INTRODUCTION

Benzo(a)pyrene, a prototypic PAH¹ found widespread in the environment, causes cancer in laboratory animals and is a suspected human carcinogen (1). It is well established that the cancer causing effects of benzo(a)pyrene are dependent upon its metabolism to reactive cytotoxic and genotoxic metabolites (2, 3). The metabolism of benzo(a)pyrene to an array of phenols, quinones, diols, and diol epoxides has been well documented (2). Although there is the possibility that several of these contribute to the carcinogenic insult, it is generally appreciated that formation of benzo(a)pyrene-7,8-dihydrodiol-9,10-epoxide, a highly potent DNA-binding metabolite, is an important event in the initiation of tissue neoplasms. Benzo(a)pyrene-7,8-dihydrodiol-9,10-epoxide is formed by two rounds of oxidative metabolism, with BPD as an intermediate. In addition to its undergoing cytochrome P450- or peroxidase-catalyzed bioactivation to benzo(a)pyrene-7,8-dihydrodiol-9,10-epoxide, BPD also undergoes metabolism by competing alternate pathways to form less toxic metabolites, for example, oxidation by dihydrodiol dehydrogenase to form benzo(a)pyrene-7,8-dione (4, 5) or conjugation by UGT to form glucuronides or by sulfotransferase to form sulfate derivatives (2, 6, 7).

The protective effect of UGTs against toxicity mediated by benzo(a)pyrene has been demonstrated in several studies (8-10). Reduced glucuronidation of benzo(a)pyrene derivatives in UGT-deficient homozygous j/j and heterozygous j/+ RHA rats in vivo correlated with increased covalent binding to hepatic DNA and microsomal protein compared to UGT-normal congenic RHA +/+ controls (8). The correlation could also be demonstrated in vitro after incubations of benzo(a)pyrene with liver microsomes (8, 9) and lymphocytes (9) from the UGT-deficient and UGT-normal animals. In another study, cultures of skin fibroblasts from UGT-deficient Gunn rats exhibited reduced glucuronidation of benzo(a)pyrene metabolites and corresponding enhanced benzo(a)pyrene covalent binding and micronucleus formation (10). These studies support the hypothesis that the level of UGT-mediated detoxification is an important determinant of cell or tissue susceptibility to toxicities mediated by benzo(a)pyrene (11-13).

An important goal in understanding tissue and organism susceptibility to benzo(a)pyrene toxicity is to identify and characterize the enzymes mediating detoxification of benzo(a)pyrene metabolites. In the case of BPD, although it is known that BPD glucuronides are formed in vitro by liver microsomes from rat (8, 14) and human (15), and primary hepatocyte cultures from rat (3) and human (6, 7), little information currently exists regarding the identity of the UGTs responsible for BPD glucuronidation. In a study of four human isoforms (UGT1*6 (a.k.a., UGT1A6),² UGT2B7, UGT2B10, and UGT2B11), only UGT2B7 exhibited detectable glucuronidating activity toward BPD (15). In another study, none of 5 rat isoforms surveyed (UGT1A6, UGT2B1, UGT2B2, UGT2B3, and UGT2B6) were active in BPD glucuronidation (16). In addition, while it is known that many UGT activities are inducible, few data exist regarding the regulation of BPD UGT. In attempting to classify BPD as a Group 1 (PAH-inducible) or Group 2 (phenobarbital-inducible) substrate, Bock et al. (17) reported that BPD transferase activity was weakly inducible by either inducing agent and concluded that it was not clearly associated with either one of the two substrate groups.

Recently we have shown (36) that rats treated with the chemopreventive hepatic enzyme-inducing agent oltipraz (4-methyl-5-pyrazinyl-1,2-dithiole-3-thione) have marked elevations (i.e. up to 10-fold) of liver microsomal BPD UGT activity (18). BPD transferase activity was also significantly induced by treatment with -naphthoflavone, a PAH-type inducing agent, whereas phenobarbital had no effect. These studies suggested that BPD UGT activity was more responsive than was previously thought, that BPD is a Type 1 substrate (i.e. increased predominately by PAH-type inducing agents), and that the gene(s) encoding the responsible isoenzyme(s) are regulated by both PAHs and dithiole thiones. Further studies using Gunn rats which have an inactive UGT1 gene locus (18, 19) suggested that the oltipraz-inducible BPD UGT is a UGT1 family member.

