Cholesterol Epoxide Is a Direct-Acting Mutagen (Lipid Oxidation/Mutagenesis/Carcinogenesis) A

Proc. Nati. Acad. Sci. USA Vol. 81, pp. 4198-4202, July 1984 Medical Sciences

Cholesterol epoxide is a direct-acting mutagen (lipid oxidation/mutagenesis/carcinogenesis) A. SEVANIAN AND A. R. PETERSON Institute for Toxicology, Department of Pathology, and the Cancer Center, University of Southern California, Los Angeles, CA 90033 Communicated by Bruce N. Ames, March 15, 1984*

ABSTRACT A 24-hr treatment of V79 Chinese hamster impair lipogenesis (21), as this compound has been reported lung fibroblasts with 12.4 ,uM cholesterol 5a,6a-epoxide in- to be cytotoxic (14, 22). duced 8-azaguanine-resistant mutants at frequencies 4.6- to In the present study, we have measured Chol epoxide up- 11.8-fold higher than the spontaneous mutation rate. We show take and apparent metabolism as well as its genotoxicity in that cholesterol epoxide, which is produced by in vivo choles- parallel experiments with V79 Chinese hamster cells. Using terol oxidation, is a weak direct-acting mutagen. Cholesterol established procedures (23, 24), we found that Chol epoxide epoxide was found to be accumulated by cells and transformed is mutagenic in Chinese hamster V79 cells, and we describe to cholestane-3.8,5x,6.8-triol, which was more toxic and a relationships between this mutagenicity and cytotoxicity, in- more potent inhibitor of DNA synthesis than the epoxide but, hibition of DNA synthesis, and transformation of Chol epox- at concentrations less than 17.8 ,.M, was not significantly mu- ide to cholestane triol. tagenic. Consideration of the rates of cholesterol epoxide conversion to cholestane triol shows that this conversion can result in abolition of the mutagenicity of the epoxide. Conditions un- MATERIALS AND METHODS der which conversion of the epoxide to the triol is low, as in the Chemicals, Radiochemicals, and Cell Culture Supplies. case of low epoxide hydrolase activity, favor mutagenicity Chol, Chol epoxide, and cholestane triol were purchased whereas rapid conversion to triol favors cytotoxicity. from Steraloids (Wilton, NH). Chol epoxide was found to be contaminated with =8% Chol and -2% cholestane triol as Cholesterol 5a,6a-epoxide, a recognized oxidation product determined by thin-layer and gas chromatographic analysis of cholesterol (Chol) (1-4), may also be a metabolic interme- (25). We therefore used purified Chol epoxide and choles- diate in bile acid biosynthesis (2, 5) and has been found to tane triol for all experiments. Both compounds were purified accumulate in such cases as hypercholesterolemia in humans by silicic acid column chromatography using benzene/ethyl (6). The enzymatic oxidation of Chol is coupled with hydra- acetate (3:2) as the eluting solvent. Alternatively, Chol epox- tion reactions that form cholestane-3f3,5a,6f3-triol (7). ide and radiolabeled Chol epoxide and cholestane triol were (Throughout this paper, the terms Chol epoxide and choles- prepared by reaction of [4-1 C]Chol (New England Nuclear), tane triol will refer to the Sa,6a and 3,1,5a,613 isomers, re- or unlabeled Chol (Sigma), with monoperphthalic acid as de- spectively, unless otherwise specified.) Since Chol epoxide scribed (25). The purity and characteristics of labeled and possesses an electrophilic oxirane group, it could be expect- unlabeled Chol epoxide were checked by gas chromatogra- ed to be genotoxic and carcinogenic (8, 9). The detection of phy/mass spectrometry using a Hewlett-Packard model Chol epoxide in UV-induced skin cancer (9) and cholestane 5992 GC/MS fitted with a 6 ft (1.86 m) 3% OV-101 packed triol in human colon cancer (10) suggests that the etiology of column and operated isothermally at 270°C. Further charac- these cancers may be associated with Chol epoxide, which is terization was accomplished by thin-layer radiochromatog- not xenobiotic but appears to originate in vivo or be assimi- raphy (26) and by HPLC using a Perkin-Elmer series 4 liquid lated as a food contaminant. Petrakis et al. (11) have recently chromatograph equipped with a 150 x 4.6 mm 3-,m Spheri- reported that human breast fluid can contain enormous lev- sorb silica column (Chromanetics, Jessup, MD). The eluting els (up to 780 ,M) of Chol epoxide. The origin of this Chol solvent for HPLC was 4.2% isopropanol in hexane at a flow epoxide is unknown although its levels are directly related to rate of 1.0 ml/min. Detection was accomplished by differen- breast fluid cholesterol content, which increases progres- tial refractometry. The 5a,6a- and 5/3,6,f3epoxides were elut- sively with advancing age. ed in 7.5 and 8.4 min, respectively (k = 2.75 and 3.20), which Studies have shown that Chol epoxide induces chromo- permits further purification of Chol 5a,6a- and 5,8,6,B3epox- somal damage, DNA repair (12), and oncogenic transforma- ides. In these experiments, further purification of the Chol tion (13, 14) of mammalian cells. Conversely, negative find- epoxide was not attempted and the substance was used at ings were obtained with Chol epoxide in the Ames Salmonel- the 95% purity level described below. A portion of the Chol la mutagenicity assay (15, 16), and Chol epoxide did not epoxide was converted to cholestane triol by reaction for 24 induce tumors nor promote N-methyl-N'-nitro-N-nitroso- hr at room temperature with 5% perchloric acid in tetrahy- guanidine carcinogenesis in rats and mice (15, 17, 18), al- drofuran/water/acetone, 4:1:0.5 (vol/vol/vol). Reactants and though careful determinations of Chol epoxide uptake and products were separated chromatographically after neutral- transformation were not made in several of these studies. ization of the reaction mixture with 0.1 M KOH (26). Furthermore, the proportions of the Sa,6a isomer and its di- The sources of N-methyl-N'-nitro-N-nitrosoguanidine and astereomer, Chol 5P,6p-epoxide, both of which may be [6-3H]thymidine and of the V79 Chinese hamster lung fibro- formed enzymatically or by peroxidation (1, 2, 4) were not blasts and the media and plasticware used in their culture indicated. Nonabsorption (19) or low levels of uptake and have been described (24). All reagents were of spectrophoto- metabolism of Chol epoxide (20) would be expected to affect metric grade. its genotoxicity, while conversion to cholestane triol could Abbreviations: Chol, cholesterol; zsGr, 8-azaguanine resistant. The publication costs of this article were defrayed in part by page charge *Communication of this paper was initiated by Charles Heidelberger payment. This article must therefore be hereby marked "advertisement" and, after his death (January 18, 1983), completed by Bruce N. in accordance with 18 U.S.C. §1734 solely to indicate this fact. Ames. 4198 Downloaded by guest on September 26, 2021 Medical Sciences: Sevanian and Peterson Proc. NatL. Acad. Sci. USA 81 (1984) 4199

