-------
176 EUGENE SAWICKI
Since there are such a tremendous number of organic
compounds in our environment, a much better separation of
these compounds is desirable before their analysis. One
method is by increasing the reliability and resolution of
capillary gas chromatography. The possibilities as reported
by G. Grob and K. Grob are shown in Table 2.
Table 2
Resolution of GC Peaks
(Water extract - OV-1 Columns - Grob and Grob)
I.D.
3
60
35
(m
X
X
X
x mm)
2
0.6
0.28
Co
Packed
Glass
Glass
lumn
capillary
capillary
No. Peaks
118
320
490
A reliable routine automated system of qualitative and
quantitative analysis of the highly complicated mixtures in
our polluted environment could be developed from a promising
method, termed HISLIB, that compares combined gas chromato-
graphic/mass spectrometric profiles of new environmental mix-
tures with historical libraries of GC/MS data on related mix-
tures (34). The presence of several components is established
by matching retention indexes and mass spectra after removal
of column bleed, contamination, and other types of background
and after resolution of overlapping GC components. The system
is quantified with the help of internal standards by comparing
relative concentrations of components.
The simplification and standardization of routine screen-
ing methods for the key genotoxicants are desirable. These
key chemicals are in high production, are present in the
atmosphere in high concentration, or are carcinogenic to humans.
Examples of some of the genotoxicants carcinogenic to humans
are given in Table 3, and some of the carcinogens found in
the polluted atmosphere are listed in Table 4.
Although a fairly large amount of information is available
on the carcinogenic and mutagenic activities, the reaction with
DNA, and the metabolic properties of the atmospheric carcino-
gens, little work has been done with the various families of
-------
ATMOSPHERIC GENOTOXICANTS
177
Table 3
Genotoxicants Carcinogenic to Humans
Genotoxicant
4-Aminobiphenyl
Arsenic
Asbestos
Auramine
Benzene
Benzidine
Bis-chloromethyl
ether
Cadmium
Chimney soot
Chloroprene
Chromate
Coal hydrogenation
vapors
Coal tar pitch
Coke oven
effluents
Creosote oils
Cutting oils
Hematite
Isopropyl oil
Mineral oil
Mustard gas
S-Naphthylamine
Nickel
Petroleum wax
Radium
Radon and radon
daughters
Rubber plant
effluents
Shale oil
Soots, tars and
oils
Vinyl chloride
Wood dust
Target
Bladder
Lung
Lung, pleural
cavity
Bladder
Bone marrow
Bladder
Lung
Prostate
Scrotum
Lung, skin
Lung
Skin
Lung, skin
Bladder, lung
Lung, skin
Lung, scrotum
Lung
Nasal cavity,
larynx
Scrotum
Lung, larynx
Bladder
Nasal cavity,
lung
Scrotum
Lung
Lung
Brain
Lung
Lung, scrotum
Brain, liver,
lung
Nasal cavities
Pathway
Inhalation,
Inhalation
Inhalation
Inhalation,
Inhalation,
Inhalation,
Inhalation
Inhalation,
Skin
Inhalation,
Inhalation
Skin
Inhalation,
Inhalation
Inhalation,
Inhalation,
Inhalation
Inhalation,
Skin
Inhalation,
Inhalation,
Inhalation
Skin
Inhalation
Inhalation
Inhalation
Inhalation
Inhalation,
Inhalation,
Inhalation
oral
oral, skin
skin
oral, skin
oral
skin
skin
skin
skin
skin
skin
oral, skin
skin
skin
-------
178
EUGENE SAWICKI
Table 4
Carcinogenic Air Pollutants
Compound1
Acrylonitrile
Aldrin
Anthanthrene
Arsenic (III)
Asbestos
BaA
BaCAR
BaP
BbFT
BcACR
BcCAR
Benzene
Benzyl chloride
Be (II)
BeP
BHC
Bis-chloromethyl ether
BjFT
C
Carbon tetrachloride
Cd (II)
Chloroform
Chloromethyl methyl ether
Chloroprene
Chromium (VI)
Remarks
C,h(inh),r(oral)
C,m,r(oral)
C.m(skin)
C,h(oral,skin)
C,h,r(inh)
C,m(oral,skin,sc)
C,m(sc,skin),r(skin)
C,9 species(it,oral,skin)
C,m(sc,skin)
C,m(skin),r(bi)
C
C,h(inh),m(inh),r(oral,inh)
C,r(sc)
C,mk(inh),r(inh),rb(iv)
C,m(skin)
C,m(oral)
C,h(inh),m(inh,sc,skin)
C.m(skin)
C,m(sc,skin)
C,ha,m,r(inh,oral)
C,h(inh),r(im,sc)
C,m(oral),r(oral,sc)
C.h(inh),m(skin),r(inh,skin)
C,h(inh,skin)
C,h(inh),r(im,ipl)
1 A. = anthracene, ACR = acridine, B = benzo, BHC = a-benzene
hexachloride, C = chrysene, CAR = carbazole, DB = dibenzo,
DDD = l,l-dichloro-2,2-bis(p-chlorophenyl) ethane, DDE = 1,1-
dichloro-2,2-bis(p-chlorophenyl) ethylene, DDT = 1,1,1-
trichloro-2,2-bis(p-chlorophenyl) ethane, FT = fluoranthene,
IND = indeno, P = pyrene, and PEP = pentaphene. Thus, BaP =
benzo(a)pyrene while IND 1,2,3-cdP - indeno(l,2,3-cd) pyrene.
zbi = bladder implantation, C = carcinogenic, d = dog, gp =
guinea pig, h = human, ha = hamster, im = intramuscular,
inh = inhalation, ip = intraperitoneal, ipl = intrapleural,
it = intratracheal, iv = intravenous, m = mice, pn = prenatal
exposure following iv injection in pregnant female, r = rat,
and sc = subcutaneous.
-------
ATMOSPHERIC GENOTOXICANTS
179
Table 4 (continued)
Dimethylnitrosamine
Dimethyl sulfate
p-Dioxane
Ethylene dibromide
Hematite
Heptachlor
IND 1,2,3,-cdP
Kepone
Lead (II)
Lindane
Methyl iodide
Mirex
Nickel (III)
Perchloroethylene
Propylene oxide
Quinoline
Styrene oxide
o-Toluidine
p-Toluidine
Trichloroethylene
Vinyl chloride
Vinylidene chloride
DBaeP
DBahA
DBahACR
DBahP
DBaiP
DBajACR
DBalP
DBcgCAR
DBh.rstPEP
ODD
DDE
DDT
Dieldrin
Diethylnitrosamine
1,1-Dimethylhydrazine
C, 16 species
C,r(inh,pn,sc)
C.gp(oral),r(oral)
C,m(oral),r(oral)
C.m(inh)
C,m(oral)
C,m(skin)
C,m,r(oral)
C,m,r(oral)
C,m(oral)
C,m(ip),r(sc)
C,m(oral)
C,h(inh),m,r(im)
C,m(oral)
C,r(sc)
C,r(oral)
C,m(skin)
C.m.r
C,m
C,m(oral)
C,h(inh)
C,m(inh)
C,m(sc,skin)
C,6 species(oral,it,sc,skin)
C,m(sc,skin)
C,m(im,skin),r(im,skin)
C,ha(sc,skin),m(sc,skin)
C,m(sc,skin)
C,m(sc)
C,d(bi),ha(inh,it),m(bi,im,
ip,iv,oral,sc,skin),r(im,sc)
C,m(sc)
C,m
C,m
C,m(oral)
C,m(oral)
C, 16 species
C,m(oral)
-------
180 EUGENE SAWICKI
genotoxicants that usually are associated in the polluted
atmosphere (Table 5). Analytical methodology for these
families as a family of one (e.g., total aliphatic hydro-
carbons) or as a family of individuals, needs to be developed
further, perfected, and used in cancer prevention studies.
In the same way, the carcinogenic and mutagenic activities,
reaction with DNA, and the metabolic properties of these
families need to be determined, especially in ways that mimic
the human situation.
Table 5
Families of Genotoxicants in Air
Aldehydes
Aliphatic amines
Aromatic amines and precursors
Asbestos
Azaarenes (mono and dicyclic)
Azaarenes (polycyclic)
Benzene derivatives
Epoxides
Halogenated alkanes
Halogenated alkenes
Halogenated ring compounds
Long chain aliphatic acids
Long chain aliphatic alcohols
Long chain aliphatic hydrocarbons
Long chain aliphatic esters
Metals and their compounds
Nitrosamines
N0x
Olefins
Oxidants (0,, NO2, PAN)
PAH (Di- and tricyclic)
PAH (tetra-, penta- and hexacyclic)
SO
The most extensive and important interfacing that man
has with his environment is through his respiratory membrane.
Each day this membrane, which has a surface area as large as
a tennis court, is exposed to a volume of contaminated air that
would fill a 15 m swimming pool (18). From the genotoxic
viewpoint, it would be of value to know the relative amounts
of atmospheric organic gases, vapors, and particles in contact
-------
ATMOSPHERIC GENOTOXICANTS 181
with this epithelial tissue. (The reason for dividing air
pollutants into these three states is primarily because of
sampling protocol.)
Normally, gases such as SO2 are almost completely
assimilated by the nose. When these gases are absorbed on
respirable particles, they can penetrate the lower respira-
tory tract more easily. In addition, any circumstance dis-
posing to breathing through the mouth is likely to increase
exposure of the lung to pollutants. Respirable particles
constitute that portion of the inhaled particles which pene-
trate to the non-ciliated portions of the lung. The vapors,
being liquids or solids in the neat state, are more readily
retained in the respiratory tract than are the gases, other
conditions being equal. In dogs, the respiratory retention
of inhaled benzene and toluene in the upper and lower respi-
ratory tracts was very high (15).
Regarding atmospheric vapors, we certainly need a
reordering of priorities in our research on environmental
genotoxicity. Organic vapors can be present in many indus-
trial atmospheres in four orders of magnitude larger than
the total airborne particulates (33). Some of the carcino-
gens in the vapor state can be present in five to six orders
of magnitude larger than the carcinogens in the solid state.
The high amount of genotoxic vapors produced in the United
States is shown in Table 6 (3). This means that genotoxic
vapors such as toluene, trichloroethane, trichloroethylene,
vinyl chloride, and benzene are emitted in much larger
amounts compared to benzo(a)pyrene, and BaP is emitted into
the air in greater amounts as compared to DDT or the PCB.
With an increase in coal consumption to 665 million
tons in 1976 and a projected increase to 1.27 billion tons
in 1985 (6), the pollutants in airborne particulates contam-
inated with coal combustion products may become of greater
importance. The same tenuosity and delicacy that qualify
the air-blood barrier in our lungs for the rapid exchange
of oxygen and carbon dioxide reduce its effectiveness as a
barrier to inhaled genotoxic gases, vapors, and particles.
Particles greater than 10u are taken out by the filter
system of the nose. Particles of 2-10y settle on the walls
of the trachea, the bronchi, and the bronchioles. Particles
in the range of 0.3-2y reach the alveolar ducts and alveoli
while those less than 0.3u, if not taken up by the blood,
are cleansed from the lungs with air. The deposition can be
as low as 10-15% for particles in the size range 0.5-l.Oy
diameter. Above In deposition is stated to increase quite
rapidly to become effectively quantitative at 5 of 6u (16).
-------
182
EUGENE SAWICKI
Table 6
USA Production of Some Genotoxicants in 1976 (Anderson)
Production
Ranking
5
13
14
15
16
18
19
21
22
23
26
31
33
37
41
45
Chemical
Ethylene
Benzene
Propylene
Toluene
Ethylene dichloride
Xylene
Styrene
Ethylbenzene
Vinyl chloride
Formaldehyde
Ethylene oxide
p-Xylene
Cumene
Phenol
Propylene oxide
Acrylonitrile
109 Ibs
22
10.6
9.8
8.2
7.9
7.3
6.3
6.1
5.7
5.6
4.2
3.2
2.7
2.2
1.80
1.52
.1. i.4. \_/ .L v^. u< o v*
%, 1966-1976
7.0
4.3
7.6
6.9
8.2
11.9
7.1
6.6
8.7
4.2
6.1
19.9
11.6
5.4
9.7
7.8
Large dust particles containing genotoxicants are
usually filtered out by the nose. An example of such a
material is wood dust. Those who work closely with wood dust
have a higher than normal incidence of cancer of the nose
and sinuses (1). The somewhat smaller particles deposit
themselves in the bronchi, then ride the ciliary escalator
to exit from the lungs within hours of deposition. In most
cases, these particles are then swallowed and tend to end up
in the stomach. This is essentially what has been postulated
to explain a high incidence of gastric cancer in Carbon and
Emory Counties, Utah (24). The coal miners who are affected
breathe in large particles of coal dust in the mines and
coal soot at home where coal is used as a fuel. Among
other characteristics, the coal soot has a high PAH content.
With respirable particles, their alveolar clearance is a
much slower process consisting of an initial phase lasting
twenty-four hours involving phagocytosis, an intermediate
phase of continued transport for three to ten days, and a
prolonged phase lasting one hundred days or longer (17).
Because of this longer period of contact, the situation can
be conducive to carcinogenesis.
-------
ATMOSPHERIC GENOTOXICANTS 183
This is probably the reason why lung cancer in uranium miners
originates from deposition of finer dust in the conducting
airways or in the acini, the respiratory units that together
constitute the pulmonary compartment of the lung.
On the basis of a significant amount of circumstantial
evidence, it would appear that 80-90% of human cancers are
derived from contacts with environmental factors. One school
of thought believes that most human cancers are associated
with personal pollution. The two main etiological factors
are believed to be cigarette smoking and diet, which are
believed to account for most of the cancers in the digestive
tract, the respiratory tract, and the endocrine-sensitive and
reproductive organs.
There are several difficulties with this type of belief.
The cancer patients who smoked cigarettes and/or ate "badly"
did not otherwise live in a chemical vacuum. If we hurdle
this body-count type of reasoning, we can face the fact that
most cigarette smokers do not get lung cancer. This is prob-
ably because other factors (genetic and environmental) are
involved in this carcinogenesis as shown by the tip of the
"cocarcinogenic" iceberg—the effects of asbestos and radia-
tion on the cancer rate of cigarette smokers. Evidence has
been presented which indicates that lung cancer in cigarette
smokers is derived from the families of carcinogens and co-
carcinogens in cigarette smoke, a genetic factor(s), and an
urban factor(s) (37). The author suggests a strong syner-
gistic interaction between cigarette smoking and the constitu-
tional host susceptibility to lung cancer. In addition, it
is premature to argue that the carcinogenicity of the polluted
environs (air, water, industrial, and food pollution) compared
to the personal type of pollution as denoted by cigarette smoke
is relatively negligible since the production of chemicals is
continually increasing at a rapid rate as are our exposures to
these chemicals. Because of a latency period of about twenty
to sixty years, the results of this type of pollution have not,
as yet, hit us with full force. This situation will be aggra-
vated further with increasing industrialization in other
countries and with increasing world population (Table 7).
This is shown by the data in Table 6 and the report that the
production of organic chemicals in the non-communist world
increased from seven million tons in 1960 to sixty-three mil-
lion tons in 1970 and is predicted to increase to 250 million
tons in 1985 (30).
The areas that have had the highest priority in our
research studies on polluted atmospheres have usually had
-------
184 EUGENE SAWICKI
Table 7
Doubling of World Population
Year, AD No. Years Billions of People
1
1600
1850
1930
1975
2010
1600
250
80
45
35
0.25
0.5
1.0
2.0
4.0
8.0
large numbers of chemical and petroleum refining industries
concentrated near a large body of water. Examples of such
areas are the Kanawha River in the Kanawha Valley of West
Virginia; the Arthur Kill in the Rahway, Newark, Jersey City
area; the Delaware River in the Philadelphia-Camden area;
the Mississippi River in the New Orleans-Baton Rouge area;
the Gulf Coast in the Texas, Louisiana, Mississippi State
areas; the Niagara Falls area; and San Francisco Bay area.
The areas of prime interest usually have a high order of
pollution, a high production of chemicals and derived prod-
ucts, a high cancer rate, or an emergency hazardous chemi-
cal(s) situation.
One of the families of compounds found in such areas is
the group of benzene derivatives. The activity of some of
them is shown in Table 8. However, the data on the genotoxic
activity of this family is very sparse. Thus far, there is
no reliable carcinogenicity bioassay system for these com-
pounds. This is particularly disturbing because of the large
variety of benzene derivatives which are found in the polluted
atmosphere (Table 9).
A combination of genetic factors and chronic exposure
to benzene are vital factors in the etiology of leukemia (2).
However, there are probably other key chemical factors
involved in this problem. This is because people exposed
to benzene are also exposed to other chemicals in their
environment, and the benzene which they are in contact with
is usually impure or may even be impure toluene or xylene.
-------
ATMOSPHERIC GENOTOXICANTS 185
Table 8
Genotoxic Benzene Derivatives
Compound
Benzene
Toluene
Styrene
Hexachlorobenzene
Species Exposure
Rat Inhalation
Rabbit Inhalation
Human Inhalation
Rat, Mouse Inhalation
Human Inhalation
Rat Ingestion
Rat Inhalation
Yeast Host-mediated
Rat Intraperitoneal *
Hamster Ingestion5
Genotoxic
Effect
Clastogen1
Clastogen1
Clastogen
Leukemogen
Leukemogen
Carcinogen
Clastogen1
Mutagen
Comutagen
Carcinogen
References
10
21
39
4
39
3 23
10
22
12
9
'Chromosome lesions in bone marrow cells.
tentative link of high doses of benzene to leukemia in rats
and mice.
'Male and female rats fed benzene down to 50 mg/kg of body
weight resulted in some zymbal gland and dermal tumors. Pre-
liminary data which need confirmation.
"induction of 2,4-diaminoanisole mutagenicity in vitro.
'Resulting in hepatomas, haemangioendotheliomas, thyroid
adenomas, and a shortened lifespan.
-------
186 EUGENE SAWICKI
Table 9
Some Atmospheric Benzene Derivatives
Benzene Dichlorobenzenes
Acetophenone Diethylbenzenes
Anisole Diethyl Phthalate
Benzaldehyde Dimethyl Phthalate
Benzonitrile Dipropyl Phthalate
Benzophenone Ethylbenzene
Benzyl Bromide Ethyltoluenes
Benzyl Chloride Fluorobenzene
Benzyl methyl ether Hexachlorobenzene
Biphenyl Hexafluorobenzene
Biphenylene Hexylbenzenes
Biphenyl ether Methylstyrenes
Bromobenzene Pentylbenzenes
Bromotoluenes Perfluorotoluene
Bromoxylenes Phenylethanol
Butylbenzenes Propylbenzenes
Chlorobenzene Styrene
Chlorotoluenes Toluene
Chloroxylenes Trichlorobenzenes
Cumene Trimethylbenzenes
Xylenes
Large aliphatic hydrocarbons are another family of com-
pounds classified as cocarcinogens present in the environment
in fairly high concentrations. Methods are available for
their analysis as a family or as individuals. They and the
polynuclear aromatic hydrocarbons have been discussed in pre-
vious papers (31,32). The importance of cocarcinogenicity is
demonstrated in those reports that large aliphatic hydrocarbons
can increase the carcinogenicity of some PAH a thousand-fold;
they can cause lung tumors when painted on mice whose pregnant
parent had been previously injected with BaP, and they can
cause some noncarcinogenic PAH to become carcinogenic (30).
The large number of halogenated aliphatic (about 100
found so far) and ring (about 50 found so far) compounds in
the polluted atmosphere means that we have to redirect some
of our bioassay studies into investigations of integrated
genotoxic effects of both the individual members of a carcin-
ogen family and the various families. We must look for addi-
tive, multiplicative, and initiation-promotion types of
-------
ATMOSPHERIC GENOTOXICANTS 187
insults. To complicate further the situation, some members
of the family can have more than one cancer pathway or dif-
ferent organotropic effects.
Another family of genotoxic compounds found in the pol-
luted atmosphere is the aromatic amines. At least three
possible cancer pathways can be deduced for these compounds,
through ring epoxidation, N-hydroxylation, and nitrosation
with NO to form the diazonium salt. The genotoxic activities
of some monocyclic aromatic amines are given in Table 10. The
high bladder cancer rate in some American counties may be due
to the presence of aromatic amines and their azo dye precursors
in the polluted environment. In preliminary work we have found
some of these amines in the polluted atmospheres of some coun-
ties that have high bladder cancer rates. The analysis, atmo-
spheric concentrations, and genotoxic properties of many of
these pollutants have been considered (30-33).
Doll (11) has discussed various industrial genotoxicants,
many of which have been found in the polluted atmosphere in
large numbers and sometimes in fairly high concentrations. The
genotoxicants are important for a number of reasons. Their
hazard to humans is derived from their effect on sentinel
individuals (workers who are a high risk group because of this
contact and a chemical and genetic background which is conducive
to carcinogenesis). They could cause potential problems to
the workers concerned. Many of these agents find their
way into the atmosphere through leakage, accident, dumping,
or use, so that large numbers of people are exposed to
them.
It is commonly thought that a carcinogen either causes
cancer (100% effect) or it doesn't (0% effect). If it takes
several exposures to cause cancer, what effect would too few
exposures have on the body? There could be other mutagenic
effects (besides the carcinogenic one) leading to an accumu-
lation of metabolic errors, thus causing a decrease in the
quality of life and/or a life-shortening effect. One effect
of contacts with an industrial carcinogen could result in a
germinal mutagenic effect that would be passed on to future
generations.
It is highly unusual for one chemical or mixture to cause
cancer in humans, unlike carcinogenesis in animals. Man must
come in contact with huge amounts of a carcinogen for the risk
of cancer to be 100%. A 100% risk was reported for lung cancer
in miners of radioactive ore and for bladder cancer among some
aromatic amine producers (19). For example, in one small group
of 19 men employed in distilling 2-naphthylamine, the risk proved
-------
188
EUGENE SAWICKI
Table 10
Genotoxic Properties of Anilines
Amine
Aniline
o-Aminoacetophenone
2 , 4-Diaminoanisole
4,4' -Diaminodipheny 1
ether
4,4' -Diaminodiphenyl-
methane
2 , 4-Diaminotoluene
8-Methoxylkynurenic
acid
4,4' -Methylene-bis-
(2-chloroaniline)
4,4' -Methylene-bis-
(2-methylaniline)
2-Nitro-p-phenylene-
diamine
4-Nitro-o-phenylene-
diamine
Phenacetin
o-Toluidine
p-Toluidine
Genotoxic
Species Effect1
Mouse, rat, man
S . typhimurium
Mouse
S . typhimurium"
Rodents
Mouse, rat
Rat
Rat
Drosophila
melanogaster
Mouse
Mouse, rat
Rat
Mouse
S . typhimurium
Mouse
S . typhimurium
Human
Rat
S . typhimurium
Rat
NC2
M3
C
M
C?
C
C
C
M
C
C
C
C
M
C
M
C
C
M3
C
References
25
41
13
5
14
35
20
7
8
27,36
36
38
38
38
38
26
28
25
28
1C = carcinogenic, M = mutagenic, NC = noncarcinogenic
2See discussion of negative results in body of paper
3In presence of norharman and S-9 mixture
"Using liver microsomal fraction from rats pretreated with
hexachlorobenzene
-------
ATMOSPHERIC GENOTOXICANTS 189
to be 100% (11). In the majority of cases, other factors are
necessary; this applies to vinyl chloride, asbestos, coke oven
effluents, benzidine, and even cigarette smoke. Thus, the
epidemiological study of industrial carcinogens has shown the
importance of cocarcinogenic factors. Because of our reliance
on state-of-the-art animal models and our ignorance of the
human chemical environment, epidemiological investigations
are controversial and are sometimes doomed to failure. An
example of such a controversial investigation involves the
relationship between industrial vinyl chloride and cancer
(40). Even with the most "thoroughly" investigated human
carcinogens, asbestos fibers, and cigarette smoke, the
epidemiological data for their effect on humans is said to
be inadequate for choosing between the additive, multiplicative,
or cocarcinogenic asbestos models to explain the results
(29).
Because of the vital importance of industrial chemicals
to our civilization, the boundaries of carcinogenicity and
noncarcinogenicity of our atmospheric and other environmental
pollutants should be determined. This means that we need much
more knowledge on the carcinogenicity, cocarcinogenicity, and
anticarcinogenicity of the numerous chemicals, families of
chemicals, and mixtures in our environments. The first step,
once we know what chemicals are in our environments, would be
to decrease exposure to the potent human carcinogens and to
those genotoxic chemicals or families present in highest con-
centrations in our environment. The carcinogenicity of any
chemical, whether it is classified as a human or animal car-
cinogen, is actually potential carcinogenicity for which
boundaries have not yet been determined. Although negative
results always have a question of uncertainty about them, so
do positive results. Newer carcinogenic results supersede
all previous negative results, but negative data should not
be discarded. It is still meaningful since it tells us some-
thing about the boundaries of the carcinogenicity and noncar-
cinogenicity of that particular chemical. Finally, it is
significant to remember that while industrial chemicals can
be a curse, they are humanity's best hope for increasing the
meaning and the quality of life.
The background against which we study mutagenic problems
is particularly complicated, annoying, and frustrating because
of our continually changing chemical environment and the con-
sequent alterations in the relative amounts and relative
importances of the various types of cancers and other genotoxic
manifestations. We need to know much more about the genotoxic
properties of pure chemicals, key environmental mixtures, key
families of chemicals, and non-ionizing and ionizing radiations.
-------
190 EUGENE SAWICKI
To accumulate this knowledge on these genotoxicants, further
improvements need to be made and data accumulated on rodent
bioassays for carcinogenicity and short-term bioassays for
carcinogenicity, germinal mutation, teratogenicity, athero-
sclerosis, and aging. To show some of the possibilities in
the investigations of genotoxic materials, the short-term
bioassays for carcinogenicity should be considered. They
can be used in the following:
• Predicting the carcinogenicity of pollutants of
unknown activity that are highly toxic and/or are
present in high concentrations in the environment.
• Setting priorities for chemicals to be tested in
mammals.
• Identifying active fractions and chemicals in
environmental mixtures.
• Identifying mutagenic metabolites in human body
fluids.
• Determining the possible carcinogenicity of atmo-
spheric, aqueous, or other mixtures with which
humans are in contact.
• Identifying mutagenic metabolites in plants.
• Determining the cocarcinogenicity (or comutagenicity)
of environmental and natural chemicals.
• Determining the anticarcinogenicity (or antimuta-
genicity) of environmental and natural chemicals.
• Investigating the mechanism of carcinogenesis.
• Determining the effect of pollution control
activities.
• Selecting relatively safe chemicals to replace the
hazardous ones that are currently of great importance
to our modern industrial society.
-------
ATMOSPHERIC GENOTOXICANTS 191
ACKNOWLEDGMENT
Much of the analytical data on the organic vapors was
obtained mainly from the studies of a research group led by
Dr. Edo Pellizzari at the Research Triangle Institute,
Research Triangle Park, NC.
REFERENCES
1. Acheson ED, Hadfield EH, Macbeth RG: Carcinoma of the
nasal cavity and accessory sinuses in woodworkers.
Lancet 1:311-312, 1967
2. Aksoy M, Erdem S, Erdogan G, Dincol G: Combination of
genetic factors and chronic exposure to benzene in the
etiology of leukemia. Human Hered 26:149-153, 1976
3. Anderson EV: Top 50 chemicals regain output lost in
1975. Chem Eng News 55:37, 1977
4. Anonymous: More benzene data. Chem Week 17, August 17,
1977
5. Anonymous: Hair dyes a hazard? Chem Week 17, Dec 21,
1977
6. Anonymous: Increased use of coal deemed safe through
1985. Chem Eng News 56:22, 1978
7. Blijleven WGH: Mutagenicity of four hair dyes in
Drosophila melanogaster. Mutat Res 48:181-186, 1977
8. Bryan GT: Neoplastic response of various tissues to the
systemic administration of the 8-methyl ether of xanthur-
enic acid. Cancer Res 28:183-185, 1968
9. Cabral JRP, Shubik P, Mollner T, Raitano F: Carcinogenic
activity of hexachlorobenzene in hamsters. Nature 269:
510-511, 1977
10. Dobrokhotov VB: The mutagenic action of benzene, toluene
and a mixture of these hydrocarbons in a chronic test.
Gig i San No. 1:32-34, 1977
11. Doll R: Strategy for detection of cancer hazards to man.
Nature 265:589-596, 1977
-------
192 EUGENE SAWICKI
12. Dybing E, Aune T: Hexachlorobenzene induction of 2,4-
diaminoanisole mutagenicity in vitro. Acta Pharmacol
Toxicol 40:575-583, 1977
13. Dybing E, Thorgeirsson SS: Metabolic activation of 2,4-
diaminoanisole, a hair dye component: I. Role of cyto-
chrome P-450 metabolism in mutagenicity in vitro. Bio-
chem Pharmacol 26:729-734, 1977
14. Dzhioev FK: On carcinogenic activity of 4,4'-diaminodi-
phenyl ether. Vopr Onkol 21:69-73, 1975
15. Egle Jr JL, Gochberg BJ: Respiratory retention of inhaled
toluene and benzene in the dog. J Toxicol Environ Health
1:531-538, 1976
16. Giacomelli-Maltoni G, Melandri C, Prodi V, Tarroni G:
Deposition efficiency of monodisperse particles in human
respiratory tract. Am Ind Hyg Assoc J 33:603-610, 1972
17. Gibb FR, Morrow PE: Alveolar clearance in dogs after
inhalation of an iron-59 oxide aerosol. J Appl Physiol
17:429, 1962
18. Green GM, Jakab GJ, Low RB, Davis GS: Defense mechanisms
of the respiratory membrane. Am Rev Resp Disease 115:
479-514, 1977
19. Hueper WC, Conway MD: Chemical carcinogenesis and cancers.
Springfield, Charles C. Thomas, 1964, p 74
20. Ito J, Hiasa Y, Yoniski Y, Marugami M: The development
of carcinoma in livers of rats treated with m-toluylene-
diamine and synergistic and antagonistic effects with
other chemicals. Cancer Res 29:1137-1145, 1969
21. Kissling M, Speck B: Further studies on experimental
benzene induced aplastic anemia. Blut 25:97, 1972
22. Loprieno NA, et al.: Mutagenicity of industrial com-
pounds: styrene and its possible metabolite styrene
oxide. Mutat Res 40:317-324, 1976
23. Maltoni C in Anonymous: Research "bombshell" hits ben-
zene. Chem Week 33, November 2, 1977
24. Matolo NM, Klauber MR, Gorishek WM, Dixon JA: High inci-
dence of gastric carcinoma in a coal mining region.
Cancer 29:733-737, 1972
-------
ATMOSPHERIC GENOTOXICANTS 193
25. Nagao M, Yahagi T, Honda M, Seino Y, Matsushima T,
Sugimura T: Comutagenic action of norharman and harman.
Proc Japan Acad 538:95-98, 1977
26. Rathert P, Melchior H, Lutzeyer W: Phenacetin: a car-
cinogen for the urinary tract? J Urol 113:653-657, 1975
27. Russfield AB, Homburger F, Boger E, Van Dongen CG,
Weisburger EK, Weisburger JH: The carcinogenic effect
of 4,4'-methylene-bis-(2-chloroaniline) in mice and
rats. Toxicol Appl Pharmacol 31:47-54, 1975
28. Russfield AB, Homburger F, Weisburger EK, Weisburger JH:
Further studies on carcinogenicity of environmental
chemicals including simple aromatic amines. Toxic Appl
Pharm 25:446-447, 1973
29. Saracci E: Asbestos and lung cancer: An analysis of
the epidemiological evidence on the asbestos-smoking
interaction. Int J Cancer 20:323-331, 1977
30. Sawicki E: The genotoxic environmental pollutants. In:
Proceedings of the First Symposium on Management of Resi-
dues from Synthetic Fuels Production (Schmidt-Collerus
JJ, Bonomo FS, eds.). Denver, University of Denver,
1976, pp 122-165
31. Sawicki E: Analysis of atmospheric carcinogens and their
cofactors. In: Environmental Pollution and Carcinogenic
Risks (Rosenfeld C, Davis W, eds.). Lyon, IARC Scientific
Publications No. 13, 1976, pp 297-354
32. Sawicki E: Chemical composition and potential genotoxic
aspects of polluted atmospheres. In: Air Pollution and
Cancer in Man (Mohr U, Schmahl D, Tomatis L, eds.).
Lyon, IARC Scientific Publications No. 16, 1977, pp 127-157
33. Sawicki E: Analysis of atmospheric pollutants of possible
importance in human carcinogenesis. Presented at the con-
ference on "Modern Measurement of Environmental Pollutants"
at the University of Rochester Medical School, May 23, 1977,
in press, 1978
34. Smith DH, Achenbach M, Yeager WJ, Anderson PJ, Fitch WL,
Rindfleisch TC: Quantitative comparison of combined gas
chromatographic/mass spectrometric profiles of complex
mixtures. Anal Chem 49:1623-1632, 1977
-------
194
EUGENE SAWICKI
35. Steinhoff D, Grundmann E: Zur cancerogenen wiskung von
4,4'-diaminodiphenylmethan und 2,4'-diaminodiphenylmethan.
Naturwissenschaften 5:247-248, 1970
36. Stula EF, Sherman H, Zapp Jr JA, Clayton Jr JW: Experi-
mental neoplasia in rats from oral administration of
3,3'-dichlorobenzidine, 4,4'-methylene-bis(2-chloro-
aniline), and 4,4'-methylene-bis(2-methylaniline).
Toxicol Appl Pharmacol 31:159-176, 1975
37. Tokuhata GK: Cancer of the lung: Host and environmental
interaction. In: Cancer Genetics (Lynch HT, ed.).
Springfield, Charles C. Thomas, 1976, 213-232
38. Venitt S, Searle CE: Mutagenicity and possible carcino-
genicity of hair colourants and constituents. IARC
Scientific Publications No. 13, INSERM 52:263-272, 1976
39. Vigliani EC, Fornia A: Benzene and leukemia. Env Res
11:122-127, 1976
40. Wagoner JK, Infante PF, Saracci R, Duck BW, Carter JT:
Vinyl chloride and mortality? Lancet ii:194-195, 1976
41. Zharova El: Characteristics of blastomogenesis induced
with tryptophan metabolites. Patol Fiziol Ekspter 17:
54-58, 1973
-------
STATE-OF-THE-ART
ANALYTICAL TECHNIQUES
FOR AMBIENT VAPOR PHASE
ORGANICS AND VOLATILE
ORGANICS IN AQUEOUS
SAMPLES FROM
ENERGY-RELATED ACnvrTIES
Edo D. Pellizzari
Chemistry and Life Sciences Group
Research Triangle Institute
Research Triangle Park, North Carolina
-------
197
The presence of organic components in the ambient air
is a fact of life in a modern society, since volatile organic
compounds are ubiquitous. Automobile exhaust, fossil fuel
burning, and the chemical industry contribute many organic
compounds to the air. It is not unreasonable to expect that
products from reactions of these chemicals with N02 and S02,
by photochemical (1-5) or other processes, will be also
observed in the atmosphere (6). However, many organic con-
stituents are suspected to enter the environment directly
by'industrial pollution (7). Carcinogenic and mutagenic
compounds find frequent use as intermediates in organic
synthesis, e.g., in the preparation and use of plastics,
fabrics, dyes, resins, cosmetics, Pharmaceuticals, etc.
Organic solvents, heavily used in industry, are also sources
of high levels of organic vapors.
Comprehensive studies on levels of carcinogenic agents
in air and correlation of this information with health effects
in humans are mandatory if we are to understand better the cur-
rent genetic diseases, as well as problems in carcinogenesis
and mutagenesis. While immediate and life-threatening effects
of some of these compounds are obvious, the consequences of
chronic low levels of exposure are often not known for many
years. Qualitative/quantitative analysis of the atmosphere
is vital to establish the etiology of cancers and other
diseases. It is essential to understand the organic com-
position of the atmosphere because of the existence of anti-
and cocarcinogenic factors. Statistical studies demonstrate
that the incidence of cancer aassociated with the respiratory
system is elevated where high air pollution occurs. Thus,
-------
198 EDO D. PELLIZZARI
an analytical technique that provides information on the
identity and quantity of organic constituents of ambient
air is highly desirable.
An assortment of methods are described in the literature
for the collection and analysis of volatile organics from the
atmosphere. In fact, the variety of methods is a problem
because most techniques are too restrictive, i.e., they focus
only on one or a "few" substances at any given time; only a
"narrow window" is examined. More recently we have developed
and perfected techniques that provide for a polypollutant
approach and yield a more representative and quick chemical
analysis of the surrounding atmosphere (8-18). The polypol-
lutant method is based upon the use of a solid sorbent fol-
lowed by capillary gas chromatography/mass spectrometry/
computer analysis for qualitative and quantitative deter-
minations.
COLLECTION AND ANALYSIS
Because organic constituents of the air are present
usually at ppt to ppm levels in a vast amount of a diluting
medium (air and water vapor), it is generally not practical
to perform in situ analyses of all these organic compounds.
There is no widely applicable method of detection that could
distinguish each compound from all others at such low concen-
trations. Therefore, to register enough sensitivity, vapors
must be concentrated from a large volume of air.
There are four basic steps necessary to successfully
analyze organic vapors in air. They are:
• The collection/concentration of vapors.
• Their transfer to an analytical system.
• Their separation and identification.
• The ability to measure the quantities of each of
the components of interest.
Concentration Techniques
With regard to collection/concentration, there are several
possible techniques (Table 1). Cryogenic sampling is excellent
for extremely volatile compounds such as acetylene, NO , SO ,
X A.
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES 199
Table 1
Methods of Concentrating Organic Vapors from Air
Selectivity
Method Mechanism Based On:
Cryogenic Traps Low Temperature Temperature
Condensation
Solvent Impingers Dissolution in Solvent Polarity,
Liquid Temperature
Sorbent Cartridges Adsorption on Structure of
Solid Surface Sorbent, Temperature
or freons, using liquid nitrogen, oxygen, or solid C02/acetone
as the cooling medium. However, there are several drawbacks.
This method condenses considerable quantities of water vapor.
The reactive gases NOX, SOX, and ozone can cause artifacts by
reacting with amines, oxygenates, olefins, etc. Also, in
inaccessible field locations, setting up and maintaining
cryogenic traps can be difficult, and cryogenic samples are
difficult to store or ship. Furthermore, the use of oxygen
(or nitrogen which condenses oxygen) is rather dangerous.
Solvent impingers are also used, but handling and shipping
volumes of solvent also presents problems. Artifacts during
collection are also prevalent. The use of a solid sorbent,
such as Tenax GC, can achieve reasonably effective concen-
trations of organic components. Sorbent cartridges by con-
trast, can be made clean, lightweight, and compact. All the
vapors collected from a large volume of air can be delivered
to an analytical system by thermal desorption of the trapped
vapors. During collection of vapors on a sorbent, each
adsorbed compound is in equilibrium with its vapor in the
air stream, so that it moves slowly through the sorbent bed.
The partition ratio between the air stream and the sorbent
surface for each compound is unique, depending on the tem-
perature and the structure of the sorbent. The ratio deter-
mines selectivity for the compound because it controls the
rate at which the compound moves through the bed. Compounds
are quantitatively collected until sufficient air has passed
through the bed to elute them. This elution volume, or break-
through volume, must be known as a function of temperature
for any compound which is to be collected for quantitative
analysis by this technique.
-------
200 EDO D. PELLIZZARI
Recovery Methods
Once collected, the concentrated vapors must be recovered
from the collection system and delivered to an analytical
instrument. One method employs cryogenic traps and solvent
impingers. If cryogenic traps are used, a substantial quan-
tity of water is collected. Solvent impingers use liquid as
the collection medium. In either case, the collected com-
pounds must be separated from a substantial volume of liquid.
This could be done by inert gas purge, by solvent extraction
and concentration, or, for highly volatile compounds, by low
temperature vacuum distillation. These procedures are tedious,
sample throughput is low, and substantial losses of the col-
lected compounds may occur. Generally, only a small aliquot
of the sample can be analyzed, and thus the sensitivity is
poor.
The other method employs sorbent cartridges. The vapors
can be recovered from a sorbent bed by solvent extraction or
by thermal desorption. The disadvantages of solvent extrac-
tion have already been mentioned. Thermal desorption is done
by simply heating the sorbent in an inert gas stream.
Sampling System
After considering all of the above factors, a "polypol-
lutant" technique was developed using the sorbent cartridge
and pumping apparatus shown in Figure 1. In sampling, air
is drawn first through a glass fiber filter to remove partic-
ulates, and then through the cartridge. A manifold can be
used to collect replicate cartridges. A 12-volt DC pump is
used which is powered by an automobile battery, or, if a 110
volt AC is available, by a built-in battery charger. Thus,
the system is portable, weighing about 20 pounds. The air
stream exhaust passes through a rotameter and a gas meter.
Typical sampling rates vary between 1 and 10 liters per
minute, although the volume sampled generally is between
20 and 200 liters.
Choosing a Sorbent. There are several criteria which a sor-
bent must fulfill if it is to be acceptable for ambient air
pollution studies. These criteria are:
• The sorbent must withstand repeated use without
deterioration and must not contaminate the sample.
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES
201
lOon PYREX TUBE -
-GLASS WOOL PUGS
SORBENT 860 »^
2.2 q 35/60 MESH TENAX GC CARTRIDGE
o
GAS
METER
GLASS
FIBER
FILTER
Figure 1. Organic vapor collection system.
• In order to collect a compound, it must adsorb
substantially all the vapor passing into it.
• The adsorbed vapors must be completely released
upon thermal desorption.
• The sorbent should possess a sufficient breakthrough
volume or retention volume if the method is to be
used for quantitative analysis.
• The sorbent must not catalyze in situ reactions on
its surface during and after vapors have been
adsorbed on its surface. In other words, it should
not be involved in hydrolysis, rearrangement, syn-
thesis, or decomposition of compounds.
Before Tenax was selected, a number of sorbents—carbon,
porapaks, chromosorbs, etc.—were evaluated according to the
first four criteria. Tenax has been found to be, at least
for the time being, the best compromise. Thermal and storage
-------
202
EDO D. PELLIZZARI
stability of Tenax were observed by repeatedly desorbing it
at 270°C after varying storage intervals. The vapors thus
produced were analysed chromatographically. Ethylene oxide
and styrene were often observed, but the amounts were too
small to interfere with use of Tenax for sampling organic
vapors from air.
Collection efficiences for sorbents for representative
vapors have been evaluated by purging a small quantity of
vapor from a 2 liter bulb as shown in Figure 2. An air
stream carries the vapors from the bulb to the flame ioni-
zation detector. A cartridge is interposed between the bulb
and the detector.
The detector response depicts an exponential decrease
in the amount of vapor coming from the bulb when no cartridge
is in place. This is depicted in Figure 3. Complete collec-
tion has been demonstrated for a variety of representative
chemical classes.
FLOW METER/REGULATOR
r-GLOVE BOX
\
L
EXHAUST TO
CRYOGENIC SAFETY
TRAPS
FLAME
IONIZATION
DETECTOR AMPLIFIER
Figure 2. Monitoring system for vapors in cartridge
effluents.
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES
203
l.6r
1.4
~g
* 1.2
e
<
<
1.0
g 0.8
H
O
LJ
t-
UJ
0.6
ui
2
0.4
Q2
A no cartridge
B cartridge in line
I I I
40 2
TIME ( MIN. )
Figure 3. Elution profile of cartridge effluent.
Percent recovery of collected vapors has been determined
by comparing gas chromatographic responses when cartridges
loaded with a known amount of vapor were desorbed with the
responses obtained by direct injection into the chromatograph
of the same amount of vapor. Examples of some results are
given in Table 2. Other data have also been published for a
variety of representative compounds. In each case the results
of immediate analysis were compared with the results obtained
-------
204 EDO D. PELLIZZARI
Table 2
Percent Recovery of Vapors After Storage
Storage Period (Weeks)
035
1-Nitropropane 95 + 2 93 + 3 50 _+ 9
Chlorobenzene 95 _+ 2 80 _+ 4 50 + 8
Phenyl methyl ether 95 +_ 2 95 +_ 2 70 + 8
N-Ethylaniline 95 + 2 95 + 2 70 _+ 6
Nitrobenzene 95 _+ 2 95 + 3 50 +_ 9
Aniline 95+2 95+2 80+5
4'-Fluoroacetophenone 95+2 80+4 90+4
after a period of storage. During the first three-week storage
period the cartridges were subjected to round trip shipment by
air freight to test the effectiveness of the storage containers
at high altitude. Little change in percent recovery has been
seen after three weeks for most vapors. However, two weeks
of further storage results in significant losses for some
compounds.
In a similar manner, a representative group of model
compounds has been used to determine four general methods
of breakthrough volumes. They are:
• Cartridges have been purged into a monitoring
system as discussed earlier, and the response of
the flame ionization detector was observed.
• Disappearance of vapor from cartridges during
purging under laboratory and field conditions has
been determined.
• Appearance of vapor in backup cartridges during
purging also has been determined. One loads car-
tridges with a compound and then samples air that
does not contain that compound. Backup cartridges
are changed periodically during sampling. The
volume of air required for half of the vapor in
the backup cartridges to appear is then determined.
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES 205
This approach has been used to test the possibility
of premature breakthrough or displacement chromato-
graphy when sampling in the presence of high ambient
levels of hydrocarbons, such as occurs with auto
exhaust.
• Elution volumes may be determined on a gas chroma-
tographic column packed with a known quantity of
sorbent. The breakthrough volumes determined by
all of these methods generally are in substantial
agreement.
Table 3 lists the breakthrough volumes for a few organic
compounds. We have determined this characteristic for approxi-
mately 125 compounds and extrapolated the results to more than
400.
Table 3
Breakthrough Volumes at 26.7°C for Tenax GC 35/60M
Per gram Per 6 cm
of Tenax Cartridge
Benzene 17 i 37.4 i
Carbon Tetrachloride 7 a 15.4 i
Dimethyl Amine 16 £ 35 i
Dimethyl Nitrosamine 74 SL 163 I
1,2-Dichloroethane 10 £ 22 I
Finally, the sorbent must not catalyze in situ reactions.
One such possiblity of forming dimethylnitrosoamine (DMN) on
the surface of Tenax shall be discussed.
In order to examine this reaction, a gas flow system as
shown in Figure 4 has been used. Clean air passes through
an oxide of nitrogen/ozone generator where it mixes with
selected concentrations of nitric oxide, nitrogen dioxide and
ozone, then is humidified. The air then passes through a
glass reaction tube. At the upstream end of this tube
dimethylamine (DMA) can be introduced from a thermostatted
chamber containing a permeation tube. At the downstream end,
the air is analyzed for oxides of nitrogen and ozone and
sampled for organic vapors.
-------
206
EDO D. PELLIZZARI
SCRUBBER
TRAIN
"'"•fl
supRy
•JUTECH NO
sampler
°>
anolynr
"°,
onolyitf
f \
P^ " J
/ tCTTATED ^r
OftlERITE CARBON
MOLECULAR
SIEVES J
RCrTOMTTER:
1^
1
(54.2 pgm/
— .TEFLON MEJUBRANE
=F1LTCR
:]
SUONG OVER ^"v t^*"*
'*• l^- rv
rt~~h ~*
— /J
'] HUMIOFER I
TENAX 1
CARTROGE (GEUUN A/E PREF1LTER) I
1 M>^> - — /T^ — 1
1 CLASS REACTION TUBE /
J (Ua«t.d.iUm) PERMEATION
V
iL
i
\MK1M
\CHAM
"J
Figure 4. Schematic of apparatus for studying artifacts on
solid sorbents.
Table 4 presents experiments that have been conducted.
The sampling cartridges were preloaded with DMA and used to
sample air containing oxides of nitrogen, ozone, and 151 ppb
of deuterated DMA. DMN could be formed only on the cartridge
in this case, while deuterated DMN could form in the flow tube
and on the cartridge. Each experiment was repeated in reverse
fashion, with the cartridges being preloaded with deuterated
DMA and used to sample 131 ppb of DMA in the tube. The iden-
tity of the amine which was preloaded on the cartridge is
indicated in Table 4 by the letter d and h by the amount of
DMN formed only on the cartridge.
The ratio of DMN formed from the amine which passed down
the tube to the DMN formed from the amine preloaded on the
cartridge is consistently larger than one. This means that
some of the DMA must have been converted before reaching the
sorbent bed. The length of the flow tube was changed from
15.8 cm to 130 cm without producing any significant change
in the amount of amine attributable to formation in the tube.
This would suggest that the reaction occurred on a surface in
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES
207
Table 4
Relative Amount of DMN Formation
Flow Tube vs. Sorbent Bed
DMN (nM)
Flow Tube
Length
X
Diameter
(cm)
Concentration
(ppb) T
°3
NO
N02
Flow Tube
15.8 x 3.5
130 x 3.5
95
25
105
2
90
7
90
7
0
10
0
60
0
30
0
30
530
545
515
530
540
505
545
480
Flow Tube
4
35
485
Flow Tube
15.8 x 7.5 3
130 x 7.5 3
70 x 7.5 2
260
270
275
225
230
240
otal
DMA
(nM)
25°C
404
404
404
404
404
404
404
404
70°C
404
25°C
404
404
404
Formed
only on
Car-
tridge
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
212d
379d
189h
081h
200d
0 d
159h
148h
243h
137d
lOOd
125d
Formed in
Flow Tube
and on
Cartridge
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
310
459
437
412
256
0
287
312
625
391
256
391
Ratio
C + T
1
1
2
5
1
1
2
2
2
2
3
C
.46
.21
.31
.09
.28
-
.52
.11
.57
.85
.56
.13
-------
208 EDO D. PELLIZZARI
the inlet to the cartridge such as on the glass fiber filter
which was used. The flow tube was replaced with one having a
larger diameter. The flow rate of the air was increased to
maintain the linear velocity at 520 cm/sec, and the upstream
DMA permeation tube was replaced with a faster one in an
attempt to maintain the same DMA concentration in the tube.
The new permeation tube had a higher rate than expected, and
the resulting DMA concentration was actually 466 ppb or 3.5
times that in the smaller tube. This would be expected to
lead to a three-fold increase in the production of DMN in the
tube if it were formed in a homogeneous, gas phase reaction.
Then the ratio of DMN formed in the tube and cartridge to the
DMN formed only on the cartridge would become substantially
larger than three since the ratio was already larger than one
before the change was made. If the reaction were heterogeneous,
then less DMN would be produced in the tube, since the surface/
volume ratio was decreased from 1.14 to 0.53 in going from the
smaller to the larger tube. This would cause the ratio to
decrease, since the amount of amine from the tube drawn into
the cartridge was kept the same.
The results actually obtained indicated a less than three-
fold increase in the ratio. Furthermore, the ratio was substan-
tially independent of the length of the tube. This means that
while some conversion of amine may have occurred in the flow
tube, most of the DMN formation that happened in front of the
cartridge must have taken place on the glass fiber filter and
the glass fiber plug which was used to anchor the Tenax in
the cartridge. Fortunately, the percent conversion of DMA to
DMN was very small.
Similar experiments have been conducted with molecular
chlorine and olefins, specifically ethylene. No chlorinated
products have yet been detected. Many other types of artifact
reactions still need to be examined to precisely define the
limitations of Tenax or any other sorbent as a collection
medium for ambient air pollutants. In fact, these experi-
mental concepts which have been outlined must be applied to
any collection technique before it is used in this capacity.
Once a collection device has been thoroughly tested,
the next step is to interface it to an analytical system.
Instrumental Analysis
«
An inlet-manifold has been used to thermally desorb the
vapors at 270°C from cartridges in a helium stream from which
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES
209
vapors are trapped at -196°C (Figure 5). Then the trap is
switched into the carrier gas stream of a capillary gas
chromatograph and rapidly heated to 250°C. Figure 6 depicts
a schematic of the gas chromatography/mass spectrometer/
computer system that was used for this analysis. The mixture
is resolved by using glass capillary columns. The detector,
a mass spectrometer, is coupled to a computer which stores
full scan data on disc or magnetic tapes. After data acqui-
sition is complete, the system plots normalized mass spectra
indexed to a total ion current chromatogram. Mass fragmento-
grams may also be derived from the acquired mass spectra.
are:
The advantages of this collection and analysis system
It is easy to transport and operate the samplers
in the field, even under adverse weather conditions.
Little water is collected.
PURGE GAS
ALUMINUM
HEATING BATH
=J«£=r
COMPRESSION SPRING
TEMPERATURE
CONTROLLER
VALVE POSITION A
(SAMPLE OESORPTION)
CARRIER
GAS
PURGE VjJX
GAS ^|J
ALUMINUM
HEATING BATH
ax-nm TWO
POSITION VALVE
HEATING CARTRIDGE
CARRIER GAS
TO GLC
CAPILLARY
HEATING « COOLING BATH
Nl CAPILLARY TRAP
VENT
VALVE POSITION S
(SAMPLE MJECmMI
CARRIER
GAS
Figure 5. Inlet-manifold for recoverying vapors from sampling
cartridges.
-------
210
EDO D. PELLIZZARI
Sample
Inlet/
manifold
Telephone
T
Cornell U.
P EM/STIRS
Search Sys-
tem
and/
or
Cypliernetics
tine shared
PDF/ 10
and/
or
RTI IBM 370
Mass Spec
Library-Search
Program
Figure 6. Schematic of gc/ms/comp system.
• The entire sample is delivered for analysis.
• Early complete chromatographic resolution of
individual compounds is attained.
• Resolution of individual compounds by mass frag-
mentography is virtually complete.
• Vapors present in the ppb to ppt range can be
quantitated.
• Gas chromatographic retention times and mass spectra
can be used for positive identification.
The disadvantages are:
• The volume of data produced is very large - 1000
mass spectra per sample, or 8,000 per day, or 40,000
mass spectra per week. Fast data processing systems
are needed to assist in the identification of the
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES 211
components represented by their mass spectra. Cur-
rent computer mass spectral search systems are
costly and inaccurate. However, efforts to improve
these problems are in progress in several labora-
tories.
Some compounds are not seen because they are too
volatile to be collected effectively on Tenax GC.
Other sorbents such as XAD, Porapaks, Chromosorbs,
etc., while they have higher adsorption affinities,
do not meet the criteria outlined earlier, i.e.,
they collect too much water, exhibit artifact
reactions, have poor thermal stability, or may yield
poor recoveries.
APPLICATION OF POLYPOLLUTANT METHOD
Qualitative Analysis
Figure 7 depicts an example of a volatile organic vapor
profile of ambient air taken in an industrial area in Deer
Park, Texas. Even though glass capillary columns were used
for effecting separation of components, complete resolution
was not achieved. The arrows in this figure indicate compo-
nents in the sample that have been identified as halogenated
hydrocarbons. The composition of this profile is given in
Table 5. The list of components are in their order of elution
from the gas chromatographic column (from left to right in
Figure 7). A large assortment of chemical classes are repre-
sented: oxygenated and halogenated hydrocarbons, alkanes,
alkenes, and alkyl aromatics. The highly volatile organics
such as ethylene, acetylene, ethane, and pentane are not
efficiently collected by the Tenax cartridge sampler, nor
are compounds such as methyl chloride, methyl bromide, vinyl
chloride, formaldehyde, or acetaldehyde. Thus, the highly
volatile end of this profile is skewed. Other techniques of
collection would have to be employed to trap these substances.
Altogether, 150 compounds were identified. This particular
sample was taken during the summer when the ambient air tem-
perature was about 98°F. This condition favors the vapori-
zation of relatively high molecular weight compounds. In
this case a seventeen-carbon hydrocarbon was detected. Pre-
sumably other materials present in the air were associated
with the particulate fraction which was filtered out by the
glass fiber filter in front of the Tenax cartridge. There-
fore, some skewing of the upper end of the window is observed
which is influenced by the ambient air temperature. The
-------
212
EDO D. PELLIZZARI
Table 5
Volatile Organic Vapors in Ambient Air From Deer Park, TX
Chroma to-
graphic
Peak No.
1
2
3A
4
4A
5
6
7
8
8A
8B
8C
9
9A
9B
9C
9D
10
10A
11
12
12A
12B
12C
13
13A
14
14A
14B
15
ISA
16
16A
17
Elution
Temp.
40
42
45
46
47
48
50
52
53
55
56
56
57
57
58
58
59
60
60
61
61
62
83
64
65
67
68
69
70
71
71
72
73
74
Compound
CO,
dichlorodif luormethane
1-butene
chloroethane + acetaldehyde
isopentane
trichlorof luormethane
acetone
isopropanol + dichlorometnane
freon 113 (BKG) + chloropropene
isomer
C.HgO isomer
dichloroethylene
isobutanal
1 , 1-dichloroe thane
2-methylpentane
dichloropropene (tent.) isomer
3-methylpentane
n-butanal
hexafluorobenzene (eS)
methyl ethyl ketone
ii-hexane
chloroform
2-butanol
C4HgO isomer (tent.)
perfluorotoluene (eS)
1 , 2-dichloroethane
methyl eye lopentane
benzene
carbon tetrachloride +• CyH^g
isomer
cyclohexane
2-methylhexane
2 , 3-dimethylpentane
3-methylhexane
dichloropropane + ^-7Hi4
isomers
dichloropropene isomer
Chroma to-
graphic
Peak No.
17A
17B
18
19
19A
20
21
22
22A
22B
23
24
24A
25
25A
26
26A
27
28
28A
29
30
30A
31
31A
32
33
34
34A
35
35A
36
36A
37
37A
38
Elution
Temp.
74
75
76
78
79
80
82
83
84
85
85
87
88
89
89
90
92
93
94
95
96
97
98
100
102
106
107
108
108
109
110
110
111
111
112
112
Compound
£-pentanal
tnchloroethylene +• C^H^
isomer
^-heptane
CaH.R isomer
o lo
C0H,,; isomer
o ID
methylcyclohexace
4-methyl-2-pentanone
C0H10 isomer
O 10
CgHlg isomer
CgH16 isomer
1,1, 2-trichloroethane
toluene
CQH.,, isomer
O ID
C8H18 lsomer
methylethylcyclopentane isomer
raethylethylcyclopentane isomer
C0H,,, isomer
o ID
n-hexanal + C0H. ,- isomer
— o 10
C8H18 lsomer
CgHlg isomer
n-octane
n-butyl acetate (tent.)
CgH-8 isomer
CgH18 isomer
ethylcyclohexane
ethylbenzene
£-xylene
CgH2Q isomer
CgHlg isomer
CgH20 isomer
C^H^^O isomer
styrene
C10H22 lsomer
o-xylene
n-heptanai
CgH^g isomer
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES
213
Table 5 (continued)
Chromato-
graphic
Peak No.
39
40
40A
40B
40C
41
41A
41B
41C
42
43
43A
43B
44
44A
45
46
47
47A
48
49
49A
50
50A
51
52
52A
52B
53
54
55
56
56A
56B
57
57A
58
Elution
Temp.
113
114
116
116
117
118
119
120
121
122
123
123
124
125
126
127
128
129
129
130
131
132
134
135
135
137
137
138
139
141
142
143
144
144
145
146
147
Compound
dichloropropene
n-nonane
C~H. c, isomer
y 1 0
^10ri20 Lsomer
isopropylbenzene -«- C.r)H22
isomer
C10H22 lsomer
C-Hj-0 isomer (tent.)
C10H20 lsomer
CgHlg isomer
^10^22 lsonler
benzaldehyde
n-propyl benzene
C10H22 isoraer
p_-ethyltoluene
C10H22 lsomer
1,3, 5-trimethylbenzene
C11H24 lsomer
C10H22 isomer
C.gHjg isomer
^10^22 lsomer
o-ethyltoluene -t- n-octanal
C10H20 1Somer
n-decane + dichlorobenzene
isomer ( tent. )
C4-alkyl benzene isomer
<"10H20 lsoraer
1,2 ,3-t rime thy Ibenzene
C4-alkyl benzene isomer
C, , H24 isomer
C11H24 lsoraer
C H92 isomer
C4-alkyl benzene + C^H^
isomers
acetophenone
C.-alkyl benzene isomer
C H,,,, isomer
CgH,gO isomer (tent.)
Clo"l8 lsomer
C4-alkyl benzene isomer
Chromato-
graphic
Peak No.
59
60
60A
60S
61
61A
61B
62
62A
62B
63
64
64A
65
65A
65B
66
66A
67
67A
68
68A
69
69A
70
72
72A
73
74
76
77
78
78A
79
81
82
82A
82B
83
84
Elution
Temp.
148
150
151
152
153
154
155
155
156
156
157
157
158
159
159
160
161
163
164
165
166
167
168
170
171
181
182
185
187
191
199
200
212
222
226
227
228
238
240
240
Compound
C4-alkyl benzene
ti-nonanal
C11H12 lsomer
C^-alkyl benzene
^-undecane
Cj-alkyl benzene
ji-penty Ibenzene
isomer
isomer
isomer
tetramethylbenzene isomer
C-2H24 isomer
C11H20 lsomer
C12H25 + Valkyl
isomers
<'12H26 lsomer
C11H22 lsomer
benzene
methylindan isomer
C4-alkyl benzene
Cg-alkyl benzene
C12H26 isomer
Cg-alkyl benzene
isomer
isomer
isomer
2-decanone •*- naphthalene
C12H24 lsomer
^-decanal
C12H24 lsomer
ri-dodecane
^12H24 lsomer
Cj_H,,g isomer
^n^28 isomer
C11H22° lsomer
C13H26 lsomer
n-tridecane
C12H24 isomer (tent.)
C14H28 isomer
m-tetradecane
C15H32 LSOmer
diethyl phthalate
C16H34 isomer
ji-hexadecane
C15H32 lsoraer
C15H30° isomer (tent. )
ri-hexadecane
C18H38 lsoraer
-------
214
EDO D. PELLIZZARI
30000 _.
10000
10100 10150 10200 10250 10300
MASS SPECTRUM NO.
10360
10600
Figure 7. Profile of volatile organics in ambient air from
Deer Park, TX.
ambient air temperature, of course, would determine those
compounds which would partition between the vapor and aerosol
state. A different procedure is needed for the collection of
organics associated with particulates.
This phenomenon notwithstanding, the collection and
analysis procedure presents a wider window unlike the mono-
pollutant methods. The procedure is extremely versatile and
applicable to many chemical classes with the exception of
the peroxides and the hydroperoxides.
Quantitative Analysis
Once the compounds have been identified, their quantities
need to be determined. There are basically two approaches to
the extraction of quantitative information. One is to prepare
standard curves relating instrument response vs concentration
for each compound of interest. The primary deterrent to
extensive use of this method is that a calibration curve for
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES 215
each of the compounds listed for this sample would be prohib-
itive because of the amount of instrument/operator time
required. The sample throughput would be extremely low.
Thus, from several points of view, it would be most desirable
to extract quantitative information from full-scan data, i.e.,
mass fragmentography exchanging a level of accuracy for
breadth of information. Before analysis, approximately 200 ng
of two standards, perfluorobenzene and perfluorotoluene, are
loaded on all the cartridges.
In the second approach, it is necessary to determine
Relative Molar Response (RMR) factors. Successful use of the
RMR method requires a knowledge of the exact amount of refer-
ence standard added and the exact amount of compound added.
D1/D Runknown/Molesunknown
KMrC
unknown/standard Rstandard/Molesstandard
R is a system response; it may be a peak area (a total ion
current peak with a value determined by either integration
or triangulation) , a peak height, or the area of the peak
produced by a particular ion. The ionic peak areas are
especially useful in those situations where even 100 m of
SCOT column cannot resolve chromatographic peaks. The value
of the RMR is determined from at least three independent
analyses.
Since
Moles , = g ^ .
compound compound' compound
where GMW = gram molecular weight, the number of grams of
unknowns can be calculated from the RMR factors and values
observed in the sample analysis in a straightforward manner
R • GMW • g
unknown unknown "standard
_
Unknown Rstandard -
Usually, two or three characteristic ions are selected for a
given compound to avoid overlap with the ions of other com-
pounds, since the ratio of one ion to another is known from
the mass spectrum (either from a compendium or determined in
the laboratory) , RMR factors can be calculated quite readily
in those cases where the most intense ion of the spectrum is
saturated. Quantitative data for all the organic vapors in
an air sample can be obtained in a single sample analysis
once the response factors are known.
-------
216
EDO D. PELLIZZARI
CO
1— 1
X5
oS
EH
j
G.
0}
TJ
W
cd
I-H
„
C
0
M
0
IT;
C
05
O
•H
C
CO
&fl
f-l
o
SH
0>
Si
O
•o
c
ai
T3
•P
aS
C
0)
to
0
r™^
^
*•«
o
W
i— !
0)
>
0
in
•H
1
— i
eg
01
1 1 CO 1
o
1 1 ^ 1
co"
CO
1 1 X 1
CO
in
1 1 E- 1
t t H 1
X
n
i i in^ i
^
CO
eg ill
a
O C7)
O 1 io 1
* 2
z
o o -
ffl = CU
CS i) S
J3 OJ C r-*
= ) C 3J ^
S 7 o* i^ E S
< i - & -
3 CM "^ O >
1 1
1 |
30
["*•
1 O
O
o
i in
i) a?
0 0
JH U
0 O
rH rH
a o
-^ -H
"O "O
1 1
— J ^H
t c- r- r- 6-> 1 1 t t- t 1
O O co :o
O CO V CO
i r~ eg t ill ii
f- ^H
31 C^
CO —
1 CO H 1 1 1 1 1 II
CO
o 30 o —• tn
O P» O OJ CO
-• w t- T c-
1 1 H 1 1 1 1 1 1 -
m o
CO •-•
i i s-< i in i i i o i i
OJ OJ
"3*
1 ^H t* 1 1 1 1 1 E- 1 1
CO -H O O O O
TOJ O CO Ol 01
f 1 CO H CO 1 1 O 1 t i-»
to
S £ 2
1 H -t 1 O 1 1 1 II
m"
X CD O
in i t- i c- i i i oj i i
-,
§ S S
i 01 H i iii ii
^H **
0 ^ 0) 01 0 >. 0)
P— c^OJ!Ut-j=aJc:
a>3Gaj>»cco*jea>
ar-43jQ,— aJOr-lO)3jCi.
aIJ:^0*J'-1+J— 0-*^0
J=^-^->WO>33OI-'3^
-» ** 4; 0.0^2^3-^ Oja^u
U^wOJ-tOOfHr-tOC*
O •*-» L.O!nl--fcJ-C[-'l-'
e i =0-^00 i ooo
- -^—13-Cjc: . u S ^
•UTH cj-a-jT3'o-H*-''aT:
i H e- i i i i i ' i " i
1 | £-1 1 1 1 1 1 1 1 1 1
1 1 1 f 1 1 t 1 1 1 1 1
1-1 CO CO CO X ^ CO
OJ CO • CO 01 CO O
CO CO E-i OJ Ol Ol 1 1 1 1
CO OJ
C"« CO
OJ •
1 CO I 1 1 ( 1 1 1 i I
m
OJ
1 i 1 1 1 1 1 1 1 1 [
t 1 1 I 1 1 1 1 1 1 1 1
CD en o
CO i-l O EH
t lOt Itl H E-*Hl
O)
1 f 1 1 1 1 1 i 1 1 1 1
rp f.
CO tD
1 1 CO 1 I t I I I E- 1
1 1 1 1 1 1 1 1 1 1 1 1
1 t I t ) 1 1 ! 1 1 1 1
1 1 1 1 1 1 1 1 ( i 1 I
1 1 1 1 J 1 1 1 1 I 1 t
1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 < 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 i 1 1
0) 1>
C 1 0) C
aj — * c d
J= -H 0) JS 0)
*J X -i *J C
D a. T3 3) 3J 0) 0)
0 O «J O Q. C C
>"ii 4-> 3) &« Q 1) Q)
oa, 3 c o h a> -H -H
rHO ^3lr-HQ,C -O "O D
-Z '/] | 2,.£ O ^ 0) al aj a
SO - fci cS 0 •*-» fll 3 3 -O Q,
«(H -Ha^^dO^^ X3 'HO
*JO IO+^-CO-^0 O'-^ S^ fci /-s
*j !s j-o*j'^ooo o^-- 1-10^-
CN 1 fct r- i^OJ-^-CO^Qjj^D CJOOJ
- O3 QJ — -in "icjr-tygoS ^JE
-t*Joji ---^otHW^n >."ow
- -H (J*-«flJD3J 5) -4S^
^-t J3 j^ r^ i-H t— ( Q, Q, 4J Q, > •<-*
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES
217
3
C
•H
-P
c
o
o
CO
o
o
o
CO C-
CO CO
3 I (£> I
— r- I I I I I I I
I
=i
O
Ol
u
o
a)
aj D
a! al
J ,-H « 0)
i >»i— «->
* t. >. al 0) 0)
3>v>»«*i-* -^J-aJU>,«JO((
jD-M-tJ^caO) 01 O.S « t.r-.rH
033*JJ=W£ 5**^= y:>':>v
H i i Q..C o -c *^ e v o y
cjci a i>3 a 03 s >-H « 3
-H | 1 t-t 2 £ r-« >>
w >% >,^: >>js >>,-» 3 >, >,
_ s D>»*i-*->o^3 y ja js
t i -H -t e -i u ^ i o> i i
" "' " T3 S -^ Cl M G| Ct
-------
218 EDO D. PELLIZZARI
Using the latter approach to quantitation, this method
has been used to obtain quantitative data for several samples
taken in the Houston, Pasadena, Deer Park, LaPorte, and Free-
port, Texas areas. Examples of these data are shown in Table
6. The range of concentrations were from a few ng/m to
several ug/m3. An inspection of the vertical columns of data
quickly reveals which samples were downwind from industrial
facilities. These data also can be categorized into two
general groups of pollutants, those that are ubiquitous,
occurring in upwind and downwind samples and those that are
"site" specific.
Table 7 summarizes the total weight of halogenated and
oxygenated compounds found in the Texas study.
ANALYSIS OF VOLATILE ORGANICS IN AQUEOUS SAMPLES
Many of the concepts and instrumental techniques used
in ambient air analysis can also be used in the identification
and quantitation of volatile organics (VGA) in aqueous effluent
samples from energy-related activities. The VOA technique
employs an inert gas, helium, which is bubbled through the
sample to transfer the volatile compounds from the aqueous
phase to the gaseous phase and then trapped on a Tenax car-
tridge (19,20). The sample is heated between 40° to 95°C
during the purging. Figure 8 depicts one of the many config-
urations which has been used.
Several other configurations have been reported. Regard-
less, foaming of the sample remains a serious problem. A
paucity of data has been published on the percent recovery of
chemicals from aqueous samples using these devices. The
recovery of carbon-14 labeled acetone, acetonitrile, benzene,
and toluene has been examined for several aqueous samples
from energy-related processes such as in situ coal gasifica-
tion (19). For compounds that are highly soluble in water,
e.g., acetonitrile, the recovery was very low, about 10%. On
the other hand, the recovery of hydrocarbons, aromatics, and
alkyl-aromatics was >_ 80%. In general, the purging of volatile
organics from an aqueous medium utilizing an inert gas is
quantitative for compounds with boiling points <210° and <2%
solubility, and for compounds with boiling points of <150°
with a solubility of <10% in water.
The VOA method has been applied to the analysis of
aqueous samples from energy-related activities using the
instrumental methods described earlier (Figure 6). Figure
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES 219
Table 7
Estimated Minimum Total Ambient Air Levels
of Volatile Organic Chemical Classes
Chemical
Class
HL1
HL2
HL3
PL1
PL2
DSL1
DSL2
DDL1
DTL1
DTL2
DTL3
DTL4
FL2
FL3
LL1
LL2
LL3
Halogenated
hydrocarbons
16.7371
1,100
12,518
294
5,433
128,948
6,604
19,409
2,078
25,398
716
15,990
32,028
40,026
13,165
37,926
7,832
Oxygenated
compounds
370
1,966
-
20
-
134
60
6,947
3,734
120
60
1,020
2,726
2,596
4,601
9,634
6,253
1 3
Values are in ng/m .
-------
220
EDO D. PELLIZZARI
TCtUX CARTRIDGE
TEFLON ADAPTER
GLASS «OOL PLUG
THCRUOUCTED
— ROUND BOTTOM fLASK
Cnoml)
- FRITTED GLASS TIP
Figure 8. VOA apparatus.
9 depicts a profile of volatile organics from a low Btu
gasification process for coal. The window representing
only the volatile organics is quite complex. The gas chroma-
tographic resolution is insufficient. However, when mass
fragmentography is employed, complete deconvolution is
achieved (Figure 10). As with ambient air analysis, the use
of ion chromatography is very important for quantification.
The identity and concentrations of the components in
this sample are shown in Table 8. Many alkanes, alkyl-
aromatics, thiophenes, pyridines, indanes, indenes, furnas,
etc. , were present. In this case approximately 134 chemi-
cals were identified and quantified. The range of component
concentrations in this sample was from a few ppb to 817 ppb
for naphthalene.
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES
221
Mass Spectrum No.
Figure 9. Profile of volatile organics in aqueous sample
from low btu gasification of coal (MERC, ERDA).
••t •*•
H« II
Figure 10. Ion chromatograms of sample in Figure 9.
CONCLUSIONS
Even though significant strides have been made in the
development of techniques for volatile organic analysis of
ambient air and aqueous samples; improvements are still
warranted. A master scheme would be useful to serve as a
guide to these analyses. Standardization is needed but it
should be tempered with flexibility for modification of
methods as improvements are made.
-------
222
EDO D. PELLIZZARI
Table 8
Volatile Organics in Aqueous Condensate (-2L) from Low BTU
Gasification of Rosebud Coal (MERC, ERDA)
Chromato-
graphic
Peak No.
1
2
4
6
a
9
10
12
14
15
16
17
18
19
20
21
22
23
24
25
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Elution
Temp.
(°C) Compound ppb
49 N, + 02
50 C02
54 C.Ho isomer
59 C4Hio lsomer
63 acetone
-
-
NQ
xg
100+20
70 perfluorobenzene (ef)
71 CeH,2 lsomer
74 perf luorotolue
77 C-H.2 isomer
79 benzene
81 thiopene
31 ^6^14 lsomer
85 C^H, . isomer
o 14
37 CgH,4 isomer
96 C.,H.-; isomer
i * b
100 toluene
154166
ne , eS)
T
308+134
4.5+1
57110
90+17
84+33
T
381183
101 methyl thlophene isomer T
104 C-H,,, isomer
i ifa
106 C-.H,,, isomer
7 Ib
110 C7H14 isomer
116 -a^io isomer
121 ethylbenzene
123 C0H,,- isomer
B * b
T
33117
T
20113
3817
40120
124 m- and £-*ylene 210+67
124 dimethyl thiophene isomer 32+24
126 C8Hlg -corner
T
127 styrene and/or
cyclooctatetraene 1213
-28 o-xylene
129 cgH,g isomer
1618
2315
131 dimethylthiophene isomer 47^10
132 CgH^0 isomer
133 anisole
T
4.611.4
136 C.,-alkyl benzene isomer T
137 CgH20 isomer T
141 C_-alkyl benzene isomer T
143 benzaldehyde
T
143 C3-aikyl benzene isomer 107110
144 cgH18 isomer
144 CqHog isomer
67+33
Chromato-
graphic
Peak No.
46
47
48
49
50
51
52
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
36
Elution
Temp.
("C)
147
147
148
148
149
149
150
152
154
154
155
155
158
158
160
161
162
164
165
165
166
167
168
168
168
168
170
170
171
171
172
173
176
179
180
180
180
181
182
183
Compound
PPb
Cg-alkyl benzene isomer T
methylpyndine isomer
CgHjo isomer
C9H20 isomer
benzof uran
methylpyridine isomer
47+10
33120
T
59140
38111
Cg-alkyl benzene isomer 144118
methylpyridine isomer
C10H22 isomer
CgH.Q isomer
unknown
methylanisole isomer
1617
T
T
T
T
C.-alkyl benzene isomer 3011104
C4-alkyl benzene isomer 33+20
indan
indene
112153
114140
C4-alkyl benzene isomer T
C^-alkyl benzene isomer 1718
C10H2Q isomer
C'10H22 isomer
10 + 2
1313
C.-alkyl benzene isomer T
C10H2Q isomer
T
C4-alkyl benzene isomer T
C10H22 lsomer
C^-alkyl Denzene isome
CdH_-benzene isomer
C^Hg-benzene isomer
C4H7-benzene isomer
C11H22 Lsomer
r 82144
T
T
73160
methylbenzofuran isomer 127+50
C. . Kr>4 Lsomer
C-=HQ-benzene iscmer
D y
C4H-benzene isoroer
methyllindan tsomer
CgHg-benzene isoraer
C11H22 lsomer
methylindan isomer
methyl indene isomer
methylind&n isomer
C-H_-benzene isomer
o a
177+37
367174
T
1917.4
3716
39119
57+37
24120
80+37
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES
223
Table 8 (continued)
Chromato-
graphic
Peak Mo.
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
Elution
Terap.
(°C)
184
184
185
186
187
188
189
189
189
190
190
191
192
193
196
196
196
196
199
200
200
201
201
202
Compound
methylindene isomer
ethyl phenol isomer
ppb
55+37
NQ
Cq-alkyl benzene isomer T
Cc-alkyl benzene isomer T
C11H24 lsomer
naphthalene
dimethylindan isomer
T
817+445
47+24
C5-alkyl benzene isomer 84+23
'"'12^24 isomer
C12H26 isomer
dimethylindan isomer
CgH13-benzene isomer
dimethylbenziraidazole
CgH_-benzene isomer
CgH.,, -benzene LSOmer
C-H, ,,-benzene isomer
b ij
dimethylindan isomer
CgH. --benzene isomer
C^H- -benzene isomer
C^Hg-benzene isomer
CaH, --benzene isomer
b i J
dimethylindole (tent.
C11H14 lsomer
C^H. --benzene isomer
96+35
74+34
82+44
T
(tent.) 35+11
38+11
T
6.7+3.3
T
3.3+2.0
32+31
17+7
3.3+2.0
Chromato-
graphic
Peak No.
Ill
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
134
Elution
Temp.
(°C) Compound ppb
203 C=H, , -benzene isomer T
fa i j
204 cnHi4 isomer 117.3+6.7
205 CgH13-benzene isomer T
206 a-methylnaphthalene 350<;83
206 C13H2g isomer 38+5
207 CCH, ,,-benzene isomer T
b 1 o
208 CaH, --benzene isomer T
b lo
208 C?H15-benzene isomer T
210 a-methylnaphthalene 143+67
210 CeH. . -benzene isomer T
b 11
211 CgHj^-benzene isomer
211 C?H15-Denzene isomer 10+4.7
211 C^H. 3-benzene isomer
214 CgH. ^-benzene isomer T
216 C10H.,, isomer
13 l8 ~ 39+11
217 C14H26 lsomer
219 c.nH.,, isomer
1 o 1O
220 C14H24 isomer 11+6
221 C14H3Q isomer
224 C15H24 !-Somer f
225 C0H, ^-benzene isomer T
8 17
240 dibenzofuran T
REFERENCES
1.
