TY - JOUR
AU1 - Hu, Zhuohan
AU2 - Wells, Peter G.
AB - Abstract UDP-glucuronosyltransterases (UGTs) catalyse the glucuronidation and elimination of most xenobiotics and, thereby, may prevent their alternative bioactivation to carcinogenic and teratogenic reactive intermediates. Previous studies have shown that glucuronidation, bioactivation, and covalent binding of the carcinogen/teratogen benzo[a]pyrene (BP) in rat lymphocytes accurately reflected those processes in hepatic microsomes from the same animals. Accordingly, lymphocytes from 12 normal human volunteers were incubated with BP metabolites to determine UGT variability and its potential toxicological relevance. Over 200-fold interindividual variability was observed in both the glucuronidation and covalent binding of BP metabolites, with decreasing total glucuronidation among subjects correlating with a decreased UGT-modulated reduction in covalent binding (R2 = 0.8272, p < 0.01) and, in six subjects, enhanced cytotoxicity (r = −0.9338, p < 0.001). Decreased glucuronidation of both BP diols (r = −0.9106, p < 0.001) and BP diones (r = −0.9625, p < 0.005), but not BP monophenols, correlated with enhanced cytotoxicity. These results provide the first evidence for substantial interindividual variability in UGT activities for BP metabolites among the normal population and suggest that UGT-deficient individuals may be at increased risk from the reactive intermediate-mediated effects of BP and related xenobiotics. glucuronidation, benzo[a]pyrene, cancer, birth defects, bioactivation, reactive intermediates One of the primary goals of molecular epidemiology is to identify individuals with high susceptibility to the carcinogenic effects of xenobiotics such as benzo[a]pyrene (BP), due to inherited and/or acquired environmental host factors (Harris, 1991). Many carcinogens, including BP, require multiple enzymatic steps for bioactivation to their ultimate carcinogenic form, usually a highly reactive electrophilic or free radical intermediate that respectively covalently binds to or oxidizes cellular macromolecules such as DNA, RNA, and protein (Fig. 1). The amount of reactive intermediate is regulated by a balance among alternative pathways of elimination prior to bioactivation, the activity of the bioactivating pathway, and subsequent pathways for detoxification of the reactive intermediate. Accordingly, individual differences in these pathways may be an important determinant of carcinogenic risk (Harris, 1991). Such differences have been observed for the bioactivating cytochromes P450 and the detoxifying glutathione S-transferases (GST), with interindividual variations in activity in human tissues and cells of more than 100-fold (Guengerich, 1991; Harris, 1991). The UDP-glucuronosyltransferases (UGTs) are a superfamily of isoenzymes that catalyze the conjugation of UDP-glucuronic acid with most xenobiotics and endogenous compounds, including bilirubin, facilitating their elimination (Guillemette, 2003; Mackenzie et al., 1997; Tukey and Strassburg, 2000). Since UGT-catalyzed elimination precedes and therefore can avoid bioactivation, UGTs may be important cytoprotective enzymes (Fig. 1), although in some cases, such as with aromatic amines, glucuronidation can introduce a good leaving group and favor the formation of electrophilic reactive intermediates, especially under acidic conditions such as in the bladder (Kadlubar et al., 1977). The potential toxicological relevance of UGTs is illustrated by in vivo studies using UGT1A-deficient Gunn and RHA rats, which, when treated with the analgesic drug acetaminophen (paracetamol), had decreased acetaminophen glucuronidation and correspondingly enhanced acetaminophen bioactivation, covalent binding, and hepatic/renal toxicity (de Morais and Wells, 1988, 1989; de Morais et al., 1992a). In humans, decreasing acetaminophen glucuronidation correlated with enhanced bioactivation, although not sufficient for hepatic or renal toxicity using a single intravenous administration of a therapeutic dose (de Morais et al., 1992b; Wells et al., 2004). Epidemiological studies have demonstrated hepatic and renal toxicity in people taking chronic therapeutic doses of acetaminophen, possibly due in part to UGT deficiencies (Perneger et al., 1994; Whitcomb and Block, 1994). Similarly, with in vitro and in vivo studies with BP, UGT-deficient Gunn and RHA rats had decreased glucuronidation of BP metabolites and correspondingly enhanced bioactivation and covalent binding to DNA and protein (Hu and Wells, 1992). Pregnant UGT-deficient Gunn rats had substantially enhanced BP-initiated embryotoxicity at a subcarcinogenic dose (25 mg/kg ip) that had no effect on UGT-normal Wistar controls (Wells et al., 1989). In cultured skin fibroblasts derived from RHA rats, homozygous UGT-deficient (j/j) cells compared with UGT-normal (+/+) controls had reduced glucuronidation of BP metabolites and correspondingly enhanced BP covalent binding and micronucleus production (Vienneau et al., 1995), which is thought to reflect the potential for carcinogenic initiation (Kim and Wells, 1996). Similarly, DNA oxidation and micronucleus formation initiated by the tobacco-specific nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), BP, and the anticonvulsant drug phenytoin and its HPPH metabolite were enhanced in both heterozygous and homozygous UGT-deficient fibroblasts compared to UGT-normal controls (Kim and Wells, 1996; Kim et al., 1997b). The overall results with