t A>,
United States
Environmental Protection
Agency
Health Effects
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S1-86/005 Nov. 1986
Project Summary
RECEIVED
NOV251986
ENVIRONMENTAL PROTECTION AGENCY
LIBRARY, REGION V
lexicological Testing of
Organic Substances from
Concentrated Drinking and
Waste Waters
Linda K. Snow and Bruce C. Casto
The U.S. Environmental Protection
Agency (EPA) has been given the re-
sponsibility for regulating the release of
toxic chemicals into the environment.
Accordingly, the EPA presently con-
ducts an extensive and comprehensive
research program to determine the ad-
verse effects of environmental factors
on human health. A considerable
amount of this research activity is di-
rected toward toxicological testing and
test development. The Health Effects
Research Laboratory of the EPA, in
Cincinnati, Ohio, has specifically fo-
cussed on the task of determining,
through the use of appropriate short-
term assays, the potential health haz-
ards of the complex mixtures of organic
compounds found in drinking and
waste waters. A battery testing ap-
proach is being employed by the EPA at
Cincinnati for the toxicological testing
of these waters, progressing from rou-
tine detection systems in microbial
cells to more complex systems in mam-
malian cells and subsequently to short-
term animal testing.
In support of the EPA's inhouse re-
search and testing, a study was con-
ducted employing a battery of five
short-term assays for the toxicological
testing of coded concentrated drinking
and waste water samples. The five as-
says performed were: (1) the in vivo
toxicity test, (2) Salmonella ty-
phimurium mutagenesis assay, (3) Sen-
car mouse skin biossay, (4) rat liver
focus assay, and (5) lung adenoma as-
say. The final report for this project
gives the results of the testing of a total
of 40 coded water samples in one or
more of the above assays.
This Project Summary was devel-
oped by EPA's Health Effects Research
Laboratory, Research Triangle Park, NC,
to announce key findings of the re-
search project that is fully documented
in a separate report of the same title
(see Project Report ordering informa-
tion at back).
Introduction
The U.S. Environmental Protection
Agency (EPA) has been given the re-
sponsibility for regulating the release of
toxic chemicals into the environment
under the Toxic Substances Control Act,
Resource Conservation and Recovery
Act, Federal Insecticide, Fungicide and
Rodenticide Act, and the Clean Air and
Water Acts. Along with the regulatory
aspects of the above Acts, the EPA con-
ducts an extensive and comprehensive
research program to determine the ad-
verse effects of environmental factors
on human health. The purpose of these
research programs is to provide infor-
mation that is essential to formulate
regulatory and compliance policies to
protect public health and the environ-
ment.
With the advent of the Toxic Sub-
stances Control Act and evidence show-
ing the relationship between environ-
mental agents and adverse health
effects (cancer, toxicity, developmental
and genetic defects, pulmonary dis-
ease, etc.), the EPA has directed a con-
-------
siderable amount of activity into toxico-
logical testing and test development.
Due to the large number of potentially
hazardous chemicals being introduced
into the environment each year, the
backlog of untested chemicals already
present, and the prohibitive cost and
time necessary to assess chronic effects
of these agents in long-term animal
testing, the EPA has been particularly
interested in the development, utiliza-
tion, and evaluation of time- and cost-
effective short-term in vivo and in vitro
test systems.
In accordance with the above objec-
tives, the Health Effects Research Labo-
ratory of the EPA, in Cincinnati, Ohio,
has specifically focussed on the task of
determining, through the use of appro-
priate short-term assays, the potential
health hazards of the complex mixtures
of organic compounds found in drink-
ing and waste waters. Preparation of
these mixtures in sufficient quantity for
lexicological testing has been made
possible by the development of concen-
tration (reverse osmosis) and isolation
(using various macroreticular resins)
techniques that concentrate water and/
or waste water contaminants by 400- to
1000-fold. The identification, characteri-
zation, and possible interaction of these
materials presents a formidable prob-
lem. Moreover, it is recognized that man
is exposed to these compounds as com-
plex mixtures and thus, that risk to man
associated with these exposures can
best be determined by toxicologies!
testing in animals or in vitro testing at
biological levels of organization that
may simulate the response in man.