Calv
ert JG. Pitt
IS JN ! Photrx-'honi
1 C:-t--T>T7 M/31TT \T^-**Tr T~U«
- - . -, ^- *.* «A** -*-*_, v J. J- • il^,VY J. \_/ J. -CV ,
-------
224 EDO D. PELLIZZARI
5. Gould RF: Photochemical smog and ozone reactions. In:
Advances in Chemistry Series 113, Washington, DC, Amer
Chem Soc, 1976, p 285
6. Matz J: Z Ges Hyg Ihre Grenzebiete 18:903, 1972
7. Fishbein L: Chromatography of Environmental Hazards.
New York, Elsevier Pub Co, 1972, p 499
8. Pellizzari ED: Development of method for carcinogenic
vapor analysis in ambient atmospheres. Research Tri-
angle Park, Environ Prot Agency, EPA-650/2-74-121, 1974,
pp 148
9. Pellizzari ED: Development of analytical techniques
for measuring ambient atmospheric carcinogenic vapors.
Research Triangle Park, Environ Prot Agency, EPA-600/2-
75-075, 1975, pp 187
10. Pellizzari ED: The measurement of carcinogenic vapors
in ambient atmospheres. Research Triange Park, Environ
Prot Agency, EPA-600/7-77-055, 1977, pp 288
11. Pellizzari ED: Analysis of organic air pollutants by
gas chromatography and mass spectroscopy. Research
Triangle Park, Environ Prot Agency, EPA-600/2-77-100,
1977, pp 104
12. Pellizzari ED: The measurement of carcinogenic vapors
in ambient atmospheres. Research Triangle Park, Environ
Prot Agency, Contract No 68-02-1228, in preparation
13. Pellizzari ED, Bunch JE, Berkley RE, Bursey JT: Identi-
fication of n-nitrosodimethylamine in ambient air by
capillary gas-liquid chromatography-mass spectrometry-
computer. Biomed Mass Spec 3:196-200, 1976
14. Pellizzari ED, Carpenter B, Bunch JE, Sawicki E: Col-
lection and analysis of trace organic vapor pollutants
in ambient atmospheres - a technique for evaluating the
concentration of vapors by sorbent media. J Environ
Sci Tech 9:552-555, 1975
15. Pellizzari ED, Bunch J, Carpenter B, Sawicki E: Col-
lection and analysis of trace organic vapor pollutants
in ambient atmospheres - studies on thermal desorption
of organic vapor from sorbent media. J Environ Sci
Tech 9:556-560, 1975
-------
ANALYTICAL TECHNIQUES FOR ORGANICS IN AQUEOUS SAMPLES 225
16. Pellizzari ED, Bunch JE, Berkley RE, McRae J: Collec-
tion and analysis of trace organic vapor pollutants in
ambient atmospheres - the performance of a Tenax GC
cartridge sampler for hazardous vapors. Anal Letters
9:45-63, 1976
17. Pellizzari ED, Bunch JE, Berkley RE, McRae J: Deter-
mination of trace hazardous organic vapor pollutants in
ambient atmospheres by gas chromatography/mass spectrom-
etry/computer. Anal Chem 48:803-807, 1976
18. Pellizzari ED, Bunch JE, Bursey JT, Berkley RE, Sawicki
E, Krost K: Estimation of n-nitrosodimethylamine levels
in ambient air by capillary gas-liquid chromatography/
mass spectrometry. Anal Letters 9:579-594, 1976
19. Pellizzari ED: Identification of components of energy-
related wastes and effluents. Athens, Environ Prot
Agency, Contract No 68-03-2368, in preparation
20. Pellizzari ED, Castillo NP, Willis S, Smith D, Bursey
JT: Identification of organic constituents in aqueous
effluents from energy-related processes. Fuel Chemistry
23:144-155, 1978
-------
STRATEGY FOR COLLECTION
OF DRINKING WATER
CONCENTRATES
Carl C. Smith
Department of Environmental Health
College of Medicine
University of Cincinnati
Cincinnati, Ohio
-------
229
In the first chapter of a recent book on water analysis,
Aaron Rosen reviewed the state of water analysis in 1950
and pointed out the need for a new and entirely different
approach to sampling (1). Thus in 1950, Braus, Middleton,
and Walton (2), using a column of activated carbon to filter
5000 gallons or more of water, were able to recover from
2 to 4 g of organic pollutants by extracting the carbon
with ethyl ether and an additional 2 to 10 g by a second
elution with ethanol. The process was subsequently scaled
up eighty-fold (3); the ether was replaced with chloroform
and water-free extracts of 150-1700 g were obtained. As
recently as 1974, this same procedure was used to prepare
the starting material for a detailed analysis of the organic
pollutants in New Orleans drinking water (4).
Once large samples of organic pollutants became avail-
able, various specialized analytical procedures were applied.
These included infrared spectroscopy and, in particular,
various chromatographic procedures including column, thin-
layer, and gas/liquid chromatography. The introduction
first of gas chromatography followed by coupled gas chromato-
grapny-mass spectrometry led to the detection of an increasing
array of compounds that have been identified in various
drinking water samples. The list published by EPA in 1976
included 398 compounds and the current number exceeds 700 (5).
The problems encountered in using activated carbon are
detailed in Rosen's review and include pore size, pore volume,
surface groups, content of extractable organic matter, and
desorption with different organic solvents. The bases for
-------
230 CARL C. SMITH
these problems were studied at length and led to the final
evolution of a standard material and a standard elution
procedure using chloroform and ethanol (6). In spite of
some inherent difficulties with the procedure, the U.S.
Public Health Service issued Drinking Water Standards -
1962, which specified a limit of 200 ug/1 for the concen-
tration of carbon chloroform extractable or CGE compounds
in order to "avert, if possible, the health hazard of
unidentified and unnumbered industrial organic pollutants"
(7).
At the same time other methods for concentrating and
fractionating various contaminants in water supplies were
being developed. One of the earliest approaches, and one
still employed with many variations, is the sparging or
purge and trap procedure of Bellar and Lichtenberg (8).
Other methods for isolating the purgeable components from
drinking water included those of Rook (9), who was the
first to demonstrate the origin of the trihalomethanes in
drinking water, Mieure et al. (10), who continously swept
the head space gas through a porous polymer trap, and the
method of Zlatkis (11), in which the volatile compounds
in the sample were thermally extracted into the head space.
The latter procedure was used by Dowty et al. (12,13) in
analyzing drinking water. Kopfler et al. (14) applied a
revision of the Bellar and Lichtenberg method to the deter-
mination of volatiles for the National Organics Reconnais-
sance Survey (NORS). The GC/MS techniques for identifying
and quantifying the volatile organics in this study are
described by Lingg et al. (15).
A liquid-liquid procedure for determining halomethanes
in drinking water was developed by Glaze and coworkers (16).
In this procedure, 120 ml samples were collected in serum
bottles in a manner which excluded any head space. When
the sample was returned to the laboratory, 3 to 5 ml of
pentane was added to the bottle using two syringes (one to
add pentane, the other to accept the displaced water) and
the bottle was shaken for 15 minutes at 500 rpm on a
gyratory platform shaker. Glaze et al. suggested that this
procedure offers several advantages over the Bellar and
Lichtenberg procedures: (1) multiple samples can be
processed; (2) the GC procedure is shorter because thermal
desorption is eliminated; and (3) the isothermal GC proce-
dure cuts analysis time.
-------
STRATEGY FOR COLLECTION OF DRINKING WATER CONCENTRATES 231
Another approach for detecting and/or quantifying
volatile impurities in water samples is the closed loop
stripping procedure developed by Grob and coworkers (17).
Some difficulties were experienced in certain laboratories
in applying this method until some additional information
on the equipment and procedure was published in 1976.
The first paper (18) stressed the need to carefully control
certain parameters including extraction (water) temperature;
filter characteristics and amount; stripping flow rate,
duration and temperature; and, finally, desorption from
the special filter. The second paper (19) discussed the
application of narrow and wide bore capillary columns for
the GC analyses and contrasted the typical separation (118
peaks) obtained with a 3 m/2 mm column packed with OV-1
on Gaschrom Q to the greatly increased number of peaks
detected (490) when a 35 m/0.28 mm glass capillary column
coated with an OV-1 film was used.
Investigators at the U.S. Environmental Protection
Agency have prepared a slightly larger model of the Grob
device (20). It will accept a 4 1 water sample and should
provide sufficient materials for analysis and limited bio-
logical testing.
Another approach for monitoring contaminants in water
samples is that described by Junk et al. (21). The Ames,
Iowa, group has primarily used the XAD-2 resin. This resin
is considered to have low polarity, and consists of smooth
white spheres, 0.25-0.5 mm in diameter with surface area of
300 m2/g and average pore size of 90 $. The general flow
chart used is shown in Figure 1.
For grab samples the design shown in Figure 2 is used.
Clear, decanted water samples are fed through the column
at a rate of 25 to 50 ml/min. Sediment, if there is some, is
transferred to the reservoir with several rinses of organic-
free water. The column is eluted with ethyl ether, and 1 to
5 ul of the dried ether extract is injected into a GC. A
typical run from the extract of the Ames, Iowa, municipal
water supply is shown in Figure 3.
Junk et al. (21) also employ another type of resin
column, shown in Figure 4, to process 55 gal samples from
numerous cities throughout the country. This apparatus
provides more material and, of course, has the great ad-
vantage of being rugged, relatively stable, and can be
shipped for subsequent analysis.
-------
232
CARL C. SMITH
SORB
2
g XAD-2
j
ELUTE
25
ml EtjO
i
DRY
H
^1-100 liters\
V H23 )
IDENTIFY(?;
QUANTIFY
SEPARATE
GC/TID+EC
CONCENTRATE
TO 1 ML
IDENTIFY
SEPARATE
GC/MS-COM.
CONCENTRATE
TO 0.1 ML
or Liq. N2 distillation
free evaporate
Figure 1. Flow chart of general resin sorption scheme of
Junk et al. (21).
This group has also been testing raw and finished
water samples from 14 major U.S. water utilities on a
monthly basis for the presence of mutagenic materials,
using the Ames spot test (22). The ether extracts from
their XAD columns are concentrated to 1 ml, 0.1 ml of DMSO
is added and the solution allowed to evaporate until the
ether is gone. Dr. Bonita Glatz, a member of their group,
has applied these DMSO solutions to discs which are applied
to dishes containing Ames strains TA98, TA100, TA1535, TA-
1537 and TA1538 with and without the addition of liver
microsomes from Aroclor 1254-induced rates (S-9). The most
sensitive of the Ames strains for detecting mutagens in
water samples was TA100, a finding confirmed by Loper,
Lang, and Smith (23). Also, they found almost no increase
or decrease in activity in the presence of the rat liver
S-9 fraction. This finding is also in general agreement
with the results presented previously by Loper et al. (23).
The authors obtained a positive mutagenic response with
extracts equivalent to 15 liters of water sample. This
finding is somewhat surprising, since from our results one
would expect a doubling of the mutagenie activity with
either TA98 or TA100 when the extract of about 1.5 liters of
water is applied to the plate. Some possible explanations
for these apparent differences will be considered in the
comments on the reverse osmosis procedure.
-------
STRATEGY FOR COLLECTION OF DRINKING WATER CONCENTRATES
233
Figure 2. Device for extracting organics from grab samples
(21). A: 5-1 reservoir, scaled down; B: glass wool plugs;
C: 24/40 ground glass joint with PTFE sleeve; D: 8 x 140
mm glass tube packed with ~2 g, 40-60-mesh resin; E: PTFE
stopcock.
These methods developed by Junk, Fritz, Svec, and Cris-
well, including the application of the Ames Test, are of
great interest and are described iji more detail in another
chapter in this symposium (22). With the acquisition of
larger (2000 1) samples, more quantitative Ames tests and
extensive fractionation procedures are possible.
The last procedures to be described for concentrating
and fractionating contaminants in potable water samples are
the reverse osmosis methods developed by Kopfler and coworkers
at the Cincinnati Laboratory of the USEPA (14).
-------
234
CARL C. SMITH
o
h-
U
ORGANIC MATERIAL FROM
AMES MUNICIPAL WATER
30 M » .25mm WCOT
COLUMN CARBOWAX 20
Figure 3. Organic material from Ames municipal water (21).
The need for these procedures can be appreciated when
one remembers that at present only about 10-15% of the
organic material in drinking water has been identified.
The volatiles such as chloroform and the other halomethanes,
although easily detected and measured either by the Purge-
and-Trap or the Liquid-Liquid extraction procedures, still
represent less than 10% of the total organic compounds in
water. Because of the shortcomings of all the previous
methods [see Kopfler et al. (14)], the reverse osmosis (RO)
procedure was developed. It appeared to be the only method
that would provide sufficiently large samples of the whole
array of non-volatile constituents to support the various
biological and chemical screens (Figure 5) devised by Dr.
Robert Tardiff and other members of the EPA staff (24).
Some idea of the total amount of organic compound present
-------
STRATEGY FOR COLLECTION OF DRINKING WATER CONCENTRATES
235
Figure 4. Device for extracting organics from composite
samples (21). A: standard garden hose coupling; B: PTFE
washer; C: 12.7 mm ID PTFE tubing; D: glass wool plugs;
E: 12.7 mm OD x 9 cm long glass tube packed with ~2 g,
40-60-mesh resin.
can be obtained by determining total organic carbon (TOC).
In Cincinnati, this varies typically from 1 to 2 mg/1
[Figure 6 (25)]. Considering carbon to be about 50% of
the weight of the organic contaminants present, 1500 liters
will contain 3 to 6 g of organic material. The scheme
shown in Figure 7 has been applied repetitively to water
samples from 5 cities; these were chosen from the group
studied by Keith et al. (26) and shown in Table 1.
The salt brines obtained by lyophilization of the reject
from each membrane were extracted as shown in the diagram.
Originally, only the total ROC-OE and XAD-eluate were pro-
vided but with the demonstration of mutagenicity in these
extracts as described by Loper et al. (23) , the separate
fractions were provided and these have been examined for
-------
236
CARL C. SMITH
PROTOCOL FOR BIO-SCREEN OF ORGANIC
CONCENTRATES FROM TAP WATER
Assay
RANGE-FINDING
(LOse MOUSE)
MUTAGENESIS
(SALMONELLA)
MAMMALIAN CELL
TRANSFORMATION
IN VIVO
CARCINOGEN BIO-
ASSAY (NEONATE)
TERATOGEN
ASSAY (RAT)
CHEMICAL
CHARACTERIZATION
(GC/MS)
Sample/City at 2 month intervals
1
X
?
2
X-f
X-f
?
3
X
?
4
X
?
5
X-f
X-f
?
6
?
?
X-?
Figure 5. Bioscreen of Tardiff et al. (24).
mutagenicity, cell death (toxicity) and in a few instances
for capability of transforming clones of the BALB/3T3 cell
line (27).
From the data in Table 2 one can see that the concentra-
tions of ROC-OE materials as well as XAD adsorable components
vary from city to city. Sometimes they are TS^a^ectedly high
as in the case of Miami. In all cases recovery -appeared to
account for 35%-40% of the total organic carbon (TOC). Al-
though this represents the best recovery reported so far
using large initial water samples, there are some drawbacks
in terms of an ideal methodology:
-------
STRATEGY FOR COLLECTION OF DRINKING WATER CONCENTRATES
237
IOC • mj/l
- — — - Specific Conductance -jumho/cm
Stpltmbtt I Ociobn I No.emoei
TOC AND SPECIFIC CONDUCTANCE - CINCINNATI TAP WATER
Figure 6. Variation over a 3 month period of the TOC and
specific conductance of Cincinnati tap water (25).
1. Most, if not all, the volatile organics are
removed by the procedure.
2. At present there is no way of ascertaining how
well the various compounds in the original water sample
are represented in the final concentrates.
3. It has been shown that some of the components
of the final concentrates are derived from the metal and
plastic which comprise the plumbing, membranes, etc.,
required by the process.
4. Stability of the extracts is not thought a problem.
Previous studies within EPA and elsewhere have shown that
C12 must be removed to stop further reaction to synthesize
trihalomethanes.
5. Although analyses on successive samples appear to
agree, too little information has been accumulated so far
to determine this beyond doubt.
-------
238
CARL C. SMITH
WATER SAMPLE
RO. Cellulose Acetate
R.O. Nylon
Cellulose Acetate
Concentrate
Nylon Concentrate
r
Pentane pH7
i
r
Methylene
Chloride pH7
i
r
Methylene
Chloride pH2
'
\
XAD-2
pH2
\
discard
80%
20% sample
\
Ethanol
Elution
ROC-
OE
80%
/
20% sample
XA D
Eluate
/
Ethanol
Elution
Pentane
i
pH7
I
Methylene
Chloride pH7
i
p
Methylene
Chloride pH2
i
XA
Pi-
r
D-2
12
\
discard
Figure 7. Diagram of the reverse osmosis (ROC-OE) and XAD
eluate devised by Kopfler et al. (14).
6. In spite of the large samples provided by the pro-
cedure, the projected studies in which we hope to fractionate
in greater detail the ROC-OE and XAD eluate using the Ames
Test as a detector are still limited by sample availability.
In view of our findings (27) that all of the city water
supplies examined to date appear to contain significant
amounts of mutagenic material, several areas need further
study:
1. Methods need to be developed which will permit one
to determine the relative mutagenic risk to the human popu-
lation consuming the water of the (1) volatile and (2) non-
volatile fractions of the organic contaminants. At present
EPA, understandably, has focused its initial attention on
-------
STRATEGY FOR COLLECTION OF DRINKING WATER CONCENTRATES
239
Table 1
Type and Location of Various Types of Water Supplies (26)
Source Character
Water Plants
Mississippi River at New
Orleans
1,
2.
3.
•Uncontaminated upland water 1.
2.
Ground water
1.
2.
C/2
-
O
Z
Contaminated by agricultural 1.
runoff 2.
Contaminated by industrial 1.
waste 2.
Contaminated by municipal 1.
waste 2.
Carrollton (City of New
Orleans)
Jefferson Parish #1
(east bank)
Jefferson Parish #2
(west bank)
Seattle, WA
New York, NY
Miami, FL
Tucson, AZ
Ottumwa, IA
Grand Forks, ND
Cincinnati, OH
Lawrence, MA
Philadelphia, PA
Terrebonne Parish, LA
Table 2
Composition of Organic Concentrates from Five Cities
Tap Water
City & Sample Processed, Liters
New Orleans IB
Miami 2
Philadelphia 1
Ottumwa 1
Seattle 1
7,800
2,270
5,800
6,000
13,000
ROC-OE, G*
1.0
1.0
1.8
0.9
0.4
XAD
Eluate, G
5.6
8.8
3.9
6.4
3.0
*Reverse osmosis concentrate, organic extract
-------
240 CARL C. SMITH
the halomethanes and has issued proposed controls on these
suspected carcinogens. The proposed standard for total chloro-
form and related trihalomethanes (CHCl2Br, CHClBr2, and CHBr3)
is 0.1 mg/liter or 100 parts per billion. The Environmental
News for January 25, 1978 contained the preliminary data
which later appeared in the Federal Register (28). On the
basis of limited but repeated studies (National Organic
Monitoring Survey) some 36 cities were named as either having
a possible problem with trihalomethanes (THM) or synthetic
organic compounds (SOC) in their drinking waters. Also listed
are some 40 cities already employing granular activated carbon
(GAG) in their water treatment plants.
It is obvious that there is need to develop reproducible
methodology for collecting and measuring various contaminants
in drinking water. Although the strategy for collecting
samples, the best designs for trapping and eluting organics,
and the most efficient and economical procedures for analyz-
ing these samples are yet to be determined, it should be
pointed out that the EPA is working diligently on these
problems and has engaged the help of groups with expertise
to (1) examine many water sources for a significant number
of organics; (2) to collect in a computerized file GC-MS
data on all known organic water contaminants; and (3) to
develop methodology to determine the chemical identity of
as many of the presently unidentified components as possible.
Studies on mutagenicity of individual halo-organics and other
synthetic organic compounds detected in drinking water are
being carried out as well as the attempts to identify active
components of the non-volatile and presently unknown com-
poments isolated by the ROC-XAD procedures described above.
New and improved procedures for isolating the THM and
quantifying them usually by improved GC procedures are
being described in scientific journals and at various
meetings such as the 1978 Pittsburgh Conference on Analytical
Chemistry and Applied Spectroscopy in which no less than ten
papers in the general area are being presented.
Finally, Leland J. McCabe recently presented (29) a
very thought provoking and penetrating analysis of the
overall drinking water problem in relation to the current
information on trihalomethanes and mutagenicity of drink-
ing water concentrates. His comments, compilations and
suggestions deserve serious consideration by those concerned
with the impact of the Safe Drinking Water Act of 1974.
-------
STRATEGY FOR COLLECTION OF DRINKING WATER CONCENTRATES 241
ACKNOWLEDGMENTS
Special thanks are due to Dr. Robert Tardiff, Dr. Fred
Kopfler, and Ms. Geraldine Wolfe for their assistance with
the manuscript and to Dr. Colin Chriswell for permission to
use certain figures.
REFERENCES
1. Rosen AA: The foundations of organic pollutant analysis,
In: Identification and Analysis of Organic Pollutants
in Water (Keith LH, ed.). Ann Arbor, MI, Ann Arbor
Science Publishers Inc., 1976, pp 3-14
2. Braus H, Middleton FM, Walton G: Organic chemical
compounds in raw and filtered surface waters. Anal
Chem 23:1160-64, 1951
3. Middleton FM, Pettit HH, Rosen AA: The megasampler
for extensive investigation of organic pollutants in
water. Proceedings 17th Industrial Waste Conference,
Engineering Ext Ser 112, 454-460, Purdue University,
Lafayette, IN May 1-3, 1962
4. US Environmental Protection Agency: New Orleans area
wat/<3r supply study. Draft analytical report, Lower
Jjlrississippi River Facility, Slidell, LA, Nov, 1974
Kopfler FC, Coleman WE: Personal communication
6. Middleton FM, Rosen AA, Burttschell RH: Manual for
recovery and identification of organic chemicals in
water. Part II. US Public Health Service, Robert A.
Taft Sanitary Engineering Center, Cincinnati, OH,
May, 1957
7. US Public Health Service: Drinking Water Standards -
1962. Washington, DC
8. Bellar TA, Lichtenburg, JJ: Determining volatile
organics at microgram-per-liter levels by gas chroma-
tography. J Am Water Works Assoc 66:739-744, 1974
Rook JJ: Formation of haloforms during chlorination
<->f natural waters. Water Treatment Exam 23:234-243,
tS.
0}
-------
242 CARL C. SMITH
10. Mieure JP, Mappes GW, Tucker ES, Dietrich MW: Separa-
tion of trace organic compounds from water. In: Identi-
fication and Analysis of Organic Pollutants in Water
(Keith LH, ed.). Ann Arbor, MI, Ann Arbor Science
Publishers, Inc. , 1976, pp 113-133
11. Zlatkis A, Lichtenstein HA, Tishbee A: Concentration
and analysis of trace volatile organics in gases and
biological fluids with a new solid adsorbent. Chromato-
graphia 6:67-70, 1973
12. Dowty B, Carlisle D, Laseter J: Halogenated hydrocarbons
in New Orleans drinking water and blood plasma. Science
187:75-77, 1975
13. Dowty B, Carlisle D, Laseter J: New Orleans drinking
water sources tested by gas chromatography - mass
spectrometry (Occurrence and origin of aromatics and
haologenated aliphatic hydrocarbons). Environ Sci
Technol 9:762-765, 1975
14. Kopfler FC, Coleman WE, Melton RG, Tardiff RG, Lynch SC,
Smith JK: Extraction and identification of organic
micropollutants: reverse osmosis method. Ann N Y Acad
Sci 298:29-30, 1977
15. Lingg RD, Melton RG, Kopfler FC, Coleman WE,
DE: Quantitative analysis of volatile organic
pounds by GC-MS. J Am Water Works Assoc 69:605-61
1977
16. Henderson JE, Peyton GR, Glaze WH: A convenient
liquid-liquid extraction method for the determination
of halomethanes in water at the parts-per-billon level.
In: Identification and Analysis of Organic Pollutants
in Water (Keith LH, ed.). Ann Arbor, MI, Ann Arbor
Science Publishers Inc., 1976, pp 105-111
17. Grob K, Grob G: Organic substances in potable water
and in its precursor. II. Applications in the area of
Zurich. J Chromatogr 90:303-313, 1974
18. Grob K, Grob G: Glass capillary gas chromatography in
water analysis: how to initiate use of the method.
In: Identification and Analysis of Organic Pollutants
in Water (Keith LH, ed.). Ann Arbor, MI, Ann Arbor
Science Publishers, Inc., 1976, pp 75-85
-------
STRATEGY FOR COLLECTION OF DRINKING WATER CONCENTRATES 243
19. Grob K, Zurcher F: Stripping of trace organic sub-
stances from water equipment and procedure. J Chroma-
togr 117:285-294, 1976
20. Kopfler FC: Personal communication
21. Junk GA, Richard JJ, Fritz JS, Svec HJ: Resin sorption
methods for monitoring selected contaminants in water.
In: Identification and Analysis of Organic Pollutants
in Water (Keith LH, ed.). Ann Arbor, MI, Ann Arbor
Science Publishers, Inc., 1976, pp 135-153
22. Chriswell CD: Chemical and mutagenic analysis of water
samples. In: Application of Short-Term Bioassays in
the Fractionation and Analysis of Complexes in Mixtures
(Waters M, ed.). Williamsburg, Virginia, Feb. 21-22,
1978
23. Loper JC, Lang DR, Smith CC: Mutagenicity of complex
mixtures from drinking water. In: Water Chlorination:
Environmental Impact and Health Effects, Vol. 2 (Jolley
RL, Gorchev H, Hamilton DH, Jr., eds.). Ann Arbor,
MI, Ann Arbor Science Publishers, Inc., 1978, pp 433-
450
24. Tardiff RG, Carlson GP, Simmon V: Halogenated
organics in tap water: a toxicological evaluation.
Proceedings of the Conference on the Environmental
Impact of Water Chlorination (Jolley RL, ed.). Oak
Ridge National Laboratory, Oak Ridge, TN, Oct. 22-24,
1975, pp 213-227
Kopfler FC, Melton RG, Mullaney JL, Tardiff RG: Human
exposure to water pollutants. In: Fate of Pollutants
in the Air and Water Environments. Part 2 (Suffet IH,
ed. ). New York, John Wiley & Sons, 1977, pp 419-433
26. Keith LH, Garrison AW, Allen FR, Carter MH, Floyd TL,
Pope JD, Thruston AD, Jr.: Identification of organic
compounds in drinking water from thirteen US cities.
In: Identification and Analysis of Organic Pollutants
in Water (Keith LH, ed.). Ann Arbor, MI, Ann Arbor
Science Publishers, Inc., 1976, Ch 22, pp 329-373
27. Loper JC, Lang DR, Schoeny RS, Richmond BB, Gallagher
PM, Smith CC: Residue organic mixtures from drinking
water show in vitro mutagenic and transforming activity,
in press
-------
244 CARL C. SMITH
28. USEPA: Interim primary drinking water regulations:
Control of organic chemical contaminants in drinking
water. Federal Register Vol. 43 (No. 28): pp 5756-
5780, Feb. 9, 1978
29. McCabe LJ: Health effects of organics in drinking
water. Water Quality Technology Conference. Kansas
City, MO, Am Water Works Assoc, pp 1-11, Dec. 4-6, 1977
\
-------
SECTION 3
CURRENT RESEARCH
-------
-------
-------
-------
SHORT-TERM BIOASSAY OF
COMPLEX ORGANIC
MIXTURES: PART I,
CHEMISTRY
M.R. Guerin, B.R. Clark, C.-h. Ho,
J.L. Epler, and T.K. Rao
Arr'.ytical Chemistry and Biology Divisions
Oak Ridge National Laboratory
Oak Ridge, Tennessee
-------
-------
249
INTRODUCTION
A multidisciplinary program to elucidate the health
and ecological effects of advanced fossil fuels use is
currently underway at the Oak Ridge National Laboratory.
The development of short-term bioassays applicable to the
characterization of complex mixtures is an important aspect
of the program. Studies reported here have been designed to
identify constituents responsible for the mutagenicity of
energy related materials. Observations made in the course
of these studies may prove useful for designing biotesting
methods suitable for the quantitative routine application to
complex mixtures.
PREPARATION OF COMPLEX MIXTURES FOR SHORT-TERM BIOTESTING
The preparation of complex mixtures for in vitro
bioassay is complicated by at least two concerns: (a) the
relevance of the material applied to the test system and,
(b) the compatibility of the material with the test system.
Chemical "relevance" is achieved when the test system is
dosed with a material whose chemical composition mimics
that which reaches the natural (man, plant, animal, soil,
water, etc.) point of impact. Difficulties with "compati-
bility" are encountered when the material being bioassayed
contains constiutents which interfere with the test
organisms' ability to respond to the effect of interest.
Physical properties of the test material may inhibit the
release, for example, of its mutagenic constituents to the
-------
250 ; M.R. GUERIN ET AL.
test bacteria. High concentrations of mildly toxic con-
stituents or small quantities of highly toxic constituents
can mask the more subtle effect of mutagenic constituents.
Complex mixtures are generally entities of poorly de-
fined and continuously changing chemical composition. If,
by virtue of sample history (generation, storage, handling),
the material is altered in its content of bioactive con-
stituents or undergoes other changes which affect the bio-
logical test system, the results of the bioassay may be
invalidated. Whole animal inhalation bioassays of cigarette
smoke, for example, are designed with considerable attention
(11,12,15,18) to the compatibility of the "smoke" being bio-
assayed to that which is freshly generated by the cigarette
under conditions comparable to those representative of the
human situation. Pellet implantation, while highly success-
ful for the application of pure compounds to respiratory
tract epithelium, is questionably applicable to the study of
complex mixtures because (19) the constituents may be released
to the test organism at differing rates depending on their
chemical properties.
Methods for the preparation of complex materials for
bioassay must meet different requirements from those de-
signed for chemical characterization. Methods for quantita-
tive chemical analyses seek to recover 100% of the individual
constituent being determined without regard to the remaining
constituents. Differing degrees of recovery of several
constituents being quantitatively determined in a single
material are acceptable for the purpose as long as the
degrees of recovery are known. For bioassay purposes,
however, quantitative recovery is important only because
it provides the maximum amount of material for bioassay
from the starting material. The critical objective is
to produce a material whose constituents have been recovered
equally. Recovering the constituents to an equal degree
ensures that the composition of the material bioassayed is
the same as that which was sampled.
"Compatibility" with the test system becomes of greater
concern when the researcher moves from whole animal models
to in vitro bioassay systems. Mechanisms of selective ad-
sorption, metabolism, and of detoxification, for example,
embodied in whole animal models may be absent in in vitro
test systems. Observation of a highly toxic reaction to a
test material need not imply the absence of constituents
capable of producing the more subtle biological effects.
-------
SHORT-TERM BIO ASSAY OF COMPLEX ORGANIC MIXTURES 251
The requirements of relevance and compatibility are
themselves incompatible. Any steps taken to remove toxic
constituents or to otherwise make the material compatible
with the test system necessarily involves a change in
physical or chemical nature of the test material. The ob-
jective of methods development here has been to prepare
materials in a form suitable for biotesting but with a
minimal or at least interpretable impact on the relevance
of the test material.
The approach used most successfully to date is to
separate the constituents of the mixture into a manageable
number of chemically distinct fractions. Each fraction is
subsequently bioassayed to determine whether mutagenic
constituents of any type are present independent of syner-
gism, toxic interferences, or other complicating factors
due to the interaction of constituents of widely differing
properties. If it is assumed (or experimentally demonstrated)
that the mutagenicities of the fractions are additive, the
activities of the fractions may be summed to estimate the
activity of the starting material. The primary chemical
requirement is that all of the constituents present in
the starting material are recovered in the fractions to
be biotested.
Acid-base extractive fractionation and gel chromota-
graphic fractionation have been used in an attempt to
identify constituents responsible for the mutagenicity of
synthetic crude oils and to assess their utility for pre-
paring the oils for bacterial mutagenesis testing.
CHEMICAL CLASS FRACTIONATION
Procedures
Acid-base fractionation involves the liquid-liquid
partitioning of the sample between an immiscible organic
solvent and an alkaline or acidic aqueous phase. The
procedure (6,20) used here first subjects the sample to
partitioning between ether and IN NaOH. Acidic constitu-
ents of the sample concentrate in the aqueous phase while
alkaline and neutral constituents concentrate in the organic
phase. The organic phase is subsequently contacted with
IN HC1 which preferentially extracts the alkaline constitu-
ents. The constituents of the aqueous phase can be back-
extracted into ether following pH adjustment. Four fractions
(acids, bases, neutrals, and any,insoluble residue which
may be formed) thus result for biological testing.
-------
252 M.R. GUERIN ET AL.
Additional discrimination is possible by subjecting
the primary fractions to further separation. The procedure
(Figure 1) used here was chosen because of the extensive
literature (1,2,3,4,14,24,25,26) available dealing with its
application to chemical and biological characterizations
of condensed tobacco smokes. Basic constituents are
further separated into those which are water soluble at
pH 11, those which are ether soluble, and those which are
insoluble under these conditions of separation. Acidic
constituents are divided into weakly acid (presumably
phenolic), strongly acidic/water soluble, strongly acidic/
ether soluble, and two insoluble residues. Neutral con-
stituents are divided into any number of subfractions by
Florisil column chromatography or other chromatographic
methods.
The extractive fractionation procedure is advantageous
in that most fractions contain chemically similar constitu-
ents of predictable types, the procedure is applicable to
large sample sizes (25,26), and that it has been demon-
strated (1,2,14) to be effective for elucidating the
biological properties of a complex mixture. An additional
advantage is that it can be applied to both hydrophilic
(e.g., an aqueous effluent) and lipophilic (e.g., a crude
oil) materials. A practical disadvantage is that the
procedure is highly labor-intensive and therefore both
costly and time-consuming when a high discrimination is
required. A potentially more serious concern is that con-
tact with highly acidic and highly alkaline environments
can lead to chemical reactions which may alter the nature
and quantities of bioactive constituents present, thus
invalidating the bioassay.
A new procedure (Figure 2), optimized to fractionate
materials which are predominantly lipophilic, has been
developed (13) for comparison with the extractive method.
The sample is dissolved in hexane and added to a column
of Sephadex LH-20 gel previously equilibrated with methanol/
water (85/15, vol/vol). Lipophilic constituents are eluted
from the column using hexane. Hydrophilic constituents
are subsequently eluted from the column using methanol
and/or acetone. In the second step of the procedure, the
lipophilic constituents are separated into "polymeric,"
"sieved," and hydrogen bonding fractions by eluting the
lipophilic fraction from a column of Sephadex LH-20 using
tetrahydrofuran. The "sieved" material is subsequently
separated according to aromaticity by a column of Sephadex
LH-20 eluted with isopropanol. Because the pore size of
-------
SHORT-TERM BIOASSAY OF COMPLEX ORGANIC MIXTURES
253
ORGANIC
ETHER (OR MeCI2)
1 /V NoOH
AQUEOUS
HEXANE
HEXANE/BENZENE
8/1
BENZENE/ETHER
METHANOL
Figure 1. Acid-base extractive separation of complex
mixtures (21).
the gel is somewhat smaller when equilibrated with isopro-
panol than tetrahydrofuran, additional "polymeric" material
elutes prior to the aliphatic constituents in this third
step of the procedure. The gel chromatographic procedure
thus produces the following size fractions: hydrophilic,
polymeric, hydrogen bonding, aliphatic, simple aromatic,
and polyaromatic.
The primary advantage of the gel chromatographic pro-
cedure is that separation is affected by gentle mechanisms.
The gel acts as an inert support for the methanol/water
phase in the first step of the procedure yielding separation
of the lipophilic constituents from the hydrophilic constitu-
ents by an essentially continuous liquid-liquid partition
mechanism. Molecular sieving and hydrogen bonding are the
primary mechanisms associated with the second step. Final
separation of the "sieved" subfraction obtained in the second
-------
254
M.R. GUERIN ET AL.
SAMPLE OF OIL
STEP I
GEL SWOLLEN IN 85 V% MeOH/15 V% H20 EQUILIBRATED WITH
HEXANE. SAMPLE ELUTED WITH HEXANE.
LIPOPHILIC
FRACTION
STEP II
HYDROPHILIC
FRACTION
GEL SWOLLEN WITH AND ELUTED BY TETRAHYDROFURAN
POLYMERIC
FRACTION
"SIEVED"
FRACTION
I
HYDROGEN-BONDING
FRACTION
STEP III
GEL SWOLLEN WITH AND ELUTED BY ISOPROPANOL
POLYMERIC
FRACTION
ALIPHATIC
HYDROCARBON
FRACTION
I
ONE AND TWO
RING AROMATIC
POLYNUCLEAR
AROMATIC
FRACTION
Figure 2. Gel chromatographic separation of complex
mixtures (14).
-------
SHORT-TERM BIO ASSAY OF COMPLEX ORGANIC MIXTURES 255
step is effected by a combination of size exclusion, ir-bonding
to the gel matrix, and hydrogen bonding. Since this is a
chromatographic procedure, additional precautions such as
excluding light, deoxygenating solvents, and operating under
inert gas blankets could be taken to minimize the likelihood .
of chemical changes occurring.
Discrimination and Reproducibility
Class fractionation procedures are optimized to recover
all of the starting material rather than to isolate quanti-
tatively one class of constituents. This factor, combined
with the extreme chemical complexity of most natural or
anthropogenic mixtures, generally results in fractions which
themselves contain a variety of chemical types. As an
example, the "aliphatic" fraction of Shale Oil B (10) obtained
using the gel chromatographic procedure was further separated
(8) into acids, bases, and neutrals by acid-base partition.
The neutral constituents were then chromatographed on alumina
(neutral, activity I) using stepwise elution with hexane,
benzene, methylene chloride, and methanol. Essentially all
(-98% by weight) of the "aliphatic" fraction was recovered
in the "neutral" subfraction as would be expected. Only
-50% by weight of the "neutrals" were found in the hexane
eluate from alumina with an additional -10% by weight being
found in the benzene eluate. Up to 40% by weight of the
mass of the aliphatic fraction would, therefore, be expected
to contain aromatic or polar moieties or to consist of very
high molecular weight species.
Gas chromatographic retention, mass spectra, and
reference to a detailed study (7) of shale oil composition
suggest (8) the presence of a complex variety (Table 1) of
hydrocarbons in the aliphatic fraction derived from Shale
Oil B. Aromaticity is noted in the benzene elutables and
might be expected to increase in proportion to aliphatic
character in fractions eluted by methylene chloride and
methauol. The aliphatic fraction would thus be considered
primarily "aliphatic" in character but also contains con-
stituents with aromatic moieties. Fractions are not dis-
tinctly different in composition but rather are "enriched"
in a given class of constituents relative to the other
fractions.
Preparative scale class fractionations have generally
been carried out (20,25) in a semiquantitative manner.
As is illustrated in Table 2 for studies carried out here,
-------
256
M.R, GUERIN ET AL.
Table 1
Some Constituents of the "Aliphatic" LH-20
Fraction of a Shale-Derived Oil (8)
Alumina/cyclohexane elutables of neutral subfraction
n CIQ - n C30 (C-,0 - C23 predominating)
Phytane, Pristane
Miscellaneous (
Terpanes
tricyclics
pentacyclics
Steranes
)0 mw = 558
n= 2-8, mw = 276-360
mw = 412
mw = 288-400
Alumina/benzene elutables of neutral subfraction
.CH9-/f-CH,
£ 4
-CH-
R = G - C
1Q
NMR Average
(CH3)2
R
C10H21
R
-------
SHORT-TERM BIOASSAY OF COMPLEX ORGANIC MIXTURES
257
Table 2
Typical Reproducibility and Recoveries of Preparative
Scale Class Fractionation
Extractive Fractionation
Sephadex Gel
Chromatographic Fractionation
Designation
NaOH Insol.
WAj
WAE
SAj
SAE
Bla
BIb
BE
Bw
N/Hexane
N/Hex-Bz
N/Bz-Ether
N/MeOH
Total
Fraction
% By Weight
0.9
0.2
1.9
0.2
1.1
1.6
0.2
2.2
7.3
74.2
4.9
4.7
2.4
102
Rel. Std.
Deviation (%)
34
86
18
41
50
82
29
38
16
2
47
17
17
9
Designation
o
Lipophilic
Hydrophilic3
Total
Polymeric4
Sieved5
H-Bonding5
Total5
Fraction
% By Weight
92.4
6.5
100.2
1.2
92.5
7.7
101
Rel. Std.
Deviation (%)
1.4
12
1.5
34
17
23
19
Coal-derived crude oil "D" (10), four repetitions, sample sizes of 4.4-11.9
grams (20) .
Shale-derived oil "B" (10), five repetitions, sample sizes of 17.5-302.4
grams (13).
Shale-derived oil "B", three repetitions, sample sizes of 17.5-302.4 grams
(13).
From lipophilic fraction of coal-derived oil "D" (10), five repetitions,
4 gram samples (13).
DFrom lipophilic fraction of coal-derived oil "D" (10), eight repetitions,
4 gram samples (13).
-------
258 M.R. GUERIN ET AL.
quantitative reproducibility is particularly poor for
fractions constituting less than 5% by weight of the starting
material. It must be recalled, however, that the purpose
of these (1,2,6) studies has been to identify constituents
contributing to the biological activity of the starting
material rather than to quantify that activity. Little
attention has yet been given to improving the quantita-
tive reproducibility of fractionation procedures. Reduced
sample sizes, simplified fractionation procedures, and a
reduction in the number of manual operations may contribute
to the development of a procedure sufficiently reproducible
for quantitative biotesting.
Mutagenicity
Methods used and significant observations concerning
the mutagenicity of fractionated samples are summarized
elsewhere (6) in these proceedings. Table 3 summarizes
the results of mutagenicity testing of fractions from the
extractive and chromatographic separations of Shale Oil B
for an evaluation of the efficacy of the fractionation pro-
cedures. The mutagenicities of the individual fractions
and their calculated contributions to the mutagenicity of
the starting material are tabulated. The basic fractions
are seen (Table 3) to exhibit the highest mutagenic activi-
ties for Shale Oil B separated by the extractive fractiona-
tion method. The neutral fraction is of substantially
lower specific activity but makes the largest contribution
to the calculated activity of the starting material because
it constitutes the largest part ( 87% by weight) of the
material. The gel chromatographic procedure produces three
fractions of high specific activity. These fractions are
those which would be expected to concentrate constituents
suggested as mutagenic by the extractive fractionation
procedure. Basic constituents would be expected to concen-
trate in the hydrophilic and hydrogen-bonding fractions
while polycyclic aromatic hydrocarbons and azaarenes would
be expected to accumulate in the polyaromatic fraction. The
aliphatic fraction, a somewhat refined version of the neutral
fraction provided by the extractive procedure, is seen to be
the major contributor to the calculated activity of the
starting material by virtue of its quantitative contribution
(60% by weight) to the mass of the material. The gel chroma-
tographic data thus tends to confirm the utility of the
acid-base extraction fractionation procedure.
-------
SHORT-TERM BIOASSAY OF COMPLEX ORGANIC MIXTURES
259
CO
0
r-l
3
73
0
O
0
f\
G
0
•H
P
Ci
G
0
•H
p
O
crf
SH
fa
>,
JD
1^4
CO -P
•H
0 >
rH -H
JO P
crf O
EH
crf
E
0
JH
t~j
CJ
73 >>
0 p be
p .H £
J3 > ^
bc-H >
.,_! p 0
0 O in
fe <
CM
>>
P bfl
•H £
> ~"-^
•H >
P 0
0 rH
•«
C
o
•H
P
0
crf
£H
fa
G
O
•H
•P
crf
G
O
•H
-P
O
crf
J_l
fa
0
>
•H
-p
0
Crf
rH
P
X
pa
*
73 >>
0 P bO
-P -H g
bO-H >
•H -P 0
0 O rH
Ss ,
P bO
•H £
> --.
•H >
p 0
0 rH
. O
rH 03
0 I
iH rH
N/ V
m CM
oo m
>H
/•—> x-s
rH i-L
<5 <;
3s 3s
v_^ v— '
rH CM
73 73
•H -H
0 O
<3 .
rH
O
OH
0
O
^-\
IH
<;
02
v_x
rH
73
•H
O
bO
G
O
IH
P
CO
00 CO
O
rH
o ^
CO CM
rH
O
•H
P
Crf
E
O
0 rH
•H Crf
-P -H
crf Q
.C —
a o
•H G
rH 0
rH rH
\s \/
o o
m co
rH rH
X-S X-S
H S
1>
CO
rH
x-\
erf
rH
03
V^ '
rH
O
•H
CO
crf
03
0> O CO
•tf O C55
CO rH
^
00
CD
CM
•>
rj/l
O
O CM
CM CM
CM
rH
O
•H
•P
Crf
S
O
rH
Crf
1
crf
rH
-P
0
EH
CM 00 -H |> 00
CO 0) l> 1
rH
O CM CO CM
o m CM rH
00 C5 CM rH
73
x~> 0
O x~\ /--\ -p
HH H S Crf
03 03 03 G
S ' \_x V^' O
•H
CM CO •* rH P>
Crf O
O O O rH rJ Crf
• H -H -H -P
rH
Crf
a
crf
CO
r^
-P
C
O
E
CD
CO
P
0
r-3
0*
•H
rH
Crf
0
-P
Crf
rH
Crf
a
0
CO
«
•
03
::
rH
•H
O
0
73
^
rH
O
73
0
j>
•H
rH
0
73
1
0
rH
erf
,c
02
-
•
x— \
CO
•^f
in
CM
rH
5_i
0
rH
O
O
£H
O
1
CO
•t
00
>
p 0
•H £
> P
•H
-P !>>
O ,Q
crf
73
00 0
G P
•H 3
^- -I ^
^> T*^
rH -H
a P
•H CO
-P G
rH 0
3 0
S
p
£>^ r"1
r> bC
•H
73 0
0 £
P
3 73
a 0
£3 ^
O 0
0 >
0
rH 0
Crf 0
•H rH
SH
0 rH
P crf
Crf -P
E 0
p
bO
G >
0 J2 •
a x-v
G CO
CO O v-'
P -H
C P G
crf 0 O
P crf -H
SH SH -P
0 «H O
> crf
0
-------
260 ! M.R. GUERIN ET AL.
The studies yielding the results given in Table 3 were
carried out for qualitative rather than quantitative purposes,
The weighted activities are tabulated and summed, however,
to illustrate the degree of additivity and quantitative repro-
ducibility accompanying these procedures. Summation of
weighted activities from the extractive method indicates that
Shale Oil B exhibits a mutagenicity of 178 rev/mg. The gel
chromatographic method yields a calculated result of 300
rev/mg as compared to an average of 223 rev/mg from three
tests of the unfractionated oil carried out at the same time.
Considering that the oil sample used for the gel chromato-
graphic study was first subjected (13) to azeotropic dis-
tillation and other substantial differences in procedures,
agreement between the methods is good.
The efficacy of summing mutagenicities of fractions to
determine the mutagenicity of the starting material is yet
to be systematically studied. Additivity observed (Table 4)
to date has, however, been generally good.
Table 4
Comparison of Mutagenicity Calculated by Summing
Fractions With That Experimentally Determined
Revertants per milligram
Separation A1 B2 C3 D" E5 F6
Calculated
Determined
300
223
260
223
138
196
167
100
252
350
112
109
'Shale oil separated by gel chromatographic procedure (13).
2Shale oil separated into lipophilic and hydrophilic
fractions.
3Lipophilic fraction separated into polymeric, sieved,
hydrogen bonding fractions.
"Sieved fraction separated into polymeric, aliphatic,
aromatic, polyaromatic fractions.
'Shale oil separated into isopropanol and acetone
elutables from Sephadex LH-20.
6Neutral fraction separated into hexane, hexane/benzene,
benzene/ether, and methanol elutables from Florisil.
-------
SHORT-TERM BIO ASS AY OF COMPLEX ORGANIC MIXTURES 261
IDENTIFICATION OF MUTAGENIC CONSTITUENTS
Acid-base fractionation has consistently shown (6) basic
constituents of synthetic crude oils to exhibit a high
mutagenicity relative to those from petroleum crude oils.
The basic fractions of condensed cigarette smoke have also
been reported (14) to exhibit high mutagenicities. Recent
reports which demonstrate a high correlation of factors
such as tobacco type (17), degree of fertilization (18), and
stalk position (16) with the mutagenicity of the resulting
cigarette smoke condensate further suggest (9) that nitro-
genous constituents are important contributors to muta-
genicity. Candidate nitrogenous constituents are known
to be present in synthetic oils (21,22), cigarette smoke
condensate (23), and airborne particulate matter (5).
The general approach of fractionation and mutagenicity
testing has been used (6) to define better the nature of the
constituents responsible for the observed mutagenic activity.
A procedure (Figure 3) optimized to isolate mutagenic con-
stituents of the ether soluble basic (BE or ESB) fraction has
been developed. Approximately one gram of the EAB from Shale
Oil B and coal synthoil C ["synfuel A" (10)] were eluted from
a basic alumina column using benzene followed by ethanol.
For both oils, the benzene fraction contained (Table 5) from
75-80% by weight of the mass but no more than 2% of the muta-
genic activity. The ethanol fraction, containing essentially
all of the mutagenic activity, was then separated into iso-
propanol and acetone subfractions using Sephadex LH-20 gel.
For both oils, the acetone subfraction was found to contain
approximately 90% of the mutagenic activity of the ESB in
approximately 10% by weight of its mass.
The acetone subfraction constitutes approximately 0.5%
by weight of the original oils. The subfraction is extremely
complex in spite of the 20-fold enrichment. The gas chroma-
togram (Figure 4) resulting from the analysis of acetone
subfraction from coal synthoil C reveals the presence of at
least one hundred individual constituents.
The compositions of the benzene, isopropanol, and
acetone subfractions from both oils are being examined.
Table 6 summarizes observations to date based on gas
chromatographic and mass spectral analysis. All of the
fractions are seen to contain nitrogen heterocyclics.
Pyridines and quinolines predominate in the inactive
benzene fraction while constituents of greater structural
-------
262
M.R. GUERIN ET AL.
ORNL-DWG 78-911
ETHER SOLUBLE
BASE FRACTION
(~1 g)
BASIC ALUMINA
500 ml BENZENE | 700 ml ETHANOL
I I
BENZENE
ELUATE
ETHANOL
ELUATE
SEPHADEX LH-20GEL
250ml ISOPROPANOL I 600 ml ACETONE
1 1
ISOPROPANOL
ELUATE
ACETONE
ELUATE
Figure 3. Isolation of mutagenic constituents of ether
soluble bases.
complexity are enriched in the isopropanol and acetone
fractions. A general feature of these materials is that
they contain a large variety of alkyl-substituted and
partially hydrogenated derivatives of the parent hetero-
cyclics.
The primary difference between the acetone, isopro-
panol, and benzene fractions is that the acetone (highly
mutagenic) fraction contains higher molecular weight
azaaarenes. Constituents such as benzacridines and aza-
benzpyrenes are found in the acetone fraction. Perhaps of
importance is that the exceptionally mutagenic acetone
fraction from the coal-derived oil contains a larger variety
of these constituents than does the shale-derived acetone
fraction. Mass spectral evidence suggests the presence of
azacoronene in coal derived fraction.
-------
SHORT-TERM BIOASSAY OF COMPLEX ORGANIC MIXTURES
263
Table 5
Distribution of Mutagenic Activity in ESB Subtractions
Shale-Derived Oil
Subfraction
Benzene
Isopropanol
Acetone
TOTAL
Average
Weighted
Activity1
0
1
92
93
B
Average
Relative
Weight2
78
13
9
100
Coal-Derived Oil C
Average
Weighted
Activity1
2
0
88
90
Average
Relative
Weight2
76
12
12
100
Percentage of mutagenic activity [TA98, S-9, Aroclor 1254
(6)] of the ether soluble bases (ESB) fraction accounted
for in the subfraction.
Percentage by weight of the ESB.
SUMMARY
Petroleum substitutes produced from coal and shale are
among the materials requiring biological evaluation to assess
environmental and health impacts of new energy technologies.
Intractability and the presence of highly toxic constituents
are among the physical and chemical properties of products
and process streams which complicate short-term biotesting
for subtle health effects. An effective approach is to
separate the starting material into chemically well defined
fractions. Bioassay results obtained for the separated
fractions may be summed to estimate the biological activity
of the starting material. Biological activities of individual
fractions provide evidence as to the types of constituents
responsible for the biological activity of the starting
material.
Liquid-liquid partition from strongly acidic and alka-
line solutions has proven viable for the testing of coal-
and shale-derived oils. A theoretically more gentle separa-
tion procedure based on Sephadex LH-20 gel chromatography
has been found viable for lipophilic materials.
-------
264
M.R. GUERIN ET AL.
c
o
•H
o
aS
3s -t -t
JZ 3 r-.
as CT 0
u 0 C
+•> L, ^
> Cf
I—I — tH
r: 73 -H S,
CU CTJ
I
o o
CU T3 -H »
rs o o a
- ca = 0
Li N CU N
U £2 T3 C
c 0 c c
73 .Q •-. y
CO
CU
1— 1
£2
CSS
EH
CQ
g
aS i
CQ
0)
rH
£5
3
O
CO
Li
0
.C
V 2
W o
F-
0 o
_^ 4
J3 '•
•P S
o -
o
CO »
•P 5
C w
CD
CU
c
0
CU
y
:j
L>
Ss
a
0
[*
# >»
CU ^
C -4
4 O
•o
c in
aB 1
s :n
< O
!
c
c
a
a.
o
L.
a
o
w
3 !
-P
•H
-P
CO
c
CO
c
•5
-H
L.
>t
a
0
L,
•a
'A >>
O £
C -H
B C
- CJ
at 1
< CJ
o
CJ
03
0) CJ
C C
-. st
•a -a
•<-) C
L. w
£ *
Q O
t
o o"
CO
01
C CU
•a B)
-« -u
Li C
>l —
N
-H a
U Q
I
cj^cj"
3S
C
0
c
s a*
« o
•U L,
C 73
< Li
C*J CD
0 U
Q
C
^H
0
q
3
CU L,
q -a
q 4
H LI
*J U
< 5-
cf'cf
1 I
O ;j
c
"o
c
c
0
c
3
?
T3
>,
aJ
W
E«
1!
C
C
?
73
>.
L,
-o
H
J]
CU
c
o
c
5* -
NO
U -
1 ^
^H 3
CJ O-
CO
tu
c
0
c
3
CJ*
'N
^
M
C
^
^
>.
a cu cu
0 C C
L, -H W
>. o -
x: q L.
-» - >.
O 3 a
^•^ >.
cj >» q
1 C »
cj > a
c
c
73
Li
a
CJ
u2
'_
CJ
a
c
^
a
>,
a
iJ
£
CJ
j
c
w
"a cu
al C
C Q
s o
Ed 3
0 rt
L, N
>. O
I|
CJ O
cu
CU
"3
^;
ti
"a
c
3j
0
L|
73
>>
rt
cj°
1
0
CU
cu c
q w
Q J3 TJ
3 I t
3 TH U
^ n at
t- .F- 0
N 0 CU T3
%
cj* tT^2 S
1 | Li +J
— i •-» y cu
CJ CJ < H
s
o cu
L. C
O Q
— , 33
•a «J c
0 «J 73
N M -*
C 53 Li
J5 5 <
vVJ r lfjf *H
N -^ -r^
< Oh
y
' o" o s
I I =
-i Mil
u on
a c
o a
N A
O
u
-------
SHORT-TERM BIOASSAY OF COMPLEX ORGANIC MIXTURES
265
_ s
! ru
£3
O
•H
-P
O
-P
X!
bO
CO -^
O
0 O
G YH
O
-P G
0 O
O
aj O
O
'O ^^
0
X! iH
O -H
•H CO
h X
G 0
0 Q
UD C
a o
•H -H
> -P
O O
r-H 0)
iH -P
O 0
»
CO iH X!
aj CO
aJ a O
«H rH O O
O bC »H O
>
0 CO
aj
0)
bO aj
•H O
n o
w CO
o -H
CO
o 0
a P
o 3
rH G
G -H
o a
-------
266 M.R. GUERIN ET AL.
Studies suggest that alkaline constituents of petroleum
substitutes are major contributors to their Ames Test activ-
ity. Subfractionation of ether soluble bases from a shale-
derived and coal-derived oil has concentrated the bioactive
constituents in a fraction constituting approximately 0.5%
by weight of the starting oil. Nitrogen heterocyclics are
found to be the predominant constituents of this active sub-
fraction.
REFERENCES
1. Bock FG, Swain AP, Stedman RL: Carcinogenesis assay of
subfraction of cigarette smoke condensate prepared by
solvent-solvent separation of the neutral fraction.
J Natl Cancer Inst 49:477-483, 1972
2. Bock FG, Swain AP, Stedman RL: Composition studies on
tobacco. XLIV. Tumor-promoting activity of subfractions
of the weak acid fraction of cigarette smoke condensate.
J Natl Cancer Inst 47:429-436, 1971
3. Bock FG, Swain AP, Stedman RL: Bioassay of major
fractions of cigarette smoke condensate by an acceler-
ated technic. Cancer Res 29:584-587, 1969
4. Chamberlain WJ, Stedman RL: Fractionation of tabacco
smoke condensate for chemical composition studies. In:
The Chemistry of Tobacco and Tobacco Smoke (Schmeltz I,
ed.). New York, Plenum Press, 1972, pp 99-105
5. Dong MW, Locke DC, Hoffmann D: Characterization of
aza-arenes in basic organic portion of suspended
particulate matter. Environ Sci and Tech 11(6):612-
618, 1977
6. Epler JL, Clark BR, Ho C-h, Guerin MR, Rao TK: Short-
term bioassay of complex organic mixtures: Part II,
mutagenicity testing. Proc of Symposium on Application
of Short-Term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures (these
proceedings)
7. Gallegos EJ: Terpane-sterane release from kerogen by
pyrolysis gas chromatography-mass spectrometry. Anal
Chem 47(9):1524-1528, 1975
-------
SHORT-TERM BIOASSAY OF COMPLEX ORGANIC MIXTURES 267
8. Goeckner NA: Western Illinois University, Macombe,
Illinois, Oak Ridge National Laboratory Faculty Parti-
cipant, 1977-1978, private communication
9. Griest WH, Guerin MR: Influence of tobacco type on
smoke composition. In: Recent Advances in Tobacco
Science, Vol 3. Tobacco Smoke: Its Formation and
Composition. Proc 31st Tobacco Chemists' Research
Conference, 1977, pp 121-144
10. Guerin MR, Epler JL, Griest WH, Clark BR, Rao TK:
Polycyclic aromatic hydrocarbons from fossil fuel con-
version processes. Proc Second International Symposium
on Polynuclear Aromatic Hydrocarbons (in press)
11. Guerin MR, Maddox WL, Stockely JR: Tobacco smoke in-
halation exposure: concepts and devices. Proc
Tobacco Smoke Inhalation Workshop on Experimental
Methods in Smoking and Health Research. DHEW Pub No
(NIH) 75-906:31-44, 1976
12. Guerin MR, Stokely JR: Proband-machine animal inter-
actions in inhalation bioassays. Proc Tobacco Smoke
Inhalation Workshop on Bioassay Models and Inhalation
Toxicology. DHEW Pub (in press)
13. Jones AR, Guerin MR, Clark BR: Preparative-scale
liquid chromatographic fractionation of crude oils
derived from coal and shale. Anal Chem 49:1766-1771,
1977
14. Kier LD, Yamasaki E, Ames BN: Detection of mutagenic
activity in cigarette smoke condensates. Proc Nat
Acad Sci USA, Vol 71(10):4159-4163, 1974
15. McGill HC: The human model. Proc Tobacco Smoke In-
halation Workshop on Bioassay Models and Inhalation
Toxicology, DHEW Pub (in press)
16. Mizusaki S, Okamoto H, Akiyama A, Fukuhara Y: Relation
between chemical constituents of tobacco and mutagenic
activity of cigarette smoke condensate. Mutat Res
48(3/4):319-325, 1977
17. Mizusaki S, Takashima T, Tomaru K: Factors affecting
mutagenic activity of cigarette smoke condensate in
Salmonella typhimurim TA1538. Mutat Res 48(l):29-36,
1977
-------
268 M.R. GUERIN ET AL.
18. Reist PC: Particle size and its role in physical and
chemical interactions of tobacco smoke aerosol. Proc
Tobacco Smoke Inhalation Workshop on Bioassay Models
and Inhalation Toxicology. DHEW Pub (in press)
19. Rubin IB, Guerin MR: Chemical evaluation of the beeswax
pellet implantation bioassay model for studies of
environmental carcinogens. J Natl Cancer Inst 58(3):
641-644, 1976
20. Rubin IB, Guerin MR, Hardigree AA, Epler JL: Fractiona-
tion of synthetic crude oils from coal for biological
testing. Envir Res 12:358-365, 1976
21. Schiller JE: Nitrogen compounds in coal derived liquids.
Anal Chem 49(1):2292-2294, 1977
22. Shultz JL, White CM, Schweighardt FK, Sharkey AG:
Characterization of the heterocyclic compounds in coal
liquefaction products. Part I: nitrogen compounds.
Pittsburgh Energy Research Center Report PERC/RI-
77/7, 1977
23. Snook ME, Arrendale RF, Higman HC, Chortyk OT: Iso-
lation of indoles and carbazoles from cigarette smoke
condensate. Anal Chem 50(1):88-90, 1978
24. Swain AP, Bock FG, Cooper JE, Chamberlain WJ, Strange
ED, Lakritz L, Stedman RL, Schmeltz I, Russell RB:
Further fractionations of cigarette smoke condensate
for bioassays. Weak acid and neutral subfractions
and combinations of active fractions. Beitr Tabak-
forsch 7:1-7, 1973
25. Swain AP, Cooper JE, Stedman RL: Large scale fractiona-
tion of cigarette smoke condensate for chemical and
biologic investigations. Cancer Res 29:579-583, 1969
26. Swain AP, Cooper JE, Stedman RL, Bock FG: Composition
studies on tobacco XL. Large scale fractionation of
the neutrals of cigarette smoke condensate using ad-
sorption chromatography and solvent partitioning.
Beitr Tabakforsch 5:109-114, 1969
-------
SHORT-TERM BIOASSAY OF
COMPLEX ORGANIC
MIXTURES: PART II,
MUTAGENICITY TESTING
J.L. Epler, B.R. Clark, C.-h. Ho,
M.R. Guerin, and T.K. Rao
Biology and Analytical Chemistry Divisions
Oak Ridge National Laboratory
Oak Ridge, Tennessee
-------
271
INTRODUCTION
The feasibility of using short-term mutagenicity assays
to predict the potential biohazard of various crude and com-
plex test materials has been examined in a coupled chemical
and biological approach. The principal focus of the research
has involved the preliminary chemical characterization and
preparation for bioassay, followed by testing in the Salmon-
ella histidine reversion assay described by Ames (1). The
mutagenicity tests are intended to (a) act as predictors of
profound long-range health effects such as mutagenesis and/
or carcinogenesis, (b) act as a mechanism to rapidly isolate
and identify a hazardous biological agent in a complex mix-
ture, and (c) function as a measure of biological activity
correlating baseline data with changes in process conditions.
Since complex mixtures can be fractionated and approached in
these short-term assays, information reflecting on the actual
compounds responsible for the biological effect may be ac-
cumulated. Thus, mutagenicity tests will (d) aid in identi-
fying the specific hazardous compounds involved and in estab-
lishing priorities for further valid testing, testing in
whole animals, and more definitive chemical analysis and
monitoring.
Our work has emphasized test materials available from
the developing synthetic fuel technologies (2). However,
the procedures are applicable to a wide variety of industrial
and natural products, environmental effluents, and body
fluids. The general applicability of microbial test systems
has already been demonstrated with, for example, the use of
the assay as a prescreen for potential generic hazards of
-------
272 J.L. EPLER ET AL.
complex environmental effluents or products, e.g., tobacco
smoke condensates (3), natural products (4,5), hair dyes
(6), soot from city air (7), fly ash (8), and, in our work
with synthetic fuel technologies, oils and aqueous wastes
(9,10).
BIOASSAY METHOD
For the study of application of mutagenicity testing
to environmental effluents and crude products from the syn-
thetic fuels technology, we performed preliminary screening
with the highly sensitive Ames histidine reversion strains
known to repond to a wide variety of known mutagens/carcino-
gens. The working hypothesis was that sensitive detection
of potential mutagens in fractionated complex mixtures could
be used to isolate and identify the biohazard. In addition,
the information could be helpful in establishing priorities
for further testing, either with other genetic assays or
carcinogenic assays.
The Salmonella strains used in the various assays are
listed below. All strains were obtained through the cour-
tesy of Dr. Bruce Ames, Berkeley, California.
Salmonella typhimurium Strains
TA1535 hisG46, uvrB, rfa (missense)
TA100 hisG46, uvrB, rfa (missense plus R factor)
TA1537 hisC3076, uvrB, rfa (frameshift)
TA1538 hisD3052> uvrB, rfa (frameshift)
TA98 hisD3052, uvrB, rfa (frameshift plus R factor)
In the screening of fractionated materials, the two
strains TA98 and TA100 were generally employed. Standard
experimental procedures have been given by Ames et al. (7).
Briefly, the strain to be treated with the potential muta-
gen(s) is added to soft agar containing a low level of histi-
dine and biotin along with varying amounts of the test sub-
stance. The suspension containing approximately 2 x 108
bacteria is overlaid on minimal agar plates. The bacteria
undergo several divisions with the reduced level of histidine,
thus forming a light film of background growth on the plate
and allowing the mutagen to act. Revertants to the wild-
type state appear as obvious large colonies on the plate.
The assay can be quantitated with respect to dose (added
amount) of mutagen and modified to include "on-the-plate"
treatment with the liver homogenate required to activate
metabolically many compounds.
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING 273
Fractions and/or control compounds to be tested were
suspended in dimethyl sulfoxide (supplied sterile, spectro-
photometric grade from Schwarz/Mann) to concentrations in
the range of 10-20 mg/ml solids. The potential mutagen was
in some cases assayed for general toxicity (bacterial sur-
vival) with strain TA1537. Normally, the fraction was
tested with the plate assay over at least a thousand-fold
concentration range with the two tester strains TA98 and
TA100. Revertant colonies were counted after 48 h incuba-
tion. Data were recorded and plotted versus added concen-
tration, and the slope of the induction curve was determined,
It is assumed that the slope of the linear dose-response
range reflects the mutagenic activity. Positive or ques-
tionable results were retested with a narrower range of
concentration. All studies were carried out with a parallel
series of plates plus and minus the rat liver enzyme prepara-
tion (7) for metabolic activation. Routine controls demon-
strating the sterility of samples, enzyme or rat liver S-9
preparations, and reagents were included. Positive controls
with known mutagens were also included in order to recheck
strain response and enzyme preparations. All solvents used
were nonmutagenic in the bacterial test system.
SAMPLES
Samples tested and their sources are listed below:
• Coal-liquefaction product from a process under
development, courtesy of the Pittsburgh Energy
Research Center (Synfuel A), or Coal A from ORNL
repository.
• Coal-liquefaction product from the COED Pyrolysis
Process, courtesy of FMC (Synfuel B), or Coal B
from ORNL repository.
• Louisiana-Mississippi sweet crude oil, courtesy
of Dr. J.A. Carter of the Analytical Chemistry
Division, Oak Ridge National Laboratory.
• Composite crude oil sample from materials ob-
tained through the courtesy of Dr. Dee Latham of
the Laramie Energy Research Center.
• A crude shale-oil sample (B) from the above-
ground simulated in situ oil shale retorting
process.
-------
274 J.L. EPLER ET AL.
• The aqueous product water consisting of the centri-
fuged water of combustion from the same process
(both samples 5 and 6 courtesy of the Laramie
Energy Research Center).
• A coal-gasification aqueous condensate from a
process under development, courtesy of Pittsburgh
Energy Research Center.
• A separator liquor from a coal-liquefaction pro-
cess, courtesy of FMC.
The authors recognize the possibility that these samples
may bear no relationship to the process as it may exist in
the future, nor should it be construed that these materials
are representative of all natural crudes, synthetic, or shale
oil processes. They are used here simply as appropriate and
available materials for the research.
BIOASSAY RESULTS
Class Fractionation
Oil Samples. The bulk of the samples listed above were
subjected to the fractionation scheme described by Swain et
al. (11), as modified by Bell et al. (12). The scheme is
described in detail as applied to oils in Rubin et al. (13)
and in the first part of this presentation. As an example,
a summary of the results from a sample of Synfuel A-2 (9) is
given in Table 1. Subfractionation results are shown with
the neutral fraction chromatographed on a Florisil column.
The column was eluted with the solvents shown and, with this
sample, collected in one fraction. The data includes the
analytical weight analysis of the sample (column 1) along
with the specific mutagenic activity (slope of dose-response
curve) of each fraction (column 2). The product of these
(column 3) represents a weighted value of each fraction rela-
tive to the contribution to the starting test material.
Mutagenic activity is seen in both the acidic and basic
fractions along with the neutral subfractions. However, the
major contributors to the mutagenicity appear to occur in
the basic fractions, with activities also consistently pres-
ent in the neutral materials.
A comparison of these activities and the total mutagenic
potential of the various oil and aqueous samples is given in
Table 2-. Reasonable reproducibility is seen in similar
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING
275
Table 1
Distribution of Mutagenic Activity
of Synthetic Oil1 (Synfuel A-2)
Relative
Weight
(% of
Fraction2 total)
NaOHj
WAj
WAE
SAI
SAE
SAW
Bla
BIb
BE
Bw
Neutral
TOTAL
Neutral Subfractions
Hexane
Hexane/ benzene
Benzene/ether
Methanol
Subtotal
Initial sample, g
Chromatographed, g
20.9
2.2
4.9
<0.1
0.4
0.4
6.8
0.1
2.0
0.6
69.2
107.6
72.7
5.0
19.8
2.3
99.8
26.166
10.664
Specific
Activity3
(rev/mg)
1,700
180
1,260
30
130
120
38,700
1,270
36,200
570
583 (570) 5
340
710
1,360
1,460
Weighted
Activity"
(rev/mg)
356
4
62
0
1
1
2,633
1
725
3
403
4,189
244
35
270
34
583
•
-------
276 J.L. EPLER ET AL.
'All assays carried out in the presence of crude liver S-9
from rats induced with Aroclor 1254.
2I = insoluble (fractions a and b), E = ether soluble, W =
water soluble, WA = weak acid, SA = strong acid, B = base.
3rev/mg = revertants/mg, the number of histidine revertants
from Salmonella strain TA98 by use of the plate assay with
2 x 10* bacteria per plate. Values are derived from the
slope of the induction curve extrapolated to a milligram
value. NT = not tested.
*Weighted activity of each fraction relative to the starting
material is the product of columns one and two. The sum of
these products is given as.a measure of the total mutagenic
potential of each material.
sComparable to "specific activity," but based on the activity
of the total neutral fraction rather than the summation of
the individual fraction.
samples, e.g., Synfuel A-l and A-2, and Synfuei B-l and B-2.
Synfuel A-3 represents the same material without prior cen-
trifugation of the solids. The consistency of activities
seen in all oils considered is illustrated. On a relative
scale, the synthetic fuels show more mutagenic activity than
the natural crude "control" samples shown. Shale oil appears
to be only slightly higher than the natural crudes. Refer-
ences are given to the complete published compilations on
these samples. See Table 2.
Each determination represents the slope of the dose-
response curve. All testing was carried out in the presence
of the rat-liver microsomal activation system. Slight
mutagenic activity without enzyme treatment was occasionally
noted.
The routine screening employed strains TA100 (missense)
and TA98 (frameshift); however, complete strain-specificity
tests were carried out with selected materials. Fractions
giving a positive response with straing TA98 were, in
general, also positive with the other frameshift strains,
TA1537 and TA1538. Additionally, positive results were rou-
tinely noted with the sensitive missense strain TA100; how-
ever, reversion of the missense strain TA1535 was rare.
TA98 appeared to be the best general indicator of mutagenic
activity of these materials. Furthermore, liver preparations
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING
277
CM
CO C
o
s: -H
-p -P
•H cd
> C
0
CO vi
•p p
iH O
3 aj
Ws_.
rH
CU ft*
prj
CO
M CO
C ccj
•H r— 1
-P O
CO
0) 1
&~*
CO
•P rH
•H ft
o s
•r^ (t)
C CO
0)
hO
aj
rj
fe-J
"^
O • CU
EH -p (H
• bO
O • S
CU -P -^-
ft 0 >
rH CO < CU
h ^
3
CU -P
2; ^
'~*
),--) >^,/
cu
oi
• bfl
0 • £
CU -P -^
a o >
en < 0
£*
«*^
W
CQ
•P
* ^
rH^-
CO
oi
•—>
0 • S
cu -P ~-^
a o >
CO < CU
^
v^
*
co -P
*
•s*
l—i V-X
cu
CB
cu
i— i
a.
E
aj
CO
03
v^
,_4
rj*
CM
^
CM
CM
oo
o
m
rH
CM
O
O
0
rH
0
,_!
1
0)
•a
^
u
0)
>r^
CO
o
a
£
0
CJ
O O3
^_^
C- CD
T}* [*•
rH
CO O
CD 03
rH
CN t>
"tf1 O
oo oo
0 O
O 00
in rH
CM CM
O 0
O 0
O CM
rH rH
O 0
CM
1
CU
•a
3
!H
O CU
•a
0) 3
•P !H
•rH CJ
CO
O CO
a s
£ 1
0 <
CJ J
O 03
, .
00 CM
t> CO
rH O
"*
CM [>•
rH in
r> co
CD CO
00 O
CM O
m o
03 03
•*
co
CM
rH O
!> CM
0 O
CD ^
rH CM
CD CO
0 O
rH
1
rH ^
•H
O rH
CU
0) 3
rH IN
OS C
JS >,
vi en
O3 O3
O3 CO
00 O
rH CO
TJ. t-
CO ^J*
00 O3
in o
rH
CM ^
O3 CO
co m
O 0
o o
CM CO
• •
CD CO
CO *sj*
O rH
CM CO
0 O
CM rH
rH O
rH
^ CO
O O
CM CO
1 1
i-H rH
CU CU
3 3
>»
CO W
03 03
CD ^J1
rH CO
in i<
o in
CO CO
in \f
CO CO
CM 03
oo oo
o o
o o
m oo
K M
rH CO
co co
CM rH
O O
-gi CD
O rH
-H CM
1 1
CQ CO
rH rH
CU CU
3 3
» >»
00 OT
»
"H
3
a
3
r-
rH
O
in
rH
O
m
co
m
o
o
O3
CO
m
^•4
O
3
O1
•H
J
SH >
•P &
ai
h CO
OS •
ftrH
CU
w
^
^,
rH
rH
CM
O
o
rH
03
rH
O
O
o
•
^
in
0
o
00
0
CO
cu
at
CO
a
cu
•a
a
0
^
SH ^-*,
CU f
•H
m
^
rH
C-
CM
O
O
in
co
rH '^V
0) p-»
L A 1
OS O 6*
* r>
CO rH *
CO cd
CU £ O
0 CO •
0 -^rH
OH
C
0
+J
(^
08
a
3>
a,
*CJ
cu
o
^
c
•H
1
{L«
0
rH
0
o
u
<
•p
•H
*
a
o
*M
•P
>
•rH
•P
0
aj
O
•H
rH
O
£3
o!
•P
0)
.C
•p
•H
S
00
03
SH
C
•tH
aS
£4
•p
CO
-------
278 J.L. EPLERETAL.
from rats induced with Aroclor 1254 (a gift from Monsanto)
showed the best general applicability. However, individual
differences in effectiveness do occur; for example, variously
induced preparations show obvious differences between basic
fractions and, e.g., the neutral/methanol fraction (9). An
Aroclor-induced preparation reacts best with the neutral
fraction (polynuclear aromatic hydrocarbons?), while a pheno-
barbital-induced preparation works more efficiently with the
basic fraction (heterocyclic nitrogen compounds?).
Primary candidates for the mutagens (and carcinogens?)
responsible for activity in the basic fractions include
quinoline, substituted quinolines, alkyl pyridines, acridine,
naphthylamines, aza-arenes, benzacridines, and aromatic
amines; in the neutral fractions, potential threats may con-
sist of benzanthracenes, dibenzanthracenes, substituted
anthracenes, benzopyrenes, benzofluorenes, pyrene, substi-
tuted pyrenes, and chrysenes (see Ho et al., 15). Thus,
work with these pure compounds is being carried out concur-
rently.
Reproducibility of results was shown by comparison of
data from similar samples. Although discrepancies exist
from fraction to fraction, the general trend is apparent,
and the sum of activities appears to be roughly reproducible.
Again, when the major component, neutral fraction is assay-
able as with the Synfuel A, the summation of the subfraction
values of the neutrals reflects the approximate additivity
of the individual mutagenic determinations. For example,
570 revertants/mg with a direct assay of the neutrals from
Synfuel A-2 compares with 583 revertants/mg based on the
summation (Table 1).
An overview of the results points to a number of con-
sistencies: (1) all crudes and synthetic fuels showed some
mutagenic potential; (2) the neutral and basic fractions
showed activities regardless of the source of the sample;
and (3) the relative total mutagenic potentials varied over
two orders of magnitude. Whether these results reflect a
comparative biohazard of processes still under development
is not the point in question here. The results simply show
that biological testing—genetic reversion assays in this
case—can be carried out with the newly developed tester
systems, but only when coupled with the appropriate analyt-
ical separation schemes. Conceivably, the use of this
approach could provide rapid information concerning health
effects.
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING 279
Aqueous Samples. Table 2 also lists sample results from
a group of aqueous samples subjected to the class fractiona-
tion procedure (Stedman procedure). In general, greater
activity is seen in the more polar, more water soluble frac-
tions rather than in the nonpolar neutral materials. Cau-
tion has to be extended on work with any aqueous material
because of the high potential for instability. Although we
have used organic extraction here, techniques with resin con-
centration, e.g., XAD-2, may prove useful with aqueous sam-
ples (16,17). Only in exceptional cases is the mutagenic
activity directly observable in an unconcentrated sample.
