Fungal Oxidation of Benzo[a]pyrene and (&)-trans-7,8-Dihydroxy-7,8-dihydrobenzo[a]pyrene EVIDENCE FOR THE FORMATION OF A BENZO[a]PYRENE 7,8-DIOL-9,10-EPOXIDE*

The oxidation of benzo[a]pyrene by intact cells of the filamentous fungus Cunninghamella elegans resulted in the formation of a complex mixture of polar products which were detected by high pressure liquid chromatography. One of the products had a retention time on high pressure liquid chromatography identical with that given by (+)-7/3,8a,9a,lO/3-tetrahydroxy-7,8,9,10- tetrahydrobenzo[a]pyrene. In addition the absorption and mass spectra given by the fungal metabolite were consistent with this structural assignment. residue was analyzed by HPLC. The HPLC separation conditio^ were identical with those used for the isolation of hydrolysis products of diol epoxide-I and diol epoxide-2. Experiments with (*)-[3H]benzo[a]pyrene 7,8-diol (4.04 X IO6 dpm, 5 nmol in 50 p1 of dimethylforrnamide) were conducted as described above for nonradioactive dihydrodiol. HPLC analysis was performed as described for unlabeled benzo[a]pyrene 7,8-diol metab- olites except that the HPLC eluant was collected and the tritium contained in all fractions was quantitated by liquid scintillation count- ing techniques.

In contrast to the vast literature on the metabolism of benzo[a]pyrene by mammals little information is available that relates to the products formed from this substrate by microorganisms. Bacteria oxidize benzo[a]pyrene to a mixture of cis-benzo[a]pyrene 7,8-and 9,lO-diols (25 Organism and Growth Conditions-The isolation and characterization of C. elegans has been described previously (27). Small scale fermentations were conducted in 125-ml Erlenmeyer flasks which contained 30 ml of Sabouraud dextrose broth. After 72 h, cultures were harvested as described previously (28) and transferred to 30 ml of fresh medium. Benzo[a]pyfene was added to each of three flasks to give a final concentration of 0. 5  Extraction and Detection of Transformation Products-After 48 h, the mycelium from each flask was removed by filtration. The culture filtrates were pooled and extracted with 3 volumes of ethyl acetate. The organic layer was dried over anhydrous sodium sulfate, and the solvent removed in uacuo at 30°C in the dark. The residue was redissolved in 500 p1 of methanol and 250 pl was applied to a preparative silica gel thin layer plate (20 X 20 cm; 250 p thickness; Analtech). The chromatogram was developed in a solvent system consisting of benzene:ethanol(91). Benzo[a]pyrene metabolites were located on chromatograms by viewing under ultraviolet light (254 nm) and by comparing the chromatographic mobilities to synthetic benzo[a]pyrene derivatives. The area of the chromatogram that contained reaction products more polar than benzo[a]pyrene 9,lO-diol was designated the pre-9,lO-diol region. The areas of silica gel corresponding to the pre-9,lO-diol region were removed and extracted with methanol. After removal of the solvent the residue was stored at -20°C prior to analysis by high pressure liquid chromatography (HPLC). Anaiyses were conducted on a Waters model 440 high pressure liquid chromatograph fitted with two coupled pBondapak C~S columns (3.9 mm X 39 cm). The metabolites were separated by gradient elution. The initial solvent composition was 40% methanol and 60% water. The final solvent concentration was 95% methanol and 5% water. A linear gradient was employed (1 h) with an initial flow rate of 0.8 ml/min at 2000 p.s.i.g. In experiments with [''Cc] benzo[a]pyrene, fractions (0.4 ml) were collected at %-min intervals and added to tubes containing 5.0 ml of Aquasol-2. The radioactivity present in each fraction was determined in a Beckman LS-250 liquid scintillation counter.
Ultraviolet and visible spectra were determined on a Beckman model 25 recording spectrophotometer. Mass-spectral analysis was performed on a Finnigan model 2100 mass spectrometer at 70 eV ionizing voltage with a solid probe. Samples were dissolved in 10 pl of tetrahydrofuran. Spectra were recorded at a probe temperature of 240°C and an ion source temperature of 325°C.
Separation of Hydrolysis Products of Diol Epoxide-1 and Diol E p o x~e -2 -S~t h e t i c diol epoxide-I and diol epoxide-2 were hydrolyzed as described by Yang et al. (15). HPLC analysis of the tetraol hydrolysis products was carried out as described previously (13) using a DuPont Zorbax ODS column (6.2 mm X 25 cm).
