1979. Changes in mammalian sperm morphology after
x-ray and chemical exposures. Genetics (Suppl.) 92 : s 105-s 119.
Wyrobek, A.J, (in press). Methods for human and murine sperm
assays. In: Short-Term Tests for Chemical Carcinogens. H.F.
Stich and R.H.C. San, eds. Springer Verlag: New York.
Wyrobek, A.J., J. Brodsky, L. Gordon, G. Watchmaker, and E. Cohen,
(in press a). Sperm studies in anesthesiologists.
Wyrobek, A.J., and W.R. Bruce. 1975. Chemical induction of sperm
abnormalities in mice. Proc. Natl. Acad. Sci. USA
7 2:4425-4429.
Wyrobek, A.J., and W.R. Bruce. 1978. The induction of sperm-shape
abnormalities in mice and humans. In: Chemical Mutagens:
Principles and Methods for Their Detection, Vol. 5. A.
Hollaender and F.J. de Serres, eds. Plenum Press: New York,
pp. 257-285.
Wyrobek, A.J., J.G. Burkhart, M.C. Francis, L.A. Gordon, R.W. Kapp,
G. Letz, H.V. Mailing, J.C. Topham, and M.D. Wharton. (in
press b). An evaluation of the mouse sperm morphology assay
and sperm assays in other animals: a report for the Gene-Tox
program. Mutation Res., Rev. Genet. Toxicol.
Wyrobek, A.J., J.G. Burkhart, M.C. Francis, L.A. Gordon, R.W. Kapp,
G. Letz, H.V. Mailing, J.C. Topham, and M.D. Wharton. (in
press c). Chemically induced alterations of spennatogenic
function in man as measured by semen analysis parameters: a
report for the Gene-Tox program. Mutation Res., Rev. Genet.
Toxicol.
Wyrobek, A.J., M. DaCunha, L. Gordon, G. Watchmaker, 3. Gledhill,
B. Mayall, J. Gamble, and M. Meistrich. 1980. Sperm
abnormalities in cancer patients. Cancer Res. 21:196.
Wyrobek, A.J., and B.L. Gledhill. (in press). Human semen assays
for workplace monitoring. In: Proceedings of the Workshop on
Methodology for Assessing Reproductive Hazards in the
Workplace, NIOSH.
Wyrobek, A.J., L. Gordon, and G. Watchmaker. (in press d). The
effects of 17 chemical agents including 6 carcinogen/non-
carcinogen pairs on sperm shape abnormalities in mice.
In: Short-term Tests for Carcinogens: Report of The
International Collaborative Program. F.J. de Serres and J.
Ashby, eds. Elsevier/North Holland: Amsterdam.
-------
506
ANDREW J. WYROBEK
Wyrobek, A.J., G. Watchmaker, K. Wong, and D. Moore II.
(in press e). Sperm abnormalities in carbaryl exposed
workers. Environ. Hlth. Perspect.
-------
ASSESSING CARCINOGENIC RISK RESULTING FROM COMPLEX
MIXTURES
Roy E. Albert
Institute of Environmental Medicine
New York University Medical Center
New York, New York
The evaluation of carcinogenic risks from complex mixtures,
in contrast to pure substances, adds a dimension of uncertainty to
a situation already frought with uncertainties and complexities.3.
The uncertainties of evaluating complex mixtures are likely to be
lost in the overall uncertainties of the risk assessment process.
There is considerable controversy about the risk assessment of
carcinogenic substances these days. Some liken the assessment of
carcinogenic risks to the theater, in that a willing suspension of
disbelief is required. To others, like myself, the assessment of
carcinogenic risks reflects Mark Twain's definition of work: "It
is something which a body is obliged to do." I believe that the
risk assessment of carcinogens is something that one is obliged to
do in a regulatory setting in order to make regulatory judgements
in as rational a manner as possible. The responsibility of those
doing the risk assessment is to make the most sensible use of
current science, paying proper regard to caucioning those who are
making the decisions about the uncertainties in the process.
One of the problems with the assessment of carcinogenic risks
either of pure substances or complex mixtures is that it is a
relatively new field. Before 1970, the only well-established risk
assessment area involved ionizing radiation. It is worth noting
that the standards for permissible exposure to ionizing radiation
were set not on the basis of risk assessment but on the
traditional basis of applying safety factors to observed levels of
carcinogenic responses, particularly to cancer induction in the
bone, lung, and bone marrow. It was in the field of ionizing
radiation that the dominant concept of risk assessment emerged,
namely, the linear non-threshold extrapolation model. This model
was based on the correlation between carcinogenesis and
507
-------
508
ROY E. ALBERT
mutagenesis and on the recognition that the linearity of dose
response is applicable to mutagenesis and consistent with
linearity in the dose response for leukemia, particularly among
Japanese atom-bomb survivors.
