Retrospective View of the Value of
Short-Term Genetic Bioaasays in
Predicting the Chronic Effecta of Diesel Soot
(U.S.) Health Effects Research Lab.
Research Triangle Park, NC
Aug 86
PB86-240934
* ;'"r 1 fTCf--' *"-*^»--* -••Y^iiftj-^fcy
s^,ii.ffiSK7'?^^
K^
-------
TECHNICAL REPORT DATA
If lease read /mtrucnons on iHe reverse before completing)
1. REPORT NO.
EPA/600/D-86/189
z.
3. RECIPIENT'S ACCESSION NO.
PB86-2409347AS
4. ' ITLE AND SUBTITLE
A RETROSPECTIVE REVIEW OF THE VALUE OF SHORT-TERM
GENETIC BIOASSAYS IN PREDICTING THE CHRONIC EFFECTS
OF DIESEL SOOT
5. REPORT DATE
August 1986
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Joellen Lewtas and (Catherine Williams
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
USEPA, HERL/GTD/GBB, RTP, NC 27711
10. PROGRAM ELEMENT NO.
ANNA1E
11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS
Health Effects Research Laboratory - RTP, NC
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCV CODE
EPA/600/11
15. SUPPLEMENTARY NOTES
16. ABSTRACT
• In retrospect, it is now safe to conclude that short-term mutagenicity assays
were not only useful but instrumental in: (1) indicating that diesel soot was potentially
carcinogenic and should be evaluated in chronic animal cancer bioassays, (2) identifying
V02~PAHs as potential carcinogens in this very complex mixture, (3) providing initial
evidence that the mutagens were btoavailable, and (4) estimating the relative importance
of various sources and fuels and other factors which can influence human exposure to
carcinogens. This is not to say that short-terra bioassays used alone can accomplish all
af this. However, used in combination with chemical/analytical methods and toxicological
tools, short-term genetic bioassays have become a critical component of many environmental
lealth studies. Although substantial advances in our knowledge of the toxicology of
diesel emissions have been made since 1978 when the initial observation that the organics
extracted from diesel soot were mutagenic, a number of important questions remain not
only for diesel emissions but for other combustion sources as well. Are the chemicals
*hich induce positive results in the short-term bioassays the same agents which cause
tumors in chronic animal bioassays? Which phase of the diesel emissions (gaseous or
particulate) is carcinogenic in the animal inhalation studies? With advances in our
understanding of the molecular mechanisms involved In producing chronic effects such as
cancer, it is possible that new genetic tools and short-term bioassays will continue to
contribute to our ability to answer these and other questions as they arise.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTOR?
b.lDENTlFIEPS/OPEN ENDED TERMS
c. COSATi Field/Group
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (Tliii Report/
JNCLSSSIFIED
21. NO. OF PAGES
20. SECURITY CLASS iThit pagtl'
UNCLASSIFIED
22. PRIC
EPA Fern 2220-1 (R*«. 4-77) PREVIOUS EDITION i< OBSOLETE
-------
PB86-2<*093<*
EPA/600/D-86/189
August 1986
A RETROSPECTIVE VIEW OF THE VALUE OF SHORT-TERM GENETIC BIOASSAYS IN
PREDICTING THE CHRONIC EFFECTS OF DIESEL SOOT
by
Joellen Lewtas and Katherine Williams
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
HEALTH EFFECTS RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK.' NC 27711
-------
NOTICE
This document has been reviewed in accordance with
U.S. Environmental Protection Agency policy and
approved for publication. Mention of trade names
or commercial products does not constitute endorse-
ment or recommendation for use.
11
-------
A RETROSPECTIVE VIEW OF THE VALUE OF SHORT-TERM OENETIC
BIOASSAVS IN PKEDICTING THE CHRONIC EFFECTS OF DIESEL SOOT
JOELLEN LEV.TAS AND KATHERINE WILLIAMS
Genetic Bioas5ay Branch. Genetic Toxicology Division. Health Effects Research
Laboratory, U.S. Environmental Protection'Agencv. MD 68, Research Triangle
Park. NC 27711 (U.S.A) -
The organic matter extractable from diesel soot particles was first reported to be
mutagcnic in bacteria in 1978 (1-3). This finding was rapidly confirmed by many
groups (-1-6). These organics were also found to cause gene mutations. DNA
damage, and chromosomal effects in several mammalian cell systems (7-11).
Bioassay directed fractionation and characterization studies using bacterial
mutagenicity assays indicated that compounds more polar than polynuclear
aromatic hydrocarbons (PAHs) »ere responsible for most of the nutagenicity in
these organics (1, 4-6. 12). By 1982 a number of potent nitrated PAHs (NO-,-
PAHs) had been identified in diesel soot (13-16).
••
Short-term genetic bioassays were used in studies designed to determine the
bioavailability and metabolism of diesel soot mutagens (4, 17. IS). Concern that
bacterial mutagenesis assays may "overestimate'* (he mutagenic activity of diesel
soot due to the presence of NOi-PAHs. led to studies on the mammalian
metabolism and DNA-binding of NO-,-PAHs alone and associated with diesel soot.
These studies showed that the bacterial mutagens and NO,-PAHs were rapidly
released from diesel particles. Metabolism studies in both whole animals and
mammalian target cells (e.g.. tracheal ceils) have demonstrated that NOi-PAHs
are metabolized by both oxidathe and reductive pathways to produce metabolites
that bind covjlently to DNA (19-21). In diesel soot extracts, the concentration of
certain NO->-PAHs is highly correlated with the mutagenicity and tumor-initiating
activity of the extracts (22). Soots from many other combustion sources (e.g.,
wood stoves and gasoline automobiles) contain substantially less NO-,-
PAHs than diesel emissions (15).
Short-term genetic bioassays have been used in a series of studies designed to
--"determine (he comparative potency and characteristics of various diesel soot
extracts compared (o other combustion sources (7, 23. 24). Bacterial and
mammalian cell assays have been compared to mouse skin tumor initiation assays.
These studies suggest that the relative potency of diesel soot extracts in bacterial
or mammalian cell assays can be used to piedict the relative carcinogenic potency
in rodent assays and in humans under certain conditions and assumptions.
19
Comparative studies of (he mutagenic emission rates of various automobiles and
fuels provide a simplified method for directly comparing alternative sources and
technologies.