acetaminophen, BP, NNK, and phenytoin in UGT-deficient models indicate that UGTs are cytoprotective and genoprotective, preventing reactive intermediate-mediated protein and DNA lesions. Since the toxicologic relevance of hereditary UGT deficiencies for acetaminophen in rats was reflected in humans, a lymphocyte model involving incubation with BP metabolites and uridine diphosphate glucuronic acid (UDPGA) was developed for the human study of BP and related carcinogenic environmental chemicals. Initial studies in RHA rats demonstrated the in vivo relevance of the lymphocyte model (Hu and Wells, 1994). Among UGT-normal (+/+) and heterozygous (j/+) and homozygous (j/j) UGT-deficient animals, the progressive decrease in glucuronidation of BP metabolites and converse increase in BP covalent binding in lymphocytes from +/+, j/+, and j/j rats accurately reflected those processes in hepatic microsomes taken from the same animals (Hu and Wells, 1994), as well as reflecting enhanced in vivo developmental toxicity (Wells et al., 1989) and in vitro genotoxicity (Kim and Wells, 1996; Vienneau et al., 1995). Thus, it was reasonable to assume that human lymphocytes might be similarly predictive of in vivo processes and, hence, potentially of value in carcinogenic risk assessment. This lymphocyte study shows that decreased UGT activity in humans can lead to enhanced bioactivation, covalent binding, and cytotoxicity of BP, and suggests an important cytoprotective role for UGTs in modulating human toxicities due to exposure to environmental chemicals. MATERIALS AND METHODS Human subjects. All subjects gave written, informed consent, had no history of hepatic or renal disease, and were not taking medication at least within two weeks of the beginning of the study. Five female and seven male volunteers between the ages of 18 and 39 years were studied, and their smoking history was recorded. These studies were approved by the Institutional Review Committee for the University of Toronto. Animals. Male Sprague-Dawley rats, 300-320 g (Charles River Canada Ltd., St. Constant, Quebec), were treated with the cytochromes P4501A1/2 (CYP1A1/2) inducer beta-naphthoflavone (BNF) and used as a source of CYP1A1-induced hepatic microsomes. All animals were allowed to acclimatize for 1 week in separate hanging cages with a wire mesh floor. Food (Rodent Chow, Ralston Purina Co., St Louis, MO) and tap water were provided ad libitum, and a 12-h light and dark cycle was maintained automatically. Chemicals. [7,10-14C]Benzo[a]pyrene (specific activity 58.5 mCi/mmol) was purchased from Amersham Canada Ltd. (Oakville, Ontario). The reference standards for BP metabolites were obtained from the Chemical Repository at IIT Research Insititute (Chicago, IL). Uridine diphosphate glucuronic acid (UDPGA) and beta-glucuronidase (type 8-1) were obtained from the Sigma Chemical Co. (St Louis, MO), and nicotinamide adenine dinucleotide phosphate, reduced form (NADPH) from Boehringer Mannheim Canada Ltd. (Dorval, Quebec). All other reagents used were of analytical or HPLC grade. Measurement of bilirubin glucuronidation in vivo. Bilirubin glucuronidation was estimated in six subjects by measuring plasma concentration of unconjugated bilirubin using a standardized commercial kit (No.123 919; Boehringer Mannheim). A 0.2-ml aliquot of plasma from heparinized blood was collected after centrifugation at 800 × g for 5 min at room temperature and mixed with kit reagents. Unconjugated bilirubin in plasma was expressed as μmol/l. Preparation of human lymphocytes. Sixty milliliters of whole blood was collected in heparinized sterile tubes from each subject. Within 1 h after blood collection, lymphocytes were isolated in Ficoll-Paque medium by density gradient centrifugation (Hu and Wells, 1994). The final cellular pellet contained >95% lymphocytes. Preparation of rat hepatic microsomes. For the preincubation procedure, hepatic microsomes were prepared from male Sprague-Dawley rats (300-320 g) 24 h following the final treatment with the CYP1A1-inducing agent BNF in corn oil (80 mg/kg ip for 3 days). Animals were killed by CO2 asphyxiation, the livers were homogenized in 4 volumes of ice-cold 1.15% KCI, and the homogenate was centrifuged at 9,000 × g for 25 min at 4°C. The supernatant was then recentrifuged at 105,000 × 9 for 1 h at 4°C. The microsomal pellet was suspended in 1.15% KCI solution to yield a protein concentration of about 20 mg/ml, as estimated by the method of Bradford (Bradford, 1976), with bovine serum albumin as the standard. Preincubation of benzo[a]pyrene in vitro. To provide the lymphocytes with sufficient oxidized BP metabolites, it was necessary to preincubate BP with hepatic microsomes prior to incubation with lymphocytes (Hu and Wells, 1994). Radioactively labeled [14C]BP, 0.25 μCi/4.81 nmoles (final BP concentration, 16 μM) was used in the preincubation to determine BP glucuronidation and covalent binding, while unlabelled BP (final concentration, 5 mM) was used in the preincubations to determine BP cytotoxicity. Radiolabeled or unlabeled BP was preincubated for 60 min with hepatic microsomes from BNF-induced rats (1.0 mg protein) in buffered medium containing final concentrations of 130 mM NaCI, 5.2 mM KCI, 1.3 mM KH2PO4, 10 mM Na2HPO4, and 1.3 mM MgSO4 (pH 7.4), supplemented with 10 mM NADPH. The preincubation was terminated by adding 0.3 ml of chilled methanol to each sample. Microsomal protein was precipitated for 30 min at −20°C, and the tubes were then centrifuged