Liquid Chromatographic Fractionation
In the initial studies with coal liquefaction products,
the crude oils were fractionated using the scheme originally
developed for cigarette smoke condensates (Stedman procedure),
The scheme yields class separations based on the relative
acid-base properties of the components. The samples are par-
titioned between ethyl ether and 1 N NaOH in a single-stage,
continuous procedure to yield an aqueous acid fraction and
organic phase base and neutral fractions. The organic frac-
tion is extracted with 1 N HC1 to yield an aqueous basic
fraction and an organic basic fraction and an organic neutral
fraction. The neutral material is subsequently subfraction-
ated on a Florisil column. These primary subfractions are
then subjected to mutagenicity testing.
Realizing the potential for modification of the com-
ponents within the procedure, we moved to consideration of a
number of other methods. The fractionation procedure using
Sephadex LH-20 can provide a gentle and large-scale class
separation for (initially) crude oils from shale oil and coal
liquefaction processes. The procedure involves three steps
using the gel in different modes:
• Lipophilic-hydrophilic partitioning.
• Molecular size separation.
• Aliphatic-aromatic separation.
The procedure (18) was designed by Jones, Guerin, and Clark
of the Analytical Chemistry Division. Using fractions pre-
pared as above, we have started a comparison of this pro-
cedure and the Stedman procedure for usefulness in prepara-
tion for bioassay. The preliminary mutagenicity studies
-------
280 J.L. EPLER ET AL.
confirm the suitability and utility of the method. Table 3
summarizes some of the results from shale oil. The method
appears to be generally applicable to complex organic mix-
tures and achieves the goal of presenting a gentle and rapid
separation scheme, useful with large-scale samples.
Subfractionation of Neutral Components. The polycyclic
aromatic hydrocarbons (PAH), presumably occurring in the
neutral fractions of the various schemes noted, have been
listed as major contributors to mutagenicity of the test
materials. With natural crudes, these components appear to
account for the bulk of the activity. With synthetic crudes,
the contributions of both the basic and neutral fractions
must be considered. Further work is needed to define (iso-
late and identify) the mutagenic components of these impor-
tant classes.
We have carried out a preliminary study with synfuel
PAHs subfractionated and detected by the short-term mutagen-
icity assay. As shown in Table 4, shale oil (sample B) can
be separated into lipophilic and hydrophilic fractions with
the Sephadex LH-20 partition chroinatography described pre-
viously (18,19). Further separation of the lipophilic
fraction is achieved by neutral alumina and LH-20 using
various solvents. The various subfractions can then be
assayed for mutagenicity with the Salmonella histidine re-
version system. As seen in Table 4, activity seems to peak
in the 4- and 5-ring subfractions, those containing known
carcinogens/mutagens as benzo(a)pyrene, benzo(c)phenanthrene,
and 3-methylcholanthrene.
Subfraction of Basic Components. Again, considering the
results with class fractionation procedures, we developed a
procedure (20) specifically designed for subfractionation of
the basic materials, now realized to be a major contributor
to mutagenic activity. An elution sequence using alumina and
Sephadex LH-20 gel with a combination of solvents isolates 90%
of the mutagenic activity from basic compounds into 0.5 wt%
fraction of crude oil.
A basic alumina column eluted first with benzene and
then ethanol isolates the mutagenic components of the ether
soluble base fractions (ESB) of synthetic crude oils into a
fraction of about 25% of the ESB. A further separation is
achieved by eluting the ethanol isolate through a Sephadex
LH-20 gel column with isopropanol followed by acetone. About
90% of the basic mutagenic activity is recovered in the ace-
tone subfraction which comprises -0.5 wt% of the crude oil.
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING 281
Table 3
Sephadex LH-20 Fractionation of Shale Oil Coupled With
Mutagenicity Testing
Test
Material-Fraction
Crude Oil
Hydrophiiic
Lipophilic
• Polymer
• Hydrogen Bonding
• Sieved
•• Polymer
•• Aliphatics
Aromatics
• 1 and 2 Ring
• 3 and 4 Ring
• Polynuclear
% of Total
100
6
93
5
5
84
1
60
14
5
4
Specific*
Activity
(rev/mg)
233
1300
196
54
1040
100
0
180
24
132
1220
*Slope of dose-response curve with Salmonella strain TA98
plus rat-liver preparation induced with Aroclor 1254.
Development of this separation scheme was made possible using
the Ames microbial mutagenesis assay as the detector during
exploratory liquid chromatographic separations. Table 5
lists some of the preliminary data from these studies.
COMPARATIVE MUTAGENESIS
In order to validate and compare the results accumulated
in the Ames system with complex test materials from synthetic
fuel technologies, we selected specific fractions or sub-
fractions on the basis of their activity in the histidine
-------
282 J.L. EPLERETAL.
Table 4
Subfractionation of Neutral Components From Shale Oil:
Distribution of Polycyclic Aromatic Hydrocarbons
and Mutagenic Activity
Specific Activity*
rev/mg of Fraction
Subfraction
Aromatic Fraction
I (polymeric)
II (1-ring)
III (2-ring)
IV (3-ring)
V (4-ring)
VI (5-ring)
VII (<5-ring)
TOTAL
Weight %
100
5.7
47.0
33.7
8.0
2.7
0.6
0.5
98.2
Without
S-9
60
0
0
0
0
1600
2600
600
62
With
S-9
170
0
0
0
1000
4000
3800
1500
214
*Number of histidine revertants from Salmonella strain TA98
by use of plate assay with 2 x 108 bacteria per plate.
Values derived from slope of the induction curve. "With S-9"
indicates test carried out in presence of crude enzyme prep-
arations from rats induced with Aroclor 1254.
reversion assay for further testing in the various other
tests designed to detect mutagenicity. Preliminary results
have been published in the Proceedings of the Second Inter-
national Conference on Environmental Mutagens, Edinburgh,
1977 (21). For the purposes of a qualitative comparison,
the results are given in Table 6. The selected fractions
or subfractions used were basic and neutral isolates from
synthetic crude oils from coal liquefaction processes [Syn-
fuel A and B as described in Epler et al. (9)]. With
Drosophila (22) and in the mammalian cell gene mutation
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING 283
Table 5
Subfractionation of Basic Fraction:
Distribution by Weight and Mutagenic Activity
Shale oil Synfuel A-3
wt% rev/mg* wt% rev/rag*
Basic fraction (A) 100 2,500
Alumina
Benzene (B) 78 600
Ethanol (C)
LH-20
Isopropanol (D) 12 0
Acetone (E) 10 20,000
100 30,000
76 0
- -
12 0
12 222,000
*Assayed with Strain TA98 with Aroclor-induced preparation.
assay (23), the detection has been a function of newly de-
veloped fractionation schemes (e.g., the use of LH-20) (39,40),
that result in higher specific activity (more highly purified)
mutagenic subfractions. In general, the results validate the
initial screening carried out in the Salmonella assay, but
these other systems have not as yet been used to test ex-
haustively materials that are negative in the Ames system.
Note also, however, that the preliminary results of Generoso
(personal communication) show that the crude synthetic fuel
does induce dominant lethals in mice although the basic frac-
tion alone appears to be negative.
For the comparative studies with microbial systems given
here, we selected four Synfuel fractions. The results with
the frameshift strain TA98 with metabolic activation were
considered. Fractions 6 [strong acid, water soluble (SA )];
7 [base insoluble, fraction A (B )]; 9 [base, ether soluble
(B )]; and 14 (neutrals/methanol) were selected on the basis
of their ability to revert the frameshift alleles of the
Ames system. In order to validate the mutagenicity results
obtained from the Salmonella histidine-reversion system, we
-------
284
J.L. EPLER ET AL.
Table 6
Comparative Mutagenesis of Fractions from
Synthetic Crude Oils1
Test System
Salmonella
E. coli
Yeast
Drosophila
CHO cells
Assay
his~*his
arg~-arg
gal^gal*
his"— his
CAN8— canr
SLRL
6-thioguanin
Basic Neutral Crude2
Fraction Fraction Synfuel
+ + NT
+ + NT
+ + NT
+ + NT
+ - NT
e + NT NT
Human
leukocytes
Mouse
Carcinogenesis
Resistance
Chromatid
Aberrations
Dominant
Lethals
Skin Painting3
P
P
NT
'For references to published work or work in progress, see
text. The fractions used were generally those from Synfuel
A-3 or Synfuel B-2. + = mutagenic; - = nonmutagenic;
NT = not tested; and P = in progress.
2Crude synfuels are generally too toxic to test in most
systems.
3Work of J.M. Holland, Oak Ridge National Laboratory, in
progress.
extended the treatment with the selected test fractions to
the E. coli 343/113 system of Mohn (24). The results ob-
tained in the forward (gal") and reverse-mutation (arg+)
assays with E, coli support the results obtained with
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING 285
Salmonella. Both the basic fraction (#9) and the neutral
subfraction (#14) are mutagenic upon metabolic activation
with Aroclor-induced rat-liver homogenate (S-9).
Further validation of the bacterial results was obtained
by assaying for both forward mutation and reverse mutation in
the yeast system (21,25). The Synfuel A fractions tested
were weakly mutagenic and were effective without metabolic
activation. Some antagonistic effects were encountered when
metabolic activation was incorporated. The most active
fraction, the ether soluble bases (BE), also reverted the
putative frameshift marker, hom3-10. This fraction may con-
tain acridines and other nitrogen heterocyclics. Unpublished
results from our group have pointed to similar effectiveness
without activation in the Salmonella system when suspension
tests rather than plate assays are used with crude mixtures.
To ascertain the comparative effectiveness in the human
leukocyte chromatid aberration assay, we treated selected
test fractions from Synfuel B. The test materials used were
the neutral subfractions and represent largely polycyclic
aromatic hydrocarbons. The coal fractions were suspended in
DMSO at a concentration of 20 mg/ml total solids. Two hours
of control treatment with 5% DMSO and treatment with the
four subfractions (neutrals as hexane, hexane/benzene, ben-
zene/ether, and methanol subfractions) over a concentration
range of 0.1-1.0% (20 yg-200 ug) were ineffective in pro-
ducing chromatid aberrations (1600 cells scored). However,
metabolic activation was not included with any exogenous
enzyme source nor are the assumed constituents (PAH's) ef-
fective as chromosome breaking agents. Work with other
fractions and the inclusion of metabolism is in progress.
Selected test fractions from Synfuel B were assayed in
the Drosophila sex-linked recessive-lethal (SLRC). Fraction
13 (neutral/benzene/ether) is slightly effective as a muta-
gen for Drosophila at the higher concentrations fed.
Several other syncrude (similar crude) fractions which
were scored as mutagenic in the Salmonella assays were tested
in Drosophila. (All of the fractions require metabolic ac-
tivation in the Salmonella assay.) Of the five fractions
tested, only 12 and 13 gave any indication of an effect.
Additionally, the highly active (in Salmonella) basic sub-
fraction from the procedure previously described was tested.
This basic material showed a significant dose-dependent re-
sponse in the Drosophila SLRL assay. (See discussion by
Nix and Brewen, this proceedings, reference 22.)
-------
286 J.L. EPLER ET AL.
In conclusion, short-term tests with bacterial and fun-
gal mutagenicity assays appear to detect effectively the
mutagenic potential of complex environmental or industrial
effluents; however, chemical fractionation is necessary to
reduce toxicity and concentrate hazardous materials. Exten-
sion of the results to higher organisms, i.e., mammalian
cells, Drosophila, and the mouse appears to be valid but
needs more testing.
CONCLUSIONS
In these initial feasibility studies, the purpose has
not been to reflect on whether a relative biohazard exists
in comparison with other materials or processes. The results
show that biological testing, within the limits of the spe-
cific system used, can be carried out with complex organic
materials but perhaps only when coupled with the appropriate
analytical separation schemes. An extrapolation to relative
biohazard at this point would be, at least, premature. The
primary use that such combined chemical and biological work
may serve is to aid in isolating and identifying the speci-
fic classes or components involved. A number of precautions
are listed below.
The detection or perhaps the generation of mutagenic
activity may well be a function of the chemical fractionation
scheme used. The inability to recover specific chemical
classes or the formation of artifacts by the treatment could
well corrupt the results obtained, in addition to the possi-
bility of an inability to detect the specific biological
endpoint chosen. Along with the obvious bias that could
accompany the choice of samples and their solubility or the
time and method of storage, a number of biological discrepan-
cies can also enter into the determinations. For example,
concomitant bacterial toxicity can nullify any genetic damage
assay that might be carried out; the choice of inducer for
the liver enzymes involved might be wrong for selected com-
pounds; the choice of strain could be inappropriate for
selected compounds; and additionally, the applicability of
the generally used Salmonella test to other genetic endpoints
and the validation of the apparent correlation between muta-
genicity and carcinogenicity still remains a point of sig-
nificant fundamental research. Furthermore, the short term
assays chronically show negative results with, certain sub-
stances, e.g., heavy metals and certain classes of organics.
Similarly, compounds involved in or requiring cocarcinogenic
phenomena would presumably go undetected.
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING 287
However, as a prescreen to aid the investigators in or-
dering their priorities, the short-term testing appears to
be a valid testing approach with complex mixtures. Over-
interpretation at this stage of research especially with
respect to relative hazard or negative results should be
avoided.
REFERENCES
1. Ames BN, Lee FD, Durston WE: An improved bacterial
test system for the detection and classification of
mutagens and carcinogens. Proc Natl Acad Sci USA 71:
782-786, 1973
2. Klass DL: Synthetic crude oil from shale and coal.
Chem Technol August:499-510, 1975
3. Kier LD, Yamasaki E, Ames BN: Detection of mutagenic
activity in cigarette smoke condensates. Proc Natl
Acad Sci USA 71:4159-4163, 1974
4. Nagao M, Yahagi T, Kawachi T, Seino Y, Honda M,
Matsukura N, Sugimura T, Wakabayashi K, Tsuji K,
Kosuge T: Mutagens in foods, and especially pyrolysis
products of protein. In: Progress in Genetic Toxi-
cology (Scott D, Bridges BA, Sobels FH, eds.).
Elsevier/North-Holland Biomedical Press, 1977, pp 259-
264
5. Hardigree AA, Epler JL: Comparative mutagenesis of
plant flavonoids in microbial systems. Mutat Res, in
press
6. Ames BN, Kammen HO, Yamasaki E: Hair dyes are mutagenic:
Identification of a variety of mutagenic ingredients.
Proc Natl Acad Sci USA 72:2423-2427, 1975
7. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian-
microsome mutagenicity test. Mutat Res 31:347-364,1975
8. Chrisp CE, Fisher GL, Lammert JE: Mutagenicity of
filtrates from respirable coal fly ash. Science
199:73-75, 1978
-------
288 J.L. EPLER ET AL.
9. Epler JL, Young JA, Hardigree AA, Rao TK, Guerin MR,
Rubin IB, Ho C-h, Clark BR: Analytical and biological
analyses of test materials from the synthetic fuel
technologies. I. Mutagenicity of crude oils determined
by the Salmonella typhimurium/microsomal activation
system. Mutat Res, in press
10. Epler JL, Rao TK, Guerin MR: Evaluation of feasibility
of mutagenic testing of shale oil products and effluents.
Environ Health Perspect, in press
11. Swain AP, Cooper JE, Stedman RL: Large scale fractiona-
tion of cigarette smoke condensate for chemical and
biologic investigations. Cancer Res 29:579-583, 1969
12. Bell JH, Ireland S, Spears AW: Identification of
aromatic ketones in cigarette smoke condensate. Anal
Chem 41:310-313, 1969
13. Rubin IB, Guerin MR, Hardigree AA, Epler JL: Fractiona-
tion of synthetic crude oils from coal for biological
testing. Environ Res 12:358-365, 1976
14. Epler JL, Larimer FW, Rao TK, Nix CE, Ho T: Energy-
related pollutants in the environment: The use of
short-term tests for mutagenicity in the isolation and
identification of biohazards. Environ Health Perspect,
in press
15. Ho C-h, Clark BR, Guerin MR: Direct analysis of organic
compounds in aqueous byproducts from fossil fuel con-
version processes: Oil shale retorting, synthane coal
gasification and COED coal liquefaction. J Environ Sci
Health All (7):481-489, 1976
16. Yamasaki E, Ames BN: Concentration of mutagens form
urine by adsorption with the nonpolar resin XAD-2:
Cigarette smokers have mutagenic urine. Proc Natl Acad
Sci USA 71(8):3555-3559, 1977
17. Brown JP, Brown RJ, Roehm GW: The application of short-
term microbial mutagenicity tests in the identification
and development of nontoxic, nonadsorbable food addi-
tives. In: Progress in Genetic Toxicology (Scott D,
Bridges BA, Sobels FH, eds.). Elsevier/North-Holland
Biomedical Press, 1977, pp 185-190
-------
COMPLEX ORGANIC MIXTURES: MUTAGENICITY TESTING 289
18. Jones AR, Guerin MR, Clark BR: Preparative-scale
liquid chromatographic fractionation of crude oils
derived from coal and shale. Anal Chem 49:1766-1771,
1977
19. Guerin MR, Epler JL, Griest WH, Clark BR, Rao TK:
Polycyclic aromatic hydrocarbons from fossil fuel con-
version processes. In: Carcinogenesis, Vol. 3 (Jones
PW, Freudenthal RJ, eds.). New York, Raven Press, in
press
20. Ho C-h, Clark BR, Guerin MR, Rao TK, Epler JL: Detec-
tion by bioassay in the liquid chromatographic isolation
of mutagenic constituents of synthetic crude oils. Anal
Chem, in press
21. Epler JL, Larimer FW, Nix CE, Ho T, Rao TK: Comparative
mutagenesis of test material from the synthetic fuel
technologies. In: Progress in Genetic Toxicology
(Scott D, Bridges BA, Sobels FH, eds.). Elsevier/North-
Holland, 1977, pp 275-284
22. Nix CE, Brewen BS: The role of Drosophila in chemical
mutagenesis testing. Symposium on Application of Short-
Term Bioassays in the Fractionation and Analysis of Com-
plex Environmental Mixtures. Williamsburg, Virginia,
1978
23. Hsie AW, O'Neill JP, Sebastian JRS, Couch DB, Brimer PA,
Sun WNC, Fuscoe JC, Forbes NL, Machanoff R, Riddle JC,
Hsie MH: Mutagenicity of carcinogens: Study of 101
individual agents and 3 subfractions of a crude syn-
thetic oil in a quantitative mammalian cell gene muta-
tion system. Symposium on Application of Short-Term
Bioassays in the Fractionation and Analysis of Complex
Environmental Mixtures. Williamsburg, Virginia, 1978
24. Mohn GR, Ellenberger J: The use of Escherichia coli
K12/343/113 (X) as a multipurpose indicator strain in
various mutagenicity testing procedures. Mutat Res,
in press
25. Larimer FW, Ramey DW, Lijinsky W, Epler JL: Mutageni-
city of methylated N-nitrosopiperidines in Saccharomyces
cerevisiae. Mutat Res, in press
-------
QUANTITATIVE MAMMALIAN
CELL GENETIC TOXICOLOGY:
STUDY OF THE CYTOTOXICITY
AND MUTAGENICITY OF
SEVENTY INDIVIDUAL
ENVIRONMENTAL AGENTS
RELATED TO ENERGY
TECHNOLOGIES AND THREE
SUBFRACTIONS OF A CRUDE
SYNTHETIC OIL IN THE
CHO/HGPRT SYSTEM
Abraham W. Hsie, J. Patrick O'Neill,
Juan R. San Sebastian, David B. Couch,
Patricia A. Brimer, William N.C. Sun,
James C. Fuscoe, Nancy L. Forbes,
Richard Machanoff, James C. Riddle,
and Mayphoon H. Hsie
Biology Division
Oak Ridge National Laboratory and
University of Tennessee-Oak Ridge
Graduate School of Biomedical Sciences
Oak Ridge, Tennessee
-------
293
As science and technology advance, an extraordinary
quantity of natural and synthetic chemicals is introduced
continuously into our environment. Through conventional
animal tests, some of these environmental chemicals have been
found to be either highly toxic, mutagenic, carcinogenic, or
teratogenic. Epidemiological studies have shown .that among
these harmful chemicals, a few also exhibit such detrimental
effects in the human population. Because of the high cost
and length of time required for the animal experiments, such
tests have been confined to only a very small fraction of
these environmental agents. Thus, the biological effects of
the great majority of these chemicals, including ingredients
of our daily foods and drugs, remain either incompletely
tested or unknown.
During the past few years, evidence has accumulated
that a high percentage (80-90%) of human cancer is linked
to exposure to industrial and environmental chemicals iden-
tifiable as carcinogens (23,44). Since the expense of ani-
mal tests preclude their routine use to identify environ-
mental carcinogens, many short-term assays have been devel-
oped as initial carcinogen screening tests. Studies of
mutagenesis and DNA-repair in microorganisms, especially
Salmonella typhimurium and Escherichia coli, have estab-
lished that approximately 90% of chemical carcinogens cause
mutation induction or DNA damage in these bacteria (2,3,26,
27,38,39,42,45,46). Such findings imply that the microbial
tests are useful to identify not only potential mutagens but
also carcinogens in the environment.
-------
294 ABRAHAM W. HSIE ET AL.
In view of the intrinsic limitation of the microbial
assay to respond to certain classes of chemicals, such as the
apparent failure of the Salmonella assay to demonstrate that
carcinogenic halogenated hydrocarbons and metallic compounds
are mutagenic (27), it appears that no single test system
will give 100% correlation between mutagenicity and carcino-
genicity. The use of a battery of tests rather than any
single test in isolation has thus been proposed to reduce
the probability of false negatives (i.e., known carcinogens
are not mutagenic) and false positives (i.e., known noncar-
cinogens are mutagenic) (4,37).
It has been recognized that studies of mutagenesis in
prokaryotes may not reveal some fundamental mechanisms of
mutagenesis in mammals, because mammals differ from prokar-
yotes in their level of organization and repair of DNA,
mechanisms of metabolism of chemicals, and other related
functions. Some bacterial mutagens such as caffeine and
hydroxylamine do not appear to be mutagenic in mammalian
cells, while agents such as nickel and beryllium compounds
are mutagenic in mammalian cells but not in the Salmonella
system (Couch, San Sebastian, and Hsie, unpublished, 27).
In addition, it is well known that chromosomal abnormality
is a major cause of inheritable human diseases and is often
associated with the process of malignancy. The great major-
ity of chemical carcinogens are known to induce chromosomal
aberrations (1,24) or sister-chromatid exchange (1). Dieth-
ylstilbestrol, a synthetic hormone associated with cancer in
women, causes chromosomal aberrations in cultured mammalian
cells (24), but does not cause mutation induction in Sal-
monella (27). Clearly, mammalian cell systems offer advan-
tages over bacterial systems for studying genetic toxicity
at the chromosome and chromatid level.
Since the observation that treatment of mammalian somat-
ic cells with conventional mutagens such as ethyl methane-
sulfonate (EMS) and N-methyl-N'-nitro-N-nitrosoguanidine
(MNNG) causes an increase in the number of cell variants that
differ from parental cells in either nutritional requirement
(5,36) or drug sensitivity (5), there has been much interest
in utilizing a quantitative mammalian cell mutation system
for studying mechanisms underlying the process of mammalian
mutation and, additionally, for assessing the genetic hazard
of environmental agents to the human population. Several
mammalian cell mutation systems, especially those utilizing
resistance to purine analogues such as 8-azaguanine (AG) and
6-thioguanine (TG) as a genetic marker (6), have been devel-
oped for such purposes. The selection for mutation induction
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 295
to purine analogue resistance is based on the fact that the
wild-type cells containing hypoxanthine-guanine phosphoribosyl
transferase (HGPRT) activity are capable of converting the
analogue to toxic metabolites, leading to cell death; the pre-
sumptive mutants, by virtue of the loss of HGPRT activity,
are incapable of catalyzing this detrimental metabolism and,
hence, escape the lethal effect of the purine analogue (Table
1).
Table 1
CHO/HGPRT Mutation Assay1
(1) Enzyme system:
(or TG, AG) (or TG , AG)MP
(2) Mutation induction and selection for variants and
revertants:
or chemical agents)
genotype HGPRT+ HGPRT"
phenotype TGS, TGr,
aminopterin positive aminopterin
negative
(b) Variant selection is based on resistance to TG
(c) Selection of revertants is based on growth in the
presence of aminopterin.
(3) Characterization of TGr variants:
(a) Direct enzyme assay for conversion of [ 3H]hypox-
anthine to [ 3H]IMP.
(b) Cellular incorporation of [ 3H]hypoxanthine into
cellular macromolecules as revealed by either direct
radioactivity measurement or autoradiographic deter-
mination .
(c) Sensitivity of clonal growth to aminopterin (10 yM)
in medium F12FCM5 which contains hypoxanthine (30
yM) , glycine (100 yM) , and thymidine (3 yM) .
'Table II of ref. 21.
-------
296 ABRAHAM W. HSIE ET AL.
The near-diploid Chinese hamster ovary (CHO) cell line
has been chosen for our study because a mutation assay,
referred to as the CHO/HGPRT system, has been well defined
(7-10,16-22,29-33). We have used CHO cells because these are
perhaps the best characterized mammalian cells genetically
(35,40). They exhibit high cloning efficiency, achieving
nearly 100% under normal growth conditions, and are capable
of growing in a relatively well-defined medium on a glass or
plastic substratum or in suspension with a population doub-
ling time of 12-13 hr. In addition, the cells have a stable,
easily recognizable karyotype of 20 or 21 chromosomes (de-
pending on the subclone) (11) and are suitable for studying
mutagen- or carcinogen-induced chromosome and chromatid aber-
rations (1) and sister-chromatid exchanges (1,24) (Table 2).
METHODS AND MATERIALS
Cell Culture
All studies to be described have employed a subclone of
CHO-Kj cells (25), designated as CHO-K,-BH% (16). It was
isolated following selection in F12 medium containing aminop-
terin (10 uM) (16). Cells are routinely cultured in Ham's
F12 medium (Pacific Biological Co.) containing 5% heat-
inactivated (56°C, 30 min), extensively dialyzed fetal calf
serum (Pacific Biological Co.) (medium F12FCM5) in plastic
tissue culture dishes (Falcon or Corning Glass Works) under
standard conditions of 5% C02 in air at 37°C in a 100% humid-
ified incubator. These cells grow in medium which contains
aminopterin as well as in regular medium with 5 or 10% dia-
lyzed fetal calf serum with a population doubling time of
12-13 hr. Cells are removed with 0.05% trypsin for subcul-
ture, and the number is determined with a Coulter counter
(model B, Coulter Electronics).
Treatment with Chemicals
We have standardized treatment procedures which are found
to be suitable for various chemicals (16,29). Briefly, CHO
cells are plated at 5 x 10s cells/25 cm2 bottle in medium F12-
FCM5. After a 16- to 24-h growth period (cell number = 1.0-
1.5 x 10s cells/plate), the cells are washed once with saline
G, and sufficient serum-free F12 medium is added to bring the
final volume to 5 ml after the addition of various amounts of
microsome preparation (up to 1 ml) and 50 yl of chemical,
usually dissolved in dimethyl sulfoxide. Chemicals and/or
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 297
Table 2
Characteristics of CHO cells1
(1) Exhibit a stable karyotype over 20 years with a modal
chromosome number of 20 which has a distinctly recogniz-
able morphology.
(2) Have a colony-forming capacity of nearly 100% in a
defined growth medium.
(3) Grow well in either monolayer or suspension with a. rela-
tively short population doubling time of 12-14 hr.
(4) Are genetically and biochemically well characterized,
with many genetic markers available, including auxotro-
phy, drug resistance, temperature sensitivity, etc.
(5) Respond well to various synchronization methods, includ-
ing the mitotic detachment procedure, which facilitate
cell cycle study.
(6) Are useful in somatic cell hybridization experiments be-
cause they readily hybridize with different cell types,
including human cells; when the CHO-human cell hybrid is
formed there is subsequent rapid, preferential loss of
human chromosome, which facilitates the assignment of
marker genes to specific chromosomes or linkage groups
in the human karyotype.
(7) Respond quantitatively to various physical and chemical
mutagens and carcinogens with high sensitivity.
(8) Adapt to mutation induction either through coupling with
a microsome activation system or through host (mouse)
mediation.
(9) Are capable of monitoring induced mutation to multiple
gene markers, chromosome aberration, and sister chroma-
tid exchange in the same mutagen-treated cell culture.
1 Table I of ref. 21.
-------
298 ABRAHAM W. HSIE ET AL.
microsomes are omitted from some plates to provide controls.
The microsomal preparation has been prepared in this labora-
tory according to the method of Ames et al. (3) from livers
of Aroclor 1254-induced male Sprague-Dawley rats; the micro-
some mix for biotransformation contains (per ml) 33 umoles
KC1, 8 ymoles MgClz, 4 ymoles NADP, 5 umoles glucose-6-phos-
phate, 100 nmoles phosphate buffer (pH 7.4), and 0.2 ml micro-
some fraction. Cells are then incubated for 5 h and washed
3 times with saline G before 5 ml of F12FCM5 are added. Fol-
lowing overnight incubation, cells are trypsinized and plated
for cytotoxicity and specific gene mutagenesis to be described
below. Treatment with physical agents has been described in
detail elsewhere (17,19,29,30).
Cytotoxicity
The effect of chemicals on the cellular cloning effi-
ciency is determined by use of the treated cells described
above. For an expected cloning efficiency higher than 50%,
200 well-dispersed single cells are plated, and for an expec-
ted survival lower than this, the number of cells plated is
adjusted accordingly to yield 100-200 surviving colonies
after standard incubation in medium F12FCM5 for 7 days. At
the end of the incubation period, the plates are fixed with
3.7% formalin and stained with a dilute crystal violet solu-
tion before the colonies are enumerated. A cluster of more
than 50 cells growing within a confined area is considered
to be a colony. Control cells, which do not receive treat-
ment with mutagen, usually give 80% or higher plating effi-
ciency under this condition. Neither the solvent-microsome
mix nor these agents individually affect the cellular cloning
efficiency. The effect of carcinogen on the cloning effi-
ciency is expressed as percent survival relative to the
untreated controls.
Specific Gene Mutagenesis
The CHO/HGPRT system has been defined in terms of medium,
TG concentration, optimal cell density for selection (and,
hence, recovery of the presumptive mutants), and expression
time for the mutant phenotype (16,29). For the determination
of mutation induction, the treated cells are allowed to ex-
press the "mutant phenotype" in F12 medium for 7-9 days, at
which time mutation induction reaches a maximum which is
maintained thereafter (as long as 35 days examined) for
several agents (EMS, MNNG, ICR-191, X ray, and UV) irrespec-
tive of concentration or intensity of the mutagen (29-32).
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 299
Routine subculture is performed at 2-day intervals during
the expression period, and at the end of this time the cells
are plated for selection in hypoxanthine-free F12FCM5 con-
taining 1.7 ug/ml (10 uM) of TG at a density of 2.0 x 10s
cells/100 mm plastic dish (Corning or Falcon), which permits
100% mutant recovery in reconstruction experiments (29). We
find the use of dialyzed serum particularly important, pre-
sumably due to potential competition between hypoxanthine
and TG for transport into the cells and for catalysis by
HGPRT (29). After 7 to 8 days in the selective medium, the
drug-resistant colonies develop; they are then fixed, stained,
and counted. Such a protocol permits the maximum yield by
various physical and chemical agents of TG-resistant variants,
>98% of which have highly reduced HGPRT activity (7-10,16-22,
29-33). Mutation frequency is calculated based on the number
of drug-resistant colonies per survivor at the end of the
expression period.
RESULTS
Characteristics of the CHO/HGPRT System: Evidence of the
Genetic Basis of Mutation at a Specific Locus
Conclusive, direct proof of the genetic origin of muta-
tions in somatic cells should theoretically rely on demonstra-
tion that the affected hereditary alteration has resulted in
a modified nucleotide sequence of the specific gene, causing
modified coding properties which result in the production of
altered protein with changes in the amino acid sequence. In
the absence of such proof, one must rely on indirect criteria
which are consistent with the concept that the observed pheno-
typic variations are genetic in nature. Such criteria include
stability of altered phenotype, mutagen-induced increase in
occurrence of stable variants, biochemical and physiological
identification of the variant phenotype, chromosomal locali-
zation of the affected gene, etc. (6,35,40,43).
Over the past four years, we have used the assay proto-
col described (16,29) and have found in approximately 400
experiments that the spontaneous mutation frequency lies in
the range of 1-5 x 10 ~6 mutant/cell. Various physical and
chemical agents are capable of inducing TG resistance.
Among all chemical mutagens examined, mutation induction
occurs as a linear function of the concentration (7-10,16-22,
29-33). For example, mutation frequency increases approxi-
mately linearly with EMS concentration in this near-diploid
-------
300 , ABRAHAM W. HSIE ET AL.
cell line, conforming to the expectation that mutation induc-
tion occurs in the gene localized at the functionally mono-
somic X chromosome. However, in the tetraploid CHO cells,
EMS does not induce an appreciable number of mutations, even
at very high concentrations, as predicted theoretically (18).
We have been unable to detect any spontaneous reversion
with 13 TG-resistant mutants, all of which contain low, yet
detectable, HGPRT activity. More than 98% of the presumptive
mutants isolated either from spontaneous mutation or as a
result of mutation induction are sensitive to aminopterin,
incorporate hypoxanthine at reduced rates, and have less than
5% HGPRT activity (29). Studies in progress have also shown
that mutants containing temperature-sensitive HGPRT can be
selected, suggesting that mutation resides in the HGPRT
structural gene (O'Neill and Hsie, unpublished observations).
The CHO/HGPRT system appears to fulfill the criteria for
a specific gene locus mutational assay (Table 3) and should
be valuable in studying mechanisms of mammalian cell muta-
genesis and as a system to determine the mutagenicity of
various physical and chemical agents.
Table 3
CHO/HGPRT Mutation Assay: Genetic Basis of Mutation at
HGPRT Locus in TG-Resistance Selection1
(1) Spontaneous mutation frequency at 1-5 x 10~6 mutant/cell.
(2) Mutation induction by physical and chemical agents with
linear dose-response relationship.
(3) Frequency of spontaneous reversion at less than 10~7
reversion/cell.
(4) Failure to induce mutation in near-tetraploid cell lines,
(5) Altered HGPRT activity in mutants.
(a) 1179/1189 (98.4%) mutant colonies are aminopterin
negative.
(b) 121/122 (99.2%) mutant colonies show reduced hypox-
anthine incorporation by autoradiography studies.
(c) 81/83 (97.6%) isolated mutant clones show reduced
HGPRT enzyme activity.
'Table III of ref. 21.
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 301
Mutagenicity of 70 Individual Energy-Technology-Related
Environmental Agents
Polycyclic Hydrocarbons (Total of 27). Some of the most
ubiquitous environmental organic pollutants in our environ-
ment are polycyclic hydrocarbons, many of which are carcino-
genic. Coal- and synthetic-fuel-related energy technologies
and gasoline-driven engines often generate high levels of
polycyclic hydrocarbons which are detectable in urban air and
water. We have studied the mutagenicity of benzo(a)pyrene
[B(a_)P] and its 19 metabolites, including 11 phenols, 3
epoxides, 3 diols, and 2 diolepoxides. For comparison,
benzo(£)pyrene [B(e_)P] and pyrene were added to this study.
Also included were benz(a)anthracene (BA) and 4 related com-
pounds (7,12-dimethyl BA, anthracene, and two phenolic deriv-
atives of BA). The carcinogenic polycyclic hydrocarbons B(a_)P,
BA, and 7,12-dimethyl BA require metabolic activation to be
mutagenic. The weak carcinogen (B(e?)P is less mutagenic than
B(&")P. The noncarcinogenic polycyclic hydrocarbons, pyrene
and anthracene, are nonmutagenic even with metabolic activa-
tion. B(a_)P-4,5-epoxide and B(a)P-7,8-diol,9-10-epoxide(syn)
are mutagenic. Since CHO cells cannot activate procarcinogens
such as B(a_)P, these cells appear to be most useful in screen-
ing for the mutagenicity of metabolites such as those of
B(a^)P (Hsie and Brimer, unpublished). Because of the limited
availability of B(a)P derivatives, some of the experiments
remain to be pursued in detail.
Metallic Compounds (Total of 15). The carcinogenic and muta-
genic potential of certain toxic metallic compounds has become
an environmental concern, especially with the increasing
large-scale coal mining and coal firing of power plants. We
found that MnCl2-4H?0, FeSOH-7H20, CoCl2-6H20 and cis-
Pt(NH3)2Cl2 (an antitumor agent) are mutagenic, while NiCl2-
6H20, BeSO,,-4H20, and CdCl2 are weakly mutagenic. Determina-
tion of metal mutagenicity is apparently complicated by the
ionic composition of the medium. For example, we found that
the mutagenicity and cytotoxicity of MnCl2 were abolished by
the excess of MgCl2. The unusual environment required for
demonstration of mutagenicity of MnCl2 makes assessment of
its biological hazard difficult. This too may account in
part for varying results obtained in studying the mutagenicity
of AgN03, CaCl2, Pb(CH,COO)2- 3H20, RbCl, H2Se03, TiCl „, and
ZnSO,,'7H20 (10; Couch, San Sebastian, Forbes, and Hsie,
unpublished).
-------
302 ABRAHAM W. HSIE ET AL.
Nitrosamines and Related Compounds (Total of 16). Nitro-
samines are potent carcinogens for various animal species.
They are of environmental concern because it is known that
oxides of nitrogen produced at high temperature in internal
combustion engines and in coal-fired power plants can react
with atmospheric water to form nitrosamines. Nitrosamines
can also be formed in the human stomach by a reaction between
a common meat preservative, sodium nitrite, and various sec-
ondary and tertiary amines, many of which are often used as
counter or prescription drugs.
Nitrosamines generally require metabolic activation to
be cytotoxic and/or mutagenic. In addition to investigating
two common aliphatic nitrosamines, dimethylnitrosamine (DMN)
and diethylnitrosamine (DEN), we have studied the mutagenicity
of 11 cyclic nitrosamines, including 3 nitrosopiperidines, 3
nitrosopyrrolidines, 3 nitrosopiperazines, and 2 nitrosomor-
pholines. Our studies also include the nitrosamine-related
chemicals dimethylamine, formaldehyde, and sodium nitrite.
We have found that all 9 carcinogenic nitrosamines (DMN, DEN,
2-methyl-l-nitrosopiperidine, 3,4-dichloro-l-nitrosopiperi-
dine, nitrosopyrrolidine, 3,4-dichloronitrosopyrrolidine,
1,4-dinitrosopiperazine, 1,5-dinitrosohomopiperazine, nitro-
somorpholine) are mutagenic and all 4 noncarcinogenic nitro-
samines (2,5-dimethylnitrosopiperidine, 2,5-dimethylnitro-
sopyrrolidine, 1-rnitrosopiperazine, nitrosophenmetrazine)
are nonmutagenic. Formaldehyde and sodium nitrite are non-
mutagenic, and dimethylamine is mutagenic at high concentra-
tions (San Sebastian, Couch, and Hsie, unpublished). Varia-
ble carcinogenicity data on the latter three chemicals exists
in the literature.
Quinoline Compounds (Total of 5). One class of potential
environmental contaminants from fossil-fuel energy is hetero-
cyclic compounds such as quinolines. Quinoline, a known
carcinogen, is" mutagenic with metabolic activation. Another
carcinogen, 4-nitroquinoline-l-oxide, is highly mutagenic;
its mutagenicity decreases when assayed in the presence of
the activation system. The carcinogenicity of 8-hydroxy-,
8-amino-, and 8-nitroquinoline is riot known, but these com-
pounds exhibit variably weak mutagenicity in preliminary
experiments (San Sebastian and Hsie, unpublished).
Physical Agents (Total of 7). The mutagenicity of both ioniz-
ing radiation such as X ray and nonionizing physical agents
such as UV light has been demonstrated. Fluorescent white,
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 303
black, and blue lights are slightly cytotoxic and mutagenic.
Sunlamp light is highly cytotoxic and mutagenic, exhibiting
the biological effects within 15 sec of exposure under condi-
tions recommended by the manufacturer for human use. Cyto-
toxic and mutagenic effects are observed after five min of
sunlight exposure; responses vary with hourly and daily vari-
ations in solar radiation. In view of man's constant expos-
ure to various light sources, demonstration of their genetic
toxicity suggests that daily exposure to these light sources,
especially sunlight, should be minimized (17,19,30). The
demonstration that the CHO/HGPRT system is capable of quanti-
fying the cytotoxic and mutagenic effect of sunlight recom-
mends it as a model mammalian cell system for studies of the
genetic toxicology of sunlight per se and of the interactive
effects between sunlight and other physical and chemical
agents, leading ultimately to a better understanding of the
effects of sunlight on humans and the environment.
Mutagenicity of 39 Other Chemicals
Direct-Acting Alkylating Agents and Related Compounds (Total
of 11).Included are 10 alkylating agents {2 alkyl sulfates
[dimethyl sulfate (DMS), diethyl sulfate (DBS)], 3 alkyl
alkanesulfonates [methyl methanesulfonate (MMS), EMS, and
isopropyl methanesulfonate (iPMS)], 2 nitrosamidines [MNNG
and N-ethyl-N'-nitrosoguanidine (ENNG)], 3 nitrosamides [N-
methyl-N-nitrosourea (MNU), N-ethyl-N-nitrosourea (ENU), and
N-butyl-N-nitrosourea (BNU)]J and a structural analogue of
MNNG, N-methyl-N'-nitroguanidine (MNG). Among the alkyl
sulfates and alkanesulfonates, cytotoxicity was found to
decrease with the size of the alkyl group: DMS>DES; MMS>EMS
>iPMS. The mutagenicity based on mutants induced per unit
mutagen concentration was DMS>DES; MMS>EMS>iPMS. However,
when comparisons were made at 10% survival, mutagenic potency
was: DES>DMS; EMS>MMS>iPMS. Among the nitroso compounds,
the order of the mutagenicity based on 10% survival was
MNNG>ENNG>MNU>ENU>BNU. This is the same order of potency as
observed for mutation induction per unit concentration of
mutagen. MNG is not mutagenic (7-9; Couch, San Sebastian,
and Hsie, unpublished).
Heterocyclic Nitrogen Mustards—ICR Compounds (Total of 10).
A series of heterocyclic nitrogen half-mustards, the ICR-
compounds, has been developed at the Institute for Cancer
Research as antitumor agents. Apparently, the biological
activities of these compounds are associated with their
-------
304 ABRAHAM W. HSIE ET AL.
ability to intercalate and covalently bind nucleic acid.
Ten ICR-compounds (ICR-191, -170, -292, -372, -340, -191-OH,
-170-OH, -292-OH, -372-OH, and -340-OH) have been studied.
The 2-chloroethyl side chain of the first 5 compounds (e.g.,
ICR-191, etc.) has been replaced by a hydroxy group in the
latter 5 (e.g., ICR-191-OH, etc.). The 10 compounds differ
in the heterocyclic nucleus (methoxyacridine for ICR-191 and
-170, benz(a)acridine for -292, and azaacridine for -372 and
-340) and the alkylating side chain (the same secondary amine
for ICR-191 and -372, and the same tertiary amine for -170,
-292, and -340). Those with 2-chloroethyl side chains are
highly mutagenic, with the tertiary amines 3 to 5 times more
mutagenic than the secondary amines. The 5 hydroxy deriva-
tives are nonmutagenic, but remain highly toxic, indicating
that although the 2-chloroethyl group (nitrogen half-mustard)
is needed for mutagenicity, its replacement with a hydroxy
group does not alter cytotoxicity. Cytotoxicity and muta-
genicity of ICR-compounds appear to be dissociable (32,33;
Fuscoe, O'Neill, and Hsie, unpublished).
Aromatic Amines (Total 5). Many aromatic amines are human
carcinogens. We have shown that the carcinogens 2-acetylami-
nofluorene and its N-hydroxy- and N-acetoxyl derivatives are
mutagenic, while fluorene, a noncarcinogenic analogue, is
nonmutagenic. l-hydroxy-2-acetylaminofluorene appears to be
mutagenic at a very high concentration in one preliminary
experiment (Hsie, Sun, and Brimer, unpublished).
Miscellaneous Compounds (Total of 13). Three commonly used
organic solvents (acetone, dimethyl sulfoxide, and ethanol)
are noncarcinogenic and do not appear to be mutagenic. All
four metabolic inhibitors (cytosine arabinoside, hydroxyurea,
caffeine, and cycloheximide) are nonmutagenic in a preliminary
study without coupling with the metabolic activation system.
Hydrazine and hycanthone appear to be direct-acting mutagens.
N6,02 -dibutyryl adenosine 3':5'-phosphate, an analogue of
adenosine 3' :5'-phosphate and an important effector of growth
and differentiation in many biological systems, is not muta-
genic. The pesticides captan and folpet are mutagenic. The
mutagenicity of an artificial sweetener, saccharin, appears to
be variable; its determination is complicated by the require-
ment of high concentrations to yield any biological effect
(O'Neill and Hsie, unpublished).
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 305
Correlation of Mutagenicity in the CHO/HGPRT Assay with
Reported Carcinogenicity in Animal Tests
Among a total of 109 chemical and physical agents stud-
ied, at different stages of completion, 56 have been reported
to be either carcinogenic or noncarcinogenic in animal stud-
ies. Mutagenicity in the CHO/HGPRT assay of 54 of these
agents correlated with documented animal carcinogenicity.
The concurrence (i.e., known carcinogens are mutagenic and
noncarcinogens are nonmutagenic in CHO/HGPRT assays) of each
class of agents so far tested is 100% except for nitrosamines
and relatives (93%) and ICR compounds (83%) (Table 4). The
existence of a high correlation [54/56 (96.43%)] between
mutagenicity and carcinogenicity speaks favorably for the
utility of this assay in prescreening the carcinogenicity of
chemical and physical agents. However, this result should
be viewed with caution, since so far only limited classes of
chemicals have been tested and some of the preliminary results
remained to be confirmed.
A possible false negative was formaldehyde, which has
been shown to be either carcinogenic or noncarcinogenic
depending on the way test animals are exposed to it. An
apparent false positive was ICR-191, a potent mutagen for
microorganisms and CHO and other mammalian cells, which has
been shown to be noncarcinogenic in a recent study.
A Study of EMS Exposure Dose: Differential Effects on
Cellular Lethality and Mutagenesis
Earlier, we found that EMS-induced mutation frequency to
TG resistance in cells treated for a fixed period of 16 h is
a linear function over a large range of mutagen concentrations
(0.013-0.8 mg/ml), including both the shoulder region (0-0.1
mg/ml) and the exponentially killing portion (0.1-0.8 mg/ml).
To investigate whether EMS-induced mutagenesis can be quanti-
fied further, cells were treated with several concentrations
of EMS for intervals of 2-24 h. Mutation induction increased
linearly with EMS concentrations of 0.05-0.4 mg/ml for incuba-
tion times of up to 12-14 h. However, cell survival decreased
exponentially with time over the entire 24-hour period. This
difference in the time course of cellular lethality vs. muta-
genicity might be due to the formation of toxic, nonmutagenic
breakdown products in the medium with longer incubation times,
or it might reflect a difference in the mode of action of EMS
in these two biological effects. Further studies using vary-
ing concentrations (0.05-3.2 mg/ml) of EMS for 2-12 h showed
-------
306
ABRAHAM W. HSIE ET AL.
Table 4
Correlation of Mutagenicity1 in the CHO/HGPRT Assay with
Reported Carcinogenicity2 in Animal Tests3
Total No.
Agent* Studied
Energy- technology-
related substances
Polycyclic
hydrocarbons
Metallic compounds
Nitrosamines and
relatives
Quinolines
Physical agents
Subtotal
Other chemicals
Direct-acting
alkylating agents
and relatives
ICR compounds
Aromatic amines
Miscellaneous
compounds
Subtotal
27
15
16
5
7
70
11
10
5
13
39
Concurrence*
6/6 (100%)
4/4 (100%)
14/15 (93.33%)
2/2 (100%)
3/3 (100%)
29/30 (96.67%)
11/11(100%)
5/6 (83.33%)
4/4 (100%)
5/5 (100%)
25/26 (96.15%)
False
Negatives6
0
0
1/15 (6.57%)
0
0
1/30 (3.33%)
0
0
0
0
0
False
Positives7
0
0
0
0
0
0
0
1/6 (16.67%)
0
0
1/26 (3.85%)
All agents
109
54/56 (96.43%) 1/56 (1.79%) 1/56 (1.79%)
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 307
Table 4 (continued)
'Agents studied are found to be either mutagenic
regardless of "mutagenic potency") or nonmutagenic. The
mutagenicity is assayed either directly or coupled with a
metabolic activation system in vitro or in vivo. In the S-9
coupled assay the microsome used was prepared from livers of
Aroclor 1254-induced male Sprague-Dawley rats. The effects
of other inducers or of conditions of the activation system
have not been investigated extensively and are under study.
2Agents studied are denoted as either carcinogenic, non-
carcinogenic, or uncertain based primarily on published data
from USPHS (44) and IARC (23), regardless of "carcinogenic
potency." Carcinogenicity data on many compounds is not yet
available. The search for such data is admittedly neither
exhaustive nor updated.
3In part from Table VII of ref. 21.
"The data are compiled from all agents studied, exclud-
ing those whose carcinogenicity is either unknown or uncer-
tain. Thus, only 56 out of 109 agents studied are compiled
table.
'Known carcinogens are mutagenic in CHO/HGPRT assays,
e.g., MNNG, ICR-292, Ni, B(a)P, hycanthone, UV.
sKnown carcinogens are nonmutagenic in CHO/HGPRT assays,
e.g., formaldehyde.
7Known noncarcinogens are mutagenic in CHO/HGPRT assays,
e.g., ICR-191.
that the manifestation of cellular lethality and mutagenesis
occurs as a function of EMS exposure dose in that the bio-
logical effect is the same for different combinations of
concentration multiplied by duration of treatment which yield
the same product. From these studies the mutagenic potential
of_EMS can be described as 310 x 10~* mutants (cell mg ml~l
h)"1. Thus, the CHO/HGPRT system appears to be suitable for
dosimetry studies which are essential for our understanding
of the molecular mechanisms involved in mammalian mutagenesis
(31).
-------
308 ABRAHAM W. HSIE ET AL.
Screening for the Mutagenicity of Fractionated Synthetic Fuel
In addition to studying the mutagenicity of individual
environmental agents such as polycyclic hydrocarbons, quino-
lines, nitrosamines, metallic compounds, etc., we have found
that the CHO/HGPRT assay can detect the cytotoxicity and muta-
genicity of a crude organic mixture, in this case three sub-
fractions of a crude synthetic oil (fractionated by M.R.
Guerin of the Analytical Chemistry Division, ORNL) supplied
by the Pittsburgh Energy Research Center. The acetone efflu-
ent (which contains tentatively identifiable heterocyclic
nitrogen compounds) derived from the basic fraction is most
mutagenic in the presence of a metabolic activation system
(Table 5) (Hsie and Brimer, unpublished). Earlier, it
appeared that the extreme toxicity of the unfractionated
crude fuel prevented meaningful mutagenicity studies in the
CHO/HGPRT system (Hsie and Brimer, unpublished). The chemi-
stry (14), mutagenicity in microbial systems (12), and
environmental testing (13) of the Synfuel are presented
elsewhere in this proceedings.
Preliminary Development and Validation of the CHO Genetic
Toxicity Assay for.the Simultaneous Determination of Cyto-
toxicity, Mutagenicity, Chromosome Aberrations, and Sister
Chromatid Exchanges
We have so far shown that CHO cells are useful for
studying the cytotoxicity and mutagenicity of various indi-
vidual physical and chemical agents and a crude organic mix-
ture. The CHO cells and other hamster cells in culture were
also found to be suitable for studying carcinogen-induced
chromosome and chromatid aberrations (1,15,14,18) and sister
chromatid exchanges (1,34,41). In our preliminary studies,
we have found that these assays are useful in evaluating the
cytogenetic effects of B(a_)P and DMN when CHO cells are
coupled with the standard microsome preparation described
earlier (San Sebastian and Hsie, unpublished).
The successful development and validation of the multi-
plex CHO cell genetic toxicity system will be extremely valu-
able from both the scientific and economic points of view in
genetic toxicology, because this system will allow the simul-
taneous determination of four distinct biological effects:
cytotoxicity or cloning efficiency measures the reproductive
capacity of a single cell to develop into a colony; single
gene mutagenesis involves changes in the nucleotide sequence
of DNA of a specific gene resulting in the acquisition of a
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY
309
Table 5
Cytotoxicity and Mutagenicity of Subfractions1
of Synfuel A Basic Fractions
Subf raction2
Benzene
Isopropanol
Acetone
Controls
EMS
B(a)P'
Solvent
Concentration
(ug/ml)
0.25
1
2.5
10
25
50
100
0.25
1
2.5
10
25
50
100
0.25
1
2.5
5
10
25
50
100
200
8
Relative
Cloning
Efficiency (%)
Without With
S-9 S-9
92
102
117
109 91
90
71
<0.2 0.2
95
94
103
108 95
102
82
<0.2 58
89
101
93
100
58 96
22 56
<0.3 4
<0.2 0.2
100 100
Observed
Mutation
Frequency
(TG mutants/
106 cells)
Without With
S-9 S-9
4
<1
1
<1 1
16
13
25
1
1
6
<1 4
16
7
2
6
5
<1
9
13 22
6 46
15 49
135
279
557
4 9
'Unpublished data of Hsie and Brimer.
2See ref. 14 for details about the chemical separation of
Synfuel A.
-------
310 ABRAHAM W. HSIE ET AL.
novel or altered phenotype; chromosome aberrations involve
microscopically identifiable changes in the number and/or
structure of the chromosome; and sister chromatid exchange
measures the extent of double-strand exchange in the DNA
duplex after breaks and rejoining of subunits of chromatids
each of which consists of one DNA duplex.
SUMMARY AND CONCLUSIONS
Conditions necessary for quantifying mutation induction
to TG resistance, which selects for >98% mutants deficient in
the activity of HGPRT in a near-diploid CHO cell line, have
been defined. Employing this mutation assay, we have deter-
mined the mutagenicity of diversified agents, including 11
direct-acting alkylating agents, 16 nitrosamines, 10 hetero-
cyclic nitrogen mustards, 15 metallic compounds, 5 quinolines,
5 aromatic amines, 27 polycyclic hydrocarbons, 12 miscella-
neous compounds, and 7 ionizing and nonionizing physical
agents. The direct-acting carcinogen MNNG is mutagenic,
while its noncarcinogenic analogue N-methyl-N'-nitroguanidine
is not. Coupled with the rat liver S-9 activation system,
procarcinogens such as nitrosopyrrolidine, B(a)P and 2-
acetylaminofluorene are mutagenic while their analogues 2,5-
dimethylnitrosopyrrolidine, pyrene, and fluorene are not.
The mutagenicity of the 56 agents documented to be either
carcinogenic or noncarcinogenic correlated well [54/56
(96.43%)] with the reported animal carcinogenicity. A pos-
sible false negative was formaldehyde and a false positive
was ICR-191. Preliminary studies on a synthetic crude oil
show that the acetone effluent (tentatively identifiable as
heterocyclic nitrogen compounds) derived from the basic frac-
tion of Synfuel A is the most mutagenic fraction. Thus the
assay appears to be applicable for monitoring the genetic
toxicity of crude organic mixtures in addition to diverse
individual chemical and physical agents. The quantitative
nature of the assay enables a study of EMS exposure dose:
the mutagenic potential of EMS can be described as 310 x 10~6
mutants (cell mg ml"1 h)-1. It is also feasible to expand
the CHO/HGPRT system for quantifying cytotoxicity and muta-
genicity to determination of chromosomal aberrations and
sister-chromatid exchanges in cells treated under identical
conditions. Thus it is possible to study simultaneously
these four distinctive biological effects.
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 311
REFERENCES
1. Abe S, Sasaki M: Chromosome aberrations and sister
chromatid exchanges in Chinese hamster cells exposed to
various chemicals. J Natl Cancer Inst 58:1635-1641,
1977
2. Abrahamson S, Lewis EB: The detection of mutations in
Drosophila melanogaster. In: Chemical Mutagens: Prin-
ciples and Methods for Their Detection (Hollaender A,
ed.)- New York, Plenum Press, 1973, pp 461-488
3. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian-
microsome mutagenicity test. Mutat Res 31:347-364, 1975
4. Bridges BA: Short term screening tests for carcinogens.
Nature 261:195-200, 1976
5. Chu EHY, Mailing HV: Chemical induction of specific
locus mutations in Chinese hamster cells in vitro.
Proc Natl Acad Sci USA 61:1306-1312, 1968
6. Chu EHY, Powell SS: Selective systems in somatic cell
genetics. Adv Hum Genet 7:189-258, 1976
7. Couch DB, Hsie AW: Dose-response relationships of cyto-
toxicity and mutagenicity of monofunctional alkylating
agents in Chinese hamster ovary cells. Mutat Res 38:399,
1976
8. Couch DB, Hsie AW: Mutagenicity and cytotoxicity of
congeners of two classes of nitroso compounds in Chinese
hamster ovary cells. Mutat Res, in press
9. Couch DB, Forbes NL, Hsie AW: Comparative mutagenicity
of alkylsulfate and alkanesulfonate derivatives in
Chinese hamster ovary cells. Mutat Res, in press
10. Couch DB, Hsie AW: Metal mutagenesis: Studies of the
mutagenicity of manganous chloride and 14 other metallic
compounds in the CHO/HGPRT assay. In: Proceedings of the
Annual Meeting of the Environmental Mutagen Society.
San Francisco, March 9-13, 1978, p 74
11. Deaven L, Peterson D: The chromosome of CHO, an aneu-
ploid Chinese hamster cell line: G-band, C-band, and
autoradiographic analysis. Chromosoma 41:129-144, 1973
-------
312 ABRAHAM W. HSIE ET AL.
12. Epler JL: Short-term bioassay of complex organic mix-
ture: Part II, Mutagenicity testing, this proceedings
13. Gehrs CW: Short-term bioassay of complex organic mix-
ture: Part I, Chemistry testing, this proceedings
14. Guerin MR: Short-term bioassay of complex organic mix-
ture: Part III, Environmental testing, this proceedings
15. Heddle JA, Bodycote J: On the formation of chromosomal
aberrations. Mutat Res 9:117-126, 1970
16. Hsie AW, Brimer PA, Mitchell TJ, Gosslee DG: The dose-
response relationship for ethyl methanesulfonate-induced
mutations at the hypoxanthine-guanine phosphoribosyl
transferase locus in Chinese hamster ovary cells. Somat
Cell Genet 1:247-261, 1975
17. Hsie AW, Brimer PA, Mitchell TJ, Gosslee DG: The dose-
response relationship for ultraviolet light-induced
mutations at the hypoxanthine-guanine phosphoribosyl
transferase locus in Chinese hamster ovary cells. Somat
Cell Genet 1:383-389, 1975
18. Hsie AW, Brimer PA, Machanoff R, Hsie MH: Further evi-
dence for the genetic origin of mutations in mammalian
somatic cells: The effects of ploidy level and selec-
tion stringency on dose-dependent chemical mutagenesis
to purine analogue resistance in Chinese hamster ovary
cells. Mutat Res 45:271-282, 1977
19. Hsie AW,Li AP, Machanoff R: A fluence-response study
of lethality and mutagenicity of white, black, and blue
fluorescent lights, sunlamp, and sunlight irradiation
in Chinese hamster ovary cells. Mutat Res 45:333-342,
1977
20. Hsie AW, Machanoff R, Couch DB, Holland JM: Mutagenicity
of dimethylnitrosamine and ethyl methanesulfonate as
determined by a quantitative host-mediated CHO/HGPRT
assay. Mutat Res, in press
21. Hsie AW, Couch DB, O'Neill JP, San Sebastian JR, Brimer
PA, Machanoff R, Riddle JC, Li AP, Fuscoe JC, Forbes NL,
Hsie MH: Utilization of a quantitative mammalian cell
mutation system, CHO/HGPRT, in experimental mutagenesis
and genetic toxicology. Chemical Industry Institute of
Toxicology (CUT) Workshop on "Strategies for Short-Term
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 313
Testing for Mutagens/Carcinogens," Research Triange Park,
North Carolina, August 11-12, 1977; also in CRC Uni-
science Monographs in Toxicology (Butterworth BE, ed.)»
in press
22. Hsie AW, O'Neill JP, Couch DB, San Sebastian JR, Brimer
PA, Machanoff R, Fuscoe JC, Riddle JC, Li AP, Forbes NL,
Hsie MH: Quantitative analyses of radiation- and
chemical-induced lethality and mutagenesis in Chinese
hamster ovary cells. Radiat Res, in press
23. IARC, IARC Monograph on the Evaluation of Carcinogenic
Risk of Chemicals to Man, Vols. 1-10. Lyons, IARC,
1972-76
24. Ishidate M Jr, Odashima S: Chromosome tests with 134
compounds on Chinese hamster cells in vitro: A screening
for chemical carcinogens. Mutat Res 48:337-354, 1977
25. Kao FT, Puck TT: Induction and isolation of nutritional
mutants in Chinese hamster ovary cells. Proc Natl Acad
Sci USA 60:1275-1281, 1968
26. McCann J, Ames BN: Detection of carcinogens as mutagens
in the Salmonella/microsome test: Assay of 300 chemi-
cals: Discussion. Proc Natl Acad Sci USA 73:950-954,
1976
27. McCann J, Choi E, Yamasaki E, Ames BN: Detection of
carcinogens as mutagens in the Salmonella/microsome
test: Assay of 300 chemicals. Proc Natl Acad Sci USA
72:5135-5139, 1975
28. Natarajan AT, Tates AD, Van Buul PPW, Meijers M, De
Vogel N: Cytogenetic effects of mutagens/carcinogens
after activation in a microsomal systen in vitro. I.
Induction of chromosome aberrations and sister chroma-
tid exchanges by diethylnitrosamine (DEN) and dimethyl-
nitrosamine (DMN) in CHO cells in the presence of rat
liver microsomes. Mutat Res 37:83-90, 1976
29. O'Neill JP, Brimer PA, Machanoff R, Hirsch GP, Hsie AW:
A quantitative assay of mutation induction at the
hypoxanthine-guanine phosphoribosyl transferase locus
in Chinese hamster ovary cells: Development and defini-
tion of the system. Mutat Res 45:91-101, 1977
-------
314 ABRAHAM W. HSIE ET AL.
30. O'Neill JP, Couch DB, Machanoff R, San Sebastian JR,
Brimer PA, Hsie AW: A quantitative assay of mutation
induction at the hypoxanthine-guanine phosphoribosyl
transferase locus in Chinese hamster ovary cells (CEO/
HGPRT system): Utilization with a variety of mutagenic
agents. Mutat Res 45:103-109, 1977
31. O'Neill JP, Hsie AW: Chemical mutagenesis of mammalian
cells can be quantified. Nature 269:815-817, 1977
32. O'Neill JP, Fuscoe JC, Hsie AW: Mutagenicity of hetero-
cyclic nitrogen mustards (ICR compounds) in cultured
mammalian cells. Cancer Res, in press
33. O'Neill JP, Fuscoe JC, Hsie AW: Structure-activity
relationship of antitumor agents in the CHO/HGPRT
system: Cytotoxicity and mutagenicity of 8 ICR com-
pounds. Proceeding of the Annual Meeting of the Ameri-
can Society for Cancer Research, Washington, DC, April
5-8, 1978
34. Perry P, Evans HJ: Cytological detection of mutagen/
carcinogen exposure by sister chromatid exchange.
Nature 258:121-125, 1975
35. Puck TT: The Mammalian Cell as a Microorganism. San
Francisco, Holden-Day,, Inc, 1972, pp 219
36. Puck TT, Kao FT: Treatment of 5-bromodeoxyuridine and
visible light for isolation of nutritionally deficient
mutants. Proc Natl Acad Sci USA 58:1227-1234, 1967
37. Purchase IEH, Longstaff E, Ashby J, Styles JA, Anderson
D, Lefevre PA, Westword FR: Evaluation of six short
term tests for detecting organic chemical carcinogens
and recommendations for their use. Nature 264:624-627,
1976
38. Rosenkranz HS, Gutter B, Speck WT: Mutagenicity and
DNA-modifying activity: A comparison of two microbial
assays. Mutat Res 41:61-70, 1976
39. San RHC, Stich HF: DNA repair synthesis of cultured
human cells as a rapid bioassay for chemical carcinogens
Int J Cancer 16:284-291, 1975
40. Siminovitch L: On the nature of heritable variation in
cultured somatic cells. Cell 7:1-11, 1976
-------
QUANTITATIVE MAMMALIAN CELL GENETIC TOXICOLOGY 315
41. Stetka DG, Wolff S: Sister chromatid exchange as an
assay for genetic damage induced by. mutagens/carcinogens,
II. In vitro test for compounds requiring metabolic
activation. Mutat Res 41:343-350, 1976
42. Sugimura T, Sato S, Nago M, Yahagi T, Matsushima T,
Seino Y, Takeuchi M, Kawachi T: Overlapping of carcin-
ogens and mutagens. In: Fundaments in Cancer Preven-
tion (Magee PN et al., eds.). Tokyo, University of
Tokyo Press/Baltimore, University Park Press, 1976,
pp 191-215
43. Thompson LH, Baker RM: Isolation of mutants of cultured
mammalian cells. Methods Cell Biol 6:209-281, 1973
44. USPHS, Survey of compounds which have been tested for
carcinogenic activity. USPHS Publication No. 149,
1972-73
45. Vogel E: The relation between mutational pattern and
concentration by chemical mutagens in Drosophila. In:
Screening Tests in Chemical Carcinogenesis (Montesano R,
Bartsch H, Tomatis L, eds.). IARC Scientific Publica-
tion No. 12, Lyons, IARC, 1976
46. Zimmerman SK: Procedures used in the induction of
mitotic recombination and mutation in the yeast
Saccharomyces cerevisiae. Mutat Res 31:71-86, 1975
-------
ENVIRONMENTAL TESTING
CW. Gehrs, B.R. Parkhurst, and D.S. Shriner
Environmental Sciences Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee
-------
319
INTRODUCTION
Environmental toxicology is a term that conveys different
images to different people. In its broadest sense environ-
mental toxicology encompasses all of the research necessary
to evaluate the potential ecological effects, to determine
the ultimate fate in the environment, and to identify crit-
ical pathways to man that might occur as the result of re-
lease of a particular material (Table 1). Evaluation of
the potential toxicity of the material can be accomplished
through three types of testing: short-term bioassays (envi-
ronmental screening); subacute organismic, population, com-
munity, and ecosystem evaluation; and mechanistic studies.
The latter two types are resource intense in that they gen-
erally require substantial manpower and time commitments.
However, they are also essential for developing predictive
capabilities regarding the potential for long-term chronic
environmental effects resulting from the release of a com-
plex effluent stream. Environmental screening, on the other
hand, requires less manpower with results often obtained in
less than a week's time.
Research sponsored by the Division of Biomedical and Environ-
mental Research, U.S. Department of Energy, under contract
W-7405-eng-26 with Union Carbide Corporation. Publication
No. 1187, Environmental Sciences Division, ORNL.
-------
320
C.W. GEHRS ET AL.
Table 1
Description of Various Segments of
Environmental Toxicology Research
Environmental Toxicology
t
Effects
1) Short-term Bioassays
LC
a
5 0
GR5
2) Subacute Testing
organism, population,
ecosystem
3) Mechanistic Studies
chemical structure/
biological activity,
reproductive impairment
physiological responses
t
Transport
1) Abiotic Processes
hydrolysis, photolysis,
sorption/sedimentation
2) Biotic Processes
microbial degradation,
uptake, bioaccumulation,
bioconcentration, trans-
formation, tissue dis-
tribution
LCs0, the concentration of original test material that
will result in mortality of 50% of test organisms
in a certain time (usually 48 to 96 hours).
DGR50, the concentration of original test material that
will reduce the growth of the test organisms during
a certain time (usually 48 hours).
Unfortunately, most environmental screening does not
provide the types of data necessary either for predicting
potential ecological effects or for estimating potential en-
vironmental risks. This paper is limited to a discussion of
short-term environmental bioassays and of potential uses of
information obtained from this type of testing. The inte-
gration of chemical characterization and fractionation into
the testing protocol is discussed, as well as the rationale
for selecting appropriate test systems. Two of the systems
currently employed are described.
-------
ENVIRONMENTAL TESTING 321
Short-term bioassays have the potential for providing
toxicity data for three sets of users. They can (1) provide
guidance to control technologists and waste management per-
sonnel regarding which components of a complex mixture are
biologically active and, hence, may require removal before
effluent discharge; (2) provide guidance to the environmental
scientist in determining which components of a complex mix-
ture require further evaluation in subacute and mechanistic
studies; and (3) provide semi-quantitative hazard assessment
data when comparisons are made to standard reference com-
pounds or toxicity data from other complex mixtures.
USE OF A BATTERY OF TESTS
A major consideration in the environmental testing of
complex chemical mixtures is that such mixtures contain
chemicals of a wide variety of classes and an equally wide
range of concentrations potentially capable of causing addi-
tive, antagonistic, or synergistic responses in test organ-
isms or at specific receptor sites. Such interactions may
be a function of the combined dose of toxicants or of the in-
herent genetic susceptibility of a particular target organism.
The genetic susceptibility of an organism may be due to (1)
the presence or absence of effective barriers to absorption
or translocation of a chemical; (2) the selective accumula-
tion of the chemical in a bound or inactive form, or in tis-
sues remote from a receptor site; (3) the presence or absence
of the ability to detoxify the substance through biotrans-
formation; (4) the existence or absence of specific target
or receptor systems in exposed cell systems; or (5) some
combination of the above factors (1). In addition, since a
specific compound may influence different biologic receptor
sites within an organism through a variety of pathways, the
problem of environmental hazard assessment is extremely
complex.
To establish with absolute certainty the degree of haz-
ard posed by a particular material would require testing of
all potential target organisms for their various responses.
In the light of the large number of complex mixtures entering
the environment each year which require evaluation for their
potential hazard, it is readily apparent that the costs, both
of time and money, for such an option prevent the making of
assessments in a realistic time frame. However, at the other
end of the spectrum, predictions concerning potential hazards
based on the response of a single species to a chemical mix-
ture have an unacceptably low level of certainty or dependa-
bility (because of differential species responses, etc.).
-------
322 C.W. GEHRS ET AL.
The task then is to arrive at some intermediate point
in cost and effort that permits an acceptable estimate of
hazard potential. One method of increasing the level of
confidence is to use a series of test systems and develop a
hazard estimate based on pooled results from these systems.
This group of systems would ideally attempt to represent as
broad a spectrum as possible of taxa of organisms, tissue
ages, growth forms, routes of exposure, and environmental
variables. In this manner, ""we anticipate being able to ob-
tain estimates of potential hazard which will reflect better
the range of response variability expected from a normal
population of organisms.
CRITERIA FOR TEST SYSTEM SELECTION
For a specific system to be functional in environmental
screening it must be of short duration, require minimal quan-
tities of material, and be a standard test system, i.e., it
must employ an organism for which a large base of toxicolog-
ical data is available.
The test system must be of short duration, not only to
minimize manpower costs, but also to prevent confounding
interpretation of results that might occur from chemical
changes in the aqueous media. Chemical separation to pro-
vide relatively discrete fractions is costly, time consum-
ing, and produces only small amounts of material (2). It
is impractical to use this approach, for example, for fish
requiring several gallons of water per replicate, or plants
requiring similar quantities of test media. The final cri-
terion, using a test system for which a large data base
already exists, is essential if even a semiquantitative
assessment is to be made.
DESCRIPTION OF TEST SYSTEMS EMPLOYED
Bioassays of varying types have been used successfully
over a wide range of applications. Prominent among these ap-
plications are tests with herbicides, insecticides, and plant
hormones that have dealt essentially with effects of specific
compounds. When employing a bioassay test in the screening
of complex mixtures of chemicals, there are two basic assump-
tions that are made: (1) the species used will show an in-
jury response in proportion to the concentration of the bio-
logically active chemical species; and (2) the responses
obtained are reproducible (3).
-------
ENVIRONMENTAL TESTING 323
The two species used in the experiments discussed in
this paper are the zooplankter, Daphnia magna, and the radish,
Raphanus sativus. The parameter used in the zooplankton sys-
tem is the 48-hr LC50. Although these aquatic organisms are
easily cultured in the laboratory, they have been found to be
sensitive to aquatic pollutants. Four D. magna are placed in
80 ml of each toxicant test solution in 100-ml beakers cov-
ered with watch glasses. The small numbers of test organisms
and the small volumes of media used are necessitated by the
small quantities of test materials (often less than one gram)
that are derived from the chemical fractionation procedure.
Water temperature is maintained at 22 +_ 0.5°C by placing the
beakers in an environmental chamber. Photoperiod is main-
tained under a 12-hr light/dark regime. Toxicant solutions
are prepared with filtered spring water (pH 7.8, alkalinity
119 mg/liter, hardness 140 mg/liter). All tests are run in
triplicate. Serial dilutions with each concentration being
60% of the previous one are made for each test material.
Controls of spring water without added toxicant are included.
The range of dilutions are selected to bracket the 48-hr
LC50, which is obtained by computerized PROSIT analytical
procedures (4). The presence of toxic interactions between
fractions is determined using the additivity index of
Marking and Dawson (5).
Radish seed responses to chemical mixtures are expressed
in terms similar to the LCSO value used in the Daphnia stud-
ies. The value obtained represents the concentration of the
original mixture (or concentration of specific components in
the original mixture) which reduced the yield of a specific
species by 50% (3,6,7), and is expressed as GR50 (concentra-
tion for 50% growth reduction). Measurement parameters used
as estimates of yield include percent germination time to root
emergence, fresh weight, and root elongation. A series of
seed germination tests (according to the procedure discussed
below) with water blank control solutions adjusted to a range
of pH's are performed at different pH levels and serve as con-
trols. The pH of each test material is determined when the
material is received. Seed germination tests are then per-
formed on each test material, which consists of seeds in petri
dishes containing filter paper moistened with the toxicant.
Germination percentages are determined, and length, fresh
weight, and time to root emergence are measured and compared
with control seeds. Toxicants showing inhibitory effects
on germination percentages or root growth are diluted and
tested as above to determine a GR5fl. Toxicants showing no
effects in the original test are not tested further. Speci-
fic fractions of material are also tested to determine GR
-------
324 C.W. GEHRS ET AL.
ISOLATION OF TOXIC COMPONENTS
The sequential procedure employed to isolate and iden-
tify the toxic components of complex mixtures is shown in
Figure 1. In the first step the mixtures are screened for
acute toxicity (i.e., the Bodean test). Mixtures not found
to be acutely toxic do not undergo further testing. This
does not mean that potential problems do not necessarily
exist with those materials, but identification of such
potential problems would need to await chemical screening.
Those mixtures found to be acutely toxic are separated by
chemical extraction into their organic and inorganic compo-
nents and tested for acute toxicity. If acute toxicity is
found in the organic components, the component is fraction-
ated into acid, base, and neutral fractions. At this time,
the relative amount of each fraction in the total organic
component is also determined. The toxicity of each fraction
is determined and its contribution to the toxicity of the
original mixture calculated by relating the relative toxicity
of the fraction to its concentration in the original mixture.
If further isolation of the actual toxic compounds of the
organic fractions is desired, subfractionation and testing
can be performed. Up to 14 such subfractions have been sep-
arated from synthetic fuel process effluents for use in
mutagenesis testing, and these same subfractions could be
produced for acute toxicity testing (8). Ultimately, tests
with specific compounds can be used to evaluate the toxicity
of individual chemical components of the mixture.
A different approach is used to identify the toxic com-
ponents of the inorganic fraction of the complex mixture.
If the inorganic fraction is found to be acutely toxic, it is
chemically characterized to determine its composition. The
toxicities of the individual components are ascertained and
the contribution of each component to the toxicity of the
original mixture assessed by the same method used for the
organic fraction (Figure 1).
In the last step of this environmental screening an ef-
fort is made to determine whether the testing procedure has
accounted for all of the toxicity of the original complex
mixture. The individually identified inorganic components
present in the original mixture are combined to produce a
"reconstituted" mixture. The acute toxicity of this mixture
is tested and compared to the toxicity of the original
mixture.
-------
ENVIRONMENTAL TESTING
325
COMPLEX MIXTURE
INITIAL SCREENING
TOXIC
(INITIAL SEPARATION)
NONTOXIC
TOXIC
(FRACTIONATION)
/ I \
NONTOXIC
TOXIC
(CHARACTERIZATION; NA, A.A.)
ACID
| BASE
NEUTRAL
QUANTIFICATION AND RELATIVE TOXICITY DETERMINATION
Figure 1. Flow diagram that shows testing patterns used to
identify toxic components.
The procedure outlined above has been used to identify
the toxic components of several types of synthetic fuel aque-
ous effluents. The data from two are presented in this paper
as examples. They include one in which inorganic components
were the most toxic and one in which organic compounds were
the most toxic. Only data from the Daphnia system are pre-
sented. Both of the effluents tested are complex mixtures of
hundreds of individual organic and inorganic compounds which
are byproducts of coal conversion processes. The first mate-
rial tested was an untreated process-water effluent from the
solvent refined coal (SRC) pilot plant in Ft. Lewis, Washing-
ton. The second material tested consisted of two effluents,
untreated and biologically treated hydrocarbonization (HCZ)
process-water from the process development unit (PDU) at the
Oak Ridge National Laboratory.
The results of the application of the acute toxicity
testing protocol to the SRC effluent revealed the total
effluent to be acutely toxic to Daphnia, with a 48-hr LC50
estimated to be at a dilution of 15.7% (9). The inorganic
-------
326 C.W. GEHRS ET AL.
portion of the effluents was found to be nontoxic; i.e., con-
centration was necessary before an LC50 could be produced.
Further testing required chemical separation. Of the three
fractions extracted from the organic portion of the effluent,
the neutral fraction had the highest toxicity with a 48-hr
LC50 of 9 mg/liter (Table 2). The 48-hr LC50s of the acid
and base fractions were 29.5 mg/liter and 45.8 mg/liter.
respectively. Based on the concentrations of the individual
fractions in the SRC effluent, the toxicity contribution of
each fraction to the toxicity of the whole SRC effluent was
calculated (Table 2). These toxicity contributions ranged
from 53% for the acid fraction to 3.6% for the base fraction.
One of the values of being able to test the whole efflu-
ent as well as the various fractions is that toxic interac-
tions between fractions can be detected and quantified.