-------
Studies to evaluate the mutagenicity of whole diesel emissions containing both
the soot particles and gases hare been conducted in plants, insects, and mammals
(8, 25). Bacterial assays were used to evaluate the gaseous organics that could be
collected by adsorption or condensation. These studies demonstrate that the
gaseous components also contain mutagens. In vivo mutagenesis studies of rodents
after relatively short inhalation exposures, however, show much less nu'tagenicity
and demonstrated no heritable effects (25). This suggests that either the
oiutagenic components of diesel emissions did not reach the gonads under these
exposure conditions, that these assays are insensitive to the mutagens present in
diesel emissions, or the effects were helow the level of detection.
Incomplete combustion of many types of fuels result in the production of soot.
The International Agency for Research on Cancer's (IARC) Monograph on Soot
(26) concludes that there is sufficient evidence that soot is carcinogenic to
humans. Very few studies of humans exposed to diesel soot, however, have clearly
shown evidence of increased cancer risk. Many soot extracts have also been shown
to be carcinogenic in experimental animals; however, few rodent inhalation studies
of combustion emissions have demonstrated carcinogenicity. New studies, reported
in this volume, showing that diesel emissions are carcinogenic to rodents after
chronic inhalation are consistent with the IARC conclusion that soot is a human
and animal carcinogen. These rew results are also consistent with the positive
short-term genetic bioassay results reported over five years earlier. In this paper
we take a retrospective vie* of the short-term genetic bioassay data developed
over the past eight years and their value in predicting chronic carcinogenic
effects.
WHAT WERE THE EARLY BIOASSAY CLUES REGARDING THE RISK OF
DIESEL SOOT?
We found the e.vtractable organics (tar) from diesel particle emissions to
consistently cause positive responses in short-term mutagenesis and carcinogenesis
bioassays (8, 11) as summarized in Table I. Other investigators have reported
similar samples to be mutagenic in human cells (9, 27). We have recently
evaluated the genetic activity (potency) data from diesel and gasoline automotive
emission.'; in a number of bioassays using the genetic activity profile method of
Waters (28, 29). These profiles provide a graphic representation of bioassay data
that facilitates both visual and computer-assisted comparative assessment. The
concept of a genetic activity profile for a chemical originated from the need to
represent in a single two-dimensional configuration the qualitative and
quantitative data from a large number of genetic bioassay systems (currently more
than 200). The x-axis values of the profiles shown here (Fig. 1) represent the
bioassays in an endpoint/pliylogenetic sequence, and the y-axis values represent
-------
TABLE I
SUMMARY OF THE RESULTS UF SHORT-TESM BIOASSAYS USED IN THE EVALUATION OF DIESEL
PARTICLE EXTRACTS
Results for
3iesel
Genetox Particle
Assays Code Organics
Mutagenesis Bioassays
Gene Mutation Assays
Bacterial
Salmonella typhimurium SA9 +
Escherichia coli WP2 EC2 *
Escherichia coli K12 ECK +
Mammalian cell
Mouse lymphoma, L5178Y/TK*/- locus G5T +
Chinese hamster ovary, CHO, HGP3T locus GCO «•
Chinese hamster lung, V79, HGPRT locus G9H (-)
Mouse embryo fibroblasts, Balb/c 3T3, Ouar GIA (*)
DMA Damage Assays
Yeast
Saccnaromyces cerevisiae
03 mitotic recombination assay . SC3 (+)
03 preincubation mitotic recombination assay SCP
04 mitotic gene conversion astiy SCG
07 induced mitotic crossing over SCH (+)
07 reverse mutation SCR
07 gene conversion assay SCG
Mammalian Cell
Unscheduled ONA repair in liver cells UPR *
Sister chromatid exchanges in CHO cells SIC *
ONA strand breaks in SHE cells . OIA
Chromosomal Aberrations
Mammalian Cells
CHO cells CIC *
Human lymphocytes CHL (+)
Carcinogenesis Bioassays
Oncogenic Transformation Assays
Chemical Transformation
Mouse embryo fibroblasts, Balb/c 3T3
- Mouse embryo fibroblasts, C3H10T1/2
Syrian hamster embryo, SHE, focus assay
Viral Enhancement of Transformation
SA7 virus enhancement in SHE cells
Skin Tumor Initiation
SENCAR mice
C57 Black mice
C3H/HeJ mice
T8M
TCM
TFS
TEV
SKT
(*)
() indicates the response was either: weak, not reproducible, or was observed in
only one sample
-------
Meroedes
Diesel
10
6
cn
LU
o 6
1
cn
a: A
< 4
LU
2-
^j
2
•
12
.'
.-
. •
t
_7
.j
Jl
•3
' j
n
0
1
*
I
1
SJIUI
*
u
•j
^
•r.
h'
*
1
* *
y
t
ii
^
'i;
'• '1
iH
k
F"
A
JB
3;
. i
?
,\
.7^
•J
'*i
•H
-S
J
•>!
'1
r.'i
3
P1+ MA
i^iijtJ
- MA
airu^i— oxno-coiarcoucrcjaErs:
-------
the potency or linear slope of the dose response.
These results all suggest (hat diescl soot contains mutagens and carcinogens and
that inhalation of diesel soot is very likely to pose an excess cancer risk to whole
animals and humans. The negative results reported in the first several inhalation
carcinogenesis studies (30-33) led to questions about the use of short-term
bloassays and inhalation studies such as: 1) Were the carcinogens associated with
the particles not bioavailable? 2) Did the short-term mutagenesis bioassays
"overestimate" the potential carcinogenicity of the diesel emissions? 3) Were
there mitigating factors present in the whole emissions that had anti-carcinogenic
properties? or 4) Were the animal inhalation models or study designs employed
simply inappropriate?
Several of these questions were addressed by using genetic bioassay methods as
described below.
ARE THE MUTAGENS BIOAVAILABLE?
Initial studies reported that the mutagens associated with diesel particles were
not "bioavailable" when incubated with lung fluids or other physiological fluids
(34, 35). Subsequent studies demonstrated that the mutagens and specific
components (e.g. 1-nitropyrcne) were removed from particles'after incubation with
lung fluids and lung cells (36, 17); however, the proteinacious fluids were directly
antimulagenic by a protein binding mechanism in the Ames assay (36). Other
investigators have demonstrated that the antimutagenic activity of 59 with diesel
extracts was non-enzymatic (37).
A number of studies have been reported that directly assess the mutagenicity and
bioavailability of mutagens from whole particles. Cultured cells attached to a
surface will readily engulf small particles; and in two independent studies, whole
diesel particles induced mutations in CHO cells (38) and human fibroblasts (27).