It has been proposed that testing of
environmental agents proceed through
a battery of well-defined and controlled
short-term in vivo and in vitro assays
that are reproducible between laborato-
ries, possess a high degree of correla-
tion with the chronic in vivo effects of
various chemical compounds, and pro-
vide information relevant to effects ob-
served in man. Ideally, these assays
should generate no false-negative re-
sults and preferably only a low percent-
age of false-positives, be responsive to
diverse classes of chemical agents, and
lend themselves to quantitation. In ad-
dition, the assays should be relatively
inexpensive, rapid and tehcnically sim-
ple. Such a battery approach is
presently being employed by the EPA at
Cincinnati for the lexicological testing
of drinking/waste waters progressing
from routine detection systems in mi-
crobial cells to more complex systems
in mammalian cells and subsequently
to short-term animal testing.
Procedure
Four coded groups of concentrated
drinking or waste water samples were
screened for toxicological activity in
one or more of the bioassays. One
group - the Eight Sample Series - was
composed of concentrated water sam-
ples. Two groups - the Nine Sample
Series and Eighteen Sample Series -
were composed of organic acetone
samples. These latter samples were
concentrated by removal of the acetone
solvent followed by the addition of ster-
ile water, to bring the volume to one-
fifth that of the starting volume, and the
addition of sterile 10% Emulphor*
(1 volume to 4 volumes of sample). The
remaining group - the Five Sample
Series - consisted of wastewater con-
centrates dissolved in 2% Emulphor.
The in vivo toxicity test was used to
evaluate the toxicity of the Eight Sample
Series and the Five Sample Series in
CD-1 mice. For each study, CD-1 mice
(approximately 8 weeks of age) were ex-
posed to test samples (or control) for a
period of 30 days. Two dose groups, re-
ceiving 1X and 1/4X or 1/10X concentra-
tions, were employed per test sample.
The control group of mice received ster-
ile deionized water or sterile 2%
Emulphor. Each test or control group
contained a total of 20 mice - 10 male
and 10 female. For the Light Sample
Series, the test samples (or control)
were administered in the drinking
water, with only about a 3- or 4-day sup-
ply of water (approximately 35 mis/day/
cage) placed in the bottle twice weekly.
For the Five Sample Series, test sam-
ples (or control) were administered by
gavage, 0.3 ml per mouse, on each
Monday, Wednesday, and Friday of the
30-day treatment period (total of 13
doses).
A cage-side observation of each ani-
mal for overt signs of illness or death
was performed twice daily and a de-
tailed health status check was per-
formed on each mouse once a week at
the time of the weekly weighing. At the
end of the 30-day treatment period, all
mice were weighed, sacrificed and ne-
cropsied. Animals found dead or mori-
bund during the treatment period were
•Mention of trademarks or commercial products
does not constitute endorsement or recommenda-
tion for use.
also necropsied. The liver, lungs, kid-
neys, spleen, brain, and gonads of each
animal were examined, weighed, and
preserved in individually labelled speci-
men bottles containing 10% buffered
formalin.
The Salmonella typhimurium muta-
genesis assay was used to evaluate the
mutagenicity of the Nine, the Eighteen,
and the Five Sample Series. In each ex-
periment, a series of samples were
tested in strains TA98 and TA100 at four
sample doses in duplicate with and/or
without S9 activation. In each experi-
ment, samples were tested within an
eight-fold range with the highest sam-
ple dose not exceeding 400 uJ per plate.
Although the actual amount of test sam-
ple added during individual experi-
ments varied, the volume added to each
test plate was held constant in each ex-
periment by appropriate dilution of the
samples with 2% Emulphor. Test plates
were incubated in the dark at 37°C for 48
hours and revertants then counted
using an automatic colony counter. The
background lawn of the test plates was
examined with a stereomicroscope to
evaluate toxic effects.