at 1,000 × g for 5 min at 4°C. The methanol in the supernatant was evaporated under nitrogen, and the water phase of the supernatant was stored at −20°C for up to 30 min for lymphocyte incubations. Human lymphocyte incubations. Lymphocytes (5 × 106/ml) were resuspended in buffered medium containing 115 mM NaCI, 4.25 mM KCI, 1.2 mM KH2PO4, 1.2 mM MgSO4, 2.6 mM CaCI2, 25 mM NaHCO3, 0.6% HEPES (pH 7.4). The lymphocyte incubation was started by adding the supernatant from the hepatic microsomal preincubation and incubating, with or without 10 mM UDPGA, at 37°C and 5% CO2 for 16 h to determine cytotoxicity, and for 4 h to determine glucuronidation and covalent binding. The incubation was stopped by precipitating lymphocytes by centrifugation at 800 × g for 5 min. The supernatant was collected for measuring BP glucuronide conjugates, and the pellet for BP covalent binding to protein. Measurement of BP glucuronidation. A 200-μl aliquot of the supernatant was injected directly into a high-performance liquid chromatograph (HPLC) (Series 4; Perkin-Elmer Canada Ltd.) equipped with a 200-μl loop injector, a 15-cm reverse-phase C-18 Spherisorb 5-μ column (Jones Chromatography Ltd, Mid Glamorgan, UK) and connected in series to a UV detector set at 254 nm, and a radioactive flow detector for HPLC (model 171; Beckman Instruments Canada Inc., Mississauga, Ontario). The separation program used a flow rate of 1 ml/min and a linear methanol/water gradient as follows: At step 0, before sample injection, the column was equilibrated for at least 5 min with 46% methanol; after injection of the analytical sample, step 1 lasted 45 min, during which time the methanol concentration was increased with a linear gradient from 46 to 95%; step 2, the concentration of methanol was changed linearly over 10 min from 95 to 100%. During the last 5 min, the column was washed with 100% methanol. The glucuronidation of BP was expressed as percentage of total radioactivity. BP glucuronides were identified as previously described and by comparison with synthetic standards (Hu and Wells, 1992, 1994). Measurement of BP covalent binding. After termination of the incubation, the lymphocyte pellet was washed repeatedly with hot methanol until all removable radioactivity was gone. Washed pellets were dissolved overnight at 40°C with 0.5 ml of 1.0 M NaOH and neutralized with 1.0 M HCI, and radioactivity was counted in a liquid scintillation spectrometer (model LS 5000TD; Beckman) using a standard liquid scintillation cocktail (Ready Protein +, Beckman). Counting efficiency was 93-97%. The covalent binding of BP to microsomal protein was expressed in pmoles/mg protein/h. UGT-modulated covalent binding of BP metabolites indicated the magnitude of reduction in BP covalent binding when the UGT cosubstrate UDP-glucuronic acid (UDPGA) was added to the incubation. This parameter was calculated as follows: [(BP covalent binding without UDPGA − BP covalent binding with UDPGA)/BP covalent binding without UDPGA] × 100. Measurement of cytotoxicity. Essential and optimal conditions for evaluating BP cytotoxicity in lymphocytes were established, and BP cytotoxicity subsequently was determined in lymphocytes from six subjects. A 50-μl aliquot of lymphocyte incubation media (final cell concentration, 5 × 106/ml) was mixed with 50 μl of 0.2% trypan blue in normal saline solution for 3 min. At least 200 cells per assay were counted using a hemocytometer, and trypan blue exclusion allowed the determination of percentage viability of lymphocytes (Hinman et al., 1990). Statistics. Statistical analysis of, and correlations among, BP glucuronidation and covalent binding and cytotoxicity were performed using a standard, computerized statistical program (SAS Institute Inc., Cary, North Carolina), with p < 0.05 as the minimal level of significance. RESULTS Contribution of BP Preincubation to the Measurement of Cytotoxicity In control studies there was no consistent difference over time in cytotoxicity of lymphocytes exposed to BP itself, without incubation, or its vehicle (DMSO), from 6 to 12 h (Fig. 2). In contrast, when lymphocytes were incubated with the supernatant from BP preincubated with rat hepatic microsomes, there was over a three-fold increase in cytotoxicity at 6 h (p < 0.01), increasing to over a seven-fold cytotoxic enhancement within 10-12 h (p < 0.01) (Fig. 2). This study indicated a necessity for preincubation in studies of BP cytotoxicity in human lymphocytes, as previously had been observed with rat lymphocytes (Hu and Wells, 1994). Glucuronidation of Bilirubin and BP The mean (± SE) plasma concentration of unconjugated bilirubin in six human volunteers was 4.71 ± 1.02 μmol/l, which was well within the physiologically normal range [below 17.0 μmol/l, (Roy Chowdhury et al., 2001)] (data not shown), indicating that all subjects had normal UGT activity for glucuronidating bilirubin. The major BP glucuronides were conjugates with 3-hydroxy-benzo[a]pyrene (3-OH-BP), 9-hydroxy-benzo[a]pyrene (9-OH-BP), benzo[a]pyrene-6,12-dione (BP-6,12-dione), benzo[a]pyrene-3,6-dione (BP-3,6-dione), benzo[a]pyrene-4,5-dihydrodiol (BP-4,5-diol), and benzo[a]pyrene-7,8-dihydrodiol (BP-7,8-diol), for which the mean respective levels among subjects with detectable concentrations (number of subjects with detectable glucuronides/total number tested) were 0.43% (8/12), 0.10% (4/12), 0.24% (8/12), 0.50% (3/12), 3.4% (10/12), and 0.46% (9/12). Most importantly, a greater than 200-fold interindividual