The linear non-threshold concept has proven to be a powerful
tool for predicting small excess risks of cancer at levels that
can't be confirmed or denied by direct observation either in
animals or in epidemiologic follow-up studies of exposed
populations. Both animal studies and the epidemiologic follow-up
methods are too insensitive to detect levels of risk below a few
percent, levels which are far too high to be tolerated willingly.
Consequently, the linear non-threshold concept of dose response
can be considered a two-edged sword. It provides the impetus Co
regulate, because we are essentially accepting the position that
any exposure, however small, will produce an excess of cancer, but
it leaves us in a quandary as how to regulate, because it puts us
in a position of trying to figure out which excess risks are
tolerable under given circumstances. The non-threshold concept
was first incorporated into the regulation of carcinogens in the
Delaney amendment, which bans food additives that show evidence of
carcinogenic action. The Delaney amendment reflects the
scientific concept that there is no safe level of carcinogen
exposure and the political judgement that no excess risk of
cancer from food additives is tolerable.
In the 1970's the regulatory movement was accelerated with
the formation of the U.S. Environmental Protection Agency (E?A),
the Occupational Safety and Health Administration (OSHA), and the
Consumer Protection and Safety Commission (CPSC). Many laws
dealing with the regulation of carcinogens were formulated using
different approaches to the control of carcinogens: the Clean Air
Act calls for protecting everyone with a margin of safety—clearly
impossible under a non-threshold concept; the Federal Insecticide,
Fungicide, and Rodenticide Act calls for weighing risks and
benefits; the Toxic Substances Control Act refers to making
regulatory judgments based on reasonable risk; the Occupational
Safety and Health Act calls for using "best available technology"
combined with economic considerations. A number of other laws
call for the use of "best available technology."
These different regulatory approaches use risk assessment in
different ways and to different degrees. Some of them simply call
for the characterization of an agent as a carcinogen with the
application of "best available technology," thereby applying
considerable economic and technical pressure. Others call for
weighing risks and benefits, whereby an agent is not only
characterized as a carcinogen, but its public health impact is
estimated as a basis for evaluating the benefits and costs of
r egulat i on.
-------
ASSESSING CARCINOGENIC RISK FROM COMPLEX MIXTURES
509
In 1976, EPA developed its own guidelines for Che risk
assessment of carcinogens in response to a great deal of criticism
of ics approach to regulating pesticides. The EPA guidelines for
risk assessment are based on the evidence approach. Risk
assessment was viewed as an exercise trying to answer two
questions: First, is the agent likely to be a human carcinogen?
Second, if the agent is a carcinogen, how much cancer is likely to
be produced? The former question requires a qualitative judgment,
the latter a quantitative one.
The qualitative judgement is based on the sum of evidence
concerning the quality, scope, and kinds of tumorigenic responses.
The judgements can range from the characterization of an agent
with strong evidence, based on good epidemiologic data backed by
animal data, to the opposite end of the spectrum where only a
single test, a single strain, a single sex, or a single species
may have a marginal response. The guidelines regard short-term
tests as being in a suggestive category when they stand alone, yet
recognize their great value in providing support for animal
bioassays or epidemiologic evidence of carcinogenicity. It is
difficult to know when a scientific approach achieves sufficient
consensus to provide the basis for regulatory action. So far,
short-term bioassays haven't achieved this stature. I am not sure
when they will. A tremendous amount of work is certainly aimed in
that direction. It may very well be that within a forseeable
time, short-term bioassays will have sufficient stature to provide
a very strong impetus toward regulation on their own merits.
The quantitative assessment looks at how much cancer is
likely to be produced, and assumes some background knowledge about
exposure. From our experience in the Carcinogen Assessment Group,
I think that exposure assessment is one of the weakest areas in
EPA. I am sure that this situation also exists in other agencies.
A quantitative assessment is based on an estimate of the exposure
as well as the use of models for extrapolation from high doses to
low doses; in many cases it also involves extrapolation from
animals to humans. The U.S. Environmental Protection Agency
started off in irs guidelines calling for the use of more than one
extrapolation model. Over the last four years, the assessments
have focussed on the linear non-threshold extrapolation model.
While no one model has a resoundingly solid scientific foundation,
the linear non-threshold model has more scientific plausibility
than other models and also tends to provide conservative estimates
of risk. This model has been used almost exclusively in EPA's
Carcinogen Assessment Group over the past four years. Given the
uncertainties in the quantitative assessment of risk, the linear
non-threshold model can be regarded as providing a plausible upper
limit of estimated risk.