The Sencar mouse skin bioassay was
used to evaluate the tumor initiating po-
tential of the Nine Sample Series. Six-
to-eight-week-old female Sencar mice
were dosed three times a week (M,W,F)
for tv u weeks intragastrically with
either 0.5 r.,1 of the test article or 0.5 ml
of the test article vehicle. Samples were
tested undiluted with one exception
(tested at 1/2X concentration). Positive
controls received a single intragastric
administration of urethane in deionized
water (500 mg/kg b.w.) at the time when
all other animals received the last dose
of the test article or test article vehicle.
Test and control groups contained 40 or
20 animals. Two weeks after the last ini-
tiating dose, TPA in acetone (1 u-g/
0.2 ml/animal) or acetone only (0.2 ml)
was applied topically to the shaved
backskin of the animals three times a
week for 20 weeks. The positive control
group receiving urethane was treated
similarly but with TPA only.
Starting weekly (at the time of the first
topical application of promoter or pro-
moter vehicle), the number, type, and
location of each backskin tumor was
recorded on an individual animal basis.
Such tumors were classified as papillo-
mas or carcinomas, and whether or not
they were cumulative tumors (persis-
tent for three consecutive weeks) or ac-
tual tumors (present for any length of
-------
time). A necropsy was performed on
each animal found dead or sacrificed
during the course of the study. Gross
observations on all external and inter-
nal lesions were recorded for each ani-
mal, including the number of skin papil-
lomas and carcinomas observed, and
the specified tissues were routinely ex-
amined and preserved in 10% neutral
buffered formalin. Representative 5|j,
Hematoxylin and Eosin stained slides
were produced from paraffin embedded
blocks of each gross lesion found at ne-
cropsy and these slides were subse-
quently examined and evaluated by a
veterinary pathologist.
The rat liver focus assay was used to
evaluate the tumor initiating potential
of the Nine and the Five Sample Series.
For each study, male Sprague-Dawley
rats were subjected to two-thirds partial
hepatectomies and then dosed once in-
tragastrically with either 3 ml of test
sample (undiluted), 3 ml of test sample
vehicle (2% Emulphor), or positive con-
trol (diethylnitrosamine [DEN], 50 mg/
kg body weight) precisely 24 hours after
the hepatectomies. Each test or control
group contained 10 animals. Seven to
nine days after the partial hepatec-
tomies, all rats received 0.05% pheno-
barbitol (PB) in their drinking water for a
period of 8 weeks after which they were
held for an additional week on drinking
water without PB and sacrificed. Cry-
otomy sections from the livers of the
animals were then stained and micro-
scopically evaluated for: (1) the number
of gamma glutamyl transpeptidase pos-
itive (GGT+) foci with nine or more nu-
clei per liver section; (2) the total num-
ber of morphometry points over areas
of GGT+ Zone 1 induction; (3) the total
number of morphometry points over
the section of liver evaluated; and
(4) the radius in cm of each GGT+ focus
with nine or more nuclei.
The lung adenoma assay was used to
evaluate the tumor initiation potential
of the Nine and the Five Sample Series.
For each study, Strain A/J mice (approx-
imately 8 weeks of age) were dosed in-
tragastrically with 0.25 ml of test sample
or sample vehicle (2% Emulphor) 3X per
week for 8 weeks. A positive control
group received a single per os adminis-
tration of urethane (10 mg/mouse). Two
dose groups, receiving 1X and 1/2X or
1/2X and 1/4X sample concentrations,
were employed per test sample. In addi-
tion to the vehicle and positive control
groups, an untreated control group was
also included to determine the inci-
dence of spontaneous lung tumors.
Each test or control group contained a
total of 40 mice (20 male and 20 female).
At the end of the 24th experimental
week (when mice were 32 weeks of
age), all mice were weighed, sacrificed
by C02 overdose, and subjected to
gross necropsy. Any animals found
moribund or dead were also necrop-
sied, except for those animals where
pathological examination was pre-
cluded due to severe postmortem auto-
lysis. Lung adenoma counts were made
for each animal necropsied. Lungs were
perfused by tracheal cannulation with
10% buffered formalin and placed in
separate labelled vials for subsequent
counting of the lung adenomas. Individ-
ual lung lobes were observed under a
dissecting microscope and adenoma
counts recorded on forms depicting the
individual lung lobes. Dual counts were
made by two technicians working inde-
pendently of each other's observations.