The objective of this study was to identify the isozyme corresponding to the UGT1-associated, oltipraz-inducible BPD UGT. The strategy used was to prepare and screen a cDNA library prepared from oltipraz-induced RALA255-10G LCS-3 cells for UGT1 cDNAs. RALA255-10G LCS-3 cells, an adult rat liver-derived immortalized cell line, exhibit a high basal BPD transferase activity, which is increased >2-fold by treatment with oltipraz. In contrast to rat liver, a tissue with complex and abundant expression of various transferases from both the UGT1 and UGT2 families, RALA255-10G LCS-3 cells do not express detectable UGT2 family RNAs and have very limited expression of UGT1 subtypes. We, therefore, anticipated that less intensive screening would be required to identify the BPD UGT cDNA.

EXPERIMENTAL PROCEDURES

The RALA255-10G LCS-3 cell line was kindly provided by J. Chou (National Institute of Child Health and Human Development, Bethesda, MD). AHH-1 cells and the expression vector pMF6 were kindly provided by Dr. Charles Crespi (Gentest Corp., Woburn, MA). BPD (()-trans-isomer) and other benzo(a)pyrene metabolites were from the NCI Carcinogen Repository (Midwest Research Institute). [¹⁴C]UDP-glucuronic acid (320 mCi/mmol) was from DuPont NEN.

Female F344 rats, 6-8 weeks old (Harlan Sprague-Dawley, Raleigh, NC) were treated with either oltipraz (300 mg/kg/day for 2 days) suspended in vehicle (30% (w/v) polyethylene glycol 8000 dissolved in water) or vehicle alone. Twenty four h after the final dose, animals were euthanized and total liver RNA was prepared (20).

RALA255-10G LCS-3 cells are an SV40 temperature-sensitive tsA255 immortalized adult rat hepatocyte clonal cell line (21-23). The line proliferates rapidly when grown at 33 °C in -modified Eagle's medium (1 ×) (Mediatech) and 4% fetal bovine serum (Hyclone, Salt Lake City, UT) in a 5% CO₂ incubator. The cells were trypsinized twice weekly (1:8 split ratio). For UGT assays, RALA255-10G LCS-3 cells were trypsinized, centrifuged at 1000 × g (4 °C), and washed in ice-cold phosphate-buffered saline. The washed cells were resuspended in 0.1 M potassium phosphate buffer (pH 7.5) to a final density of 1 × 10⁶ cells/43 µl and stored at 70 °C.

AHH-1 cells were grown in suspension as described (24) using RPMI 1640 containing 9% horse serum and maintaining the cell density between 0.1 and to 0.8 × 10⁶ cells/ml. Cells were collected, washed, resuspended, and stored as described above.

Cellular BPD UGT activity was assayed using cells suspended in 100 mM potassium phosphate (pH 7.5) (1.16 × 10⁶ cells or ~0.14 mg of total protein/50 µl). Frozen cells were thawed and lysed by three consecutive freeze-thaw cycles (5 min on dry ice and 5 min in a room temperature water bath) followed by brief sonication to disperse aggregates. Aliquots (43 µl) were placed in fresh tubes containing 2 µl of ()-trans-BPD dissolved in dimethyl sulfoxide (2.5 mM). UDP-glucuronic acid (150 nmol), containing 0.1 µCi of [¹⁴C]UDP-glucuronic acid, was added and reactions were carried out at 37 °C. Under these conditions, formation of glucuronide was proportional to cell concentration and the time of incubation. Parallel incubations omitting BPD were performed as reaction blanks. The reaction products were separated by thin layer chromatography as described (14). The zone containing fluorescent product was removed, dissolved in 0.1 ml of concentrated hydrofluoric acid, and counted by liquid scintillation. The counting efficiency was >90%. For qualitative analysis, autoradiographs of plates were generated by spraying the plates with a surface autoradiography enhancer (EN³HANCE^TM, DuPont NEN) and exposure to x-ray film for 7-14 days.