Metabolism of ffj-Benzo[alpyrene 7,8-Di?tydrodwl-CelIs of C. elegans were incubated as described above in which (rt)-benzo[a] pyrene 7,8-dihydrodiol (2 mg dissolved in 0.5 ml of dimethylformamide) replaced benzo[a]pyrene as the substrate. After 12 h, the culture fittrates were extracted with 3 equal volumes of ethyl acetate and the organic layers were dried over anhydrous NaD04. The ethyl acetate was removed in uacuo at 3OoC in the dark. Each residue was analyzed by HPLC. The HPLC separation conditio^ were identical with those used for the isolation of hydrolysis products of diol epoxide-I and diol epoxide-2. Experiments with (*)-[3H]benzo[a]pyrene 7,8-diol (4.04 X IO6 dpm, 5 nmol in 50 p1 of dimethylforrnamide) were conducted as described above for nonradioactive dihydrodiol. HPLC analysis was performed as described for unlabeled benzo[a]pyrene 7,8-diol metabolites except that the HPLC eluant was collected and the tritium contained in all fractions was quantitated by liquid scintillation counting techniques.

RESULTS AND DISCUSSION
When C. elegans was grown on Sabouraud dextrose broth in the presence of benzo[a]pyrene, several oxidation products were detected by thin layer chromatography. Metabolites more polar than benzo[a]pyrene 9,lO-diol were further analyzed by HPLC. Fig. 3 shows that a complex mixture of products were detected. Our previous studies have shown that many of these compounds are sulfate conjugates of hydroxylated benzo[a]pyrene derivatives (26).
One product (Compound VI) gave an identical absorption spectrum to those given by the tetraols formed by the addition of water to diol epoxides-1 and -2 (Fig. 4). Although the absorption spectrum is only indicative of the presence of a pyrene ring it suggests that the fungal metabolite was reduced at positions 7,8,9, and 10 of the benzo[a}pyrene molecule. When ['*C]benzo[a]pyrene was incubated with C. elegans and the polar metabolites analyzed as described above, Compound VI accounted for 1.3% of the total polar benzo[a]pyrene metabolites.
Diol epoxides-1 and -2 are readily hydrolyzed to tetraols (Fig. 2). Thus, diol epoxide-1 undergoes predominantly cis addition of water to the epoxide group by a reaction mechanism that is acid-catalyzed at low pH and spontaneous above  (-) and trans-tetra012 formed by hydrolysis of (+)-diol epoxide-2 (---). The conditions used for hydrolysis of the diol epoxide and separation of its hydrolysis products are described under "Materials and Methods." pH 5.0. In contrast, diol epoxide-2 undergoes a trans addition of water to the epoxide group at low pH. At pH d u e s above 7.0 the spontaneous hydrolysis mechanism results in an increase in the amount of cis hydroxylation observed (24). These results explain the ratio of tetraols formed from synthetic diol epoxides-1 and -2 (Fig. 5A). The tetraols formed from diol epoxide-1 (cis-tetraol 1 and trans-tetraol 1) show the predicted ratio (14,23). However, diol epoxide-2 gave more of the cis hydrolysis product; a result that suggests that the synthetic sample had undergone spontaneous hydrolysis. Nevertheless, the retention time of the fungal metabolite (Fig. 5B) was identical with that of trans-tetraol 2 which can be formed by the trans addition of water to the epoxide group of diol epoxide-2. Further evidence for the identity of the compound formed from benzo[a]pyrene by C. elegans was provided by mass spectral analysis (Fig. 6). The metabolite gave a parent ion at m/e 320 and a fragmentation pattern almost identical with that reported for trans-tetra012 (23). The results suggest that   Benzofalpyrene 7,8-Diol-9,10-epoxide by Fungi the metabolite formed from benzo[a]pyrene by C. eleguns is trans-tetraol2. It is of interest to note that only trans-tetra01 2 has been identified as a benzo[a]pyrene metabolite. Hydrolysis of diol epoxide-2 should also produce cis-tetraol 2 (see Fig. 2). Under the HPLC conditions used in the initial isolation experiment (Fig. 3) cis-tetraol 2 would elute with Compound VII. Fractions that eluted between 33 to 39 min were pooled and resolved into four components by the chromatographic conditions used to separate the diol epoxide hydrolysis products. None of the compounds in this fraction had spectral