Differences between the two observers
were reconciled by a third independent
count. Randomly selected tumor-
bearing lungs from each group (10%)
were evaluated histologically for confir-
mation of adenomas and adenoma
counts.
Results and Discussion
In Vivo Toxicity Tests
In both the Eight Sample and Five
Sample Studies, data collected on the
mean body weights and organ weights
for the male and female mice in each
test or control group indicated that none
of the test samples produced any overt
toxic effects in the coded sample
groups, at either concentration, when
compared to the control groups. In
agreement with these findings, results
of the detailed health status checks re-
vealed no observable toxic effects.
Salmonella Typhimurium
Mufagenesis Assays
The Nine Sample Series was tested in
five separate experiments designed to
determine the stability of the concen-
trated drinking water samples (in 2%
Emulphor) during the concurrent test-
ing in the animal assays (Sencar, rat
liver focus, and lung adenoma). There
were no significant differences in muta-
genic response between drinking water
samples concentrated and stored in 2%
Emulphor at 4°C for 62 weeks, 33 weeks,
11 weeks, 3 weeks, and 2 weeks. The
relative mutagenic potency of the five
positive samples remained consistent;
four of the nine samples were negative
in all assays. It was also observed that
the presence of the S9 activation system
decreased or eliminated the positive re-
sponses observed without S9 addition.
Another experiment was designed to
compare the stability of those samples
stored in 2% Emulphor for 11 weeks and
those freshly concentrated from origi-
nal acetone samples stored at -20°C for
the same period. Similar mutagenic re-
sponses were observed between the
two series, thus indicating that the
mutagenic compounds in these nine
samples were stable under two differ-
ent storage conditions.
The Eighteen Sample Series was
evaluated for stability of mutagenic ac-
tivity in two separate experiments. Re-
sults indicated that the samples positive
in the first experiment lost some muta-
genic activity over the 34-week storage
period prior to retest, with a decrease
from seven to four positive samples for
TA98 and from five to four for TA100,
even though the sample dose range
was increased in the retest. Ten sam-
ples were negative in both experiments.
Except for one sample, it was again ob-
served that the presence of an S3 acti-
vating system decreased the positive
responses.
The Five Sample Series was evalu-
ated in a single experiment, designed to
compare the mutagenic activity of these
samples with activity observed in the
animal bioassays (in vivo toxicity, rat
liver focus, and lung adenoma). Three
of the five samples demonstrated a pos-
itive mutagenic response at one or
more sample doses. For all three posi-
tive samples, the presence of the S9 ac-
tivating system decreased or negated
the positive responses observed with-
out S9 addition.
Sencar Mouse Skin Bioassay:
Initiation/Promotion
The Nine Sample Series was tested in
three separate and independent experi-
ments. Survival and body weight rec-
ords indicated that the test compounds
did not cause acute or delayed toxicity
resulting in death or altered body
weights.
The gross scoring of skin tumors con-
ducted during these studies proved to
be quite accurate both in terms of the
ability to detect a tumor and to classify
them as papillomas or carcinomas.
Overall, histopathological evaluations
-------
revealed only seven cases where skin
lesions were scored incorrectly.
During these studies, a total of only
six carcinomas were detected. One was
detected grossly at week 30 and the re-
maining five were detected microscopi-
cally. The extremely small number of
detected carcinomas and their late ap-
pearance eliminated the use of skin car-
cinoma formation as a parameter for
evaluating the initiating potential of the
test articles. Time until a 50% incidence
of cumulative tumors was also elimi-
nated as a valid test parameter, since
the test article groups never attained
higher than a 33% incidence.
Control groups receiving test article
initiation followed by acetone promo-
tion or initiator and promotor vehicles
(2% Emulphor and acetone) showed no
skin tumor formation.
Animals in the positive control
groups (receiving urethane with TPA
promotion) showed cumulative tumor
incidences after 30 experimental weeks
of 65% to 70%. Animals in the control
groups receiving initiator vehicle (2%
Emulphor) and TPA promotion demon-
strated a weaker response for cumula-
tive tumor incidences at 30 weeks (9%
to 36%).