Treatment of Cells for DNA Synthesis, Cytotoxicity, and purity of [14C]Chol epoxide was substantiated by HPLC. Mutagenesis Assays. For a single measurement of cytotoxic- The results of these analyses suggest that the biological ef- ity, 200 cells in each of four 60-mm dishes, or for a single fects measured are probably attributed to the a-isomer. measurement of DNA synthesis, 105 cells in each of two 100- The highest concentrations of Chol epoxide and Chol used mm dishes, were treated as follows. Cells were incubated in in our experiments were 62 ,bM and 25 ,tM, respectively. medium containing 0.5% acetone, which was used as the de- Concentrations greater than these produced precipitates and livery solvent and served as the negative control. N-methyl- therefore were not used. The highest concentration of cho- N'-nitro-N-nitrosoguanidine (positive control) at 6.8 /LM, lestane triol used, 24.8 .uM, caused cells to detach from the Chol, Chol epoxide, and cholestane triol were added to Dul- dishes; therefore, concentrations below 24.8 uM were used becco's medium containing 5% dialyzed fetal calf serum and to study its effects on V79 cells. 5% dialyzed calf serum at 370C. Cells were then incubated The time course of Chol epoxide (62 ,tM) incorporation for intervals of 2-24 hr. The 60-mm and 100-mm dishes con- over a 24-hr interval by cultures of V79 cells in which Chol tained 5 ml and 10 ml, respectively, of medium, which was epoxide produced significant mutagenesis is shown in Fig. aspirated after incubation. The monolayers were then 1A. It is assumed that any of the compound not washed from washed twice with complete medium and incubated with the monolayer was assimilated into a cellular compartment. fresh complete medium. A linear increase was found over the first 8 hr of incubation Assay of DNA Synthesis, Cytotoxicity, and Mutagenesis. and thereafter the rate of uptake decreased. Also shown in These assay procedures have been characterized (23, 24, the figure is the total amount of cholestane triol recovered 27). DNA synthesis was assayed by measuring [3H]thymi- from the cells and medium combined. By 24 hr the amount of dine incorporation into the cold-trichloroacetic acid-insolu- cholestane triol recovered from the medium represented ble fraction of cells that had been incubated in medium for 45 -35% of the total formed. Incubation of Chol epoxide in cul- min and then in medium containing [3H]thymidine (5.0 ture medium alone revealed very low levels of hydrolysis kCi/ml) for 30 min (23). Cytotoxicity was assayed by the (0.15 nmol/24 hr), and the cholestane triol produced nonen- plating-efficiency method (24). The frequencies of 8-aza- zymatically was subtracted to give the corrected enzymati- guanine-resistant (zsGr) mutants were determined using an cally dependent rate. After 24 hr of treatment, -5% of the expression time of 6 days with replating as described (24). epoxide was converted to cholestane triol. The concentra- Concentrations of cholestane triol or Chol epoxide that in- tion dependent uptake of Chol epoxide is shown in Fig. 1B. hibited DNA synthesis by -40% without producing precipi- Using a 24-hr incubation interval, a dose-dependent increase tates in the dishes were determined from single measure- in Chol epoxide uptake was observed. The formation of cho- ments of DNA synthesis. Using three different concentra- lestane triol appears to be time dependent over the concen- tions in this range, at least two measurements of DNA tration range of Chol epoxide tested. Approximately 5-11% synthesis, cytotoxicity, and mutagenesis were made on dif- of the Chol epoxide taken up was converted to cholestane ferent occasions using different batches of cells and media. triol. The total amount of glutathione conjugates isolated Uptake and Transformation of Chol Epoxide to Cholestane from the cells or medium amounted to <10% of the choles- Triol. Studies of uptake and transformation of Chol epoxide tane triol formed (data not shown). These findings indicate were initiated by adding 0.5 XCi of [14C]Chol epoxide (spe- that Chol epoxide hydrolase activity was low in these cells cific activity = 19.4 ;iCi/Amol) to 60-mm dishes. Final con- but accounted for the only readily discernible transformation centrations of these steroids were adjusted up to a maximum of Chol epoxide. of 62 uM by addition of nonradioactive compound in ace- The transformation of Chol epoxide to cholestane triol in tone, and dishes were incubated for intervals up to 24 hr. cultures of V79 cells in which the Chol epoxide did not pro- After incubation, the medium was collected and saved, the duce significant mutagenesis is shown in Fig. 2. The appar- monolayers were washed twice with fresh medium, and the ent difference between the cultures studied and described in washings were pooled. The cells were scraped from the dish- Figs. 1 and 2 was the batch of serum used in the culture es and collected by centrifugation at 500 x g for 10 min. The medium. There is uncertainty as to whether serum differ- stability of Chol epoxide and cholestane triol in complete ences accounted for the disparate behavior of the cells; nev- medium was examined by incubating the medium alone for ertheless, the results shown in Fig. 2 and reported in Table 3 similar time intervals. Lipids were extracted as described show clear differences in Chol epoxide conversion to choles- (25). Isolation, quantitation, radioactivity measurements of tane triol as well as in Chol epoxide toxicity and mutagene- the steroids, and the methods for measuring epoxide hydra- sis. The pattern of Chol epoxide uptake by the cells was sim- tion were carried out as described (26). The degree of conju- ilar to that shown in Fig. 1, although the amount accumulat- gate formation was estimated by recovery of radioactive glu- ed was -30% less over the same time period. Moreover, 26- tathione or mercapturic acid conjugates (28). The extent of 29% of the Chol epoxide taken up was converted to choles- Chol epoxide hydration (dependent on epoxide hydrolase, tane triol. E.C. 3.3.2.3) or glutathione conjugation was calculated from Inhibition of DNA Synthesis. Incorporation of [3H]thymi- measurement of the conversion of [14C]Chol epoxide to dine into DNA was determined 45 min after treatment of the products. cells with Chol, Chol epoxide, and cholestane triol, and this incorporation protocol is considered to be a measure of the RESULTS effects of Chol, Chol epoxide, and cholestane triol on DNA synthesis (23, 27). Twenty-four-hour treatment with _11.9 Metabolism of Chol Epoxide. The Chol epoxide used in ,uM cholestane triol produced a significant inhibition of these studies was found by HPLC to be 95% 5a,6a and 3% [3H]thymidine incorporation into DNA (Table 1); however, 5/3,6g. The only other identifiable contaminant eluted with a at this concentration 2- to 4-hr treatment had no effect on retention time identical to Sa-cholestan-3p-ol-6-one (6-keto- DNA synthesis. Twenty-four-hour treatment with _24.8 ,uM cholestanol, 10.5 min). The retention times observed by Chol epoxide produced a significant inhibition of DNA syn- HPLC (see above) and by gas chromatography (retention thesis. Maximal inhibition was found at 62 ,uM (the highest time for Chol epoxide of 12.4 min) were identical among the concentration studied) and was significant only after 16 hr of commercial samples and those synthesized by using the treatment. In this regard, Chol epoxide was found to be less monoperphthalic acid reaction. The identity of Chol epoxide potent than cholestane triol in inhibiting DNA synthesis. was further confirmed by its electron-impact mass spectro- Chol did not affect [3H]thymidine incorporation (data not graphic fragmentation patterns (data not shown). The radio- shown). Downloaded by guest on September 26, 2021 4200 Medical Sciences: Sevanian and Peterson Proc. NatL Acad Sci. USA 81 (1984)

