microsome assay with the mouse dermal carcinogenesis bioassay for complex petroleum hydrocarbon mixtures

FUNDAMENTAL

AND

APPLIED

TOXICOLOGY

5,

382-390 (1985)

Lack of Concordance of the Salmonella/Microsome Assay with the Mouse Dermal Carcinogenesis Bioassay for Complex Petroleum Hydrocarbon Mixtures STEVEN T. CRACK&* C. CLIFFORD CONAWAY,~ AND JUDITH A. MACGRJSOR$ *American Petroleum institute, Washington, D.C. 20005; tTexaco Research Center, Beacon, New York 12508; and *Chevron Environmental Health Center, Richmond, California 94802 Lack of Concordance of the SalmoneI/a/Microsome Assay with the Mouse Dermal Carcinogenesis Bioassay for Complex Petroleum Hydrocarbon Mixtures. CRAGG, S. T., CONAWAY, C. C., and MACGREGOR, J. A. (1985). Fundam. Appl. Toxicol. 5, 382-390. Typical petroleum hydrocarbon mixtures were tested directly, without extraction, in the Salmonella/microsome mutagenesis assayin order to determine if the assaywould be useful to predict their carcinogenic activity. The carcinogenic activity of each sample had been previously characterized in the in vivo mouse dermal carcinogenesis bioassay. The series of samples evaluated offered several advantages. They spanned a wide boiling point range, were well characterized chemically, had been tested for carcinogenic activity in a single laboratory, and varied in potency in vivo from inactive to highly active. Mutagenicity testing was performed in several welLestablished contract laboratories that routinely perform the assay. These laboratories were the main contracting laboratories for these assays at the time and had previously tested petroleum samples for clients. Initially, the 61% laboratory tested 13 samples in five strains of Salmonella typhimurium with and without rat liver S-9 (Arochlor 1254 induced), utilizing both plate and suspension techniques. None of the 13 samples exhibited a mutagenic response, even though 9 of the 13 were slightly to highly dermally carcinogenic in mice. Because of the unexpected results, it was decided to repeat the mutagenicity assaysin two other laboratories. Six of the thirteen samples were selected, ranging in carcinogenic potency from negative to highly active. Again, none were mutagenic in the second contract laboratory. In a third facility, only one sample of the six exhibited a definite mutagenic response. However, the response was observed with a sample having only weak carcinogenic activity and, unusual for petroleum hydrocarbons, occurred without activation. When this sample was recoded and retested in the same laboratory, negative results were obtained. The results from this multiple laboratory study clearly indicate that the methods employed routinely to perform the Salmonella/microsome assay are not useful to predict the dermal carcinogenic activity of typical complex petroleum mixtures. Various explanations for these results and some further experimental approaches are discussed. Q 1985 Society of Toxicology.

Attempts are presently being made in the “Gene-Tox Program” to assessthe ability of the Salmonella/microsome test to predict the carcinogenic activity of chemicals (Green et al., 1980). Extensive efforts have also been made in the past (McCann et al., 1975; Andrews et al., 1978; Ames et al., 1981). However, most of this validative research has been concerned with single, pure compounds. Far less has been done to evaluate this shortterm procedure with regard to complex mix0272X690/85

$3.00

CopyTi%t Q 1985 by the Society of Toxicology. All rights of repmduclion in any form resewed.

tures. In the literature which does exist, almost never have mixtures been employed which have been previously evaluated for carcinogenic activity. Lack of this crucial element has precluded a fair assessment of the ability of the Salmonella test to detect carcinogenic activity for complex hydrocarbon mixtures. To further complicate matters, rather thau test complex mixture& directly in the Salmonella assay, most of the work performed 382