Initiation with any of the nine test arti-
cles followed by promotion with TPA
yielded tumor incidences in animals
(cumulative skin tumors by 30 experi-
mental weeks) that were not signifi-
cantly different from the vehicle/TPA
controls (range of 8% to 33%).
When the average latency period for
the appearance of all cumulative tu-
mors on all tumor bearing animals
within, a group was calculated, it was
determined that none of the test articles
caused a reduced latency compared to
the TPA-only controls.
Rat Liver Focus Assays
Results in both the Nine Sample and
Five Sample Studies demonstrated a
significant difference between the posi-
tive control and vehicle control groups
for three designated test parameters
(number of GGTase positive foci/cm2 of
liver evaluated, number of GGTase pos-
itive foci/cm3 of liver evaluated, and per-
centage of fiver evaluated exhibiting
GGTase positive foci). When, however,
the test groups in each study were com-
pared to the vehicle control groups,
there was no significant difference be-
tween any of the groups and the vehicle
control.
The use and significance of the per-
centage of liver evaluated exhibiting
GGTase positive zonal induction (Zl) as
a parameter to measure initiation by the
test articles was questionable in both
studies, since there were no significant
differences in Zl between the positive
and vehicle control groups in either
study. In the Nine Sample Study, how-
ever, there were significant differences
in Zl between the vehicle control and
some test groups and between the pos-
itive control and some test groups. In
the Five Sample Study, none of the test
groups showed a significant increase in
the amount of GGTase positive zonal in-
duction. Results of both studies sug-
gested that the ability of a compound to
initiate GGTase positive foci does not
correlate with its ability to initiate pro-
duction of GGTase positive zonal induc-
tion.
In the Nine Sample Study, the percent
of deaths within one week of the intra-
gastric administrations of the test arti-
cles indicated possible acute toxicities
for five of the compounds. No sample
toxicity was observed in the Five Sam-
ple Study.
Lung Adenoma Assays
With the Nine Sample Series, the inci-
dence of lung adenomas was 100% for
both the male and female positive con-
trol groups, with cumulative lung nod-
ule indexes of 11.78 and 11.66, for
males and females, respectively. Ani-
mals in the untreated and vehicle con-
trol groups showed a much weaker re-
sponse, with tumor incidences ranging
from 5 to 10% and cumulative lung nod-
ule indexes from 0.05 to 0.10.
The incidence of lung adenomas in
animals treated with the nine coded
samples varied from 0 to 30% among
males and 0 to 28% among females
while the average number of tumors per
animal ranged from 0 to 0.35 in the
males and 0 to 0.33 in the females.
When these responses were compared
to those observed in the untreated and
vehicle control groups, there were no
indications that any of these samples
were initiators of lung adenomas.
With the Five Sample Series, the inci-
dence of lung adenomas was 30% for
the male positive control group and
35% for the female group with a cumu-
lative lung nodule index of 0.40 for both
males and females. These responses
were much lower than those observed
in the Nine Sample Study. Animals in
the untreated and vehicle control
groups yielded tumor incidences rang-
ing from 5.6 to 21% and cumulative lung
nodule indexes from 0.06 to 0.42.
The incidence of lung adenomas in
animals treated with the five coded
samples varied from 10 to 35% among
males and 0 to 25% among females
while the average number of tumors per
animal ranged from 0.10 to 0.35 in the
males and 0 to 0.30 in the females.
When these responses were compared
to those observed in the untreated and
vehicle groups, there were no indica-
tions that any of these samples were
initiators of lung adenomas.
Correlation between gross adenoma
counts and histological confirmation
was 89% in both studies. None of the
test samples in either series exhibited
any overt toxicity, as indicated by
changes in body weight or mortality
data.
Conclusions and
Recommendations
Four coded groups of concentrated
drinking or waste water samples, identi-
fied as the Eight Sample Series, the
Nine Sample Series, the Eighteen Sam-
ple Series, and the Five Sample Series,
were screened for toxicological activity
in one or more of the following
bioassays: (1) in vivo toxicity test,
(2) Salmonella typhimurium mutagene-
sis assay, (3) Sencar mouse skin
bioassay, (4) rat liver focus assay, and
(5) lung adenoma assay.