Testing of whole effluents (before fractionation) and recon-
stituted effluents enables gross evaluation of the role of
the fractionation scheme in producing artifacts in the efflu-
ent components. The results of an experiment using effluents
from an SRC facility (Table 3) suggest that neither toxic
interactions or artifact formation occurred in the SRC efflu-
ents. The 48-hr LC50s of the whole effluent and the recon-
stituted effluent were found to be 15.7% and 15.5%, respec-
tively, calculated as a percent dilution of the effluent
(Table 3). These values were not significantly different (t,
P = 0.05), and indicated both that (1) the chemical fraction-
ation procedure did not alter the toxicity of the fractions
and (2) all of the toxicity of the effluents was accounted for
Table 2
Acute Toxicity of Organic Components
of Untreated SRC Effluents
Components
Concentration in effluent (mg/1)
Composition of total effluent (%)
Daphnia (48-hr LC50 in mg/1)
Relative contribution to
Acid
99.5
28.5
29.5
53 .0
Neutral
24.8
7.1
9.0
43.4
Base
10.6
3.0
45.8
3.6
effluent toxicity (%)
-------
ENVIRONMENTAL TESTING
327
CO
rH
rQ
oS
EH
p
fl
C^
3
rH
^
p
•H
0
*rH
X
0
EH
rH
aJ
P
O
EH
0
P
CQ
C!
o
•rH
P
O
aJ
rH
EH
O
03
CO
f. ,
O
CO
c
o
•H
p
3
JD
•H
rH
P
C
O
o
X
o>
73
B
0)
•H
•P
•H
73
•o
p
C
CU
3
O rH
W VI
Ed
•o
CD
•P -P
3 B
•p cu
•H 3
P rH
CO VI
C VI
O td
O
CD
a
-P
- B
CD x-\
rH fc«
0 ^
£
4J
0 -H
•P 0
•r-4
B X
O O
•H EH
p
3 -P
J2 C
•H CU
h 3
•P rH
C VI
O vi
CJ fid
cj cu
OS 3
CO rH
VI
VI
IS
•o
cu
•p -p
3 B
•H 3
•P rH
CO VI
O fid
o
CD
CC
^
rH
~-~ C
ts> o
a!
B M
O -P
•P CU
ai O
U C
•P 0
B CJ
CP
O »
B •»
0 CJ
CJ J
B w
O J3
•H 1
•P 00
O •*
aS
i* -P
fa ai
•P
C
CJ 3
QC rH
CO VI
a
•a
CD
•P +J
3 B
•P 0)
•H 3
P rH
CO Vi
B VI
O fid
o
cu
ai
o
M
CJ
£
1
CO
•*
B
O
p
o
OS
^
fa
co m
•M CO
m
in in
rH CO
in
CD 0
• •
rH
CM CO
• •
m rH
rH
m oo
en in
IN •*
"O CD
•H CO
O a)
^J CQ
CO
o
o
o
O) CO
-H CO
CM O
CM CO
05 tO
CO CO
CO
00 t>
co w
CO
o
ai o
o
o
o
•H
rH d
ai as
S-, bC
P H
3 O
CU S
Z rH
CU
aS
CD
0
•p
T3
0
£
3
CO
aS
CO
'
rH
~^.
bfl
£
O
O
o
rH
•
CO
CD
•H
S-.
CU
to
CO
p
fl
•H
B
tratio
c •
CD »
0 CJ
B J
0
0 M
J3
P 1
CO CO
eg > CD
-P >
•r-* -H
rH -P
ai cS
•P >
(H rl
O CD
S > C
CO OS
!>» !>v
•P -P
O >» O
•H -P -rl
X -H X
0 O O
•P iH -P
CD O CD
> -p >
•r4 *H
•P CD P
73 -rH 73
73 P 73
ai -H ai
•a
E T3 B
aS ai at
P CD P
rH
CD a. co
rH £ CO
O -H CD
£ CO rH
II II II
o o o
A j V
XXX
CD CD CD
73 73 73
B B B
•H -H 'H
CD CD CD
> > r>
•H *H *H
•P +•> -P
•H -H -rl
73 73 73
73 73 73
w
.
>>»
Jp
CJ
B J
CD s.
•iH CO
bfi CD
•H
CO P
•H -H
O
•P -H
a x
CD O
3 -P
rH
VI 0)
VI £,
CD P
CD . -H
O
CU
rH
O
£1
=c
cu
j3
>J
. . CD CD
•P £ O
•rH -P B
0 O
•H CD O
X !H
O ai B
•P O
•H
CD IS! -P
•P 1
o
ai
1 H
O 1 vi
p
P3 CU
B fi
O P
•H <
•p a
O CD
al h 73
fc. CD B
a!
VI
O
C
o
•H
4^
3
n S *H
•H ^1 ^
1^
4H
B
0
O
cu
S
EH
••
CO
£ -H B
N tSI O
•H
+ -P
' O
1 ai
1 ti
1 VI
+ rH
CS •
£ -H 3 -P
CO CQ 73 B
•H CD
+ > 3
•H rH
£ -H 73 VI
< < C Vl
•H CU
-------
328 C.W. GEHRS ET AL.
in the four fractions. Using the method of Marking and Dawson
(5), the additivity index was 0.03 for the whole SRC effluent
and 0.00 for the reconstituted effluent (Table 3). These
values were not significantly different (t, P = 0.05), indi-
cating that the toxicities of the individual fractions were
directly additive within the effluents. As a comparison, if
each fraction acted independently, an additivity index of
0.95 would be predicted.
The second example of the environmental screening ap-
proach deals with effluents from the Oak Ridge National Labo-
ratory hydrocarbonization unit (HCZ). In the first step of
the toxicity screening of the HCZ effluent, the toxicities of
the untreated and treated effluents were compared. The treat-
ment process was determined to have reduced the toxicity of
the effluent by 99% in the Daphnia system. Further testing
showed that the inorganic portion of the effluent contributed
99.5% of the toxicity (Table 4), with the remaining 0.5% con-
tributed by phenols. Of the inorganic constituents, ammonia
was the principal toxic agent, contributing 96% of the toxi-
city.
The determination of interactions between the HCZ efflu-
ent components demonstrated a less than additive or antagon-
istic behavior. This indicated that the sum of the toxici-
ties of the effluent components calculated individually was
greater than when they were present together in the effluent.
Table 4
Acute Toxicity Data for Hydrocarbonization
Qrganics Inorganics
Concentration in effluent (rag/1) 110 2846
Relative quantity (%) 3.7 96.3
Daphnia (48-hr LCso in mg/1) 774 31.7
Relative contribution to 0.5 99.5
effluent toxicity (%)
-------
ENVIRONMENTAL TESTING 329
SUMMARY AND CONCLUSION
The previous discussion has shown how coupling of chemi-
cal separation and fractionation with environmental testing
is able to identify those materials most biologically active,
at least, in acute toxicity. In the case of the SRC effluent
the primary activity was attributed to the acidic organic
fraction, where the phenolics are located. Because of the
relative removal efficiency of phenolics through chemical
stripping and biological treatment (10), near field acute
toxicity from effluent releases would not be expected from
aqueous effluents released from a facility similar to the SRC
pilot plant. The ultimate conclusion concerning potential
effluents from the HCZ process development unit is the same,
although the biologically active component was the inorganic
NHa .
Environmental testing can help the control technologist
and waste management engineer identify materials of potential
environmental consequence. These screening activities are
not intended to, and should not, replace subacute toxicity
and mechanistic toxicity studies for developing data for pre-
dictive purposes with respect to chronic low level effects
in the far field environment.
REFERENCES
1. Loomis TA: Essentials of Toxicology. Philadelphia,
Lea and Febiger, 1970, p 162
2. Guerin MR, Clark BR, Ho CH, Epler JL, Rao TI: Short-
term bioassays of complex organic mixtures: Part 1.
Chemistry. Application of Short-term Bioassays in the
Fractionation and Analysis of Complex Environmental
Mixtures (1978) EPA-60019-78-027.
3. Santelmann PW: Herbicide bioassay. In: Research
Methods in Weed Science. Southern Weed Science Society,
177, pp 79-88
4. Finney DJ: Statistical Methods in Biological Activity.
London, Griffin Press, 1971, 2nd Edition
-------
330 C.W. GEHRS ET AL.
5. Marking LL, Dawson VK: Method for assessment of toxic-
ity of efficacy mixtures of chemicals. Investigations
in Fish Control No. 67, U.S. Department of Interior,
Fish and Wildlife Services, Washington, D.C., 1975
6. Sheets TJ: Effects of soil type and time on herbicidal
activity of CDAA, CDEC, and EPTC. Weeds 7:442-448,
1959
7. Upchurch RP: The influence of soil factors on the
phytotoxicity and plant selectivity of diuron. Weeds
6:442-448, 1959
8. Rubin IB, Guerin MR, Hardigree AA, Epler JL: Fraction-
ation of synthetic crude oils from coal for biological
testing. Environmental Research 12:358-365, 1976
9. Parkhurst BR, Gehrs CW, Rubin IB: The value of chemi-
cal fractionation for identifying the toxic components
of complex aqueous effluents. Proceedings of ASTM 2nd
Annual Symp. on Aquatic Toxicology. Cleveland, in
press-
10. Herbes SE, Southworth GR, Gehrs CW: Organic contami-
nants in aqueous coal conversion effluents: Environ-
mental consequences and research priorities. Proceed-
ings of Trace Substances in Environmental Health-X,
(Hemphill DD, ed.), University of Missouri, Columbia,
1977, pp 295-303
-------
INTEGRATING
MICROBIOLOGICAL AND
CHEMICAL TESTING INTO
THE SCREENING OF AIR
SAMPLES FOR POTENTIAL
MUTAGENICTTY
Edo D. Pellizzari, Linda W. Little,
Charles Sparacino, and Thomas J. Hughes
Research Triangle Institute
Research Triangle Park, North Carolina
Larry Claxton and Michael D. Waters
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
-------
333
A recent review of respiratory carcinogenesis (10) notes
that presence of chemical carcinogens in the atmosphere,
especially in certain urban and industrial environments, was
described more than two centuries ago in Pott's studies in
chimney sweeps and a century ago in the work of Harting and
Hesse on uranium miners. In the last two decades, with the
availability of sophisticated air sampling devices, analyti-
cal chemistry techniques, and bioassay procedures, the iden-
tity and carcinogenicity of many air pollutants, especially
those organic compounds that can be extracted by solvents
from particulates, has been documented (10).
Considerable work remains to be performed on those
phases which are more difficult to collect, identify, and
bioassay, including the insoluble portions of particulates
(10) and the volatile organic components (17). A particu-
lar problem needing study, notes Van Duuren (17), is the
role of aromatic hydrocarbon compounds, such as pyrene and
fluoranthene, which are noncarcinogenic but act as potent
cocarcinogens. In this regard, Weisburger (18) points out
the necessity for studies to "delineate the carcinogenic
risk of specific and rationally selected mixtures which may
affect man."
The most direct measure of human risk would be provided
by studies with human beings. Whereas in water pollution
"the fish is the final arbiter of toxicity" (4), here the
human being is the final arbiter of carcinogenicity. In-
formation on human effects is usually obtained after the
fact in case histories or epidemiological studies.
-------
334 EDO D. PELLIZZARI ET AL.
According to Kluyver's principle of the unity of bio-
chemistry, a wide variety of organisms' effects at the sub-
cellular level should show a great deal of similarity. Thus,
other mammals, especially rodents, have been employed in
assessment of carcinogenic risks to man. However, definitive
bioassays are expensive because they require months to years
for completion, large numbers of experimental animals, and
large amounts of test sample.
Ames and associates (1,2,3,12,13) have developed and
validated a bacterial mutagenesis assay which detects as
mutagens approximately 85 percent of the known carcinogens
that have been tested. The test is rapid and economical,
requires little space, and is sensitive to nanogram or
microgram levels of many mutagens. By addition of mammalian
microsomal enzymes to the test system, the test can detect
mutagens requiring metabolic activation.
Because of these advantages, the test has been recom-
mended by EPA as one of bioassays to be used in "Level 1
Environmental Assessment" (6,8), designed by the EPA Indus-
trial Environmental Research Laboratory to be a cost-effec-
tive approach in screening emissions to determine which
"have a higher potential for causing measurable health or
ecological effects." Thus, this test should receive priority
for further assessment (8). The Ames test has been employed
in detection of airborne mutagens by several investigators,
including Talcott and Wei (16), Pitts et al. (15), and
Flessel (7).
For the past eight months, an EPA-sponsored program
has been underway at Research Technical Institute to develop
a protocol that will define a minimal biological and chemical
methodology to serve as a screen for the potential carcino-
genicity of complex air pollutant mixtures. This protocol
includes sampling, fractionation, chemical identification,
and mutagenicity method development. Specifically, the pro-
gram involves:
• Collection at selected locations of ambient
air samples containing particulate and vapor
phase material.
• Mutagenesis testing of crude particulate and
vapor phase materials.
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES 335
• Treatment of crude particulates by separating
into major chemical classes through organic
fractionation procedures.
• Chemical characterization of fractions showing
mutagenic activity.
• Mutagenesis testing of vapor phase components.
Studies to date have focused on particulate phase components.
Sampling
Collection of a particulate sample is accomplished
using the Battelle Maxi-sampler which partitions the sample
into three particle size ranges, > 3.5 urn, 1.7-3.5 ym, and
< 1.7 vim, the latter two ranges representing respirable
particles. The Maxi-sampler can collect a relatively large
amount of particulate, a gram or more per 24-hour period.
This is important for ultimate bioassay and chemical identi-
fication procedures because of the sensitivity requirements
associated with these processes.
FRACTIONATION
The well documented complexity of air particulate
extracts, and the desirability of reducing the number of
compounds that require identification, indicates the need
for a suitable prefraction method. We have adopted a pro-
cedure that is reproducible, mild, and effective in which
particulate extract is divided into a relatively small number
of fractions, each with a similar chemical make-up.
The fractionation scheme proposed and used initially
in this program involved sonication of the particulate with
two solvents, cyclohexane and methanol. This approach has
been shown to be effective in removing significant amounts
of polar materials that are not removed by the more usual
treatment with a single nonpolar solvent such as cyclo-
hexane or benzene. It was hoped that the materials extracted
by the two solvents could then be further treated separately
to give two types of fractions for testing, namely, nonpolar
and polar. The complete fractionation scheme yielded 13
separate samples for bioassays (Figure 1).
-------
336
EDO D. PELLIZ2ARI ET AL.
F-ARTICULATE
'• SOMC-TE '\
I 2 FILTER
RE
SIOUE FIL1
H SONICATE IN MtOM
21 FILTER
1
INORGANIC
RESIDUE
RATE
COMB
11 RE
2) RE
31 AC
COMBINE FILTRATES
II REMOVE SOLVENT
2) REDISSOLVE IN CHjCL2
31 ACID WASH SEQUENCE'
AQUEOUS
'DISCARD)
1) 8ASIFY TO pH 10
21 EXTRACT 3X WITH CH2CU2
CH2CL2
1) BASE WASH ScuUeNCE*
CH2CL2
CHjCl.2
11 REMOVE SOLVENT
2) WEIGH
ORGANIC BASES
11 ACIDIFY TO pH 3
21 EXTRACT 3X WITH
CH2CL?
CH2CL2
II REMOVE SOLVENT
2) WEIGH
AQUEOJS
IOISCAROI
1) REMOVE SOLVENT
2) REDISSOLVE IN CSH1Z
ORGANIC ACIDS
1NSOLUBLES
C6H12
11 WASH WITH MEOM/H2O
| 1) WASH 3X WITH
21 M6N02
I
MEOH/HjO
II FREEZE DRY
21 WEIGH
POLAR NEUTRALS
I li REMOVE SOLVENT
i 2l WEIGHT
11 REMOVE SOLVENT
2) WEIGHT
iCID «ASH SEQUENCE 2 X WITH •!) HjSOj 1 X WITH 20' HjSOj
BASE rtASH SEQUE\Ct 3 X .MTH • 'J \aOH
Figure 1. Initial Fractionation Scheme.
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES 337
It quickly became evident that the amounts of material
available from the fractionation of the particulate extracts
for bioassay were in some cases too low. It was also found
through preliminary gc/ms results that the difference
between the compositions of the polar and nonpolar fractions
was small, i.e., essentially the same compounds were found
in both sample types. For example, the nonpolar and polar
organic base fractions were virtually identical in terms of
components identified. Hill et al. (9) have shown that the
composition of methanol vs cyclohexane extracts of air par-
ticulate samples are largely duplicative although there are
real differences. We have thus modified the procedure by-
combining the cyclohexane and methanol extracts to produce
a single crude particulate extract which is then fractionated
through a solvent partition scheme to produce six fractions
instead of 13 as before. This, of course, will provide more
sample per fraction and thus reduce somewhat the burdens of
high sensitivity on the bioassay.
The new partition scheme also incorporates other modi-
fications which increase its usefulness and reliability.
Attempts to validate the initial fractionation procedure with
a known sample were disappointing. For example, a sample
containing 53.5 mg of quinoline as the only organic base
present was partitioned with 0.1 normal HC1. Recovery of
the base from the acid phase produced only 6.4 mg (12%) of
quinoline. Hence, a new procedure was adopted utilizing
several sulphuric acid (10% and 20%) washes. When this pro-
cedure was applied to the 53.5 mg sample of quinoline, the
recovery of base was quantitative. Other modifications to
this scheme were based on published work. The entire scheme
is depicted in Figure 2. The efficacy of the scheme was
assessed by subjecting a mixture containing known amounts of
compounds to the partition procedure. The mixture consisted
of benzoic acid, phenol, quinoline, hexadecane, phenanthrene,
and ethylene glycol. The compounds were chosen to represent
the five classes of materials produced by the partition
scheme. All these materials have been found in air partic-
ulate samples except ethylene glycol; no information on the
composition of the polar neutral fraction is available.
Ethylene glycol was included as a likely component of this
fraction based on its known chemical properties. The experi-
ment was conducted using both large and small mass samples.
Recoveries were determined gravimetrically. The results are
depicted in Table 1.
-------
338 EDO D. PELLIZZARI ET AL.
FROONATON SCHEME
(TUB) SAMPLE (RWTCJLATES)
(I) SONCATE aaOHEXMtE
(2) FILTER
SOUOS
I
TOUWORGAXCS AMD INGRGMKS
(I) SCNCHE-MaOH
(2) RUHR
•ECN-POLAfi ORGANIC
EXTRACTON WITH AQUEOUS ACTS
|
1
gjjQj y^ AQUEOUS LAYER CYCLOHEXANE
I INCH-POLAR BASES rr-9il I NON-POLAR AQOS AND NEUTRALS cr-e;
POLAR ORGAMCS *NO NORGAMCS 1 ' \ '
(I) EVAWRATDN LZVQ- 3 EXTRACTON *,TH AOUEOUS BASE
ia exTRAcno* «rrH ca,o,
j!3) OLTER
OCLOHEXANE AOUEOUS LAYER
CH2O2 | NON-POLAR NEUTRALS ITHO)| JNON-POLAH AaDS(T-lllj
I
FPOLAB acAMcsT^sn PARTITIOK LEVEL 3
QrmAcn» wrro AOUEOUS AOJS cicu»«xA«-.AWH/HO(4ii
AOJEOUS LAYER
[POLAR BASES fT-H| I >^AR A^5^-MEUtRAtjTP4|| OCUWEXANE MMH/IUX4 II
LEVEL 3 EXTRACmC WTTH JOUEOUS BASE | fHUfFHi (T-B1| | AROMATCS IT-IjTJ
i - ; 1 1 I
[POLAR NEUTRAUCMIJ AQUCOU5 LAYER LEVEL 3 LEVEL 3
LEVEL 3 IPOLAR AQDS (T-TH
LEVELS
Figure 2. Final fractionation scheme.
As a further check on the procedure, TLC scans were con-
ducted on each fraction to ascertain the extent, if any, of
spillover of one compound into other fraction(s). No such
spillover was detected.
The first sample to be examined under this program was
collected at South Charleston, West Virginia, during August,
1977, using the Maxi-sampler. Sampler plate scrapings were
made at EPA and some four grams of material (less than 1.7 ym
diameter) were received by Research Triangle Institute (RTI).
This material was neither scraped nor stored by EPA under
ideal conditions. However, the aim at this stage of the pro-
gram was to develop the necessary extraction, partition, and
identification methodology; the sample was thus considered
acceptable for this use. Later samples have been received
which have been properly handled both at EPA and RTI.
For identification of the fractional components, the
following gc/ms run parameters were employed. The gc/ms
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES 339
Table 1
Partition Scheme Validation Results
Amount Amount Amount
Added Recovered Recovered
Compound Class (nig) (nig) (nig)
Large Initial Mass
Benzoic Acid/Phenol
Quinoline
Hexadecane
Phenanthrene
Ethylene Glycol
Small Initial Mass
Benzoic Acid/Phenol
Quinoline
Hexadecane
Phenanthrene
Ethylene Glycol
Org. Acid
Org. Base
Nonpolar
Neutral
PNA
Polar
Neutral
Org. Acid
Org. Base
Nonpolar
Neutral
PNA
Polar
Neutral
41.2/39.0
42.4
45.1
45.4
44.6
0.8/6.4
5.5
8.6
4.3
5.0
74.0
40.2
10.0
34.4
43.4
7.0
5.3
6.9
4.3
3.6
92.3
94.8
22.2
75.8
97.3
97.2
96.4
80.2
100.0
72.0
studies were carried out using an LKB 2091 gas chromatograph/
mass spectrometer. The samples were chromatographed on an
OV-101 capillary column (25 m, SOCT, LKB) using a linear
temperature program. The column was held at 100°C for 2
minutes after injection, and then heated to 240°C at a rate
of 8°/min. Carrier gas flow rate was 2.0 ml/min with a split
ratio of 10:1. Injector temperature was 240°C. Mass spectral
scans were taken every 2 sec scanning from 50-492 mu. Total
ion current and mass plots were generated for interpretation.
-------
340 EDO D. PELLIZZARI ET AL.
Identifications were achieved by comparison of data from
generated mass plots with the Aldermaston 8 peak index of
mass spectra. Components which have been identified thus
far are shown in Table 2.
MUTAGENICITY TESTING
Materials and Methods
Bacterial Strains. Salmonella typhimurium strains used are
the histidine deficient mutants used to detect frameshift
reverse mutations (TA98, TA1537, and TA1538) and base pair
substitutions (TA100 and TA1535). All were obtained from
Dr. Bruce N, Ames, Biochemistry Department, University of
California at Berkeley.
Preparation of Liver Homogenate S~9 Fraction. Male Sprague-
Dawley or Craig-Dawley rats induced with Aroclor 1254 are
used, in preparation of liver homogenates. For each prepara-
tion a minimum of 3 rats is used. Induction involves a sin-
gle intraperitoneal injection of Aroclor 1254 in corn oil,
at a dose of 500 mg/kg, 5 days prior to sacrifice (3). The
S-9 fraction was prepared and stored in 2-5 ml aliquots at
-80°C for no longer than one month.
Presentation of Test Materials. Test materials are dissolved
in spectral grade dimethyl sulfoxide (DMSO), Schwartz-Mann or
Fisher brand. Other solvents are under investigation.
Test Procedure. For routine testing the plate incorporation
and spot test methods of Ames and associates are employed (3).
A well test procedure has been devised which allows detection
of toxicity, mutagenicity, and activation requirements, thus
decreasing the total amount of material required (11).
INTERFACING CHEMICAL AND BIOLOGICAL TESTING WITH PARTICULATES
Crude samples and fractions thereof, as described earlier
in the section on chemical fractionation procedures, must be
presented to the Salmonella assay system for determination of
mutagenicity.
For the Level 1 assessment mutagenicity test, the IERL-
RTP procedures manual (6) recommends that each sample undergo
a mutagenesis test, and if possible, a toxicity test, with the
mutagenesis test to include 4 tester strains; with and without
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES
341
Table 2
Preliminary Identifications of Partition Fractions
Polar Bases
Nicotine
Di-butylphthalate
Butylbenzyl phthalate
Di-2-et-hexyl phthalate
Butyl benzoate
Butyl-a-methyl benzylamine
Di-i-bu phthalate
2-et-hexyl mercaptan
2-nitro-4,6-dichlorophenol
Phenyl benzoate
Polar Neutrals
2,6-di-t-butyl-p-cresol
n-pentylthiol-n-butyrate
2,5-dimethyl-4-isopropenyl-
2,3-hexadien-5-ol
Butylbenzylphthalate
Di-n-octyl phthalate
Nonpolar Neutrals
Diphenyl diacetylene
Butryl benzyl phthalate
2 ,6-di-4-butyl-p-cresol
2,4-dimethylundecane
**Fluoranthene (17)
**Pyrene (17)
Triphenylene
Di-n-octylphthalate
Perylene
*Benzpyrene (5,17)
Methyl heptadecanoate
Benzanthrone
Naphtho-(2,1-b)thianaphthene
Triphenyl phosphate, 5-hydroxyimidazole
*Benz(a)anthracene (5), naphthacene
1,2-dichloro-3,3,4,4-tetrafluorocyclobutene
Di-2-ethylhexylphthalate
Polar Acids
p-fluoroacetophenone?
1-methoxycarbonyl pyrrolizine?
2-i-pr-l,3-dioxolane?
Nonyl-6-naphthol?
Allopeucenin?
Methylene dichloride
3-cyclohexyleicosane
2,6-dimethyl-3-heptanol
2,4-dichlorobenzaldehyde
2-formyl-3,4-dihydropyran
2,6-dimethyl-2,5-heptadien-4-one
Methyl tridecanoate
Butyl benzyl phthalate
Di-2-et-hexyl phthalate
Diamylphthalate
Methyl-15-ethylheptadecanoate
Methyl-2,4-dichlorobenzoate
Dimethylphthalate
2,3-dihydro-2-methylbenzofuran
2,7-dimethylbenzo(b)thiophene
Methyl caproate?
Methyl-9-dodecenoate
Methyl palmitate
Methyl octadecanoate
Di-n-propyl phthalate
Di-n-octyl phthalate
* carcinogen
**cocarcinogen
-------
342 E DO D. PELLIZZARI ET AL.
the S-9 activation system; plate incorporation tests at con-
centrations of 0.01, 0.1, 1, and 10 mg/plate; all in dupli-
cate. Sample size requirement for the initial mutagenesis
testing is thus 178 mg. Repeat studies over narrower concen-
tration ranges, taking into account positive results in the
initial test, would require additional sample. Finally,
those compounds or samples producing positive results should
undergo testing to determine dose response curves.
With the samples received to date, sample size has been
a critical limiting factor. Typically, one to four grams of
test material were obtained with 90-95% being inorganic, thus
only 50-400 mg of material have been available for further
fractionation, identification, and mutagenesis testing. The
small amounts available after fractionation are shown in the
first column of Tables 3-5.
Tables 3-5 show the results obtained in spot and/or pour
plate tests of fractions obtained with the fractionation
scheme initially used (Figure 1), the amount tested being
dependent on the amount available. Positive results were
obtained in most Level 1 fractions (Table 3). On further
fractionation, activity was seen with both polar acids and
polar bases (Table 4). Some activity was seen with nonpolar
organics (Table 5) but interpretation is difficult due to
small test samples. In a few cases, dose response "curves"
were attempted. It must be stressed here that where sample
size was small, a negative result cannot be interpreted as
having much significance. There is no way to determine where
we are on a dose response curve, and if only spot tests can
be done we do not know if the samples have components which
cannot be evaluated in the spot test [for example, benzo-
(a)pyrene].
Approaches to Resolving the Problem of Sample Size
To resolve the problem of limited sample size, a number
of options have been considered:
• Develop a tighter fractionation scheme with less
fractions, as described above.
• Develop a modified system of assay that will give
move information per unit of sample (11). (See
Figure 3.)
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES
343
Induced microsomes
+ agar
Uninduced microsomes
+ agar
Sample + agar
PROCEDURE:
1.
2.
3.
4.
Prepare base layer.
Cut wells into base layer.
Fill wells as indicated on diagram.
Overlay with agar containing microorganisms.
TEST RESULTS:
If toxic, clear zone around B. If not toxic,
uniform lawn.
If sample requires activation, line of cells between
A + B, weaker line between B + C.
If no activation is required, ring of cells around B.
ADVANTAGES:
Figure 3.
results.
Allows determination of induction requirement in-
duced and uninduced, mutagenicity, and toxicity on
one plate instead of separate plates for each case.
Do not have to worry about take-up of compound by
filter or run-off of compound onto plate.
Use of wells instead of spot test to optimize test
-------
344
EDO D. PELLIZZARI ET AL.
Table 3
Ames Testing of Samples from Level 1 Fractionation
of Particulates (WV, Sample 2308)
Fraction
Description
Chemical Type
+
Total Sample
Amount
Results,
Avg. No. Colonies/Plate
Dose No Induced Mutagenic
ug/plate Strain Microsomes Microsomes Ratio
A. SPOT TESTS
T. Nonpolar Organics
(23.6 rag)
790
790
790
790
790
98
100
1535
1537
1538
19*
0
2
5
3
TNTC*
0
2
7
6
—NOT TESTED—
Polar Organics
(~0.78 mg)
Polar Organics
(7.5 mg)
26
26
26
26
26
250
250
250
250
250
98
100
1535
1537
1538
98
100
1535
1537
1538
12*
0
5
2
4
Polar Organics
(12.0 rag)
(portion insoluble
in methanol)
T,, Polar Organics
(12.9 mg)
400
400
400
400
400
430
430
430
430
430
98
100
1535
1537
1538
98
100
1535
1537
1538
0
TNTC**
2
5
1
25*
0
3
9
28*
0
0
2
7
1
15
0
6
6
16*
B. POUR PLATE TESTS OF POSITIVE FRACTIONS
T1 Nonpolar Organics 790 98
(23.6 mg)
T., Polar Oragnics 250 98
(7.5 mg)
T_ Polar Organics 400 1535
Ja (12.0 mg) 100
(portion insoluble
in methanol)
T_ Polar Organics 430 98
(12.9 mg)
89*
63
43
46
106*
1.9
1.3
1.3
0.9
2.3
* Postive Results.
**TNTC = Too numerous to count.
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES
345
Table 4
Ames Testing of Samples from Level 2 Fractionation
(Polar Organics) of Particulates
(WV, Sample 2308)
Description
Chemical Type
•f
Total Sample Dose
Fraction Amount ug/plate
A. SPOT TESTS
T4 Polar acids 33
and neutrals
(1.0 mg)
T_ ** Polar bases 30
Da (0.9 mg)
T ** Polar bases 1440
DD (43.2 mg)
Tg Polar neutrals 70
T_ Polar acids 50
(1.5 mg)
T_, Polar acids 300
(9.0 mg)
Results,
Avg. No. Colonies/Plate
Strain
98
100
1535
1537
1538
98
100
1535
1537
1538
98
100
1535
1537
1538
98
100
1535
1537
1538
98
100
1535
1537
1538
98
100
1535
1537
1538
No
Microsomes
0
0
7
5
3
0
0
1
7
1
35*
0
2
16*
34*
5
0
3
4
2
0
0
2
5
3
TNTC*
TNTC*
2
TNTC*
TNTC*
Induced
Microsomes
0
0
2
4
2
0
0
1
3
1
28*
0
5
7
5
0
0
2
5
3
0
0
3
6
2
TNTC*
TNTC*
TNTC*
TNTC*
TNTC*
Mutagenic
Ratio
B. POUR PLATE TESTS OF POSITIVE FRACTIONS
T_h** Polar bases 1440
(43.2 mg)
T7, Polar acids 300
(9.0 mg)
98
1535
100
98
154*
36
47
70
3.3
1.0
0.7
1.7
* Positive results.
TNTC = Too numerous to count.
**5a and 5b are replicates, differing only in amount of starting material.
-------
346
EDO D. PELLIZZARI ET AL.
Table 5
Ames Testing of Samples from Level 3 Fractionation
(Nonpolar Organics) of Particulates
(WV, Sample 2308)
Description
Chemical Type
+
Total Sample Dose
Fraction Amount ug/plate
Results,
Avg. No. Colonies/Plate
No
Strain Microsomes
Induced Mutagenic
Microsomes Ratio
A. SPOT TESTS
T8
T8
O
T9
T10
J. W
Tll
T12
T13
Nonpolar acids and
neutrals (in DMSO)
Nonpolar acids and 6
neutrals (0.2 mg)
Nonpolar bases 20
(0.6 mg)
Nonpolar neutrals 13
(0.4 mg)
Nonpolar acids 580
(17.4 mg)
Paraffins 56
(1.7 mg)
Aromatics 43
(1.3 rag)
98
100
1535
1537
1538
98
100
1535
1537
1538
98
100
1535
1537
1538
98
100
1535
1537
1538
98
100
1535
1537
1538
98
100
1535
1537
1538
98
I'OO
1535
1537
1538
0
0
2
9
0
0
0
3
3
1
0
0
3
3
0
0
0
3
9
3
0
0
3
4
0
0
0
4
5
2
0
0
3
4
2
•0
0
TNTC*
6
0
0
0
3
6
2
0
0
2
4
0
0
0
5
4
2
0
0
3
4
1
0
6
4
5
1
0
TNTC*
2
6
4
B. POOR PLATE TESTS OF POSITIVE FRACTIONS
Nonpolar acids and
neutrals (in DMSO)
T,„ Aromatics
(1.3 mg)
43
1535
100
1535
100
32
47
18
35
1.0
0.9
0.7
0.5
* Positive results
**TNTC = Too numerous to count.
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES 347
• Develop a scheme of priorities to be used with
amounts of sample which do not allow complete test-
ing, in order to maximize the amount of information
that can be derived (Table 6).
• Request a larger sample.
Although the last option would appear to be the simplest
and most obvious approach, technical difficulties are encoun-
tered in operation of the sampling system if longer sampling
periods are used. Therefore, the other options have been
adopted.
Interfacting Chemical and Biological Testing with Vapors and
Gases
To date, only preliminary experiments have been conducted
on approaches to quantitatively collect and present to the
bioassay vapors and gases sorbed to Tenax and carbon. The col-
lection system includes a Tenax GC cartridge backed up by a
carbon cartridge to collect highly volatile materials that
break through the Tenax. In the laboratory, Tenax-sorbed com-
pounds may be transferred onto carbon with the assumption that
the carbon can then be incorporated into the Ames assay system
along with sufficient solvent to release the vapors into the
plate.
Initial tests with model compounds, in which we are not
limited by the availability of sample, indicate certain prob-
1 ems:
• Failure of the carbon to release some sorbed com-
pounds at concentrations of solvent tolerated in
the Ames test system.
• Where release is possible, assurance that the
desorbed materials have the opportunity to contact
the organisms before loss to the headspace in the
plate.
Also, we must assure that with complex environmental
samples, our method of presentation does not selectively
release some of the sorbed materials, essentially "narrowing
the window," by what in effect amounts to an additional
fractioning or partitioning.
-------
348 EDO D. PELLIZZARI ET AL.
Table 6
Developing Priorities for Mutagenesis Procedures
To get the maximum information with limited quantities
of test fractions, the following order of priorities is
suggested, the most important being listed first:
(1) Minimum of four strains with the well test and with
induced microsomes.
(2) Minimum of two doses (100 and 500 yg/plate), four
strains, well test with induced microsomes.
(3) As in 1, but with 5 strains.
(4) As in 1, but with 5 strains and non-induced as well
as induced microsomes.
(5) As in 4, but with 2 doses.
(6) Pour plate tests with >1 strain using induced micro-
somes. If plate is positive, repeat with no micro-
somes.
(7) As in 6, but with 2 or more test sample concentra-
tions.
Strains in order of decreasing importance are:
98, 1535, 1537, 1538, 100.
In all cases tests will be conducted in duplicate.
Depending on the amount of sample available, as many as
possible of the tests will be conducted.
Only one concentration of the sample fraction will be
made for these tests. To obtain lower test concentrations,
less sample volume per plate will be used. Pour plate tests
will be conducted at concentrations twice those of the well
tests, i.e., at 200 and 1000 ug/plate.
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES 349
It might appear immediately obvious that we should con-
sider suspension tests. Liquid suspension tests have been
conventionally performed under conditions where sample size
has not been limiting. Previous investigators have tested
gaseous compounds for mutagenicity in both plate incorpora-
tion and liquid suspension tests, as recently reviewed by
Malaveille and co-workers (14) in France. They examined the
failure of liquid suspension tests to detect mutagenicity of
certain compounds which we're easily detected in soft agar
plate incorporation tests. In tests with vinylidine chloride,
they found that microsomes maintained their viability for up
to 9 hr in soft agar, contrasted to about an hour in liquid
suspension. Hence, for mass screening it would appear that
a plate incorporation test would be preferable. Malaveille
et al. (14) has devised a method for exposing soft agar
plates to a test gas using a desiccator. However, this pro-
cedure involves a substantial quantity 2% (v/v), of the test
material. Methods of this kind and novel techniques requir-
ing small quantities of sample are currently under investi-
gation.
SUMMARY
The goal of the research described in this paper is to
adapt the Ames assay to mass screening, qualitatively and
quantitatively, of both particulate and volatile components
of complex air samples.
At present, problems are encountered due to
• Technical limitations on the amount of sample which
can be obtained initially, thus limiting the amount
of sample available for biotesting and pushing the
assay to its limits of sensitivity.
• Difficulties in presenting the "entire sample" to
the assay organisms without loss of some components
or introduction of artifacts.
These problems have been addressed
• By development of a better chemical fractionation
scheme, producing a smaller number of fractions
with more material in each fraction.
-------
350 EDO D. PELLIZZARI ET AL.
• By developing a list of priorities in mutagenesis
testing (strains, activation requirements, etc.)
so that the maximum information can be obtained
from a small sample.
• By developing a well test modification of the Ames
test which allows testing of a number of parameters
on a single plate.
The combined chemical fractionation/mutagenesis screening
approach allows rapid identification of those fractions with
mutagenic activity, i.e., those that should receive priority
for further chemical identification.
REFERENCES
1. Ames BN: A bacterial system for detecting mutagens and
carcinogens. In: Mutagenic Effects of Environmental
Contaminants (Sutton HE, Harris JF, eds.). New York,
Academic Press, 1972
2. Ames BN, Durston WE, Yamasaki E, Lee FD: Carcinogens
are mutagens: A simple test system combining liver
homogenates for activation and bacteria for detection.
Proc Natl Acad Sci USA 70:2281-2285, 1973
3. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella mammalian
microsome mutagenicity test. Mutat Res 31:347-364, 1974
4. Brown VM: The prediction of the acute toxicity of river
waters to fish. Proc Fourth Brit Coarse Fish Conf, Liver-
pool Univ, 1969
5. Dipple A: Polynuclear aromatic carcinogens. In: Chemi-
cal Carcinogenesis, ACS Monograph 173 (Searle CE, ed.).
Washington, DC, American Chemical Society, 1976, pp 245-
314.
6. Duke KM, Davis ME, Dennis AJ: IERL-RTP Procedures
Manual: Level 1 Environmental Assessment Biological
Tests for Pilot Studies. EPA-600/7-77-043. Washington,
DC, US Environmental Protection Agency, 1977
7. Flessel CP: Mutagenic activity of particulate matter
in California hi-vol samples. Berkeley, CA, Third Inter-
national Symp Air Monitoring Quality Assurance, May 18-19,
1977
-------
INTEGRATING TESTING INTO SCREENING OF AIR SAMPLES 351
8. Hammersma JW, Reynolds SL, Maddalone RF: IERL-RTP Proce-
dures Manual: Level 1 Environmental Assessment, EPA-600/
2-76-160a. Washington, DC, US Govt Printing Office, 1976
9. Hill HH, Chan KW, Karasek FW: Extraction of organic
compounds from airborne particulate matter. J Chromat
131:245-252, 1977
10. Hoffman D, Wynder EL: Environmental respiratory carcino-
genesis. In: Chemical Carcinogenesis, ACS Monograph
173 (Searle CE, ed.). Washington, DC, American Chemical
Society, 1976, pp 324-365
11. Hughes TJ, Little L, Pellizzari E, Sparacino C, Claxton
L, Waters M: Application of agar diffusion wells for
microbial mutagenesis testing of air pollutants.
Presented at Environmental Mutagen Society meeting,
poster session, March 11, 1978
12. McCann J, Ames BN: Detection of carcinogens as mutagens
in the Salmonella/microsome test: assay of 300 chemicals,
part II. Proc Natl Acad Sci USA 73:950-954, 1976
13. McCann J, Choi E, Yamasaki E, Ames BN: Detection of
carcinogens as mutagens in the Salmonella/microsome
test: assay of 300 chemicals. Proc Natl Acad Sci USA
72:5135-5139, 1975
14. Malaveille C, Planche G, Bartsch H: Factors for effi-
ciency of the Salmonella/microsome mutagenicity assay.
Chem-Biol Interactions 17:129-136, 1977
15. Pitts JN Jr, Grosjean D, Mischke TM, Simon VF, Poole D:
Mutagenic activity of airborne particulate organic pol-
lutants. Toxicol Letters 1:65-70, 1977
16. Talcott R, Wei E: Brief communication: airborne muta-
gens bioassayed in Salmonella typhimurium. J Natl Can-
cer Inst 58:499-451, 1977
17. Van Duuren BL: Tumor-promoting and cocarcinogenic
agents in chemical carcinogenesis. In: Chemical Car-
cinogenesis, ACS Monograph 173 (Searle CE, ed.). Wash-
ington, DC, American Chemical Society, 1976, pp 24-51
18. Weisburger JH: Bioassays and tests for chemical carcin-
ogens. In: Chemical Carcinogenesis, ACS Monograph 173
(Searle CE, ed.). Washington, DC, American Chemical
Society, 1^76, pp 1-23
-------
CHEMICAL AND
MICROBIOLOGICAL STUDIES
OF MUTAGENIC POLLUTANTS
IN REAL AND SIMULATED
ATMOSPHERES
James N. Pitts, Jr., Karel A. Van
Cauwenberghe, Daniel Grosjean,
Joachim P. Schmid, and Dennis R. Fitz
Department of Chemistry and Statewide
Air Pollution Research Center
University of California
Riverside, California
William L. Belser, Jr., Gregory B. Knudson,
and Paul M. Hynds
Department of Biology and Statewide
Air Pollution Research Center
University of California
Riverside, California
-------
355
INTRODUCTION
In the early 1940's the organic extracts of ambient par-
ticulate matter (POM) collected from urban air in the United
States were found to be carcinogenic when administered sub-
cutaneously to mice (1,2). Subsequently, this effect was
also observed in experimental animals injected with extracts
of ambient POM collected from Los Angeles photochemical smog
(3) and in seven other U.S. cities (4). Similar results now
have been found with ambient samples collected in various
urban centers throughout the world. This carcinogenicity is
customarily attributed to certain polycyclic aromatic hydro-
carbons (PAH), e.g. benzo(a)pyrene (BaP), benz(a)anthracene
and aza-arenes such as benzocarbazoles in the neutral frac-
tion of the organic particulates and benzacridines and
dibenzacridines in the basic fraction.
Several researchers, however, have found significant
discrepancies between the observed biological activity of
POM and the amounts of carcinogenic polycyclics determined
to be present. This is true not only of samples of ambient
particulate matter but also from the exhaust from spark-
ignition engines in light duty motor vehicles (3-10).
Thus, for example, Gordon and co-workers reported that, with
airborne particles collected in the Los Angeles area, the
benzene extract had 100 to 1000 times the cell transformation
-------
356 JAMES N. PITTS ET AL.
activity that could be attributed to its known BaP content.
Furthermore, the methanol extract, while containing only
about 1/30 of the BaP in the total sample, showed an activity
comparable to the benzene extract (10).
Mohr et al. recently showed that auto exhaust had a pro-
nounced carcinogenic effect on the lungs of Syrian golden
hamsters (100% rate of multiple pulmonary tumors). The
authors point out that "considering the relatively low total
dose of BaP contained in the condensate, this pronounced neo-
plastic response cannot be explained alone by the effects of
this well known carcinogenic hydrocarbon" (8).
We, among others, have felt for some time the need to
try to identify the chemical species responsible for this
" excess carcinogenicity" if one is to obtain, a reliable esti-
mate of the health impact on man of POM from whatever source
—ambient air, auto or diesel exhaust, or fly ash from coal
fired power plants. In short, one must fully characterize
the dose parameters in dose-response curves. The problem
is, for atmospheric particulates, characterization of the
"dose" requires a detailed knowledge of the physical and
chemical nature of the species present in POM at the site
of impact upon the biological target. Complications arise
because primary organic pollutants may, and do, undergo a
variety of chemical transformations in the presence of light,
oxygen, water, and a variety of copollutants. Thus, ambient
POM from polluted urban atmospheres is a highly complex mix-
ture consisting of hundreds, probably thousands, of different
compounds.
Because of this complexity and also because of the costs
and time involved in animal tests for suspected carcinogens,
to date results from experiments directed to identify the
chemical structures of the compounds responsible for this
"excess carcinogenicity" have been relatively limited.
Therefore, we were most interested in applying to this prob-
lem the relatively inexpensive microbiological assay for
mutagenic activity developed by Ames and coworkers (11-13)
for fast screening of compounds for potential carcinogenic
activity. This assay is a reverse mutation system employing
histidine-requiring mutants of the bacterium Salmonella
typhimurium. It is now generally recognized as a useful,
though by no means exclusive, screening test for chemical
mutagens in complex environmental samples.
Specifically, we have been using the Ames test to screen
POM samples collected from real and simulated atmospheres for
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES
357
mutagenic activity. Additionally, we have been using results
of this test to provide microbiological clues to the chemical
nature of the compounds responsible for the "excess" carcino-
genicity, prior to their final characterization. Such clues
include:
• The type of mutation induced. Thus, many frame-
shift mutagens detected by strains TA1537, TA1538,
and TA98 are planar molecules, capable of interca-
lation between bases of the DNA strand (e.g., 9-
aminoacridine); on the other hand, many base pair
substitution mutagens, detected by strain TA1535,
are alkylating agents (e.g., B-propiolactone).
• The distinction between compounds which are
directly mutagenic and those which are promutagens
requiring metabolic activation. Thus, BaP is a
promutagen requiring treatment with S-9 liver
homogenate, whereas its metabolite, 6-hydroxy-BaP,
does not.
• The position of substitution in the compound will
have a pronounced effect on the observed mutageni-
city. For example, of the 12 hydroxy-isomers of
BaP, only five phenols are directly acting frame-
shift mutagens. The 6- and 12-OH-BaP have strong
activities; the 1-, 3-, and 7-OH-BaP are also muta-
genic, but much weaker. The remaining seven iso-
mers are nonmutagenic (14-16).
Soon after initiating our combined chemical-microbio-
logical experiments on POM (17,18), the need for standardi-
zation of the Ames test, including the number of cells per
plate, plate volume, concentrations of rat liver S-9 homoge-
nate, etc., became increasingly apparent. Therefore, along
with the development of our HPLC separation and GC-MS iden-
tification procedures, we conducted a series of experiments
designed to give us a better understanding of the effects
of certain variables on its reproducibility. Thus, in this
paper we shall first deal with the chemical aspects of the
problem and then discuss some factors involved in the intra-
and interlaboratory standardization of this microbiological
assay.
-------
358 JAMES N. PITTS ET AL.
MUTAGENIC ACTIVITY OF AMBIENT PARTICULATE MATTER
In 1975 we first reported the mutagenicity of the
organic extracts from ambient particles collected at several
sites in the Los Angeles Basin, as detected by the Ames
assay system (19). This phenomenon has now been reported in
studies at Ohmura and Fukuoka, Japan (20); Kobe, Japan (21);
Buffalo, New York, and Berkeley, California (22); New York
City, New York (23); and Chicago, Illinois (24).
More recently, we collected samples of airborne partic-
ulates at 11 urban sites in California's South Coast Air
Basin (25,26). Using strains TA1537, TA1538, and TA98, all
samples exhibited direct frameshift-type mutagenic activity,
i.e., they did not require metabolic activation. Addition
of the microsomal activation system (S-9 solution) did not
significantly increase the activity of the majority of the
samples tested. No activity was observed in any of the
assays with strain TA1535, which, as noted earlier, is rever-
ted by base pair substitution mutations.
Finally, in a size-resolved sample collected in down-
town Los Angeles using a Sierra Hi-Vol cascade impactor, all
mutagenic activity was found to be associated with the parti-
cles of diameter 1.1 micron or less. This is consistent with
the well documented occurrence of PAH such as BaP in the
respirable range of ambient particulates (27-30).
From these data we concluded that urban POM must contain
direct mutagens in addition to carcinogenic PAH, such as BaP,
which require metabolic activation. This is consistent with
the numerous observations of "excess" carcinogenicity in ani-
mals or in cell transformation activity, as discussed above,
and with the low average concentrations of BaP measured in
the Los Angeles Basin (31).
We then formulated the hypothesis that some of these
direct mutagens in ambient particulates might be formed in
the atmospheric transformations of particulate BaP and other
PAH by gaseous species such as ozone, nitrogen dioxide,
peroxyacetyl nitrate (PAN), singlet molecular oxygen (O
and free radicals present in photochemical smog such as OH
and H02 (17,18,26) .
Support for this idea can be found in several reports of
the carcinogenic activity of polar fractions of organic par-
ticulates (4,5,10,32,33), of products of ozonized gasolines
(34), of products of oxidation of aliphatic hydrocarbons (35),
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 359
and of the toxicity of the photooxidation products of a com-
mercial fuel oil (36). Furthermore, it is known that some
of the oxygenated metabolites formed from BaP in mammalian
cells are directly mutagenic (14-16) and given the oxidizing
potential of photochemical smog, it seemed reasonable that
at least some analogous transformations might occur in pol-
luted air.
One complicating factor we faced initially was a direct
conflict in the literature over the chemical reactivity of
PAH in ambient particulates. Thus, two references stated
that "they are chemically inert and thus are removed from
the air only by rain or the slow sedimentation of the par-
ticulate" (37,38). However, we found this position diffi-
cult to reconcile with earlier literature data on the photo-
chemical transformations of PAH adsorbed on a variety of
support materials such as filters, silica gel, and carbon
(soot) particles (3,39-43). These data show that certain
key PAH (e.g., BaP) can be quite reactive. As discussed
below, our experiments fully support the latter observations,
FORMATION OF MUTAGENIC POLLUTANTS FROM PAH IN REAL AND
SIMULATED ATMOSPHERES
In order to test our hypothesis, experiments were
carried out in which several PAH, deposited on glass fiber
filters, were exposed to gaseous pollutants both in real and
in simulated atmospheres. We shall briefly summarize the
results; details are presented in two papers (17,18).
Exposure of Benzo(a)pyrene to Ambient Photochemical Smog
In this series of experiments, BaP was exposed to the
gases present in ambient photochemical smog and the muta-
genicity of the resulting products determined. The experi-
mental set-up consisted of two conventional washed and fired
Gelman A/E glass fiber filters mounted in series in a high
volume sampler. The upstream filter, a "blank," collected
all ambient particulates and allowed only the gaseous pol-
lutants present in photochemical smog to pass through to the
second filter. The latter was coated with BaP (~2 mg) that
could interact with the gaseous pollutants.
After several days of exposure, the BaP coated filters
were extracted by ultrasonication and the concentrated
organic extracts tested with Salmonella strains TA98 and
-------
360 JAMES N. PITTS ET AL.
TA100. They now showed direct mutagenicity indicating the
formation of products from the promutagen BaP.
In another experiment, the second BaP coated filter
was extracted and further separated into fractions by TLC on
silica gel plates. Each band was recovered in methanol and
analyzed by methane chemical ionization mass spectrometry.
Some of the tentatively identified products were, in order
of decreasing polarity: BaP-dihydrodiol(s) (mol. wt. 286),
BaP-diphenol(s) (mol. wt. 284), BaP-phenol(s) (mol. wt. 268),
and BaP-quinones (mol. wt. 282) (17,18).
Exposure of Benzo(a)pyrene to 0} and Peroxyacetyl Nitrate
In order to obtain specific information about the reac-
tions with BaP of certain key single pollutants present in
photochemical smog, another series of experiments was car-
ried out under conditions similar to those used with ambient
smog. In these, glass fiber filters coated as above with
BaP were exposed in the dark to clean, particle-free air
containing 11 ppm 03 (exposure time 24 hours at a flow rate
of 3 cfm), or 1.1 ppm PAN (16 hours, 3 cfm). Control runs
with BaP exposed to pure air (24 hours, 3 cfm) and with
blank filters exposed to 03 or PAN were also included. No
mutagenicity was observed in any of the control runs, but
all the other exposures produced direct mutagenicity.
After exposure, the products and unreacted BaP were
separated by TLC, and the major bands analyzed by mass spec-
trometry. They were also tested separately for mutagenic
activity, both with and without metabolic activation. For
comparison purposes, a sample of BaP was also incubated for
30 minutes at 37°C with the liver S-9 homogenate solution
and the metabolites formed in this microsomal activation
system analyzed and tested for mutagenic activity.
We found that BaP reacted readily with these 1-10 ppm
levels of 03 and PAN in air to form a variety of oxygenated
products. As expected, the TLC bands containing the unreac-
ted BaP were not directly active and required metabolic ac-
tivation. The band containing the BaP-quinones was complex
and contained, in addition to the inactive quinones (16), a
directly active compound of molecular weight 284, to date
still unidentified.
Products of the treatment of BaP with the S-9 mix
appeared as a series of TLC bands, one of which was complex.
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 361
This complex band was also seen with ambient smog but not
with 03 or PAN. It contained directly active mutagens. The
Rf-values and the molecular weight of its components (mol.
wt. 268) were consistent with those of isomers of hydroxy-
benzo(a)pyrene. Positive identification will require com-
parison with the proper reference compounds.
Exposures of Benzo(a)pyrene and Perylene to N02
Exposure of BaP to 1.3 ppm of N02 (24 hours, 1 cfm),
containing traces of nitric acid (~10 ppb) resulted in the
appearance of only one major TLC band. This contained a
directly active mutagen whose Rf and molecular weight of 297
were consistent with the nitrobenzo(a)pyrene (nitro-BaP)
structure.
This band containing nitro-BaP was further resolved
into two bands, one yellow and one orange, using TLC with
toluene as the solvent. Comparison of the mass spectra and
ultraviolet-visible spectra with those of authentic samples
synthesized according to Dewar (44) , allowed us to assign
the structure 6-nitro-BaP to the component present in the
yellow TLC band; the orange TLC band was a mixture of the
1-nitro and 3-nitro isomers.
In more recent studies, we have shown that the nitra-
tion of BaP by ppm levels of N02 in air is acid catalyzed by
ppb levels of HN03. Furthermore, we obtained yields of -~18%
of nitro-derivatives from eight-hour exposures of BaP to
only 0.25 ppm N02 (containing ~3 ppb HN03) in air. The
value of 0.25 ppm is the air quality standard for N02 (one
hour average) in California; during the late fall and winter
months it is commonly exceeded in downtown Los Angeles,
Pasadena and the coastal regions of the South Coast Air
Basin.
Since exposure of BaP, a known carcinogen and activata-
ble mutagen, to ppm and sub-ppm levels of N02 resulted in
the formation of directly mutagenic nitro-derivatives (see
discussion below), it seemed interesting to see if, under
the same conditions, similar products could also be formed
from a "noncarcinogenic" PAH, perylene (39,45). This isomer
of BaP is also present in ambient POM and in POM emissions
from a variety of combustion sources.
Thus, perylene, deposited on glass fiber filters as with
with BaP, was exposed to 1 ppm of N02 for 24 hours at a flow
-------
362 JAMES N. PITTS ET AL.
rate of 1 cfm. The major resulting TLC band (brick-red
color on silica gel) consisted of 3-nitro-perylene, identi-
fied by its mass spectrum and by comparison of its UV spec-
trum with literature data (46,47).
Figure 1 presents the mutagen test data on the parent
BaP, the filter derived nitro-BaP isomers and authentic
nitro-BaP isomers synthesized by the Dewar method. Panel A
shows a standard dose-response curve for BaP. Without acti-
vation we observed ~4 revertants/nmole; with S-9 activation
we observed ~120 revertants/nmole. The latter figure is in
good agreement with data of Ames for BaP (48).
Panel B shows results comparing the polluted air-filter
generated 6-nitro-BaP with the authentic laboratory prepared
6-nitro-BaP. In the absence of metabolic activation, this
isomer is more than six times as active as the parent BaP,
giving values of ~20 revertants/nmole for both the filter
generated and authentic 6-nitro-BaP. With metabolic activa-
tion, this isomer is more than three times as mutagenic as
the parent BaP, giving values of ~390 revertants/nmole and
~420 revertants/nmole for the filter generated and authentic
6-nitro-BaP, respectively.
Panel C shows the test data for the mixture of 1- and
3-nitro-BaP isomers. Note that the range of concentrations
used is lower than in the other tests, since we had only a
limited quantity of the filter generated material available.
Clearly this mixture of isomers is highly active. Thus, in
the absence of metabolic activation, the 1- and 3-isomer mix
yields approximately 40 times as many reyertants as BaP
itself and approximately five times as many revertants as
the 6-nitro-BaP. The values are ~140 revertants/nmole and
~210 revertants/nmole for the filter generated and authentic
1- and 3-nitro-BaP mix, respectively. Activation of the 1-
and 3-isomers with S-9 gave values of ~3500 rev/nmole and
~5200 rev/nmole for the filter generated and authentic sam-
ples, respectively.
The shape of the dose response curve in this case is a
puzzle. Maximum mutagenesis occurs at very low doses, fol-
lowed by some inhibition. The curve begins to rise again
at higher concentrations. At the moment we have no explana-
tion for this, although our intuition is that this may be a
function of the relative abundance of the two isomers in the
mixture. Resolution of this issue will depend upon separat-
ing the 1- and 3-isomers and subjecting them to individual
tests.
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES
363
r
(-
1200-
)
K
z
<
K
cr
UJ
UJ
cc
300-
1200-
900-
300-
-S9
10
10
+S9
I20O-
300
PERYLENE
23 15
NANOMOLES OF COMPOUND TESTED
Figure 1. Dose-response curves obtained for BaP (A), 6-
nitro-BaP (B), 1-nitro- and 3-nitro-BaP (C), and 3-nitro-
perylene (D) in the Ames mutagenicity test with the most
sensitive strain, TA100 (A,B), TA98 (C), and TA1538 (D) with
(+S-9) and without (—S-9) metabolic activation. Each point
represents the mean of at least triplicate plates. Sub-
tracted spontaneous revertant backgrounds were: TA100:118-
130 (-S-9), 130-152 (+S-9); TA98:24 (-S-9), 44 (+S-9);
TA1538-.12 (-S-9), 41 (+S-9) . Also shown (full circles) are
the corresponding curves for perylene (D) and for the nitro-
BaP isomers (B,C) synthesized according to Dewar (44).
-------
364 JAMES N. PITTS ET AL.
These data clearly demonstrate that nitration of the
promutagen BaP produces a direct mutagen and additionally
increases its mutagenic potency on S-9 activation. Further-
more, and not unexpectedly, the position of substitution in
the compound (6 versus 1 or 3) has a pronounced effect on
its mutagenic activity.
As shown in Panel D of Figure 1, with the Ames rever-
sion assay, perylene itself proved to be nonmutagenic, at
low levels, either in the presence or absence of rat liver
homogenate.* However, the 3-nitro-isomer proved to be muta-
genic in this test, giving ~40 rev/nmole without metabolic
activation and ~100 rev/nmole with metabolic activiation.
Thus, addition of a nitro-group in the 3-position converts
perylene into a direct mutagen. Furthermore, on activation
the 3-nitro-perylene is almost as potent as BaP.
SOME FACTORS AFFECTING THE REPRODUCIBILITY OF THE AMES TEST
Obviously, an increasing number of laboratories are
beginning to apply the Ames Salmonella reversion test to
mutagenicity studies of environmental samples (11-13). In
surveying the literature, and as a consequence of our expe-
rience in conducting the Ames test, we found there is a
pressing need for standardization of the procedures employed
in this assay for mutagenicity. Thus we present here some
data from our laboratory that may provide some basis for our
plea for the establishment of a set of standard conditions
for application of the Ames test.
In working with the standard set of Ames tester strains,
TA1535, TA1537, TA1538, TA98, and TA100, we have observed,
as have others, that there is strain specificity in the
response to different mutagens. Thus TA1537 may give a high
number of revertants with a given sample, while TA1538 and
TA98 are low. Hence, in working with environmentally derived
samples of unknown composition, we feel it is essential to
include all five tester strains in the preliminary screen.
In addition, although TA100 is widely recognized to be
strongly responsive to many frameshift mutagens, this tester
strain is often reported as specifically detecting base pair
substitution mutagens. Within its limitations TA100 is a
versatile strain, the fact of which we should not lose sight.
*Perylene has since been reported to be mutagenic in a for-
ward mutation assay (49).
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 365
To satisfy ourselves on the reproducibility of the Ames
test, we have examined the effect of growth media, cell den-
sity, agar plate volume, and S-9 concentration with the fol-
lowing results.
Effect of Growth
We have grown our tester cultures in a variety of media,
including L-broth, nutrient broth, and Vogel and Bonner
enriched medium with glucose as a carbon source. Except for
slight variations in cell numbers, there is no observable
effect in response to mutagens with different growth regimes.
Effect of Cell Density
We have conducted experiments on the effect of cell den-
sity on mutation frequency with interesting results. Using
four of the Ames tester strains, we tested a single frame-
shift mutagen (hycanthone), and for the fifth tester strain
(TA1535) we used N-methyl-N'-nitro-N-nitrosoguanidine (NTG) .
We grew overnight cultures of each of the five strains to
approximately the same turbidities. These were assayed by
dilution and plate count. Each culture was tested undiluted
and at 1:2, 1:4, 1:8, and 1:16 dilutions. The undiluted cul-
ture was tested at 0.2 and 0.1 ml per plate and the diluted
culture was added at 0.1 ml per plate. Hycanthone was added
at 15, 30, and 50 yg/plate.
The highest numbers of revertants were obtained at 2
x 10* cells per plate, with slightly fewer at 1 x 10" cells
per plate. Over a range from 5 x 107 cells down to 6 x 106
cells per plate, the numbers of revertants were about 40%
lower than the value at 1 x 10* cells per plate, but with
no significant variations within this range.
Competition for the trace of histidine in the top agar
is a prime factor in explaining these results. With only
enough for two or three rounds of replication, adding too
many cells will lead to the rapid exhaustion of histidine
and hence less opportunity for mutation to occur. Similarly,
lower numbers of cells allow for more background growth and
can lead to an incorrect assessment of a compound's activity.
Based upon these observations, it appears that 1 x 108 cells
per plate is probably optimal. This is consistent with the
observation of Rosenkranz (50).
-------
366 JAMES N. PITTS ET AL.
Since the generation time of these organisms is less
than 30 minutes in rich medium and inoculation sizes vary,
we strongly recommend that the titer of overnight cultures
be adjusted by optical density measurements to about 1 x
10' cells/ml, and that 0.1 ml inocula delivered to the test
plates will then be close to the optimum cell density.
Effect of Agar Volume
Another variable in the experimental protocol is agar
volume in the plates. This was also mentioned in Dr. Rosen-
kranz' presentation (50). In a sample of unknown composi-
tion, some compounds are certain to be water soluble. Dif-
fusion caused by this solubility into different volumes of
agar in the base layer could provide considerable variabil-
ity in the dose of mutagen seen by the cells in the top
layer.
Until recently we had been hand pouring our base agar
layers. In preparation of large numbers of plates, consid-
erable variation in agar volume occurs. Therefore, we con-
ducted a controlled experiment in which we checked a random
selection of hand-poured plates against plates poured by the
Manostat automatic plate pourer. The machine was set to
pour plates containing 15, 20, 25, or 30 ml agar per plate.
We used two mutagens, hycanthone and 2-aminofluorene,
each at a single dose. For each mutagen we plated 30 repli-
cates on the hand-poured plates and 30 replicates for each
of the four volumes of the machine-poured plates. The tester
strain was TA98, 2-ajninofluorene was used at 0.5 jag/plate and
hycanthone was used at 50 ug/plate. In Table 1 are presented
the average number of revertants per plate.
From these data we conclude that the volume of the agar
layer can affect the dose of the mutagen seen by the cells.
At higher constant volumes, the number of revertants per
plate falls off.
These data were also subjected to statistical analysis
and the standard deviation and variance on the 30 replicates
of each sample from the hycanthone data calculated. The
variances of the constant volume samples were pooled and com-
pared to the variance in the hand-poured plate sample using
the F test. The variation among the•hand-poured plates was
significant at the p = 0.05 level. Thus, using variable vol-
ume hand-poured plates, in the smaller number of replicates
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 367
Table 1
Volume of Agar Per Plate
Hand
15 ml 20 ml 25 ml 30 ml Poured
Hycanthone 270 223 189 188 167
(Avg. rev/plate)*
2-aminofluorene 666 687 531 490 504
(Avg. rev/plate)
*Average of 30 replicates in each case.
normally used (three to five plates per sample) in quanti-
tative tests, could introduce substantial error into the
results.
Based upon these studies, we recommend that when pos-
sible constant volume plates containing 20 ml base agar be
used. Although the average number of revertants observed
on the 15 ml plates was higher in the hycanthone experiment,
we recommend use of 20 ml plates for other reasons. Specif-
ically, some of the 15 ml plates appeared to be drying out
during the 48-hour incubation period, and we see this as
introducing another potential problem in quantitation of
results.
S-9 Suppression and Optimal Concentrations
In a number of our experiments, we observed suppression
of reversion frequency when S-9 was added; therefore we set
out to examine the effect of S-9 concentration on reversion
frequency. Two different activatable mutagens were used,
2-aminofluorene and BaP. Each mutagen was tested at three
concentrations and S-9 was tested at four concentrations.
In this experiment, we used TA100 as the tester strain.
The S-9 liver homogenate was prepared according to Ames et
al. (12). Sprague-Dawley rats were given a single i.p.
injection of Aroclor 1254 at a dose of 500 mg/kg. The rats
were starved for 12 hours and sacrificed on the fifth day
past injection. The results of this experiment are shown
in Figures 2 and 3.
-------
368
JAMES N. PITTS ET AL.
60O -
550 -
O.OI 0.02 Q03 004 005 QO6 0.07 O.O8 009 0.1
AMOUNT OF S9 PER PLATE (ml)
Figure 2. Dose response curve of S-9 for benzo(a)pyrene
with Salmonella typhimurium strain TA100.
Examination of the curve for the low concentration of
BaP shows that optimum activation occurs at 0.01 ml of the
homogenate, and that increasing concentrations of S-9 sup-
press the appearance of revertants. At the highest concen-
tration of BaP optimum activation occurs at 0.05 ml S-9 per
plate with suppression at higher concentrations of S-9
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES
369
2/ig
0.01 0.02 0.03 Q04 0.05 0.06 0.07 0.08 0.09 0.1
AMOUNT OF S9 PER PLATE (ml)
Figure 3. Dose response curve of S-9 for 2-aminofluorene
with Salmonella typhimurium strain TA100.
(Figure 2). The optimum S-9 concentration for the inter-
mediate dose of benzo(a)pyrene (data not presented) was at
0.02 ml S-9 per plate with suppression at higher concentra-
tion of S-9. It is interesting that at the high and low
concentrations of 2-aminofluorene (Figure 3) the optimum
S-9 concentration for activation was identical and very low,
-------
370 JAMES N. PITTS ET AL.
Thus, when screening unknown samples for mutagenicity,
we recommend the use of two concentrations of S-9: a "low"
(0.01 ml/plate) and a "high" (0.05 ml/plate) concentration.
Thus, if a mutagen of the 2-aminofluorene type is present,
it will not be suppressed by high S-9 to give a false nega-
tive, and if a mutagen such as BaP is present, a much better
idea of the effective concentration range for further quan-
titation will be obtained. Before finally settling on the
exact concentrations of S-9 to use, we plan to test several
other activatable mutagens to confirm these findings.
Experimental Application
In our program, working with ambient air samples, we
are faced with a problem in application of the test. Dr.
Little (51), in her presentation, has emphasized this prob-
lem. The samples are often small and in low concentrations.
Because of this, we essentially have "one shot" at a test
effect. These samples are complex mixtures of chemicals and
may comprise several mutagens as well as several cytotoxic
compounds. Because of these properties of the samples, we
quickly rejected use of a spot test as a preliminary screen-
ing procedure. This rejection was on two grounds. First,
many of the mutagens are non-diffusable. Second, if a cyto-
toxic compound is present and diffuses, it would eliminate
the positive reversions in the vicinity of the spot. This
would lead to a high frequency of false negative test results.
This led us to develop a preliminary screening test
based upon the agar layer method. For this we use all five
tester strains. Each sample is tested at three concentra-
tions over a thousand-fold concentration range. We use two
levels of S-9 for activation: low = 0.01 ml and high =
0.05 ml S-9 per plate. We pour only a single plate for
each test. From this screen, we cannot draw any quantita-
tive conclusions, but we do establish a base line for the
quantitative test. We learn whether a sample is mutagenic
or not, what level of S-9 to use and what concentration
range of sample to use for the quantitative test.
From this assay we select the most responsive strain
or strains for the quantitative test.
For attempts to quantify mutagens, we use the most
sensitive strains, select the high or low S-9 concentration,
and test four concentrations of the sample based upon the
highest number of revertants found in the preliminary test.
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 371
These concentrations are selected to cover one log of con-
centration surrounding the optimum.
Each day that samples are tested, it is essential that
the full battery of controls be run. These provide the
basis of reproducibility. It is necessary to constantly
monitor the strains for presence of the plasmids and the
mutations. In each experiment, spontaneous reversion fre-
quencies must be tested. Furthermore, it is essential that
the mutant strains be tested against known mutagen standards
as an internal control on their response. It is desirable
to quantify this response each time to control inherent bio-
logical variability.
In summary, our experiments support the following
recommendations for standard application of the Ames test
to samples of unknown chemical composition.
1. Preliminary test
a. Grow overnight cultures of all strains.
b. Adjust cultures to fixed optical density
corresponding with 1 x 109 cells per ml.
c. Use 0.1 ml of this cell suspension for assay
plates.
d. Use 20 ml constant volume agar plates where
possible.
e. Use the 2.0 ml soft agar overlay method to
eliminate false negatives due to toxic chem-
icals.
f. Test appropriate sample dilutions (we use a
3-log concentration range).
g. Use two levels of S-9 (high and low) for
metabolic activation, to avoid suppression.
h. Include all five tester strains because of
strain specificity in response to mutagens.
i. Test five colonies from each "positive"
sample for true reversion to exclude drug
induced phenocopies.
-------
372 JAMES N. PITTS ET AL.
2. Quantitative tests
a. Select the most responsive strain or strains.
b. Use three or four replicates plated on con-
stant volume (20 ml) agar plates where pos-
sible.
c. Use either high or low concentration of S-9
as determined by the preliminary screen.
d. Hold cell densities constant at approximately
1 x 10* cells per plate.
As demonstrated in many of the papers of this proceed-
ings, assays of complex samples do not generate straight
line dose response curves. Dr. Commoner (52) has emphasized
the problem of making quantitative judgments from the com-
plex curves. Our position is that a peak value derived from
such a nonlinear dose-response curve at least provides a
minimal estimate of the mutagenicity of the sample. True
quantitation depends upon subfractionation of these samples
to isolate the mutagenic agents.
CONCLUSIONS
Directly active mutagens are formed upon exposure of
BaP of ambient photochemical smog as well as to sub-ppm levels
of several major gaseous components, NO 2, 03, and PAN (17,18).
However, we would like to emphasize that our studies
were conducted with PAH deposited on the surface of glass
fiber filters. Whether PAH adsorbed on the surface of air-
borne particles (soot, fly ash, etc.) will react in a simi-
lar fashion in the atmosphere is a complex problem. Thus,
the atmospheric reactions of PAH may be influenced by many
factors typical of surface chemistry as well as by pollutant
levels, particle size, sunlight intensity, atmospheric mix-
ing, and transport time. Similarly, little is known about
the extent of possible reactions of PAH on glass fiber fil-
ters widely employed for decades to collect ambient partic-
ulates; our results suggest that they may indeed be signif-
icant. Therefore, the determination of possible filter
"artifacts" is of major importance since historically most
evaluations of the carcinogenic and mutagenic activity of
organic particulates have been based upon filter samples.
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 373
Finally, control experiments on the Ames Salmonella
reversion test have resulted in a series of findings which
support a standardized protocol for application of the test
to ambient air samples and possibly to samples from other
sources. These include recommendations for control of cell
density, agar volume, S-9 concentration and strains used.
ACKNOWLEDGMENT
Much of this paper is based on "Atmospheric Reactions
of Polycyclic Aromatic Hydrocarbons: Facile Formation of
Mutagenic Nitro-Derivatives," Science (17), and "Photochemi-
cal and Biological Implications of the Atmospheric Reactions
of Amines and Benzo(a)pyrene," Philosophical Transactions of
the Royal Society of London, in press (18). These papers
should be consulted for details.
We want to thank Dr. T.M. Mischke and Dr. T.L. Gibson
of the Department of Chemistry, University of California,
Riverside, who were involved with the chemical aspects of
collection and analysis of the urban particulates, and Dr.
V.F. Simmon and Mr. D. Poole, Stanford Research Institute,
who kindly carried out the Ames tests during our initial
screening program of urban aerosols collected in the Los
Angeles Basin.
We also want to express our appreciation to the Univer-
sity of California and to the Federal agency who generously
funded this research—the National Science Foundation-
Research Applied to National Needs (Grant No. ENV73-02904-
A04, Dr. R. Carrigan, Project Officer).
The contents do not necessarily reflect the views and/
or policies of the NSF-RANN nor does mention of trade names
or commercial products constitute endorsement or recommenda-
tion for use.
REFERENCES
1. Leiter J, Shimkin MB, Shear MJ: Production of sub-
cutaneous sarcomas in mice with tars extracted from
atmospheric dusts. J Natl Cancer Inst 3:155-165, 1942
-------
374 JAMES N. PITTS ET AL.
2. Letter J, Shear MJ: Quantitative experiments on the
production of subcutaneous tumors in strain A mice with
marginal doses of 3,4-benzpyrene. J Natl Cancer Inst
3:455-477, 1943
3. Kotin P, Falk HL, Mader P, Thomas M: Aromatic hydro-
carbons. 1. Presence in the Los Angeles atmosphere
and the carcinogenicity of atmosphereic extracts. Arch
Indust Hyg 9:153-163, 1954
4. Hueper WC, Kotin P, Tabor EC, Payne WW, Falk HL,
Sawicki E: Carcinogenic bioassays on air pollutants.
Arch Pathol 74:89-116, 1962
5. Epstein SS, Mantel N, Stanley TW: Photo-dynamic assay
of neutral subfractions of organic extracts of partic-
ulate atmospheric pollutants. Environ Sci Technol 2:
132-138, 1968
6. Rigdon RH, Neal J: Tumors in mice induced by air par-
ticulate matter from a petrochemical industrial area.
Texas Reports on Biology and Medicine 29:110-123, 1971
7. Freeman AE, Price PJ, Bryan RJ, Gordon RJ, Gilden RV,
Kelloff GJ, Huebner RJ: Transformation of rat and ham-
ster embryo cells by extracts of city smog. Proc Nat
Acad Sci 68:445-449, 1971
8. Mohr U, Reznik-Schuller H, Reznik G, Grimmer G, Misfeld
J: Investigations on the carcinogenic burden by air
pollution in man. XIV. Effects of automobile exhaust
condensate on the Syrian golden hamster lung. Zbl Bakt
Hyg, I Abt Orig B 163:425-432, 1976
9. Grimmer G: Analysis of automobile exhaust condensates.
In: Air Pollution and Cancer in Man, IARC Scientific
Publication #16 (Mohr U, Schmahl D, Tomatis L, eds.).
Lyon, France, International Agency for Research on Can-
cer, 1977, pp 29-39
10. Gordon RJ, Bryan RJ, Rhim JS, Demoise C, Wolford RG,
Freeman AE, Huebner RJ: Transformation of rat and
mouse embryo cells by a new class of carcinogenic com-
pounds isolated from city air. Int J Cancer 12:223-
227, 1973
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 375
11. Ames BN, Durston WE, Yamasaki E, Lee FD: Carcinogens
are mutagens: A simple test system combining liver
homogenates for activation and bacteria for detection.
Proc Nat Acad Sci 70:2281-2285, 1973
12. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian-
microsome mutagenicity test. Mutat Res 31:347-364, 1975
13. Ames BN, McCann J: Carcinogens are mutagens: A simple
test system. In: Screening tests in chemical carcino-
genesis, IARC Scientific Publication #12 (Montesano R,
Bartsch H, Tomatis L, eds.). Lyon, France, International
Agency for Research on Cancer, 1976
14. Jerina DM, Lehr RE, Yagi H, Hernandez 0, Dansette PM,
Wislocki PG, Wood AW, Chang RL, Levin W, Conney AH:
Mutagenicity of benzo(a)pyrene derivatives and the des-
cription of a quantum mechanical model which predicts
the case of carbonium ion formation from diol epoxides.
In: In Vitro Metabolic Activation in Mutagenesis Test-
ing (de Serres FJ, Fonts JR, Bend JR, Philpot RM, eds.).