MUTAGEN/CARCINOGEN

NONCORRELATION:

to date has been oriented toward basic research, not applied screening. As such, it has entailed extraction of complex mixtures with different solvents, followed by testing of the extracts rather than the material itself. This approach has often been coupled with elaborate chromatographic or distillation separations and testing of the resulting multiple fractions. In a screening mode, however, such an approach totally defeats the utility of the assay. Moreover, a strong case can be made to avoid fractionating if possible, because when parts of the material are evaluated separately, their summed activity need not reflect that of the whole (Pelroy, 1980). In addition, Epler et al. (1978, 1980), have noted that complex separation schemes may well create or inhibit mutagenesis. The purpose of this research effort was to assess the ability of the Salmonella/microsome test, as conducted in contract laboratories, to reflect the carcinogenic activity of petroleum hydrocarbon mixtures in order to validate the use of this test as a predictive tool. To accomplish this, typical petroleum hydrocarbon mixtures were tested by the Ames procedure in three separate contract laboratories using standard methodology (Ames et al., 1975). Mixtures were not sep arated by solvent extraction or other means. In addition, the carcinogenic activity of all samples had been previously characterized in the mouse dermal carcinogenesis bioassay. Test materials were coded before submission to each laboratory so that all testing was performed under blind conditions. The lab oratories were given no guidance initially in testing the materials as part of the goal was to evaluate the state-of-the-art in large contract laboratories. Subsequently, one of the laboratories was instructed to follow certain modifications regarding sample handling techniques. MATERIALS

AND

METHODS

Petroleum mixtures. Dktillation fkactions were derived from two crude oils of divergent chemical composition.

MIXTURE8

383

The first parent oil was a low sulfur, U.S. Gulf Coast crude of predominantly naphthenic composition which is designated as crude “c’ while the second was a paraffinic, hi sulfur foreign crude, “D”. Both oils were fractionated into typical refinery streams corresponding to light ends or petroleum ether (D-l), naphthas or gasoline components (C-2 and D-2), kerosene (C-3 and D-3), gas oil (C-4 and D-4), heavy oils (C-5 and D-5) and residua (C-6 and D6). As crude “C” contained no light ends, the 13 samples included 11 distillate fractions and the 2 parent oils. The lowest boiling fractions contain hundreds of compounds while literally tens of thousands comprise the high boiling samples. Samples have proven stable as evidenced by similar tumotigenic potencies in repeat testing 4 years after initial bioassay. Fractionation procedure. Crudes were fractionally distilled into boiling ranges which correspond to the refinery streams listed above. Up to a temperature of 371°C (700°F), crudcs were separated in a Podbielniak Series 400 fractionater unit. The adjacent residue boiling above 371°C was flash-evaporated to 577’C in a vacuum equilibration flash vaporizer (Artisan Metals, Inc.), to obtain the 371-566“C fraction as well as the residue boiling above this temperature. A more complete de scription is available in Eppolito (1977). PAH analyses. Samples were subjected to a series of chromatographic procedures designed to separate polynuclear aromatic hydrocarbons (PAH) from crude oils (Hoe1 et al., 1979). The first step consisted of serial column chromatography using increasingly activated Woelm alumina. After this, concentrated eluents were streaked on cellulose acetate TLC plates with PAH reference compounds spotted alongside. Following TLC, samples were subjected to reverse phase HPLC. PAH were identified by comparison to standard UV spectra and quantified by correcting for recovery of radiolabeled benzo[a]pyrcne (BaP) which was added to samples as a tracer prior to analysis. Further details of this procedure arc available in Hoe1 et al. (1979). Salmonella assay. Two contract laboratories (X and Z) employed all five strains of Salmonella typhimurium obtained from Dr. Bruce Ames of the University of California-Berkeley. These strains were: TA-1535, TA1537, TA-1538, TA-98, and TA-100. Laboratory Y used four Salmonella strains, omitting TA- 1538. All petroleum mixtures in each laboratory were tested with and without rat 8-9 (Arochlor 1254induced) supematant containing liver microsomes. Each laboratory employed both suspension and plate assay procedures a&r the method of Ames et al. (1975). At least two replicate plates were evaluated for each sample and were tested over a dose range of two to three orders of magnitude using at least five dose levels. Where possible, high dose levels were included which exhibited some toxicity. Criteria for a positive test varied among the laboratories; however, all required a repeatable dos+response relationship, with the highest dose considered registering at least three