variation in UGTs for BP metabolites was observed in lymphocytes from the 12 normal volunteers, with BP glucuronides ranging from 0.01% (the lower limit of detection) to 5% of the total BP metabolites (Fig. 3). UGT-Dependent Modulation of BP Covalent Binding to Protein When lymphocytes were incubated without UDPGA, the cofactor for UGTs, a 220-fold interindividual variation in BP covalent binding was observed among 12 human subjects (Fig. 4, insert). There was a lower, 143-fold interindividual variation in BP covalent binding when lymphocytes were incubated with UDPGA. Comparing BP covalent binding in each subject, with and without UDPGA, addition of UDPGA reduced BP covalent binding (p < 0.05) in lymphocytes from 7 of 12 subjects, with the reduction varying from 7 to 93% (Fig. 4, insert). Among all subjects, there was no correlation between UGT activity for BP metabolites and the amount of BP covalent binding with UDPGA (data not shown). However, there was a significant correlation between UGT activity for all BP metabolites and the magnitude of reduction in BP covalent binding by the addition of UDPGA (R2 = 0.8272, p < 0.01) (Fig. 4). This reduction in BP covalent binding by UDPGA was termed “UGT-modulated BP covalent binding.” UGT-modulated BP covalent binding also was analyzed with respect to the glucuronidation of particular BP metabolites, including BP-diols, BP-diones, and BP-phenols. Total BP-diol glucuronides, including BP-4,5-diol and BP-7,8-diol, correlated highly with the UGT-modulated reduction in BP covalent binding (r = 0.9683, p < 0.001) (Fig. 5). Among the BP-diol glucuronides, both BP-7,8-diol glucuronide and BP-4,5-diol glucuronide independently correlated with the UGT-modulated reduction in BP covalent binding, with respective coefficients (R2) of 0.9144 (p < 0.005) and 0.6978 (p < 0.025) (Fig. 5). For BP-diones and BP phenols, neither the formation of total BP-dione glucuronides nor total BP phenol glucuronides correlated with UGT-modulated BP covalent binding, nor did the individual glucuronides of BP-6,12-dione, BP-3,6-dione, 3-OH-BP, or 9-OH-BP (data not shown). Modulation of BP Cytotoxicity by UGTs BP itself was not cytotoxic compared to DMSO controls (Fig. 2). Time-dependent BP-initiated cellular necrosis required preincubation with rat hepatic microsomes to produce oxidized BP metabolites, and this preincubation step was employed in all subsequent studies. In lymphocytes from six subjects, BP-initiated cytotoxicity was similar without UDPGA (about 80%); however, after the addition of UDPGA, there was a substantial interindividual variation in the UGT-dependent reduction in cytotoxicity, ranging from 0% (two subjects) to 14-48% reductions (p < 0.05) (four subjects) (Fig. 6, insert). Importantly, decreasing UGT activity for total BP metabolites correlated highly with increasing BP cytotoxicity (r = −0.9338, p < 0.001) (Fig. 6). UGT-modulated BP cytotoxicity also was analyzed with respect to the glucuronidation of particular BP metabolites, including BP-diones, BP-diones, and BP-phenols. Decreasing glucuronidation of total BP-diols correlated highly with enhanced BP cytotoxicity (r = −0.9106, p < 0.001) (Fig. 7). Among the BP-diol glucuronides, both BP-7,8-diol and BP-4,5-diol glucuronidation independently correlated with BP cytotoxicity, with respective coefficients (r) of −0.9499 (p < 0.001) and −0.7559 (p < 0.05) (Fig. 7). For BP-diones, similar to the glucuronidation of BP-diols, decreasing glucuronidation of total BP-diones correlated highly with enhanced BP cytotoxicity (r = 0.9625, p < 0.001) (Fig. 8). However, the independent glucuronidation of either BP-6,12-dione or BP-3,6-dione did not correlate with BP cytotoxicity (data not shown), since several subjects produced only one of the two BP-diones. There was no correlation between the glucuronidation of BP phenols and BP cytotoxicity (data not shown). DISCUSSION Toxicological Relevance of UGTs in Humans These results in human lymphocytes provide cellular confirmation of animal studies in vivo and in hepatic microsomes indicating a cytoprotective and genoprotective role for UGTs in modulating BP bioactivation and toxicity. Perhaps more importantly, these data provide human evidence for the cytoprotective importance of UGTs in humans exposed to potentially toxic environmental chemicals like BP. Previous studies have shown that lymphocytes from UGT-deficient rats can predict accurately the glucuronidation, bioactivation, and toxicity of BP observed in vivo and in hepatic microsomes from the same animals (Hu and Wells, 1992, 1994; Wells et al., 1989); therefore it is likely that the human lymphocyte model is similarly predictive of in vivo human toxicologic risk. Although the human volunteers had normal bilirubin UGT1A1 activity, the glucuronidation and covalent binding of BP metabolites in lymphocytes from these subjects varied over 200-fold, and decreased BP glucuronidation correlated significantly with both decreased UGT-modulated reductions in BP covalent binding and enhanced BP cytotoxicity. An increasing number of functionally important polymorphisms have been discovered that decrease the activity of UGT isoenzymes by reducing either protein levels or catalytic activity (Guillemette, 2003; Wells et al., 2004). In rats and humans, mutations in one or more of the common exons may compromise the function of a range of UGT isoenzymes, whereas mutations to a variable exon may result in a deficiency of a single isoenzyme (Bock, 1991; Guillemette, 