The most convincing studies on the bioavailability of mutagens from diesel
particles have been conducted in vivo either by inhalation or iniratracheal
instillation. The inhalation studies (39—11) are described in more detail in this
volume; however, they generally support the Unavailability of mulagens observed
lo the short-term bioassay studies. We have found that C-nitropyrene vapor
coated onto diesel particles and intratracheally instilled into rats was readily
released from the particles both in the lung and CI tract (18).
WHAT ARE THE MUTAGENS IN DIESEL SOOT?
The discovery that orgaalcs from diesel soot were mutagenic (1) resulted from
studies to fractionate and chemically characterize the mutagenic constituents. We
Initially used a fractlonation scheme developed by Swain et al (42) for cigarette
smoke condensate. Ths moderately polar and highly polar fractions contained
-------
most of the bacterial mutagenic activity as shown in Table II. The moderately
polar fraction, which was the most mutagenic in Salmonella, induced mutation and
oncogenic transformation in Balb/c-3T3 cells but did not induce mitotic
recombination la Saccltaromyces cerrerisiae D3 (Table III). The fraction which
contains the most polar and highly oxygenated species, was also mutagenic In
mammalian cells in the absence of S9 and Induced mitotic recombination and
oncogenic transformation. Conventional gas chromatography/mass spectroscopy
identified many ftuorenones and methylated ftuorecones as major constituents of
these mutagenic fractions. None of these or other identified constituents
accounted for the direct-acting frameshift mutagenic activity observed. Studies
with nitroreductase-deflcient strains of Salmonella lyphimtirium showed a
substantial reduction in the mutagenicity suggesting that nitrated corn-pounds
contributed to this direct-acting mulagenicity (43). Nitrated polycyclic aromatic
hydrocarbons (NOj-PAHs) are potent direct-acting frameshift mutagens initially
detected in xerographic toners (44). A series of NOj-PAHs were later identified
and quantilated in diesel extracts In order to estimate their contribution to the
mutagenic activity 31* diesel paniculate emissions (15, 45). These studies showed
that NO2-PAHs, di-NO,-PAHs, and hydroxy-NOj-PAHs together account for
much of the mutagenicity observed in Salmonella lyphimurium. Particulate
emissions from catalyst-equipped gasoline-engine vehicles using unleaded fuel
contain significantly less of these NC>2-PAHs (15). The mutagenic activity of both
leaded- and unleaded-gasoline emissions is substantially Increased with the
addition of an exogenous metabolic activation (MA) system, suggesting that the
unsubstituted PAHs may play a more important role than do NO,-PAHs in the
mutagenicity and carcinogenicity of gasoline emissions (8, 11).
Although NOt-PAHs were identified in extracts of diesel particles and urban
air paniculate matter, the quantification of these compounds at low levels has
posed problems for analytical chemists because the conventional analytical
techniques for quantifying PAHs are relatively Insensitive to nitro-substituted
PAHs. The dinitropyrene homers are so highly mutagenic in the Ames (TA98 -59)
assay that trace concentrations of these compounds, if present, could account for
a major proportion of the observed mutagenic activity. Therefore, a
capillary column GC/MS analytical technique using on column injection
and negative chemical ionization (NCI) detection was developed to detect
these and other nitro-PAHs at very low concentrations (15). This technique »at
applied to extracts of soot particles from diesel and gasoline vehicles and
urban air particles. Over twenty different NO->-PAHs were identified in the
diesel engine extracts. 1-Nitropyrene was the NO^-PAH detected in greatest
quantity in the diesel extracts (107-1590 ppm relative to the weight of ilu-
extract), followed by the nitrophenanthrcne/anthracene isomers. The onl>
-------
TABLE II
DISTRIBUTION OF THE MASS AND BACTER1AL/MUTAGENIC ACTIVITY OF FRACTIONATED DIESEL
PARTICLE OP.3ANICS
SpecificDistribution of
Mutagenic Activity Mutagenic Activity
Fraction Mass (n
-S3
Organic acids
Organic bases
Ether insolubles
Paraffins
Aromatics
Moderately Polar
Highly Polar
Unfractionatad (DCM)
»
TABLE III
14.9
0.3
3.9
36.7
6.9
5.0
26.9
—
193
43.8
53.9
Neg.
49.5
7520
629
2557
248
132
80.9
"eg.
30.1
2620
798
1625
4.9
0.02
0.36
0.0
0.60
64.9
29.2
9.5
0.10
0.80
0.0
0.54
33.5
55.4
COMPARISON OF THE 8IOASSAY ACTIVITY OF THE TOTAL ORGANICS AND TWO FRACTIONS IN
FIVE SHORT-TERM 8IOASSAYS
Bioassay
Salmonella typhimurium
(revertants/ug)
Mitotic Recombination in
Total
Extract3
-59
2.6
0.3
+S9
1.6
0.1
Moaerately Polar
Fraction
-S9
7.5 •
Neg.
+39
2.6
Neg.
Hignly Polar
Fraction
-S3
0.6
0.2
0.8
0.5
Saccharomyces cerevisiae 03
(mitotic recombination/
xlO'6)
Gene Mutation in L5178Y NT NT 5.7 0.7
Mouse Lymphoma Cells
(mutation frequency/
ug/ml x 10-6)
Gene Mutation in Balb/c 0.6 0.05 1.2 1.6 1.6 1.4
3T3 Cells
(mutation frequency/
x 10-6)
Oncogenic Transformation 0.3 0.06 1.4 0.6 0.8 1.0
in Balb/c 3T3 Cells
(transformation frequency/
ug/ml xlO"5) *
aTotal extract and fractions are those shown in Table II.
. 7
-------
dinitro-PAHs for which analytical standards were available, the dinitropyrene
isomers, were delected in one diesel extract sample at sub-ppm concen:r?:ions (0.4-
0.6 ppm).
NO,-PAHs. ARE THEY THE CULPRITS?