A B .4 160 -5 u

la 140 '0 '5 120

r_ , 80 ,y 0 100

-5 80

0 °2 60 + - = *m- 40

u 20 f z ~~~~~~~0 0 12 16 20 24 6.2 21 62 Time, hr Chol epoxide, gM

FIG. 1. Uptake and Chol epoxide metabolism to cholestane triol in V79 cells. Cells were treated with 106 dpm of [14C]Chol epoxide and the level of uptake (o) and extent of conversion to cholestane triol (e) were determined at various intervals. (A) Treatment with 62 A&M Chol epoxide. (B) Effect of various Chol epoxide concentrations on uptake and conversion to cholestane triol over a 24-hr incubation period. Values represent mean ± SEM of four experiments.

Cytotoxicity. Chol (25 uM) was not cytotoxic. The effects fractions of V79 cells treated with Chol epoxide or choles- of Chol epoxide and cholestane triol on the survival of V79 tane triol decreased exponentially with increasing time of cells are shown in Fig. 3. The surviving fraction of V79 cells treatment. decreased exponentially on treatment of the cells for 24 hr Mutagenesis. The z8Gr mutants produced by the protocol with Chol epoxide and cholestane triol. Using these data, the used for these measurements of mutagenesis have been thor- dose that produces a decrease in the surviving fraction equal oughly characterized and shown to be hypoxanthine/guan- to l/e in the linear region of the survival curve of Chol epox- ine phosphoribosyltransferase-deficient mutants (24). In the ide and cholestane triol was calculated to be 106 ± 15 juM present experiments, the spontaneous mutation frequency and 43 ± 4 uM, respectively, indicating that cholestane triol was 0.5 ± 0.24 mutants per 105 survivors (nine experiments). was more than twice as toxic as Chol epoxide. The surviving The mutation frequencies of cultures treated with Chol were not significantly different from the spontaneous mutation frequency. The mutation frequencies produced by treating cultures with Chol epoxide and cholestane triol are given in Table 2. z Treatment of V79 cells for 2-24 hr with 62 AM Chol epoxide 0 produced mutants at frequencies that were reproducibly 4.6- x to 11.8-fold greater than the spontaneous mutation frequencies. Although the mutation frequency was not dose depen- r-o dent nor time dependent over the entire range of doses and C_ E treatment times, significantly fewer mutants were produced 0