-------
510
ROY E. ALBERT
This assessment approach has been used in EPA for Che last
four years on about 100 agents. More recently, an Interagency
Regulatory Liaison Group looked at the assessment of carcinogenic
risks and formulated a position document. The agencies
represented were EPA, OSHA, CPSC, and the Food and Drug
Administration. The position taken in this interagency document
can be read to be essentially consistent with the approach chat I
already mentioned. Realistically, however, it reflects
considerable resistance to quantitative risk assessment. The
document, to a considerable extent, gives only qualified supporc
to quantitative risk assessment and can be cited as support by
agencies that wish to do or not do quantitative assessment,
depending on their point of view. More recently the Federal
Regulatory Council published a cancer policy: its approach to the
assessment of carcinogens supports the approach that has been
taken by the EPA and IRLG, but with more qualification of the
quantitative aspects of risk assessment than has been given by
EPA.
The assessment of complex mixtures is no different, at least
in principle, to the pattern that I have described already. There
is certainly nothing unusual about complex mixtures of
carcinogens. After all, the first demonstration of chemical
carcinogenesis in animals was in the 1920's by the Japanese
Yamagiwa and Ichikawa. They painted coal tar on rabbits' ears and
first displayed the action of chemical carcinogens. One hundred
and fifty years earlier the first epidemiologic observations on
cancer induced by environmental chemicals dealt with scrotal
cancer caused by soot in chimney sweeps. Soot is a complex
mixture. From the beginning, the field of chemical carcinogenesis
has been firmly embedded in the problem of complex mixtures.
Probably the most important exercise that has been undertaken
by the EPA Carcinogen Assessment Group involving complex mixtures
has concerned diesel exhaust parciculaces. The approach here has
been to peg the evaluation of diesel particulates to some fairly
solidly established epidemiologic dose-response data for agents
Chat are chemically similar Co diesel exhausC, namely, lung cancer
among coke-oven workers, cigarette-smoke-induced lung cancer, and
lung cancer in workers who have used coal tar and asphalt roofing
materials. The approach has been Co develop a cross comparison of
the potency of diesel particulate exhaust and the other three
materials on the basis of short-term in vitro assays, mouse skin
tumorigenesis studies, and intratracheal intubation studies in
hamsters. If we can bracket the carcinogenic potency of diesel
exhaust with respect to Che materials that have been observed to
produce lung cancer in humans, we can use the human epidemiology
data as surrogates for diesel particulates data in quantitative
risk assessment. I think that this approach covers about as much
as can be done in terras of dealing with complex mixtures of chis
-------
ASSESSING CARCINOGENIC RISK FROM COMPLEX MIXTURES
511
sort. The 9ame principle could be applied to the comparison of
complex mixtures with pure materials using animal studies;
certainly the use of short-term assays, both in vitro and in vivo,
is particularly important when complex mixtures have a variable
c omposi t ion.
The diesel story is certainly not yet finished. The research
program is still in the stage of producing data. In addition to
the efforts involving cross comparisons of different materials
with diesel particulates for carcinogenic potencies, there are
also studies dealing with exposure Co the diesel exhaust per se.
One should wait until much of the experimental data is in hand
before making any firm quantitative risk assessments.
The risk assessment process is receiving its major challenge
in the quantitative area, although the publicity over the recent
saccharin epidemiology studies is probably going to provide
ammunition to those who are opposed to the use of assessment even
in the qualitative area, namely, the applicability of rodent
bioassays to human cancers. I must express ray dismay at the way
in which the publicity about the saccharin studies failed to
highlight the fact that the duration of exposure to saccharin is
completely inadequate to permit a proper characterization of its
carcinogenicity in humans. It is perfectly clear that even if one
were dealing with an agent that produced cancer, if the effects on
human populations were evaluated before the end of the latent
period, no effects would be found.
But the most serious challenges to risk assessment come in
the quantitative area. I have mentioned the kind of objections
that have been made by some agencies within the IRLG; also, the
Federal Regulatory Council tended to downplay quantitative risk
assessment on the basis of its uncertainty. Industry doesn't
particularly like the EPA brand of quantitative risk assessment
because of the use of the conservative linear non-threshold
dose-response model. I think the position of industry is that if
one is going to do quantitative risk assessment, which they think
is not a bad idea, it would be better to use models that are less
conservative, namely that show a curvilinear dose response and
yield much lower risk estimates. The Occupational Safety and
Health Administration is against quantitative risk assessment,
because it feels that the law doesn't require it and it would only
interfere with its regulatory program. There will be a decision
made on this question by the Supreme Court in the case of benzene.