properties similar to those of cis-tetraol 2. It seems likely that this hydrolysis product is formed in amounts too low to be detected by the experimental conditions. If this supposition is correct it suggests that diol epoxide-2 formed by C. elegans may undergo an acid-catalyzed hydrolysis reaction inside of the cell. The pH of the culture medium changes from 5.8 to 6.4 during the experiment and these conditions would favor the formation of cis-tetraol2. It is also of interest that cis-tetraol 1 was not detected as a product formed from either benzo[a]pyrene or benzo[a]pyrene 7,8-diol (see below). This would be the major tetraol formed from diol epoxide-1 irrespective of the pH of hydrolysis (24). These observations suggest that C. eleguns may oxidize benzo[a] pyrene in a stereospecific manner to (-)-benzo[a]pyrene 7,8diol. If the (+)-enantiomer of benzo[a]pyrene 7,8-diol were formed one would expect, by analogy with mammalian systems, that epoxidation would yield mainly diol epoxide-1 and hence cis-tetraol 1 as a major hydrolysis product. However, diol epoxide-1 is extremely susceptible to nucleophilic attack due to anchimeric assistance by the proximate cis-hydroxyl group (22). Thus, if diol epoxide-1 is formed inside the cell it may react with cell nucleophiles and not be excreted into the culture medium. Future experiments with fungal microsomes should establish whether or not diol epoxide-1 is formed from benzo[a]pyrene and benzo[a]pyrene 7,8-diol. Although the results strongly indicate that C. elegans oxidizes benzo[a] pyrene to a diol epoxide it is not possible at this time to eliminate the alternative possibility that trans-tetra012 could be formed by the hydrolysis of the analogous benzo[a]pyrene 9,10-diol-7,8-epoxide. Results of studies on the fungal oxidation of benzo[a]pyrene 9,lO-diol will be reported at a later date. When 3H-labeled (+)-trans-7,8-diol was incubated for 12 h with intact cells of C. elegans 4% of the substrate was converted into metabolites. Analysis by HPLC gave the results shown in Fig. 7. Compound I had an identical retention time to that given by trans-tetraol2. Calculations based on radioactivity measurements show that trans-tetra012 accounts for 25% of the trans-7,8-diol metabolites whereas Compounds I11 and IV accounted for 23 and 14%, respectively.

Formation of
The experiment was repeated in the absence of radioactive benzo[a]pyrene 7,8-diol in order to isolate Compounds I, 111, and 1V for character~tion purposes. Compound VI1 is unchanged benzo[a]pyrene 7,8-diol and no attempt was made to isolate Compounds 11, V, and VI. The absorption and mass spectra of Compound I were identical with those given by trans-tetraol 2. Similar observations were reported for the oxidation of benzo[a]pyrene 7,8-diol by microsomes prepared from livers of control and phenobarbital-induced rats. However, these liver preparations also formed significant amounts of diol epoxide-1 from both (+)-and (+)-benzo[a]pyrene 7,8diol. In contrast, liver microsomes from rats treated with 3-methylchol~threne oxidized (-)and (+)-benzo[a Jpyrene 7,8diol to diol epoxide-2 as the major product (14). The only detectable tetraol formed from (+)-benzo[a]pyrene 7,8-diol by C. elegans was trans-tetra01 2 which indicates that diol epoxide-2 may be the only stereoisomer formed by the fungus. The formation of diol epoxide-1 could be undetected for the reasons cited previously and it will be of interest to see if the fungal enzymes show the same inducible properties and enantiomeric specificity as the enzymes present in mammalian microsomes.
Mass spectral analysis of C o m~u n d s I11 and IV revealed that each compound gave a parent ion at m/e 302 and a base peak at m/e 284 (p-HzO). These observations suggest that Compounds I11 and IV are d~ydrodiol derivatives of a benzo[a]pyrene phenol. Further evidence was obtained by observing the spectral changes that occurred when each compound was treated with acid (Fig. 8). The actual structures of Compounds I11 and IV which are presumed to be new monohydroxylated derivatives of trans-benzo[a]pyrene 7,8-diol have yet to be determined.
From the data presented we propose the metabolic sequence in Fig. 9 for the oxidation of benzo