Amsterdam, Elsevier, 1976
15. Jerina DM, Yagi H, Hernandez 0, Dansette PM, Wood AW,
Levin W, Chang RL, Wislocki PG, Conney AH: Synthesis
and biological activity of potential benzo(a)pyrene
metabolites. In: Polynuclear Aromatic Hydrocarbons:
Chemistry, Metabolism and Carcinogenesis. (Freudenthal
RI, Jones PW, eds.). New York, Raven Press, 1976, pp
91-113
16. Wislocki PG, Wood AW, Chang RL, Levin W, Yagi H,
Hernandez 0, Dansette PM, Jerina DM, Conney AH: Muta-
genicity and cytotoxicity of benzo(a)pyrene, arene
oxides, phenols, quinones and dihydrodiols in bacterial
and mammalian cells. Cancer Res 36:3350-3357, 1976
17. Pitts JN Jr, Van Cauwenberghe KA, Grosjean D, Schmid JP,
Fitz DR, Belser WL Jr, Knudson GB, Hynds PM: Atmospheric
reactions of polycyclic aromatic hydrocarbons: facile
formation of mutagenic nitro-derivatives. Science,
Vol 202, 1978, pp 515-519
18. Pitts JN Jr: Photochemical and biological implications
of the atmospheric reactions of amines and benzo(a)-
pyrene. Philos Trans Royal Soc, in press, 1978
-------
376 JAMES N. PITTS ET AL.
19. Pitts JN Jr, Doyle GJ, Lloyd AC, Winer AM: Chemical
transformations in photochemical smog and their appli-
cations to air pollution control strategies. Second
Annual Report, National Science Foundation-Research
Applied to National Needs, Grant No. AEN73-02904-A02,
1975, p V-8
20. Tokiwa H, Tokeyoshi H, Morita K, Takahashi K, Soruta N,
Ohnishi Y: Detection of mutagenic activity in urban
air pollutants. Mutat Res 38:351-359, 1978
21. Teranishi K, Hamada K, Watanabe H: Mutagenicity in
Salmonella typhimurium mutants of the benzene-soluble
organic matter derived from airborne particulate matter
and its five fractions. Mutat Res 56:273-280, 1978
22. Talcott R, Wei E: Brief communication: Airborne muta-
gens bioassayed in Salmonella ryphimurium. J Natl
Cancer Inst 38:449-451, 1977
23. Daisey JM, Hawryluk I, Kneip TJ, Mukai FH: Mutagenic
activity in organic fractions of airborne particulate
matter. Presented at the Conference on Carbonaceous
Particles in the Atmosphere, Berkeley, California,
March 20-22, 1978
24. Commoner B, Madyastha P, Bronsdon A, Vithayathil AJ:
Environmental mutagens in urban air particulates. J
Toxicol Environ Health 4:59-77, 1978
25. Pitts JN Jr, Grosjean D, Mischke TM, Simmon VF, Poole
D: Mutagenic activity of airborne particulate organic
pollutants. Presented at: Symposium on mutagenic
activity of airborne particulate organic pollutants,
American Chemical Society, August 1977, and to be
published in: Biological Effects of Environmental Pol-
lutants (Lee SD, ed.). Ann Arbor, Michigan, 1978, in
press
26. Pitts JN Jr, Grosjean D, Mischke TM, Simmon VF, Poole
D: Mutagenic activity of airborne particulate organic
pollutants. Toxicol Lett 1:65-70, 1977
27. Kertesz-Saringer M, Meszaros E, Varkonyi T: On the
size distribution of benzo(a)pyrene containing particles
in urban air. Atmos Environ 5:429-431, 1971
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 377
28. Natusch DFS, Wallace JR: Urban Aerosol Toxicity: The
influence of particle size. Science 186:695-699, 1974
29. Pierce R, Katz M: Dependency of polynuclear aromatic
hydrocarbons content on size distribution of atmospheric
aerosols. Environ Sci Technol 9:347-353, 1975
30. Friedlander SK, Miguel A: Atmos Environ, in press, 1978
31. Gordon RJ, Bryan RJ: Patterns in airborne polynuclear
hydrocarbon concentrations at four Los Angeles sites.
Environ Sci Technol 7:1050-1053, 1973
32. Wynder EL, Hoffman D: Some laboratory and epidemic-
logical aspects of air pollution carcinogenesis. J Air
Pollut Control Assoc 15:155-158, 1965
33. Asahina S, Andrea J, Carmel A, Arnold E, Bishop Y,
Joshi S, Coffin D, Epstein SS: Carcinogenicity of
organic fractions of particulate pollutants collected
in New York City and administered subcutaneously to
infant mice. Cancer Res 32:2263-2268, 1972
34. Kotin P, Falk HL, McCammon CJ: The experimental induc-
tion of pulmonary tumors and changes in the respiratory
epithelium in C57BL mice following their exposure to an
atmosphere of ozonized gasoline. Cancer 11:473-481,
1958
35. Kotin P, Falk HL, Thomas M: Production of skin tumors
in mice with oxidation products of aliphatic hydrocar-
bons. Cancer 9:905-909, 1956
36. Larson RA, Hunt LL, Blankenship DN: Formation of toxic
products from a #2 fuel oil by photooxidation. Environ
Sci Technol 11:492-496, 1977
37. Fishbein L: Atmospheric mutagens. In: Chemical Muta-
gens: Principles and Methods for Their Detection
(Hollaender A, ed.). New York, Plenum Press, 1976,
Volume 4, pp 219-339
38. Berry RS, Lehman PA: Aerochemistry of air pollution.
Ann Rev Phys Chem 22:47-84, 1971
39. National Academy of Sciences: Particulate Polycyclic
Organic Matter, Washington, DC, National Academy of
Sciences, 1972
-------
378 JAMES N. PITTS ET AL.
40. Falk HL, Markul I, Kotin P: Aromatic hydrocarbons.
IV. Their fate following emission into the atmosphere
and experimental exposure to washed air and synthetic
smog. AMA Arch Ind Health 13:13-17, 1956
41. Tebbens BD, Thomas JF, Mukai M: Fate of arenes incor-
porated with airborne soot. J Amer Ind Hyg Assoc 27:
415-422, 1966
42. Tebbens BD, Mukai M, Thomas JF: Fate of arenes incor-
porated with airborne soot: Effect of irradiation. J
Amer Ind Hyg Assoc 32:365-372, 1971
43. Barofsky DF, Baum EJ: Exploratory field desorption
mass analysis of the photoconversion of adsorbed poly-
cyclic aromatic hydrocarbons. J Amer Chem Soc 98:8286-
8287, 1976
44. Dewar MJS, Mole T, Urch DS, Worford EWT: Electrophilic
substitution. Part IV. The nitration of diphenyl,
chrysene, benzo(a)pyrene and anthanthrene. J Chem Soc
3573-3575, 1956
45. Pfeiffer EH: Oncogenic interaction of carcinogenic
and noncarcinogenic polycyclic aromatic hydrocarbons in
mice. In: Air Pollution and Cancer in Man, IARC
Scientific Publication #16 (Mohr M, Schmahl D, Tomatis
L, eds.). Lyon, France, International Agency for
Research on Cancer, 1977
46. Dewar MJS, Mole T: Electrophilic substitution. Part
II. The nitration of naphthalene and perylene. J Chem
Soc 1441, 1956
47. Hopff H3 Schweizer HR: 251. Zur Kenntnis des Coronens
Dienanlagerungen in der Perylen-und Benzperylenreihe.
Helv Chim Acta 42:2315, 1957
48. McCann J, Choi E, Yamasaki E, Ames BN: Detection of
carcinogens as inutagens in the Salmonella/microsome
test: Assay of 300 chemicals, Proc Nat Acad Sci 72:
5135-5139, 1975
49. Kaden DA, Thilly WG: Genetic toxicology of kerosene
soot. Presented at: The Workshop on Unregulated
Diesel Emissions and Their Potential Health Effects.
Washington, DC, April 27-28, 1978
-------
STUDIES OF MUTAGENIC POLLUTANTS IN ATMOSPHERES 379
50. Rosenkranz HS: The use of microbial mutagenesis assay
systems in the detection of environmental mutagens in
complex mixtures. Presented at: Symposium on Applica-
tion of Short-Term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, Williams-
burg, Virginia, February 21-22, 1978
51. Pellizzari ED, Little LW: Integrating of microbiologi-
cal and chemical testing into the screening of air sam-
ples for potential mutagenicity. Presented at: Sympo-
sium on Application of Short-Term Bioassays in the
Fractionation and Analysis of Complex Environmental
Mixtures, Williamsburg, Virginia, February 21-22, 1978
52. Commoner B, Vithayathil AJ, Dolara P: Mutagenic analy-
sis of complex samples of air particulates, aqueous
effluents and foods. Presented at: Symposium on Appli-
cation of Short-Term Bioassays in the Fractionation and
Analysis of Complex Environmental Mixtures, Williams-
burg, Virginia, February 21-22, 1978
Notes Added in Proof:
p. 360
We recently found that the half-life of BaP in air con-
taining only 0.1 ppm ozone was less than one hour, and that
certain of these products were direct mutagens. This ozone
oxidation may be the most important fate of BaP on the surface
of particulate matter.
Subsequent experiments using HPLC separation suggest
that some quinones may have been formed on the TLC plate.
p. 364
Unpublished results from our own lab and that of Dickson
and that of Eisenstadt (private communications) show that,
at high S-9 levels (40% v/v) and higher concentrations of
terylene than used above, is an activatable frameshift muta-
gen in the Ames test.
-------
APPLICATION OF BIOASSAY
TO THE CHARACTERIZATION
OF DIESEL PARTICLE
EMISSIONS
J. Huisingh, R. Bradow, R. Jungers,
L. Claxton, R. Zweidinger, S. Tejada,
J. Bumgarner, F. Duffield, and M. Waters
Health Effects Research Laboratory and
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
V.F. Simmon
SRI International
Menlo Park, California
C. Hare and C. Rodriguez
Southwest Research Institute
San Antonio, Texas
L. Snow
Northrop Services, Inc.
Research Triangle Park, North Carolina
-------
383
PART I. CHARACTERIZATION OF HEAVY DUTY
DIESEL PARTICLE EMISSIONS
INTRODUCTION
A wide variety of combustion sources produce soot,
i.e., carbon aerosols containing variable quantities of
organic matter. The most significant transportation-
related sources of such materials are diesel engines.
Diesel power has been used for railway locomotives, long
haul trucks, and earthmoving equipment for many years.
However, recently a strong trend has developed toward
use of diesel engines in urban service vehicles and also
taxicabs. In the near future substantial numbers of
diesel-powered automobiles may be used by the general
public.
These comparatively new developments not only in-
crease the present rather small contribution from this
source to ambient air particulate matter, but also shift
the potential locale of the soot emission to more densely
populated urban core areas. In fact, diesel engines have
the greatest fuel economy advantages over gasoline engines
in the low speed-light load situations characteristic of
urban stop-and-go driving (1,2).
-------
384 J. HUISINGH ET AL.
Some years ago the Environmental Protection Agency's
Office of Research and Development recognized that this
issue might come into prominence as petroleum-based fuels
became scarce. Consequently, considerable efforts were
made to develop procedures suitable for measuring diesel
particle concentrations and composition (3,4). Subse-
quently, these methods have been used to describe the
emission rates and general chemical character of the com-
bustion products of a wide variety of small and large
engines (2,4,5,6).
The most interesting aspect of these particles is the
associated organic matter which varies widely in both emis-
sion rate and composition (3,7). Generally, the sources
of these organic compounds appear to be unburned fuel and
lubricant. However, there seems to be some partitioning of
organic material between the gas phase and the particle-
bound phase. Consequently, the soot-bound organic material
is higher in average molecular weight than the fuel (3).
The weight percentages of nitrogen and sulfur are also
higher in soot organics than in the fuel. Further, there
is substantial oxygen incorporation in the material, cer-
tainly as a result of partial combustion (3,4).
Diesel exhaust particulate, as well as other fossil
fuel combustion products, are known to contain the carcino-
genic and mutagenic chemical benzo(a)pyrene, among many
other potentially hazardous and less well characterized
components. Due to the potential proliferation of diesel
powered vehicles, it is critical to identify those compo-
nents which constitute a possible public health risk to
facilitate their control.
To reduce the immensity of the organic analytical task,
chemical fractionation and analysis were guided by short-
term bioassays. In this way, crude fractions containing
biological activity would be identified and prioritized
for analytical efforts to characterize components in the
most active fractions. This procedure also allows iden-
tification of relatively inactive materials and conserves
resources which might otherwise be devoted to analysis of
less important substances. The initial bioassays employed
included cytotoxicity in mammalian cells and mutagenicity
in bacteria. Since the fractions tested were not found
to be highly toxic but were mutagenic, subsequent efforts
concentrated on the use of bacterial mutagenesis bioassay
in Salmonella typhimurium to guide fractionation.
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 385
This paper represents early, but very promising,
results using such a procedure. Also described are engi-
neering, chemical fractionation and analysis, and bioassay
procedures currently being employed.
ENGINEERING PROCEDURES
Test methods for both heavy duty diesel truck engines
and diesel passenger cars have been previously described in
detail (3-5) . The procedures used for the heavy duty engine
experiments and the rationale for those procedures are out-
lined here.
Truck diesel engines, because of their very small
speed range, tend to operate at or near constant speed
much of the time. Consequently the current heavy-duty
test procedure uses a series of 13 steady state operating
speed-load conditions (modes) to simulate overall urban
use. Independent gas analysis is made of each mode and
weighing factors are used to arithmetically compose a
cycle value. For the particulate sampling, however, it
is more convenient to vary the time-in-mode to achieve a
single physically composited filter sample. All heavy
duty samples used in the present work were such time
averaged 13 mode composites collected on glass fiber
filters as previously described (4,8).
In order to obtain reasonable samples of particle-
bound organics, it is important to consider the nature of
the emission process. In the tailpipe of an operating
diesel engine, the temperatures are sufficiently high
(>200°C) that organic materials are generally in the gas
phase. Thus, soot filtered at these temperatures contains
very little extractable organic material; approximately
one percent by weight can be extracted with methylene
chloride, for example. However, when particles and
gaseous exhaust enter the ambient air, as from automobiles
and trucks, the mixture is quickly cooled and diluted.
During this process, the overall temperature is reduced to
the point that carbon particles begin to absorb organic
material. Still further dilution may reduce the gas phase
hydrocarbon concentrations to the point that further absorp-
tion ceases to occur. Thus, the particle composition may be
stabilized at some point in the exhaust-air dilution process,
This process has not been examined experimentally with real
-------
386
J. HUISINGH ET AL.
vehicles, but considerable work has gone into laboratory
simulation of this process which is assumed to occur in the
ambient air.
A number of investigators have used air dilution tunnel
techniques to achieve this simulation (3,4,9,10), and a
wide variety of systems have been shown to be reasonably
effective in at least some sampling applications. Figure 1
presents the dilution-tunnel system used for these studies.
In the highest load modes of the 13-mode test procedure, ex-
haust volumes are so large that the capacity of a system
scaled to dilute the whole exhaust would have to be immense,
perhaps 500,000 to 1,000,00 liters/min. Consequently, heavy-
duty engine exhaust is first variably split, then diluted
10- to 15-fold in the dilution tunnel. The individual mode
dilution ratios are determined by the ratios of gas concen-
trations in the dilute and raw exhaust streams for C02 and
NO. L-irge samples of particulate material were obtained from
this apparatus by filtering the whole dilution tunnel con-
tents at a flow rate of 12,000 liters/min.
TO
MUFFLER
BACKPRESSURE
REGULATOR
Figure 1. Dilution tunnel system for collection of heavy
duty diesel particulate.
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 387
Two test engines were chosen for this program, both
high production, naturally-aspirated, medium duty truck
power plants. The first, engine No. 1, was a typical city
bus engine, a Detroit Diesel two-stroke-cycle 6V-71 in-line
6 cyclinder engine. Engine No. 2 was a 4-stroke cycle,
V-8, Caterpillar 3208, an engine now widely used in urban
service vehicles.
CHEMICAL FRACTIONATION PROCEDURES
The procedure used for the extraction and separation of
the organic components present in diesel exhaust particulate
are outlined in Figures 2-5. Diesel exhaust particulate
collected by filtration on glass fiber filters was extracted
for six hour periods, first with dichloromethane (DCM) fol-
lowed by acetonitrile (ACN). The majority of organic material
was removed by the DCM extraction with some additional organic
and inorganic material obtained by the subsequent ACN extrac-
tion (see Table 1). Initial characterization studies have
dealt entirely with the DCM extracts. Fractionation of the
DCM extracts for mutagenesis testing was carried out by pro-
cedures similar to those employed by Swain, et al. (11) for
cigarette smoke condensate.
Table 1
Diesel Particulate Extracts
Engine #1 Engine #2
2-Stroke 67-71 4-Stroke 3208
Particulate emission rate 86.7 g/hr 42.5 g/hr
Total particulate collected 118.0 g 195.13 g
DCM extract 64.05 g 47.33 g
ACN extract 10.57 g 19.65 g
The solvent partitioning steps employed to obtain acid,
basic, and neutral fractions are outlined in Figure 2. A
sample of DCM extract was evaporated, weighed and recon-
stituted in ether. A small amount was ether insoluble and
removed by filtration (INT fraction). The ether solution
was extracted with 0.1N Na2C03 to obtain the acid fraction
(ACD) and then with IN H3PO,, to obtain the basic fraction
-------
388
J. HUISINGH ET AL.
FILTERS
SOXHLET
EXTRACTION
CH2CI2 (DCM)
DCM EXTRACT
1. EVAPORATE & WEIGH RESIDUE
2. REDISSOLVE IN ETHER
ETHER SOLUTION WITH
SOME INSOLUBLES
I EXTRACT WITH BASE
FILTERS
SOXHLET
EXTRACTION CH3CN (ACN)
ACN EXTRACT
\
AQUEC
\
US PHASE ETHER
1. ACIDITY
2. EXTRACT ETHER
'1
ACID FRACTION AQUEOUS PHASE
(ACD) (DISCARD)
1 -0.3
II -2.0
8-5.15% AQUEOUS PHASE
(1. ADDBAS
2. EXTRAC1
t
BASIC FRACTION
BAS
*
SOLUTION ETHER INSOLUBLES
(INT)
c.ThMU H3K,4 ..0.12- 0008
II 1 18-0.700
*
ETHER SOLUTIONS
E NEUTRALS (NUT)
1 -53.38-51.81
* . I
AQUEOUS PHASE |
(DISCARD)
1 • 0.03 • 0.03
II -0.09-0.05
Figure 2. Isolation and fractionation organics from diesel
exhaust particulates.
(BAS). The remaining ether solution containing the neutral
fraction (NUT) was then further fractionated by chromatography
on silica gel (Bio-sil-A, 100-200 mesh) as shown in Figure 3.
Elution was initiated with hexane, which removed the paraf-
fins (PRF). When fluorescent materials, as observed under
long wave-length UV, reached the bottom of the column, the
aromatic fraction (ARM) was collected. The eluting solvent
was then changed to 1% ether in hexane at which time a
narrow yellow band moved down the column. This band did not
fluoresce and quenched the bluish fluorescence ahead of it.
The third or transitional fraction (TRN) was the yellow band
of material which was also eluted with 1% ether in hexane.
The remaining polar oxygenated compounds (OXY) were removed
by elution with 50% acetone/methanol.
The percentage that each fraction represented of the
total exahust particulate originally collected is given in
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS
389
EXHAUST AND FUEL SILICA GEL CHROMATOGRAPHY FRACTIONS
(NEUTRALS AND EM-239 FUEL FOLLOWED SAME SCHEME)
FUEL
F-ARM h*-
_F-TRNj-*-
7-oxvJ-*-
SILICA
GEL
CHROMATOGRAPHY
HEXANE ELUTION
NO FLUORESCENCE
1% ETHER/HEXANE
INCIPIENT FLUORESCENCE
CONTINUED
1% ETHER HEXANE
STRONG FLUORESCENCE
50/50
ACETONE/METHANOL
ELUTION
MODERATE FLUORESCENCE
NEUTRALS
7.40 - 6.31
II-7.43-5.64
Figure 3. Silica gel chromatography fractionation of the
neutral organics from both diesel exhaust particulate and
uncombusted diesel fuel.
Table 2. The yield for each fraction was slightly affected
by the base extraction employed in the initial solvent
partitioning. The largest variations were seen in the ACD
and OXY fractions and probably were due to incomplete ex-
traction of phenols and other weak acids by 1 N Na2C03.
In addition to the DCM neutrals, a sample of diesel fuel
was also chromatographed on silica gel in analogous fashion
(Figure 3). The amounts obtained for each fraction are given
in Table 3. Samples of all of the above fractions were pre-
pared for mutagenesis bioassay by removing the solvents by
evaporation and reconstituting in dimethylsulfoxide (DMSO).
On the basis of initial mutagenesis test results, further
fractionation of the TRN and OXY fractions was accomplished
by high pressure liquid chromatography (HPLC). The TRN
fraction was chromatographed on a NH2-bounded phase column
-------
390 J. HUISINGH ET AL.
Table 2
Fractionation DMC Extracts
(% of Total Particulate)
Engine #1 Engine #2
Fraction 2-Stroke 6V-71 4-Stroke 3208
ACD
BAS
INT
NUT
PRF
ARM
TRN
OXY
l.ON N2C03 O.IN KOH
0.83 1.93
0.03 0.03
0.12 0.008
53.38 51.81
36.26 34.35
6.72 7.51
3.22 2.96
7.40 6.31
Table 3
Fraction of Diesel
Fraction %
PRF
ARM
TRN
OXY
l.ON Na0COo
£t O
2.08
0.09
1.18
19.27
8.89
1.76
1.20
7.43
Fuel EM 239
of Fuel
74.04
21.51
0.69
0.33
O.IN KOH
5.15
0.05
0.70
17.30
8.94
1.61
1.22
5.64
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS
391
TRANSITION SUBFRACTIONS:
N-TRN
CHROMATOGRAPHY
ON NH2 - BONDED PHASE
4% CH2CI2I HEXANE
ELUTION
i i if
TRN-li
FrA
TRN-II
FrB
TRN-II
FrC
TRN-II
FrD
Figure 4. Chromatographic subfractionation of the neutral
transitionals.
(Varian Associates) with 4% methylene chloride in hexane
(Figure 4). Four complex fractions (A,B,C, and D) were
collected for additional mutagenic testing.
The OXY fraction was further separated by gel permea-
tion chromatography on 100 A y Styragel (Waters Associates)
which has an exclusion limit of approximately 700 molecular
weight. Two fractions arbitrarily divided into high (GPC-
1) and low (GPC-2) molecular weight were collected using
dichloromethane as eluent (Figure 5). These two fractions
were of approximately equal mass and were submitted for
mutagenesis testing along with the original OXY material
(NEAT OXY).
Detailed characterization of the various fractions and
subfractions aimed at identifying specific mutagens has been
undertaken by gas chromatography on glass capillary columns
with mass spectrometric detection. At present, some in-
formation is available concerning classes and types of com-
pounds. The aromatic fraction (ARM) contains most of the
PNA hydrocarbons such as benzo(a)pyrene. The TEN fraction
contains substituted PNA's, phenols, ethers and ketones such
as fluorenone and its methyl and dimethyl isomers. The known
mutagen 2-aminofluorene has also been tentatively identified.
-------
392
J. HUISINGH ET AL.
N - OXY SUBFRACTIONS:
N
-OXY
T I
GEL PERMEATION
ON 100 A juSTYRAGEL
IN CH2CI2
OXY-II
GPC-1
OXY-M
GPC-2
1
OXY-II
NEAT
Figure 5. Gel chromatographic subfractionation of the neutral
oxygenates.
It has been extremely difficult to work with the OXY
fraction. Due to its very polar nature (up to 10% oxygen by
elemental analysis), this material does not gas chromato-
graph well. Investigations are currently under way using
HPLC fractionation schemes coupled with direct probe and
field desorption mass spectrometry in both electron impact
and chemical ionization modes.
BIOASSAY
The Ames Salmonella typhimurium/microsome mutagenesis
bioassay was used to indicate which fractions of diesel
exhaust were genetically active and to guide the fractiona-
tion of diesel exhaust so that biologically active components
could be isolated and characterized. The plate incorporation
procedure followed in these studies is described in detail by
Ames, McCann, and Yamasaki (12). Histidine dependent strains
of Salmonella typhimurium were obtained from Dr. Bruce Ames
of the University of California at Berkeley. These strains
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 393
were routinely checked for their genotypic characteristics
and for the presence of the plasmid, as described by Ames et
al. (12). Positive controls for each tester strain and the
activation system as well as negative solvent controls were
included with each experiment. In order to detect chemicals
which are mutagenic only after metabolism by a mammalian
enzyme system, a rat liver metabolic activation (MA) system
is used in the bioassay. The Aroclor 1254-stimulated meta-
bolic activation system was prepared as described by Ames et
al. (12). Chemicals which are mutagenic without the meta-
bolic activation system are referred to here as direct-acting
mutagens.
Exhaust particulate samples from the heavy duty diesel
engines were solvent extracted and fractionated as previously
described. Bioassays were performed following removal of
the solvent and addition of dimethyl sulfoxide (DMSO) to
dissolve the mixture. After preliminary range-finding tests
all samples were evaluated as described below except where
sample size was limiting. The fractions were examined with
five tester strains of Salmonella typhimurium (TA1535, TA1537,
TA1538, TA98, TA100) with and without the liver metabolic
activation (MA) system. The experiments were conducted in a
dose response fashion (6-8 doses/fraction/tester strain) and
each experiment was repeated where sample size permitted.
Seven fractions from each engine were tested initially at
SRI International. Subsequent subfractions and selected
samples of the initial fractions were tested at Northrop
Services Inc. and EPA (HERL/RTP) Laboratories. Identical
samples tested in separate laboratories produced a similar
mutagenic response.
In these initial investigations, extracted and fraction-
ated samples were stored for several months prior to bioassay.
Subsequent studies (reported in Part II) have shown that
the mutagenicity of an unfractionated diesel extract was
slightly reduced as a result of storage. No data are avail-
able on the effect of storage on the fractionated samples.
Figure 2 gives the original fractionation scheme used
with heavy duty diesel exhausts. A summary of the bioassay
results on these fractions is provided in Table 4. A frac-
tion was considered positive if it gave a maximum response
that was 2.5 times greater than the spontaneous rate for the
particular strain used and if it gave a positive linear dose
response in the major portion of the curve.
-------
394
J. HUISINGH ET AL.
rtfi
Q
rH
^Q
erf
EH
•P
CO CO
CD 0>
•H
be s:
C 0
W -i-l
-P
O erf
EH O
ft
O 0
!H O
>
-p
CD erf
CO rH
0) rH
•r-l 0)
Q C
O
>CO
rH
erf
c
H
rH
CO
cu
•rH
Q
CO
d
PQ
0>
rH
0
0
o
? *
EH
.
rH
0)
C
•H
bfl
C
W
ST*
*
g oooSC
+ gj 28
CO CM C-
8
^ S
< 31 *
<: o ^ co
S 55 C-D^
i • •
CO CN 00
.— .
8*
2 in C1* co
ass
X — t CO
01
S^
*
S r- to co
sss
CM
^
S ?s
< ^-^
-t- in o ^
CO CN) Ol CM
in — < co
< ^
- o
^ ,H TH'CO
i to m •»— '
S8S
5T
Ol *
S ? CO 2-
CM ^ ro
i> CM rH in
CO
in
og
q o. O
CO O
•* N
u< dj
^ a
SB«
• S, -^^
•« S
cams
J
a
3
CN C^J * *
r*- '•"'» co -"^ • o • *** s c* O ©• O
in CN in CN %—' • *— 'TH co • c3i *
(0 ^-O J-O -; JO CM (O
x— • ^v
SCO
o) ^^
• '-v • O
CO CTJ O CO S C
Z S CJCMCMMg S
•"* t?
I CT 1 S"
— q --q
CO <3) ^^ X CJi ^
S z 2^22^2^ fe
-> ^-x ^^ ^^
•!T ___ CO IN »^ -i J^
O CO • *~* • O • 1O
g g in co o o co' J-H -H co in in CN CN
,->. ^-v ^-, s~.
2§ 3,7 Iff ol
<3 X-N
•*> tO CM ^~^ "CO fll '"^ Ol '"^
Sc O co ^8 ^^ OC-
^,J,O in o -^ c» 8 "^ CNto^ COCD^
•fCOrH S rH <-* rH rH trtCOin OJrHCN
^ £?
-— s ^) CM *~^ CO O
So Oo Sc§ O[2 ^S
•-'^ IrtO^ ^ •-* ^ CMCM^ COt^'—
•qi^CO &H COrHCM IDrHtO
bo
M
i
£
in •
•0 CM* r>
(- aj O
O D
O *-* -C
• 0 g -rJ
•P CO O
*^ CO* O -H
CO CU S. •*-»
! u ^
i ^ p -0
o -c — • S1
«N '- 1 0)
• 1* J^
-C C -<-> __
cji ^ £5 ofl 3J
flf JM -8 4J C
of 8 u H Si
^ ° +3 !£ S
CO C t-
-fj 0 *0 0)
C tH C > CD
£ ai V O
1 [3 a ^i
^ -H "H H
cu — a 3)
5 w T3 0)
C C S3 -M
=*-< «J o O •"
O * +- U CO
V 5 ^ x
~ o o P # a
O -P rH > 3S
•>H co cfl cp S ca
c c bfl « c 5
1-1 ^_i ^ ci a
1 o fcj -p 2 H
Co 01 CO ti Li
i* O O *+H
'S "" 5 8 J H
w t O c
ac C yi 0 Q) T3
0 0 > -i-t 35 0
•HOlOW*'al§3
illgf^I-a
0 clI'Ss^ 1 ™
I 1 1 * 5 5 M
a 1 .... I I "1
§fl) ^~V r*% -P +J SO.
£C tt ^ -H -H 4-)
^-- C5 * 3K cfl CP
1 3 _o a
'o'^rHasSSrHcS
O 03 CO S ( + * "M
•H ea co * 10 to c-
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS
395
•o
0)
c
•H
c
o
o
0)
oJ
EH
HO
C
W
iH
CO
•H
Q
I—I
O
U
JH
O
PH
o;
c
bO
C
w
•—— co -j •— inm— 01 o> — to •»
3 n ^H
*-tO 'O 'CO *O •
05 CO — • —CO ^^ • —
— — '^'l^-' oioi—
• oo co in x
lO GO
in c*«
8S
2SS
•CO Ol CO CO i-4 £*•
in m ^
01 o 01 c
[> 00 *-' CO C
co in ^ co i
co q ^
01 O 00
CO -H CO
ci co "-' q
t*- in c*- 01
CO CM CO ©
i-H ^ 01 in o
m TH c^ CM co
Ol CM CM CM TT
s§
jr-*co t-HCMO inooi rrcom
J CO »~t CM >H CM H
010^-^ co ^ *—' in m *—' oi'*^-'
8{
- m *•
SQ CO rr '
•v n CM
S5
O -M
al CJ
• M «^tti *i-HCfl • £
3 co S co co S co co 2 coco!
CO CO S CO CO S CO CO S CO CO S COCO
< 3 < a
co co S co co S
S
TJi CO CO CO
CM CM Ol O
in co co -
« ra « .
Q-
:«
-------
396 J. HUISINGH ET AL.
The total DCM extract and ACD, TRN, and OXY fractions
from the 2-stroke cycle bus engine were positive, both with
and without activation, in strains TA1537, TA98, and TA100.
In addition the total DCM extract was positive with TA1538
both with and without activation. The TRN fraction in this
engine was also positive with TA1538 when the activation
system was added.
The total DCM extract and ACD, INT, TRN, and OXY frac-
tions of the 4-stroke cycle truck engine were positive with
and without activation, with strains TA1537, TA98, and TA100.
TA1538 showed a positive response with and without activation
to the DCM extract and INT, TRN and OXY fractions and to the
ARM fraction without activation. When activation was used
the DCM extract and TRN and OXY fractions also showed positive
results with strain TA1535 in this engine. The BAS fraction
of this engine was positive with TA98 with the added activa-
tion system.
A comparison of mutagenic response in TA1538 of the
various fractions from the 4-stroke cycle truck engine is
shown in Figure 6. In both engines the TRN and OXY subfrac-
tions of the neutral compounds are the most mutagenic when
either the maximum fold increase (max. revertants/plate in
sample minus solvent controls) or the specific activity
(revertants/plate/yg sample) are compared. It can be noted
that each positive fraction contained direct-acting mutagens.
It also appears that the positive fractions contain compounds
that need activation before being mutagenic. However, these
fractions are complex mixtures and the metabolic activation
system may also function to detoxify certain components thus
allowing expression of the mutagenic potential of other com-
ponents. Furthermore, as the concentration of some compounds
is increased in the assay, metabolism of potentially active
compounds may be altered such that a mutagenic metabolite is
not formed.
The use of all five tester strains, as shown in Figure 7
for TRN II, yields information about the chemical structure
and reactivity of the mutagens (13-17). Strain TA1535 is
reverted to histidine independence by many mutagens which
cause base-pair substitutions. Strains TA1537 and TA1538
are reverted by many frameshift mutagens. Strains TA98 and
TA100 are more sensitive generally to mutagenic agents due
to the addition of plasmids and may respond to mutagens which
act either by base-pair substitution or frameshift mutation.
Based upon the positive responses obtained in strains TA1537
and TA1538, it appears that the active components are mainly
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 397
800
960 1000
CONCENTRATION OF COMPOUND ADDED TO PLATE IN MICROGRAMS
DOSES ARE 10, 33,100. 333.1000
Figure 6. Comparison of the mutagenic response of various
organic fractions from the 4-stroke cycle diesel truck
exhuast particulate in Salmonella typhimurium strain TA1538.
frameshift mutagens. Furthermore, the activity within strains
TA98 and TA100 appears to arise mainly from direct-acting
mutagens. Since strain TA1538 showed activity with each of
the positive fractions and also demonstrated a quantitative
difference with and without activation, it was the strain of
choice when only the use of one strain was possible due to
sample size limitations.
Since most of the mutagenic activity was present in the
neutral subfractions (TRN and OXY) it was decided to further
separate these fractions. Chromatography of the TRN fraction
yielded four subfractions (FrA, FrB, FrC, and FrD). Due to
the lack of material each of these subfractions was assayed
at only one dose in duplicate. FrA was negative, but the
other three subfractions were positive. On a per weight
basis, FrC was clearly the most mutagenic component (Table
5). The OXY fraction was subfractionated into two components,
GPC-1 and GPC-2 by gel permeation on 100& Styragel. An equal
amount of components by weight went into GPC-1 and GPC-2.
-------
398
J. HUISINGH ET AL.
1000
50
100
350
Figure 7. Comparison of the mutagenic response of all five
Salmonella typhimurium tester strains with (+) and without
(-) metabolic activation for the TEN II fraction.
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 399
Table 5
Mutagenicity of the TRN and OXY Subfractions in the
Salmonella typhimurium Plate Incorporation Test in TA1538
TRN: OXY:
Subfraction Fr B C D GPC-1 GPC-2 NEAT
ug/plate 184 370 218 410 360 360 333
Revertants/plate 50 427 >1000 290 75 2333 1306
with activation
Again, due to small sample quantities only one dose (320 ug/
plate) could be tested. Nearly all of the activity was
recovered in GPC-2; therefore, there was a two-fold concen-
tration of the active components into GPC-2. Fractionation
is continuing in order to identify the specific combustion
products that are mutagenic in this microbial system.
In order to investigate the possible source of mutagenic
compounds in the exhaust particulate, the unburned fuel was
fractionated and bioassayed. Neither the neat or fraction-
ated fuel components were found to be mutagenic. The possi-
bility that preparation and fractionation of exhaust partic-
ulate could convert otherwise inactive compounds into mutagens
seems very unlikely in view of these negative results with the
uncombusted fuel fractions. Blank filters, extracted and
carried through the fractionation and bioassayed were also
negative.
SUMMARY
Heavy-duty diesel particulate emissions from a 2-stroke
cycle and 4-stroke cycle engine were found to have 54 and 24%
organic extractable components. These organic extracts were
mutagenic in the Salmonella typhimurium/microsome bioassay.
Fractionation of these extracts yielded 53% (2-stroke cycle)
and 17% (4-stroke cycle) neutral components and substantially
smaller amounts of ether insoluble, acid and basic components.
All of these fractions showed some mutagenic activity. The
neutral components contained a major paraffinic fraction which
was not mutagenic. The other three fractions of the neutrals
were mutagenic, with the transitional and oxygenated fractions
being most mutagenic. Further fractionation and bioassay
-------
400 J. HUISINGH ET AL.
suggest that these fractions contain a minimum of four sepa-
rable mutagenic components. These mutagenic fractions consist
of the more polar neutral compounds such as substituted poly-
nuclear aromatics, phenols, ethers, and ketones.
The mutagenic activity compared among tester strains,
with and without metabolic activation, suggests that the muta-
gens are primarily direct-acting frameshift mutagens. Meta-
bolic activation, in most cases, increases the mutagenic
response suggesting either the additional presence of pro-
mutagens or the detoxification of toxic components in the
mixture.
The mutagenic activity does not appear to result from
artifacts of extraction or fractionation of the samples.
Fractions of uncombusted fuel were not found to be mutagenic,
suggesting that the mutagens are products of the combustion
process.
PART II. APPLICATION OF A MUTAGENICITY BIOASSAY
MONITORING LIGHT DUTY DIESEL PARTICLE EMISSIONS
INTRODUCTION
The premise that diesel passenger cars will be used in
the future stimulated EPA to initiate the development of
methods for monitoring emissions products.
The predicted amount of particulate matter emitted from
both gasoline and diesel cars is illustrated in Table 6.* The
estimated particulate matter emitted from gasoline cars in
1990 assuming that zero diesel cars are sold is approximately
32,000 tons. This amount of particulate from gasoline cars
decreases insignificantly as the sales of diesels increase.
In contrast, the estimated amount of particulate matter emit-
ted from the diesel cars ranges upwards dramatically, to
155,000 tons if the sales of diesels increase to the predicted
25%. Therefore, a 25% penetration of the market by diesel
vehicles could result in 181,000 tons of particulate emitted
per year by all vehicles with over 85% of the particulate
attributable to diesel cars. This estimate assumes no emis-
sion control on diesel passenger cars.
*This estimate does not reflect the influence of particle
emission standards to be imposed as a result of the 1977
Clean Air Act Amendments.
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 401
Table 6
Particulate Matter Emitted From Gasoline and Diesel Cars
Particulate Emissions Particulate Emissions
% Diesel in Tons by in Tons by
New Car Sales Gasoline Cars Light Diesel Cars
1985 - 1990 by 1990 by 1990
0 32,000 0
10 28,000 57,000
25 26,000 155,000
A means of enforceably, efficiently, and economically
monitoring mobile source emissions is needed. This pilot
study on particulate emissions from light-duty diesel vehicles
employed several analytical tools including the Salmonella
mutagenesis bioassay. The objective was to determine the
feasibility of identifying factors which influence the muta-
genicity of organic extracts of diesel particulate. Monitor-
able parameters currently used in particulate samples from a
variety of sources include but are not limited to total sus-
pended particulate, benzene soluble organics, and benzo-(a)-
pyrene (BaP). Although BaP may not be the most biologically
active component present in diesel exhaust emissions, an
analytical scheme had been developed to measure BaP concen-
tration rapidly and precisely in ambient air particulate (18-
20) so this technique was used in this study.
ENGINEERING AND CHEMICAL PROCEDURES AND RESULTS
The particulate samples used in this study were collected
using a dilution tunnel and sampling configuration described
previously (21). In this case, all of the passenger car
exhaust was diluted, but a fraction of the dilution tunnel
contents was filtered. The filter samples were collected iso-
kinetically on 20.32 by 25.4 cm (8"xlO") Gelman type A glass
fiber filters at a. flow rate of 600 liters/min. The dilution
tunnel flow was 10,000 liters/min. Each filter therefore
represents 6 per cent of the total exhaust particulate mass.
The diesel vehicles used were a Volkswagen Diesel Rabbit,
Mercedes 240D and Nissan - 4 cylinder. These diesel automo-
biles were operated on a chassis dynamometer using the
-------
402 j. HUISINGH ET AL.
following standard driving cycles: hot start Federal Test
Procedure (FTP), cold start FTP, and 85 km/hr.
The fuels used in this study included those listed in
Table 7. These fuels were blended for this study (22) to
represent a cross section of diesel fuel available to the
public, including the Gulf National Average Fuel. EM 238-F
is a No. 2 diesel smoke test fuel with medium to high cetane
rating and medium-low aromatic content. EM 239-F is a No. 2
diesel Gulf National Average Fuel with medium to high cetane
rating and low aromatic content. EM 240-F is a No. 1 diesel
jet A fuel showing a high cetane rating and very low aromatic
content. This fuel was blended for aircraft use, but it can
be used in automotive power plants. EM 241-F is a No. 2
diesel minimum quality fuel having low cetane rating and a
high aromatic content. EM 242-F is a No. 2 diesel maximum
quality fuel having a high cetane rating and a low aromatic
content.
The particulate samples were treated as outlined in
Figure 8. A one by eight inch strip of the filter was cut
and processed. The samples were extracted for six hours in
a soxhlet extraction apparatus with 100 ml cyclohexane
(Burdick-Jackson) refluxing at a rate of eight times per
hour. The apparatus was allowed to cool to room temperature
and the extract transferred to a Kuderna-Danish concentrator,
which was placed in a water bath at a constant temperature of
50°C. To speed evaporation, the solvent surface was swept
with a stream of dry filtered nitrogen. The extract was
reduced to ten milliliters after two successive washes of
the container. Fifty microliters were removed and spotted on
a one centimeter channel of a 20x20 cm 20 per cent acetylated
cellulose TLC plate. The plate with samples, standards, and
blanks was developed to the 19 cm line in a solvent mixture
of 50 ml methylene chloride and 100 ml ethanol. The plates
were air dried and placed in a Perkin Elmer MPF-3 (or MPF
44A) fluorescence spectrometer. Each channel (18 total) was
scanned using an excitation wavelength of 388nm and read at
an emission wavelength of 430nm for BaP. All extraction and
fluorescence steps were carried out under filtered light
(Kodak Yellow Chrom II). A minimum detectable limit of
0.05 ng BaP per 50 microliter of extract was determined.
Extraction efficiencies for BaP from spiked blank filters
using cyclohexane were 98 +_ 5% while recoveries from ambient
air filters were 93 +_ 5% for cyclohexane, 94 +_ 5% for ben-
zene, and 88 + 5% for methylene chloride.
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS
403
^
CD
rH
X!
oJ
EH
fa
£
-3
1
S3 rH
w ^
CO
0)
3 fa rH
fa
OT -P •
rH CO OJ fcC
0) CM !S >
03 1
1 EH
S3 oq
a*b
CM
5fc
.c
G
3
0
O
03
•H
03
>v
1— 1
ctj
C
<
O
rH
CO t>
t> oo oq
00 O O
CO
oo
rH
CD O
•<*< 00 CO
• • •
CSI O O
CO
CD
CO rH
O rH
C5 00 O
• • •
CO O O
G>
'J*
^^ t^
C3 00 rH
. . «
in o o
CO
^^ C"**
C5 CO CM
• • *
in o o
CO
in
T^
in o
CO 00 rH
• • •
^ o o
CO
i^
>> >> 3
-P -P 03 3
oS C CO
t-t CO
O Q S*.
O 00
o o>
• •
C** 0s* CO O3 ^^ ^* ^^
C^ rH 00 C>- rH 00 rH
co ^* ^t^ in co
CD in
CO O5
• •
oo in co o o ^ co
C- OJ CO t> rH 00 rH
CO ^^ ^^ tO CO
00 CO
O5 00
• •
^J* O3 C^* 00 CO CO *^
^ CD O5 in O 00 rH
co co co ^ in
O5 CO
rH CJ5
• •
05 oo co CD ^f co ^
CO O] C5 t> CM 00 rH
CO ^^ ^f lO CD
^ OQ
CO !>
r • •
^^ Tj^ CO ^^ CO CO '^
O5 N C5 O in 00 rH
CO ^ ^ CD CO
CSI O
CD 00
• •
CD 00 00 O CS] Tfi rH O5 O5 ^ 00 rH
co ^ ^ in co
C
0
•H C
•P >
03 £LI 6% &%3* O 33
•H 23 O O O CX
O HHrHinOTWs*^
CD ^ O
• • •
rH O 00
00 rH
oo in i>
• * •
CJ> O O3
m co
rH ^j* in
• • •
CO CM ^
00 rH
Tf ^J* CNI
• • •
CSI O t>
00 rH
in co CM
• • •
^ o in
t> CM
05 in co
• • •
00 rH O5
CO CM
03 03
0) 0
•P CO -H
oJ G -P
rH -H Ctf
3 --H E
•P 03 O
aS rH IH
< CO O — i
fa &% S^. 6%
O
•
CO
in
00
•
rH
^t1
^
•
f^
-------
404 J. HUISINGH ET AL.
Hi. Vol. Filter Strips
6hr. Soxhlet Extraction
with 1OOml.
Cyclohexane
Concentrate to 1Oml. in
a
Kuderna-Danish Apparatus
^k
T.L.C. SOjil. Solvent Exchange
for BaP Analysis with 1Oml. DMSO
1
to Bioassay
Figure 8. Extraction of organics from diesel exhaust partic-
ulates for benzo-a-pyrene analysis and bioassay.
The samples prepared for bioassay were solvent exchanged
with dimethysulfoxide (DMSO) in a Kuderna-Danish apparatus by
evaporating the cyclohexane to 3 ml and adding 10 ml DMSO.
The volume was then reduced to 7 ml under nitrogen and in a
50°C water bath. An additional quantity of DMSO was added to
bring the sample volume quantitatively to 10 ml. The samples
were placed in vials and frozen prior to bioassay.
BIOASSAY PROCEDURES AND RESULTS
The bacterial mutagenesis plate incorporation assay with
Salmonella typhimurium was performed according to the method
of Ames et al. (12) with the exception that the minimal his-
tidine concentration was incorporated into the base layer of
the bacterial plates rather than into the overlay. None of
the light duty diesel exhaust samples have been chemically
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 405
fractionated. Instead, we examined either a dichloromethane
(DCM) or a cyclohexane (CH) extract of the total exhaust.
The cyclohexane and dichloromethane extracted samples
were solvent exchanged into DMSO as described above and
either 0, 50, 100, 200 or 400 vl of the sample (except where
indicated) were added to each plate. Strain 1538 was chosen
for the experiments reported here due to limited sample size
and the response of the total diesel extract observed pre-
viously with this strain as reported in Part I. All assays
were performed in duplicate in the presence and absence of
metabolic activation (MA). Average revertants per plate were
calculated and adjusted by subtracting the spontaneous rever-
tants from the control plates. The fold increase was calcu-
lated at each dose by dividing the number of revertants (rev./
plate) in the treated plates by the control. The revertants/
plate were plotted against the equivalent mg of particulate
added and a linear regression was used to determine the slope.
The specific activity in revertants/plate per 1 mg diesel
particulate was calculated using this slope. The data ac-
quired from the testing of light duty diesel are summarized
in Table 8.
Variables inherent in the bioassay of organic extracts
of diesel particulate were examined. These variables included
the solvent systems and sample storage method and time. A
comparison of two solvent systems was made using samples from
two different engines each extracted with the different sol-
vents, and tested in the bacterial plate incorporation test
with and without activation. The dichloromethane (DCM)
extracts gave consistently higher numbers of revertants per
plate than did the cyclohexane (CH) extracts in either the
presence or absence of metabolic activation, as shown in
Figure 9. The DCM extracts were more mutagenic than the CH
extracts when either fold increase or specific activity was
compared (Table 8).
Since there is usually a time lapse of several days to
several weeks before generated samples can be tested, storage
may be an important factor. Equal portions of the same fil-
ters were taken for storage samples and fresh samples. Fil-
ter and extract samples were stored for eight weeks, refrig-
erated in sealed containers. The activities of the stored
samples were then compared to fresh sample activity. Fresh
samples were tested within 24 hours of collection. Results
are illustrated in Figure 10. Whether or not metabolic acti-
vation is used, there seems to be some loss of mutagenic ac-
tivity with storage. Direct acting components also seem to
-------
406
J. HUISINGH ET AL.
oo
0)
Q
d
EH
00
CO
j
OH
CD
(~* ^{) |
O C
r- 1 -H E
CJ CO 3
>>;__> -H
O rH
3
«H E
O -H
,c
CD a
CO >,
C -P
o
i—H CO
CO rH
OS 0>
o
w
EH
rH
crj
CO
CU
a
^ rH
C" 3
cd M 0
CD E -H
bfl -P
C rl
c>
CJ CO
3 03 P
O cd
P •"-) o.
•H T!
* <
1
1
1 CM
1 \ CJ
1 > cd
1 CB rH
i as a.
i ^j*
1 M
1 T3 cd
1 rH CO
1 O fn
C
1 rH
1
< -^
as >' co
.= cc 4-
P Cd
H • rH
£ -r-j CH
1 -r->
1 <
1
1
1 CM
1 -^ CU
1 • -P
1 > cd
1 CU rH
(J
HP
Si
CJ r-
W bfi -
O S o
Q -H
4-1
r-,
5
a.
rH 0 >
•• Z r-
W O
J fc. 03
a * i
S P rH
< rH 1
CO •-. 3
CL. 3.
^
r-
O)
co
10 co
* •
P* 00
*™* CO
CO n
o o
m
CM
•
^*
*H
m
^* CO
• •
rH C-
CM
rH O
^ rH
rH CO
^* CO CJ3
-i O O
rH rH rH
in in in
CO rH CM
C51 C^ CO
CM CM CM
CO •* rH
r- in i/3
CM CM rH
CO CM CM
ro co CM
C- CO CO
o o in
CM rH O
^•^
-rj
CJ
0 1
CO Cft
O CO
CO CM
•^
rH C-
CO f-
o o
i-H
CO CM rH O
^3< CM co en
CM in rH fl<
CD O 1^ C*«
o in in ^
co rH co en
rH
CM m f- o r~ o
CMCMrHrH COCMCMrH
co
en c^ CM co in oo
"r^ CO »H CO CO CM
ininmm coininm
^ji
to
c
0
4^
O
a
IH
4-1
X
4->
C
CU
rH
0
(0
5)
4-1
-
cu
o
>,
4-J
CU
c
'So
c
CB
>>
J2
•a
cu
3
CJ
a
E
:tS
CO
0)
•p
vO
0)
rO
•H
fn
0
1
at
rH
a.
M
a
to
+J
d
d
4-1
CO
a>
rH
CO
3
C
c
cd
4->
o
o.
frt
s
p;
•r-4
a?
ed
r-f
O.
CO
•p
s
+J
J-
OJ
i-.
Ql
p
• CtJ
CU rH
p a.
cd
a cu
a
.H
0) CO
a P
c
p P
C (H
p >
!H CU
CU rH
0) tn
^
O 3
j_i
cu -o
jj 0!
S P
3 03
a 3
•rj
CU T3
cm <
cd
rH
U) ..
> CU
-a P
cd
,-H
• • a
cu *^^
p •
cd >
rH CU
o. a
> -rj
CU TJ
CM CO
•
CJ
4->
a
'a
p
c
si
p
rH
CB
>
CU
^
CO
3
0
CJ
c
oi
a
Q
fn
r-^
CJ
d
rH
a
10
4-1
C
cd
4-1
rH
CB
>
»
IH
cu
CO
cd
cu
!H
O
C
rH
O
Tf
cu
4-1
cd
3
CJ
tl
a
tjj
g
CJ
c
o
(^
cu
a
CO
cu
p
C rH
o a
co cn
CO P
rH 3j
(50 P
CJ M
t. CJ
CB
G
-H
rH >*
•r-i
G -H
O >
-rH
13 P
(V O
trf
JO O
•H
CU
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS
407
0)
-2
OT 3
cS 00 O
CO S -ri
C -H
al
0.
•
cy <;
< a •
son
1 r-l
CO 00
0) «J
< a •
S O CO
+ rH
CO 63
1 1C
1 T3 oj
I rH (1)
1 b U
1 n
=7
P Q) 0)
3 OS +J
O tri
"""^ P' " CL
"O HT3
3 T
C ] ^
-p 1 > B
,- 1 4) r-l
5 1 OS 0.
o
0
s_^ 1 0)
I CO
~-> 1 "0 ed
00 1 r-l CD
1 O i-
... 1 fc. 0
a) i c
(H 1 i
XJ '.
aJ <^
H =>
31 CP
£OJ -P
P cS
1 <
1
1
1
1 ~-~ CB
i • -P
1 > n!
1 CJ r-l
i a: ex
0)
a!
CJ rH
<0 SD 3
A c r)
— ' C ^
Q -rt
cn
0,
o >
. . jj f—t
a o
J r-l CO
§5 ri
C~ -I
^ CO CM
in in in
C- CM O *H
o cn v CM
m co CM
CD M CO
m cn o o
CM tO -H -H
CM -< -H
m in
cot-**
CM 00 CO
cn to f
in
t-t O *J* CO
c^ co r* *p
cn t* ^*
f- •* C-
to co -<
•^ CM i-H O
o
•H ~
A S
£1 U
a Q
03 to 1
cn 11
• o ^
* X CM
in cn to t^
O If C* 1 X
i • • i •
to i to o i 11
r-l 1 CM 1
* *
^ ,-N CO
cncM ^xx^ rccn CMCO <"o
• • •••• •• •• ••
inco CMC- CCOOOCM om COM xx
C-f> tOtO COCnO"* CM CM rtCO
rt f- X <-t CM CM
* *
^ ^ CM
CMTT tOCO tOCMCMtD "< O Xil r-IO
CM CM XX « tO ^J1 tO XO5 CM -1 r-4 O
tox too VC-CMOJ •* .110 coe-
CM^ oco CM^mm
rH -4 -~- ^
iif*o cnooin o^^c
cncncncoo c-toxo ^-rnt-xo o^cno ^C~CMP-^ r-ltOCM—l CMCOt»f->1 iH rH iHrlr-l^l COCM-Hil
c-t*c*CM in CMCMCMCM inm inmo mm
• ••• • •••• •• ••• ••
cocMincn •vmco toMcnx CMCMC» cnmii OOCMCM
cncnoo ^*coin t*-xxi-i ir-tt incoii
CMCMCM-H 1«COCM COPOCMrH
mco m mmininco mm mm mm
-Hocotoc* totomcM cocntomt- cooxm cnmocn toi-icacn
r^^cMCMrH toinc-cM cncnocoi-i CMCOIICM CMCMCMH r-mcoi-i
COCOCMH TCOCM COCOCOil
fl"O COCO-1 VOO'S'
tO^CM^CO tOinOO X^CMinO t-NNO r-ICOfl«O tO O CO O
COCMrH rHl-H^H tOCOCMr-1
ininmin in minm
t~*cMe- mocn >-H
rl CD 00
cMtoxf comcM cMtoxcn toxcn toxcn CM CD oo
mt-x^o mcMi-io cncncn^o -Hint^-o ^mc-o iiinc-o
CO ^H O O CO rH O ClrH O CO rH O CO rH O
_
-*^ S '•*'» i^*» **** t~*\
a cj s w -H B CMS CM x
CJ Q CJ CB 1 CJ 1 O 1 CJ
tOI tDI •*! TS^I ij
-------
408
J. HUISINGH ET AL.
X™ N
^3
rti
•3
a
T^
•P
C
n
w
o
s^/
00
.
I—I
^2
a.
EH
a>i
^
oj
•^ r-l
PJ 2
33 e •-<
hO -r^
C rH
O.
.
1) <
< a •
2 0 CO
1 ""•<
CO o3
^
CJ a
1 U
i a>
1 13 d
1 rH 0)
1 r >
1 C
• *~ '
S^"
>
30S -W
0 ai
J2 • rH
-^ 'Oft
H -a
1
1
1 — 11
1 > ai
1 0) rH
i as ft
1 0)
1 (0
) °O oj
1 rH CD
1 0 ^
1 Ct* U
1 C
t HH
1
<-^
s >
O 0)
— cc; -M
— ^ CTl
•tH * rH
S'Oft
I T!
1
I
1
1 -~- (U
1 • •*-"
i > a
t 1- rH
1 — ft
i
: c
IS
QJ i-r
W e^ 3
O S O
Q -H
L,
1 ft
1 O >
' M O
J I. CO
E -rJ rH
i < rH (B
i CO •« 3
f- 00
1 ^Jl f rH
1 1
1 rH 1 CM
1 1
CM^P CMrH oico coin
coin CMO coc- cnci
rH CO CO CM CO CO
CTJP- CJJOJ inCM OOrH
• • >• •* ••
coco rHco co^j* tnc~
^ c^ rH in CM in c^* »^
rH
roooo rnoOin t-tHoo ooOJC^
incDrHO cooco oomcMO ococoo
in in in in
OJt-CM COCMOJ C^Oin COCMCO
CO rH rH rH CO t- CO
,_t
in in inmin inin m inin in
^coooin coc^^ cDOjinoj c^-rHcocn
CD'tCMCM COCMCO ,-OCMCMrH OCJinrH
,_4
CMtnOJ rHOJCM inCOrH OOrH
corHino C^COCM ^cMrHO ootNino
in in in
coc^-^r cncocn ino^ cocno
co^rcM CMrH mrH inmin
-H CM rH
in m in
» * •
t-xm— . or-o COOOCMCO coc-moo
c^-ooco^f >-mtn oj^^co CTJOJOJCO
rH CM rH
CMCDOO OOO OC 00-^ in co
CO OJ O O
00 rH C- rH
rH rH
CO CO -H
C** rH CO
CO 1" CM
CO 1* OJ 00
rH in CO CO
t- -V CM
CO TJ- CM
CO 00 OJ O
CO rH O
C1
1 U
rH 1
rH rH
rH ^
CO CM
CO ^"
t^ CD
OJ CD
^ ^
m oo
CM CO CD
00 CM ^1 O
in in
rj< co 00
O CO CM
in m
^l1 CM 00 OJ
O) 00 'J" rH
CO ^1* CO
CO f- O O
CO CO CM rH
m
CO ^* OJ
CM O CO
CN rH
m
•
rH CM t~ CO
CO <3> t- CO
CM rH
CO CO •*
rH 0 0 0
•q> CM -H O
X
o
<-H rH
rH ^
00 CM
CM OJ
co co
co m
OJ CO
CM TJ"
CO CO
m o co
C- rH rH O
CD CO CM rH
in in in
CD CO OJ
<*p in CM
rH
m
CM oj m in
o f- m CM
rH
CM O CO
0 0 T 0
^" CM rH rH
m
TI* rH C^
CM TP rH
rH
in
•
in CM 00 rH
co oo in ^j*
rH
TP CM CO
CO OJ OJ O
CO rH O
rH S
1 O
CO 1
CO CM
rH ^
CO CM
c- co
co in
in rH
CO OJ
m c-
00 -H CO
co *• o O
C^ *& CO rH
in in m
Tf CD O
CM CO ^C
rH
in
•* CO O OJ
*^ 00 CD rH
rH
CD rH
m CM co o
CO CO rH rH
oj co co
O 00 CM
CM
P- TJI — 1 00
•
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS
409
•a
CD
3
C
—I
-P
C
O
o
00
CD
c*>
gj
EH
-M
id
a 3
a! «> 0
CO E -H
C t.
at
CM
.
CO <3
< O, •
SOW
1 r-l
OT 08
.
< a •
sow
+ rH
GO 08
a>
10
T3 at
rH Q)
O ^t
fa o
C
t ,
"7
30! •<->
O a!
s: • I-H
•P-i-iG.
*<
"^ *
> ed
0) r-H
a: a.
»
a]
•0 ed
b 0
HH
<-,
E >
.COS -M
-4-j fli
•H • i— I
at
CJ f-H
a: a.
0 I
M bfl 3
O £•• O
a -H
d
a.
o T
•-S5 rH
M O
>-4 IH CO
p. Q) |
S r^ rH
*r o> c-
CO CM
00 in rH Oi
Cl 00 to Cl
rH in *H ^J*
tO C^ CJJ rH 05 O
a> m o CM in CM o
rH rH rH CM rH rH rH
rt* in in o in in ao
CM rH CM rH
in co in
oooom cOrHco<35
lO^fCMCM ^COCMrH
CO rH ^* CO CO f™
CMintMO rHCMrHO
CMrHrHrH CMrHrHrH
in in
CM rH O CO O tO
mCM rH <*
uj in
COCMrHrH rHC-TfQO
cstomTj* ooi'i'co
co T c^ oo •* t~
tocotoo tocotoo
CM rH O CM rH O
l~\ t^
CO rH 35 CM 33
eg 1 U 1 CJ
"O CM 1 CM 1
<]) Tf CM ^1< CM
1- OOCM 00 CM
D ^-^ •*-• •
S
-------
410
J. HUISINGH ET AL.
9OOr
VW 241 DCM
VW 241 CH
MERC 241 DCM
MERC 241 CH
W/0 Activation
W Activation
.2
.4
.8
1.0
1.2
mg. Organic Extract
Figure 9. Comparison of the mutagenic response of organics
extracted from diesel particulate with cyclohexane (CH) and
dichloromethane (DCM) in Salmonella typhimurium strain TA1538.
increase in toxicity with storage. If the linear portions of
the dose response curves are compared, the effect of storage
is minor. The major differences -are apparent at higher sample
concentrations where the differences may be due to toxicity
factors. In no case did mutagenicity increase with storage.
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS
411
1400
1200
1000
a.
cc
01
a.
to
soo
uu
cc
cc
co 600
5
D
Z
CC
400
200 —
-O FRESH EXTRACT. 1 day. *MA
-Q FRESH EXTRACT, 1 day. MA
. + FILTER STORED. 8 weeks. »MA
.0 FILTER STORED. 8 weeks. MA
-& EXTRACT STORED. 8 weeks. »MA
-V EXTRACT STORED. 8 weeks, MA
+ MA
0.2 0.3
Mg RELATED PARTICUUATE
0.4
0.5
Figure 10. Effect of storage on the mutagenicity of organic
extracts from diesel particulate when bioassayed in Salmonella
typhimurium strain TA1538.*
-------
412 J. HUISINGH ET AL.
Under specific testing conditions a comparison can be
made between the different fuel types and engine types.
Diesel particle samples which had been extracted and analyzed
for BaP as part of a fuel study with diesel passenger cars
(22) were selected for bioassay. The samples chosen repre-
sented five fuels, two vehicles and the widest possible range
of BaP values. These samples were all extracted with cyclo-
hexane (CH), solvent exchanged and bioassayed under identical
conditions. The data from these samples are shown in Table 9
and summarized in the histogram in Figure 11. In both vehi-
cles, the minimum quality fuel (241) resulted in emissions
with the highest tnutagenic activity. The BaP concentration
of the emissions was also highest with this fuel (241). Un-
der these engine testing modes and with the cyclohexane ex-
tract being used, the VW engine generally created a higher
specific activity than did the Mercedes engine. Since these
vehicles were tested with only one testing mode and one sol-
vent system, these results may not represent a true compari-
son of emissions characteristic of the engine being tested.
This histogram could change markedly with a change in solvent,
engine testing mode, engine type, or fuel characteristics;
however, it does demonstrate that a variety of parameters
influence the mutagenic activity.
SUMMARY
In contrast to some complex mixtures which are too toxic
to be bioassayed for microbial mutagenicity prior to fraction-
ation, e.g., synthetic fuel (23), organic extracts of diesel
particle emissions were found to be mutagenic in the Salmo-
nella typhimurium plate incorporation tests without fractiona-
tion. The mutagenic response is dependent on the organic
solvent employed to extract the particulate. The cyclohexane
extraction and benzo-(a)-pyrene analysis procedures developed
for ambient air particulate was applied to diesel particulate
emissions. Selected cyclohexane extracts, after solvent ex-
change were bioassayed directly in the plate incorporation
Salmonella typhimurium/microsome mutagenicity bioassay.
Diesel particulate emissions from light duty passenger
cars were found to have a wide range of both benzo-(a)-pyrene
content and mutagenic activity. The results from this pilot
study indicate that both methodologies are applicable to
evaluation of diesel particulate emissions. Variables which
affect these determinations, such as the extraction solvents
employed and the method of storage, need to be optimized and
subsequently standardized. This is particularly important
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS
413
00
CO
LO
i
CD ^C
G £-<
•H
tiD G
G -H
H ccj
M
G -P
cci 00
CO
CQ G
•H -H
-P
erf CQ
JU
O
•H G
5*- o
•H
CQ -P
0 aj
t~i rH
a o
£ CX
TO rH
02 O
C5 O
-P S
CD CO rH
j rj
rQ Crf CD
aj .C -P
E-< X aj
W r— i
pi (
r— |
CD S
CQ 3
CD -H
•H JH
Q 3
£
' ' '"— '
O JS
a
CD >>
tuO-P
aj
in cri
rH '•y
O r- 1
_1_1 ^
•i-^ ft
02 CD
C
01
OS
•a
o
fc<
•
"1 "~T-
S >
0)
£ OS
p
•H •
•o
<
• — „
•
>
01
i a;
CO
o
•H
C
SO
o
0)
to
B
,
0
•• 2
Cd
rH rH
a, cu
*£ P
*IJ rH
CO -rt
ct,
<
•
CO
00
*o
•
ai
&.
•
SI
e
rH
"7
>
r-t
0
CO
i
rH
01
"J
b.
in IH
cn in
co cn
iH *H
rH ®
CO CO
co in
CM CM
in co CM
1 CO t- rH O
I in CM CD i-t
in in co
i cn t- •* o
1 CM Cn rH
i cn oo co
1 CD -V i-l C-
i cn cn CD
O •* CM CM
O 00 C^ CO O
•fl< (35 O CM rH
CO TJ* ^ CM
m -H oo oo o
CM CM -H in
ao CM o m
iH 1-t
O CD CO CO in
in f ^< oo CM
oo CM o m
rH rH
m
m CM
o m CM IH o
rH O O O O
o o o o o
o o o m
•<3< CM rH
P
O
LH
•p
X
H >.
C as
cS £ Q
CO CO
CO S 01 01
"-t CJ .H q
Z C ^, O
rH f-
cn co
00 rH
iH
CM CO
CM iH
ffl rH
rH CM
o o m o
co cn CM cn o
r> co co ^ *H
CM CO CM -H
co co m oo o
CM m o t-
m co m CM
co co « co o
1" t- CM C» CM
in CD m CM
c^ cn 0s" co
co o co m o
•H in CO 00 rH
CO CM rH
CM m o o o
rH cn in tn
o t~ m CM
rH
m oo co co co
^ CM 00 CO CO
o oo m CM
rH
m co
m CM CD
in CM rH O O
o o o o o
o o o m o
o o m CM
OJ rH
•a
0)
IH CO
0 *^
P 01
co o)
*
I*
01 P
p .a
s -H ao
CJ -H -rH
Q Cu Cd
CO O
co cn
O rH
CM CM
cn oo
^ CO
00 O
-H CM
in o o m
in CD rH oo o
cn oo CM ^ »H
CM CM -H
rH CM CM > O
t> in CM c-
r-i m i
4J
01
C
•H
SO
C
O)
>»
o
T3
0)
to
3
CJ
rH
a
S.
to
CU
p
to
01
£>
•rH
0
to
•a
rH
C
E
3
rH
0
CJ
rH
0)
ai
rH
a
£_*
a
to
P
c
ai
p
^4
Q)
01
*"
o
^
11
f
C
so
ai
M
01
>
01
p
at
• rH
01 a.
rH *^^
a •
S >
at 01
CO C£
CM
,_
CU
a
CO
P
a
ai
P
SH
CU
>
CU
rH
CO
3
O
01
ai
p
C
O
a
to
to
3
C
•H
£
0)
p
a
rH
a
CO
4~>
C
ai
p
0)
j>
01
rH
""'
01
p
aj
rH
a
L>
o
a
to
•P
C
oi
tl
CU
CU
tn
01
S
3
a
•o
01
CO
3
'"-J
T3
01
xi
rH
a.
•
>
01 •
/o7
• p
•T a!
"O rH
< a
CO
^
01
P
ai
rH
a
•P
cri
In
CU
0)
rH
CO
3
0
0>
C
crj
p
C
0
a
CO
^^
^-^
CU
ai
rH
a.
to
P
ai
p
CU
0)
as
V— '
CU
CO
ai
01
!H
O
C
rH
n
rH
0
OH
•
C
o
to
CO
CU
rH
so
rH
FH
OS
CD
C
•rH
rH
C
0
•o
01
tn
>
•H
>
•rH
P
o
<
O
-------
414
J. HUISINGH ET AL.
400 -
300
o
f 200
t/5
2
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 415
Storage of samples, either on filters or as extracts
resulted in slight decreases in specific mutagenic activity
when the linear portions of the dose response curves were
compared. The absolute number of revertants per plate at
higher, non-linear concentrations was more clearly reduced
after storage. The loss in mutagenic activity was more pro-
nounced in samples tested without metabolic activation and
may have been due to an increase in direct acting toxic com-
pounds as a result of sample storage.
Although sample replication was limited in this pilot
study, it appears that the mutagenicity of the particulate
emissions is influenced by the fuel and to a lesser extent
by the vehicle. Both the BaP content and the mutagenic
activity of the emissions were the highest when the minimum
quality fuel (241) was used. This fuel has the lowest cetane
value, highest aromatic content, and highest nitrogen content
of the five fuels compared. The relationship of mutagenic
activity to other fuel variables is being explored.
Sample 1804 with the highest BaP content (26.5 ng/mg
particulate) provided only 0.1 ug BaP per plate; therefore
this sample would not contain sufficient concentrations of
BaP to be detectable in the mutagenesis bioassay (the mini-
mum detectable limit in the plate incorporation assay is
approximately 1 ug per plate). As described in Part I of
this paper, the most mutagenic fractions were not the frac-
tion (ARM) in which BaP would be found. Nevertheless, the
sample containing the highest concentration of BaP was
found to have the highest mutagenic specific activity, sug-
gesting that BaP may be a useful indicator chemical.
ACKNOWLEDGMENTS
The authors wish to acknowledge the following technical
support: D. Swanson, R. Hedgecoke, and C. Morris for the
assistance in benzo-a-pyrene analysis; J. Hein for the anal-
ysis of physical properties of fuel; P. McBride and H.G.
Shan for technical assistance in microbial mutagenesis; T.
Baines and SWRI personnel for the light duty diesel particu-
late samples from which the exhaust particulate samples were
obtained for the light duty pilot study Part II.
-------
416 J. HUISINGH ET AL.
REFERENCES
1. Springer KJ, Asby HA: The low emission car for 1975—
enter the diesel. SAE Paper No. 739133, Philadelphia,
PA, August 1973
2. Springer KJ, Stahman RC: Emissions and economy of four
diesel cars. SAE Paper No. 750332, Detroit, MI, Feb 1975
3. Braddock JN, Bradow RL: Emissions patterns of diesel-
powered passenger cars. SAE Paper No. 750682, Houston,
TX, June 1975
4. Hare CT, Springer KJ, Bradow RL: Fuel and additive
effects on diesel particulate-development and demonstra-
tion of methodology.
5. Braddock JN, Gabele PA: Emissions patterns of diesel-
powered passenger cars—Part II. SAE Paper No. 770168,
Detroit, MI, Feb 1977
6. Springer KJ, Baines TM: Emissions from diesel versions
of production passenger cars. SAE Paper No. 770818,
Detroit, MI, Sept 1977
7. Springer KJ: Investigation of diesel-powered vehicle
emissions VII. EPA Report No. EPA-460/3-76-034, Feb 1977
8. Hare CT: Characterization of diesel gaseous and particu-
late emissions. Final Report on EPA Contract No. 68-02-
1777, Sept 1977
9. Beltzer M, Compion RJ, Petersen WL: Measurement of
vehicle particulate emissions. SAE Paper 740286, 1974
10. Begeman CR, Jackson IW, Nebel GJ: Sulfate emissions
from catalyst-equipped automobiles. SAE Paper 741060,
1974
11. Swain AP, Cooper JE, Stedman RL: Large scale fraction-
ation of cigarette smoke condensate for chemical and
biological investigations. Cancer Res 29:579-583, 1969
12. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian
microsome mutagenicity tests. Mutat Res 31:347-364, 1975
-------
CHARACTERIZATION OF DIESEL PARTICLE EMISSIONS 417
13. McCann J, Choi E, Yamasaki E, Ames BN: Detection of
carcinogens as mutagens in the Salmonella microsome
test: Assay of 300 chemicals. Proc Nat Acad Sci USA
72:5135-5139, 1975
14. Ames BN, Gurney EG, Miller JA, Bartsch H: Carcinogens
as frameshift mutagens: Metabolites and derivatives of
2-acetylaminofluorene and other aromatic amine carcino-
gens. Proc Nat Acad Sci USA 69:3128-3132, 1972
15. Ames BN, Lee FD, Durston WE: An improved bacterial test
system for the detection and classification of mutagens
and carcinogens. Proc Nat Acad Sci USA 70:782-786, 1973
16. Ames BN, Durston WE, Yamasaki E, Lee FD: Carcinogens
and mutagens: A simple test system combining liver
homogenates for activation and bacteria for detection.
Proc Nat Acad Sci USA 70:2281-2285, 1973
17. McCann J, Spingar NE, Kobori J, Ames BN: Detection of
carcinogens as mutagens: Bacterial tester strains with
R factor plasmids. Proc Nat Acad Sci USA 72:979-983,
1975
18. Swanson D, Morris C, Hedgecoke R, Bumgarner J, Jungers,
R: A rapid analytical procedure for the analysis of
benzo(a)pyrene in environmental samples, in press.
EMSL, MD-78, Research Triangle Park, North Carolina
19. Human population exposure to coke oven atmospheric emis-
sions, pp 64-67, EPA draft report, OAQPS (J Manning, MD-
12), US Environmental Protection Agency, Research Tri-
angle Park, North Carolina
20. Human population exposure to coke oven atmospheric emis-
sions, p 48, EPA draft report, OAQPS (J Manning, MD-12),
US Environmental Protection Agency, Research Triangle
Park, North Carolina
21. Bradow RL, Moran JB: Sulfate emissions from catalysts
cars—A review. SAE Paper No. 750090, 1975
22. EPA Contractor 68-02-2417 with Southwest Research
Institute
-------
418 J. HUISINGH ET AL.
23. Epler JL, Young JA, Hardingree AA, Rao TK, Guerin MR,
Rubin IB, Ho CH, Clark BR: Analytical and biological
analysis of test materials from the synthetic fuel
technologies. I. Mutagenicity of crude oils determined
by the Salmonella typhimurium/microsomal activation sys-
tem. Mutat Res, in press
-------
MEASUREMENT OF
BIOLOGICAL ACTIVITY OF
AMBIENT AIR MIXTURES
USING A MOBILE
LABORATORY FOR IN SITU
EXPOSURES: PRELIMINARY
RESULTS FROM THE
TRADESCANTIA PLANT TEST
SYSTEM
L.A. Schairer and J. Van't Hof
Biology Department
Brookhaven National Laboratory
Upton, New York
C.G. Hayes and R.M. Burton
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina
Frederick}. de Serres
National Institute of
Environmental Health Sciences
Research Triangle Park, North Carolina
-------
421
A variety of short-term bioassays has been developed to
assess the mutagenicity of industrial chemicals. Many of
these assays work well when used under laboratory conditions
but are not suitable for monitoring ambient air under field
conditions. To facilitate exposures of biological systems
to ambient air pollution in natural or industrial sites a
plan was implemented to design, assemble, and test a mobile
laboratory. The Tradescantia plant test system was chosen
for these initial field studies because of its high sensi-
tivity to both physical and chemical mutagens and its versa-
tility and adaptability to monitoring the mutagenicity of
gaseous pollutants. Positive results to date support the
further development of the mobile laboratory and Tradescantia
system as a useful method for monitoring biological activity
of complex environmental mixtures in situ.
Several species of the family Commelinaceae, of which
Tradescantia is a member, have features particularly well
suited for certain radiation and chemical mutagen studies.
The effects of chemicals and/or ionizing radiation that are
easily measured include the following:
• Chromosome aberrations in microspores, root tips,
and stamen hairs
• Somatic mutations in petals and stamen hairs in
clones heterozygous for flower color
-------
422 L.A. SCHAIRER ET AL.
• Pollen abortion
• Cell sterility in stamen hairs
Of the four features mentioned, somatic mutation in stamen
hairs is the most versatile as it requires the least compli-
cated techniques and is more sensitive than the other end-
points to both physical and chemical mutagens. The pattern
and magnitude of response of phenotypic changes in pigmenta-
tion in stamen hair cells have been studied after treatment
with X rays (10), gamma rays (4,8), 3H-8 rays (Schairer LA,
unpublished data), nitrogen ions (12), monoenergetic neutrons
(13), and low gravity of space flight (7). X-ray and neutron
dose-response curves as well as those for chronic gamma expo-
sures show straight-line relationships over wide dose ranges
with no evidence of a threshold dose even at levels as low as
250 mrad of X rays, 10 mrad of 0.43 MeV neutrons and 33 mR/h
of *37cesium gamma (8,10).
The significant mutagenic response to an accidental
exposure to a gaseous chemical (5) as well as the high
radiosensitivity were factors that prompted the use of
Tradescantia as a test system to assay for the mutagenicity
of various chemicals and air pollutants (9,11). Newly devel-
oped chemical exposure and dosimetric techniques verified
the high sensitivity of the Tradescantia stamen hair system
to gaseous chemical mutagens and these demonstrated its
potential for monitoring ambient air pollution for mutageni-
city (2,6,9).
Individual compounds or air pollutants can best be
studied in the laboratory, but the mutagenicity of unusual
and even unique ambient mixtures in urban or industrial sites
must be assayed in the field. Perhaps the greatest advantage
the stamen hair system affords over other test organisms is
its versatility and adaptability to field studies.
THE TRADESCANTIA STAMEN HAIR SYSTEM
The stamen hair system has been described in detail
elsewhere (1,11) so only certain features will be reviewed
here. The plant used exclusively in the field studies to
be described here is clone 4430, an interspecific hybrid
(T_. subacaulis x T. hirsutiflora) produced at Brookhaven
(Figure la). This clone is a hybrid between pink- and blue-
flowering parents with blue being dominant over pink. The
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES
423
\ 1 44| I | I
05
10cm
Figure l(a). Normal stock plant of Tradescantia clone 4430
showing several mature inflorescences.
-------
424 L.A. SCHAIRER ET AL.
visible marker used in this test system is the phenotypic
change in pigmentation from blue to pink in mature flowers.
The pigmentation change (hereafter called mutational or pink
events) is induced in young developing floral tissue and is
expressed 5 to 18 days later as isolated pink cells or groups
of pink cells in the stamen hairs of mature flowers (Figure
Ib, c). The pink events are essentially nonlethal so large
mutant sectors indicate genetic injury early in the develop-
ment of that tissue.
The stock plants are easily maintained by vegetative
propagation and flower continuously throughout the year in
controlled-environment growth chambers. The material treated
consists of a group of unrooted, fresh cuttings containing
young inflorescences which contain flower buds in a range of
developmental stages as shown in Figures Ib and 2. Following
exposure to either chemical or physical mutagens, the cuttings
are grown in aerated Hoagland's nutrient solution under stan-
dard conditions and the flowers are analyzed each day as they
bloom for approximately three weeks after treatment. Induced
pink-event rates are expressed as the mean of the rates for
several consecutive peak response days, usually days 11 to 15
for acute X rays and 7 to 12 for acute chemical exposures
(Figure 2). Detailed descriptions of laboratory techniques
for radiation and chemical exposures and calculating mutation
rates are given elsewhere (7,9,11). The only modification
that has been adapted in the scoring method is that any inter-
rupted series of pink cells within one hair is considered to
be the result of a single mutational event (1) . This conser-
vative approach has only a slight effect upon the mutational
frequency at the levels described in this paper. The tech-
niques for field exposures are new and, although described
briefly by Schairer et al. (3), they are reviewed below.