The National Academy of Science, in a yet-to-be-released document
on how to regulate pesticides, also takes a crack at quantitative
risk assessment. This particular committee would go so far as to
agree that one could extrapolate from animals to humans, as a
basis for characterizing carcinogens in terms of relative potency,
but they balk at the use of extrapolation models to predict
-------
512
ROY E. ALBERT
responses at low levels of exposure. It has been pointed oat, to
no effect, that several other committees in the National Academy
of Science have used the linear non-threshold extrapolation model
for estimating risks. Probably the most vocal group to express
opposition at the present time to quantitative risk assessment is
the environmentalists. In a joint comment by the Environmental
Defense Fund and the Natural Resources Defense Council to EPA on
its newly proposed Air Cancer Policy, they say that "EPA must
abandon its arbitrary and unlawful reliance on quantitative risk
assessment methods. Quantitative risk estimates can be wrong by a
factor of five million times or more and thus are too unreliable
and imprecise to play a rational role in determining the levels of
control applied to a hazardous pollutant. Such estimates may have
a role in the grossest form of priority setting, but that will be
valid only if the agency much more explicitly recognizes the
uncertainties of the estimates and commits itself to not using
them in any way in subsequent standard setting."
That quote is quite a forceful expression of opinion. The
notion that quantitative risk estimates can be wrong by a factor
of five million stems from the National Academy of Sciences
Saccharin Report, which listed the risk estimates from saccharin
using a variety of extrapolation models. One can pick the
extrapolation model that yields the kind of result that one wants.
It is quite easy to pick a set of models that give estimates that
vary by many orders of magnitude. Although I do believe that it
is fair to say that the use of the linear non-threshold
extrapolation model may somewhat overestimate risk, it is not
likely to produce much of an underestimate of risk. Certainly,
for example, the estimation that one percent of cancer is due to
background ionizing radiation can't be too small by five million
times. That wouldn't be very likely.
Clearly, these expressed reservations about the use of
quantitative risk assessment reflect the very real weakness in the
scientific foundation underlying quantitative risk assessment.
This problem should be one of the major priority areas for
research related to carcinogen assessment. The regulation of
carcinogens is a public health program, and it is difficult to
mount a public health program without having some idea of the
likely benefits in relation to costs. Hopefully, we will make
rapid progress in gaining the knowledge necessary for doing risk
assessments with greater confidence.
-------
TECHNICAL REPORT DATA
iPlc'J^e read instructions on the rt verse bijCsv ':orr.piCling}
1 . REPORT NO.
EPA-600/9-82-004
Proceedings
4. TITLE AND SUBTITLE
Short-Terr. 3ioassP.ys in toe Analysis of '^oxplex
Environmental Mixtures 11
:?OR" DiT =
March 19S?"
6. PERFORMING ORGANIZATION CO C £
AUTHOR1S)
Michael D. Waters, Shahbeg 5. Sandhu, Joel 1er Lewtas
Huisingh, Larry Claxton, and Stephen N'esnow
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Genetic Toxicology Division
Environnenta1 Protection Agency
Research Triangle Park, North Carolina 27711
10. PROGRAM ELEMENT NO.
11. CONTRACT,GRAN T NO.
12. SPONSORING-AGENCY NAME AND ADDRESS
Office of Research and Development
Healtn Effects Research Laboratory
US Environmental Protection Agency
Research Triangle Park, NC 27711
(RTP,NC)
13 Tv©E OF REPORT AND PERIOD COVERED
Proceedings of Symposium
in. SPONSORING AGENCY CODE
Health Effects Research Lab.
EPA
Research Triangle Park, NC
15. SUPPLEMENTARY NOTES
16. ABSTRACT
'he present proceedings of the
Environmental Protection Agency1 s'Second
Symposium on the Application of Short-term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, held in Williamsburg, VA, March
4-7, 1980, :r,eludes 37 papers as wel1 as the Keynote Address. The papers are
divided according to the environmental media wherein snort-term oicassays are
appliec--axbient air, water, and soil--and the sources of environmental pollution-
mobile source emissions, stationary source emissions, and industrial emissions
.and effluents. A separate section is devoted to the problems of health hazard
and risk assessment.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS

b.IDENTIFIERS/OPEN ENDfcD TERV:
c. COSATl Field/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
!9. SECURITY C.ASS Hi,is Rezori
UNCLASSIFIED
; 21 . NO. OF PAGES
20 SECURITY CLASS iTIiispatt..
UNCLASSIFIED
?2 PRICE
Aia.
EPA Form 222 C — 1 (Rev. 4-77) p^evio-s editcn s c s s c l e "
-------