Northern electrophoresis was carried out in gels containing 0.9% agarose, 20 mM MOPS (pH 7.0), 8 mM sodium acetate, 1 mM EDTA, and 0.6 M formaldehyde. The RNA was transferred by overnight capillary blotting to a nylon membrane. Blots were cross-linked in a Stratalinker (Stratagene, La Jolla, CA) and allowed to dry at room temperature for several hours.

Plasmids with the rat c-jun and cyclophilin cDNAs were obtained from the American Type Culture Collection and Dr. Phillip Hylemon (Medical College of Virginia), respectively. A plasmid containing a fragment of the 5 unique exon of UGT1A6 (783-bp insert, bases +28 to +810 of the 3-methylcholanthrene-inducible 4-nitrophenol UGT cDNA (25)) was generated by polymerase chain reaction-mediated amplification followed by cloning into pBluescript and analysis by restriction mapping. A similar strategy was employed to clone complementary DNA representing the UGT1 common region (pR16E3, 637-bp insert, bases +970 to +1606 of the 3-methylcholanthrene-inducible 4-nitrophenol UGT sequence (25)) and NAD(P)H quinone oxidoreductase (pRQR, 1389-bp insert, bases +32 to +1420 of the sequence reported by (26)). Inserts were labeled to >1 × 10⁹ dpm/µg DNA with [-³²P]dCTP using a random priming strategy (27). Standard prehybridization and hybridization conditions were used (28). Autoradiograms were generated by exposing blots to x-ray film for 2 days at 70 °C in cassettes with ³²P-intensifying screens.

Complementary DNA was synthesized from 2 µg of poly(A)⁺ RNA isolated from RALA255-10G LCS-3 cells treated for 24 h with oltipraz (100 µM), using a uniZAP cDNA synthesis kit (Stratagene). The library was packaged using Gigapack Gold^TM -phage packaging extract (Stratagene) and plated, yielding >5 × 10⁶ independent phage clones. After amplifying once, a total of 1 × 10⁶ phage were screened using the pR16E3 common region fragment as probe. Fourteen UGT1 positive phages were plaque-purified, amplified, and subjected to the "in vivo excision" reaction as described by Stratagene. Plasmid DNA from each clone was isolated and characterized by restriction analysis. The longest clones were sequenced using Sequenase^TM (U. S. Biochemical Corp.).

The LC14 cDNA was ligated into the SalI site of pMF6 (24). Colonies containing plasmid with the LC14 insert in the correct (pMF6-LC14-3) or reverse (pMF6-LC14-23) orientations were selected and amplified. Log phase AHH-1 cells were electroporated with pMF6-LC14-3 or pMF6-LC14-23. Forty-eight h later, transfected cells were selected by addition of hygromycin B (400 µg/ml) to the cultures. A hygromycin B-resistant cell population expressing LC14 under the control of the herpes simplex virus thymidine kinase promoter was obtained after 3-4 weeks in culture. The cells were assayed for BPD UGT activity as described above. BPD UGT activity was found to be stably expressed for a period of at least 2 months.

Glucuronidation of Benzo(a)pyrene Metabolites and Susceptibility to -Glucuronidase

Glucuronidation activities of LC14-transfected AHH-1 cells toward other benzo(a)pyrene metabolites were determined as described above for BPD UGT. The sensitivity of the products to -glucuronidase was examined by adjusting the pH of the final reactions to 6.8, incubating with 12.5 units of E. coli Type VII-A -glucuronidase (Sigma) for 24 h at 37 °C, and analyzing the products by silica thin layer chromatography.