384

CRAGG, CONAWAY,

times solvent control background. Furthermore, negative and positive control revertant values were required to fall within prespecified ranges for an experiment to be considered valid, while pattern agreement among strains was considered important but not essential. In all three laboratories, positive controls consisted of 9aminoacridine, 2-nitrofluorene, 2-anthramine, and BaP which were used with selected strains and conditions of metabolic activation. Negative solvent controls were run with and without activation. Sterility checks and evaluation for toxicity were conducted by all three laboratories. Demal carcinogenesis bioassay. Each of the 13 samples was administered to a group of 50 male C3H/HeJ mice, beginning at 8 weeks of age, for 18 months at Kettering Laboratory, University of Cincinnati. During this treatment period, 50 mg of undiluted sample was dropped onto the clipped, interscapular region of each mouse, twice per week. After the treatment period, mice were observed for the remainder of their lives. The average latency period and the percentage developing tumors were recorded for each group. All tumors were diagnosed microscopically. A clipped-only group and a toluenetreated group constituted the negative controls while positive control mice received 0.05 and 0.15% BaP solutions in toluene.

AND MAC GREGGR

RESULTS Mutagenesis results from all three laboratories are summarized in Table 1. Data in this table reflect the highest revertant counts obtained for a sample (tested in at least two separate assays in each laboratory) regardless of test conditions (including bacterial strain, solvent used, dose, presence/absence of S-9, etc.). Numerical values are expressed as the fold-increase over solvent controls (i.e., the maximum revertant counts found, under any conditions, divided by the solvent control count). As can be seen in Table 1, the majority of tests yielded reversion rates which never exceeded a threefold increase over background under any circumstances in repeat testing. In only 4 of 27 instances did revertant counts equal or exceed the spontaneous background rate by threefold, and even here, repeat testing

TABLE 1 SALMONELLA

MUTAGENESISREWLTSWITH

PETROLEUM

FRACTIONS

Ames test results” Crude sample

Distillation range (“C)

LabX

Cb c-2 c-3 C-4 c-5 C-6 Db D-l D-2 D-3 D-4 D-5 D-6

OPC->577 OP-177 177-288 288-371 371-577 >577 OP->577 OP-49 49-177 177-288 288-371 371-577 >577

<3.0 <3.0 <3.0 <3.0 4.0d <3.0 <3.0 <3.0 <3.0 <3.0 13.0 <3.0 <3.0

LabY

Lab Z( 1)

Lab Z(2)

<3.0 <3.0

38.2 13.0

13.0 <3.0

<3.0

<3.0

13.0 <3.0 3.1d

3.0d <3.0 <3.0

‘Values reported under X, Y, and Z represent the maximum obtained increase over background. Values are expressed as the average revertants per plate, divided by that of the solvent control. The experiment, dose, strain, and presence/absence of activation reported in this table are those which gave the highest increase over background in a single assayfor the indicated test material and am not necessarily comparable among samples. b Represents entire unfractionated crude oil. c OP = Overpoint: Similar to initial boiling point. d Not reproducible when retested in the same laboratory in a second assay(i.e., value did not equal or exceed three times background when retested-sample was not mutagenic).

MUTAGEN/CARCINOGEN

NONCORRELATION:

in the same laboratory rendered three of the four values as spurious. Thus, all three contract laboratories concluded that none of the samples (except C-2) were mutagenic when tested under the prescribed conditions. Sample C-2 produced up to a 38-fold increase over background in Laboratory Z. It exhibited a strong dose-response relationship over five dose levels ranging over two orders of magnitude (0.33 to 33 ~1 of a 33% solution of the sample in DMSO). This positive response occurred only in strain TA1537 without S-9 activation (Table 2). After Laboratory Z reported this positive, sample C-2 was recoded and returned to this same laboratory for blind reevaluation. Another recoded sample, C-3, was also resubmitted. This time, Laboratory Z concluded that both samples were nonmutagenic. However, a different solvent and concentration range were used for sample C-2 in the repeat

385

MIXTURES

bioassay (Table 2). As mentioned previously, the laboratory was purposely not directed in the choice of solvent or doses as the goal of this research was to evaluate the correlation of the Ames test with carcinogenesis, for complex mixtures routinely tested in commerical laboratories.