THE MOBILE MONITORING VEHICLE
The vehicle selected for the mobile monitoring project
was a 24-foot Clark mini-van trailer. The trailer shown in
Figure 3 was insulated and air conditioned to permit year-
round operation of the laboratory. In order to maintain a
semiclean environment for these studies, the trailer air was
recirculated through activated charcoal and HEPA particulate
filters. Three Model M-13 growth chambers (Environmental
Growth Chambers, Chagrin Falls, Ohio) were installed. One of
the chambers serves as a clean air control, the second is used
for ambient air exposure and the third is used as a backup
unit for either control or ambient air exposures (Figure 4).
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES
425
from
1"Y> O •*- •> ' V»
, A 1 d. '- U j.
Figure i(c). Enlargement of
stamen hairs with pink mutant
events indicated by shading.
Mutant events in the flower
color locus are not usually
lethal; chains of pink cells
represent daughter cells of
the initial mutated cell.
-------
426
L.A. SCHAIRER ET AL.
E e E
a. a. a.
a. a. a.
CO -P
-P C
C CD
CD £
^ > -P
3 CU erf
O CD
I G -P
CD T-(
OH 'H
bO O
C
•H • CL)
> CD E
0 r-l -H
rH 3 -P
rH 03
O O -P
SH a erf
X
CO CD CO
SH CD
•H JH N]
etf CD -H
s: -P co
C erf T3
CD 3
E CO £3
erf >>
-P erf bC
CO T3 C
erf
C
erf
O
CO
CD
•a
erf
CXI T3
r-l C
I O
CO
CO CD
r-* O
erf O
CD
G erf -P
•H 0)
a •-
CO CO
o> .c f-i •
*-> O CD erf
erf erf S CO
H 0 O CO
M rH •!-(
C «-( O
o c/3 co
-P iJ in erf
erf 3
-P CO -P D
3 ^} erf £
£ O £ -P
CD
^ CO G bfi
C oj —i G
•P bfl O
C O erf
-P S3
>> C
CXI CD r-l S
?- erf C
CD
- O
erf 'SI
O CD
« M
erf a
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES
427
a;
.c
P
T3
bD G
C aj
•H
£ rH
O 0)
.G rH
CO -H
cd
^~*> fn
C? "^
S 0)
(1)
rH ^
O O
•H -H
erf
hC t-\
a +->
•H
fn ^
O 0)
4-> -P
•H rH
C -H
O
-P -P O
•H tH
"H T3 bD
O G
O 0
-POX:
G -P
O f-i
rH -H rH
bO£
0) G O
•H -H X
> P 0
C
3 -W
• O Oj
n E 0
0) 0) ""
tn P 0)
3 O 0)
be e M
•H 0) J=
SH f-i -P
-------
428
L.A. SCHAIRERET AL.
Figure 4. Interior of motai]e monitoring vehicle (MMV) show-
ing rear exposure chamber (with cuttings) and a control cham-
ber on the right. Round air filter cannisters are mounted on
brackets above chamber door.
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES 429
Ambient air is drawn into the fumigation chamber through a
four-inch glass duct at continuous flow rates up to about
18 cubic feet per minute, a maximum of one air change every
two minutes. Each chamber is equipped with an air filter
train composed of activated charcoal and HEPA particulate
filters. This filter train is used to scrub the air contin-
ually in the chamber serving as the concurrent control. The
total external electrical power requirement for the trailer
air conditioning and chamber operation is a 100 amp, 220 volt
service.
FIELD EXPOSURE TECHNIQUE
Field exposures were accomplished in the following man-
ner: fresh cuttings of Tradescantia clone 4430 were made
from stock plants grown in controlled environment chambers at
Brookhaven National Laboratory; they were hand-carried to the
test site by car or airplane; cuttings were placed in the
chambers in glass containers filled with Hoagland's nutrient
solution, and exposures were made for a ten-day period. At
the end of the exposure the cuttings were taken back to
Brookhaven National Laboratory for posttreatment analysis of
the flowers as they bloomed each day. The peak mutation
response period following a ten-day exposure is 11 to 17 days
after the start of the exposure. The mean of the mutation
rates for the seven-day scoring period resulted in an observed
rate for a given test site based on an average stamen hair
population between 300,000 and 400,000. A population of 300
cuttings in each ambient air and control chamber will yield
enough data to resolve as small as a 10% increase in pink
events over the background frequency.
CHEMICAL EXPOSURES UNDER LABORATORY CONDITIONS
Exposures to a standard chemical mutagen, the alkylating
agent 1,2-dibromoethane (DBE), in the gaseous state, showed
that the number of mutational events increased linearly with
the product of concentration and hours of exposure to DBE, at
least over the range from 2 to 144 hours. These data may be
expressed in terms of total dose by plotting induced mutation
frequency against the product of concentration (ppm) and dura-
tion of exposure (hours) (Figure 5). For purposes of compari-
son, a standard curve for X-ray effect is shown in rads.
Slope and shape of the curve for DBE induction of color change
resemble those for radiation injury.
-------
430
L.A. SCHAIRER ET AL.
100^
03
si
\ 2
O) O
Z: O
i
0.
o
0.01
I [ I I 11 ll| I I I I i i IT|
CLONE 4430
DBE EXPOSURE
o 2 hr o 4hr
*6hr «!2hr
• I8hr *30hr
x 48hr » !44hr
irn
II Mill
CLONE 4430
250-kVpx RAYS
10 100 1000
CHEMICAL CONCENTRATION
(ppm) x TIME (hours)
I L_l_i I I I III I
10 100 1000
RADIATION DOSE (rods)
Figure 5. Stamen hair mutation frequencies from several
experiments are plotted against total dose of 1,2-dibromo-
ethane (DBE) (ppm x hours of exposure). A linear response
curve fits all data points from 2- to 144-hour exposures.
The standard acute X-ray curve is shown for comparison.
Although a large percentage of the effort of this group
has been spent on the development of the mobile monitoring
vehicle, a number of chemicals have been tested in the labo-
ratory to validate the system as a monitor for gaseous muta-
gens. Typical dose-reponse curves for several chemicals
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES
431
are shown in Figure 6. Chemicals such as the gasoline addi-
tives 1,2-dibromoethane (DBE) and trimethyl phosphate (TMP)
were found to be potent mutagens while SO2, N02, vinyl
chloride, and freon-12 were weak mutagens according to this
test system. Other chemicals or air pollutants tested are
listed in Table 1. The concentration listed is the lowest
value tested that showed a significant mutagenic response.
CO
cr
<
o
o
V.
CO
Z
O
0.5
0.2
O.I
X rays
-50 rods
1,2 -Dibromoethane
(DBE)
Trimethyl
Phosphate
(TMP)
10 rods
CONTROL
Freon-12
\
Vinyl Chloride
j I
i i
5 10 20 50 100 200
GAS CONCENTRATION (ppm)
500
Figure 6. Typical dose-response curves for pink events in
Tradescantia clone 4430 are shown following 6-hour exposures
to various gaseous compounds.
-------
432
L.A. SCHAIRERETAL.
Table 1
Summary of Mutation Response Data for Various Chemicals
Used on Clone 4430 in Terms of Lowest Concentration
Giving Significant Effect
Chemical
Air Pollutants
Ozone (O3)
Sulfur Dioxide (S02)
Nitrogen Dioxide (N02)
Nitrous Oxide (N^O)
Industrial Chemicals
Ethyl Methanesulfonate (EMS)
1 ,2-dibromoethane (DBE)
Trimethylphosphate (IMP)
Trichloroethylene (TCE)
Vinyl Chloride (VC)
Vinylidene Chloride (VDC)
Vinyl Bromide (VB)
2-Bromoethanol (2BE)
Freon-12 (Fr-12)
Freon-22 (Fr-22)
Hexamethylphosphoramide (HMPA)
Benzene
Caffeine
Atrazine
Sodium Azide
1 , 1-dibromoethane
Dimethylamine Hydrochloride
Vapona
Exp.
Time
(hr)
6
6
6
6
6
6
144
6
6
6
24
6
24
24
6
6
6
6
6
Chronic
Chronic
3
6
2
6
Mm.*
Cone.
(ppm)
5.0
40
50
250
5
1
0.14
13
0.5
75
25
86
22
50
24
392
194
9
4000
10~4M
0.045g/pot
10~4M
58
10"2M
Sat?
Hairs
Scored
(xlO3)
48
41
24
29
20
258
148
32
44
34
56
30
100
49
33
32
66
48
43
39
93
19
56
16
81
Total
Pink
Events
153
170
87
115
246
1088
1119
115
148
133
281
130
338
201
131
103
249
314
292
142
260
96
219
83
278
Pink Events
per 100
Hairs
(-Control)
.098 '
.222
.112
.117
1.012
.118
.315
.125
.112
.112
.151
.064
.057
.159
.107
.095
.100
.277
.287
.047
.0
.269
.073
.151
.0
i SE
.040
.041
.056
.055
.133
.027
.035
.051
.036
.046
.041
.056
.028
.048
.046
.059
.039
.051
.063
.040
.0
.055
.039
.080
.0
Stat.
Sig.
2%
1%
5%
1%
1%
1%
1%
2%
IS
2%
1%
Insig
5%
1%
2%
Insig
2%
1%
1%
Insig
Insig
1%
Insig
Insig
Insig
•Minimum concentration used which showed a significant increase over background mutation rate.
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES 433
RESULTS OF EXPOSURE TO AMBIENT AIR POLLUTION
The first field trials for the mobile monitoring vehicle
(MMV) were conducted in the summer of 1976. A location was
sought which had high levels of a mixture of pollutants and
was within about a two-hour drive from Brookhaven National
Laboratory.
The first test site selected was Elizabeth, NJ beside a
NJ air pollution monitoring station. The NJ Turnpike, toll
plaza, petroleum refineries, Newark Airport, and other indus-
trial pollution sources surrounding this test site are shown
diagrammatically in Figure 7. When two-week exposures were
made in July and October 1976 and January 1977, the data
indicated increases in mutation frequencies, following expo-
sure to ambient air, which were significant at the 1% level
for all three periods (Table 2). In the third two-week
exposure, January 1977, two chambers were exposed to ambient
air to demonstrate that the induced effects observed in the
previous two runs were real and not a unique chamber effect
in the third control chamber. Data from the ambient air
samples were not different from each other, but both were
significantly higher than the concurrent control. Apparently
no unique chamber effect exists between chambers, even under
field conditions.
Wind direction is an important factor in the location
of a mobile monitoring unit. The high induced mutation rate
in July occurred with prevailing southwesterly winds, while
the October run had prevailing northwesterly winds (Figure
7). Pollution sources were certainly different in these two
exposures, but a much more sophisticated air monitoring
facility and a detailed map of industrial and natural pollu-
tion sources in the greater Elizabeth area would be required
to identify the environmental mutagen(s) and its probable
source.
These data were encouraging and supported the use of
the Tradescantia test system as a field monitor for air pol-
lution. To continue the study, a series of exposures was
planned in collaboration with the U.S. EPA Epidemiology and
Measurements Sections. Test sites were selected because of
high cancer mortality or presumed exposure to high levels
of carcinogens. The MMV experiments were to look for bio-
logical activity, while an EPA mobile monitoring van made
real-time measurements of the pollution levels. Organic
vapors were collected on Tenax absorbers for subsequent
identification by Dr. Edo Pellizzari. The sites selected
-------
434
L.A. SCHAIRER ET AL.
s>»
\
*»• \
CO bfl
0 S3
O -H
!H ?H
3 O
O -P
CO -H
c
0) O •
£ S CO
-p -p
^ sa
, fc
£ O (P
O CO O,
0 0
o
0 -p
C
•H
CO OJ O
T3 «H
OJ -H
-P W G
•H 0) >
CO X2 O
X!
-p -a co
CO CD
CD -P a>
-P erf !H
O aj
>> O
CD r-l CO
CO T3
^ CO C
CD (A -H
"-3 > BS
CD
C
-H
> O 0
0 -H £H
•^ -P a,
> s
rH
O rH •
•H O ^
4^ 0i fe
erf O
£ ^ fc
bfl-P 0)
(rf C 0)
•H »
• oj fn
O O
X! -P
0) bO «J
S-l -H ^
3 X! O
bC ^3
•H
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES
435
Table 2
Mutagenicity of Ambient Air at Elizabeth, NJ
as Measured by Tradescantia Stamen Hairs
Treatment
No. No. No. Pink
Flowers Hairs Events Events/Hair +_ S.E,
Control
Ambient Air
7/20-8/3/76
726 299,475 1182
658 268,464 1386
Ambient Air Minus Control
.00395 + .00013
.00516 + .00016
.00122 + .00021*
Control 892 350,824 1487 .00424 + .00012
Ambient Air 890 358,047 1727 .00482 + .00012
9/27-10/11/76 Ambient Air Minus Control .00058 + .00012*
Control (1) 689 266,023 872
Ambient Air (2) 742 291,161 1146
Ambient Air (2) Minus Control
Ambient Air (3) 617 231,557 873
Ambient Air (3) Minus Control
Ambient Air (2+3) 1359 522,718 2019
1/21-2/4/77
Ambient Air (2+3) Minus Control
.00328 + .00012
.00394 + .00013
.00066 + .00017*
.00377 + .00014
.00049 + .00018*
.00386 + .00009
.00058 + .00015*
*Significant at the 1% level.
-------
436
L.A. SCHAIRER ET AL.
for this phase of the study were: Charleston, WV, Birming-
ham, AL, Baton Rouge, LA, Houston, TX, Upland, CA, Magna, UT,
and Grand Canyon, AZ. The latter site at Grand Canyon served
as a clean air control study.
The results of these field exposures are summarized
graphically in Figure 8. The pollution sources indicated
here are only general categories under the heading of the
major industries in the areas and do not imply a known cor-
relation between mutation response and specific industrial
effluent. Statistically significant increases in mutant
event frequencies above control levels were observed at
Elizabeth, Charleston, Baton Rouge, and Houston. The
MUTAGENICITY OF AMBIENT AIR AS MEASURED BY TRADESCANTIA IN THE MOBILE MONITORING VEHICLE
X
I
10
o
^
z
UJ
UJ
t-
|
5
^
Z
a.
POLLUTION SOURCES
PETROLEUM & AUTOMOTIVE STEEL PETROLEUM
T CHEMICALS PETROLEUM
-
T
I
\-
"i
\
\
\
\
\
^
t
i
\
\~
\
\
\
\
\
\
\
\
\
i
r
"~~[Ti N
N
\
\
\
T \
jl \
i
T
-fc
\
\
\
\
\
\
\
\
\
\
t
.L
T
^<
\
\
\
\
\
\
\
k
t
17
\
\
\
\
\
\
\
\
i
CONTROL ^^
AMBIENT AIR 1
t s
E. ^4-^
COPPER
AUTOMOTIVE 3
PHOTOCHEMICAL
K
\
\
\
\
\
K
1976 1977
T
t
i
8
\
\
\
\
\
^
T
T
1
\
\
\
\
\
\
\
\
\
\
CLEAN AIR
CONTROL
MEAN ~
CONTROL
y
T J-lT
ski Eli
$ti ^
n 11 .
Aug. Oct. Jan. Mar. Apr. May July Aug. Sept. Oct. Nov. Dae.
ELIZABETH CHARLESTON BATON ROUGE UPLAND UPLAND GRAND CANYON
NJ WV LA CA AZ
BIRMINGHAM HOUSTON MAGNA
AL TX UT
Figure 8. The mutagenicity of ambient air as measured by
Tradescantia in the mobile monitoring vehicle is summarized
for the eight test sites visited.
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES 437
remaining locations, especially the clean air site at Grand
Canyon, showed no significant response to ambient air.
These data are also shown in Table 3 arranged by pollution
source and presented as a pollution-induced increase in
mutation rate as percent of control. Locations associated
with petroleum refining and mixed chemical processing gave
increases ranging from 31% down to 17%. The real-time mea-
surements of both organic and inorganic compounds are being
analyzed at the present time, and when completed, these
results may provide more specific identification of compounds
common to those sites showing induced mutations. If suspect
compounds are identified they can be tested individually
under controlled laboratory conditions using existing tech-
niques .
It should be emphasized that a negative response in a
single exposure of a test organism may provide inadequate
assurance of absence of a health hazard. As pointed out in
the Elizabeth experiment, the prevailing wind direction
changed from summer to fall and the induced mutation frequency
dropped from 31% to 18%. Wind direction, amount of precipita-
tion, industrial complex work schedule, etc., all have a
direct bearing on the pollution mixture and level at a fixed
monitoring location.
CONCLUSION
The body of evidence is growing for a meaningful extra-
polation from cytological and genetic effects in microorgan-
isms, cell cultures, plants, insects, and mammals to health
hazards in man. The high correlation between mutagenicity
and carcinogenicity supports the use of visible genetic mark-
ers in test organisms as monitors for carcinogens. The
observation of similar chromosome aberrations in both gametic
and somatic tissues gives cytological evidence for the effec-
tiveness of somatic mutation markers as an assay for chemical
mutagenicity and hence health hazard potential. The Trades-
cantia stamen hair system encompasses the cytogenetic and
somatic potential to make the system a useful tool for muta-
genicity monitoring of ambient air pollution mixtures or iso-
lated fractions. This plant is uniquely adapted to field ex-
posures, hardy enough to tolerate a broad range of environmen-
tal conditions, and requires no elaborate sterile culture
conditions. The data presented above demonstrate the high
sensitivity of the system to gaseous compounds and the rela-
tively short time from start of exposure to definition of re-
sults (3 weeks). In the absence of hard genetic evidence for
-------
438
L.A. SCHAIRER ET AL.
CO
03
p
CO
rH
crf CO
•H I*
!H -H
P Crf
co as
3
-a c
C 03
... p
crf
CO P
3 CO
o
•H Crf
!M -H
Crf -P
> S3
CO Crf
P 0
03 crf CO
rH 03
£3 rn TJ
crf -H cd
EH >
03 £5
!Q -a
£ 03
< in
3
Ed
C 0 •
•H TH CO
| -^
C CO +|
0 -P
•H G CO
-P 03 JH
3 > -H
rH H Crf
rH X
O
ft
G
O
•H 03
•P O
3 rn
rH 3
rH Q
O CO
ft
03
O 3
CO
03 O
p a
erf X
G Pd
C
O
crf
O
O
03
•H
CO
en co
O 00
CO TH
TH CO
CM CM
O O
+ 1 +
CM O
CM CO
-H 0
6cJ d$
03
03 P 03
rH 0 rH
0 £ 0
H O r-(
P -P P
03 3 03
ft < ft
^0 CO
£•>. £*^
C5 01
iH tH
>>
rH P
3 O
"3 O
2
-
,fl
p
03
£!
erf
N
•H
rH
M
0 O tH
o m oo
TH tH tH
m CD t>
iH TH TH
0 0 O
+1 +1+1
00 tH CM
m CD t>
0 0 O
c2J
03 03
> £ > £ £
•H 3 -H 3 3
-P 03 -P 03 03
O rH O rH rH
£ O £ O O
0 H O SH f-(
P P -P -P -P
3 03 3 03 03
O t>
Ol O Ol
tH Oi tH
tH
G >. r-t
crf erf 3
i-J S >-3
J
03 X
ciD &H
3
O «
Cfi G
0
G P
O CO
P 3
erf 0
ea as
tH
t>
tH
CD
tH
O
+ 1
tH
t>
O
CO
rH
erf
O
•H
G
03
rf
CJ
t>
t>
01
iH
O
erf
3=
«
G
O
-P
CO
03
r-t
^
erf
^
O
m co
m CM
I
CN. ^jt
TH tH
0 0
+ 1+1
•^ o
CM t-
O O
1
£-i
rH 03
03 ft
03 a
-P 0
CO O
t>
t^ rjN.
O) t>
tH Oi
TH
I
r^
•H •
f-, P
ft O
<3 O
<
fc r^
£ crf
crf P
.fl 3
bfl
c -
•H Crf
£ G
!H bfl
•H crf
pa s
tH 00
CM CO
1 1
m i>
tH TH
0 O
+ 1 +1
oo m
O TH
O 0
i i
iH iH
£ > e
•H 03 -H 03
p .c P s:
0 O O O
£ O £ 0
O -P O -P
-P O -r3 O
3 £1 3 J3
f~ C^
C^ O5
O) tH
rH
-p
bfl ft
3 03
c^
O3 Oi
TH tH
. .
> o
O 03
"Z, Q
G
0
G
crf
CJ
•a
G
crf
^
a
-------
MEASUREMENT OF BIOLOGICAL ACTIVITY OF AIR MIXTURES
439
extrapolation from plants to man, at least this system can
become part of a battery of tests which can provide early
warning of the potential health hazard of exposure to mixed
air pollutants.
ACKNOWLEDGMENTS
This work was supported jointly by the U.S. Department
of Energy, National Institute of Environmental Health Sciences,
and U.S. Environmental Protection Agency. The authors acknow-
ledge with thanks the special efforts of: Mr. N.R. Tempel for
MMV assembly, deployment, and instrumentation and Mr. R.C.
Sautkulis for supervision of stock plants and field exposures;
Mr. W. Barnard, R. Baxter, and R. Ballard for aerometric in-
strumentation development and field operation; and Mr. J. Dame
and D. Brashear of Xonics, Inc. for operation of the CHAMP-van
and assistance in the field operation. The many hours of
flower analysis by Mr. E.E. Klug, Ms. A. Nauman, Ms. M.M. Naw-
rocky, Ms. V. Pond, Mr. R.C. Sautkulis, and Ms. R.C. Sparrow
also gratefully acknowledged.
REFERENCES
1. Mericle LW, Mericle RP: Genetic nature of somatic muta-
tions for flower color in Tradescantia, clone 02, Radia-
tion Botany 7:449-464, 1967
2. Nauman CH, Klotz PJ, Sparrow AH: Dosimetry of tritiated
1,2-dibromoethane in floral tissues of Tradescantia.
Mutat Res 38:406, 1976
3. Schairer LA, Van't Hof J, Hayes CG, Burton RM, de Serres
FJ: Exploratory monitoring of air pollutants for muta-
genicity activity with the Tradescantia stamen hair
system. Environmental Health Perspectives, in press
4. Sparrow AH, Baetcke KP, Shaver DL, Pond V: The rela-
tionship of mutation rate per roentgen to DNA content
per chromosome and to interphase chromosome volume.
Genetics 59:65-78, 1968
5. Sparrow AH, Schairer LA: Mutational response to Trades-
cantia after accidental exposure to a chemical mutagen.
EMS Newsletter 5:16-19, 1971
-------
440 L.A. SCHAIRER ET AL.
6. Sparrow AH, Schairer LA: Response of somatic mutation
frequency in Tradescantia to exposure time and concen-
tration of gaseous mutagens. Mutat Res 38:405-406, 1976
7. Sparrow AH, Schairer LA, Marimuthu KM: Radiobiologic
studies of Tradescantia plants orbited in Biosatellite
II. In: The experiments of Biosatellite II, (Saunders
JF, ed.)« NASA Special Publication 204, 99-122.
Scientific and Technical Information Office, NASA,
Washington, DC, 1971
8. Sparrow AH, Schairer LA, Nawrocky MM, Sautkulis RC:
Effects of low temperature and low level chronic gamma
radiation on somatic mutation rates in Tradescantia.
Radiation Res 47:273-274, 1971
9. Sparrow AH, Schairer LA, Villalobos-Petrini R: Compari-
son of somatic mutation rates induced in Tradescantia
by chemical and physical mutagens. Mutat Res 26:265-276,
1974
10. Sparrow AH, Underbrink AG, Rossi HH: Mutations induced
in Tradescantia by small doses of X-rays and neutrons:
analysis of dose-response curves. Science 176:916-918,
1972
11. Underbrink AG, Schairer LA, Sparrow AH: Tradescantia
stamen hairs: a radiobiological test system applicable
to chemical mutagenesis. In: Chemical Mutagens: Prin-
ciples and Methods for Their Detection, Vol. 3
(Hollaender A, ed.). New York, Plenum Press, 1973,
171-207
12. Underbrink AG, Schairer LA, Sparrow AH: The biophysical
properties of 3.9-GeV nitrogen ions. V. Determinations
of the relative biological effectiveness for somatic
mutations in Tradescantia. Radiation Res 55:437-446,
1973
13. Underbrink AG, Sparrow RC, Sparrow AH, Rossi HH: Rela-
tive biological effectiveness of X-rays and 0.43-MeV
monoenergetic neutrons on somatic mutation and loss of
reproductive integrity in Tradescantia stamen hairs.
Radiation Res 44:187-203, 1970
-------
PHYSICAL AND BIOLOGICAL
STUDIES OF COAL FLY ASH
Gerald L. Fisher and Clarence E. Chrisp
Radiobiology Laboratory
University of California
Davis, California
-------
443
In our initial studies of the potential health impact
of energy technologies, we have performed physical, chemical,
and mutagenic studies with coal fly ash. Although the vast
majority (95-99%) of the fly ash produced in coal combustion
for electric power generation is retained in the power plant,
we (5) have estimated that 2.4 million metric tons of fly ash
were emitted in the atmosphere from U. S. coal-fired electric
plants in 1974. Because the principal particulate emission
control technologies, electrostatic precipitators (ESP) or
wet scrubbers, have low collection efficiency for smaller
particles (34), much of the released fly ash is in the "re-
spirable" size range (aerodynamic diameters <10 m) (11).
This fine particle fraction presents the greatest potential
health hazard because fine particles have the longest atmo-
spheric residence times, and thus the greatest potential for
ultimate human inhalation (21), and are generally most effi-
ciently deposited in deep lung and least efficiently removed
by mucociliary transport (35) .
FLY ASH COLLECTION
To obtain sufficient quantity of size-classified fly ash
for detailed physical and biological testing, a specially de-
signed in-stack fractionator was constructed (20). The appa-
ratus was mounted in the stack breeching downstream from the
electrostatic precipitator (ESP) of a modern western U. S.
power plant burning high ash, low sulfur pulverized coal. At
the time of stack sampling, ESP hopper fly ash was also col-
lected. The apparatus consisted of a heated enclosure con-
taining two ryclone separators in series followed by a 25-jet
-------
444 GERALD L. FISHER AND CLARENCE E. CHRISP
centripeter (virtual dichotomous impactor). The stack gasses
were drawn through the inlet probe into the heated enclosure,
which was maintained at 95°C to prevent moisture condensation
associated with the high dew point of the stack effluent.
The serial arrangement of the two cyclones and the centripeter
provided in situ size-classification of four size fractions.
The two cyclone fractions had volume median diameters (VMDs)
of 20 (cut 1, coarsest) and 6.3 jm (cut 2) and the centripeter
fractions had VMDs of 3.2 (cut 3) and 2.2 pro (cut 4, finest)
(Table 1). All fractions had geometric standard deviations
(crg) of approximately 1.8. The fractionator was operated for
30 days at a flow rate of 30 cfm. Approximately 16 kg of mate-
rial was classified with approximately 67%, 16%, 7%, and 10%
of the mass in cuts 1, 2, 3, and 4, respectively. The size
distributions of the four sized fractions were compared (after
conversion to aerodynamic equivalent size) to samples col-
lected isokinetically from the stack (5). This approach al-
lowed for direct comparison of the size-fractionated material
to fly ash representative of normal stack emissions. The
comparison indicated the enhancement of fine particles and
the depletion of coarse particles in cuts 3 and 4 relative
to the isokinetically collected sample. Cut 1 was enhanced
in coarse particles, while cut 2 approximated the isokinetic
sample fairly well from 1.4 to 20 pm. Specifically, cuts 3
and 4 displayed six- to ten-fold and ten- to twenty-fold
increases, respectively, in the relative mass contributions
from 1 to 2 urn, while cut 1 contained less than one-tenth the
relative mass in this size interval when compared to the iso-
kinetic data. Therefore, with regard to subsequent chemical
and biological studies, it is important to note the size-
classification procedure resulted in extensive enhancement of
the fine particles (1-2 um) in cuts 3 and 4, relative to the
total particulate emission.
Physical and Morphological Studies
The average particulate density in the four size frac-
tions was found (5) to correlate negatively (p < 0.05) with
the VMDs (Table 1). A detailed morphological analysis of
particle types indicated that the variation in density could
be explained by the size dependence of the relative abundance
of the morphological classes.
We used light microscopy to define eleven major classes
of particulate morphology (5). On the basis of opacity and
particle shape, a fly ash morphogenesis scheme was developed
(Figure 1). The morphological classes included particles
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH
445
Table 1
Physical Properties of Size-Classified
Stack-Collected Coal Fly Ash
Percent Mean
Volume Geometric of Total Particle
Median Standard Mass Density
Fraction Cut # Diameter Deviation Collected (g/cm3)
First cyclone 1 20
Second cyclone 2 6.3
Centripeter- 3 3.2
large fraction
Centripeter- 4 2.2
small fraction
1.8
1.8
1.8
1.9
67
16
7
10
1.85
2.19
2.36
2.45
SHAPE
NON-OPAQUE
OPACITY
MIXED
EXPOSURE
AMORPHOUS
o
SILICATC .
IHIIICJitl.il
INCREASING
Figure 1. Morphogenesis scheme indicating probable relation-
ship between particle morphology and chemical composition.
Opacity and shape are used as primary characteristics for
morphological classification.
-------
446 GERALD L. FISHER AND CLARENCE E. CHRISP
that appeared amorphous and either opaque or non-opaque with
relatively limited exposure to combustion conditions within
the boiler. With further exposure to combustion conditions,
these particles developed somewhat rounded surfaces and con-
tained vesicles. Continued exposure to combustion conditions
resulted in formation of spherical particles derived from mol-
ten inorganic minerals or soot particles from incomplete coal
combustion.
We have defined five classes of spherical particles, the
most abundant morphological type. Solid, non-opaque spheres
and hollow, non-opaque spheres (cenospheres) are predominantly
aluminosilicates derived from clay minerals within the coal
(3). Spheres may range in color from water-white through yel-
low to dark red to opaque. Opaque spheres are mostly magne-
tite and are easily identified in microscopic studies by
taking advantage of their magnetic properties (5)« Some
spheres contain large numbers of smaller spheres (Figure 2).
These plerospheres are most abundant in the coarser fly ash
fractions. Careful examination of the plerospheres indicates
that often the encapsulated spheres within the plerosphere
are themselves plerospheres. We (3) have demonstrated that
the gases within the plerospheres are H20 and C02. On the
basis of the morphological appearance, bulk chemical compo-
sition and gaseous content we have postulated a mechanism to
account for the sphere-within-sphere structure.
As a noncombustible particle is progressively heated,
a molten layer develops on the outer surface. During that
time, mineral decomposition from CaC03 or clay minerals may
result in C02 or H20 evolution. This gas formation serves
as the driving force to separate the molten surface from the
solid particulate core. Further gas formation causes the
surface of the core to boil away resulting in microsphere
formation within the molten shell. The plerosphere is fro-
zen after the particle is carried out of the combustion zone.
We have calculated the time require for formation of a ple-
rosphere of 50 ym diameter to be on the order of 1000 usec.
We have also observed crystals on the surface of and
within fly ash spheres. Analysis of some of the large sur-
face crystals by electron microprobe indicated high concen-
trations of calcium and sulfur with no other elements de-
tected. On the basis of the SEM appearance of these crys-
tals, we (3) concluded that they were anhydrite (CaSO^) or
gypsum (CaSO,,-2H20) resulting from interaction of surface
formed or deposited H2SO,, with particulate calcium oxide.
Interiorized crystals generally appeared to radiate from one
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH
447
Figure 2. Micrographs of plerospheres indicating the sphere
•within sphere structure of these fly ash particles. The
plerosphere in the light photomicrograph (left) is 20 um in
diameter; the scanning electron micrograph (right) depicts
an 80 ym diameter plerosphere.
or two points on the sphere surface through the sphere.
These "quench" crystals have been reported to form from
heterogeneous nucleation at the surface of molten silicate
droplets during rapid quenching (14).
We have quantified the relative abundance of the mor-
phological types of particles in the four fly ash fractions
(Table 2). The relative abundance of most particle types
appears to be positively correlated with particle size. In
contrast to this observation, non-opaque spheres were corre-
lated negatively with particle size. The most striking dif-
ferences in frequency distributions were observed between
cut 1 and cut 4. Cut 1 was composed of 41% cenospheres and
-------
448 GERALD L. FISHER AND CLARENCE E. CHRISP
Table 2
Frequency (%) Distribution of Particle Classes in
Size-Classified Coal Fly Ash
Cut 1 Cut 2 Cut 3 Cut 4
Particle Class (20 um) (6.3 um) (3.2 um) (2.2 u
Combined amorphous,
opaque and non-opaque
Combined vesicular,
opaque and non-opaque
Sooty
Cenosphere
Plerosphere
Opaque sphere
Non-opaque sphere
Sphere with crystals
7.4
14.7
1.3
41.4
0.5
1.6
25.6
6.8
2.4
6.9
0.6
26.2
0.2
0.9
56.0
6.8
0.8
2.9
0.3
13.2
—
0.3
79.2
3.2
0.3
3.0
0.3
7.9
—
0.2
87.2
0.9
26% non-opaque spheres while cut 4 was composed of 8% ceno-
spheres and 87% non-opaque spheres. The greater amount of
solid spheres and lesser amount of vesicular particles ap-
pears to explain the observed trend of increased average
particle density with decreased particle size.
Elemental and Chemical Analysis
Because of the observed morphological heterogeneity,
we initiated elemental analysis of individual particles. In
our initial study (29), we used three-color X-ray mapping
techniques with a scanning electron microscope (SEM). We
analyzed fly ash provided by the NBS as a standard reference
material (NBS-SRM 1633) for 12 trace elements. Analysis of
fly ash particles with similar SEM morphologies indicated
extreme elemental heterogeneity, i.e., morphologically similar
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH 449
particles were found to contain high concentrations of Ti, S,
Al, K, Ca, or Fe. Further studies are now underway to evalu-
ate elemental composition of the eleven light-microscopically
defined morphological classes. Preliminary results indicate
that the pigmentation in non-opaque spheres from water-white
to yellow to red is associated with iron concentrations (4).
Analysis of opaque, amorphous particles indicates these par-
ticles are composed primarily of low atomic number elements,
reflecting the organic components of coal. Particles rich
in Ni, Cr, Zn, or Mn have been observed.
Detailed elemental analyses of the four fly ash frac-
tions were performed by instrumental neutron activation anal-
ysis (INAA) and atomic absorption spectrophotometry (AAS).
Prior to analysis of the fly ash fraction, the accuracy and
precision of the two techniques were evaluated using NBS fly
ash (SRM-1633) (26). The AAS analysis involved a room tem-
perature digestion in hydrofluoric acid followed by addition
of a saturated boric acid solution (32). This digestion tech-
nique resulted in quantitative dissolution of all elements
except selenium and barium. Comparison of INAA and AAS deter-
mination of Al, Ba, Co, Cr, Fe, K, Mn, Na, Ni, Ti, and Zn
indicated excellent agreement between the two techniques as
well as with previously published literature values (27).
Be, Cu, Cd, Mg, Ca, and Pb analyses by AAS and As, Ce, Cs,
Eu, Hf, La, Rb, Sb, Sc, Se, Sm, Sr, Ta, Tb, Th, U, V, W, and
Yb analyses by INAA also agreed well with previously published
literature values.
Summary tables of the analytical results are presented
for those elements displaying concentrations independent of
particle size (Table 3) and dependent on particle size (Table
4). For elements analyzed by both INAA and AAS, the data re-
ported are the results of the analytical technique with the
smaller coefficient of variation. Data from atomic absorption
analyses are the average of two independent determinations;
the INAA data are the weighted averages of three independent
determinations. Concentration dependence on particle size was
determined qualitatively with the criteria that consistent
concentration trends beyond experimental uncertainty were ob-
served for each fraction, although significantly higher con-
centrations of the element may have been observed in the
finest fraction relative to the coarsest fraction. The
enhancement factor is defined as the ratio of the element
concentration in cut 4 to its concentration in cut 1.
-------
450
GERALD L. FISHER AND CLARENCE E. CHRISP
Table 3
Elemental Concentrations Independent of Particle Size
Element
Technique
Cut
(VMD =
1
20 urn)
Cut
(VMD =
2
6.3
um)
Cut
(VMD =
3
3.2 um
Cut
) (VMD =
4
2.2
um)
Al
Fe
Ca
Na
K
Ti
Mg
AAS%
INAA"
AAS
AAS
AAS
AAS
AAS
Concentration in %
13.8(0.1)
2.5(0.1)
2.12(0.14)
1 .19(0.13)
0.74(0.01)
0.62(0.05)
0.47(0.01)
14.4(0.1)
2.9(0.2)
2.23(0.08)
1.75(0.05)
0.80(0.07)
0.76(0.05)
0.56(0.01)
14.2(0.8)
3.0(0.1)
2.30(0.14)
1.83(0.06)
0.82(0.08)
0.77(0.11)
0.60(0.02)
14.1(0.3)
3.2(0.1)
2.38(0.09)
1.85(0.03)
0.81(0.03)
0.78(0.06)
0.63(0.01)
Concentration in ug/g
Sr
Ce
La
Rb
Nd
Th
Ni
Sc
Hf
Co
Sm
Dy
Yb
Cs
Ta
Eu
Tb
INAA
INAA
INAA
INAA
INAA
INAA
AAS
INAA
INAA
INAA
INAA
INAA
INAA
INAA
INAA
INAA
INAA
410(60)
113(4)
62(3)
51(3)
45(4)
25.8(0.6)
25(3)
12.6(0.5)
9.7(0.4)
8 .9(0.2)
8.2(0.3)
6 .9(0.3)
3.4(0.4)
3.2(0.1)
2.1(0.1)
1 .0(0.1 )
0.90(0.05)
540(140)
122(5)
68(4)
56(4)
47(4)
28.3(0.6)
37(1)
15.3(0.6)
10.3(0.3)
16.3(0.8)
9.1(0.4)
8.5(0.9)
4.1(0.4)
3.7(0.2)
2.3(0.2)
1.2(0.2)
1.06(0.06)
590(140)
123(6)
67(11)
57(3)
49(7)
29(1)
43(4)
15.8(0.6)
10.5(0.3)
19(1)
9.2(0.4)
8.1(0.3)
4.0(0.2)
3.7(0.2)
2.5(0.3)
1 .2(0.2)
1 .10(0.07)
700(210)
120(5)
69(3)
57(8)
52(6)
30(2)
40(2)
16.0(0.2)
10.3(0.5)
21(1)
9.7(0.4)
8.5(0.8)
4.2(0.3)
3.7(0.2)
2.7(0.1)
1 .3(0.4)
1 .13(0.06)
Concentration dependence witn particle size was determined qualitatively
with the criteria that consistent concentration trends beyond experimen-
tal uncertainty were observed for each fraction.
?
"AAS values are the averages of two independent determinations; the
ranges are given in parentheses.
INAA values are the weighted averages of three independent determina-
tions; uncertainties (in parentheses) are the largest of twice the
weighted standard deviation, the range, or an estimate of the accuracy.
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH
451
^
0)
r— 1
rQ
Hi
IU
C-i
C '
CD
N
•rH
CO
0)
r-i
O
•H
i \
^™*
rH
_J
/"(
Pj
C
t-1
+J
C!
CD
T3
S
CD
ft
CD
Q
CO
c
O
•H
^j
cri
*»u
rH
4»3
jg
CD
O
C
o
o
r— t
aj
4_i
C
cu
Q)
rH
-*4
«v*
C
OJ
£ M
0) O
O •*-*
c y
a ai
— CL,
C
w
X— *
£
3.
CM
•
Tl< CM
•i-1 II
CJ Q
E
>
s— '
,— ^
£
CM
•
CO CO
+•> II
CJ Q
5
>
~
/— N
£
CO
•
CM CD
P II
CJ Q
S
£>
^^
x~v
E
11
0
n csi
+•> II
3
CJ Q
•A
>
> —
CD
3
CP
• H
C
y
cu
EH
p
S
CU
E
CD
rH
a
ino^coorHinrHcooocoini'
•* co m
rHrHOCnC-C-in^COCOCOCMOJCMrHrH
rH rH rH
/-N
CO
rH
O
CM t- O
O CO ^"^'^ O '"" s '""* CJi
COCMCgCMCOCMfflCnCOrJ'Tj.M'^
T3
CD
-p
ai
O
-H
•a
c
•H
CO
CO
CD
rH
C
.
bfi
^J'OOOCMO'3'OCOOOC^CnrHO
^CJ^COCMCMinc^C^CMCMI^
[s. rH tH rH CM CO
/^v
CO
rH
o
^* c*- o
• • *~s
O '^ O '"•> -~ •• O
CO-HCMCnOCMinCN'fl1— ITCOO
CMrHOOt^COCOOOCO^CMCOO
OC^COrHrH^^CM^C^CD
CO rH CM CM
gl
CM
C
•H
C
o
•- J
oi
P
a
CD
y
c
o
CJ
0
^^ °
CO "31 CO O
• • • ,^\ s^
o /~*' o rH ^ -^^ ^^ in
rHOicncocooooocQcnoocQ^o
coinin cMrHCDt^rHin
rH rH rH rH
,— V
rH
o
O
c7 cor7c7tfT aT o
• • • • • • • — /
O — 1 O O CM ^ ^rH CO
O'rHCMC'-OTJ'rHrHCO'tfOOCOrH
• x^ * *^s • • • • *^s * * v_x • ^* •
OOOC5COCMCOC5COCOCOCOCOO
CO rl »H ^ ^» QQ ^
<; <2 <2 «5 <; ^x
T3CCUCO.O Oai^2 Siai
CJNOO^CJCQ
<~N
in
0
-H CO CO
t--'o aT
CO rH O
rH CO
y— y
CO
•
o
•* in c~
C- C5 CO
o c-
rH CM
s—\
CM
•
o
-H in m
G> 00 rH
co co
CM
*~\
•
o
rH CO C-
•^x • ^-x
co co a
in o
CM
CO CO C/2
< <; <:
^ •*'
^O 0)
c^ o o
CM ^
CO
CM
^~* o
i-H •
• o
& s~**~S
00 O O
CM CM
^
^•s.
»
o
r;
aj y
•H ai
tH Q
(11
w
•H 0
SH <4H
y
T3
CD CD
si >
4-1 Si
CD
.c to
P o
•H O
fll
y^
cu &
•rl >,
4-> i— 1
aj C
P -iH
•H ai
ai Si
3 CD
a- y
c
73 3
CD
C rH
E •"->
Si a
CD CD
+-1 E
CD -rl
•a si
cu
'j> a,
ai x
» CD
cu -a
N C
•H O
to >>
[1)
UJ
CD .Q
rH
y to
•H T3
•p fl
!H CD
cd Si
Q.4->
— c
•P 0
•H -H
^s. 1
to ^ i erf
• •
O5 O O
CM CO
w
< o o
^J KH HH
X-N
&^
s^/
x-s
5*4 w
\^f f
•* , O
CD Si
y P
c c
CD CU
•a y
c c
CD O
a. y
CD
T3 P
c
a cu
O P
•H M
-P 'tH
si to
Si C
p o
c y
O -P
a ai
0 -3
o •*->
c
•rl
c
CD
>
•H
bfl
CD
^
ai
C/}
CD
=
aS
UJ
g;
-*-*
. A
U3
c
0
• rl
p
ai
•H
E
u
CD
p
• CD
rH TJ
p p
3 C
y CD
•a
c c
•rl CU
a
ai T3
^2 C
P -H
o o
p *
p
Tj<
p
a! CD
Si £
4-> P
c
CD CU
y Si
o
y co
cu
C
y -H
aj -P 00
rl Oi <
tH OS
•P CD
CD T3
•a
73
•P Si
C aj
CD T3
C ai
CD P
a, co
CD
•0 T3
C CD
•H P
CD "Sfl
CD -rl
Si CU
-C S
p
CD
O1J
+^
CO CD
CD y
ai ^
Si P
CD
> 0 rl >>
•H OS O
O rH Oi
S Si
CU 3
cu — y
-S 4-> O
P ai
CD
CD Si CD
ft at Si
ai P
CO to
> -P a!
S £
< CD -rl
< Si P
Z as CO
rH O, CU
f u
si
.p
•tH
?
CO
.p
y
ai
rH
-p
X
CD
rl
CD
4->
a!
^
m
CD
SH
fi
4H
•
4-> CD
tO T3
•O CD
a! O
C >,
Oi rH
CU c
E 0
0 £
r! rj
P —i
cu "s
Si
ai Si
CO <4H
CU
3 -
rl rH
a
> 4H
3
>> y
Q*
-------
452 GERALD L. FISHER AND CLARENCE E. CHRISP
The major element composition of the fractionated fly
ash is relatively independent of particle size with the excep-
tion of silicon, which appears to decrease with decreasing
particle size. Greater than 92% of the mass of the fraction-
ated fly ash can be accounted for by oxides of Si, Al, Fe, and
Ca. The more volatile elements (or their oxides), Cd, Zn, Se,
As, Sb, Mo, Ga, Pb, and V display clear-cut increases in con-
centration with decreasing particle size, in agreement with
the vapor-condensation mechanism of Natusch and Wallace (24).
It is important to note, however, that refractory elements
also display concentration trends inversely dependent on par-
ticle size. Therefore, processes other than vapor condensa-
tion are involved in the concentration-size relationship.
The elements U and Cr are associated with the organic frac-
tion of coal (22) and may be released in the combustion pro-
cess as fine particles that may agglomerate with other par-
ticles. The elements Fe, Mn, Ba, and Sr (22) may in part be
present as carbonate minerals which decompose to form fine
particles during coal combustion and again agglomerate with
other particles. Copper is probably present in part as the
sulfide arid Be as the alumiriosilicate in the coal (22). Thus,
mineral decomposition and elemental distribution may in part
explain the elemental trends of the high boiling chemical
species.
Analyses of H20 extracts of the fly ash fractions by
ion chromatography (10) indicated an inverse concentration
dependence on particle size for sulfate and fluoride (Table
4). Sulfite was not detected in the samples by either ion
chromatography or thermometric titration calorimetry.
Filtration studies with neutron activated fly ash indi-
cated that the elements Mo, Ca, Se, Ba, Co, As, and Sb dis-
play significant solubility at physiological pH (6). The
elements displaying the greatest solubilities relative to the
initial fly ash concentrations were Mo, Ca, and Se with rela-
tive solubilities of 55%, 30%, arid 20%, respectively.
Analysis of the organic compounds in the fly ash has been
initiated using gas chromatography with high resolution glass
capillary columns and mass spectrometry (17). Chromatograms
clearly demonstrate the presence of over 120 well-resolved
peaks. To date, the following polynuclear aromatic hydro-
carbons have been tentatively identified based on retention
data and Kovat's Indices of Standard Compounds, and/or mass
spectral data: dibenzofuran, pyrene, 1,2-benzoanthracene,
20-methylcholanthrene, benzo g,h,i(gi) perylene, naphthalene,
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH 453
1-methylnaphthalene, fluorene, phenanthrene, anthracene,
fluoranthene, and benzoanthracene 7,12-dione. Work is pre-
sently underway to substantiate these observations and to
identify further the organic compounds in fly ash.
MUTAGENICITY TESTING OF COAL FLY ASH
This report will review recently published data (2) and
illustrate our approach to biological testing of a complex
mixture. Certain metals which are carcinogenic in man or
animals (8,9,12,15,16,19,28,33) were shown to be concentrated
in stack fly ash as described earlier in this report. The
presence of carcinogenic substances as a result of fossil
fuel combustion has been suspected since scrotal cancers were
first observed in chimney sweeps in 1775 by Percival Pott (30).
Subsequently, organic compounds from coal tar products proved
to be carcinogenic (18).
Because a high positive correlation between carcinogeni-
city of substances for animals or man and mutagenicity in a
bacterial test system has been shown by Ames (25), we de-
cided to use this simple and economical test for the detection
of putative carcinogens on the surface of cut 4 fly ash.
Briefly, all five strains of histidine requiring auxotrophs
or Salmonella typhirourium. TA1535, TA100, TA1537, TA1538,
and TA98, kindly supplied to us by B.N. Ames, were used in
testing cut 4 of fly ash collected from the stack of a coal
burning power plant. The genetic background and testing
methods for these strains have been previously described (1).
Care was taken in the selection of the proper solvent
for the extraction of possible mutagens from the surface of
fly ash. Several laboratory solvents were tested for toxic-
ity and for mutagenicity. One must be careful to distinguish
between toxicity and mutagenicity in this test system. It
is necessary to incorporate a small amount of histidine into
the medium so the bacteria may undergo several replications.
Resultant tiny colonies are seen as a background lawn. How-
ever, if a solvent or mutagen is toxic, some of the bacteria
may be lysed, leaving others with a greater amount of histi-
dine per bacterium. This may be enough so that visible col-
onies are formed that may be mistaken for his"1" revertants.
If small colonies are seen, it is necessary to examine the
plate under a microscope to see if the background lawn is
sufficient. If not, either the solvent or test mutagen is
toxic. In addition, laboratory solvents can also be a
-------
454 GERALD L. FISHER AND CLARENCE E. CHRISP
source of mutagens, either because of solvent impurities in
manufacture, or contamination with mutagens in the laboratory
environment.
In initial studies cyclohexane, a nonmutagenic, nonpolar,
organic solvent was used. Cut 4 fly ash was extracted with
four 10 ml volumes of cyclohexane at room temperature and the
supernatant was passed through a 0.45 pm filter to remove fly
ash particles. Results of pour plate tests are shown in
Table 5. His revertants were seen with strains TA98 and
TA1538, but not with TA1537, TA1535, and TA100. These re-
sults indicated the probable presence of nonpolar, organic,
frameshift mutagens.
Two media were selected for further studies with cut 4
fly ash. Dulbecco's phosphate buffered saline was used be-
cause it has the pH and tonicity of physiological fluids.
Horse serum was selected because serum has a chemical con-
stituency similar to lung alveolar fluid and forms soluble
complexes with some carcinogenic heavy metals (13). Fly ash
samples were incubated with each of these media for a minimum
Table 5
Number of TA1538 His+ Revertants/Plate
S-9 Not Added S-9 Added
Test Mixture Fly Ash Control Fly Ash Control
Cyclohexane 62 + 2 5 + 2 152 + 8 27 +_ 5
extract
Serum filtrate 154 + 32 10 + 2 202 + 18 12 +_ 5
Saline filtrate 17 +_ 3 4 + 1 40 + 9 16 + 2
S-9 is the supernatant fraction of Aroclor-induced rat liver
homogenate, centrifuged at 9000 g. Positive controls were
spot tests with 4-nitro-quinoline-N-oxide without S-9 and
with 2-aminofluorene and S-9 added. The mean number of spon-
taneous revertants per plate was 7 +_ 1 without S-9 and 20 + 1
with S-9. The numbers given present the mean number of colo-
nies +_ the standard deviation on 3 replicate plates. Concen-
trations of fly ash are equivalent in all 3 test mixtures
(78 mg/ml). Filtrate (100 yl) was added to 2 ml of soft top
agar before plating.
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH 455
of one week at 37°C. After incubation, the fly ash mixtures
were centrifuged at 35,000 g and the supernatants were passed
through a 0.45 urn membrane filter to remove particulate mat-
ter. Media controls of serum or saline were treated in the
same fashion as the fly ash mixtures. No mutagenic activity
was found with spot tests, but his"1" revertants were found with
the pour plate technique (Table 5). This was evidence that
the mutagen or mutagens did not readily diffuse into the media
from the paper discs. Again, only the frame shift mutants TA-
98 and TA1538 showed his revertants. More revertants were
seen with strain TA1538 than TA98, so the former was used in
subsequent studies. A small increase in his revertants was
seen when optimal concentrations of rat liver homogenates from
rats treated with polychlorinated biphenyl (Arochlor 1254) was
added to pour plates (Table 5). Repetition of these tests has
shown that there is a small but highly significant (p < 0.001)
increase in his"1" revertants with metabolic activation. At
first the fly ash was autoclaved before incubation with the
various solvents in order to avoid bacterial contamination.
Later it was found that the fly ash was sterile and auto-
claving prior to incubation did not change the number of re-
vertants.
A dose response curve for mutagenicity of cut 4 fly ash
filtrates in strain TA1538 is shown in Figure 3. Serum fil-
trates had approximately a ten-fold greater activity than
saline filtrates. All mutagenic activity was found in the
aqueous fraction after extraction of saline filtrates with
cyclohexane. Solubility of substances responsible for muta-
genic activity in saline, a polar solvent, suggested the
presence of a polar organic or an inorganic mutagen. In ad-
dition these data imply that horse serum might be a useful
extract for complex mixtures of mutagens.
Reproducibility of the Ames test with fly ash serum fil-
trates was examined. The ratio of his"1" revertants to spon-
taneous revertants ranged from 20 to 60 when fly ash serum
filtrates were incubated at different times and the same fil-
trates stored and tested on different days. This variability
was greater than that observed when samples were incubated
at the same time and tested on the same day.
It is well known that serum protein can bind to both
organic (31) and inorganic (13) compounds. Fly ash serum fil-
trates were fractionated on a Sephadex G-25 column with a
cut-off of 25,000 daltons. Figure 4 shows the protein pat-
tern for three fractions collected from the column. Approxi-
mately 80% of the mutagenic activity was associated with the
-------
456
GERALD L. FISHER AND CLARENCE E. CHRISP
UJ
cu
or
UJ
o_
CO
oc
UJ
CE
CO
cr
UJ
CD
260
240
220
200
180
160
140
120
100
80
60
40
20
0
MUTAGENICITY OF SERUM AND SALINE FILTRATES
25 12 31 78
INCUBATION CONCENTRATION OF FLY ASH IN MEDIUM (mg/ml)
Figure 3. Mutagenicity of fly ash serum and saline. Fil-
trates with strain TA1538. The number of his revertants
per plate is the mean of 5 to 20 determinations minus the
mean of the appropriate background revertants (serum or
saline). The background reversion was defined as the group
mean of the spontaneous revertants and the appropriate media
control after it was determined that the number of his"1" re-
vertants in all negative controls was not significantly dif-
ferent from that of spontaneous revertants. The means (+_
SEM) of the background revertants 5.8( + 0.4), 6.9(+- 0.9),
4.0(;+ 0.6) for the spontaneous revertants, serum controls,
and saline controls respectively. Filtrate (100 wl) was
added to 2 ml of soft top agar before plating. Plates were
incubated for 2 days at 37°C. The vertical bars are 1 SEM.
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH
457
ELUTION OF HORSE SERUM FROM 625M COLUMN
40i i ., i i i ;i i i i i i ! i i i i i i i i
35
30
25
J20
10
0
0 6
I— FRACTION
12 I 18 24 I 30 36 42
—H— FRACTION 2—-I— FRACTION 3 —H
TUBE NUMBER
Figure 4. Pattern of elution of horse serum from a molecu-
lar weight exclusion column with a cutoff of 25,000 daltons.
The first fraction contains 95% of the serum protein, the
secod less than 5%, and the third only low molecular weight
compounds.
first fraction which contained 95% of the total serum pro-
tein. This indicated that most of the substances accounting
for mutagenic activity were probably bound to serum proteins,
Mutagenicity of EDTA-Treated Fly Ash Filtrates
The Ames test has not been very useful in testing known
carcinogenic heavy metals for mutagenicity; however, a few
have been shown to be mutagenic in this system (7). It was
-------
•158 GERALD L. FISHER AND CLARENCE E. CHRISP
decided that if heavy metals were responsible for any muta-
genic activity, a metal chelator such as ethylenediamine-
tetraacetic (EDTA) might remove this activity. EDTA-treated
and untreated serum filtrates were fractionated on a column
as illustrated in Figure 4. EDTA (2 mM) was added to one
portion of serum filtrate and stirred overnight at 4°C be-
fore elution on the column. A second portion was prepared
in the same manner without prior treatment with EDTA. Each
of these two filtrates was eluted with three void volumes of
double distilled water. As mentioned previously, the first
fraction contained most of the total serum protein (Figure
4). The second had the remaining protein and a small amount
of low molecular weight compounds, while the third fraction
contained only low molecular weight components. Each of the
three fractions was lyophilized and reconstituted with double
distilled water before testing. Regardless of prior treat-
ment with EDTA, the total mutagenic activity in the fractions
was lower than that in the original filtrate (Table 6). Of
the total net activity after subtraction of background re-
vertants (5.0 + 1.0), 79%, 18%, and 3% were present in the
first, second, and third untreated fractions, respectively.
Of the total net activity after subtraction of appropriate
control values 83%, 0%, and 17% were found in the three EDTA-
treated fractions, respectively. The significant increase
(p < 0.01) in the activity of the low-molecular-weight frac-
tion of the EDTA-treated serum filtrate lends credence to
the hypothesis that EDTA acted by chelating heavy metals from
serum proteins. Although it appears that metal chelation
is responsible, it is also possible that the EDTA may act to
increase bacterial cell permeability to mutagens. In addi-
tion, there may be synergism between metals and organic com-
pounds. The fact that the mutagenic activity of the frac-
tions is less than the total, regardless of EDTA treatment,
is partially explained by the necessity to subtract the con-
trol revertants from each fraction.
Studies are underway to evaluate the carcinogenic po-
tential of coal fly ash as well as the possible role of fly
ash inhalation in respiratory disorders.
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH
459
Table 6
Column Chromatograpy
Number of TA1538 His+ Revertants/Plate
Fly Ash
Unfractionated 162 + 18
serum filtrate
Serum filtrate 78 + 11
fraction 1
Serum filtrate 21+4
fraction 2
Serum filtrate 7+2
fraction 3
Fly Ash + EDTA Control
261 +25 8+2
94+10 7+1
11 + 4 11 + 2
22+3 4+1
Concentrations of fly ash were 78 mg/ml. The number given
represents the number of revertants +_ the standard deviation
on 5 replicate plates. Filtrate (100 ul) was added to 2 ml
of soft agar before plating.
REFERENCES
1. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella mammalian
microsome mutagenicity test. Mutat Res 31:347-363, 1975
2. Chrisp CE, Fisher GL, Lammert JE: Mutagenicity of fil-
trates from respirable coal fly ash. Science 199:73-75,
1978
3. Fisher GL, Chang DPY, Brummer M: Fly ash collected
from electrostatic precipitators: Microcrystalline
structures and the mystery of the spheres. Science
192:553-555, 1976
4. Fisher GL, Hayes T: unpublished data
5. Fisher GL, Prentice BA, Silberman D, Ondov JM, Biermann
AH, Ragaini RC, McFarland AR: Physical and morphologi-
cal studies of size-classified coal fly ash. Environ
Sci Tech, in press, 1978
-------
460 GERALD L. FISHER AND CLARENCE E. CHRISP
6. Fisher GL, Silberman D, Heft RE, Ondov JM: Fly ash fil-
terability, differential solubility and elemental dis-
tribution studies. In: Radiobiology Laboratory Annual
Report, University of California, Davis California 34-
40, 1977
7. Flessel CP: Metals as mutagens. Adv Exp Biol Med 91:
117-128, 1978
8. Furst A, Schlauder M, Sasmore DP: Tumorigenic activity
of lead chromate. Cancer Res 36:1779-1783, 1976
9. Furst A: An overview of metal carcinogenesis. Adv Exp
Biol Med 91:1-12, 1978
10. Hansen LD, Fisher GL: unpublished data
11. Hatch TF, Gross P: Pulmonary Deposition and Retention
of Inhaled Aerosols. New York, Academic Press, 1964
12. Heath JC: Carcinogenic action of metals. Brit Emp
Cancer Campaign Rep, Part 11:389, 1963
13. Heath JC, Webb M, Caffrey M: The interaction of carci-
nogenic metals with tissues and body fluids: Cobalt and
horse serum. Br J Cancer 23:153-166, 1969
14. Hurt J, Biechnicki DJ: Ultrafine-grain ceramics from
melt phase. In: Ultrafine-Grain Ceramics (Burke JJ,
Reed NL, Weiss V, eds.). Syracuse, Syracuse University
Press, 1970, pp 286-287
15. International Agency for Research on Cancer. Evaluation
of Carcinogenic Risk of Chemicals to Man, Vol I. Lyon,
184, 1972
16. International Agency for Research on Cancer. Some In-
organic and Organic Metallic Compounds, Vol II. Lyon,
181, 1973
17. Jennings WG, Sucre L, Fisher GL, Raabe OG: Analysis of
the organic constituents of coal, fly ash, coke and
coal tar. In: Radiobiology Laboratory Annual Report,
University of California, Davis, California, 1977, pp
24-33
-------
PHYSICAL AND BIOLOGICAL STUDIES OF COAL FLY ASH 461
18. Kubota H, Griest WH, Guerin MR: Determination of car-
cinogens in tobacco smoke and coal-derived samples -
trace polynuclear aromatic hydrocarbons. In: Trace
Substances in Environmental Health IX (Hemphill DD, ed.),
Columbia, University of Missouri, 1975, pp 281-289
19. Lau TJ, Rackett RL, Sunderman FW: The carcinogenicity
of intravenous nickel carbonyl in rats. Cancer Res 32:
2253-2258, 1972
20. McFarland AR, Bertch RW, Fisher GL, Prentice BA: A frac-
tionator for size-classification of aerosolized solid
particulate matter. Environ Sci Tech 11:781-784, 1977
21. Mercer TT: Aerosol Technology in Hazard Evaluation.
New York, Academic Press, 1973, pp 21-62
22. Murchison D, Westoll, TS: Coal and Coal-Bearing Strata.
New York, American Elsevier, 1968, p 418
23. Natusch DFS: Potentially carcinogenic species emitted
from fossil fuel power plants. Environ Health Perspec-
tives, in press, 1978
24. Natusch DFS, Wallace JR: Urban aerosol toxicity: The
influence of particle size. Science 186:695-699, 1974
25. McCann J, Choi E, Yamasaki E, Ames BN: Detection of
carcinogens as mutagens in the Salmonella microsome
test. Assay of 300 chemicals. Proc Nat Acad Sci 72:
5135-5139, 1975
26. Ondov JM, Ragaini RC, Heft RE, Fisher GL, Silberman D,
Prentice BA: Interlaboratory comparison of neutron
activation and atomic absorption analyses of size-
classified stack fly ash. Proc NBS 8th Materials
Research Symposium, Gaithersburg, MD, 1977, pp 565-572
27. Ondov JM, Zoller WH, Omez I, Aras NK, Gordon GE, Ranci-
telli LA, Abel KH, Filby RH, Shah KR, Ragaini RC: Ele-
mental concentrations in the National Bureau of Standards'
environmental coal and fly ash standard reference
materials. Anal Chem 47:1102, 1975
28. Ottolenghi AD, Haseman JK, Payne WW, Falk HL, MacFarland
HN: Inhalation studies' of nickel sulfide in pulmonary
carcinogenesis. J Nat Can Inst 54:1165-1172, 197
-------
462 GERALD L. FISHER AND CLARENCE E. CHRISP
29. Pawley JB, Fisher GL: Using simultaneous three color
X-ray mapping and digital-scan-stop for rapid elemental
characterization of coal combustion by-products, J
Micros 110:87-101, 1977
30. Pott P: The Chirurgical Works of Percival Pott, Vol II.
Philadelphia, James Webster, 1819, p 291
31. Rosenor VM, Oratz M, Rothschild MA: Albumin Structure,
Function and Uses. New York, Pergamon Press, 1977, pp
143-158
32. Silberman D, Fisher GL: Analysis of coal fly ash by
atomic absorption spectroscopy. Pacific Conference on
Chemistry and Spectroscopy, October 1977, Anaheim,
California
33. Stone GD, Shimkin MB, Troxell MC, Thompson TL, Terry LS:
Test for carcinogenicity of metallic compounds by the
pulmonary tumor response in strain A mice. Cancer Res
36:1744-1747, 1976
34. Vandergrift AE, Shannon LF, Gorman PG: Controlling fine
particles. Chem Eng 80:107-114, 1973
35. Yen, HC, Phalen RF, Raabe OG: Factors influencing the
deposition of inhaled particles. Environ Health Per-
spect 15:147-156, 1976
-------
MUTAGENICITY OF SHALE OIL
COMPONENTS
R.A. Pelroy and M.R. Petersen
Biology Department
Battelle-Northwest
Richland, Washington
-------
465
Raw shale oil is a complex chemical mixture differing
from most crude petroleums in having comparatively high
concentrations of basic (nitrogen-containing) and phenolic
compounds, in addition to having neutral compounds and
polynuclear aromatic hydrocarbon (PNA) constituents more
commonly found in crude oils (7). In the work described
below, we have investigated the mutagenicity of a pilot
plant sample of a crude shale oil (designated L01) and
two subfractions derived from this material.
Although the Ames assay has been widely used for muta-
genic screening of pure chemicals (1,3,4), its use for bio-
assay of complex chemical mixtures has been more limited.
Cigarette smoke condensate (2,5), complex mixtures of
polycyclic compounds associated with airborne pollutants
(9,10), and to a lesser extent, some synthetic fuels,
have been assayed in this way (8).
In the work to be reported at this symposium, we have
directed our attention to two problems that can arise
during the Ames testing of complex chemical mixtures.
First, we have estimated the ability of known chemical
mutagens (premutagens requiring metabolic activation) to
express themselves in the chemical environment to be re-
presented by a raw shale oil or its subfractions. Second,
we have estimated the degree of cell killing that occurs or
is the result of exposing the Salmonella typhimurium test
strains to these complex fractions under the conditions
employed for the Ames assay, and the possible importance
of such killing on the sensitivity of this assay.
-------
466 R.A. PELROY AND M.R. PETERSEN
MUTAGENIC PROPERTIES OF A SHALE OIL SAMPLE
The raw shale oil, L01, was fractionated into five sub-
fractions: acidic (phenolic), basic, neutral, PNA, and a
complex residual mixture defined as a tar fraction. The
raw shale oil was mutagenic in the standard Ames assay.
The basic and PNA subfractions contained most of the muta-
genic activity recoverable after separation of the shale
oil into its various chemical classes. In all cases,
mutagenicity was dependent on metabolic activation cata-
lysed by postmitochrondrial, microsomal enzymes. As
shown in Figure 1, the mutational response of S_. typhimurium
TA100 was comparatively low for the crude product, the
basic, and PNA fractions. Comparable results were obtained
for the other test strain that we used for most of this
work, £>. typhimurium TA98. In general, the mutational
responses for the basic and PNA range from 0.1 to 1 rever-
tant colony per ug per 109 test cells added to the assay
system. In some instances, the response curves for the
two subfractions were linear for a greater concentration
range than shown in Figure 1. However, nonlinear muta-
genic responses shown here are typical for both the crude
product and its subfractions.
Mutagenicity of Pure Chemical Plus Complex Fraction
Mixtures
A potential problem in interpreting the results of
the standard Ames test of complex chemical mixtures is the
possibility that the mutagenicity. of the whole will be
significantly different than the sum of the individual
components.
One method of estimating the importance of chemical
composition on the Ames assay is to add a known mutagen
or premutagen to a complex fraction, and then compare
the mutagenicity of the mixed-system (chemical + fraction)
with the mutagenicity of the chemical alone. This experi-
mental approach was followed with the raw shale oil, the
basic and PNA subfractions as complex materials, and
2-aminoanthracene, benzo(a)pyrene (benzopyrene) and 7,9-
dimethylbenz(c)acridine (dimethylbenzacridine) as known
premutagens.
In these experiments, the concentration of the pure
chemical was held constant at a value sufficient to yield
a strong mutational response when assayed alone, i.e.,
-------
MUTAGENICITY OF SHALE OIL COMPONENTS
467
i/i
"el
500
400
300 ~
200
100
200 400 600-800 1000 2000
|jg crude fraction
Figure 1. Mutagenicity of shale oil (L01) and the basic
and PNA fractions derived from L01. Salmonella typhimurium
TA100 was the test strain and each sample plate contained
50 yl of the S-9 enzymes.
1 ug 2-aminoanthracene, or 20 yg for benzopyrene and di-
benzanthracene per assay plate (Figure 2). The concentra-
tion of the S-9 enzymes for the mixing experiments was
determined on the basis of that required for activating
the raw shale or its subfractions to form mutagens against
TA100. The data comparing S-9 requirements for the three
pure chemicals, the shale oil, and the four subfractions
is shown in Figure 3. For all of the mixing experiments
reported here, a constant value of 50 yl of S-9 per plate was
used. It should be noted (Figure 3) that the optimum S-9 con-
centrations for metabolic activation of dimethylbenzacridine
and benzopyrene, and for the crude fractions were approxi-
mately the same, while the optimum concentrations of S-9
-------
468
R.A. PELROY AND M.R. PETERSEN
12,000
10,000
5 8,000
CD
OS
6,000
4,000
2,000 -
dimethyl benzacridine
(50 int S9)
benzopyrene
(50 uj S9)
1 2 5 10
20 30
ug chemical
40
50
Figure 2. Mutagenicity of three chemicals as a function of
concentration of pure or complex chemicals.
for activation of 2-aminoanthracene (alone) was considerably
less.
Three patterns of response were observed in the mixing
experiments, depending on the chemical in question. For
2-aminoanthranene, addition of the raw shale oil or either
the basic or the PNA fractions derived from the crude
product, led results in a sharp increase in the number of
revertants formed from TA100 (Figure 4) over that expected
for the sum of the fraction plus chemical.
-------
MUTAGENICITY OF SHALE OIL COMPONENTS
469
12000
2-ami noanthracene
100
d i methy 1benzacr i dIne
60
100
sq
Figure 3. Mutagenicity of three chemicals and two complex
fractions as function of S-9 concentration.
The increase in the mutagenicity of the mixture was
greater than four times the maximum response for any one
of the crude fractions assayed separately, and was equal to
about 17% of the maximum mutagenic response observed for
2-aminoanthracene assayed alone at its optimum S-9 concen-
tration (Figure 2).
In contrast to the results for 2-aminoanthracene the
mutagenicity of benzopyrene steadily diminished with increas-
ing concentrations of the three crude mixtures (Figure 5).
In each case, the mixture yielded approximately the same num-
ber of revertant colonies per plate as the crude fraction
alone and the mutagenicity of benzopyrene was marked.
-------
470
R.A. PELROY AND M.R. PETERSEN
2 aminoanthracene (1 jig] vs:
S 2000 -
1000
600
1000 2000 200 600 1000 200
ug crude fraction
600
1000
Figure 4. Mutagenicity of 2-aminoanthracene (1 pg) and in-
creasing concentrations of shale oil (L01), basic, or PNA
fraction. The concentration of the S-9 enzymes was constant
at 50 ul per assay plate.
benzopyrene (9 [ig] vs:
§ 2000
1000
200
600
1000 2000
200 600
yg crude fraction
1000 200
600
1000
Figure 5. Mutagenicity of benzophyrene (9 ug) and increasing
concentrations of shale oil (L01), basic, or PNA fraction.
Conditions same as Figure 4.
-------
MUTAGENICITY OF SHALE OIL COMPONENTS 471
The mixing experiments for dimethylbenzacridine showed
a third pattern. Here the mutagenicity of the mixture was
only slightly less than the sum of responses for the chemical
alone and crude fractions assayed at various concentra-
tions (Figure 6). Addition of shale oil had the least
effect on the combined system, while the basic and PNA
fractions showed little inhibitory effect up to approxi-
mately 200 ug per assay plate.
Toxicity to Test Cells
In the standard Ames assay, the level of cell killing
due to formation of toxic metabolites or due to chemical
composition is not directly measurable. Since complex
hydrocarbon mixtures are generally toxic to bacteria, the
Ames assay of shale oil should take this into account.
In the work described here we have used an indirect method
to estimate the toxicity that occurs during mutagenesis
caused by the pure chemicals and complex fractions studied
above in the mixing experiments.
A revertant of TA100 was isolated from an assay plate.
This organism, designated TA100 rev, was added to the
standard Ames assay system at a range of dilutions from
10 ~H to 10 "7 from nutrient broth cultures containing
approximately 2 x 109 viable cells per ml. Because TA100
was wild type with respect to the biosynthesis of histidine,
it was able to grow on the assay plates used in the Ames
assay (i.e., on a glucose mineral base containing biotin
for which TA100 rev was still auxotrophic).
Addition of TA100 rev to the standard Ames assay
system showed that survival of this strain differed greatly
depending on the complex material or pure chemical being
assayed. For the three pure chemicals studied above, only
2-aminoanthracene showed a strong killing effect on TA100
rev (Figure 7). The concentration dependence for 2-amino-
anthrancene induced toxicity closely followed the concentra-
tion dependence observed for mutagenesis (Figure 2), so, at
least in qualitative terms, loss of viability for TA100
rev was correlated with decreased mutational response by
the histidine auxotroph, TA100. On the other hand, neither
benzopyrene nor dimethylbenzacridine gave rise to killing
of TA100 rev over the concentration range used in the
standard Ames mutagenesis assays.
-------
472
R.A. PELROY AND M.R. PETERSEN
dimethybenzacridine (5 fig] vs:
8 1000
<
evertan
V/l
8
a) shale oil
b) basic
v_
c) PNA
200
600
1000 2000
200 600 1000
pg crude fraction
200
600
1000
Figure 6. Mutagenicity of dimethylbenzacridine and increasing
concentrations of shale oil (L01), basic, or PNA fraction.
Conditions same as Figure 4.
Contrasting results were also obtained for the crude
fractions. Shale oil (L01) and the PNA fraction showed
little or no killing of TA100 rev in the Ames assay system
at concentrations (per plate) approaching 1,000 yg for PNA
fraction and nearly 2,000 yg for this raw shale oil (Figure
8) .
The basic fraction, however, was highly cytotoxic for
TA100 rev (Figure 8). The highest level of toxicity was
observed for the complete assay system which contains all
the necessary components for metabolic activation. Omission
of reduced pyridine dinucleotide phosphate (NADPH2), required
for metabolic activation, reduced the killing of TA100 rev.
For example, at approximately 400 yg basic fraction, killing
of TA100 rev was nearly five times greater for assay plates
containing NADPH2 relative to those assay plates without this
cofactor. At approximately 600 yg per plate, this relative
increase was 18-fold and at slightly less than 800 yg per
plate NADPH2 dependent killing was more than 25 times greater
than toxicity observed for the assay system minus the cofac-
tor.
-------
MUTAGENICITY OF SHALE OIL COMPONENTS
473
1.0
0.5
0.2
0.1
benzopyrene (50 \il S9)
-8
dimethylbenzacridine (50 \ii S9)
aminoanthracene (5 M£ S9)
I i
i
i
1 2 5 10
20
30
Mg chemical
40
50
Figure 7. Survival of Salmonella typhimurium TA100 rev vs.
premutagen. Concentration per plate of untreated (control)
cells, Sc; exposed cells, S. The titer of TA100 rev on
control plates was 1.7 x 109 cells per ml of nutrient broth
culture. The concentration of S-9 per assay plate is in-
dicated in the figure.
In previous work we showed that formation of metabolite
mutagens from 2-aminoanthracene and benzopyrene in the
presence or absence of crude fractions is limited to the
initial stages of the Ames assay, i.e, within the first 90-
120 min (6). Thus, the extensive killing demonstrated here
for TA100 rev exposed to basic fraction might seriously re-
duce the mutagenic response for the system for this material,
-------
474
R.A. PELROY AND M.R. PETERSEN
1.0
0.1 —
0.001
400
600 800
ng crude fraction
1000
2000
Figure 8. Survival of Salmonella typhimurium TA100 rev vs.
complex fractions. The assay system for the basic fraction
was complete for one set of plates and lacked a NADPH2
generating system in a second set. The other samples Con-
tained the NADPH2 generating system. S-9 concentration was
fired at 50 ul per plate.
In summary, of the three chemical premutagens tested,
2-aminoanthracene and dimethylbenzacridine expressed more
of their mutagenicity in the presence of shale oil than did
benzopyrene. The mutagenicity of the latter compound was
strongly suppressed by each of the complex fractions tested.
The basic fraction in addition to being mutagenic was highly
toxic to a revertant strain of _S. typhimurium TA100 over the
same concentration of crude fraction required for muta-
genesis of the auxotrophic parental strain. Toxicity by
the basic fraction was enhanced in the presence of a complete
system for metabolic activation.
-------
MUTAGENICITY OF SHALE OIL COMPONENTS 475
REFERENCES
1. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian-
microsome mutagenicity test. Mutat Res 31:347, 1975
2. Kier LD, Yamasaki E, Ames B: Detection of mutagenic
activity in cigarette smoke condenstates. Proc Natl
Acad Sci 71:4159, 1974
3. McCann T, et al.: Detection of carcinogens as mutagens:
Bacterial tester strains with R factor plasmids. Proc
Natl Acad Sci 72:979, 1975
4. McCann J, et al.: Detection of carcinogens as mutagens
in the Salmonella/microsome test: Assay of 300
chemicals. Proc Natl Acad Sci 72:5135, 1975
5. Mizusaki S, Takashima T, Tomura K: Factors affecting
mutagenic activity of cigarette smoke condensate in
Salmonella typhimurium TA1538. Mutat Res 48:29, 1977
6. Pelroy RA and Petersen MR: Use of Ames test in evalu-
ation of shale oil fractions. Environ Health
Perspectives, in press
7. Petersen MR, Fruchter J, Laul JC: Characterization
of substances in products, effluents and wastes from
synthetic fuel production tests. Quarterly report for
the US Energy Research and Development Administration.
Battelle, Pacific Northwest Laboratories, Richland,
WA 99352. BNWL-2131, 1976
8. Rubin I, et al.: Fractionation of synthetic crude oils
from coal for biological testing. Environ Res 12:358,
1976
9. Talcott R, Wei E: Airborne mutagens bioassayed in
Salmonella typhimurium. J Natl Cancer Inst 58:449,
1977
10. Tokiwa H, et al.: Detection of mutagenic activity in
particulate air pollutants. Mutat Res 48:237, 1977
-------
MUTAGENIC ANALYSIS OF
DRINKING WATER
Colin D. Chriswell, Bonita A. Glatz,
James S. Fritz, and Harry J. Svec
Iowa State University
Ames, Iowa
-------
479
As recently as ten years ago relatively little was known
about organic contaminants in drinking water. The carbon ab-
sorption methods (3,12) and other techniques were used to
provide an indication of the amount of organic matter in
water. However, only a handful of the individual compounds
had ever been identified. During the past ten years it has
become possible to separate and identify many organic sub-
stances in drinking water using techniques such as gas chro-
matography-mass spectrometry (GC-MS). Nearly 500 compounds
have now been positively identified (9,10) and the list of
identifications is continuing to grow.
Despite the progress that has been made, much remains
to be learned about organic contaminants in water. In par-
ticular, we must elucidate the potential health effects of
these organic compounds.
Some compounds have been identified in drinking water
that may pose a threat to human health. Chloroform is
present in water from every utility using chlorine as a
disinfectant (2,13) (Figure 1), and chloroform and other
trihalomethanes are suspected carcinogens (7,11). Other
suspected carcinogens have also been identified in drinking
water, but these compounds are generally less widespread and
are rarely found at as high concentrations as the trihalo-
methanes (4,14). Continued identification and characteri-
zation efforts will undoubtedly reveal the presence of
additional potentially harmful organic contaminants. It has
become possible to use bioassay procedures such as the
Salmonella/mutagenicity assay to guide the identification
efforts towards compounds of the greatest potential interest.