RESULTS

During studies to evaluate RALA255-10G LCS-3 cells as a model to investigate the UGT gene regulation, we observed that the cells exhibit high basal BPD UGT activity which is increased by treatment with oltipraz. Fig. 1A shows the effect of treatment with 25 or 50 µM oltipraz on BPD UGT activity. These treatments resulted in a 50 and 140% increase, respectively, in the BPD UGT activity compared to control cells.

This effect was accompanied by a concentration-dependent increase in total UGT1 family transcripts, assessed by Northern analysis using a pan-probe (pR16E3) directed at the identical 3 end sequences of UGT1 family mRNAs (Fig. 1B). Quantitation by PhosphorImager analysis revealed increases of 1.6-, 2.1-, and 2.8-fold at 10, 25, and 50 µM oltipraz, respectively. Messenger RNA for NAD(P)H quinone oxidoreductase, another detoxification enzyme responding to oltipraz (29), was also dose-dependently increased over this range of concentrations (3.1-, 6.1-, and 7.5-fold, respectively). Interestingly, the UGT1A6 mRNA, which is highly responsive in rat liver in vivo, was found to be unaffected in oltipraz treated-RALA255-10G LCS-3 cells. The negative data for the c-jun and cyclophilin mRNAs demonstrate that the effect is not due to different amounts of RNA loaded in each lane. UGT2 transcripts were not detected (data not shown). These differential responses suggest that induction of an unknown UGT1 mRNA correlates with induced BPD UGT activity.

To identify the mRNA a  ZAP cDNA library was constructed from RNA prepared from oltipraz-treated RALA255-10G LCS-3 cells, and the library was screened for UGT1 clones using pR16E3 as probe. UGT1 cDNAs were found to be highly represented in the library (0.15%). Fourteen positives were randomly selected and purified for analysis by restriction endonuclease site mapping, Southern analysis, and terminal nucleotide sequencing. Insert sizes estimated by XhoI-EcoRI digestion ranged between 1100 and 2200 bp. None hybridized with UGT1A6-specific probe.

To access inducibility by oltipraz, probes corresponding to the 5 variable region were prepared by polymerase chain reaction amplification, labeled with [³²P]dCTP, and hybridized with RNA from vehicle- and oltipraz-treated RALA255-10G LCS-3 cells and rat liver. The data for 4 of these clones is presented in Fig. 2. All clones tested exhibited identical patterns of inducibility, a 40-fold increase in vivo but only a 4-fold increase in the cells. The lower magnitude of inducibility in vivo and in vitro is attributable to the presence of higher basal expression in the untreated RALA255-10G LCS-3 cells (compared to control rat liver).

Partial nucleotide sequencing of the four clones revealed nucleotide sequence identity at the 5 terminus. The complete nucleotide sequence of the longest clone, LC14, is presented in Fig. 3. The LC14 insert is 2301-bp in length, and contains 42- and 666-bp 5- and 3-untranslated regions flanking the 1596-bp UGT coding region. The transferase encoded by LC14 is 531 amino acids in length, the first 286 of which match the deduced sequence of the UGT1A7 unique exon (referred to previously as UGT1A2) reported by Emi et al. (30) with the exception of 4 differences (LC14/UGT1A7, respectively): Ile⁶-Val⁶, Gly¹⁰⁴-Ser¹⁰⁴, Gly¹⁹⁴-Pro¹⁹⁴, and Leu²⁸⁴-Val²⁸⁴. Isolation of genomic DNA spanning the UGT1A7/UGT1A6 region of the UGT1 locus and sequencing of the UGT1A7 exon have permitted us to verify the identity of the residues at positions 6 and 104 of the gene as Ile and Gly, respectively. These results indicate correspondence of the LC14 cDNA with the UGT1A7 gene product and suggest that the differences in the sequence reported by Emi et al. (30) are either due to strain variation or possible sequencing errors.

To determine whether the UGT encoded by LC14 (UGT1A7) represents a BPD UGT, the LC14 insert was expressed in AHH-1 cells and tested for activity (Fig. 4). Homogenates from cells expressing LC14 in the correct orientation (pMFLC14-3) formed a fluorescent glucuronide with the same relative migration as in the incubation with rat liver microsomes. This product was not visible in reactions with untransfected cells or cells transfected with the cDNA in the reverse orientation (pMFLC14-23).