Mutagenicity Testing Proceduresand Results: Laboratory X Because more extensive and varied procedures were performed on the samples in Laboratory X, results from this facility are reported independently. First, a qualitative spot test was performed employing strain TA- 100 with nine undiluted samples (excluding those which distill above 371°C) to determine toxicity. When 20 mg of sample was tested with and without metabolic activation, no toxicity or mutagenicity was observed.

TABLE 2 SAMPLE C-2: RESULTS OF ARRAYSIN LABORATORY X, Y, AND Z USING STRAIN TA- 1537 WITHOUT S9 ACTIVATION

Sample Solvent control 2-Aminoanthracene c-2

’ Ethanol solvent * Acetone solvent. c DMSO solvent. d Acetone solvent.

Micrograms of material per plate

Revertants per plate Lab X” 7

50 100 1 3 10 33 50 100 330 500 Loo0 3,330 5,ooo 10,ooo 33,000 100.000

989

Lab Y* 11 395

Lab z(l)=

Lab Z(2)d

10

2’

666

614 4 4 2 5 6

12 18 6 70 210 382

386

CRAGG,

CONAWAY,

AND

In the standard assay, only sample C-5 gave revertant counts which equaled or exceeded three times background. However, the laboratory concluded that this sample was not mutagenic when, upon retest, results were negative. In order to achieve optimal solubilization, Laboratory X was directed to determine the extent of dissolution of the petroleum hydrocarbon samples in various solvent vehicles. Table 3 summarizes the solvents finally selected for each sample, as well ‘as dose ranges tested, occurrence of toxicity, and number of experiments per sample. It must be noted that suspensions rather than actual solutions were obtained and tested. Test suspensions ranged in concentration from 15000 to 1: IO, sample to solvent. In previous experiments, Laboratory X had determined that the mutagenic activity of BaP in mineral oil could be markedly enhanced if 50 ml of DMSO were added to the incubation medium immediately follow-

Crude sample C

c-2 c-3 C-4 c-5 C-6 D D-I D-2 D-3 D-4 D-5 D-6

3

TESTING

IN LAJWRATORY

X

Toxicity

No. of EXptS.

MUtap. a&.

FJaP mutagenesis ah-x spiking’

Effect of mixing and sonic&ion c

104,ooo

None

1

None

Sl decrease

No effect

5,000-10,000 1040,000 IO-lO,OOO IO-10,000

None None None None

I 2 3 2

None None None None

SIdecrease LgS1 decrease

None ~2 mg/plate >5 mp/plate None 21 mg/plate None None None z-500 &me

2 2 2 2 2 2 2 2 2

None None None None None None None None None

solvent Ethanol and sonication llenzeneb Ethanol Ethanol Ethanol Acetone and sonication Benzene Benzene Ethanol Ethanol Ethanol Ethanol Benzene Benzene

GREGOR

ing addition of the sample. Thus, all petroleum hydrocarbon samples were tested in two separate experiments; one with and one without postaddition of DMSO. The addition of DMSO did not alter revextant response levels; therefore, results were reported as the average of the two experiments. When no consistent mutagenic response was detected for any of the samples, Laboratory X added BaP to all of the test materials and retested them to determine whether some constituent within the petroleum samples could be inhibiting mutagenic activity. Table 3 shows that several samples significantly inhibited the mutagenic activity of BaP. Finally, six samples were mixed overnight with DMSO and sonicated for 1 hr prior to testing and the resulting suspensions were tested in strain TA-98 with and without S-9 activation. Despite the fact that four of the six samples were carcinogenic, none of the six suspensions produced any mutagenic activity, as can be seen in Table 3.