Quantification of the concentration of the NO->-PAHs in diesel soot and
determination of their contribution to the direct-acting trutagcnicity in Salmonella
lyphimiiritim TA 98 (Table IV) shows that although 1-nitropyrene was present at
the highest concentration (107-1590 ppm) in the diesel particle extracts, it
accounted for only 3-13% of the mutagenicity. 3-Nitrofluoranlhene
present at 1 ppm to 7 ppm in the diesel samples accounted for 0.8% to 1.4% of
the mutagenicity. By using the mutagenicity values determined in separate
experiments for 2-nitrofluoreue and vnose reported in the literature for 1-
nitronaphthalene. these compounds were estimated to account for less than
0.01% of the mutagenic activity. Although the dinitropyrene isomers (1,3:
1,6; and 1.8) were detected in only one diesel sample (Auto 2) at 0.4-0.6 ppm (sum
of 1.6 ppm), their mutagenic activity (496,000; 629,000; and 870,000 rev/mg,
respectively) was high enough to account for 26% of the mutagenicity of
this sample. The total "direct-acting" mutagenic activity In Salmonella
typhimurium TA98 that can be accounted for by the 23 nitro-PAHs quantified
In Diesel 2 is 40%. This estimation is supported by the loss of 50% of the
mutagenic activity of this extract when it was assayed in TA98NID a
classical nitrorcductase-deficienl Salmonella lypliimiirinni tester strain obtained
front H. Rosenkranz (49).
The fact that dinitropyrenes at concentrations below the ppm level can account
for nearly one-third of the mutagenic activity (50) suggests that the presence of
other highly potent dinitro-PAHs may account for even more of the mutagenic
activity in the moderately polar neutral fraction. A recent application of S.
typhimurium tester strains developed to exhibit resistance to dinitropyrenes (e.g.,
TA98/l,8DNPg) lias led to even larger estimations of the contributions of
dinitropyrenes to the mutagenicity of diesel particle extracts. The lack of
quantitative data on the concentrations of the dinitropyrene isomers in such
samples has previously made it impossible to confirm whether the concentrations
of the dinitropyrenes are indeed sufficient to account for the contribution
predicted by the resistant tester strains. Pederson (51) has reported that in
several light-duly diesel particle extracts, 50% to 90% of the TA98 (-S9)
mutageniclly Is lost when the extracts are tested in TA98/1.8DNPg. Because these
strains may show a resistance to other dinitro-PAHs, it is possible that highly
potent dinitro-substituted isomers of other parent PAHs present in the particle
extracts may also contribute to the mutagenicity of the moderately polar neutral
-------
TABLE IV
CONTRIBUTION OF NO^-PAHS TO THE MUTAGENIC ACTIVITY OF PARTICLE EXTRACTS IN
SALMONELLA TYPHIHURIUM TA93 (-S9)
Extract
Sample
Diesel
Auto 1
Auto 2
Auto 3
Gasoline
Auto 4
Mutagenic
Activity
(rev/wg)
TA98 -S9
13.
3.9
3.5
1.6
1-nitropyrene nitrofluoranthene dinitrooyrene isomers
(ppm) {•.)« (ppm) (*.} (ppm) ('.)
1590 11. 7.0 1.4
589 13. 1.2 0.8 1.6 26.
107 2.7 0.9 0.8
2.5 0.1
aPercent contribution to the total mutagenicity in Salmonella based on the
concentration and mutagenicity of the individual MUg-PAH.
ifr1 '
TABLE V
NITSOPYRENt AND NITROFLUORANTHENE CONCENTRATIONS IN DIESEL PARTICLE EXTRACTS
AND CORRELATION ANALYSIS WITH HUTAGENIC AND TUMORIGENIC ACTIVITY
Particle Extract
Diesel Auto 1
Diesel Auto 2
Diesel Auto 3
Gasoline Auto 4
Correlation Coef.
r2 with 1-r;?
r2 rfith 3-KF
1-Npa
ppm
1590
589
107
2.5
3-NF&
ppm
7.0
2.9
1.2
0.9
Ames
TA98 (-S9)
rev/pgt
13.0
3.9
3.5
1.6
0.91
>0.99
Mouse
Lynphoma (-59)
MF/yg/mld
4.2
0.98
1.2
0.38
0.9P,
0.99
Tunor
Initiation
pap/mouse/-^
590
210
310
170
U.82
0.95
al-Nitropyrene
&3-NitrofIuoranthene
cRevertants per ug
^Mutation frequency (mutants p^r 106 survivors) per yg per ml
ePapillomas per mouse per ug in SENCAR mice
-------
fraction.
Good correlations (r~>0.90) were observed when the slope of the dose-response
for the mutagenic activity of this same seies of automotive particle extracts in S.
typhimurium strain TA98(-S9) was plotted versus the mutagenic activity in the
mammalian cell assays and the skin tumor initiating activity. The correlation of
mutagenic and skin tumor initiating activity with the concentration of selected
nitro-PAHs was examined for these diesel and gasoline samples. Table V" shows
the high correlations observed between the concentrations of 1-nitropyrene and 3-
nitrofluoranthene and the mutagenic activity in Salmonella typhimurium (-59),
L517SY mouse lymphoma cells (-S9), and skin tumor initiating activity in SENCAR
mice. The r correlation coefficient in the presence of S9 (not shown) was
somewhat lower. The mutagenicity and tumor initiation activity (-S9) also
correlated well (r >C.9) with the concentrations of the nilro-252 isomer, the nitro-
228 isomers, nitromethylpyrenes, and nitrofluorenes: while no correlation was
observed with several of the other less mutagenic, lower molecular weight nitro-
PAHs (nitro-phenanthrenes and nitronaphthalenes). The quantified mono-NO-,-
PAHs account for less than 20% of the direct-acting bacterial mutagenicity of
these samples. The high correlations observed between .the concentration of these
compounds and the mutagenic activity of the total extract, therefore, suggest that
the unidentified mutagens responsible for the remainder of the mutagenic activity
and possibly the mutagenic and carcinogenic activity in other bioassays is directly
related to the relative concentrations of these mono-NOt-PAHs. Because the
remaining unidentified mutagens appear to be located primarily in the chemical
fractions that are more polar than the fraction that contains the unsubstituted
and mono-NO2-PAHs, it is possible that other di-NO->-PAHs (e.g..
dinitrofluoranthenes) or other oxygenated NO-»-PAH*s (e.g., hydroxy-nilro-PAHs)
species are responsible for the unidentified mutagenic activity.
ARE THE ORCANICS FROM DIESEL SOOT SIMILAR TO OTHER
COMBUSTION ORGANICS?
Characterization of the mutagenicity of emissions from other combustion sources
shows some general similarities (46). In both wood and diesel combustion. 82-99%
of the mutagenicity was in the neutral fraction. Very little mass or mutagenic
activity was observed in the organic bases. Differences between these sources were
observed in the distribution of mass and mutagenicity in the neutral subfractions.