The amount of BPD glucuronide increased linearly for at least 24 h, resulting in 0.4 nmol of glucuronide/10⁶ cells by 24 h (Fig. 5A). A similar time course for 3-OH-BP (Fig. 5B) revealed a 20-fold more rapid rate of glucuronidation (3 nmol of glucuronide formed/10⁶ cells/8 h) with linear production of glucuronide up to 8 h. Analysis of the apparent K_m by Lineweaver-Burk plots (Fig. 5, C and D) revealed relatively high affinities of the LC14 UGT enzyme for BPD and 3-OH-BP of 15.5 and 12.3 µM, respectively.

Since the UGT is active toward 3-OH-BP and is known to be inducible by PAH-type inducing agents, we assessed its activity toward other benzo(a)pyrene monophenols and the 4,5-dihydrodiol (Fig. 6). All metabolites tested were found to be suitable substrates for UGT1A7 with the phenols representing the best substrates. All products required the presence of UDP-glucuronic acid in the reaction (data not shown) and were susceptible to hydrolysis by -glucuronidase (Fig. 7A). In a representative quantification (Fig. 7B), the radioactivity associated with the 3-OH, 9-OH, and BPD glucuronide products was reduced to background levels by treatment with -glucuronidase.

-Glucuronidase sensitivity of glucuronides formed from various benzo(a)pyrene metabolites by pMFLC14-3-transfected AHH-1 cells.

To compare the tissue distribution and oltipraz responsiveness of this isoform in hepatic and extrahepatic tissue, RNA was isolated from liver, lung, spleen, kidney, intestine, and ovary of rats treated with vehicle or oltipraz (150 mg/kg/day for 3 days) and analyzed by Northern blotting using a specific 5 variable region probe for UGT1A7. These results are shown in Fig. 8. UGT1A7 RNA was undetectable in untreated rat liver and was markedly increased in this tissue by treatment with oltipraz. In contrast, UGT1A7 was constitutively expressed at low levels in the 5 extrahepatic tissues examined. The level of expression was highest in kidney > intestine = lung > spleen = ovary. The only tissue other than liver to exhibit significant (>2-fold) inducibility by oltipraz was intestine. The RNA in ovary was decreased by treatment with oltipraz.

To compare the expression of UGT1A7 to UGT1A6, an identical blot was analyzed for UGT1A6 mRNA (Fig. 8). As seen for UGT1A7, UGT1A6 RNA was barely detectable in liver of untreated rats and was dramatically increased by oltipraz treatment. Similarly, UGT1A6 RNA was found to be constitutively expressed in the 5 extrahepatic tissues examined, and the liver and intestine were the only tissues to show signficant oltipraz responsiveness. The relative abundance of the UGT1A6 and UGT1A7 mRNAs in each tissue could not be compared because the probe lengths differed significantly.

DISCUSSION

This study describes the cloning and characterization of an oltipraz-inducible BPD UGT cDNA (LC14) isolated from RALA255-10G LCS-3 cells. It is the first report of a rat liver UGT cDNA encoding activity toward BPD, the key intermediate in the pathway resulting in the bioactivation of benzo(a)pyrene. Two unexpected findings of the study were that: 1) the oltipraz-responsive UGT corresponds to the predicted product of UGT1A7, recently identified as a PAH-responsive gene (30), and 2) the "BPD UGT" is widely active toward other benzo(a)pyrene metabolites, particularly the phenols. In view of this broad spectrum glucuronidating activity, it is not surprising that the enzyme is under the control of PAH-type inducers. The regulation of UGT1A7 expression by two presumably independent mechanisms, the dithiole thione mechanism of oltipraz (an agent with chemopreventive properties) and the Ah-receptor mechanism of PAHs suggests an important role of this enzyme in adaptive xenobiotic detoxification.