TABLE RESULTS OF MUTAGENFBIS

MAC

IO-10,000 lo-10,000 IO-10,000 lo-10,000 lo-5,ooo IO-10,000 1040,000 IO-10,000 lo-5,000

No effect 4dSl decrease No effect SI decrease S1 decrease

No eEect No effect No e&t

iLI$zz Lg-

No effect No e&a

’ AU assays performed with co-addition of 2 w BaP/plate; with and without the addition of 50 d of DMSO; and with s-9 only. Assays were perfomed in duplicate (with and without “p&addition” of DMSO) with duplicate platea per assay. b Benzene was used at concentrations above 100 mg/ml. cSampl~ WI?. mixed overnight with DMSO and aonicated 1 hr prior to incubation.

MUTAGEN/CARCINOGEN

NONCORRELATION:

387

MIXTURES

Dermal Carcinogenic Activity of the Test single carcinoma and papilloma, respectively. Samples The remaining nine samples produced suffiThe dermal carcinogenic activity of the 13 samples is reported in Table 4, as are the parts per million total PAH, and BaP, for several of the samples. The two fractions boiling in the 371-577°F range (D-5 and C-5) exhibited the most tumorigenicity, with latency intervals of 34 weeks for D-5 and 50 weeks for C-5. The percentage of animals developing tumors was also highest for these two fractions (80 to 90%). In addition, the tumors were largely carcinomas rather than papillomas. The remaining tumorigenic samples were less active, showing latency intervals in the range of 60 to 90 weeks and producing tumors in approximately 20 to 55% of the mice with a lower ratio of malignant to benign tumors. Two samples, D-l and C-6, produced no tumors, while D-4 and D-6 produced only a

cient tumor yields to conclude that they were carcinogenic, ranging over the full spectrum of activity. DISCUSSION

AND

CONCLUSIONS

The lack of mutagenic activity shown by the carcinogenic samples were obtained under valid test conditions. Specilically, no appreciable toxicity occurred, adequate replicates were used (entire experiments were even repeated, sometimes several times in each of the three laboratories), and positive control compounds yielded appropriate responses. For reasons already given, neither organic solvent extracts nor separated fractions of the test materials were tested, but rather samples were evaluated directly in the Salmonella assay with and without DMSO as a solubilizer/“extractant”, as is routinely the case in

TABLE 4 CARCINOGENIC

Cl-U& sample No treatment Toluene C c-2 c-3 C-4” c-5 C-6 D Dlb D-2 D-3 D-4 D-5 D-6 0.05% BaP 0.15% BaP

Distillation range (“C)

OP->577 OP-177 177-288 288-37 1 371-577 >577 OP->577 OP-49 49-177 177-288 288-37 1 37 l-577 >577

AGILITY

OF PETROLEUM

FRACTIONS

Average latency w=ks)

96 Mice with tumor

I-k./ benign

91 69 85 70 85 50 64 85 62 4oc 34 7oc 46 29

0 2 30 21 30 34 81 0 56 0 25 15 3 91 2 74 97

O/l 2.3 0.3 0.8 1.6 2.9 2.2 4.5 1.0 l/O 9.3 O/l 2.1 6.2

Total” PAH (mm)

1.2 lO-* 48 137

0.1 6.5 2.8 1o-4

1.7 62


’ Reported where available. Total PAH include 19 PAHs representing several class of polynuclear aromatics. b Tested at half dose (25 mg/lX/week) due to toxicity at full dose. c Not a reliable value due to very low tumor yield.

388

CRAGG, CONAWAY,

AND MAC GREGOR

contract laboratories. Other researchers, in nonscreening, basic research-oriented investigations, have obtained positive responses with this mutagenesis assay when testing organic solvent extracts and other separated fractions of complex hydrocarbon mixtures containing high levels of PAH (Pelroy and Peterson, 1978; Petrilli et al., 1980). Regarding the carcinogenicity results, activity exhibited by the lower boiling fractions of both crudes points to the existence of carcinogens other than PAHs typical in higher boiling hydrocarbon mixtures; Such hypothetical constituents might act in a largely promotional capacity which could account for the lack of mutagenicity exhibited by the lower boiling petroleum fractions. However, as discussed below, other hypotheses appear more attractive at this point. Table 5 summarizes and compares the results of the Salmone&.z/microsome test with those of the chronic dermal carcinogenesis bioassay. The two most tumorigenic samples (C-5 and D-5) also contained the highest levels of PAH. As such, they might have been expected to evoke a positive response in the Salmonella assay but did not. It is also