Bioassay-directed fractionation and chemical characterization studies also hate
been used to characterize and compare the complex organic emissions from roofing
tar pots, coke ovens, and cigarette smoke. To obtain a gross characterization of
the chemical classes present in the samples, the chemical class distribution was
determined by solvent partitioning the orga-nics into acidic, basic, neutral, and
10
-------
cyclohexane insoluble fraciions. The ceutral fractions were further separated
into the ftonpolar neutrals, aromatic* (nitre-methane soluble) and polar
neutrals. Characterization of the distribution of bacterial mulaqcnic activity
(Ames Salmonella lyphimiirium bioassay) in each of these fraciions (47. -IS)
showed significant differences between the diesel. coke o»en main, roofing
tar. and cigarette smoke condensate samples as shown in Fig. 2. In the diesel
samples, over 90% of the mutagenic actiiity was located in the aromatic and
polar-neutral fractions, and a significant portion of this activity can be
accounted for by NO-»-PAHs. The cigarette smoke condensate. coke o»en
main, and roofing tar samples did not contain detectable amounts of NO-,-
PAHs (48). Most of the mutagenicity of coke oven main sample was found in
the basic fraction (37%) and polar neutral fraction (39%). The cigarette smoke
condensate sample also had significant activity in the basic fraction (66%).
but chemical analysis indicated that the components differed significantly from
those of the coke oven main sample. The roofing tar sample contained aromatic
(14%) and polar (75%) mutagenic con-stituents that were not NO^-P.AH's. The
PAH subfraction of each of these samples accounted for only a small portion
of the mutagenicity [e.g. diesel (0.2%), cigarette smoke condensate (0.1%).
roofing tar (5%). and coke oven main (S%)|.
Although the specific mutagens in these different sources are not identical, they
all cause frameshift mutations and appear to be compounds that could be
classified as polycyclic organic matter. Chemical characterization suggests that in
addition to nitrated NC^-PAHs found in the slightly and moderately polar
neutrals, hydroxylated and carboxylated polycyclic organics are found in the
Organic acid fraction, aromatic amines and nitrogen heterocycles are found in the
organic bases, and highly oxygenated quinones. diones. and nitro-ox\ genated
compounds aro found in the polar neutral fractions.
CAN GENETIC BIOASSAYS BE USED TO ESTIMATE RISK FROM NEW-
ENGINES. FUELS OR CONTROL TECHNOLOGIES?
A comparative potency method has been developed for cancer risk assessment of
diesel particle emissions faasec1 on a constant relative potency hypothesis and using
data from a battery of short-term mutagenesis bioassays and animal
tumorigtnicity studies on a series of diesel vehicles emissions (23.24). These same
bioassays were used to evaluate three complex emissions for which human lung
cancer risk estimations were available (emissions from coke ovens, roofing tar
pots, and cigarette smoke).
The comparative potency method for cancer risk assessment is based on the
hypothesis that there Is a constant relative potency between two different
carcinogens (Cl and C2) across different bioassay systems (Bl and B2). This
11
-------
01
w /u
g
< 60
|
'§ 50
1
§ 40
to
i 30
*•»
'>
+•*
u
«* 20
o
'c
«
3 w
3
5
^o A
p O Diesel
~
-
-
-
-
n r
=
=
1
F
i
=
=
1
|
•7.
•_r
:.-.
r-
~
=
^ CSC
CD Coke Oven
Wh Roofing Tar
i
i
i
i
i
aw^s
ILlLJBi.lr
Acid Base NPN PNA1 PNA2 PNA3 PNA4 PN
Fractions
m
'//
//>
1
m
-
,-
V
"•
i;
$
\
•=
"^
:~-^.
.-:
|
|
1
1
1
|
^
^
^
,//
//
1
Cl
Flcj. 2. Distribution of mutjijenlc activity across fractions separated by liquid
partitioning (48). Fractions are as follows: organic acids (Acid), organic
bases (Base), nonpolar neutral-alphatics (NPN), polynuclear aromatic hydrocarbons
(PNA1), moderately polar fractions separated from the PNA fraction (PNA2, PNA3,
PNA4), polar neutrals (PH) and highly polar cyclohtxane insoluahles (CI).
-------
constant relative potency assumption has been tested most extensively for diesel
emissions where the largest series of emissions has been tested in seteral
mutagenesis and carcinogenesis bioassays. Tne test of this model is whether there
is a constant relationship (k) between the relative potencies in the two bioassays
beinrc compared such that:
Relative Potency (C1/C2) in Bioassavf 1) = constant (k)
Relatue Potency (C1/C2) in liioassay(2)
Based upon the data avialable on diesel emissions, it appears that this assumption
holds when comparing the relative potency in the Ames Salmonella typhimurium
(TA 98) assay, the mouse lymphoma assay,the sister chromatid exchange assay in
CHO cells, and mouse skin tumor initiation in SENCAR mice. A number of short-
term bioassays that did not result in quantitative dose-response data or in which
the responses were very weak could not be evaluated.
In order to adequately test this hypothesis for prediction of human cancer risk,
a more extensive comparative data base is needed for human lung cancer. This
hypothesis was tested for only the three complex organic emissions from a coke
o»
-------
TABLE VI
COMPARATIVE MUTAGENIC AND TUMORIGENIC EMISSION RATES
Organic Potency of Organicsb
Emission Ames House SCE Skin
Vehicles3 Rate TA98 Lymph. CHO Tumor
(+S9) (+59) (+S9) Init.
(mg/kn)
Activity
Emission Ratesc
Ames Mouse SCE Skin
TA98 Lymph. CHO Tumor
(+S9) (+S9) (+S9) Init.
(xlo5) (xlO4) (xlO3)
Diesel
Car
Car
(Mercedes)
(VW Rabbit)
Truck (Ford/Cat)
Bus
(GM)
20
52
312
362
.2
.2
.0
.0
12
6
1
0
.0
.1
.7
.1
1
0
0
0
.5 0.16 0.37
.72 0.03 0.24
.23
.35
2.4
3.2
5.3
.4
3
3
8
13
.0
.8
.5
.0
3.2 7.5
1.6 12.5
...