-------
480
COLIN D. CHRISWELL ET AL.
52O
5OO
480
460
440 U|
il
60
40
CITY I
cc
UJ
CO
<
cc
o
o
cr
JULY
876
200
\
MOV
1976
MARCH
1977
JULY SEPT
1977 1977
JULY
876
CITY 5
NOV
1976
MARCH
S77
JULY SEPT
1977 1977
JULY
1976
Figure 1. Levels of trihalomethanes found in drinking water
from fourteen cities. Upper line is total concentration of
-------
MUTAGENIC ANALYSIS OF DRINKING WATER
481
JULY
1976
JULY
876
PlTT-l-
JULY SEPT
1977 1977
CITY 9
NOV
S76
MARCH
1977
JULY SEPT
1977 1977
JULY
1976
NOV
1976
MARCH
1977
"<—T
JULY
1976
JULY
S76
CITY 13
MARCH jULY
1977 1977
SEPT
1977
JULY SEF
1977 1977
trihalomethanes expressed as chloroform equivalents. Lower
line is chloroform.
-------
482 COLIN D. CHRISWELL ET AL.
The research group at Iowa State University that I am re-
presenting has been involved in the development of analytical
methods for isolating, concentrating, and identifying organic
compounds. In the past, our group has consisted of analyti-
cal and physical chemists specializing in the areas of sepa-
rations and mass spectrometry. During the past year bacte-
riologists, immunologists, sanitary engineers, and water
utility operators have joined our project. A multidisci-
plinary effort is being undertaken to determine more about
organic contaminants in drinking water and their potential
health effects. An immediate goal is to answer three ques-
tions: (1) How prevalent are mutagenic materials in drink-
ing water? (2) What levels of mutagenic activity are
present in drinking water? (3) What are the chemical
characteristics of the mutagenic materials?
HOW PREVALENT ARE MUTAGENIC MATERIALS IN DRINKING WATER?
Since July of 1976 our group has been conducting a sur-
vey of organic contaminants in drinking water for the Ameri-
can Water Works Association. As part of that survey, organic
compounds are isolated from raw and finished water from each
of fourteen cities at monthly intervals. Aliquots of the
isolated organic materials have been assayed for mutagenic
activity.
Accumulation of Organic Compounds
Organic compounds are isolated by sorption on column
assemblies containing Amberlite XAD-2 resin in series with
Filtrasorb 200 activated carbon (Figure 2). With each sam-
pling 200 1 of water is passed through the sampling columns.
Both the primary and secondary columns are 6" x 1/2" i.d.
Accumulated organic substances are desorbed by elution with
100 ml of diethyl ether. The compounds are then further
concentrated by distilling the ether eluates to a final
volume of 1.00 ml. Of this 1.00 ml concentrate, 0.25 ml is
used for gas chromatographic and GC-MS determinations and
the remainder for mutagenic assays. Extracts obtained dur-
ing the winter months of 1976 were composited, 300 yl of
dimethylsulfoxide (DMSO) added to each composite and the
residual ether evaporated. These DMSO concentrates contained
organic materials originally present in 15 1 of water in each
10 ul of DMSO.
-------
MUTAGENIC ANALYSIS OF DRINKING WATER
483
-c
K
Figure 2.
water.
A sampler used to accumulate organic materials from
-------
484 COLIN D. CHRISWELL ET AL.
Mutagenicity Assays
Mutagenicity assays were performed using the spot test
procedure described by Ames, McCann, and Yamasaki (1).
Whatman No. 1 filter paper discs were soaked with 10 \il of
DMSO concentrate and placed in petri dishes on the surface
of agar seeded with approximately 108 cells of special
mutant strains of Salmonella tryphimurium. Strains TA98,
TA100, TA1535, TA1537, and TA1538 were used. Each sample
was tested at least twice with each strain with and without
the addition of the microsomal fraction of Aroclor 1254-
activated rat liver.
The Salmonella strains lack the ability to grow without
added histidine but may regain the ability to grow in the ab-
sence of histidine by various mutagenic agents. Strains TA-
100 and TA1535 are reverted by substances causing base-pair
substitutions. Strains TA98, TA100, TA1537, and TA1538 are
reverted by frameshift mutagens of varying specificities.
Positive tests were defined in this work as a concentration
of revertant colonies in a circular array around the site of
sample application (Figure 3). The number of colonies in a
positive test is at least twice the number appearing in re-
sponse to solvent controls. Marginal results were recorded
if only a small increase in colony count or a slight concen-
tration of colonies around the sa.mple were observed. The
liver fraction, designated S9, is added to provide many of
the key enzymes of in vivo mammalian metabolism. Thus, muta-
genic metabolites of compounds not mutagenic in themselves
may be detected. Positive (known mutagens) and negative
(solvent) controls were included for each strain in each
experiment. No positive results were reported if replicate
determinations did not agree. A positive test with any one
strain of Salmonella indicates the presence of mutagenic
substances in the water sampled.
Results
The results of these assays are presented in Figure 4.
Eleven of the fourteen finished and six of the raw water
sources exhibited some degree of mutagenic activity. The
greatest number of positive tests were obtained against
strain TA100. In contrast, the related strain, TA1535, was
not reverted by a single sample. This may in part be due to
the greater sensitivity bestowed on TA100 by the plasmid R
factor pKMlOl. In addition, TA100 is reverted by either
mutagens causing frame shift mutations or base-pair substi-
-------
MUTAGENIC ANALYSIS OF DRINKING WATER
485
TY
> c
o
<
O
UJ
UJ
LU
UJ
>
en
O
Q.
a
0)
bfl
M
O
O
ol
X
0)
0
*
OJ
=H
O
oD
•H
P
w
p
0)
p
o
0)
c
•H
0)
p
oS
r-t
a
rj >»
-P
-------
486
COLIN D. CHRISWELL ET AL.
Figure 4. Results of the assay of composited samples from
fourteen cities. Samples taken during the winter months of
1976.
tutions while TA1535 is reverted only by base-pair substitu-
tion mutagens. Addition of the liver extract did not have a
pronounced effect on the activity of water samples against
strain TA100. In fact, in two instances, samples from cities
8 to 12, activity was reduced in the presence of S9. Strain
TA98 responded to many of the same samples as did strain TA-
100. Increased reversions were observed against strain TA98
and the related strain TA1538 in the presence of the liver
extracts. Activity was noted against strain TA1537 in a few
scattered instances.
Activity was found in finished water .camples from cities
3, 8, 9, 10, and 14 without any corresponding activity exhib-
ited by raw water samples. In these instances the water
treatment process may be responsible for introducing muta-
genic factors. Raw water samples from city 4 were active
against strain TA1537 and from city 11 against strain TA98.
No activity was observed against these strains in the fin-
ished water samples. In these instances water treatment may
have either altered the nature of the mutagenic materials or
removed them.
-------
MUTAGENIC ANALYSIS OF DRINKING WATER 487
In this initial screening the prime goal was to deter-
mine if mutagenic materials were widespread in drinking water.
We found such agents are very prevalent in finished water.
We realize that the Ames test is not a perfect assay nor is
the accumulation technique used perfect. Thus, these results
may be only a conservative indication of the true prevalence
of mutagenic materials in water.
WHAT LEVELS OF MUTAGENIC ACTIVITY ARE PRESENT IN DRINKING
WATER?
The presence of any mutagenic materials in water is a
cause for some concern. However, in order to evaluate the
threat posed, the levels of mutagenic materials must be deter-
mined. In part, this is a matter of performing quantitative
mutagenicity assays rather than the spot test procedures. It
is also necessary to have confidence that all mutagenic mate-
rials are accumulated from water.
In the initial studies Amberlite XAD-2 resin was used as
a primary accumulating agent. This sorbent is effective for
recovering gas chromatographic organic compounds from water
(5,6,8). However, it does not lead to the recovery of all
organic materials from water. It is not known if mutagenic
materials are of a nature such that they are recovered using
XAD-2 resin.
To determine if other sorption techniques might remove
more effectively mutagenic materials from water, sixteen col-
umns, each containing a different test sorbent, were connected
in parallel and used to sample finished water from four dif-
ferent Iowa utilities (Table 1). Some mutagenic materials
were isolated from water using Amberlite XAD-2, XAD-4, XAD-7,
XAD-8, Duolite S-761, and L-863 resins. Mutagenic materials
were not isolated using activated carbons, weak base ion ex-
change resins, or a carbonaceous resin. The greatest amount
of mutagenic activity was found in organic materials isolated
using Amberlite XAD-4 resin. It is, of course, still not
known that this sorbent is removing all mutagenic materials
from water, but it is the most effective sorbent tested to
date.
Studies are continuing in evaluating the effectiveness
of other sorbents for removing organic mutagens from water,
comparing mutagen recoveries at different water pH levels,
and using reverse osmosis as the accumulating technique.
-------
488
COLIN D. CHRISWELL ET AL.
Table 1
Evaluation of Sorbents for Accumulating Mutagens
Sorbent
Sorbent Type
Activity of
Isolated
Organic
Compounds
Amberlite XAD-2
Amberlite XAD-4
Amberlite XAD-7
Amberlite XAD-8
Duolite S-761
Duolite L-863
Duolite S-37
Duolite A-7
Duolite ES-561
Hydrodarco
Filtrasorb 300
Nuchar WVB
Nuchar WVG
NACAR G-216
NACAR G-107
Amberlite XE-340
PS-DVB Resin
PS-DVB Resin
Acrylic Ester Resin
Acrylic Ester Resin
Phenol-Formaldehyde •
Adsorbent
PS-DVB Resin
Weak-Base Anion Exchange
Resin
Weak-Base Anion Exchange
Resin
Weak-Base Anion Exchange
Resin
Granular Activated Carbon
Granular Activated Carbon
Granular Activated Carbon
Granular Activated Carbon
Granular Activated Carbon
Granular Activated Carbon
Carbonaceous Resin
+
+
-------
MUTAGENIC ANALYSIS OF DRINKING WATER 489
WHAT ARE THE CHEMICAL CHARACTERISTICS OF THE MUTAGENIC
MATERIALS?
A typical water source will contain on the order of one
part per million of total organic carbon. Most of this or-
ganic material, such as humic material, is believed to be of
natural origin. Other materials are introduced by man's
activities. Still other materials are produced during water
treatment. Identified organic compounds constitute only a
small fraction of the total amount of organic material in a
typical water supply.
It is of extreme importance that mutagenic materials
from drinking water be identified or at least characterized.
A very good correlation exists between mutagenicity and mam-
malian carcinogenicity, but the correlation is not perfect.
Thus, mutagenic materials from water should be characterized
so they can be tested for carcinogenicity. In addition,
effective measures for the control of mutagenic materials
can only be taken when the characteristics of the mutagens
are known.
The protocol adopted for the identification or character-
ization of mutagenic materials from water is based on succes-
sive, bioassay-guided fractionations until mutagenic activity
is isolated into a limited number of fractions with relatively
few components. Fractionation procedures must preserve the
integrity of the samples, be applicable to low amounts of
materials, and separate samples into fractions containing com-
ponents of predictable characteristics.
Thus far we have evaluated fractionation procedures based
upon solvent extractions, thin-layer and column chromatography
on silica gel and alumina, column chromatography on Florisil,
high pressure liquid chromatography on Sephadex LH20, and
preparatory scale gas chromatography. Solvent extraction
procedures are not conveniently applicable to the ultra-trace
amounts of materials that can be isolated from drinking water.
We have found that organic compounds are lost or altered dur-
ing fractionation procedures using silica or alumina. Poly-
aromatic hydrocarbons are irreversibly sorbed; alkenes and
some carbonyl-containing compounds undergo condensation or
polymerization reactions.
Fractions can be readily performed by preparatory scale
gas chromatography. This technique does, however, have the
serious limitation as it is applicable only to gas chromato-
graphic compounds.
-------
490 COLIN D. CHRISWELL ET AL.
The initial step before performing a fractionation on
Florisil is to transfer the sample from a diethyl ether to
petroleum ether. Some materials precipitate during this sol-
vent change. These materials are mutagenic, are not gas
chromatographic, contain only very low levels of carbon, and
give no characteristic IR or NMR spectra. In short, we have
no idea what the material is but do know it is mutagenic.
After materials are eluted from Florisil, the Florisil is
dissolved in hydrofluoric acid to recover any very polar
materials. The very polar materials have given no indication
of mutagenic activity.
Fractionations performed using activated Florisil are
based primarily on sample component polarities. A sample is
introduced onto the top of a Florisil column and components
are sequentially eluted with solvents of increasing polari-
ties. In our current work we have found it desirable to
elute organic materials isolated from water with 2% methylene
chloride in petroleum ether, 60% methylene chloride in petro-
leum ether and 60% methylene chloride plus 2% acetonitrile
in petroleum ether. Alkanes, alkenes, arenes, and halogenated
hydrocarbons are eluted in the initial fraction. Despite the
fact that this fraction would contain any polyaromatic hydro-
carbons in samples, no mutagenic activity has been detected
in it. The second elution fraction contains aldehydes, ke-
tones, nitro-substituted compounds, nitriles, and some weaker
phenols and amines. Mutagenic activity has been detected in
this fraction. The third fraction contains alcohols, phtha-
lates, amines, and phenols. Mutagenic activity has also been
detected in this fraction (Table 2).
Fractionations of organic compounds on Sephadex LH20 are
based both on sample component polarity and molecular size.
A sample is introduced onto an LH20 column and eluted with
2-propanol. Alkanes are eluted first in order of decreasing
molecular weights. Alkanes are followed by polar organic
compounds. The last materials eluted are aromatic compounds.
The aromatic compounds are eluted in order of increasing num-
ber of fused rings (Figure 5).
Identification or characterization efforts are based on
the volatility of the isolated components. Gas chromato-
graphic components are identified based on responses towards
element specific detectors and GC-MS. We have not yet begun
to characterize the nonvolatile components.
-------
MUTAGENIC ANALYSIS OF DRINKING WATER
491
Table 2
Florisil Fractionation
Fraction
Types of Compounds
Mutagenic Activity
2% CH2C12 in
petroleum ether
Aromatic and aliphatic
hydrocarbons, halogenated
aromatic and aliphatic
hydrocarbons
60% CH2C12 in
petroleum ether
Aldehydes, ketones, nitro-
substituted compounds,
nitriles, weak phenols,
and amines
60% CH2C12 +
2% CH3CN in
petroleum ether
Alcohols, phthalates,
amines, phenols
Hydrofluoric
acid
Very polar
CONCLUSIONS
Mutagenic materials are widespread in drinking water and
may be introduced into the water during treatment processes.
This finding is cause for concern and we are continuing to
monitor both finished and raw water in an attempt to deter-
mine what processes enhance the mutagenicity of water. We
are also attempting to determine the characteristics of or-
ganic compounds in water that are responsible for the muta-
genic activity. To date we have not identified a single
compound that is responsible for a significant portion of
the observed activity. Obviously a great deal remains to be
done.
-------
492
COLIN D. CHRISWELL ET AL.
(ill!)
AROMATIC COMPOUNDS
PROFILE OF ORGAN I
MATERIALS FROM STACK
SEPHADEX LH 20- COLUMN
2-PROPANOL-ELUENT
100
76 68 60 52 44 36
ELUTION TIME, MINUTES
8
Figure o. Separation of organic materials on Sephadex.
-------
MUTAGENIC ANALYSIS OF DRINKING WATER 493
Earlier I made a transition from talking about health
effects of organic materials in water to talking about muta-
gens and potential carcinogens. A convenient procedure exists
in the Ames test for determining if mutagenic materials are
present in water. No convenient assay does exist for deter-
mining if organic materials from water possess other deleteri-
ous or beneficial properties.
REFERENCES
1. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian
microsome mutagenicity test. Mutat Res 31:.347, 1978
2. Arguello MD, Chriswell CD, Fritz JS, Kissinger LD,
Lee KW, Richard JJ, Svec HJ: Trihalomethanes in water:
A report on the occurrence, seasonal variations in con-
centrations, and precursors of trihalomethanes. Jour
AWWA, submitted
3. Buelow RW, Carswell JK, Symons JM: An improved method
for determining organics in water by activated carbon
adsorption and solvent extraction. Jour AWWA 65:57 and
65:195, 1973
4. Chriswell CD, Arguello MD, Avery MJ, Ericson RL, Fritz
JS, Junk GA, Kissinger LD, Lee KW, Richard JJ, Svec HJ,
Vick R: Proceeding of the American Water Works Associa-
tion Convention, May 1977
5. Chriswell CD, Ericson RL, Junk GA, Lee KW, Fritz JS,
Svec HJ: Comparison of macroreticular resin and acti-
vated carbon as sorbents. Jour AWWA 69:56-69, 1977
6. Chriswell CD, Fritz JS, Svec HJ: Evaluation of sorbents
as organic compound accumulators. AWWA Water Quality
Technology Conference Proceedings, Dec. 1977
«
7. EPA Statement: Chlorinated and brominated compounds are
not equal. Jour AWWA 69:5-12, 1977
8. Junk GA, Richard JJ, Grieser MD, Witiak D, Witiak JL,
Arguello MD, Vick R, Svec HJ, Fritz JS, Calder GV: Use
of macroreticular resins in the analysis of water for
trace organic contaminants. Jour Chromatogr 99:745,
1974
-------
494 COLIN D. CHRISWELL ET AL.
9. Junk GA, Stanley SE: Organics in drinking water. Part
1. Listing of identified compounds, Springfield, VA
National Technical Information Service, 1975
10. McCabe LJ: Health effects of organics in water study.
AWWA Water Quality Technology Conference, Dec. 1977
11. Report on the carcinogenesis bioassay of chloroform,
Carcinogen Bioassay and Program Resources Branch,
Carcinogenesis Program, Division of Cancer Cause and
Prevention, National Cancer Institute
12. Standard methods for the examination of water and waste
water, 13th ed., New York, NY, 1971
13. Symons JM, Bellar TA, Carswell JK, DeMarco J, Kropp KL,
Roebeck GG, Seeger DR, Slocum CJ, Smith BL, Stevens AA:
National organics reconnaissance for halogenated
organics. Jour AWWA, 69:62, 1977
14. Von Rossum P, Webb RG: XAD resins and carbon for isola-
tion of organic water pollutants, Anal Chem, 1978, in
press
-------
IN VITRO ACTIVATION OF
CIGARETTE SMOKE
COMPENSATE MATERIALS TO
THEIR MUTAGENIC FORMS
R.E. Kouri, K.R. Brandt,
R.G. Sosnowski, L.M. Schechtman
Microbiological Associates
Department of Biochemical Oncology
Bethesda, Maryland
W.F. Benedict
Children's Hospital of Los Angeles
Los Angeles, California
-------
497
INTRODUCTION
Cigarette smoke is a complex mixture composed of 5,000-
10,000 different chemicals in the particulate phase, of which
about 3,000 have been identified (1), and 1,000-2,000 chemi-
cals in the gas phase. The particulate fraction contains
many chemicals that are capable of inducing cancer in model
test systems. Among these chemicals are certain polycyclic
aromatic hydrocarbons (PAH) [e.g., benzo(a)pyrene (BP),
dibenz(a,h)anthracene, and benz(a)anthracene]; certain nitro-
samines (e.g., diethylnitrosamine and nitrosopiperidine);
and certain aromatic amines [e.g., 2-napthylamine (2-NA) and
2-aminofluorenes (2-AF)] (see review 2). These chemicals
are normally at levels approaching 0.5-20 ng/cigarette. The
particulate phase also contains chemicals that are capable
of promoting carcinogenesis (2-4). The level of these chem-
icals (e.g., catachol) are on the order of 10,000-100,000
ng/cigarette. Thus, there is a problem in determining not
only whether cigarette smoke plays an active role in smoke-
associated cancers in man, but also if this association
occurs at the level of initiation and/or promotion of cancer.
One way to assess the potential initiating role that
cigarette smoke may have is to test for the biological activ-
ity of certain cigarette smoke-derived fractions. This
'Supported in part through contracts from The Council for
Tobacco Research USA, Inc., New York, NY 10002.
-------
498 R.E. KOURIETAL.
paper shows that measurement of the mutagenic potential of
cigarette smoke condensate (CSC) materials has some very
interesting ramifications. The fractions thought to contain
many biologically active chemicals (i.e., the PAH) have rela-
tively weak, if any, mutagenic activity. The primary tissue
thought to be at risk to smoke effects is that of the lung.
However, of the lung systems employed, the mouse lung does
not activate these smoke condensates; yet another tissue that
is not believed to be at risk to smoke, the liver, can acti-
vate the smoke condensates. In yjltro activation of such a
complex mixture as smoke condensate is obviously quite diffi-
cult to interpret, yet the- studies presented here do suggest
some approaches that may be able to show direction for the
eventual understanding of the biological effects of tobacco-
related chemicals.
MATERIALS AND METHODS
Bacterial Strains
The Salmonella typhimurium strains used, TA1538 and TA98,
were obtained from Dr. B. Ames (Biochemistry Department,
University of California, Berkeley, CA) and have been described
previously (5).
Compounds
NADH, NADPH, 6-arainochrysene (6-AC) , arid BP were obtained
from Sigma Chemical Company, St. Louis, MO; 2-AF was from
Aldrich Chemical Company, Milwaukee, WI; Aflatoxin El (AfBt)
was from Calbiochem, La Jolla, CA; and 7,8-benzoflavone (7,8-
BF) was from Eastman Organic Chemical Company, Rochester,
NY. 7,8-dihydro, dihydroxy-BP (7,8-diol~BP) was provided by
Dr. D. Jerina (NIH-NIAID). Aroclor 1254 was from Analabs,
North Haven, CT; 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
was provided by Dr. A. Poland (McArdle Laboratories, Univer-
sity of Wisconsin). The CSC fractions were generated by
Meloy Laboratories according to the methods of Patel et al.
(6). Dimethyl sulfoxide (DMSO) was obtained from Schwarz/
Mann, Inc., Rockville, MD.
Rat Hepatic and Mouse Pulmonary S-9
Male Sprague-Dawley rats, weighing approximately 200 g
each were treated intraperitoneally (IP) with 0.5 ml of 200
mg Aroclor 1254/ml corn oil, in order to induce hepatic
-------
CIGARETTE SMOKE CONDENSATE MATERIALS 499
enzymes. Forty-eight hours after injection, rats were sacri-
ficed and livers excised. C57BL/6Cum mice, 6-8 weeks old,
were treated intratracheally (IT) with 0.02 ml of 6.0 ug
TCDD/ml trioctanoin in order to induce pulmonary and hepatic
enzymes. After 48 hr, mice were sacrificed and their lungs
and liver were excised. The 9000 x g post-mitochondrial super-
natant (S-9) fractions from the liver and lung tissues were
prepared as previously described (7,8).
Preparation of S-9 Mix
The S-9 mix for the suspension assay contained 1.2 mM
NADPH, 1.41 mM NADH, 136.9 mM NaCl, 2.68 mM KC1, 8.1 mM
Na.HPO,, 1.47 mM KH.PO,,, and 3.0 mM MgCl2, pH 7.4; total S-9
varied from 0.002 to 0.2 ml per ml of S-9 mix. The S-9 for
the pour-plate assay contained 3.6 mM NADPH, 4.2 mM NADH,
136.9 mM NaCl, 2.68 mM KC1, 8.1 mM NajHPO,, 1.47 mM KH^O^,
and 3.0 mM MgCl, pH 7.4, and varies from 0.2 to 0.3 ml of
S-9 fraction per ml. The liver preparations were sterile,
but the lung S-9 mix contained bacterial contaminants which
were removed by passing the S-9 mix through a sterile Milli-
pore disposable filter unit (0.45 y pore diameter). Total
protein was determined for each condition with fluorescamine
according to the method of Weigele et al. (9).
Aryl Hydrocarbon Hydroxylase (AHH) Assay
The assay for AHH activity was done according to proce-
dures outlined by Nebert and Gielen (10) and modified by
Kouri et al. (11).
Mutagenesis Assays
All pour plate incorporation mutagenesis assays were
performed according to the method of Kier et al. (12). For
suspension assays, 0.1 ml of the bacterial tester strain,
0.5 ml of S-9 mix, and the sample to be tested were incubated
in a 37°C water bath for 35 min. After incubation, samples
were taken from each condition, diluted, and spread on nutri-
ent agar plates to determine the number of bacteria at risk.
Two ml samples of molten top agar containing L-histidine
(0.05 mM) and biotin (0.05 mM) were added to each incubated
sample, mixed and poured onto Spizzizen minimal agar plates.
After 48 hr incubation at 37°C, prototrophic revertant
-------
500 R.E. KOURI ET AL.
colonies were counted on an NBS Model Gill Colony Counter
(New Brunswick Scientific, Edison, NJ). DMSO and acetone,
in the amounts used, have no toxic or mutagenic effects on
the tester strains. Preliminary studies were done using the
suspension protocol in order to assess the relationship
between the number of mutant colonies observed relative to
the number of bacteria added per plate when trace amounts
(0.05 mM) of L-histidine and biotin were present. Initial
bacterial concentrations ranging from 5 x 10s to 1 x 10 */pIa.te
resulted in only an 0.3-fold increase in the number of mutants
per plate. Thus, when L-histidine and biotin were present,
the number of revertant colonies did not really reflect the
initial number of bacteria that were added because the trace
levels of L-histidine allowed for a certain amount of growth
to occur. Therefore, a mutation frequency was calculated only
when the numbers of surviving bacteria for the various test
groups remained relatively constant. In all other cases, the
mutation data were given just in terms of number of revertant
colonies per plate.
RESULTS
A summary of the biological effects of the smoke conden-
sate from 1A1 low nicotine, normal tar content cigarettes is
shown in Table 1. The assays used included measurements of:
(a) pulmonary AHH following IT administration of fraction;
(b) competitive inhibition of BP metabolism in vitro; (c)
mutagenesis at the his locus in j3. typhimurium strains TA1538;
and (d) neoplastic transformation of C3H 10T% cells in cul-
ture. The whole condensate and reconstituted fractions were
weak inducers of pulmonary AHH, weak competitive inhibitors
of BP metabolism, mutagenic to TA.1538, and transforming to
the 10T^ cells. Mutagenesis required the presence of an ex-
ogenous metabolic activation system in the form of Aroclor
1254-induced rat hepatic S-9 preparation. Fraction Bjb (for
discussion of fraction nomenclature, see 13) contained chemi-
cals that were potent inducers and inhibitors of AHH, could
be metabolized to forms highly mutagenic to TA1538, and could
transform the 10T1^ cells. Fractions BTa, B^, and WAT also
1 Cj i
were active in most of these systems. Fractions N,, ~H and
N^.. were inducers of pulmonary AHH and could competitively
inhibit BP metabolism in vitro, but had low mutagenic poten-
tial and did not transform the 10T% cells. The N^ fraction
accounts for most of the BP content of the smoke condensate
(see footnotes in Table 1). The strong acid fractions (SA-,,
-------
CIGARETTE SMOKE CONDENSATE MATERIALS
501
Table 1
Effects of Fractions of 1A1 CSC in Various Model Systems
Fraction1
Whole CSC
Reconsti-
tuted CSC
B,a
BIb
BE
Bw
WAj
WAE
SAj
SAE
SAW
NMeOH
NCH
NNM
mg/Cig.
23.50
23.00
0.81
0.29
0.95
0.36
2.27
1.98
0.39
0.78
8.69
1.19
4.58
0.70
AHH
Ind.2
1.7
1.8
3.6
2.5
1.5
0.5
1.6
1.1
0.5
0.3
0.4
2.5
1.2
3.2
[X]/[BP]to
Give 50% Mutants/ Transfor-
Inhibition3 Plate* mation5
5.0 +++ +
5.2 +++ +
0.8 ++
0.5 +++ +
3.0 ++
>10.0
5.0 ++ +
2.0 +
>10.0 +
>10.0
>10.0
3.0 +
ND -
1.0 +
'Whole cigarette smoke condensate (CSC) has 21.0 mg nicotine,
5.70 mg phenols, 0.98 ug BP/g. Reconstituted CSC has 22.0 mg
nicotine, 5.51 mg phenols, 0.90 wg BP/g. B has 31.0 mg nic-
otine/g. WA£ has 41 mg phenols/g. NNM has 13.1 ug BP/g.
2Aryl hydrocarbon hydroxylase (AHH) inducibility = Effect of
fractions of 1A1 CSC on pulmonary AHH activity of C57BL/6Cum
mice relative to a corn oil control (11).
3BP inhibition = Competitive ±ia vitro effect of CSC fractions
on BP metabolism by hepatic microsomes from 3-MC-treated
C57BL/6Cum mice (14).
*Mutagenesis = Mutagenic activity of 1A1 CSC fractions in the
Ames assay with S. typhimurium TA1538 in the presence of
liver microsomal S-9 mix (12) .
transformation = Malignant transformation frequency in C3H
10T% Cl. eight cells treated with CSC fractions (15).
-------
502 R.E. KOURIETAL.
SAE, and SAW) actually inhibited pulmonary AHH activity and
only the SA-, fraction had an effect in any of the in vitro
bioassays.
A repeat (using a blind protocol) of the mutation stud-
ies, this time using the 2A1 CSC and fractions derived from
this condensate, is shown in Table 2. Experiments 1 and 2
are results from studies completed one year apart. The con-
densate and fractions were stored at ~70°C during the interim.
The total tar content was higher in this cigarette condensate
relative to the 1A1 condensate. However, on a per cigarette
basis, these data are very similar to those of Kier et al.
(12) using the 1A1 cigarette condensate. The nmtagenic activ-
ity of the whole condensate, reconstituted fractions and the
12 fractions were very similar to that of 1A1 condensate.
The most active fractions were Bjb, Bg, Bja, WAr and WAg.
The only discrepancies relative to the 1A1 condensate were
the higher activity of the nicotine-containing B<- fraction
and the slightly lower activity of the WAj fraction. The
mutagenic activity was stable at -7Q°C for at least one year
since both experiments yielded quite similar results.
The use of TCDD-induced mouse pulmonary tissue as an in
vitro activation system for the 2A1 condensate is shown in
Table 3. Under conditions in which the pulmonary S-9's could
efficiently metabolize BP to 3-hydroxybenzo(a)pyrene (3-OH-BP)
(the basis for the AHH assay), no metabolism of the 2A1 con-
densate to a mutagenic form could be observed. Use of pul-
monary S-9's from 3-methylcholarithrene (3-MC) treated or con-
trol mice also did not activate the 2A1 condensate material
(data not shown). Comparison of the in vitro metabolic capa-
city of TCDD-induced pulmonary S-9 with mouse or rat hepatic
S-9's in either a pour plate or suspension assay is shown in
Table 4. Under conditions in which both the rat and mouse
hepatic S-9's activated 6-AC or the 2A1 condensate to muta-
genic forms, the mouse pulmonary S-9 failed to activate either
of these chemicals. Also, addition of similar levels of total
AHH activity for both pulmonary and hepatic S-9's (by adjusting
total protein concentration) yielded conditions in which only
the hepatic S-9 activated 2A1 condensate to forms mutagenic to
strain TA98 (data not shown).
There was the possibility that even though these pulmo-
nary S-9's were capable of metabolizing BP to 3-OH-BP, some
sort of inhibitor of bacterial mutagenesis was functioning
-------
CIGARETTE SMOKE CONDENSATE MATERIALS
503
->
CO
c
o
•H
-p
o
oj
tn
b
0
CO
O
T-l
<:
CM
CM
.C
0) -P
rH -H
.0 £
«J
&H 00
o
<
H
«H
O
co
•H
03
0)
c
/n
u/
bfl
oj
-P
3
s
Q)
-P
erf
rH
ft
~^,
to
-p
a
aJ
-P
3
S
>,
-p
•H
>
•H
•P
O
CO CMCDC-COiHCnOOCMClCQC'-C}
CM iHCOCOCMCMrHlOCO Tt•>*> «• 0
rHICCMCOlO ^iH CM
-/
CD !H
•H O •
rK arH
O JH «J
-P O
d O -P
fn C Q)
O -H
X! h
«j CD >rH
o a«H
rH 0
CD rH
S 3 W
O T3
>» a o
.a £!
0) -P
13 J3 »
O CO nj
•H J5 tfl
C -P CO
Oj
> o
O -P CQ
rH -H
s-/ bD CO
C 0)
rH -H fl
<• T3 CD
-------
504
R.E. KOURI ET AL.
Table 3
Activation of 2A1 CSC
by Pulmonary S-9 Using a Suspension Assay1
Pulmonary
S-92
(mg protein)
0.72
0.72
0.72
0.72
1.44
1.44
1.44
2A1
Condensate
(ug/Tube) AHH3
1300 547.4
650
260
0
650 1071.0
260
0
BAR
(x 10')
0.78
0.76
0.79
0.86
0.81
0.88
0.72
Mutants/
Plate
15
13
12
13
21
19
13
MF*
(x 10"')
19.23
17.11
15.19
15.12
25.92
21.59
18.06
TA98 (alone)
0.78
14
17.95
lFor suspension assays, 0.1 ml of the bacterial tester
strain, 0.5 ml of S-9 mix, and the sample to be tested
were incubated in a 37°C water bath for 35 min. After in-
cubation, samples were taken from each condition, diluted,
and spread on nutrient agar plates to determine the number
of bacteria at risk (BAR). Two ml samples of molten top
agar containing L-histidine (0.05 mM) and biotin (0.05 mM)
were added to each incubated sample, mixed, and poured on
Spizzizen's minimal agar plates. After 48 hr incubation
at 37°C, prototrophic revertant colonies were counted on
an NBS Model Clll Colony Counter.
2 S-9 was derived from pulmonary tissue of C57BL/6Cum ? mice
induced by IT installation of 120 ng TCDD/0.02 ml trioc-
tanoin 48 hr prior to sacrifice.
3AHH = pMoles 3-OH-BP formed per 35 min incubation in
separate tubes which contained 25 ug BP/ml as substrate
and which were assayed under the same conditions and at
the same time as those tubes containing 2A1 condensate.
%MF = Mutation frequency, i.e., the number of his+ revertant
bacterial colonies per BAR.
-------
CIGARETTE SMOKE CONDENSATE MATERIALS
505
Table 4
Comparison of C57BL/6Cum TCDD-Induced Hepatic
and Pulmonary S-9 Mediated Metabolism of 2A1 Whole CSC
to Form(s) Mutagenic to ^. typhimurium TA98
Source of S-9
C57BL/6Cum Hepatic:
Pour plate
Suspension
C57BL/6Cum Pulmonary:
Pour plate
Suspension
Rat Hepatic:
Pour plate
Suspension
TA98 alone:
Pour plate
Suspension
Compound
(vg/Plate)
2A1
2A1
6 -AC
2A1
2A1
6 -AC
2A1
AfB1
2A1
6 -AC
(650)
(1)
(260)
(5)
(650)
(1)
(650)
(5)
v
(1300)
(1)
(650)
(0.5)
mg
Protein
3.14
3.14
3.14
3.14
1.44
1.44
1.44
1.44
4.43
4.43
2.95
0.03
AHH1
4929
4929
4929
4929
1071
1071
1071
1071
9625
9625
7371
190
BAR1 Mutants/
(x 107) Plate
132.0
189.0
0.84 75.7
0.74 131.7
12.0
20.7
0.81 21.0
0.78 39.7
202.3
775.0
0.82 97.7
0.63 329.0
14.7
0.77 14.7
MF'
(x 10~7)
90
178
26
51
120
522
19
'AHH = pMoles 3-OH-BP formed per assay tube; total time was 35 rain; 25 ug BP/ral
was substrate.
2BAR = Number of bacteria at risk.
'MF = Mutation frequency, i.e. the number of his revertant mutant bacterial
colonies per number of BAR.
-------
506 R.E. KOURI ET AL.
in these S-9's. Tables 5 and 6 show that these pulmonary
S-9's are capable of activating the 7f8-diol-BP and 2-AF to
mutagenc forms, respectively. Thus, these S~9"s were capable
of activating at least some PAH and some aromatic amines to
forms mutagenic to tester strain TA98. The activation of
2-AF was dependent on the integrity of the mixed-function
oxidase system because inhibition of AHH by tne inhibitor
7,8-BF resulted in concomitant inhibition of 2-AF-induced
mutagenesis (Table 7).
DISCUSSION
Cigarette smoke contains chemicals that have been shown
to be biologically active in a variety of model systems both
in vitro (see Tables 1 and 2; 12,14,15) and In vivo (3,9,16,
17). Of prime importance is the fact that either whole smoke
(14,18) or smoke condensate material is capable of interacting
with those microsomal monooxygenases known to play a major
role in the activation of many chemical carcinogens to their
cytotoxic (19-21), mutagenic (5,22,23), or carcinogenic (24-27)
forms. In this paper, we show that both the 1A1 and 2A1 ref-
erence cigarettes contain chemicals that are substrates for
hepatic monooxygenases and as a result of metabolism by these
hepatic tissue preparations, intermediates are generated which
are mutagenic to S_. typhimurium tester strains TA1538 and TA98
(see Tables 1 and 2). Two interesting facts emerge from these
studies: (a) the tester strain TA98 is selectively more sensi-
tive to mutagenesis induced by smoke condensate; and (b) the
fractions that contain most of the mutagenic activity are not
those known to contain the PAH, but rather should contain such
base-soluble chemicals as aromatic amines. Thus, the data
suggest that the majority (approximately 58%) of the total
mutagenic activity of these condensates is in che basic frac-
tions, and not in those fractions containing the PAH.
Another main issue of concern is the fact that mouse
pulmonary tissue fails to activate the 2A1 smoke condensates
to mutagenic forms (see Tables 3 and 4). That is, under
conditions in which these pulmonary S-9's metabolize BP to
3-OH-BP and metabolize both 7,8-diol-BP (Table 5) and 2-AF
(Table 6) to mutagenic forms, these S-9's fail to activate
either 6-AC or the 2A1 condensate (see Table 4). Thus, mouse
pulmonary tissues seem to be capable of activating certain
PAH and aromatic amines, but not others. If the 2A1 smoke
condensate does contain aromatic amines and these chemicals
are responsible for the high mutagenic activity of these con-
densates wh^n metabolically activated by hepatic S-9's, then
-------
CIGARETTE SMOKE CONDENSATE MATERIALS
507
Table 5
Activation of 7,8-diol-BP by Pulmonary S-9
Pulmonary S-9
(mg Protein)
6.57
3.28
0
0
7,8-diol-BP
(ug/Tube)
0.5
0.5
1.0
0.1
AHH1
1928
1680
-
-
BAR2
(x 107)
0.44
0.14
0.68
0.76
Mutants/
Plate
544
338
76
84
TA98 alone
0.77
20
= pMoles 3-OH-BP formed per 35 min incubation in
separate tubes containing 25.0 ng BP/ml as substrate.
BAR = Number of bacteria at risk. Because of large varia-
tion in BAR, no mutation frequency is given.
Table 6
Activation of 2-AF by Pulmonary S-9
Pulmonary S-9
(mg Protein)
6.57
3.28
0
TA98 alone
2-AF
(ug/Tube)
25
10
10
25
10
5
25
10
-
AHH1
1904
1925
1904
1452
1452
1680
-
-
BAR2
(x 107)
0.37
0.62
0.47
0.40
0.40
0.78
0.63
0.65
0.77
Mutants/
Plate
536
668
521
458
580
627
39
52
20
1AHH = pMoles 3-OH-BP formed per 35 min incubation in
separate tubes containing 25.0 ug BP/ml as substrate.
2BAR = Number of bacteria at risk. Because of large varia-
tion in BAR, no mutation frequency is given.
-------
508
R.E. KOURIETAL.
Table 7
Effect of 7,8-BF on Pulmonary and Hepatic S-9
Mediated Activation of 2-AF to Forms
Mutagenic to S. typhimuriom TA98
Rat
(0.
Mou
(2.
S-91
Hepatic
15 mg protein)
se Pulmonary
9 mg protein)
2-AF
(wg)
10
10
10
_
10
10
10
7,8-BF
Ug)
0
10
25
_
0
10
25
AHH
856
630
202
1701
1701
261
66
2
.8
.7
.3
.3
.7
.8
.5
B
(x
0
0
0
0
0
0
0
AR3
107)
.31
.64
.52
.68
.46
.71
.54
Mutants/
Plate
564
210
187
28
538
125
68
TA98 (alone)
1.07
16
1S-9's were derived from: (1) hepatic tissue from Fischer
334 d" rates (200-250 g) induced by IP administration of
500 mg Aroclor-1254/kg body weight 48 hr prior to sacri-
fice, or (2) pulmonary tissue of C57BL/6Cum £ mice (approx-
imately 20 g) induced by IT instillation of 120 ng TCDD/
0.02 ml trioctanoin 48 hr prior to sacrifice.
2AHH = pMoles 3-OH-BP formed per 35 min incubation in
separate tubes containing 25 ug BP/ml and, when necessary,
the indicated levels of 7,8-BF.
3BAR = Number of bacteria at risk; because of variation in
BAR, no mutation frequency is given.
the aromatic amines would seem to mimic the effects of 6-AC
more nearly than those of 2-AF. This is likely since 6-AC is
activated by hepatic tissue, but not by mouse pulmonary tissue;
whereas 2-AF can be activated by pulmonary tissue (see RESULTS
and Table 4).
Whether or not the inability to metabolically activate
smoke condensate is unique to the mouse pulmonary tissue
cannot be answered at this time. Kier et al. (12) reported
-------
CIGARETTE SMOKE CONDENSATE MATERIALS 509
that rat pulmonary S-9's gave only slight increases in number
of mutants with the 1A1 smoke condensate and its fractions.
Hutton and Hackney (28) reported different results using the
1R1 tobacco smoke condensate fractions and induced rat and
normal human pulmonary S-9. These authors observed no
statistically significant increase in mutagenicity of these
condensates with either of these lung-derived activation sys-
tems. We are presently comparing pulmonary tissue from mouse,
rat, and human sources for their ability to metabolically
activate CSC material to biologically active forms.
REFERENCES
1. Wakeham H. Recent trends in tobacco and tobacco smoke
research. In: The Chemistry of Tobacco and Tobacco
Smoke (Schmeltz I, ed.)» New York, Plenum Press, 1972,
pp 1-20
2. Weisburger JH, Cohen LA, Wynder EL. On the etiology and
metabolic epidemiology of the main human cancers. In:
Origins of Human Cancer (Hiatt H, Watson JD, and Winsten
JA, eds.), Cold Spring Harbor, New York, Cold Spring
Harbor Laboratory, 1977, pp 567-602
3. Bock FG, Swain AP, Stedman RL. Bioassay of major frac-
tions of cigarette smoke condensate by an accelerated
technic. Cancer Res 29:584-587, 1969
4. Van Duuren B, Katz C, Goldschmidt BM. Co-carcinogenic
agents in tobacco carcinogenesis. J Natl Cancer Inst
51:703-705, 1973
5. McCann J, Spingarn NE, Kobori J, Ames BN. Detection of
carcinogens as mutagens: Bacterial tester strains with
R factor plasmids. Proc Natl Acad Sci US 72:979-983,
1975
6. Patel AR, Haq MZ, Innerarity CI, Innerarity LJ,
Weisgraber K. Fraction studies of smoke condensate
samples from Kentucky reference cigarettes. Tobacco
176:61-62, 1974
7. Schechtman LM, Kouri RE. Control of benzo(a)pyrene-
induced mammalian cell cytotoxicity, mutagenesis and
transformation by exogenous enzyme fractions. In: Pro-
gress in Genetic Toxicology (Scott D, Bridges BA, Sobels
FH, eds.), New York, Elsevier/North Holland Biomedical
Press, 1977, pp 307-316
-------
510 R.E. KOURI ET AL.
8. Kouri RE, Schechtman LM. In vitro metabolic activation
systems. In: Short-Term In Vitro Testing for Carcino-
genesis, Mutagenesis and Toxicity (Berky J, Sherrod PC,
eds.), Philadelphia, Franklin Inst. Press, 1978, pp 423-
430
9. Weigele M, DeBernardo S, Tenji J, Leimgruber W. A
novel reagent for the fluorometric assay of primary
amines. J Amer Chem Soc 94:5927-5931, 1972
10. Nebert DW, Gielen JE. Genetic regulation of aryl
hydrocarbon hydroxylase induction in the mouse. Fed
Proc 31:1315-1324, 197?
11. Kouri RE, Rude T, Thomas PE, Whitmire CE. Studies on
bred strains of mice. Chem-Biol Interactions 13:317-
331, 1976
12. Kier LD, Yamasaki E, Ames BN. Detection of mutagenic
Acad Sci US 71:4159-4163, 1974
13. Swain AP, Cooper JE, Stedman RL. Large scale fraction-
ation of cigarette smoke condensate for chemical and
biological investigations. Cancer Res 29:579-583,
1969
14. Kouri RE, Demoise CF, Whitmire CE. The significance
of aryl hydrocarbon hydroxylase enzyme systems in the
selection of model systems for respiratory carcinogens.
In: Experimental Lung Cancer, Carcinogenesis and Bio-
assays (Karbe E, Park J, eds.), New York, Springer-
Verlag, 1974, pp 48-61
15. Benedict WF, Rucker N, Faust J, Kouri RE. Malignant
transformation of mouse cells by cigarette smoke con-
densate. Cancer Res 35:857-860, 1975
16. Lazer P, Chouroulinkov I, Izard C, Moree-Testa P,
Hemon D. Bioassays of carcinogenicity after fraction-
ation of cigarette smoke condensate. Biomedicine
20:214-222, 1974
-------
CIGARETTE SMOKE CONDENSATE MATERIALS 511
17. Stanton MF, Miller E, Wrench C, Blackwell R. Experi-
mental induction of epidermoid carcinoma in the lungs
of rats by cigarette smoke condensate. J Natl Cancer
Inst 49:867-877, 1972
18. Gielen JE, Van Cantfort J. Organ selectivity and bio-
chemical characteristics of aryl hydrocarbon hydroxylase
induction by cigarette smoke in rats and mice. IARC,
Scientific Publication No. 12:275-291, 1975
19. Gelboin HV, Huberman E, Sachs L. Enzymatic hydroxyla-
tion of benz(a)pyrene and its relationship to cytotoxi-
city. Proc Natl Acad Sci USA 64:1188-1195, 1969
20. Somogyi A, Kovacs K, Solymoss R, Kuntzman R, Conney AH.
Suppression of 7,12-dimethylbenz(a)anthracene produced
adrenal necroses by steroids capable of inducing aryl
hydrocarbon hydroxylase. Life Sci 10:1261-1271, 1971
21. Lubet RA, Brown DQ, Kouri RE. The role of 3-OH benzo-
(a)pyrene in mediating benzo(a)pyrene induced toxicity
and transformation in cell culture. Res Commun Chem
Path Pharm 6:929-952, 1973
22. Ames BN, Lee FE, Durston WE. An improved bacterial
test system for the detection and classification of
mutagens and carcinogens. Proc Natl Acad Sci USA 70:
782-785, 1973
23. Umeda M, Saito M. Mutagenicity of demethylnitrosamine
to mammalian cells as determined by the use of mouse
liver microsomes. Mutat Res 30:249-254, 1975
24. Gelboin HV, Wiebel FW, Diamond L. Dimethylbenzanthra-
cene tumorigenesis and aryl hydrocarbon hydroxylase in
mouse skin: inhibition by 7,8-benzoflavone. Science
170:169-170, 1970
25. Kouri RE, Ratrie H, Whitmire CE. Evidence of a genetic
relationship between susceptibility to 3-methylcholan-
threne-induced subcutaneous tumors and inducibility of
aryl hydrocarbon hydroxylase. J Natl Cancer Inst 51:
197-200, 1973
26. Kouri RE, Ratrie H, Whitmire CE. Genetic control of
susceptibility to 3-methylcholanthrene-induced subcu-
taneous sarcomas. Int J Cancer 13:714-720, 1974
-------
512 R.E. KOURI ET AL.
27. Kouri RE, Nebert DW. Genetic regulation of suscepti-
bility to polycyclic hydrocarbon-induced tumors in the
mouse. In: Origins of Human Cancer (Hiatt HH, Watson
JD, Winsten JA, eds.), Cold Spring Harbor, New York,
Cold Spring Harbor Laboratory, 1977, pp 811-835
28. Hutton JJ, Hackney C. Metabolism of cigarette smoke
condensates by human and rat homogenates to form
mutagens detectable by Salmonella typhimurium TA1538.
Cancer Res 35:2461-2468, 1975
Note added in proof:
We have recently found that a pulmonary S-9 preparation
from Aroclor 1254-induced mice is capable of weakly activating
6-AC to a bacterial mutagen using a pour plate assay (~2-3
fold over background). We have still not observed an increase
in bacterial mutations using this S-9 preparation and 2A1 cigar-
ette smoke condensate.
-------
MUTAGENIC,
CARCINOGENIC, AND TOXIC
EFFECTS OF RESIDUAL
ORGANICS IN DRINKING
WATER
John C. Loper and Dennis R. Lang
Department of Microbiology
College of Medicine
University of Cincinnati
Cincinnati, Ohio
-------
515
Epidemiologic studies have indicated a possible corre-
lation between pollution of drinking water and incidence of
cancer.. Much of the data for these analyses was collected
during the period 1950-1969. Considering the latency period
for clinical cancers, the findings contribute to the general
concern about long term exposure to the myriad pollutants
in our environment.
Associations have been drawn between enhanced carcino-
genesis and trihalomethane content in drinking water (2). A
complication in documenting such associations as cause-and-
effect relationships is our ignorance of the effects of most
organic compounds. In water, volatile organic compounds
including the trihalomethanes represent only about 10% of
the weight of total organic material. Of the remaining 90%,
it is estimated that 90-95% of the compounds are yet to be
identified (3). Constituent chemicals are present in very
low amounts; identification and toxicological assessment of
even a minority of the total number is a practical impossi-
bility. Moreover, such complex mixtures raise the prospect
of additive, synergistic, antagonistic, or promoter effects
similar to those discussed by others at this symposium.
Toxicological analysis of such mixtures requires some
type of initial concentration procedure. This study was
begun to test the applicability of reverse osmosis in the
concentration of drinking water residue organics. Sequen-
tial samples have been prepared from drinking water of
cities representative of United States municipal water
sources (6,12). Our studies have examined the use of two
-------
516 JOHN C. LOPER AND DENNIS R. LANG
in vitro systems for analysis of such complex mixtures: the
Salmonella/microsome system and BALB/3T3 cell transformation
(8,9). In this paper some of our results will be used to
emphasize problems and current directions in the study.
TEST SAMPLES
Residual organics were prepared for USEPA by Gulf South
Research Institute. The procedure as described by Kopfler
et al.-£4) is presented briefly here (see Figure 1). Solutes
are concentrated from repeated 200 1 samples of tap water,
maintained at pH 5.5 with the addition of HC1, by reverse
osmosis at 15°C using a cellulose acetate membrane (CA); a
Donnan softening loop is included to avoid precipitation of
salts rejected by the membrane. The CA permeate is treated
with NaOH to pH 10 and its solutes are concentrated by a
similar process using a nylon membrane, the nylon permeate
being discarded. Both the CA and the nylon concentrates are
then adjusted to pH 7 and extracted sequentially using pen-
tane and methylene chloride. The aqueous phases are adjusted
to pH <2 by addition of HC1 and methylene chloride extraction
is repeated. Twenty percent of each of the organic fractions
is saved for chemical analysis, while the bulk of the material
is concentrated and combined to generate the reverse osmosis
concentrate-organic extract fraction (ROC-OE).
The remaining concentrates are purged of excess solvent
by bubbling with N2 and are passed through columns of XAD-2
resin. After column rinses of 1M HC1 and of distilled H^p,
the organics are eluted using 95% ethanol. Eluent solutions
are dried with sodium sulfate, concentrated by vacuum distil-
lation of solvent, and pooled to generate the XAD eluate frac-
tion (XAD eluate). Both ROC-OE and XAD eluate fractions were
stored at 4°C in sealed containers before delivery and have
been maintained similarly in the dark between samplings. A
portion of each ROC-OE and XAD eluate was further fractionated
by sequential extraction with hexane, ethyl ether, and acetone
according to a method of R.G. Melton (USEPA report, Cincin-
nati, 1976). For later samples the pentane and methylene
chloride extracts of ROC were obtained as discrete fractions.
It was intended at the outset that the ROC-OE residues
should be obtained in 1 g amounts. A comparison of the ini-
tial water volumes used and yields of the samples provided
for our examination appear in Table 1. For most samples
total organic carbon in the water ranged from 6.4 to 1.7 ppm,
and volumes of 2000 to 8000 liters were sufficient to generate
-------
EFFECTS OF RESIDUAL ORGANICS IN DRINKING WATER
517
WATER SAMPLE
i
R.O. Cellulose Acetate
R.O. Nylon
Cellulose Acetate
Concentrate
Nylon Concentrate
r
Pentane pH7
i
r
Methylene
Chloride pH7
.
,
Methylene
Chloride pt-
'
r
XAD-2
pH2
2
80%
20% sample
\
Ethanol
Elution
ROC-
OE
80% i
20% sample
XAD
Eluate
/
/
Ethanol
Elution
\
discard
Pentane
i
pH7
r
Methylene
Chloride pH7
\
F
Methylene
Chloride pH2
i
XA
P*
F
D-2
^2
\
discard
Figure 1. The origin of reverse osmosis concentrates-organic
extract (ROC-OE) and XAD eluate fractions is shown in a dia-
gram of the procedure of Kopfler et al. (6). Twenty percent
of each organic extract, indicated by the short arrows, is
removed and stored for chemical analysis, and the remaining
portions are pooled and concentrated to constant weight to
form the ROC-OE. The remaining aqueous solutions are purged
of excess solvent using N2 and are passed through columns of
XAD-2. Organics eluated in 95% ethanol are dried, pooled,
and concentrated under vacuum to yield the XAD eluate.
about 1 g of ROE-OE. Seattle and Tucson drinking water yield-
ed less than 0.4 g amounts of ROC-OE from much larger sample
volumes.
-------
518
JOHN C. LOPER AND DENNIS R. LANG
rH
CD
rH
XJ
03
EH
*
H
CD
-p
^
bO
G
•H
24
G
•H
rH
Q
S
0
^-|
ryi E-] (V)
do o
S 1
CO -O
O CD O
-P 05
(13 crt y j
CO rH
(H -P -P
CD G 0
> y> ^
^^ o
rH rH CD
03 03 -P
-P O 03
o *
£_l
T3
CD
03 -P
-P 0
03 CD
Q rH
rH
O
O
SH T3 x-%
CD CD CO
•P CO rH
05 CO CD
^g fli Jjr)
O -H
ft O rH
05 rH v^
EH ft
CD
*>t rH
•p a
•H E
O oS
0"
COCO'tf(NGOa5O5T}<
^ in co 0s" oo co ^ co
CMO^THOCOCOC35
rHTHr-lrHTHOOO
CO CO C^ ^t^ ^^ Cs* C"* CO
CMCNIrHCOCOrHTHCO
inincoincoincocD
[s. f*w, fx. ^^ rjN. r%» f^ fv.
— -. . ^v, ~~v. ^^ — •» "^^ "^^
OOTHTHCMCMCMCO
rH rH rH TH
^* CM ^^ O5 rH ^* "sf* TH
O5rHCMO5^r-lrHCO
O500CD05CMOOOOO5
i>i>coTHCMininin
OS ^2 TH CM
rH rH CM
03 d
CO CO CO -H -H
G G c x: x:
OS OS OS ft ft rH
CD CD> CD Ol rH rH
rHrHrHTHCMCDOoS
f-t ?H JH *O "O ^
O O O -H -H oS oS E
E E rH rH 3
S ? > 03 03 -H -tH -P
CD CD CD -H -H J3 X3 -P
SZSZSSftQjO
m
00 O
CD CO
00
O CO
rH O
•
< o
Z TH
CO CD
c^ o
05 C-
o "^r1
in o
rji CD
m CM
TH
CM TH
03 CD
? rH
S -P
3 -P
-P 03
-P CD
O CO
o
rjH C-- =H H 1 ••
CM O -P «H CD <
a o •
CD -P C J3
O -H O
X G O
W 3
CD -a
bO G CD
• G 03 T3
< -H rH 3
ft G J2 rH
W CD S O
•P CD G
?^ HH fe 'H
^j1 m X2 o
CO CM CO CD O
O O CD G oS rH
T3 oS -P OS
•H G CD
> a o co
O O 03 03
w a ?
a CD
05 co a
CD O O
H bOrH O
0> G 3 rH
3s -H i— 1
CO rH CD
CO 3 CD bfl
< CD 0 C
Iz; a CD 0 .c
S CO XI O
03 CO -P X
CO 0) 0)
0 fl
CD O .H G
X3 rH 0
-P ft CO -H
G G
G rH O OS
CO l> O rH -H
O C- 03 -P G
*\ ^^^ oj oS oS
rH O V CD O
rH rH 03 ?H CD
T3 CD X3 rH
£ -p ft
CO -H£
•rH CO £ 03
X3 CD CO
CM rH -P rH CO
m co a G c
C- CD T3 £ O O
•H f- aS CO O
rH CM S3
CO - 3 EH
G oj -H
O rH TJ CD
•H 0 X3 •
-P T3 CO -P CD
O CD rH
05 rH
rH G TS CD OS CD 03
-P O -H CO X3 -P
-P CO CO O O oS -P
03 o CD x: x fH o
CD 3 05 -P CD -P G
CO EH *
-------
EFFECTS OF RESIDUAL ORGANICS IN DRINKING WATER ' 519
SALMONELLA/MICROSOME TESTING
The assay system has been described by Ames et al. (1),
who provided the strains TA1535, TA1538, TA98, and TA100.
Promutagen activation was conducted in soft agar overlays
using S-9 mixtures prepared from livers of rats induced with
a PCB mixture Aroclor 1254; characteristic activation poten-
tial of each homogenate preparation was verified using known
promutagens (9), and activation of 2-aminoanthracene served
as a positive control during tests of the unknowns. Samples
were dissolved in dimethyl sulfoxide (DMSO) and were delivered
in volumes of 0.01 to 0.3 ml/plate. Tests involved duplicate
platings of 5 doses over a 30-fold dose range for mutagenesis
of TA98 and TA100 in the absence of S-9 mix. This assay was
then repeated, with dose adjustments as appropriate, in an
expanded protocol including the addition of the activation
system, and optionally the strains TA1535 and TA1538. For
assays in which little or no cell killing was evident, dose
responses of net revertant colonies/mg of sample were deter-
mined from linear regression plots generated with a computer-
plotter; otherwise initial rates were used. In nearly all
cases mutagenesis for a tester strain was determined from
data which included experimental colony counts which were
at least twice those obtained from the spontaneous control
plates. Presence of characteristic pinpoint histidine-
requiring colonies and appearance of less than spontaneous
colony counts were recorded as having apparent lethal toxi-
city. Bioassays for histidine in samples were determined
turbido-metrically using strain hisDC129, a stable deletion
histidine auxotroph of Salmonella typhimurium.
BALB/3T3 TRANSFORMATION AND TOXICITY TESTING
We obtained clone 1-13 BALB/3T3 cells from Dr. Takeo
Kakunaga of the National Cancer Institute, Bethesda, Maryland,
Cells were routinely maintained at sub-confluence in anti-
biotic-free Eagle's minimum essential medium (MEM) which was
supplemented with 10% heat inactivated fetal calf serum.
Cells were incubated in a humidified atmosphere of 5% C02 in
air.
The experimental,conditions were essentially those
described by Kakunaga;(30. Cells were plated at a concen-
tration of 10* per 60 mm cell culture dish in 5 ml media and
incubated overnight. Appropriate concentrations of carcino-
gen or water sample were then added in 0.01 ml DMSO. Control
plates received 0.01 ml DMSO alone. Cultures were incubated
-------
520 JOHNC. LOPER AND DENNIS R. LANG
for 72 hours after which time the media was removed, cells
rinsed once with phosphate buffered saline (PBS), and refed
with fresh media. Cultures were maintained for an addi-
tional four weeks on a bi-weekly feeding schedule. Cells
were then rinsed with PBS, fixed with methanol, and stained
with Giemsa. Areas of piled up cells growing in a disorgan-
ized, criss-cross pattern were quantitated as foci. Prior
to fixing, cells from foci and from normal appearing areas
were cloned for isolation and storage in liquid nitrogen for
eventual testing of their in vivo tumorigenicity. Cytotoxi-
city was assayed by determining the plating efficiency of
200 cells plated in 5 ml MEM per 60 mm dish with exposure to
test compounds as described for the transformation assay.
RESULTS AND DISCUSSION
Aspects of Salmonella Mutagenesis Testing
Tests of residues and residue subfractions from each of
the samples listed in Table 1 have been conducted using two
or more strains of the Salmonella testing system; mutagenesis
was induced by residues from each drinking water sample.
Where possible we have tabulated and compared results of
repeat samples from a given city and among cities, giving
attention to (a) the amount of mutagenesis for a strain, as
expressed in terms of net revertant colonies/mg of residue
material tested; (b) the relative mutagenicity of the test
material for TA98 and TA100; and (c) the distribution of
mutagenic activity among ROC-OE and XAD eluate fractions and
their subfractions. Repeat samples have exhibited consis-
tencies of mutagenic patterns that were characteristic for
that city.
The data and our analysis have been presented elsewhere
in detail (8,9). In this paper we describe some of our
general findings in assay of these complex mixtures. All
the Salmonella mutagenesis measured to date has been direct
acting, with little or no enhancement due to the presence
of the microsome activation system. Direct mutagenic activ-
ity for TA1538 was usually similar to that seen using TA98,
while TA1535 often was unaffected in cases where TA100
showed a response. An example of these patterns appears in
Figure 2.
Many of the fractions tested gave linear dose response,
and with most of these we were able to test amounts of
material which yielded colony counts from responding strains
-------
EFFECTS OF RESIDUAL ORGANICS IN DRINKING WATER
521
40
\
5
'A 1538
400
200
TAI535
/v
750
1500
0
PLATE
750
1500
Figure 2. Strain specific mutagenic effects of Miami 2 ROC-
OE. Each point is the average of colony counts from 2 plates
Lines were drawn as linear regressions of the original data.
All assays presented in this and in the following figures
were conducted in the absence of S-9.
of two-fold spontaneous or better. Dose response data involv-
ing colony counts of this magnitude are convenient since nor-
mal appearing dose dependent increases of less than two-fold
could be due to histidine enrichment. We considered this a
possibility since large amounts of water were processed to
generate these residues. Salmonella typhimurium strain
hisDC129 is an organism which grows well in histidine enriched
media but is stably dependent upon the presence of the amino
acid for growth; the strain is thus convenient for use in
turbidometric microbiological assays of histidine. In cases
where test fractions showed marginally two-fold mutagenesis,
use of this histidine bioassay showed negligible histidine
in the samples.
In certain instances assay for mutagens in these mix-
tures was complicated by antagonistic or toxic effects. With
some fractions the lethal effects simply precluded assay for
-------
522
JOHN C. LOPER AND DENNIS R. LANG
mutagens; with others the mutagenic dose responses were non-
linear, some showing a masking of further mutagenic responses,
and some showing clear toxicity at higher dose. In such cases
the mutagenesis was scored as present or was calculated from
the initial rate of response. Plots of data from samples rep-
resentative of such mixtures appear in Figures 3-5.
For samples from the first 5 cities listed in Table 1
cell killing effects were determined by suspending TA100
cells in a fixed concentration of sample and establishing
the decrease in colony forming cells as a function of min-
utes of exposure. By this method ROC-OE fractions showed
two- to five-fold greater toxicity per mg than did the cor-
responding XAD eluates (unpublished observations). As noted
below this trend was even more pronounced in determinations
of cellular toxicity for clone 1-13 cells.
3001-
i
0.5 1.0 15
>jg/ PLATE
2.0
\
I
200 -
O-O-
100 -
100 200
jug /PLATE
Figure 3. Typical non-linear
dose response curve of muta-
genic effects of Philadelphia
2 ROC-OE on strain TA100.
Figure 4. Dose response
curve representative of
apparent mutagenic plus
antagonistic effects. The
data show effects of
Ottumwa 1 XAD eluate on
strain TA100. No calcula-
tions of net revertant
colonies/mg were attempted
for such responses.
-------
EFFECTS OF RESIDUAL ORGANICS IN DRINKING WATER
523
600 r-
I
K.
I
300
0 0.2 0.4 0.6 0.8 1.0
ml / PLATE
Figure 5. Mutagenic and toxic response of strain TA100 to
increasing volumes of aqueous concentrate from Philadelphia
drinking water. A volume of concentrate obtained by reverse
osmosis using the cellulose acetate membrane, provided to us
by EPA, was concentrated ten-fold further by lyophilization.
Test volumes were incorporated directly into the soft agar
in the standard assay procedure.
Little information is available as to the chemical agents
causing this toxicity. As reported elsewhere (9) for various
residue fractions from the first 5 cities, oxidation of the
organics by refluxing in nitric acid or in a mixture of nitric
and sulfuric acids removed bacterial toxicity. In contrast,
the ROC-OE and XAD eluate fractions obtained from the Tucson
sample showed unusually high bacterial toxicity, and the
toxicity of that XAD eluate fraction was stable to oxidation.
Spectrographic examination of the Tucson XAD eluate fraction
showed 1800 +_ 200 ppm of Hg, together with lesser amounts of
other metals. Control toxicity experiments using reagent
-------
524 JOHN C. LOPER AND DENNIS R. LANG
HgCl2 showed that, should the mercury content of the XAD
eluate be present as Hg++, it could account for all the tox-
icity of that sample. A relatively large sample of drinking
water, 27,631 liters, was processed to yield the Tucson frac-
tions and this concentration of mercury ion, calculated per
liter of original drinking water, would be well below accept-
able levels (9).
Cell Transformation and Toxicity
We have previously reported data on the transformation
of BALB/3T3 cells by the ROC-OE fraction from New Orleans Ib
sample (8,9). Clones obtained from transformed foci have
demonstrated enhanced plating efficiency in soft agar using
the technique described by MacPherson and Montagnier (10).
We have also shown that the BALB/3T3 cells differentiate
between ROC-OE and XAD eluates on the basis of cellular
toxicity. For four cities examined, toxicity on a weight
basis was ten- to twenty-fold greater for the ROC-OE samples
than for the corresponding XAD eluates (9, and unpublished
observations). Tests of transformation activity of addi-
tional samples are in progress.
We have been attempting recently to develop a mutagen-
esis assay with this same clone of 3T3 cells using ouabain
resistance as a marker. Huberman et al „ have shown about a
20:1 ratio of transformation to mutation frequency occurred
when both were measured in hamster embryo cells (4). We are
attempting to see if similar measurements can be made with
BALB/3T3 cells using the focus assay for transformation
rather than the colony assay employed by Huberrnan. Prelim-
inary data indicate that these cells may lend "hemselves to
studies of mutagenesis at the locus for the Na"~K ATPase. If
the assay can be developed, it would be of obvious utility
in describing the relative carcinogenic and mutagenic activ-
ities present in the complex mixtures obtained from drinking
water.
We are also attempting to develop a tumor promoter
assay with these BALB/3T3 cells. It has been shown by Mondal
et al. (11) that the tumor promoter tetradecanoylphorbol ace-
tate (TPA) can have stimulatory effects on transformation of
10T*3 cells by known carcinogens under conditions where there
is no transformation by either carcinogen or TPA alone. If
we are successful in extending this observation to BALB/3T3
cells, we will be able to test the promoting activity of ROC-
OE or XAD eluate fractions on transformation initiated by
-------
EFFECTS OF RESIDUAL ORGANICS IN DRINKING WATER 525
3-methylcholanthrene or other carcinogens known to be active
in this svst.em.
Relationship to Drinking Water
Many problems remain in relating the available data to
the frequency and variety of mutagens/carcinogens in the
original water samples, and several approaches are in pro-
gress or are planned to address these.
1. Although residues with specific mutagenic proper-
ties are reproducibly isolated, we do not know how represen-
tative these residues are of the total organics in water,
Mutagens may be preferentially concentrated or preferentially
lost. Some compounds may have been chemically altered during
concentration, extraction, or storage. In one study, the
mutagenic potential of New Orleans 2 ROC-OE was equally stable
over a one week period when sealed in serum vials and stored
at room temperature in DMSO, or when stored at -70°C in either
DMSO or dimethyl formamide. But more attention to limiting
oxidation throughout the procedure may be important.
From the data of Gulf South Research Institute (6),
nearly all of the drinking water TOG is retained during con-
centration to the reverse osmosis membrane reject volume.
Aqueous concentrates from the cellulose acetate membrane were
provided by EPA from New Orleans, Miami, and Philadelphia.
No mutagenesis was detected in assays of the first two of
these samples, but a further 10-fold concentration of the
Philadelphia sample gave the dose dependent mutagenic-toxic
response presented in Figure 5. We have initiated tests of
the transforming activity of this high salt material, and we
have begun examining alternate methods of extraction of the
mutagenic activity. If we are successful in establishing
the TPA promotion assay in BALB/3T3 cells, we will test the
promoting activity of these aqueous concentrates as well.
2. The complexity of these mixtures may lead to a
variety of additive and antagonistic effects. Some fractions
are too toxic for reliable determination of mutagenicity.
Fractions yielding dose responses of the types shown in
Figures 3-5 may contain components that prevent accurate
measure of the mutagens present. Microsomal activation has
not been required for the mutagenesis we have detected to
date, but here too antagonistic effects of compounds in cer-
tain of these mixtures may mask detection of mutagens requir-
ing activation. The mutagenesis/mg that we measure could be
-------
526 JOHN C. LOPER AND DENNIS R. LANG
due to a broad range of chemicals of different specific muta-
genic activity.
An initial fractionation was included in the survey
study of samples provided from the six cities, by which ROC-
OE and XAD eluates were sequentially extracted using hexane,
ethyl ether, and acetone. Mutagenic assays on these subfrac-
tions helped identify differences among samples from separate
cities and also revealed some common patterns of distribution
of active components (9). However, even these subfractions
contain a great number of components, and the identification
of the active species by direct analytical methods of GC-MS
will be impossible; chemical analysis in progress on one sub-
fraction of a Cincinnati water ROC-OE so far has revealed
several hundred compounds (E. Coleman, personal communication)
By combining selected solvents with acid, neutral, or basic
aqueous phases, and through application of HPLC, our group
will attempt separation of the bulk of the components into
smaller subfractions, monitoring progress in fractionation
using the Salmonella/microsome test. Active fractions may
be obtained sufficiently free of inactive and toxic compo-
nents to facilitate peak-to-peak identification by GC-MS.
In addition, a number of pooled component studies are
possible. Kraybill et al. have compiled a list of direct
acting mutagenic compounds known to be in finished or raw
water (7), and some of these are sufficiently non-volatile
as to be retained in ROC residues. We plan to characterize
mutagenic separation properties of such mutagens in prepared
mixtures. Using known mixtures, and available mutagenic
organic residues, a number of water reconstitution-reconcen-
tration experiments can be initiated. These studies should
allow us to define more clearly the significance of mutageni-
city and in vitro carcinogenicity found for reverse osmosis-
derived residue organics of drinking water.
This work was supported by research grant R804202 from
the USEPA.
REFERENCES
1. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian
microsome mutagenicity test. Mutat Res 31:347-363, 1975
-------
EFFECTS OF RESIDUAL ORGANICS IN DRINKING WATER 527
2. Control of organic chemical contaminants in drinking
water, Environmental Protection Agency Interim Primary
Drinking Water Regulations, U.S. Federal Register 43:
5756-5780, 1978
3. Drinking Water and Health, Report of the National
Research Council Safe Drinking Water Committee, National
Academy of Science, p 492, 1977
4. Huberman E, Mager R, Sachs L: Mutagenesis and transfor-
mation of normal cells by chemical carcinogenesis.
Nature 264:360-361, 1976
5. Kakunaga T: A quantitative system for assay of malignant
transformation by chemical carcinogens using a clone
derived from BALB/3T3. Int J Cancer 12:463-473, 1973
6. Kopfler PC, Coleman WE, Melton RG, Tardiff RG, Lynch SC,
Smith JK: Extraction and identification of organic
micropollutants: Reverse osmosis method. Ann NY Acad
Sci 298:20-30, 1977
7. Kraybill HF, Helmes CT, Sigman CC: Biomedical aspects
of biorefractories in water. In: Proceedings Second
International Symposium on Aquatic Pollutants, Oxford,
England: Pergamon Press, Ltd., in press
8. Loper JC, Lang DR, Smith CC: Mutagenicity of complex
mixtures from drinking water. In: Proceedings of the
Conference on Water Chlorination Environmental Impact
and Health Effects, Chapter 33. Ann Arbor Science Pub-
lishers, Inc., pp 433-450, 1978
9. Loper JC, Lang DR, Schoeny RS, Richmond BB, Gallagher
PM, Smith CC: Residue organic mixtures from drinking
water show in vitro mutagenic and transforming activity.
Submitted for publication
10. MacPherson I, Montagnier L: Agar suspension culture for
the selective assay of cells transformed by polyoma
virus. Virology 23:291-294, 1964
11. Mondal S, Brankow DW, Heidelberger C: Two-stage chemi-
cal oncogenesis in cultures of C3H/10T cells. Cancer
Res 36:2254-2260, 1976
-------
528 JOHN C. LOPER AND DENNIS R. LANG
12. Tardiff RG, Carlson GP, Simmon V: Halogenated organics
in tap water: a toxicological evaluation. In: Pro-
ceedings Conference on the Environmental Impact of Water
Chlorination, pp 213-227, 1975
-------
MUTAGENIC ANALYSIS OF
COMPLEX SAMPLES OF
AQUEOUS EFFLUENTS, AIR
PARTICULATES, AND FOODS
Barry Commoner, Anthony J. Vithayathil,
and Piero Dolara
Center for the Biology of Natural Systems
Washington University
St. Louis, Missouri
-------
531
INTRODUCTION
Opportunities and problems arise when the Ames muta-
genesis technique is applied to the analysis of samples, such
as those derived from the environment, which are mixtures of
unknown compounds that may or may not include mutagens. The
chief advantage of this application of the method is well
known: one can use it as a rapid, inexpensive, biological
screen capable of detecting mutagens.by their biological
effect. This makes it possible to avoid the very difficult
task of detecting and identifying all of the numerous organic
compounds that may occur in such a sample in order to compare
them with a list of known mutagens. The chief disadvantage
of this approach is that one is "flying blind," so to speak,
unaware in advance of what types of compounds are present,
their concentrations, and their possible interference with
the test.
In order to appreciate these difficulties and to devise
strategies for overcoming them, it is useful to recall cer-
tain characteristics of the Ames system:
• The various Ames strains of Salmonella are designed
specifically to respond to different classes of
organic mutagens. Therefore, in dealing with sam-
ples containing unknown mutagens one cannot know
in advance which strains the mutagens will act
upon. For the same reason there is no a. priori
basis for quantitative comparisons of mutation
rates obtained with different strains.
-------
532 BARRY COMMONER ET AL.
• The dose-response curves that relate a given
strain's response to various concentrations of a
given mutagen are almost always decidedly nonlinear,
in some cases falling to a zero response at high
concentrations. This means that a test designed
to determine whether or not mutagens are present,
if carried out at only one concentration, may
readily give a false negative result. For the
same reason, a positive value obtained at a single
concentration is insufficient to estimate the level
of mutagenic activity.
• Certain mutagens are inherently active in the
system, while other needs to be "activated" by
the "S-9" microsome preparation. However, the
latter is a complex system of related enzymes and
there is no way of knowing in advance whether the
microsome preparation will convert a particular
substance to an active rrmtagen and whether, on the
contrary, it will convert an inherently active
mutagen into an inactive substance.
« There is a certain inherent biological variability,
from time to time, in the background rate of muta-
tion of each of the Ames strains. At the same
time, as in any experimental procedure, there are
certain sources of imprecision (e.g., in volume
measurements) that also affect this value. This
raises the question of how these two sources of
variability are to enter into the computation of
the experimental results.
• Despite its considerable value, the system is
still something of a "black box" because certain
features are poorly understood. These features
include, in addition to the unresolved properties
of the microsome preparation, synergistic and/or
inhibitory interactions between mutagens concur-
rently present in the system, and possible trans-
formations of test substances by enzymes associated
with Salmonella.
In our present circumstance it is useful, while using the
Ames test, to remain alert to anomalies that may provide
useful clues for learning more about how the test works.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES 533
We will consider how the foregoing features of the Ames
system may affect the results obtained from complex, unknown
samples, and suggest some procedures which may offset the
resultant difficulties. We will present specific examples
of analyses of unknown mutagens present in samples of water,
air particulates, and food. The analyses presented are con-
cerned with one or more of the following general aims, which
commonly arise in applying the Salmonella system as a screen
to environmental and other complex samples:
• Isolation and identification of active mutagens
from complex unknown samples.
• Evaluation of the level of mutagenic activity
associated with a complex unknown sample, especially
in relation to relevant environmental parameters.
• Characterization of such samples with respect to
the presence of inherently active mutagens, muta-
gens capable of being active, and mutagens that
are inactivated by the microsome preparation.
• Comparison of mutagens detected in complex samples
with known ones.
• Application of techniques for studying the formation
of mutagens in experimental systems.
The following specific examples are discussed: (a) de-
tection and isolation of mutagens in the aqueous effluents of
petrochemical plants along the Houston Ship Channel; (b) anal-
ysis of the mutagenic activity of Chicago air particulates;
(c) analysis of a minor anomaly in the Ames test that has led
to the discovery of situations in which mutagens are produced
during conventional cooking of certain foods.
DETECTION AND ISOLATION OF MUTAGENS IN THE AQUEOUS EFFLUENTS
OF PETROCHEMICAL PLANTS ALONG THE HOUSTON SHIP CHANNEL
These studies have been carred out under a collaborative
arrangement with the Harris County Pollution Control Depart-
ment (Pasadena, Texas). First, water samples (two gallons
each) were collected directly from the effluent pipes of the
various chemical plants under the joint supervision of the
Pollution Control Department staff and the plants' personnel.
The samples were stored in our laboratory at 4°C.
-------
534 BARRY COMMONER ET AL.
A total of 24 effluent samples were collected from 16
different industrial plants (see Table 1). At two locations
samples were collected from the same outflow pipe on a series
of dates. Initially, benzene/isopropanol extracts of each
sample (usually 2 liters of water extracted successively at
pH 2.5 and pH 11) were dried, then dissolved in DMSO. Ali-
quots representing varying amounts of the original water
samples were tested against strain TA1538 with and without
the liver microsome preparation, in keeping with the proce-
dures described by Ames et al. (2). Throughout the work
described in this paper, the microsome preparation used was
the standard S-9 preparation from the livers of PCB-induced
rats. All plate counts reported are the averages of dupli-
cate plates. These techniques of sample preparation and
mutagenesis testing are not suitable for volatile compounds,
and such compounds are not involved in our studies.