"- ay _ Table 1. Inhibition of [3H]thymidine incorporation into DNA tS Conc., Duration, [3H]Thymidine ,0 80 Compound ttM hr incorporation* d0L.; 24 0.93 ± 0.09 0 Chol epoxide 12.4 601 "'o.rz 24.8 24 0.82 ± 0.20 37.2 24 0.88 ± 0.13 4)- 401 62.0 24 0.61 ± 0.10 -5 62.0 16 0.70 ± 0.10 4 1.24 ± 0.04 .0 62.0 = 62.0 2 1.16 ± 0.08 Cholestane triol 5.9 24 0.91 ± 0.03 11.9 24 0.77 ± 0.07 0 4 8 12 16 20 24 17.8 24 0.45 ± 0.16 Time, hr 23.8 24 0.42 ± 0.17 11.9 4 1.33 ± 0.20 FIG. 2. Uptake and Chol epoxide metabolism to cholestane triol 11.9 2 0.88 ± 0.22 in V79 cells. These cultures are different from those represented in Fig. 1 with respect to a demonstrated absence of mutagenesis. Results represent mean ± SEM for two to five experiments. Treatment conditions and designations are otherwise as for Fig. 1A. *Treated/control. Downloaded by guest on September 26, 2021 Medical Sciences: Sevanian and Peterson Proc. Natl. Acad. Sci. USA 81 (1984) 4201