—
Gasoline
Non-catalyst
(Ford Van)
Catalyst
(Mustang 11)
5.61 32.0 5.7 0.47 0.20 1.8 3.2 2.6 1.1
3.67 8.6 1.1
0.16 0.3 0.4
0.5
aThe methods for collecting these emissions and performing the mutagenesis Dio-
assays have been reported-elsewnere (7,3,46).
bThe mutagenicity of the organics in the Ames bioassay is expressed in revert-
ants/ug (rev/ug) in the mouse lymphoma bioassay as mutants/10^ survivor/yg/ml
(mut.freq./yg/ml). SCE as SCE/cell/ug/ml and skin tumor initiation as papil-
lomas/mouse at 1 mg.
cThe rcutagenic and turaorigenic emission rates were determined from multiplying
the mutagenicity of the organics times the organic emission rate to give
activity emission rates per km driven.
14
-------
gasoline automotive emissions to unit risk estimates for various combustion
emissions or other complex mixtures through the data base established linking
human, animal, and short-term genetic bioassay data. A simplified comparative
approach to evaluating alternative energy sources is to employ parallel bioassay
studies of the alternative (a) and conventional (c) source emissions and determine
a relative risk by direct comparison as follows: increased risk (a/'c) = relative
bioassay potency (a/c). Because there is no one conventional standard
petroleum-derived fuel or one standard combustion source, such studies need Co
consider the range of mutagenic and carcinogenic potency between different
conventional sources and fuels. The establishment of such a range could then
serve as a guide for evaluating alternative fuels or sources.
In order to evaluate a battery of short-term genetic bioassays for their utility in
testing and assessment of unregulated automotive emissions, a matrix of bioassays
and particle extract samples from various engines and fuels was constructed (54).
Three heavy-duty vehicles, each fueled by three to five grades of diesel fuel, were
operated on the '83 transient driving cycle. The light-duty automobiles were
fueled by gasoline (leaded or unleaded) or the No. 2 diesel fuel. The short-term
*
bioassays included those shown in Fig. 1 and included both mutagenesis (gene
mutation, DNA damage, and chromosomal effects in procaryotcs and eucaryoles),
oncogenic transformation, and mouse skin tumor initiation. The mutagenesis
bioassays were of three types: gene mutation in both bacterial and mammalian
cells, DNA damage assays in yeast and mammalian cells, and chromosomal effects
in mammalian cells. The carcinogenesis assays included oncogenic transformation
in two lines of mouse embryo fibroblasts and the mouse skin tumor initiation assay
in SENCAR mice. One sample from a diesel Mercedes car was tested in all 16
assays (Fig 1). The data has been compared as mutagenic emission rates.
This data base of mutagenic emission rates has been evaluated by an analysis of
variance (ANOVA) method to determine if any of the emissions produced
significantly different (p-0.05 level) responses (54). Because the ANOVA did not
detect any significant difference with or without S9 activation, all of the results
were combined for this analysis. Only the data from the following three bioassays
where adequate for such ANOVA analysis: Ames Salmonella typhimuriiim assay,
mouse lymphoma gene mutation assay, and sister chromaioid exchange assay in
CHO cells. Among the three tests, the Ames test exhibited greatest differences
between the samples as shown in Fig. 3. Samples 5 (heavy-duty Caterpillar
engine) and IS (light-duty Nissan engine), both run on minimum grade diesel fuel,
hoc" significantly higher mutagenic emission rates than the other vehicles and
fuels. The two mammalian cell bioassays showed higher standard deviations, and
none of the emissions were found to'be significantly different, possibly due to the
lack of replicate bioassays.
15
-------
ANALYSIS OF VARIANCE FOR ARTiES TEST ASSAYS
*>
£
2
V)
h-
LU
UJ
ouu
540
480-
420-
380-
300-
240-
•i fift-
lou
120-
60-
0-
.RfH-
5 T
f f
J I T
T
I
I? ' } * J J ? I
j. - x -j .
23456
1011121314
SAMPLE NUMBER
16 171819
I i'|. I. Analysis ot viirijuce for Ames test .
-------
The increased or decreased human cancer risk from combustion emissions may be
of greater concern than the absolute risk from these emissions when we consider
the effect of alternative energy technologies and fuels. The data of greatest value
In determining this increased or decreased risk are comparative bioassay data on
the conventional and alternative technology being evaluated. Such studies should
be useful in providing direction for engineers and chemists to design alternative
energy sources and fuel:; that result in less mutagenic and potentially less
carcinogenic emissions.
The comparative potency method for cancer risk assessment described in this
paper cannot be employed without uncertainty or without invoking the constant
relative potency assumption inherent in the method. It is important to recognize
that the assumptions and uncertainties are different from those employed in using
either low-dose extrapolation techniques with human data or scaling factors with
animal data. This makes it possible, therefore, to perform quantitative cancer risk
assessments using more than one approach with different methodological
assumptions. In the future, more information may be gained by comparing the
quantitative assessments for one source using several independent cancer risk
assessment methodologies.
ARE DIESEL PARTICLE EMISSIONS A SIGNIFICANT SOURCE OF HUMAN
EXPOSUE TO MUTAGENS AND CARCINOGENS?
The relative contribution of different sources to ambient particle concentrations
has been determined using either dispersion or receptor modeling. These studies
show that combustion emissions account for most of the respirable (<2.5 um)
particles in the air (55). We have recently used these methods to estimate the
contribution of various combustion sources to the airborne mutagenicity due to
these particles (56). Since a much larger data base is available on the Ames
Salmonella assay, this data base was used for the emission factors. Table VII
summarizes the findings from these studies. Using simply the information in
EPA's publication of the annual U.S. fuel use by category, we estimate that 44%
of the mutagenicity emitted into the air was derived from diesel vehicles, 22°/o
from gasoline vehicles, and 32% from residential heating. Dispersion modeling was
used to estimate-ambient concentrations in a theoretical city and in actual cities
where more detailed source inventories and meterological factors can be considered
in a more complex dispersion model. In the examples shown in Table VII, 33-51%
of the mutagenicity of these locations was estimated to arise from diesel vehicles.
Receptor modeling uses data from ambient concentrations of tracer signature
chemicals (e.g., lead for automobiles) to determine the contribution from various
sources. The data shown for Denver, Colorado was derived from a study to
17
-------
apportion the particles and organic; using receptor modeling (57). By applying
data on the mutagenicity of each of these source emissions, assuming no
atmospheric transformation of the mutagenicity, only 5% of the mutagenicity was
estimated to be derived from diesel trucks and 56% from gasoline vehicles.