noteworthy that the lone carcinogenic sample which did yield a positive response contained extremely low concentrations of polycyclic aromatic hydrocarbons (e.g., RaP < 0.1 ppb, Table 4) and was only slightly to moderately tumorigenic. Further, the mutagenic response occurred only in strain TA-1537 and did not require metabolic activation, certainly an unusual response for a petroleum sample. The foregoing suggests several conclusions. First, the Salmonellalmicrosome test, as conducted in three large contract facilities with commonly utilized methodology, does not reflect the carcinogenic activity of the complex petroleum hydrocarbon mixtures. This standard assay did not demonstrate activity even for high boiling samples which contain substantial concentrations of carcinogenic polycyclic aromatic hydrocarbons and which are highly active tumorigens when applied chronically to the skin of mice. Second, addition of DMSO which could be expected to solubilize putative mutagens, did not enhance detection of carcinogenic samples in the Salmonella assay. Third, extensive mixing, even when coupled with sonication, did not improve the response. Fourth,

TABLE 5 CORRELATION

OF

SALMONELLA TEST

RJWLTS

TO DERMAL

CARCIN~GENECITY

Ames test results Crude sample

Distillation range (“C)

careinogenecity

C c-2 c-3 C-4 c-5 C-6 D D-l D-2 D-3 D-4 ‘D-5 D-6

opa->577 OP-177 177-288 288-37 I 371-577 >577 OP->577 OP-49 49-177 177-288 288-37 1 371-577 >577

++ + ++ ++ +++ ++ + + +++ -

0 OP = Overpoint: Similar to initial boiling point.

IabX

-

LabY

LabZl

-

++t -

-

-

-

-

-

-

-

-

Lab 22

MUTAGEN/CARCINOGEN

NONCORRELATION:

toxicity did not appear to account for the poor mutagenic response of these mixtures since the lawns in the sample dishes appeared normal at most of the doses tested and the samples were evaluated for toxicity prior to selection of doses in all three laboratories. Fifth, because the Salmonella assay readily detects carcinogenic PAH when tested in pure form, and since it is widely accepted that the carcinogenic activity of the higher boiling petroleum hydrocarbons is mainly attributable to these PAHs, it is likely that a technical problem such as sample delivery accounts for the largely negative response in the Ames assay. It is possible that initiators within complex petroleum hydrocarbon mixtures are not available to the test organism. Such nonavailability may be due simply to the low aqueous solubility of the test materials or may be caused by some sort of selective binding or sequestering which is occurring within the sample. For complex mixtures then, the poor correlation of the conventional Salmonella test with the mouse dermal carcinogenesis bioassay could perhaps be improved through technical modifications. Avenues of research which might be explored include (1) testing broader dose ranges, (2) enhancing delivery of lipophilic substances to the cell by solvent selection, emulsification via nontoxic detergent-type molecules, or perhaps using vehicles such as liposomes, and (3) developing a simple, yet effective, extraction method which could reflect carcinogenic activity. If the first two approaches fail, true extraction schemes (not simply solubilization-type “extractions” as performed in this research) should be examined in a systematic manner. If such procedures are pursued, however, it would be important to bear in mind the potential problems referred to in the introduction, Particularly, the utility of the test for screening purposes could quickly be circumvented if an extraction scheme of any complexity were required, since convenience, speed, and economy are so vital for a screening-type bioassay.