Positive results were obtained with
the Salmonella screen (strains TA100
and TA98). Several mutagenic samples
were present in each of the three sam-
ple groups - Nine, Eighteen, and Five
Sample Series - tested with this assay.
However, it was observed that the rela-
tive activity of these samples, at the
concentrations tested, was somewhat
low as compared to most known muta-
genic compounds. It was also observed
that the presence of the S9 activation
system decreased or eliminated the
positive responses observed without S9
addition. Experimental results indicated
that some of the concentrated water
samples lost activity during storage at
4°C in the Emulphor vehicle (Eighteen
Sample Series) while other samples re-
tained their activity (Nine Sample
Series). It was also determined that the
mutagenic compounds in the Nine
Sample Series were stable under differ-
ent storage conditions, i.e., concen-
trated and stored in 2% Emulphor for
11 weeks at 4°C versus freshly concen-
trated from original acetone samples
stored at -20°C for the same period.
-------
No carcinogenic activity was ob-
served with any of the coded samples
tested in the Sencar, rat liver focus, or
lung adenoma assays, nor was any
toxic activity observed with coded sam-
ples tested in the in vivo toxicity test.
The weak activity of samples positive in
the Salmonella assay suggests that the
concentrations of mutagens/carcino-
gens in the samples might have been
too low for detection in the animal
bioassays. However, in view of the toxic
responses observed in the rat liver
focus assay, additional concentration,
using the same methods, might not be
productive. Rather, it might be neces-
sary to fractionate the sample concen-
trates, in an attempt to separate the
toxic components of these complex
mixtures from their mutagenic/carcino-
genic components. Sensitive short-term
tests such as the Salmonella screen
might be useful in guiding these frac-
tionation procedures.
In lieu of fractionation and additional
concentration, potentially more sensi-
tive in vitro and in vivo systems may be
employed for identifying the toxic/mu-
tagenic/carcinogenic compounds in
water samples. Such systems for de-
tecting cell transformation in vitro in-
clude: the hamster embryo focus or
colony assay, the mouse fibroblast
C3H10T1/2 CL8 assay, and the Syrian
hamster embryo/SA7 enhancement
assay. For measurement of toxic activ-
ity in vitro, inhibition of growth of pig
aortic endothelial cells may be used.
Some in vivo systems, which may prove
to be more sensitive for measurement
of activity in concentrated drinking and
waste waters, include: the sex-linked re-
cessive lethal test in Drosophila, the de-
tection of sister chromatid exchanges
(SCEs) in the liver, lung or other major
organs of mice, and the rat neonatal
survival assay.
The inclusion of this latter in vivo
assay is particularly important, since
developmental toxicity can result from
nonmutagenic/noncarcinogenic mecha-
nisms. Moreover, the test parameters in
the rat neonatal survival assay are not
limited to the assessment of gross ab-
normalities in structural development
but also include endpoints such as adult
fertility, stillbirths, pre- and post-natal
growth retardation and functional
deficits in the offspring.
This Project Summary was submitted
in fulfillment of Contract No. 68-03-
1840 by Environmental Health Research
and Testing, Inc. under the sponsorship
of the U.S. Environmental Protection
Agency. This summary covers a period
from February 1984 to September 30,
1985, and work was completed as of
September 30, 1985.
Linda K. Snow and Bruce C. Casto are with Environmental Health Research &
Testing. Inc.. Resarch Triangle Park, NC 27709.
Robert G. Miller is the EPA Project Officer (see below).
The complete report, entitled "Toxicological Testing of Organic Substances from
Concentrated Drinking and Waste Waters." (Order No. PB 86-219 136/AS;
Cost: $16.95, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S1 -86/005
^ ""^>,^.MTVfU.S.PBSUGt
•-'-c--:- n 3 ? -
-•-> "' l-
0000329 PS
U 3 ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
230 S DEARBORN STREET
CHICAGO It 60604
------- |