Pretreatment of rats with 3-methylcholanthrene has recently been shown to increase liver levels of both the UGT1A6 (i.e. A1) and UGT1A7 (i.e. A2) RNAs (30). UGT1A6 was cloned by Iyanagi et al. (25) a decade ago, but UGT1A7 has remained elusive. The notion that at least two different PAH-responsive UGTs exist was first proposed by Bock and co-workers (31, 32), who performed Western blot analysis of liver microsomal protein from control and 3-methylcholanthrene-treated rats using anti-rat UGT_3MC antibodies and observed increased staining of two proteins in the 50 kDa range. PAH-type inducers induce rat liver glucuronidating activity toward 3-OH-BP (17) and BPD (36), yet cloned and expressed UGT1A6 failed to exhibit detectable activity toward either substrate (16). The observation that 3-methylcholanthrene induces the UGT1A7 mRNA (30) together with our finding that the LC14-encoded UGT exhibits catalytic activity toward both 3-OH-BP and BPD suggests that UGT1A7 represents the source of PAH-inducible activity observed in previous studies.

The UGT isolated and characterized in this report, UGT1A7, is the fifth rat UGT1 product to be cloned at the cDNA level. Previously, cDNAs representing the products of UGT1A1 (bilirubin UGT), UGT1A5 (uncharacterized UGT), UGT1A6 (3-methylcholanthrene-inducible 4-nitrophenol UGT), and UGT1A9 (a pseudogene with an in-frame termination codon) have been reported (18, 30). The LC14 (UGT1A7) cDNA possesses the same 3 end sequence as for the other rat UGT1 cDNAs reported (18). The identical sequence is due to the sharing of exons (II through V) located at the extreme 3 end of the UGT1 gene locus (30). The 5 unique sequence of the LC14 cDNA arises from one of multiple alternative first exons located upstream of the common exons. Each first exon encodes the amino-terminal of a unique UGT isozyme. Studies of the exon organization of the rat UGT1 gene locus (30)³ confirm that the unique sequence of LC14 is encoded by the seventh 5 variable exon upstream from the common exons II-V.

This study represents the first report of the cloning and characterization of a UGT1A7 from rat or human. Isolation of the cDNA encoding this UGT isoenzyme type was hindered by its low basal expression in liver from these species. Expression of UGT1A7 in control rat liver was not even detected using reverse transcriptase-polymerase chain reaction (30). Our approach overcame this problem by using RALA255-10G LCS-3 cells, which exhibit a high basal and oltipraz-inducible BPD UGT activity. To date, the isolation of the human UGT1A7 cDNA has not been reported. Harding et al. (33) isolated and characterized the substrate specificity of a "bulky phenol"-glucuronidating cDNA, HLUG P4, which initially was thought to represent UGT1A7 (34). However, more recent mapping and sequencing studies of the human UGT1 locus indicate the origin of HLUG P4 is the UGT1A9 locus.⁴ Activities of the HLUG P4 isoform toward benzo(a)pyrene metabolites have not been characterized.

The focus of our study was on the glucuronidating activities of the LC14-encoded UGT toward benzo(a)pyrene metabolites, BPD in particular. To investigate its catalytic activity, we utilized the AHH-1 system (24). A potential disadvantage of this system is the instability of the Epstein-Barr virus-based episomal expression plasmid. However, over a 2-month period, we found no evidence of declining enzyme activities in the pMF6-LC14 transfected cells. A major advantage of the AHH-1 system is the ease of handling the suspension cultures.

A unique catalytic property of the LC14-encoded isoform is its BPD UGT activity. Although rats have been considered poor glucuronidators of BPD, we have found that BPD glucuronidation capacity depends on the exposure status of the animals (36). Our data suggest that UGT1A7 is the source of increased BPD UGT activity observed in livers of oltipraz- and 3-methylcholanthrene-treated rats. The LC14 mRNA is induced >40-fold in livers of oltipraz-treated rats, which exhibit 5-10-fold increases in liver microsomal BPD UGT activity. The lack of exact correspondence in the magnitude of the RNA and enzyme activity increases may be explained by the presence of a second isoform capable of glucuronidating BPD and which is present in control rat liver.