2003; Iyanagi et al., 1998; Mackenzie et al., 1997; Tukey and Strassburg, 2000). The clinical relevance of this lymphocyte model to particular target tissues may vary due to tissue-specific differences in isoenzyme type and activity, not only for the UGT superfamily but also for enzymes involved in xenobiotic bioactivation and reactive intermediate detoxification, as well as cytoprotection and macromolecular repair. While glucuronide conjugates may facilitate delivery of the toxic chemical to the target organ for some tissues (Kadlubar and Hammons, 1987; Kwei et al., 1992; Ramesh et al., 2001), the role of glucuronidation in BP-exposed lymphocytes herein was protective, as it was for related studies of acetaminophen hepatotoxicity and nephrotoxicity in vivo (de Morais and Wells, 1988, 1989; de Morais et al., 1992a), BP and phenytoin developmental toxicity in vivo (Kim and Wells, 1998; Wells et al., 1989), and BP, NNK, and phenytoin genotoxicity and cytotoxicity in fibroblasts (Kim and Wells, 1996; Kim et al., 1997b; Vienneau et al., 1995). Modulation of BP Covalent Binding by UGTs Although reduced BP glucuronidation and enhanced cytotoxicity in human lymphocytes correlated with a reduced UGT-modulated decrease in BP covalent binding, these correlations were not observed with the absolute level of BP covalent binding itself, contrary to results from UGT-deficient rats. The correlations with BP covalent binding in various studies of inbred rats (Gunn and RHA strains) likely were evident because, unlike outbred humans, these inbred rat strains demonstrate virtually no interindividual differences in activities of drug metabolizing enzymes other than UGTs, such as cytochromes P450 (CYPs) (bioactivation) and glutathione S-transferases (GSTs) (reactive intermediate detoxification), the balance among which constitutes an important determinant of toxicity (de Morais and Wells, 1989; de Morais et al., 1992a; Hu and Wells, 1992). In humans, however, there is considerable interindividual variation in activities of other enzymes involved in the bioactivation of BP and detoxification of its reactive intermediates (Guengerich, 1991; Harris et al., 1979; Jahnke et al., 1990; Shamsuddin et al., 1985). In the current study, we observed a 220-fold interindividual variation of BP covalent binding in the lymphocyte incubations without the addition of either UDPGA or cosubstrates for sulfotransferase (ST) and GSTs. This interindividual variation in BP covalent binding may reflect the interindividual variation in P450 isozymes, epoxide hydrolases, and possibly other enzymes, such as peroxidases (Kim and Wells, 1996; Kim et al., 1997a), involved in BP biotransformation. In 7 of 12 subjects, there was a significant reduction in BP covalent binding when UDPGA was added to lymphocyte incubations. The difference in BP covalent binding in incubations with and without UDPGA was termed the “UGT-modulated” reduction in BP covalent binding. Reduced BP glucuronidation correlated highly with a reduced UGT-modulated decrease in BP covalent binding, indicating that the cytoprotective effect of UGTs depended upon the magnitude of reduction in BP covalent binding, rather than the absolute levels of BP covalent binding. This concept is consistent with the uniform cytotoxicity observed in lymphocytes from different subjects without addition of UDPGA (Fig. 6, insert), despite the substantial variation in BP covalent binding under the same conditions (Fig. 4, insert). The lack of cytotoxic correlation with BP covalent binding itself, as distinct from the UGT-modulated reduction in covalent binding, suggests interindividual differences in other contributing cellular processes, such as reactive intermediate detoxification and macromolecular repair. The modulation of BP covalent binding by UGTs could take place at least at three different substrate levels, involving BP-monophenols, BP-diones, and BP-diols. To evaluate the independent contributions from glucuronidation of these different BP metabolites to the modulation of BP covalent binding, we analyzed the glucuronides of 3-OH-BP, 9-OH-BP, BP-3,6-dione, BP-6,12-dione, BP-4,5-diol, and BP-7,8-diol. BP-diols. UGT-catalyzed glucuronidation of BP-diols, including BP-7,8-diol and BP-4,5-diol (Fig. 5), effectively reduced BP covalent binding in human lymphocyte incubations. Importantly, in each case, the magnitude of the reduction in BP covalent binding (UGT-modulated covalent binding) declined with decreasing UGT activity for these BP metabolites. BP-diol glucuronidation, and especially BP-7,8-diol glucuronidation, is the last possible step for preventing the final bioactivation of BP to benzo[a]pyrene-7,8-dihydrodiol-9,10-epoxide (BPDE), the ultimate toxic reactive intermediate (Fig. 1). Therefore, glucuronidation of BP-diols, and particularly BP-7,8-diol, appeared to play a critical cytoprotective role in reducing BP bioactivation. BP-diones. Although BP-diones themselves are not substrates for UGTs, they can be rapidly reduced by various oxido-reductases such as DT-diaphorase (Lind et al., 1978) and cytochrome P450 reductase (Bock et al., 1980) to the corresponding quinols, which are metabolized by UGTs to form both mono- and diglucuronides (Lind, 1985). BP-monophenols. The absence of modulation of BP covalent binding by glucuronidation of BP-monophenols, including 3-OH-BP and 9-OH-BP, was not surprising, given that BP-monophenolic metabolites have no so-called Bay region in their structure, which is essential