Recently we have actually used mutagenicity as a parameter for apportionment in
a receptor modeling study in Albuquerque, Nc.w Mexico. Although this study was.
not designed tc separate diesel from gasoline vehicle emissions, 50% of the
mutagenicity of this air shed in wintertime was due to automotive emissions. In
future studies we plan to improve our ability to directly measure the contribution
of various automotive and residential heating sources to the airborne mutagenicity
and tumorigenicity.
The impact of automotive emissions on the total human exposure is generally less
than that estimated by considering only outdoor ambient air. Most individuals
spend over 80% of (heir time indoors at home or work. In a recent study of ten
homes, we found environmental tobacco smoke to be the principal source of indoor
particle-assocoated mutagenicity (59).
TABLE VII
CONTRIBUTION OF MOBILE SOURCES AND RESIDENTIAL HEATING TO THE AIRBORNE MUTAGEN-
ICITY ASSOCIATED WITH RESPIRA8LE PARTICLES
Source
Mobile Sources
Diesel
Gasoline
Emissions from
U.S. Annual
Fuel Use
44
22
Dispersion
Modeling
Theoretical Site A Si
% %
49
22
51
23
Ce B
%
38
17
Receptor
Model ina
Denver Albuquerque
% %
56
5
(TOTAL) (66) (71) (74) (55) (61) 50
Residential
Heating
32
24 25 43 39 50
W
18
-------
CONCLUSIONS ON THE VALUE OF SHORT-TERM GENETIC BIOASSAVS
In retrospect, it is now safe to conclude that short-term mutagenicity assays
were not only useful but instrumental in:
I) Indicating that diesel soot was potentially carcinogenic and should be
evaluated on chronic animal cancer liioassays.
2) Identifying NO,-PAHs as potentiil carcinogens in this very complex mixture.
3) Providing initial evidence that the mutagens were bioavailable.
4) Estimating the relative importance of various sources and fuels and other
factors which can influence human exposure to carcinogens.
This is not to say that short-term bioassays used alone can accomplish all of
this. However, used in combination with chemical/analytical methods and
toxicological tools, short-term genetic bioassays have become a critical component
of many environmental health studi-:.
Although substantial advances in our knowledge of the toxicology of diesel
emissions have been made since 1978 when the ini.'ial observation that the organics
extracted from diesel soot were mutagenic, a number of important questions
remain not only for diesel emissions but for other combustion sources *is well. Are
the chemicals which induce positive results in the short-term bioassays the same
agents which cause tumors in chronic animal bioassays? Which phase of the diesel
emissions (gaseous or paniculate) is carcinogenic in the animal inhalation studies?
With ad.vances in our understanding of the molecular mechanisms involved in
producing chronic effects such as cancer, it is possible (hat new genetic tools and
short-term bioassays will continue to contribute to our ability to answer these and
other questions as they arise.
ACKNOWLEDGMENTS
'The authors acknowledge the assistance if Dr. Larry Claxion and Ms. Jackie
Finley in the preparation of the manuscript The research described in this paper
has been reviewed by (he Health Effects Research Laboratory, U.S. Environmental
Protection Agency and approved for publication. Approval does not signify that
the contents necessarily reflect the views and policies of the Agency.
REFERENCES
1. Huisingh J Lewtas. Bradow R. Jungers R, Claxton L. Zweidinger R. Teiada S.
Bumgarner J, Duffield F. Simmon VF. Hare C. Rodriguez C, Snow L. Waters
MD (1978) In: Waters MD. Nesnow S. Huisingh J Lewtas. Sandhu SS. Claxton
L (eds) Application of Short-Term Bioassays in the Fractionati-in anil Analysis
of Complex Environmental Mixtures. Plenum Press, New York, pp 381—118
2. Toklwa H. Takeyoshi T, Takhashi K, Kachi K, Ohnishi Y (1978) Mutat Res
54:259-260
3. Wang YY. Rappaport SM, Sawyer RF, Talcolt RD, Wei EJ (1978) Cancer Lett
5:39-47
4. Siak JS. Chan JL, Lee PS (1981) Environ !nt 5:243-248
19
-------
5. Lofroth C (1981) Environ Int 5:255-262
6. OhnUhi Y. Kaciii K, Sato K, Tahara S, Takeyoslii II, Tokina H (1980) Mutal
7. Lewtas J. Bradow Rl.. .lungers RH. Harris BD, Zwcidinger RB, Cuslting KM.
GUI BE, Albert KE (1981) Environ Int 5:383-387 ** B
8. Lewtas JL (19811 In: Lewtas JL (ed) Toxicological Effects of Emissions from
Diesel Engines. Elsevier, New York, pp 243-264
9. Liber ML, Andoc BM, Hites RA, Thilly WG (1981) Environ (nt 5:281-184
10. Li AP, Royer RE (1982) Mutat Res 103:349-355
11. Lewtas J (1983) Environ Health Perspect 47:141-152
12. Choudhury DR, Doudnev CO (1981) Environ Int 5:389-392
13. Schuetzle D. Lee F.SC, Prater TJ, Teiada S (1981) In* J Environ Anal
Chcm 9:93-144
14. Xu XB. Nachtraan JP, Jin ZL, Wei ET, Rappaport SM (1982) Anal Chim
Actal36:163-I74
15. Nishioka MG. Petersen BA. Lewtas J (1982) In: Cooke M. Dennis AJ. Fisher
GL(eds) Polynuclear Aromatic Hvdrocarbons: Physical and Biological
Chemistry.Bauelle Press, Columbus. OH, pp 603-613
16. Pitts JR Jr. Lokensgard DM. Harger W. Fisher TS. Meiia V, Schuler
JJ, Scorziell CM, Katzenstein YA (1982) Mutat Res 103:241-249
17. King LC, Loud K, Tejada SB, Koban MJ, Lewtas J (1983) Environ Mutagen
5:577-588
18. Ball LM. King LC, Jackson MA, Lewtas J (1985) Environ Int
11:355-362
19. Ball LM. Kohan MJ, Inmon J, Claxton LD, Lewtas J (1984) Carcinogenesis
5:1557-1564 .