Of the 4 human and 5 rat UGT isoforms that have been screened for activity toward BPD (15, 16), only the human UGT2B7 form exhibited detectable catalytic activity toward BPD. Jin et al. (15) reported the BPD UGT activity of UGT2B7-transfected COS cells as 0.3 pmol/mg/min. In our study, LC14-transfected AHH-1 cells converted BPD to the glucuronide at a rate of 450 pmol/1 × 10⁶ cells/16 h or 3.3 pmol/mg/min assuming 1 × 10⁶ cells 0.14 mg of total protein.⁵ It is not clear if the 10-fold higher apparent rate catalyzed by UGT1A7 is due to different levels of expression in the two systems or reflect true differences in catalytic turnover of the enzymes. Our results show that UGT1A7 represents a low K_m isoform, whereas the K_m of UGT2B7 was not characterized. The high affinity of UGT1A7 for BPD supports a role for the enzyme in the formation of BPD glucuronides found in primary cultures of 3-methylcholanthrene treated-rat hepatocytes incubated with benzo(a)pyrene (3).

In addition to BPD, UGT1A7 was also found to possess broad spectrum glucuronidating activity toward 13 other benzo(a)pyrene metabolites. Compared to the human and rat isoforms tested to date, UGT1A7 is the only isoform that exhibits detectable activity toward each of 12 primary (phenol) or two secondary (diol) benzo(a)pyrene derivatives. Most of the previously tested UGT isoforms were not catalytically active toward any benzo(a)pyrene metabolites. The two most active forms were: UGT2B1 (16) which was active toward 8 of the metabolites, and UGT2B7 (15) which was active toward 10. As for BPD activity, it is not possible to draw conclusions about the relative catalytic efficiencies of the three forms in the glucuronidation of monophenols. Our results show that UGT1A7 is between 1 and 2 orders of magnitude more active in the glucuronidation of some phenols, including the 3- and 9-hydroxy derivatives, compared to the 7,8- (or 4,5-)diol. The higher apparent V_max for glucuronidation of benzo(a)pyrene phenols explains, at least in part, the greater abundance of phenol glucuronides produced in primary rat hepatocyte cultures (3). The unusually broad activity toward planar and non-planar PAH substrates and hydroxyl groups located at virtually all positions around the benzo(a)pyrene aromatic ring system indicates that the substrate binding site of UGT1A7 is fairly promiscuous and may accommodate benzo(a)pyrene metabolites in different orientations.

Another feature of the cloned UGT1A7 is that it is expressed in many extrahepatic tissues. Our findings agree with the data of Emi et al. (30) who found expression of the A2 RNA in spleen, lung, intestine, kidney, and testes, with undetectable levels in liver. Our results also are consistent with previous findings that 3-OH-BP and BPD UGT activities are detected in many extrahepatic tissues of rats (17). The molecular basis for the lower expression of UGT1A7 in rat liver is not clear. However, UGT1A6 exhibits a similar pattern of constitutive expression (35), and it has been suggested that UGT1A6 may be under the control of a transcriptional repressor.

In summary, we have identified a UGT, UGT1A7, which exhibits broad ranging activity toward various benzo(a)pyrene metabolites, including BPD, the precursor to the ultimate benzo(a)pyrene carcinogen. The data predict that UGT1A7 could exert a critical influence on benzo(a)pyrene carcinogenicity by modulating the amount of BPD entering the final bioactivation step. Furthermore, UGT1A7 appears to be widely expressed in many extrahepatic tissues and can be induced in liver and intestine by oltipraz, a prospective chemopreventive agent. Our data suggest that UGT1A7 contributes to the chemoprotective properties of oltipraz and may underly, at least in part, the recent observations implicating an unidentified UGT in various hepatic and extrahepatic cell types in genoprotective effects against benzo(a)pyrene (8-10).