for P450-catalyzed bioactivation (Dipple, 1985; Wood et al., 1979). Modulation of BP Cytotoxicity by UGTs Glucuronidation directly and significantly reduced BP cytotoxicity in lymphocyte incubations from four of the six subjects by 14 to 48%, and decreased glucuronidation of BP metabolites correlated highly with enhanced BP cytotoxicity (Fig. 6). Given that glucuronidation of BP-diols, including BP-7,8-diol and BP-4,5-diol, enhanced the UGT-modulated reduction in BP covalent binding, it was not surprising that the glucuronidation of these diols directly reduced BP cytotoxicity (Fig. 7). In contrast, contrary to the absence of a discernable effect on BP covalent binding, BP-dione glucuronidation appeared to play an important role in modulating BP cytotoxicity, wherein decreased BP-dione glucuronidation correlated highly with enhanced BP cytotoxicity (Fig. 8). BP diones (also referred to as BP quinones) have been shown to be cytotoxic by undergoing oxidation-reduction cycling (Bolton et al., 2000), in which toxic superoxide and semiquinone radicals are generated, causing inhibition of cell growth (Lorentzen and T'So, 1977) and DNA strand breaks (Lorentzen et al., 1980). In such redox cycling, the BP quinols could be removed by glucuronidation, thereby reducing the concentration of quinones and consequentially preventing the toxic redox cycling. Thus, the BP-diones appear to contribute to BP-initiated cytotoxicity via the formation of reactive oxygen species and oxidative stress, rather than via covalent binding to cellular macromolecules. Similar to its apparently neutral role in BP covalent binding, BP-monophenol glucuronidation did not modulate BP cytotoxicity, consistent with its negligible mutagenicity (Levin et al., 1978). Using human lymphoblastoid cells transfected with rat UGT1A7, Grove and coworkers observed glucuronidation of and cytoprotection against BP quinones, but not BP-7,8-dihydrodiol or 3-OH-BP (Grove et al., 2000). These observations agree with the results reported herein except for BP-7,8-dihydrodiol, the glucuronidation of which, along with BP-4,5-diol, we found correlated with an increased UGT-modulated reduction in BP covalent binding, and deceased cytotoxicity. Our correlation likely was in part due to the use of an exogenous bioactivating system. Also, this discrepancy with the transfected cellular model may reflect unappreciated differences between lymphoblastoid cell lines and the lymphocytes freshly isolated from human subjects in our study, with respect both to the level and types of UGT isozymes expressed and to interindividual variability in our model. In conclusion, glucuronidation of BP metabolites played an important cytoprotective and genoprotective role in reducing BP bioactivation and cytotoxicity in human lymphocytes. The lymphocytes were obtained from normal human subjects and showed a 200-fold interindividual variability in UGT activity for BP metabolites, with decreasing UGT activity correlating with a decreased UGT-modulated reduction in BP covalent binding, and with increased cytotoxicity. Since rat lymphocytes have been shown to accurately predict BP glucuronidation and its modulation of BP bioactivation and toxicity in vivo and in hepatic microsomes from the same animals, the data in human lymphocytes provide human evidence that UGTs may play an important cytoprotective role in modulating toxicities initiated by the electrophilic and free radical reactive intermediates of environmental chemicals such as the carcinogen/teratogen BP. The cytoprotective role afforded by the glucuronidation of BP metabolites is likely of particular importance for higher-risk populations, such as people living near hazardous waste sites, smokers, industrial (coke, aluminum plant) workers, garage mechanics, and people engaged in commercial or domestic cooking practices under poor ventilation who are exposed to polycyclic aromatic hydrocarbons (PAHs) through cooking fumes. FIG. 1. View largeDownload slide Postulated role of UDP-glucuronosyltransferases (UGTs) in the modulation of benzo[a]pyrene (BP) toxicity by reducing its bioactivation to an electrophilic reactive intermediate. UGT-catalyzed glucuronidation may similarly reduce redox cycling of hydroxylated BP metabolites and prevent the formation of toxic reactive oxygen species (not shown) (Bolton et al., 2000; Kim and Wells, 1996). FIG. 1. View largeDownload slide Postulated role of UDP-glucuronosyltransferases (UGTs) in the modulation of benzo[a]pyrene (BP) toxicity by reducing its bioactivation to an electrophilic reactive intermediate. UGT-catalyzed glucuronidation may similarly reduce redox cycling of hydroxylated BP metabolites and prevent the formation of toxic reactive oxygen species (not shown) (Bolton et al., 2000; Kim and Wells, 1996). FIG. 2. View largeDownload slide Time-dependent benzo[a]pyrene (BP) cytotoxicity in human lymphocytes. Lymphocyte preincubtion and conditions are described in the Materials and Methods. Asterisks indicate a difference compared to respective controls either without preincubation or with only dimethyl sulphoxide (DMSO) vehicle (n = 3, p < 0.05). FIG. 2. View largeDownload slide Time-dependent benzo[a]pyrene (BP) cytotoxicity in human lymphocytes. Lymphocyte preincubtion and conditions are described in the Materials and Methods. Asterisks indicate a difference compared to respective controls either without preincubation or with only dimethyl sulphoxide (DMSO) vehicle (n = 3, p < 0.05). FIG. 3. View largeDownload