The results of some initial tests of the acid extracts
of the samples are shown in Table 1 (alkaline extracts were
uniformly negative). In interpreting the significance of
these results, we have employed an approach developed earlier,
based on the comparison of 50 known organic noncarcinogens
and 50 organic compounds that previously had been shown to
be carcinogenic toward laboratory animals (3). In this com-
parison we computed a "mutagenic activity ratio" from the
E—P
quotient p—, where E is the number of mutant colonies
LAv
obtained from the experimental sample; C is the control
value (i.e., the number of mutant colonies observed when the
experimental material is not included) obtained on the day
of analysis; and C. is the "historical" control value, or
the average control value for all runs carried out during
the course of the study. The rationale for this procedure,
which was described earlier (3), is intended to take into
account daily variations in the background mutation rate as
well as those variations inherent in the method itself.
As we have shown previously, in the test of equal num-
bers of known carcinogens and noncarcinogens, the reliability
with which the two classes of compounds can be distinguished
E—C
depends on the value of ~p- which is chosen as the cut-off
CAv
point. Thus 82% of the noncarcinogens yield a mutagenic
activity ratio below 2, and 82% of the carcinogens yield a
ratio above that value. If a higher reliability of detecting
carcinogens is desired, a somewhat lower cut-off ratio is
chosen, at the risk of increasing the chance of falsely
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
535
Table 1
Analysis of pH 2.5 extracts of samples from industrial
sources in Houston Ship Channel area
Plant
A
A
A
A
a
a
B
C
C
D
D
E
F
F
F
F
F
G
ri
-
K
-
M
N
0
P
Type of Plant
Puip .mil
Pulp mill
?/pe of
Sample
black Ii4\jor
isater
Pulp mill .later
Pulp mill Water
steel in 11
water
Steel mi 11 water
Steel rrul 1 Later
Chemical wat-er
Chemical
Chemica.1.
Zhemicaj.
Chemica i
Chemical
Chemical
Chemical
Chemical
Chemical
-hercica^
Chencal
-«-.. = al
Che-ical
Industrial -faste -„:<•_-_-<-•- r_
Chemical
Chemicaj.
later
water
'vater
hater
Sludge
-ater*
water"
•Vater"
.-later*
v.ater
.<««
-•»t,r
.Vater-
^att r
nat -i
>ater
Chemical Water
chemical
Water
Date
Collected
10/3/75
10/3/75
10/3/75
10, 3/75
3-5/75
9/5/75
9/5/75
9 23, 75
3 23,75
5/26, "5
1/5,76
5, 19-75
6,26,75
1 6, 7C
1, 6/76
1/6/76
1/6/76
9/ 43/^5
9/23/75
1 / 5 / " 6
1/5/76
1/5/76
I, 5,76
1/5,76
1, 5,76
I, 5/76
. 5, 76
Date
Analyzed
10/30/75
10/20/75
10/20/75
10, 20/75
Cquiva Lent
Amount of
Sample/Plate
lull)
1
50
125
250
10,16,75 25
10, 16/75
10/16, 75
10/29/75
.0/29 75
10'U/^5
6/15/'"6
10/23/75
10, 25/75
6,22/76
6. 22,76
6, 22/76
6/22/76
10/30/75
'
6 / 1 5 '76
6/22/76
6/22/76
6/22 '76
6 22-^6
6 :2,">r,
!/" '!
'. '6/76
t / (> / 7 6
.,/!'> 71,
.' I r; 7 f,
j
L <) , 7 6 6 i 5 , 7 6
62.5
125
125
250
100
250
25
3
250
250
250
250
U°5
1 00
2 50
250
250
250
250
250
750
250
250
250
250
No. ot Colonies/Plate
(TA1538, Hver 59)
Control
31
41
41
41
42
42
42
26
26
42
27
22
17
21
21
21
21
31
2t
21
21
21
21
, ^
21
27
27
27
Experimental
36
32
47
63
41
64
17
86
32
41
36
11
35
29
40
35
38
127
568
48
36
35
59
28
29
32
37
131
63
Mutaqenic
E-C
0.2
-0.4
0.3
1.0
0
1.0
-1.1
2.6
2.4
0
2.7
0.5
0.8
0.4
0.9
0.6
0.8
4.2
24 . 2
24 . 6
1.2
0.7
0.6
1. 7
0.3
0 . 3
0.5
0 .7
4. 7
1.6
•These samples were collected from 4 <_ii: ferc-nt off lucent outlets t com the same
-------
536
BARRY COMMONER ET AL.
identifying noncarcinogens. In more general terms, ratios
of 2-3 should be regarded as at least suggestive of the
presence of mutagenic activity; values above the range of
3-5 are clearly indicative of the presence of mutagenic
acitivity.
From the values shown in Table 1 it appears that efflu-
ents from industrial waste treatment Plant G consistently
yielded significant levels of mutagenic activity. We have
carried out a systematic analysis of this effluent designed
to isolate and identify the substances responsible for the
observed activity. Figure 1 shows the procedures we applied
to a 20-gallon sample of effluent from Plant G.
Water Effluent (pH 7.0)
Benzene:isopropanol (8:2)
extraction
f
Water Layer
Adjusted to pH 2.5
followed by benzene:
isopropanol extraction
Solvent Layer
Evaporation
Residue (7R) tested
t
Water Layer
Adjusted to pH 11
followed by benzene:
isopropanol extraction
t
Solvent Layer
Evaporation
Residue (2.5R) tested
t
Water Layer
(discarded)
Solvent. Layer
Evaporation
Residue (11R) tested
Figure 1. Procedures applied to a 20-gallon sample of
effluent from Plant G.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES 537
Aliquots of the residues yielded by this scheme (7R,
2.5R, and 11R), equivalent to 200 ml of the original effluent,
were tested for mutagenic activity in the usual way, with
strain TA1538, in the presence of the microsome preparation.
The results are shown in Table 2. It is apparent that the
neutral and acidic fractions are clearly active, while the
activity of the alkaline fraction is marginal.
Table 2
Mutagenic Activity of Industrial Waste Treatment
Plant Effluent Extracted with Benzene:
Isopropanol at Different pH
Sample
Number
7R
2.5R
11R
pH of
Extraction
7
2.5
11
Number of Revertant
Colonies/Plate*
(TA1538, Liver
Microsomes)
2116
839
87
Mutagenic
Activity Ratio
E-C
CAv
95.2
37.1
3.0
*Colonies/plate for equivalent of 400 ml of the water sample.
In the next step, aliquots (representing 200 ml of the
original effluent) of the neutral (7R) and acidic (2.5R)
residues were subjected to thin-layer chromatographic (TLC)
fractionation using silica gel paper and a benzene:hexane
(1:1) solvent. A series of 1 cm sequential zones were then
cut from the developed chromatogram, each extracted in 10
percent methanol in chloroform and allowed to dry. The
successive zonal samples were then taken up in DMSO and
tested in the usual way against TA1538 in the presence of
the microsome preparation. From the numbers of mutant colo-
nies produced by each zonal sample it was possible to charac-
terize the chromatographic behavior of the mutagenically
active constituent(s).
-------
538 BARRY COMMONER ET AL.
As shown in Figure 2, following this chromatographic
procedure the mutagenic activity of both of the fractions was
found predominantly lodged at the origin. However, UV scans
of the chromatogram showed that several mutagenically inactive
components had moved away from the origin, so that this chroma-
tographic system was a useful means of initial purification of
the sample.
The zones located at the origins of the foregoing chroma-
tograms were eluted with 10 percent methanol in chloroform and
were rechromatographed using methanol:ethyl acetate:benzene
(1:10:89) as the solvent system. The results of this second
fractionation step are shown in Figure 3 for the 2.5R (acidic
fraction).
How shall we interpret this result? The most obvious
interpretation is that the material at the origin of the
first chromatogram (Figure 2) was heterogeneous, and in the
second chromatogram, resolved into three peaks (at the origin,
at RF = 0.8, and RF = 1.0). The two peaks at RF = 0.8 and
at RF = 1.0 presumably represent two different mutagens. But
this interpretation holds only if the dose-response curve is
linear. If instead the dose-response curve goes through a
maximum and falls to zero at higher concentrations, the appar-
rently double peak in Figure 3 may actually represent a single
substance. For example, Zone 8 might represent a relatively
low concentration of the mutagen, which lies on the rising arm
of the dose-response curve, Zone 9, a higher concentration
which is on the falling arm of the dose-response curve, and
Zone 10, once again a relatively low concentration which lies
on the rising arm of the dose-response curve. Thus, in actu-
ality the mutation rate values for Zones 8, 9, and 10 might
represent a^sjingle chromatographic peak centered at Zone 9.
This example is cited only to provide an illustration
of the impact that the possible non-linearity of mutagenic
response to a particular unknown substance may have on the
otherwise simple problem of interpreting chromatographic
peaks. It emphasizes once more the importance of actually
measuring dose-response curves in dealing with such samples.
MUTAGENIC ANALYSIS OF CHICAGO AIR PARTICULATES
A number of organic compounds that include carcinogens
have been found to be associated with urban air particulates.
Accordingly, analysis of such material represents another
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
539
TLC Fractionation of Industrial Plant G Effluent.
2000-
1000-
_
£ 500
_g
S 200
a
1 100
tr
50-
20-
10-
Solvent System: Benzene: Hexane (1:1)
TA 1538; Rat Liver Microsomes
• Neutral Extract (7R)
O Acidic Extract (2-5R)
-I 01 23456789 10
Chromatographic Zone
(cm. from origin)
Figure 2. Thin layer Chromatographic fractionation of the
neutral (solid lines) and acidic (broken lines) extracts of
a sample of effluent from industrial plant G. The Chromato-
graphic solvent system was benzene:hexane (50:50). Chromato-
graphic fractions were tested using TA1538 with microsome
preparation present.
-------
540
BARRY COMMONER ET AL.
900
TLC Ref ractionation of Industrial Plant 6 Effluent.
(From Zone at Origin of Acidic Extract of Benzene: Hexane
(I:I) TLC)
2
CL
\
.2
"o
$
800-
700-
600'
500
400-
O 300-
ib_
| 200-
z
100-
Solvent System: Methanol: Ethylacetate: Benzene (1:10:89)
TA 1538; Rat Liver Microsomes
01234567
Chromatographic Zone
(cm. from origin)
89 10 il
Figure 3. Thin layer chromatographic fractionation of the
zone at the origin of the acidic extraction shown in Figure 2,
The chromatographic solvent system was methanolrethylacetate:
benzene (1:10:89). Chromatographic fractions were tested
using TA1538 with microsome preparation present.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES 541
test of the research strategy for employing the Salmonella
test as a means of detecting and identifying environmental
carcinogens.
We have established a cooperative arrangement with the
City of Chicago Department of Environmental Control to carry
out mutagenic analyses on the high volume air particulate
samples that they collect daily at 25 stations in that city.
Samples are provided for us, together with data on the weight
of the collected particulates and associated meteorological
information.
As a preliminary step, analyses were made of benzene:
hexane (1:1) extracts of two square-inch samples of filters
collected concurrently from a. series of different stations
in the City of Chicago air pollution system. The results,
which are shown in Figure 4, revealed a general proportion-
ality between particulate concentration and the numbers of
revertant colonies, and identified the Washington School in
South Chicago as a site considerably more active than the rest,
In an effort to improve the efficiency of extraction, it was
then found that extracts obtained with benzene:hexane:iso-
propanol (70:10:20) yielded somewhat higher revertant colony
counts than benzene:hexane extracts, and the former solvent
was used thereafter. On the basis of these results we have
concentrated our studies on the analysis of samples from
the Washington School station, using the revised extraction
system. Also, for the reasons cited earlier, in these studies
we have relied heavily on data based on dose-response curves.
Analyses of air particulate samples collected at inter-
vals during 1975 from the Washington School site have been
carried out. Dose-response curves were obtained for each
air filter with and without the presence of the microsome
preparation from each of the following: (a) the benzene:
hexane:isopropanol extract; (b) the benzene-soluble fraction
of the benzene:hexane:isopropanol extract; and (c) the water-
soluble fraction of the benzene:hexane:isopropanol extract.
Using this procedure we have determined dose-response
curves (with strain TA1538) for samples collected at the
Washington School site for 15 days during 1975. Figure 5
shows six of the 15 dose-response curves obtained from these
samples for, respectively, the benzene:hexane:isopropanol
extracts, the benzene fractions, and the water fractions.
In each case the results obtained with microsomes present
(solid line) and microsomes absent (broken line) are shown.
-------
542
BARRY COMMONER ET AL.
Mutagenic Activity of Chicago Air Particulate Samples
z
100
200
300
400
Concentration of Air Participates (#gm/M3)
Figure 4. Number of revertant colonies produced per test
plate by benzene:hexane extracts of 2 in2 aliquots of air
particulate filters from different Chicago collection sites,
tested on strain TA1538, with the microsome preparation.
The highest value is from the Washington School site.
Approximately linear dose-response curves are exemplified by
those obtained from the February 26 and July 31 benzene frac-
tions, with microsomes present. Many of the curves exhibit
slopes that decline at higher concentrations. Instances of
toxicity at higher concentrations can be seen in the July 31
sample, benzene:hexane:isopropanol extract, without micro-
somes present. This curve also illustrates the inactivating
effect of microsomes; at two of the lower sample concentra-
tions, the numbers of colonies produced when microsomes are
present are lower than those observed in their absence.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
543
BENZENE HEXANE: ISOPROR&NOL
EXTRACT
100 ISO ZOO
Equivalent Amount of Air Particulates/Ptate (mg)
Figure 5. Number of revertant colonies (less control values)
produced per plate by increasing amounts of air particulate
extracts collected on six different dates in 1975 at Washing-
ton School site. Tested on strain TA1538, with (solid line)
and without (broken line) microsome preparation. The arrows
mark the sample concentration at which experimental minus
control values = 2 x CAV (where CAV is the historic control
value). The reciprocal of the indicated value is the rela-
tive mutagenic activity of the sample.
-------
544 BARRY COMMONER ET AL.
In order to devise a procedure capable of comparing the
mutagenic activities of different samples that takes into
account the variable shapes of the dose-response curves, we
have adopted the following procedure. To begin with, we note,
on the basis of our earlier statistical comparison of the
mutagenic activities of noncarcinogens and carcinogens, that
•p_ri
there is a minimum value of £—- which determines, with the
UAv
stated reliability figure, that the material is carcinogenic.
In the present analyses, we may regard a mutagenic activity
ratio of 2.0 or greater as indicative of the presence of
active substances in the sample, with a reliability of about
98% if microsomes are absent and of about 93% if microsomes
are present. We then determine from the sample's dose-response
curve the lowest concentration of the sample at which the
E—C
7;— = 2.0. This value, which is marked by the arrow shown
CAv
in Figure 5, can be obtained from the dose-response curve by
interpolation to determine the sample concentration at which
E-C = 2.0 x C. . The value can be determined in this way
regardless of the shape of the dose-response curve (specifi-
cally whether a maximum occurs, or whether the initial slope
is different from that at higher concentrations). Finally,
the reciprocal of the sample concentration at which the muta-
genic activity ratio is 2.0 may be defined as the relative
mutagenic activity of the sample. While this procedure does
not take into account possible synergistic interactions among
separate mutagens present in the sample, it does provide, as
a first approximation, relative measures of the mutagenic
activities of samples even if they yield dose-response curves
that differ in shape.
The relative mutagenic activities computed in this way
for the benzene:hexane:isopropanol extracts and the benzene
fractions obtained from all 15 Washington School samples are
plotted, as a function of sample date, in Figure 6. (The
corresponding plot for the water fraction is not shown since
in every sample the mutagenic activity ratios are zero.) The
reported wind direction at each date is also indicated.
The data of Figure 6 support several conclusions. First,
it is evident that in the presence of microsomes the level of
activity of the benzene fraction generally parallels that of
the original extract from which it is derived, providing that
there is little or no inherently active material present.
This situation occurs in the latter half of the year. How-
ever, the activities of the benzene fraction are generally
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
545
Mutagenic Activities of Chicago Air Participates —
Washington School Site, 1975.
0.2-
O.I5H
0.1-
±: .05H
>•»
">
Benzene Soluble Fractions
V
>
Wmd
Direction
TA 1538 With Rat Liver Microsomes
O-O TA 1538 Without Microsomes
u
•£
.1
(U
Benzene: Hexane: Isooropanol Extract
Jan
Nov Dec
Figure 6. Relative mutagenic activities (computed as indi-
cated in the examples shown in Figure 5) of different Chicago
air particulate extracts collected on different dates in 1975
at the Washington School site.
-------
546
BARRY COMMONER ET AL.
about half of those exhibited by the comparable original
extracts, suggesting that active material is lost during the
fractionation procedure. Second, it is evident that in the
first half of the year several instances occur in which the
samples exhibit considerable inherent mutagenic activity, and
that at least a good part of this activity is lost when the
microsome preparation is present. This means that some of
the mutagens that are inherently active are inactivated by
the microsome preparation. This situation, often encountered
in complex, unknown samples, creates important constraints
on the interpretation of the data. This can be seen from
the following considerations.
The basic difficulty is that measurements are made under
two different conditions relative to the microsome preparation
(i.e., with the preparation either present or absent), while
the sample may contain three different classes of mutagens
relative to their response to the microsome preparation.
Thus:
where:
Rwo ~ a + b
and R ~ a + c
W
R
R,
wo
w
a
revertant rate without microsome preparation.
revertant rate with microsome preparation.
the concentration of compounds which are
inherently mutagenic and not inactivated by
the microsome preparation.
the concentration of compounds which are
inherently mutagenic but inactivated by the
microsome preparation.
the concentration of compounds which are not
inherently mutagenic but are activatable by
the microsome preparation.
It is evident from these relationships that it is impossible,
from only the two measurements of revertant rate (i.e., R.
and R ) to determine the concentration of any one of the
classes of mutagens, except in the special case in which R
is zero, nearly zero, or at least very much smaller than R
b =
c -
wo
wo
w*
In the example shown in Figure 6, except for several
scattered points, the latter condition occurs only in August-
December, so in that period the measurements made with the
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES 547
microsome preparation present (i.e., R ) are indicative of
the concentration of activable mutagens. Most of the other
measurements made with the microsome preparation cannot be
interpreted quantitatively since there is no way of knowing
what part of the value is due to inherently active mutagens
of class (a), which also contribute to the value of R . On
W
the other hand, the values obtained in the absence of the
microsome preparation (Rwo) are interpreted as representative
of the activity of both classes of inherently active mutagens
(i.e., classes a and b).
Finally, it is evident that the inherently mutagenic
substances which are inactivated by the microsome preparation
(i.e., class b) are largely lost when the benzene fraction is
prepared. It is possible that this material passes into the
water fraction in the second step of the procedure, since
several samples (e.g., March 22 and July 31) exhibit a consis-
tently rising trend with sample concentration, even though at
the highest concentrations the value of E-C does not reach
the statistical criterion of 2.0 x C. . This suggests that
water-soluble active material is in fact present, which
would become statistically significant if larger samples
were analyzed.
Although it is premature to relate these observations
to the general data regarding meteorological conditions, it
is perhaps worth noting that most of the high concentrations
of inherently mutagenic material observed in the original
extract occurred when winds were generally from the northeast
quadrant (see Figure 7).
The foregoing observations are indicative of the expec-
ted complexity of the mutagenic materials that occur in asso-
ciation with urban airborne particulates. We have further
analyzed a particularly active sample, that for December 17,
in order to test the feasibility of using the Salmonella
technique as a means of isolating and identifying the respon-
sible substances. About 56 square inches of the air filter
was extracted in benzene:hexane:isopropanol. The extract was
dried, taken up in chloroform, and aliquots were subjected
to thin-layer chromatography according to the procedures de-
scribed earlier. The extracts of successive chromatographic
zones were then tested on strain TA1538 in the presence of
microsomes. When the original extract was fractionated in a
benzene:hexane (1:1) solvent system, two mutagenically active
components with RF values of 0 and 0.9 were detected. The
zones at RF = 0.9 and 1.0 were then combined, extracted,
dried, and rechromatographed using n-hexane as the solvent
-------
548
BARRY COMMONER ET AL.
N
360°
Benzene:Hexane: Isoproponol Extract
Without Microsomes
W 270°
90° E
Figure 7. Relative mutagenic activities of benzene:hexane;
isopropanol extracts of air particulate samples collected
from Washington School site on different dates in 1975, as
a function of concurrent wind direction (data of Fig. 6).
system. This procedure yielded a major mutagenically active
zone with an RF value of 0.8 and a minor one at the origin.
When the former was further chromatographed using iso-octane
as the solvent, as shown in Figure 8, a single mutagenically
active zone with an RF value of 0.7 was obtained. Under
ultraviolet light this zone exhibited a strong fluorescence
typical of certain polycyclic hydrocarbons. When prepara-
tions of pure benzo(a)pyrene and benzo(e)pyrene were chroma-
tographed in the iso-octane solvent system, they yielded the
same RF value as the mutagenically active component, 0.7.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
549"
.
_O
O
O
rr
•6
TLC Fractionation of Chicago Air Particulate
Extract.
90-
80-
5 70-
Q_
\
en
60
50
40
30
20
10
Solvent System: Isooctane
TA1538; Rat Liver Microsomes
-I 0! 23456789 10
Chromatographic Zone
(cm. from origin)
Figure 8. Final TLC fractionation of mutagenic activity of
material from Chicago air particulate sample (Washington
School: December 7, 1975). See text for fractionation steps,
-------
550 BARRY COMMONER ET AL.
The purified preparation obtained in this way was analyzed by
means of mass spectrometry together with a standard sample of
benzo(a) pyrene (both isomers yield identifcal spectra in such
an analysis). As shown in Figure 9 the spectrum of the active
component exhibits the strong mass peak at 252 which corre-
sponds to the mass of both the (a) and (e) isomers of benzo-
pyrene, as well as the fragmentation peaks which according
to a standard atlas are characteristic of this substance.
The presence of additional peaks, for example, at 266 and
270, 238 and 248, suggest that a small amount of some other
compound is present as well.
These results indicate that the active material isolated
by successive thin-layer chromatograms is largely a mixture of
benzo(a)pyrene and benzo(e)pyrene. Both isomers are mutagenic
toward strain TA1538 in the presence of the standard microsome
preparation (5). Consequently, benzo(a)pyrene and benzo(a)-
pyrene can be identified as two of the substances responsible
for the mutagenic activity exhibited by the original extract
of the air particulate sample.
All of the foregoing data are based on conventional high
volume samples in which particulates that vary widely in size
are trapped. Because of the tendency of small particles to
be retained in the lungs, it is of interest to determine the
distribution of mutagens in various sized urban air particu-
lates. Some preliminary results on this problem derived from
experiments conducted in Los Angeles (provided to us by Dr.
David Coffin of EPA) are shown in Figures 10 and 11. Figure
10 shows that there is an increase in mutagenic activity both
in the presence and absence of the microsome preparation with
decreasing particle size. Figure 11, which is a chromato-
graphic analysis of a sample of the smallest sized particles,
illustrates once again the value of such fractionation proce-
dures. It shows, for example, that one of the constituents
(RF = 0.1) is inherently active and not inactivated by the
microsome preparation, while one or more activatable constit-
uents is localized near the solvent front.
MUTAGENS IN FOODS
The entry of our laboratory into this area of research
illustrates one of the "black box" aspects of the Ames test,
and emphasizes the importance of paying attention to anomalies
that may arise. One such anomaly has been recognized in cer-
tain of the controls used in the test. Among the controls are
determinations of the number of revertant colonies that occur
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
551
100.0—
80.0—
40.0 —
40,0 —
20,0—
0,0—
100.0—
80,0—
60.0-
40.0—
20,0—
0.0—
Active Component from
Chicago Air Particulate
Benzo (a) pyrene major peaks
Benzo(a)pyrene Standard
150 200
m/e
250 300
Figure 9. Mass spectra of material from zone (RF = 0.7) of
chromatogram shown in Figure 8, and of benzo(a)pyrene.
-------
552
BARRY COMMONER ET AL.
O)
"c
jg
o
O
"c
a
cr
05
JQ
13
250
200-
150-
100-
50-
^ o
250-
200-
150-
100-
50-
300
250
200
150
100
50
Sample No 88A RTF 14
Sample Type. Upwind 3.5-20nM
Sample No.- 89A RTF 15
Sample Type. Upwind 1.7-3
,o
Sample No. 95A RTF 21
Sample Type: Downwind 3.5-2QMM
Sample No 96A RTP 22
Sample Type. Downwind I.
Sample No 90A RTP 16
Sample Type Upwind
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
553
0)
CO
0)
'c
_g
o
O
160-
140-
120-
100-
> 80H
LT
^ 60-|
i_
a;
| 40-
~Z.
20-
0
Sample No.: 98A
RTF 24
Sample Type: Downwind
1.7/JM
TA 1538
• With Microsomes
O Without Microsomes
| 300
250-
I 150-
tr 100-1
50-
01 5 10
Equivalent Amount of
Air Particulates/Plate (mg)
10123456789 10 I
Chromatographic Zone
(cm. from origin)
Figure 11. Chromatographic fractionation and dose-response
curve for extract of air particulate sample 98A RTF 24.
Particles were in the size range <1.7 UM. Chromatographic
solvent was benzene:hexane (1:1). Samples were incubated
with (solid lines) and without (broken lines) microsome
preparation.
-------
554 BARRY COMMONER ET AL.
on plates that contain only the bacterial inoculum and on
plates that contain the microsome preparation in addition to
the inoculum. In the course of an extended series of tests
of a number of organic compounds, we noticed a small differ-
ence between the mutation rates observed in these two controls
(3). For example, average values for 200 test plates were 13
revertant colonies per plate when only the bacterial inoculum
was present and 22 colonies per plate when a rat-liver micro-
some preparation was also present,. It also appeared that the
effect occurred preferentially with a particular strain of
Salmonella, TA1538, which according to Ames is sensitive to
substances that cause frameshift mutations. This effect,
which has been observed in other laboratories as well (1,6),
has remained unexplained.
As it occurs in the standard Ames test, the effect is so
small as to have no influence on the reliability of the test,
since active substances usually produce hundreds of mutant
colonies per plate. However, during the course of experiments
with a modified form of the Salmonella test, also based on the
Ames strains, we found that the effect could be considerably
amplified. These modified tests were conducted by incubating
Salmonella in an aerobic 5 ml culture containing nutrient
broth (Difco Laboratories), the microsome preparation, and the
substance to be tested. After various periods of incubation,
0.1 ml aliquots of the culture were removed and inoculated on
plates containing nutrient agar completely free of histidine.
The numbers of colonies that developed on these plates after
a 48-hour incubation period were indicative of the concentra-
tion of revertant cells present in the culture after various
periods of incubation.
Figure 12 describes typical data obtained from such a
liquid-culture test system when bacteria of strain TA1538
were present alone, when microsomes were present as well, and
when a typical carcinogen, activated by a microsome, 2-acetyl-
aminofluorene (AAF), was also present. It is evident that the
presence of microsomes (in the absence of AAF) increases the
number of revertant cells produced in the culture by an order
of magnitude. Similar experiments carried out with a series
of Salmonella strains, using several different types of micro-
some preparations, showed that the effect occurs only in
strains TA1538 and TA98 (which is similar to TA1538 in its
response to different mutagens). Thus, the phenomenon origi-
nally observed in standard plate tests, i.e., the specific
enhancement of the rate of mutation of strain TA1538, also
occurs in the liquid-culture system, but the effect is much
larger and therefore more capable of analysis.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
555
s
o
E
O
<0
*
04 8 2 16 20 24
Time of Incubation (hourt)
0 4 8 12 16 20 24
Time of Incubation (hours)
0 4 8 .2 >6 20 24
Time of Incubation (hours)
Figure 12. Number of revertant colonies and total number of
colonies produced from 0.1 ml inocula of a culture of strain
TA1538 obtained after increasing periods of incubation of:
bacteria alone (closed circles); bacteria with the microsome
preparation (open circles); bacteria with microsome prepara-
tion and 100 ygm of 2-acetylaminofluorene (triangles).
a: Numbers of revertant colonies, as obtained from counts
of culture aliquots inoculated on histidine-free synthetic
medium plates, b: Numbers of total colonies, as obtained
from counts of culture aliquots inoculated on plates of syn-
thetic medium supplemented with histidine. c: Ratio of
revertant to total number of colonies, computed from the
data of a and b.
-------
556 BARRY COMMONER ET AL.
As a first step in such an analysis, we undertook to
determine the functional basis for the apparent rnutagenic
effect of microsome preparations on strain TA1538. These
studies showed that material which is mutagenic toward strain
TA1538 in the presence of microsomes can be extracted by
benzene:isopropanol (80:20) and similar solvents from "Bacto
nutrient broth" (Difco Laboratories), whether fresh or follow-
ing incubation in a bacterial culture. It can be concluded,
therefore, that the effect represents the conversion of a
substance present in nutrient broth into an active mutagenic
metabolite by the enzymatic activity of microsomes. This is
confirmed by the data of Figure 13, which shows, from dose-
response curves, that such extracts of two samples of commer-
cial nutrient broth contain comparable amounts of microsome-
activatable mutagenic material, to which strain TA1538 readily
responds.
In a survey of a number of commercial bacterial nutrients,
we found that those nutrients which contain "beef extract" or
beef heart infusion contain active material, yielding from 308
to 2789 revertant colonies per gram in the presence of micro-
somes as compared with 10-36 colonies when microsomes are
absent. Comparison of the several Difco nutrients tested sug-
gests that the number of revertant colonies produced per gram
is roughly proportional to the nutrient's content of beef
extract. It appeared from these results that the mutagen is
a constituent of the beef tissue (generally muscle) used to
produce the beef extract employed in these nutrient prepara-
tions, or is derived from such a constituent during the pre-
paration process.
Beef extract used in bacterial nutrients is produced in
abbatoirs, by first preparing beef broth from beef tissue
which has been boiled for about 30 minutes in an equal volume
of water and then defatted. To prepare beef extract this
broth is then boiled down to 20 percent or less of its origi-
nal volume. The result is a dark brown paste which is used
in the manufacture of bacterial nutrients and in various
foods, such as beef bouillon cubes. "Bacto Beef Extract"
(Difco Laboratories) was tested for mutagenic activity in the
following way: samples were homogenized in distilled water
and then acidified (to pH 2.0) with HC1. Protein was then
precipitated by adding ammonium sulfate to saturation. The
samples were then filtered through glass wool, the filtrate
adjusted to pH 10 with ammonium hydroxide, extracted three
times with methylene chloride and the extract evaporated to
dryness. Aliquots representing varying amounts of the origi-
nal sample were taken up in DMSO and tested on strain TA1538
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
557
5000'
1000-
a*
•§ 500
o
5
w
0)
(T
"o 100
h_
.a
I 50-1
10
With Without
Sample: Microsomes Microsomes
Difco Nutrient Broth • O
BBL Nutrient Broth • H
n—»
0 0.2 0.4 0.6 0.8 1.0
Equivalent Amount of Sample (gm)/Plate
Figure 13. Number of revertant colonies (of strain TA1538)
produced per standard test plate by benzene:isopropanol
(80:20) extracts of increasing quantities of Difco (circles)
and BBL (squares) nutrient broth. Solid lines: microsomes
present; broken lines: microsomes absent.
with and without microsomes. The dose-response obtained
is clearly indicative of mutagenic activity in the presence
of the microsome preparation (see Figure 14). Dose-response
curves obtained with other strains show that strain TA98 is
equally active, TA1537 about one-fourth as active, while
strains TA100 and TA1535 are inactive. In all cases there
was no activity when microsomes were absent. A series of
-------
558
BARRY COMMONER ET AL.
10,000
,P O
x *«»^
X '"" »
0
0.2
0.3
0.4
0.5
Equivalent Amount of Sample
(grams dry weight/plate)
Figure 14. Dose-response curves of methylene chloride
extracts of "Bacto Beef Extract" (Difco Laboratories) tested
on strain TA1538. Ordinate: number of histidine-positive
revertants per plate. Abscissa: amount of sample used to
prepare the methylene chloride extract added per plate.
Solid lines represent plates to which microsome preparation
was added. Broken lines represent plates to which the micro-
some preparation was not added.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES 559
chromatographic analyses of Difco beef extract and of Difco
nutrient broth were carried out with hexane-acetone and with
benzene-methanol as solvents. As shown in Figure 15, the
chromatrographic mobility of the mutagenically active material
from beef extract and from nutrient broth, in benzene:methanol
(95:5) was similar. Comparable results occurred in the other
solvent system. Thus, the mutagen originally discovered in
bacterial nutrient is present in the beef extract itself.
Two commercial preparations, purchased in local stores,
"Maggi Beef Bouillon Cubes" and "B.V. Broth & Sauce Concen-
trate," which according to their labels contain beef extract,
have been tested with methods comparable to those described
for beef extract. From the dose-response curves against
various Salmonella strains, in the presence of a microsome
preparation and from chromatographic analysis 'see Figure 15),
it is evident that these preparations contain mutagens with
the characteristics of those found in bacterial nutrients and
in beef extract.
Beef broth contains no detectable mutagens whereas beef
extract, which is prepared from the broth by extensive boil-
ing does. Accordingly, we have studied the conversion pro-
cess by testing beef broth for mutagenicity at 30-minute
intervals during extensive boiling. The results, which are
reported in Figure 16, show that the mutagens are absent from
beef stock and are produced during the boiling process, espe-
cially when the preparation is reduced to a paste, at which
time the mutagenic activity rises sharply to 1572 revertants
per plate per 0.69 gm dry weight. It is apparent, then, that
the mutagens do not occur as such in beef tissue or in beef
broth, but are formed during the heating and evaporation that
occurs in the conversion of beef broth to beef extract.
Given these results, it was of obvious interest to deter-
mine whether these mutagens are formed when beef is cooked by
conventional procedures. Lean ground beef (in 100 gm, dry
weight, portions) was cooked in an electrically-heated (plate
temperature 200°C) home hamburger cooking appliance for 1.5
minutes ("rare"), 3.0 minutes ("medium"), and 5.5 minutes
("well-done"), respectively. The cooked samples and an
uncooked control were homogenized in twice their volume of
distilled water in a Waring blender and were treated in the
same way as the beef extract described earlier. Aliquots
of the final methylene chloride extracts representing 5 and
25 gm dry weight of the cooked beef (in the case of the
uncooked control, aliquots represents 5, 10, and 35 gm were
tested) were dried, taken up in DMSO, and tested in the usual
way against strain TA1538 in the presence and absence of the
-------
560
BARRY COMMONER ET AL.
10,000'
to
• 1000-
^
o
o
"c
a
k.
o>
0>
tr
100-
0)
JD
3
10
Strain TAI538-, With Microsome Preparation
Sample:
• Oifco Beef Extract
A Difco Nutrient Broth
• "B-V Concentrate"
O "Maggi" Bouillon Cubes
Solvent:
Benzene : Methanol (95:5}
0123456789 10
Chromatographic Zone
(cm. from origin)
Figure 15. Thin layer Chromatographic fractionation in
benzenermethanol (95:5) of the mutagenic material of Difco
beef extract, Difco nutrient broth, "B-V concentrate," and
"Maggi" bouillon cubes. Tested on strain TA1538 with micro-
some preparation present.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
561
O
O
i.o-
0 9-
0.8-
1575
• No. of Revertant Colonies
A % Water Content
• Optical Density
Sample: Beef Stock
34 5678
Boiling Time (hours)
10
-0
•10
-20
-------
562
BARRY COMMONER ET AL.
5000
0
Equivalent Amount of Sample
(grams dry weight/plate)
Figure 17. Dose-response curves for raethylene chloride
extracts of uncooked and cooked lean ground beef. Tests were
carried out on strain TA1538 with the microsome preparation
present (solid lines) and in its absence (broken lines). 100
gm (wet weight) samples of lean ground beef were tested before
cooking (data points indicated by crosses) and after cooking
in an electrically-heated home hamburger cooking appliance
for the following times: 1.5 minutes ("rare"; data points
indicated by triangles), 3.0 minutes ("medium"; indicated by
squares) and 5.5 minutes ("well-done"; indicated by circles).
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES 563
These data suggested a possible relation between our
observations and earlier evidence that mutagens, including
known carcinogens such as benzo(a)pyrene, are formed in meat
and fish during certain cooking procedures. Thus, Sugimura
et al. (7) report that condensed smoke from meat and fish
broiled over an open gas or charcoal flame contains material
that is mutagenic toward strain TA98, usually only in the
presence of the microsome preparation. They report that the
mutagenic activity levels are much too high to be accounted
for by the amounts of benzo(a)pyrene present in the smoke
condensates and suggest that other mutagens may arise from
pyrolysis of tissue protein and amino acids. This suggestion
is based on their observation that pyrolysis (at temperatures
of 300°-600°C) of proteins and certain amino acids produces
mutagens similar in their effects in the Ames test to those
observed in the smoke condensates (4). They also report
similar activity in material obtained from the charred sur-
face of a broiled beef steak.
In view of the foregoing results, it was of interest to
compare the mutagens that occur in beef extract and cooked
beef with those formed by pyrolysis of amino acids, and with
benzo(a)pyrene. For this purpose methylene chloride extracts
of beef extract, cooked beef, cooked beef with added benzo(a)-
pyrene, and a pyrolyzed mixture of amino acids were chromato-
graphed, using a silica-gel impregnated glass fiber sheet
(Gelman ITLC-SG) in a suitable solvent. Successive 1 cm
zones of the developed chromatograms were extracted in chloro-
form:methanol (90:10), dried, taken up in DMSO, and tested on
strain TA1538 in the usual way. Figure 18 reports such anal-
yses of methylene chloride extracts of "Bacto Beef Extract"
and of a beef patty cooked for ten minutes on a ceramic hot
plate, using benzenermethanol (95:5) as the chromatographic
solvent. From thermocouples at the surface of a patty and
in its interior, it was determined that the maximum tempera-
ture (at the end of the cooking period) at the surface of the
patty was 200°C and in the interior 80°C. The mutagens present
in the two samples exhibit identical chromatographic behavior,
with a major peak at an RF = 0.5 and a slight shoulder at
RF = 0.3. Figure 19(a) reports the results of a similar
analysis (using 100 percent hexane as the chromatographic
solvent) of methylene chloride extracts of "Bacto Beef Extract,"
of a hot-plate cooked beef patty, and of such a patty to which
25 ugm per kgm (wet weight) of benzo(a)pyrene had been added
(after cooking and extraction). All of the mutagenic activity
associated with "Bacto Beef Extract," and cooked beef remains
at the origin, while the sample of the latter in which benzo(a)-
pyrene had been added exhibits an additional peak at RF = 0.85.
-------
564
BARRY COMMONER ET AL.
800
® 700
_O
Q_
"w 600
«
O
-Q 500-
O
•*—
0 400
300H
- 200
100-
0
01 23456789 10
Chromatographic Zone
(cm. from origin)
Figure 18. Thin-layer Chromatographic fractionation of the
methylene chloride extracts of "Bacto Beef Extract" (•—•)
and hot plate-cooked lean ground beef ( H— •). Gelman ITLC-
SG sheets were used with benzene:methanol (95:5) as the sol-
vent system. Four ground beef patties (each approximately
120 gm wet weight) were wrapped in aluminum foil and cooked
on a 350°C ceramic hot plate for 10 to 12 minutes. Thermo-
couples at the surface of a patty and in its interior recorded
temperatures of 200°C and 80°C, respectively, at the end of
the cooking period. Extracts equivalent to approximately 0.2
grams of beef extract and 26 grams (dry weight) of ground
beef were applied to the chromatogram. One-centimeter zones
of the developed chromatogram were extracted with chloroform:
methanol (90:10). Aliquots were taken to dryness, resuspended
in DMSO and tested on strain TA1538 in the presence of the
microsome preparation.
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES
565
soooH
2
a. i
^ 1000-1
01
'H !
| 500,
CJ
>
v
rr 1001
a>
.a
3
1000
750
at
"I
o
a
1
500-
250-j
10-
0 I 234 56 78 9 10 II
Chromatographic Zone
(cm from origin)
0 I 2345 6789 10
Chromatoqraphic Zone
(cm from origin)
Figure 19. Thin-layer Chromatographic fraction of: (a) the
methylene chloride extracts of "Bacto Beef Extract" (•—•),
hot plate-cooked beef (•—•) and hot plate-cooked beef to
which benzo(a)pyrene was added (25 ug/Kg wet weight) (A—A).
Extracts equivalent to approximately 0.2 grams of beef extract
and 26 grams of ground beef (with and without benzo(a)pyrene)
were applied to the chromatogram. Chromatographic solvent:
100% hexane; and (b) the methylene chloride extracts of "Bac-
to Beef Extract" (•—•) and of a mixture of 18 amino acids
pyrolyzed at 350°C (*—«). Extracts equivalent to approxi-
mately 0.2 grams of beef extract and 9 mg of amino acids
(equal weights of each) were applied to the chromatogram.
Chromatographic solvent: hexane:acetone (50:50). Microsome
preparation was present.
-------
566 BARRY COMMONER ET AL.
Figure 19(b) reports a similar chromatographic analysis
[using hexane:acetone (50:50) as the chromatographic solvent]
of methylene chloride extracts of "Difco Beef Extract" and
the pyrolsis product (pyrolysis temperature 350°C) of a mix-
ture of 2 mg of each of the 18 amino acids, which according
to Matsumoto et al., (4) yield mutagenic material when
pyrolyzed. The material from "Bacto Beef Extract" exhibits
a peak at RF = 0.4, while the material from the pyrolyzed
amino acids exhibits a main peak at RF = 0.8, a minor peak
at RF = 0.6, and some residxial activity at the origin.
These analyses indicate (a) that the mutagens produced
when beef stock is heated to form beef extract are chromato-
graphically indistinguishable from those produced when ground
beef is cooked on an electrically-heated hot-plate and (b)
that the former are chromatographically distinguishable from
both benzo(a)pyrene and the mutagens produced from pyrolyzed
amino acids. Further studies of the mutagenic material
extractable by methylene chloride from "Bacto Beef Extract,"
partially purified by successive thin-layer chromatographic
separations, show the following:
• The mutagen(s) is a basic substance, extractable
by organic solvents from aqueous solutions at
alkaline pH.
• It is unaffected in its mutagenic activity or
chromatographic behavior by refluxing in 6N HC1
for six hours.
• On treatment with nitrous acid, the material becomes
inherently mutagenic (i.e., in the absence of the
micromsome preparation), suggesting the possible
formation of a nitroso group. The conditions in
which these mutagens are formed are similar to those
characteristics of the Maillard or "Browning" reac-
tions in which amino acids and sugars react to pro-
duce a variety of complex substances (8).
The foregoing experiments show that one or more substances
which are mutagenic in the Ames system (in the presence of the
microsomal preparation) are produced when beef stock is heated
and condensed to form beef extract and when ground beef is
cooked (at temperatures not exceeding 200°C) on an electric
hot-plate or a home hamburger cooking appliance. These muta-
gens are neither benzo(a)pyrene nor the mutagenic substances
produced when amino acids are pyrolyzed. This is indicated
by the chromatographic analyses reported above. Moreover,
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES 567
according to Matsumoto et al. (4), the mutagenic pyrolysis
products are formed only at temperatures in excess of 300°C,
which can readily occur in foods cooked over open flames.
In contrast, the mutagens we have detected in beef extract
are produced at temperatures that do not exceed 105°C, while
those detected in cooked ground beef are produced at tempera-
tures that do not exceed 200°C. Thus, these mutagens are
produced in conditions that occur in common cooking procedures,
including the preparation of hamburgers on electrically-heated
hot-plates at conventional cooking temperatures and times.
The mutagens found in beef extract and cooked beef are
rather active, as compared with a typical mutagen which is
also active toward strains TA1538 and TA98, 2-acetylamino-
fluorene (AAF). Tested on strain TA1538, 50 ugm of AAF
(which is in the linear portion of the dose-response curve)
yields about 4800 revertants per plate. Active material
prepared from a bacterial nutrient containing 37 percent
beef extract yielded 1367 revertants per plate containing
3.5 ugm (in the linear part of the dose-response curve) of a
preparation partially purified, by successive chromatographic
fractionation, from the original methylene chloride extract.
Accordingly, the specific activity of the beef extract muta-
gen(s) is a minimum of about 350 revertants per plate per
ugm, as compared with 96 revertants per plaice per ugm for
AAF. Based on the estimate of 350 revertants per plate per
ugm, a 3.6 gm beef bouillon cube contains a minimum of approx-
imately 0.3 ugm of mutagen and a 100 gm wet weight lean-beef
hamburger contains approximately 1 to 14 ugm of mutagen,
depending on the extent of cooking. These figures correspond
to concentrations, on a wet weight basis, of 0.1 ppm of rnuta-
gen in beef bouillon cubes and from .01 to .14 ppm in cooked
hamburgers.
If, as indicated by the observed correlation between
rnutagenicity in the Ames test and carcinogenicity, these muta-
gens—once purified and tested on laboratory animals—are
found to be carcinogens, their apparent concentration in some
foods may represent an appreciable risk to certain populations.
The relatively ordinary circumstances in which these mutagens
are formed suggest that they may arise during the course of
certain conventional cooking procedures, in addition to the
preparation of hamburgers, such as the braising of beef and
the evaporation of beef stock in the preparation of stews.
However, the sensitivity of the effect to cooking times, which
is evident in the results shown in Figure 17, suggests that it
may be possible to modify cooking procedures in ways that
reduce the formation of the mutagens.
-------
568 BARRY COMMONER ET AL.
DISCUSSION
The substantive conclusion of the foregoing results is
that mutagens occur in the effluents of certain petrochemical
plants, in Chicago air particulates, in beef extracts, and in
hamburgers. Clearly, the Ames test, is a very useful means of
detecting the occurrence of such environmental carcinogens.
It is also evident from these results that, combined with
chromatographic techniques, the method can be used to isolate
and ultimately identify mutagens which occur in such samples.
However, such qualitative conclusions—for example, the deter-
mination of whether or not a given environmental sample con-
tains a significant amount of mutagenic material—depend on
certain quantitative procedures. Specifically, the appro-
priate procedure is to determine, from a dose-response curve,
whether at any sample concentration the mutagenic activity
E—C
ratio, p— exceeds the statistical criterion previously
^Av'
established from test of standard substances. Such deter-
minations must be made separately with microsomes present
and absent. Constraints on this type of determination in-
clude the following:
• The determination relates only to substances that
are active on the particular strain of Samonella
that is used.
• A false negative result may be obtained if the
sample contains sufficient toxic or bacteriostatic
material to suppress the growth of mutants.
Subject to these constraints and to the previously stated
limits of the reliability of the test system, the Salmonella
technique can readily be used for the rapid, qualitative detec-
tion of organic carcinogens in environmental samples.
It is also evident that, subject to additional constraints,
quantitative estimation of the level of mutagenic activity is
possible, based on the analytical procedures described above.
In these procedures one determines by interpolation from the
dose-response curve the lowest sample concentration at which
the mutagenic activity ratio that is representative of statis-
tically significant mutagenic activity occurs. The sample's
mutagenic activity is expressed, in relative terms, by the
reciprocal of this sample concentration. A major constraint
on this procedure is that it is not applicable to data obtained
in the presence of microsomes, unless the sample's mutagenic
-------
MUTAGENIC ANALYSIS OF COMPLEX SAMPLES 569
activity in the absence of microsomes can be shown to be
zero, or small relative to the value obtained when microsomes
are present. Where an initial extract of the sample does
not conform to this requirement, it would be necessary to
introduce a fractionation procedure that separates inherently
active mutagens from those requiring microsomal activation
before quantitative estimates of the latter are made.
While the emphasis of this paper is on the methodologi-
cal aspects of these results, certain substantive aspects of
the results are worth noting. The results of studies of air
particulates from the Washington School site are probably re-
lated to the fact that this site, which appears to yield the
highest concentrations of carcinogens in air particulates
from the Chicago area, is located within a heavily industrial-
ized neighborhood. Steel mills, including coke-oven opera-
tions, are present. Since these operations are known to pro-
duce high concentrations of benzo(a)pyrene and other carcino-
gens, the high levels of mutagenic activity that we have ob-
served in air particulates, and direct evidence that benzo-
pyrene isomers occur in them, is not suprising. While the
data obtained from this site are insufficient to establish
firm correlations with wind direction, they do suggest that
with more detailed analyses it will be possible to define the
origins of the particulate-associated carcinogens. It would
appear, therefore, that screening procedures based on the
Salmonella mutagenesis technique can be used to determine how
the environmental distribution of the detectable carcinogens
may be associated with the local epidemiology of cancer inci-
dence, and with the activities of possible sources of the
relevant substances.
In the same way, the studies of the formation of muta-
gens in cooked beef and in beef extract, together with earlier
studies in Japanese laboratories, show that the technique can
be a very useful means of monitoring the role of cooking
practices on the formation of mutagens.
The Ames technique, suitably applied and subject to cer-
tain constraints, is a valuable means of screening environ-
mental samples for mutagens. Given the established correla-
tion between mutagenicity in this test and carcinogenicity
toward laboratory animals, those procedures form the basis
for an analysis of the role of environmental agents in the
incidence of cancer.
-------
570 BARRY COMMONER ET AL.
REFERENCES
1. Ames BN, Burton WE, Yamasaki E, Lee FD: Carcinogens
are mutagens: A simple test system combining liver
homogenates for activation and bacteria for detection.
Proc Natl Acad Sci USA 70:2281-2285, 1973
2. Ames BN, McCann J, Yamasaki E: Methods for detecting
carcinogens and mutagens with the Salmonella/mammalian-
microsome mutagenicity test. Mutat Res 31:347-364, 1975
3. Commoner B: Reliability of bacterial mutagenesis
techniques to distinguish carcinogenic and noncarcin-
ogenic chemicals. Washington DC, US Environmental
Protection Agency Publ. No. EPA-600/1-76-022, p 104,
1976
4. Matsumoto T, Yoshida D, Mizusaki S, Okamoto H: Muta-
genic activity of amino acid pyrolyzates in Salmonella
typhimurium TA98. Mutat Res 48:279-286, 1977
5. McCann J, Choi E, Yamasaki E, Ames BN: Detection of
carcinogens as mutagens in the Salmonella/microsome
test: Assay of 300 chemicals. Proc Natl Acad Sci USA
72:5135-5139, 1975
6. Nebert DW, Feton JS: Evidence for the activation of
3-methylcholanthrene as a carcinogen in vivo and as a
mutagen in vitro by P!-450 from inbred strains of mice.
In: Cytochromes P-450 and bs, Structure, Function and
Interaction (Cooper DW, Rosenthall 0, Snyder R, Witmer
C, eds.). New York, Plenum, pp 127-149, 1975
7. Sugimura T, Nagao M, Kawachi T, Honda M, Yahagi T,
Seino Y, Sato S, Matsukura N, Matsushima T, Shirai A,
Sawamura M, Matsumoto H: Mutagen-carcinogens in food
with special reference to highly mutagenic pyrolitic
products in broiled foods. In: Origins of Human
Cancer (Hiatt HH, Watson JD, Winsten JA, eds.). Book
B, Cold Spring Harbor, NY, Cold Spring Harbor Labora-
tory, pp 1561-1577, 1977
8. Tarr H: Ribose and the Maillard reaction in fish
muscle. Nature 171:344-345, 1953
-------
Poster Abstracts
-------
573
COMPARISON OF MUTAGENS: A THEORY OF RELATIVITY FOR BIOLOGY
June H. Carver, Lawrence Livermore Laboratory, University
of California, Livermore, California; F.T. Hatch
Short-term mutagenesis assays, with a battery of systems to
minimize false positives and negatives, will play a major
role in identifying potentially mutagenic agents in the
environment. Microbial tests show a high correlation between
mutagenic activity and carcinogenic potential, but do not
always compare with the observed rank order of carcinogenicity
in whole animals. In vitro tests with cultured mammalian
cells are more quantitative, but a wide range of values for
toxic and mutagenic potency have been reported with various
rodent cell systems. Our objective has been to reconcile
differences (up to two orders of magnitude) in toxicity as
measured by the D37 decrease in cell survival, as well as the
mutation rate per unit dose. We have compared forward muta-
tion frequencies for our CHO cells (azaguanine-, thioguanine-,
and ouabain-resistance) with those reported in the literature
for V79, CHO, and L5178Y cells. We have added our recent
data for azaadenine resistance (quantitatively similar to
azaguanine). When the data are expressed as the mutation
rate per locus per D37 for EMS, MNNG, 4-NQO, or UV, results
for the rodent systems compare well. Thus, the increase in
mutant frequency per unit of decrease in cell survival may
facilitate comparison of the mutagenic potency of very dif-
ferent mutagens. Fluctuations in this value for different
mutagens inducing at the same locus (or for a mutagen induc-
ing at different loci) imply that cell killing and mutation
induction do not necessarily arise from the same type of DNA
damage. The correlation between UV irradiation and the UV-
mimetic, 4NQO, for azaguanine and thioguanine resistance
suggests that the induced mutational lesions may be failures
of excision repair. The marker for ouabain resistance does
not always correlate from one system to another, making it of
less value in a battery of markers for assay of forward muta-
tion. This worked supported by the U.S. DOE Contract W-7405-
ENG-48 and by the Environmental Protection Agency.
-------
574
FUNCTIONAL CHANGES IN THE FREE-CELL POPULATION LAVAGED FROM
LUNGS OF RATS AND GUINEA PIGS DURING CHRONIC INHALATION
EXPOSURES
Finis L. Cavender, Becton, Dickinson Research Center,
Research Triangle Park, North Carolina; J. Campbell;
B.Y. Cockrell
In order to evaluate changes in the free-cell population in
the lungs of rats and guinea pigs, lavage samples were
prepared from animals after inhalation exposure for 1, 2, 3,
and 5 days as well as after 6, 12, and 24 months of expo-
sure. Particularly dynamic changes occurred in exposures to
0.25, 2.5, and 25 mg/m3 aluminum chlorhydrate (ACH). At the
time of sacrifice, tracheostomies were performed on at least
three animals per sex-species group. Phosphate buffered
saline was instilled into the lungs of each animal in situ
and was allowed to remain for 10 minutes. Five additional
instillations were withdrawn immediately. The resulting
cell suspension was centrifuged and the cells were cultured
in medium 199 containing 20% fetal bovine serum and anti-
biotics. Total cell numbers, cell viability, phagocytic
index, total ATP levels, and differential cell counts were
determined for each cell suspension. Cell numbers increased
while the phagocytic index, ATP levels, and cell viability
decreased with increasing concentrations of ACH. Although
the number of pulmonary alveolar macrophages increased,
polymorphonuclear neutrophils represented more than 60% of
cell population in rats exposed to 25 mg/m3 ACH. Eosino-
phils represented 25% of the cell population in guinea pigs
exposed to 25 mg/m3 ACH. Intracellular ACH appears as spic-
ules as determined by electron microscopy. The results
indicated that aluminum chlorhydrate caused functional
alterations in pulmonary alveolar macrophages and altered
the population dynamics of free cells in the lung.
-------
575
STIMULATION OF ADENOVIRUS TRANSFORMATION BY ENVIRONMENTAL
POLLUTANTS
Maria T. Pavlova, Brookhaven National Laboratory, Upton, New
York; Bruce C. Casto
In vitro transformation by an oncogenic simian virus SA7 was
stimulated by a variety of environmental pollutants suspected
to be potentially carcinogenic.
Hamster embryo cells (HEC) were exposed to various doses of
chemicals for 18 hours before virus inoculation. Enhance-
ment of viral transformation was calculated by comparing the
transformation frequency of treated cultures with that
obtained in untreated cultures. Treatment of HEC with 7,12-
dimethylbenz(a)anthracene (DMBA), dibenz(a,h)anthracene
(DBA), benzo(a)pyrene (BP), or 3-methylcholanthrene (MCA)
increased the frequency of SA7 transformation whereas treat-
ment with the noncarcinogenic polycyclic hydrocarbons phen-
anthrene and pyrene was ineffective. Enhancement of viral
transformation was in the range of 4-fold with DBA, 10-fold
with MCA, 13-fold with DMBA and 22-fold with BP. Treatment
with cadmium acetate and cobaltous acetate resulted in a
significant enhancement of viral transformation of 100-fold
and 600-fold respectively. However, the noncarcinogenic
metal salts CaCl2, A1SO,, and BaCl2 did not increase the
frequency of SA7 transformation. These findings suggest
that the SA7 transformation system using HEC is a useful
screening technique for potentially carcinogenic environ-
mental pollutants.
-------
576
SISTER CHROMATID EXCHANGE ANALYSIS OF HUMAN CELLS: A SHORT-
TERM BIOASSAY SYSTEM FOR ENVIRONMENTAL MUTAGENS
Donald E. Rounds, Pasadena Foundation for Medical Research,
Pasadena, California; Robert E. Guerrero
Sister chromatid exchange (SCE) analysis of hamster cell
lines, treated with known carcinogens, has given strong
support for the concept that it can be used reliably for
identifying environmental mutagens. More recently, studies
with human diploid cell types have shown that the same tech-
niques can offer data which is thought to be more relevant
to mutagenic expression in human tissues. The human cell
system has been shown to be fast (less than 10 days/test),
sensitive to nanogram quantities of test substances, inex-
pensive (less than $1000/test), and appears to be reliable,
although the data base is still limited.
The human cell types that can be used for bioassay testing
can include either fibroblasts, -lymphocytes, or epithelial
cells in primary culture. Target cells can thus be selected
for representative studies of mutagenic events that occur in
the whole organism. Most important, the sensitivity and
flexibility of the test system can serve as a bioassay of
metabolites formed after in vivo exposures to industrial or
environmental mutagens. This approach can use either cul-
tures of blood specimens from exposed subjects or a test of
the response 6f normal diploid cell lines to metabolites in
human urine specimens. These latter test systems offer an
overall response to the complex mixtures of metabolites
resulting from in vivo activation of promutagens.
-------
577
SISTER CHROMATID EXCHANGES (SCE) AS A BIOASSAY FOR EXPOSURE
TO MUTAGENIC AGENTS
Daniel G. Stetka, University of California, Lawrence Liver-
more Laboratory, Livermore, California; Anthony V. Carrano
and Jason Minkler
Induction of SCEs assays exposure of mammalian cells to
chemical mutagens. In vitro, compounds that do not require
metabolic activation (e.g., EMS) induce SCEs when used
alone; compounds that require metabolic activation [e.g.,
cyclophosphamide (CP) and benz(a)pyrene (BP) ] also incude
SCEs if cells are simultaneously treated with an activating
system which contains rat liver microsome extract (S-9 mix).
In vivo, SCE frequencies are measured in rabbit lymphocytes
before and after exposure of the animal; in this way each
animal serves as its own control. Acute effects are assayed
following single ip injections; EMS, MMS, CP, BP, mitomycin-C
(MMC), and methylcholanthrene (MC) increase SCE frequencies
within one day, but these increases are transient. Chronic
effects are assayed following repeated injections; MMC, BP,
and MC induce elevated SCE frequencies that persist for at
least several months following the final injection.
Both the in vitro and in vivo approaches are sensitive,
reliable, rapid, and inexpensive assays for exposure of
mammalian systems to chemical mutagens. (To date, all
chemical mutagens induce SCEs, but with varying efficien-
cies) . A single worker can generate one (in vivo) or two
(in vitro) 5-point dose response curves within one week.
The in vivo system possesses two additional advantages: it
is more relevant to man because whole animals are used, and
it is able to determine effects of both acute and chronic
exposure.
This work performed under the auspices of U.S. DOE Contract
No. W-7405-ENG-48.
-------
578
THE SPERM TEST: A SHORT-TERM, I_N VIVO, MAMMALIAN BIOASSAY
FOR AGENTS HAZARDOUS TO THE MALE GERM CELLS
Andrew J. Wyrobek and B.L. Gledhill, Lawrence Livermore
Laboratory, University of California, Livermore, California
The enumeration of misshapen sperm has played a long-estab-
lished role in the diagnosis of male infertility. This
method has now been modified to provide a new approach to
monitoring health hazards. The new technique is easy and
straightforward: several weeks after a male mouse is ex-
posed to a test agent, its semen is assessed by visually
scoring for abnormal forms among the sperm population. A
physical or chemical agent that induces abnormal forms has
clearly interfered with the normal differentiation of the
germ cells. Well behaved dose-response curves for over 60
agents have already been established with this bioassay.
Three large and independent bodies of evidence show that the
induction of abnormally shaped sperm signals exposure to a
mutagen. First, murine sperm shape is highly heritable and
generally unaffected by physiological factors. Second, a
strong agreement exists between an agent's ability to induce
abnormal sperm and its mutagenicity in other bioassays.
Third, induced sperm abnormalities in the mouse have been
successfully transmitted to offspring.
These studies, together with supportive findings of induced
sperm abnormalities in hamsters, rabbits, and humans illus-
trate the generality and usefulness of the sperm abnormality
assay as a short-term, ijn vivo bioassay of agents hazardous
to the male germ cells.
Work performed under the auspices of the U.S. Department of
Energy by the Lawrence Livermore Laboratory under Contract
Number W-7405-ENG-48.
-------
579
TOXICITY OF SIMPLE AND COMPLEX ENVIRONMENTAL MIXTURES
Terence E. Cody, University of Cincinnati, Cincinnati, Ohio;
Victor J. Elia and Robert T. Christian
Toxic agents may interact with each other and with chemical
and physical components in the environment and produce
mixtures having toxic characteristics not necessarily equal
to the sum of their parts. While animal toxicological
methods are too costly and time consuming for routine moni-
toring of environmental samples, toxicity can only be
assessed using a biological system. We have used cell
culture systems for the study of the toxicity of chemicals,
mixtures of chemicals and complex natural mixtures such as
coal leachates, municipal drinking water, direct reuse
water, coal fired stationary power plants emissions and
automobile exhaust emissions. Toxicities of mixtures of
individual chemicals are usually approximately equal to the
sum of the toxicities of the individual chemicals in the
mixtures but in certain cases there is an antagonism and the
toxicity is less than expected. None of the mixtures that
we have tested were more toxic than expected on the basis of
toxicity of individual components. The severity of the
toxic effect can be determined by the extent of growth
inhibition of the cultures. Recently, we have developed
methods to determine the mutagenic and mutation-promoting
properties of environmental chemicals and mixtures. Pro-
motors, if found in environmental mixtures, will be of
particular interest.
-------
580
COMPARISON OF CHEMICAL AND BIOLOGICAL DATA IN LEVEL 1
ENVIRONMENTAL ASSESSMENT
Judith C. Harris, Arthur D. Little, Inc., Cambridge,
Massachusetts; Mildred G. Broome, Philip L. Levins, James L.
Stauffer
Environmental assessment is a complex, iterative procedure
that takes into consideration available process and control
technology and environmental objectives, as well as environ-
mental data acquisition. A candidate approach is presented
here for interpreting, integrating, and reporting Level 1
organic chemical analysis results in a way that facilitates
their use in environmental assessment and/or comparison with
other experimental results.
Level 1 is the first stage in a phased sampling and analysis
strategy developed by the Process Measurements Branch,
IERL/RTP, for comprehensive characterization of multimedia
effluent and process streams in the context of environmental
assessment. Level 1 is a survey analysis which involves
physical, inorganic chemical, and organic chemical analyses
and bioassays. Level 2 is a directed, detailed analysis
based on information generated at Level 1, while Level 3 is
a monitoring phase.
The focus of this presentation is a strategy for interpret-
ing the raw data generated in Level 1 organic analysis,
which include quantitative estimates plus qualitative infor-
mation from liquid chromatography, infrared spectra and mass
spectra. The strategy allows conversion of raw data to
estimated mass emissions by organic compound class (e.g.,
heterocyclic sulfur compounds, 21 mg/cu m). In this smoothed
form, the Level 1 organic analysis data can be compared with
bioassay results and with Level 1/2 transition decision
criteria. Data on organic analysis of a typical Level 1
sample are presented and compared with the results of two
Level 1 bioassay procedures (Ames and RAM tests) . The
results are discussed in terms of the comparability of the
two types of data and their implications for a Level I/Level
2 transition decision. Some Level 2 chemical analysis data
are included for comparison.
-------
581
MUTAGENICITY OF CARCINOGENS: STUDY OF 101 INDIVIDUAL AGENTS
AND 3 SUBFRACTIONS OF A CRUDE SYNTHETIC OIL IN A QUANTITATIVE
MAMMALIN CELL GENE MUTATION SYSTEM
Abraham W. Hsie, Biology Division, Oak Ridge National Labo-
ratory and the University of Tennessee-Oak Ridge Graduate
School of Biomedical Sciences, Oak Ridge, Tennessee; J.
Patrick O'Neill, Juan R. San Sebastian, David B. Couch,
Patricia A. Brimer, William N.C. Sun, James C. Fuscoe, Nancy
L. Forbes, Richard Machanoff, James C. Riddle and Mayphoon
H. Hsie
Conditions necessary for quantifying mutation induction to
6-thioguanine resistance, which selects for >98% mutants
deficient in the activity of hypoxanthine-guanine phosphori-
bosyl transferase (HGPRT) in a near-diploid Chinese hamster
ovary (CHO) cell line, referred to as CHO/HGPRT system, have
been defined. Employing this mutation assay, we have deter-
mined the mutagenicity of diversified agents including 11
direct-acting alkylating agents, 10 nitrosamines, 10 hetero-
cyclic nitrogen mustards, 15 metallic compounds, 5 quino-
lines, 5 aromatic amines, 27 polycyclic hydrocarbons, 11
miscellaneous chemicals, 7 ionizing and non-ionizing physical
agents. The direct-acting carcinogen N-methyl-N'-nitro-N-
nitrosoguanidine is mutagenic while its noncarcinogenic
analogue N-methyl-N—nitro-N-nitroguanidine is not. Coupled
with the rat liver S-9 activation system, procarcinogens
such as nitrosopyrrolidine, benzo(a)pyrene, and 2-acetyl-
aminofluorene are mutagenic while their analogues 2,5-
dimethylnitrosopyrrolidine, pyrene, and fluorene are not.
The mutagenicity of the 49 agents documented to be either
carcinogenic or noncarcinogenic correlated well [47/49
(95.92%)] with the reported animal carcinogenicity. A
possible false negative was formaldehyde and a false posi-
tive was ICR-191. Preliminary studies show that the acetone
effluent (tentatively identifiable as heterocyclic nitrogen
compounds) derived from the basic fraction of a synthetic
crude oil (supplied by Pittsburgh Energy Research Center) is
the most mutagenic fraction. The assay, thus, appears to be
applicable for monitoring the genetic toxicity of crude
organic mixtures.
-------
582
MICROBIAL MUTAGENESIS TESTING OF AIR POLLUTION SAMPLES
Thomas J. Hughes, Research Triangle Institute, Research
Triangle Park, North Carolina; L. Little, L. Claxton, M.
Waters, E. Pellizzari, C. Sparacino
Air samples from U.S. cities with known high pollution
indices were screened for mutagenic activity using Ames1
Salmonella typhimurium reverse mutation detection system.
Objectives of the study were (1) identification of possible
sources of mutagenic pollutants, (2) determination of capa-
bility of the Ames test to detect such pollutants, and (3)
modification of the test for mass screening purposes.
Particulates collected from S. Charleston (WV) with the
Battelle Maxi-Sampler were partitioned into fractions which
were chemically identified and tested for mutagenic activity
with five tester strains (TA98, TA100, TA1535, TA1537,
TA1538). Compounds were tested for toxic and mutagenic
activity with and without metabolic activation. Because of
the minute amount of material available in each fraction
(0.2-17 mg), spot tests were used initially and fractions
showing activity were subsequently subjected to pour plate
testing. None of the fractions were toxic under the condi-
tions tested. Preliminary results suggest mutagenic activ-
ity in fractions containing organic bases, acids, and
aromatics. Marginal activity was detected with nonpolar
acids and neutrals. Each of the five tester strains gave a
positive response with at least one of the active fractions.
Of those fractions showing mutagenic activity, only the
aromatic, nonpolar acids, and neutral fractions required
metabolic activation. Sensitivity of the spot tests was
improved by increasing histidine concentration in the over-
lay. Results of pour plate tests confirmed spot tests,
however, activity was somewhat lower in pour plate tests due
to sample dilution. Results suggest that the assay can
detect mutagenic activity in small amounts of crude mixtures
and fractions, but the major problem is still availability
of the sample amounts. A system allowing testing of multi-
ple parameters (such as activation requirements and dose
response effects) on a single plate, thereby decreasing
total sample requirement is underway. Research was sup-
ported by EPA Contract No. 68-02-2724.
-------
583
X-RAY ULTRASTRUCTURAL STUDIES IN CADMIUM-COATED FLY ASH
PARTICLES
Peter Ingram, Research Triangle Institute, Research Triangle
Park, North Carolina; John D. Shelburne
Recent studies in several laboratories including our own
indicate that coal fly ash particles are extremely hetero-
geneous not only with regard to their size but also with
regard to their composition and perhaps most important their
surface chemistry. Utilizing the techniques and regimen
developed in our laboratories, the distribution and location
of elements in a model system of cadmium coated fly ash
particles was determined. The method involves making x-ray
ultrastructure maps on thin sections used for routine trans-
mission electron microscopy (TEM).
Coal fly ash particles were coated with cadmium and directly
embedded in Epon. After polymerization, gold (approximately
100 nm) sections were cut and examined by TEM, Scanning
Transmission Electron Microscopy (STEM), Scanning Electron
Microscopy (SEM), SEM backscatter, Energy Dispersive X-ray
(EDX) microprobe analysis, and EDX mapping.
EDX maps of the distribution of cadmium show a thin rim of
cadmium on the surface of the particles. Spot probes and
maps of the same particles show that the interior of most of
the particles consisted of silicon and aluminum; however,
some cadmium-coated particles consisted exclusively of iron.
With SEM backscatter these iron particles were prominent and
could readily be distinguished from the particles containing
silicon.
It is concluded that these particular techniques are espe-
cially valuable in studying the interaction of specific
particles and their surface with cell organelles at the
ultrastructural level.
-------
584
CONCENTRATION OF POTENTIAL MUTAGENIC COMPOUNDS IN TEXTILE
PLANT EFFLUENTS FOR APPLICATION TO THE SALMONELLA MUTAGEN-
ICITY TEST
Francine A. Kulik, Monsanto Research Corporation, Dayton,
Ohio; W.D. Ross
A solvent extraction method for concentrating potential
toxic and mutagenic compounds in textile plant secondary
effluents was evaluated. The method of concentration is a
modification of the extracation procedure for base-neutral
compounds described in "Sampling and Analysis Procedures for
Survey of Industrial Effluents for Priority Pollutants,"
EPA, March 1977. The use of solvent exchange between methy-
lene chloride and dimethylsulfoxide was used rather than
taking the extract to dryness and redissolving in DMSO.
The DMSO extract was applied to Ames' Salmonella typhimurium
strains for mutagenicity testing. The unconcentrated efflu-
ent was also applied to the microbial test. No activity was
seen with the neat sample and marginal mutagenic activity
was noted in TA98 and TA1538 strains in two tests. Micro-
bial toxicity was also observed in the concentrated samples.
Other methods of concentration are under investigation in-
cluding the use of XAD resins.
-------
585
MUTAGENIC ACTIVITY IN ORGANIC WASTEWATER CONCENTRATES
Stephen M. Rappaport, School of Public Health, University of
California, Berkeley, California; Monica C. Hollstein,
Michael G. Richard, Ronald Talcott
Organic wastewater concentrates from five treatment plants
in California were tested for mutagenicity using the Ames
mutagen bioassay. Four-liter samples of wastewater were
passed through columns containing two porous copolymer
resins (Amberlite XAD-2 and XAD-7) in series. Adsorbed
organic compounds were eluted from the resins with acetone,
dried, redissolved in DMSO, and bioassayed with the "soft
agar" plate test.
Of four tester strains (TA98, TA100, TA1535, TA1537) TA98
with the addition of Aroclor-induced ratliver enzymes was
the best for quantitating mutagenic responses. Extracts
from one-half of the sites were mutagenic in this bioassay.
Levels of mutagenicity were greater in both chlorinated and
unchlorinated secondary wastewater extracts than in primary
extracts, though toxicity effects were involved. Dose-
response curves were obtained for selected positive extracts,
Several extracts were separated into acidic, basic, and
neutral fractions.. Upon retesting, the basic and neutral
fractions were mutagenic while the acidic fraction showed
little activity. However, since >70% of the total mass was
found in the acidic fraction, its contribution to the gross
mutagenicity could not be ruled out.
-------
586
ASSESSMENT OF THE MUTAGENICITY OF AMBIENT AIR IN A NORTHERN
ROCKY MOUNTAIN REGION USING THE TRADESCANTIA SYSTEM
Larry Ricklefs, Montana State University, Bozeman, Montana;
D. Johnson, S. Rogers
The feasibility of using the higher plant, Tradescantia, as
one bioassay organism for air quality analysis in the North-
ern Rocky Mountain region was investigated. The Tradescantia
clone 02 has been found to be very useful in measuring low
amounts of ionizing radiation, and recently Tradescantia
clone 4430 has been successfully used to determine differ-
ences in the ambient air mutagen levels in a variety of
locations. Analysis of the unique ambient air mixtures
present in different rural, urban, or industrial locations
requires on-site exposure and data acquisition. Thus, a
number of ongoing Tradescantia colonies in different geo-
graphical locations in the United States would be a useful
addendum to the several Tradescantia monitoring projects
currently managed by the Brookhaven group. Tradescantia
clones 02 and 4430 were obtained from Lloyd Schairer in
the fall of 1976 and data collection began in June of 1977.
The Tradescantia plants were propagated using conventional
greenhouse facilities. Data were collected from plants
exposed to ambient unfiltered Bozeman air. The flowers were
collected from potted plants instead of cuttings. This
experimental design was selected to reduce to a minimum the
requirement for technical support and physical facilities.
Tradescantia clone 02 average hair count was 185 hairs/
flower over a six month period. A total of 407,835 clone
4430 control stamen hairs were scored and showed an average
mutation frequency of 0.163 pink events per 100 stamen hairs
in ambient air. A total of 197,710 clone 02 control stamen
hairs were scored and showed an average mutation frequency
of 0.155 pink events per 100 stamen hairs in ambient air.
The 4430 frequencies were found to be lower than Trades-
cantia background data reported for New York and Missouri.
A statistically significant downward trend of mutation
frequency in 4430 was also noted during the three month
scoring period.
-------
587
MUTAGENS IN AUTOMOBILE EXHAUST
Yi-Yuann Wang, School of Public Health, University of
California, Berkeley; Robert F. Sawyer and Eddie T. Wei
Particulate matter in city air contains chemicals which are
mutagenic in the Ames S_. typhimurium assay, a test system
which detects mutagens and some carcinogens. In residential
urban areas, the principal mutagens in air do not require
liver enzymes to be activated. The source of these liver-
independent (direct-acting) mutagens may be automobile
exhaust because (1) the mutagenic activities were correlated
to the lead content of air (r = 0.89, N = 28), (2) the
mutagens were in tailpipe exhausts of 5 cars and from an
experimental CFR single-cylinder, spark-ignited, internal
combustion engine using leaded-regular gasoline and (3)
these mutagens were not in fuel or unused motor oil, but
were in used motor oil (200 TA98 revertants/0.1 ml of used
motor oil).
What is the chemical identity of the exhaust mutagen(s)?
The air and exhaust samples were mutagenic in strains TA98,
TA100, and TA1537 and did not require liver enzymes for
activation. These facts indicated the mutagens were not
unsubstituted polycyclic aromatic hydrocarbons (PAH),
aromatic amines, alkylnitrosamines or aliphatic epoxides,
peroxides, and hydroperoxides. In the Ames test, nitro-
substituted polyaromatic compounds and some oxygenated
derivatives of benzo(a)pyrene are direct-acting mutagens.
PAH, especially the larger ones, are extremely sensitive to
oxidation or electrophilic substitution and possibly NOX,
HN02, or HNO., in exhaust may oxidize, add to, or substitute
in PAH. Thus, nitro-substituted PAH are possible candidates
to be the direct-acting mutagens in engine exhaust.
To test this hypothesis, we synthesized 6-nitrobenzo(a)pyrene
and found it to be a potent, direct-acting mutagen in TA98,
TA100, and TA1537 with activity comparable to that of benzo-
(a)pyrene. Although this fact suggests that nitro-substituted
PAH may be in automobile exhaust, further work is needed to
determine if these compounds are present in exhaust and to
assess their mutagenic properties in mammalian cells.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reierse before completing)
1 REPORT NO
EPA-600/9-78-027
3 RECIPIENT'S ACCESSION NO.
4 TITLE AND SUBTITLE 5. REPORT DATE
Application of Short-Term Bioassays in the Fractionatior
and Analysis of Complex Environmental Mixtures
6. PERFORMING ORGANIZATION CODE
AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
September 1978
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Toxicology Division
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
10. PROGRAM ELEMENT NO.
1NE625
11 CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Office of Research and Development
Health Effects Research Laboratory
Research Triangle Park, N.C. 27711
13 TYPE OF REPORT AND PERIOD COVERED
RTP,NC
14. SPONSORING AGENCY CODE
EPA 600/11
15 SUPPLEMENTARY NOTES
16. ABSTRACT
This report is the proceedings of a symposium convened at Williamsburg, Virginia
February 21-23, 1978. The volume consists of 24 formal presentations that amplify
the three major topics discussed during the symposium: an overview of short-term
bioassay systems; current methodology involving the collection and chemical analysis
of environmental samples; and current research involving the use of short-term
bioassays in the fractionation and analysis of complex environmental mixtures.
The purpose of these proceedings is to present the state-of-the-art techniques in
bioassay and chemical analysis as applied to complex mixtures and to foster continued
advancement of this important area of collaborative research. Complex mixtures
discussed include ambient air and water, waste water, drinking water, shale oil,
synthetic fuels, automobile exhaust, diesel particulate, coal fly ash, cigarette
smoke condensates, and food products.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Bioassay
mixtures
air
shale oil
exhaust emissions
fly ash
smoke
food
water
b. IDENTIFIERS/OPEN ENDED TERMS
short-term bioassay
c. COSATI Field/Group
06, F
18 DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (This Rfportl
UNCLASSIFIED
21. NO. OF PAGES
20. SECURITY CLASS (This pa gc
UNCLASSIFIED
22. PRICE
EPA Form 2i20-l (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
-------