20. El-Bayoumy K, Kecht SS (1983) Cancer Res 43:3132-3137
21. King LC, Jackson M, Ball LM, Lewtas J (1983) Cancer Lett 19:241-246
22. Lewtas J (1936) Fundara Appl Toxicol (in press)
23. Albert RE. Lewtas J, Nesnow S, Thorslund T, Anderson E (1983) Risk
Anal 3:101-117 -
24. Lewtas J (1985) Toxicol Indust Hcllh 1:193-203
25. Pereira MA (1982) In: Lewtas J (ed) Toxicologicaf Effects of Emissions from
Dicse! Engines. Elsetier, New York, pp 265-276
26. International Agency for Research on Cancer (IARC) Monograph programme
on the evaluation of carcinogenic risk of chemicals to humans.
Polynuclear Aromatic Compounds. Part 4. Bitumens. Coal-tar and
Derived Products, Shale Oils and Soots, 1985, Vol. 35, 271 pp
27. McCormick JJ. Zator RM. DeGal BB. Mahcr VM (1980) In: Pepiko \VE.
Danner KM,Clarke NA (eds) Health Effects of Diesel Engine Emissions:
Proceedings of an International Symposium. Vol. 1. EPA-600/9-80-057a,
Cincinnati. OH, pp413-415
28. Waters MD. Stack HF, Mavournin KH, Dellarco VL (1986) Mutal Res
167:171-188
29. Garret! NE, Stack HF. Gross MR. Waters MD (1984) Mutat Res 134:89-111
30. Pepelko WE (1982) In: Lewtas J (ed) Toxicoloqical Effects of Emissions
from Diesel Engines. Elsevier, New \ork, pp 121-142
31. Heinrich U. Peters L. Funcke \V. Pott F. Mohr U. Stober W
0982) In: Letrias ,1 (ed) Toxicotocical Effects of Emissions from
Diesel Engines. Elsevier, New York, pp 225-242
32. Kaplan ML. Springer KJ, Mackenzie WF (1983) Studies of Potential Health
2Q
-------
Effects of Lone-Term Exposure to Diesel Exhaust Emissions
[final Report ,\o 01-0750-103 (SWKI) and No 1230 (SFKE)|
Southwest Research Institute, San Antonio, TX
33. McClellan RO (1983) Environ Health PerspecC 47:283-292
34. Siak JS. Chan TL. Lee PS (1980) In: Pepelko WE. Danner R.M. Clarke NA
(eds) Health Effects of Diesel Engine Emissions: Proceedings of an
International Symposium, Vol. 1. U.S. EPA, Ciacinnsti, OH,
EPA-600/9-80-057a, pp 2-15-262
35. Brooks AL. Wolff RK. Rover RE. Clark CR. Sanchez A. McClellan RO (1980)
In: Pepclko WE. Danner KM. Clarke NA (eds) Health Effects of Diesel Engine
Emissions: Proceedings of an International Symposium,
• Vol. 1. U.S. EPA, Cincinnati, OH, EPA-600/9-80-057a, pp 345-358
36. King LC. Kohan MJ. Austin AC. Claxton LD, Hnisingh J Lewtas (1981)
Environ Mulagen 3:109-122
37. Wang YY. Talcott RE, Seid DA, Wei EJ (1981) Cancer Lett 11:265-275
38. Chescheir CM III, Garrett NE, Shelburne JD. Huisingh J Lewtas. Waters MD
(1981) Jn: Waters MD. Sandhu SS. Huisingh J Lewtas. Claxton L, Nesnow S
teds) Short-Term Bioassays in the Analysis of Complex
•Mixtures II. Plenum Press, New York, pp 337-350
39. Sun JD. Wolff RK. Abcrman HM, McClellan RO (1983) Toxlco! Appt
Pharraacol 69:185-198
40. Sun JD. Wolff RK. Kanapilly CM, McClellan RO<1984) Toxicol
Appl Pharmacol 73:48-59
41. Bond JA. Medinsky MA. Sun JD (1986) Disposition and Metabolism of Free
f. and Particle-A-.sociated Nitropyrcnes After Inhalation, Research Report
No. 2. Lovelace Biomedical anil Environmental
Research Institute. Health Effects Institute, Boston, MA
•12. Swain AP, Cooper JE, Stedman RL (1969) Cancer Res 29:579-583
43. Claxton L, Huisingh J Lewtas (1980) In: Sanders DL. Cross FT, Dagle GE,
Maliatfev JA (eds) Pulmonary Toxicology of Kespirable Particles.
COND-791002, U.S. Department of Energy, Washington, DC,
pp 453-465
44. Rosenkrani HS, McCov EC. Sanders DR. Butler, Kiriazides DK,
Mermolstein R (1980) Science 209:1039-1043
45. Schuetzle D (1983) Environ Health Perspect 47:65-80
46. Lewtas J (1985) In: Slich HF (ed) Carcinogens and Mutagens in the
Environment, Volume V, The Workplace. C'KC Press, Uoca Katon, f L,
pp 59-74 ... .-••••
47. Austin AC, Claxlon LD, Lewtas J (1985) Environ Mutagen 7:471-487
48. Williams R. Sparacino C, Petersen U. Bumgarner J, Jungers RH, Lewtas J .
(in press) Int J Environ Anal Cliem
49. Rosenkranz KS, McCoy EC, Mermelstein R, Speck WT (1981) Mutat Res
91:103-150
50. Nishioka MG. Petersen B. Lewtas J (1983) In: Rondia D. Cooke M. Haroz RK
(eds) Mobile Source Emissions Including Polvcyclic Organic Species.
Reidel Press, Dordrecht, Holland, pp 197-210
SI. Pederson TC (1983) In: ibid pp 227-246
52. Mumford J, et al, manuscript submitted
53. Meselson M. Russell K. (1977) In: Hiatl HW, Watson JD. Winsten JA (eds)
Origins of Human Cancer Book C. Cold Spring Harbor Labs, New York
pp 1473-1481
54. Use of Short Term Genotoxic Bioassays in the Evaluation of
Automotive Emissions (1983) EPA Technical Kenorl sEPA-600/1
-84/011 Available from NT1S, 5285 Port Royal Road,
Springfield VA 22161 «
55. Wolff GT (1985) Environ Int 11:259-269
21
-------
56. Lewtas J, Alfheim I, Lofrolh G (in preparation)
57. Wolf CT. Countess RJ, GroMicki PJ. Ferman MA, Cadle SH.
Muhlbaier JL (19S1) Almos Environ 15:2.485-2502
58. Lewis C, Dautngardner R, Claxton LD, Lewtas J, Stevens RK
(in preparation;
59. Lewtas J. Goto S, Williams K. Chuang JC. Pedersen BA,
Wilson NK (in press) Atmos Environ
22
-------
------- |