slide Glucuronidation of total benzo[a]pyrene (BP) metabolites in human lymphocytes. FIG. 3. View largeDownload slide Glucuronidation of total benzo[a]pyrene (BP) metabolites in human lymphocytes. FIG. 4. View largeDownload slide Relation of benzo[a]pyrene (BP) glucuronidation to the magnitude of reduction in BP covalent binding by UDP-glucuronosyltransferase (UGT) activity (UGT-modulated binding). UGT-modulated covalent binding was calculated as the reduction in BP covalent binding achieved by the addition of the UGT cosubstrate UDP-glucuronic acid (UDPGA) (insert). The insert presents modulation of BP covalent binding by UGTs in lymphocytes from 12 human subjects. The asterisks indicate a difference compared to respective control lymphocytes from the same subject incubated without UDPGA (n = 3, p < 0.05). FIG. 4. View largeDownload slide Relation of benzo[a]pyrene (BP) glucuronidation to the magnitude of reduction in BP covalent binding by UDP-glucuronosyltransferase (UGT) activity (UGT-modulated binding). UGT-modulated covalent binding was calculated as the reduction in BP covalent binding achieved by the addition of the UGT cosubstrate UDP-glucuronic acid (UDPGA) (insert). The insert presents modulation of BP covalent binding by UGTs in lymphocytes from 12 human subjects. The asterisks indicate a difference compared to respective control lymphocytes from the same subject incubated without UDPGA (n = 3, p < 0.05). FIG. 5. View largeDownload slide Relation of benzo[a]pyrene (BP)-diol glucuronidation to UGT-modulated covalent binding. Total BP-diol glucuronidation was the sum of BP-7,8-diol and BP-4,5-diol glucuronides. FIG. 5. View largeDownload slide Relation of benzo[a]pyrene (BP)-diol glucuronidation to UGT-modulated covalent binding. Total BP-diol glucuronidation was the sum of BP-7,8-diol and BP-4,5-diol glucuronides. FIG. 6. View largeDownload slide Relation of total benzo[a]pyrene (BP) glucuronidation to BP cytotoxicity. Lymphocytes were incubated with UDPGA and the total glucuronide conjugates of BP metabolites were measured. The insert presents the effect of UDPGA addition on the cytotoxicity of BP. Asterisks indicate a difference from respective control lymphocytes from the same subject incubated without UDPGA (p < 0.05). FIG. 6. View largeDownload slide Relation of total benzo[a]pyrene (BP) glucuronidation to BP cytotoxicity. Lymphocytes were incubated with UDPGA and the total glucuronide conjugates of BP metabolites were measured. The insert presents the effect of UDPGA addition on the cytotoxicity of BP. Asterisks indicate a difference from respective control lymphocytes from the same subject incubated without UDPGA (p < 0.05). FIG. 7. View largeDownload slide Relation of benzo[a]pyrene (BP)-diol glucuronidation to BP cytotoxicity. Total BP-diol glucuronidation was the sum of BP-7,8-diol and BP-4,5-diol glucuronides. FIG. 7. View largeDownload slide Relation of benzo[a]pyrene (BP)-diol glucuronidation to BP cytotoxicity. Total BP-diol glucuronidation was the sum of BP-7,8-diol and BP-4,5-diol glucuronides. FIG. 8. View largeDownload slide Relation of benzo[a]pyrene (BP)-dione glucuronidation to UGT-modulated cytotoxicity. BP-dione glucuronidation was the sum of BP-3,6-dione and BP-6,12-dione glucuronides. FIG. 8. View largeDownload slide Relation of benzo[a]pyrene (BP)-dione glucuronidation to UGT-modulated cytotoxicity. BP-dione glucuronidation was the sum of BP-3,6-dione and BP-6,12-dione glucuronides. Preliminary reports of this work were presented at the annual meeting of the Society of Toxicology in Seattle, WA, February 1992 and the 5th North American meeting of the International Society for the Study of Xenobiotics (ISSX), Tucson, Arizona, in 1993. The research was supported by grants from the Medical Research Council of Canada and the McNeil Consumer Products Co. (Guelph, Ontario). Z. H. was supported in part by a Mitchell Scholarship in Cancer Research from the University of Toronto. Current address for Z. H.: Research Institute for Liver Diseases. (Shanghai), Co. Ltd., 328 Bibo Road C101-113, Shanghai, China 201203. The authors are grateful to Dr. Louise M. Winn for final preparation of the figures. REFERENCES Bock, K. W. 1991. Role of UDP-glucuronosyltransferases in chemical carcinogenesis. Crit. Rev. Biochem. Mol. Biol.  26, 129–150. Google Scholar Bock, K. W., Lilienblum, W., and Pfeil, H. 1980. Conversion of benzo[a]pyrene-3,6-quinone to quinol glucuronides with rat liver microsomes or purified NADPH-cytochrome c reductase and UDP-glucuronosyl transferase. FEBS Lett.  121, 269–274. Google Scholar Bolton, J. L., Trush, M. A., Penning, T. M., Dryhurst, G., and Monks, T. J. 2000. Role of quinones in toxicology. Chem. Res. Toxicol.  13, 135–160. Google Scholar Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem.  72, 248–254. Google Scholar de Morais, S. M. F., Chow, S. Y. M., and Wells, P. G. 1992. Biotransformation and toxicity of acetaminophen in congenic RHA rats with or without a hereditary deficiency in bilirubin glucuronosyltransferase. Toxicol. Appl. Pharmacol.  