Concentration, AM Time, hr Table 3. Comparison of cytotoxicity of Chol epoxide under 20 40 60 10 2 conditions of induced and uninduced mutations 1.0 ,0 Cholestane Surviving z8Gr mutants triol, Group fraction Background Chol epoxide nmol/24 hr A 0.54 ± 0.066 0.5 ± 0.24 4.8 ± 1.6 17.0 + 1.9 B 0.27 ± 0.161 0.5 ± 0.24 1.0 ± 0.5 29.8 ± 4.6 V79 cells were treated with 62 AM Chol epoxide and treatment C effects were differentiated on the basis of significant (group A) and insignificant (group B; n = 6) mutations over background levels. cd produced mutation frequencies that were significantly higher ._ than the background mutation frequency, but a higher con- ._ centration of cholestane triol, which caused the cells to de- V) tach from the dishes, and lower concentrations of cholestane triol did not produce mutation frequencies significantly greater than the background mutagenesis. Comparison with the mutation frequencies produced by 6.8 MM N-methyl-N'-nitro-N-nitrosoguanidine (2-hr treatment) indicates that Chol epoxide was weakly mutagenic and that at less than 17.8 AM cholestane triol was not significantly mutagenic. Relationships Between Cytotoxicity, Mutagenesis, and Chol FIG. 3. Dose-response curves (Left) and treatment-time depen- Epoxide Hydration. The results given in Tables 1 and 2 and dence (Right) of the cytotoxicity of Chol epoxide (e) and cholestane shown in Fig. 3 were obtained with cultures in which the triol (o) in V79 cells. Treatment time for the dose-response curves uptake of Chol epoxide was relatively high and its conver- was 24 hr. The time-course data represent treatment of cells with sion to cholestane triol was relatively low. These results are 11.9 AM cholestane triol and 62 ,M Chol epoxide. Values represent compared in Table 3 with data from cells that were cultured mean ± SEM of two to five experiments; lines were fitted by linear in medium containing a different batch of serum. These cul- regression analysis. tures displayed a lower (-=30%) uptake of Chol epoxide and higher (-2-fold) conversion of Chol epoxide to cholestane by 24-hr treatment with <12.4 AM Chol epoxide and by a 2- triol In the latter hr treatment with 62 Chol epoxide. (Fig. 2). cells, the Chol epoxide was twice AM as toxic as in the former cultures and was not significantly Treatment of cells for 24 hr with 17.8 uM cholestane triol mutagenic. These observations show that the cytotoxicity of Table 2. Mutation frequencies in cultures of V79 Chinese Chol epoxide can be enhanced and its mutagenicity can be hamster cells abolished under conditions in which Chol epoxide is effi- ciently converted to cholestane triol. Frequency, z8Gr colonies DISCUSSION Conc., Treatment per 105 We have shown that Chol epoxide is a weak mutagen in Chi- Compound AM time, hr survivors P* nese hamster V79 cells (Table 2), but the conversion of Chol Acetone 0.5% 24 0.5 ± 0.24 (9) epoxide to cholestane triol appears to modify the mutagenic- N-methyl-N'- ity of Chol epoxide. Cholestane triol is more toxic (Fig. 3 and nitro-N-nitroso- ref. 14) and a more potent inhibitor of DNA synthesis (Table guanidine 6.8 2 26.0 ± 4.18 (9) 1) than Chol epoxide, and cytotoxicity was observed after Chol epoxide 2.1 24 1.5 ± 0.30 (2) <0.05 short exposures to cholestane triol, whereas longer expo- 6.2 24 1.1 ± 0.30 (2) sures were necessary for the marginal mutagenesis induced 12.4 24 3.9 ± 1.00 (3) <0.01 by 17.8 AM cholestane triol. By contrast, lower concentra- 24.8 24 4.8 ± 2.30 (2) tions of Chol epoxide and relatively short treatments with 37.2 24 5.7 ± 3.53 (2) Chol epoxide were significantly mutagenic (Table 2). 