MIXTURES

389

There is a strong possibility that the Salmonella test does not respond to carcinogenic mixtures because of the methodological difficulties outlined above. Additional research with the Salmonella/microsome assay has indicated that increased vortexing and ultrasonication, coupled with preincubation, slightly increases mutagenicity for some of the samples tested. Recent studies with samples C-5 and D-5 have shown a consistent three- to fourfold increase in TA-100 revertant colonies when $9 was increased 5 to 10 times that of the standard assay (MacGregor, Ring, and Carver, in preparation). Through further research, it is hoped that similar, relatively simple modifications may be found which will enable the Salmonella test, and other in vitro procedures, to reliably detect carcinogenic hydrocarbon mixtures. REFERENCES AMES, B. N. (1981). Correspondence re: S. J. Rinkus and M. S. Legator. Chemical characterization of 465 known or suspected carcinogens and their correlation with mutagenic activity in the salmonella typhimurium system., Cancer Res., 39: 3289-3318, (1979). Cancer Res. 41, 4192-4203. AMES, B. N., MCCANN, J., AND YAMASAKE, E. (1975). Methods for detecting carcinogens and mutagens with the Salmonella/mammalian microsomal mutagenicity test. Mutat. Res. 31, 347-364. ANDREW& A. W., THIBAULT, L. H., AND LUINSKY, W. (1978). The relationship between carcinogenicity and mutagenicity of some polynuclear hydrocarbons. Mutat. Res. 57, 31 l-326. EPLER, J. L. (1980). The use of short-term teats in the isolation and identification of chemical mutagens in complex mixtures. In Chemical Mutagenesis: Principles and Methods for their Detection (F. J. DeSerres and A. Hollander, Eds.), Vol. 6, Chap. 9, pp. 239-270. Plenum, New York/London EPLER, J. L., CLARK, B. R., Ho, C., GUERIN, M. R., AND RAO, T. K. ( 1978). Short-term bioassay of complex organic mixtures Part II. Mutagenicity testing. In Application of Short-Term Bioassay in the Fractionation and Analysis of Complex Environmental Mixtures: Environmental Science Research (M. D. Walters, S. Nesnow, J. L. Huising, S. S. Sandhu, and L. Claxton, eds.), Vol. 15, pp. 269-289 Plenum, New York/ London.

390

CRAGG, CONAWAY,

EP~OL~TO,J. (1977). Carcinogenic Potential of Petroleum Hydrocarbons: Preparation and Analysis of Fractions. American Petroleum Institute Medical Research Report. GREEN, S., AND AULETTA, A. (1980). Editorial introduction to the reports of “The Gene-Tox Program”; an evaluation of bioassaysin genetic toxicology. Mutat. Res. 76, 165-168. HOEL, D. C., AND PICKING, W. C. (1979). Carcinogenic Potential of Petroleum Hydrocarbons: Polynuclear Aromatics in Petroleum Fractions. American Petroleum Institute Medical Research Report. KING, R. W. (1983). Skin carcinogenic potential of petroleum hydrocarbons. 1. Separation and characterization of fractions for bioassay. In The Toxicology of Petroleum Hydrocarbons (H. N. MacFarland, C. E. Holdsworth, J. A. MacGregor, and M. L. Kane, eds.),

AND MAC GREGDR pp. 170- 184. American Petroleum Institute, Washington, D.C. MCCANN, J., CHOI, E., YAMASAKE, E., AND AMES, B. N. (1975). Detection of carcinogens as mutagens in the Salmonella/microsome test: Assay of 300 chemicals. Proc. Natl. Acad. Sci. (USA) 72, 5 135-5139. PELROY, R. A., AND PETERSON, M. R. ( 1978). Mutagenicity of shale components. In Application of ShortTerm Bioassays in the Fractionation and Analysis of Complex Environmental Mixtures: Environmental Science Research (M. D. Walters, S. Nesnow, J. L. Huisingh, S. S. Sandhu, and L. Claxton, eds.), Vol. 15, pp. 463-475. Plenum New York/London. PETRILLI, F. L., DERENZI, G. P., AND DEFLORA, S. (1980). Interaction between polycyclic aromatic hydrocarbons, crude oil and oil dispersants in the Salmonella mutagenesis assay.Carcinogenesis 1, 5 l-56.