62.0 24 4.8 ± 1.60 (5) <0.05 These data suggest that the mutagenicity of Chol epoxide 62.0 16 3.9 ± 1.33 (4) <0.05 can be decreased by its conversion to cholestane triol, which 62.0 8 3.6 ± 1.21 (3) <0.05 is more toxic and less mutagenic than Chol epoxide. Table 3 62.0 4 5.9 ± 1.40 (3) <0.01 shows that, in cells in which the conversion of Chol epoxide 62.0 2 2.3 ± 0.10 (2) <0.01 to cholestane triol is enhanced, the cytotoxicity of Chol ep- Cholestane triol 2.4 24 1.4 ± 0.70 (2) oxide was also enhanced and the mutagenicity of Chol epox- 5.9 24 2.5 ± 0.90 (4) ide was abolished. The cause of the enhancement of Chol 11.9 24 2.2 ± 1.00 (6) epoxide conversion to cholestane triol in these cells may be 11.9 8 0.7 ± 0.70 (4) related to the batch of serum used in these experiments and 11.9 4 1.7 ± 1.20 (4) is being investigated. Nevertheless, our data indicate that 11.9 2 0.8 ± 0.40 (3) the mutagenicity of Chol epoxide in Chinese hamster V79 17.8 24 2.6 ± 0.70 (6) <0.02 cells depends on the capacity of the cells to metabolize Chol 23.8t 24 2.2 ± 2.20 (4) epoxide to cholestane triol and can be related to their epox- Chol 25.0t 24 0.3 (1) ide hydrolase activities. We attribute our data, showing a Numbers in parentheses represent numbers of experiments. plateau in the dose-response for mutagenesis with >12.4 *Student's t test that the difference from the background mutation ,uM Chol epoxide, to a selective enhancement of cytotoxic- frequency is due to chance. ity resulting from an accumulation of cholestane triol. Con- tExtensive cell detachment. flicting data in the literature regarding the genotoxicity of tPrecipitate formed. Chol epoxide (12-14, 16-18, 29-31) may be the result of dif- Downloaded by guest on September 26, 2021 4202 Medical Sciences: Sevanian and Peterson Proc. Natl. Acad Sci. USA 81 (1984) ferences in the above enzymatic conversion of Chol epoxide 4. Watabe, T., Kanai, M., Isobe, M. & Ozawa, N. (1980) Bio- to cholestane triol by the various systems studied. The con- chim. Biophys. Acta 619, 414-419. tribution to Chol epoxide genotoxicity by minor contami- 5. Raicht, R. F. & Cohen, B. I. (1981) Biochim. Biophys. Acta nants also remains as a possibility. Painter and Howard (31) 666, 455-461. 6. Gray, M. F., Lawrie, T. D. V. & Brooks, C. J. W. (1971) Lip- reported that inhibition of HeLa cell DNA synthesis was no ids 6, 836-843. longer evident when a more highly purified compound was 7. Roscoe, H. G. & Fahrenbach, M. J. (1971) J. Lipid Res. 12, tested. It is assumed that only the a-isomer of Chol epoxide 17-23. was studied, however, the purified compound may have 8. Black, H. S. (1980) Lipids 15, 705-709. contained less cholestane triol and a different proportion of 9. Black, H. S. & Douglas, D. R. (1972) Cancer Res. 32, 2630- the P-3isomer of Chol epoxide, which can be effectively sepa- 2632. rated from the a-isomer of HPLC. This was not the case in 10. Reddy, B. S. & Wynder, E. L. (1977) Cancer Res. 39, 2533- the present study as the same batch of Chol epoxide was 2539. Nevertheless, there is the possibility 11. Petrakis, N. L., Gruenke, L. D. & Craig, J. C. (1981) Cancer used in all experiments. Res. 41, 2563-2565. of differential genotoxicity between the a- and P-isomers of 12. Parsons, P. G. & Goss, P. (1978) Aust. J. Exp. Biol. Med. Sci. Chol epoxide: our preliminary experiments suggest that the 56, 287-296. p-isomer is about 5 times more mutagenic than the a-isomer 13. Kelsey, M. I. & Pienta, R. J. (1979) Cancer Lett. 6, 143-149. while being approximately as cytotoxic as cholestane triol. 14. Kelsey, M. I. & Pienta, R. J. (1981) Tox. Lett. 9, 177-182. The content of Chol epoxide (total amount of a- and p- 15. Reddy, B. S. & Watanabe, K. (1979) Cancer Res. 39, 1521- isomers) measured in rat lung tissue has been reported to be 1524. 10-15 AsM (25). This range is computed from the quantities 16. Smith, L. L., Smart, V. B. & Ansari, G. A. S. (1979) Mutat. extracted from freshly excised lung parenchyma and ex- Res. 68, 23-30. the of tissue water content. Consequently, 17. Reddy, B. S., Weisburger, J. H. & Wynder, E. L. (1978) in pressed on basis Carcinogenesis, eds. Slaga, T. J., Sivak, A. & Boutwell, R. K. rat lung tissue, which also displays low Chol epoxide hydro- (Raven, New York), Vol. 2, pp. 453-464. lase activities (26), contains Chol epoxide at concentrations 18. Seelkopf, C. & Salfelder, K. (1962) Krebsforsch 64, 459-464. similar to the range that produces a linear dose-response for 19. Fioriti, J. A., Kanuk, M. J., George, M. & Sims, R. J. (1970) mutagenesis (0-12.4 AM) in V79 cells. These considerations Lipids 5, 71-75. suggest that Chol epoxide might induce mutagenic lesions in 20. Bowden, J. P., Muschik, G. M. & Kawalek, J. C. (1970) Lip- the cells of rat lung tissues and other tissues bearing low ids 14, 623-629. Chol epoxide hydrolase activity. The electrophilicity of Chol 21. Lo, W. B. & Black, H. S. (1971) Experientia 27, 1379-1400. epoxide and demonstrated reactivity with DNA (32) indicate 22. Imai, H., Werthessen, N. T., Subramanyam, V., LeQuesne, It is also P. W., Soloway, A. H. & Kanisawa, M. (1980) Science 207, the potential for producing mutagenic lesions. pos- 651-653. sible that a variety of agents that inhibit Chol epoxide hydro- 23. Peterson, A. R. (1980) Cancer Res. 40, 682-688. lase may in a similar manner potentiate the mutagenicity of 24. Peterson, A. R., Peterson, H. & Heidelberger, C. (1979) Can- Chol epoxide. cer Res. 39, 131-138. 25. Sevanian, A., Mead, J. F. & Stein, R. A. (1979) Lipids 14, technical in 634-643. We thank Hazel Peterson for her excellent assistance F. Biochim. the treatment and analysis of cells. We also wish to thank Drs. 26. Sevanian, A., Stein, R. A. & Mead, J. (1980) Bruce Ames and Paul Hochstein for their encouragement and help- Biophys. Acta 614, 489-500. is the of Grant BC 441 from the 27. Painter, R. B. & Howard, R. (1978) Mutat. Res. 54, 113-115. ful discussions. A.R.P. recipient 28. Hayakawa, T., Lemahieu, R. A. & Udenfriend, S. (1974) American Cancer Society. Arch. Biochem. Biophys. 162, 223-230. 29. Bischoff, F. (1969) Adv. Lipid Res. 7, 165-244. 1. Ansari, G. A. S. & Smith, L. L. (1970) Photochem. Photobiol. 30. Black, H. S. & Chan, J. T. (1976) Oncology 33, 119-122. 30, 147-150. 31. Painter, R. B. & Howard, R. (1982) Mutat. Res. 92, 427-437. 2. Aringer, L. & Eneroth, P. (1974) J. Lipid Res. 15, 389-398. 32. Blackburn, G. M., Rashid, A. & Thompson, M. H. (1979) J. 3. Kadis, B. (1978) J. Steroid Biochem. 9, 75-81. Chem. Soc. Chem. Commun., 420-421. Downloaded by guest on September 26, 2021