117, 81–87. Google Scholar de Morais, S. M. F., Uetrecht, J. P., and Wells, P. G. 1992. Decreased glucuronidation and increased bioactivation of acetaminophen in Gilbert's syndrome. Gastroenterology  102, 577–586. Google Scholar de Morais, S. M. F., and Wells, P. G. 1988. Deficiency in bilirubin UDP-glucuronyl transferase as a genetic determinant of acetaminophen toxicity. J. Pharmacol. Exp. Ther.  247, 323–331. Google Scholar de Morais, S. M. F., and Wells, P. G. 1989. Enhanced acetaminophen toxicity in rats with bilirubin glucuronyl transferase deficiency. Hepatology  10, 163–167. Google Scholar Dipple, A. 1985. Polycyclic aromatic hydrocarbon carcinogenesis. J. Am. Chem. Soc.  45, 1–15. Google Scholar Grove, A. D., Llewellyn, G. C., Kessler, F. K., White, K. L., Crespi, C. L., and Ritter, J. K. 2000. Differential protection by rat UDP-glucuronosyltransferase 1A7 against benzo[a]pyrene-3,6-quinone- versus benzo[a]pyrene-induced cytotoxic effects in human lymphoblastoid cells. Toxicol. Appl. Pharmacol.  162, 34–43. Google Scholar Guengerich, F. P. 1991. Interindividual variation in biotransformation of carcinogens: Basis and relevance. In Molecular Dosimetry and Human Cancer: Analytical, Epidemicological and Social Considerations (J. D. Groopman and P. L. Skipper, Eds.), pp. 27–52. CRC Press, Boca Raton, FL. Google Scholar Guillemette, C. G. 2003. Pharmacogenomics of human UDP-glucuronosyltransferase enzymes. Pharmacogenomics J.  3, 136–158. Google Scholar Harris, C. C. 1991. Molecular Epidemiology: Overview of biochemical and molecular basis. In Molecular Dosimetry and Human Cancer: Analytical, Epidemiological and Social Considerations (J. D. Groopman and P. L. Skipper, Eds.), pp. 15–26. CRC Press, Boca Raton, FL. Google Scholar Harris, C. C., Autrup, H., Stoner, G. D., Trump, B. F., Hillman, E., Schafer, P. W., and Jeffery, A. M. 1979. Metabolism of benzo[a]pyrene, N-nitroso-dimethylamine and N-nitrosopyrrolidine and identification of the major carcinogen-DNA adducts formed in cultured human esophagus. Cancer Res.  39, 4401–4406. Google Scholar Hinman, C. L., Jiang, X. L., and Tang, H. P. 1990. Selective cytolysis by protein toxin as a consequence of direct interaction with the lymphocyte plasma membrane. Toxicol. Appl. Pharmacol.  104, 290–300. Google Scholar Hu, Z., and Wells, P. G. 1992. In vitro and in vivo biotransformation and covalent binding of benzo(a)pyrene in rats with a genetic deficiency in bilirubin UDP-glucuronosyltransferase. J. Pharmacol. Exp. Ther.  263, 334–342. Google Scholar Hu, Z., and Wells, P. G. 1994. Modulation of benzo[a]pyrene bioactivation by glucuronidation in peripheral blood lymphocytes from rats with a genetic deficiency in bilirubin UDP-glucuronosyltransferases. Toxicol. Appl. Pharmacol.  127, 306–313. Google Scholar Iyanagi, T., Emi, Y., and Ikushiro, S. 1998. Biochemical and molecular aspects of genetic disorders of bilirubin metabolism. Biochim. Biophys. Acta  1407, 173–184. Google Scholar Jahnke, C. D., Thompson, C. L., Walker, M. P., Gallagher, J. E., Lucier, G. W., and DiAugustine, R. P. 1990. Multiple DNA adducts in lymphocytes of smokers and nonsmokers determined by 32P-postlabeling analysis. Carcinogenesis  11, 205–211. Google Scholar Kadlubar, F. F., and Hammons, G. J. 1987. The role of cytochrome P450 in the metabolism of chemical carcinogens. In Mammalian Cytochromes P-450 (F. P. Guengerich, Ed.), Vol. 2, pp. 81–130. CRC Press, Boca Raton, FL. Google Scholar Kadlubar, F. F., Miller, J. A., and Miller, E. C. 1977. Hepatic microsomal N-glucuronidation and nucleic acid binding of N-hydroxy arylamines in relation to urinary bladder carcinogenesis. Cancer Res.  37, 805–811. Google Scholar Kim, P. M., and Wells, P. G. 1996. Genoprotection by UDP-glucuronosyltransferases in peroxidase-dependent, reactive oxygen species-mediated micronucleus initiation by the carcinogens 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) and benzo[a]pyrene. Cancer Res.  56, 1526–1532. Google Scholar Kim, P. M., and Wells, P. G. 1998. Phenytoin embryotoxicity: Protection by UDP-glucuronosyltransferases. Toxicol. Sci.  42, 261 (No. 1288). Google Scholar Kim, P. M., DeBoni, U., and Wells, P. G. 1997. Peroxidase-dependent bioactivation and oxidation of DNA and protein in benzo[a]pyrene-initiated micronucleus formation. Free Radic. Biol. Med.  23, 579–596. Google Scholar Kim, P. M., Winn, L. M., Parman, T., and Wells, P. G. 1997. UDP-glucuronosyltransferase-mediated protection against in vitro DNA oxidation and micronucleus formation initiated by phenytoin and its embryotoxic metabolite 5-(p-hydroxyphenyl)-5-phenylhydantoin (HPPH). J. Pharmacol. Exp. Ther.  280, 200–209. Google Scholar Kwei, G. Y., Zaleski, J., Irwin, S. E., Thurman, R. G., and Kauffman, F. C. 1992. Conjugation of benzo(a)pyrene-7,8-dihydrodiol-9,10-epoxide in infant Swiss-Webster mice. Cancer Res.  52, 1639–1642. Google Scholar Levin, W., Wood, A. W., Wislocki, P. G., Chang, R. L., Kapitulnik, J., Mak, H. D., Vagi, H., Jerina, D. M., and Conney, A. H. 1978. Mutagenicity and carcinogenicity of benzo[a]pyrene and benzo[a]pyrene derivatives. In Polycylic Hydrocarbons and Cancer (H. V. Gelboin and P. O. P. T'So, Eds.), Vol. 1, pp. 189–202. Academic Press, New York. Google Scholar Lind, C. 1985. Relationship between the rate of reduction of benzo[a]pyrene-3,6-quinone and the formation of benzo[a]pyrene-3,6-glucuronides in rat liver microsomes. Biochem. Pharmacol.  34, 895–903. Google Scholar Lind, C., Vadi, H., and Ernster, L. 1978. Metabolism of benzo[a]pyrene in liver microsomes from 3-methylcholanthrene treated rats. Arch. Biochem. Biophys.  190, 97–104. Google Scholar Lorentzen, R. J., and T'So, P. O. P. 1977. Benzo[a]pyrenedione/benzo[a]pyrenediol oxidation-reduction couples and the generation of reactive reduced molecular oxygen. Biochemistry (Mosc.)  16, 1467–1472. Google Scholar Lorentzen, R. J., Lesko, S. A., McDonald, K., and T'So, P. O. P. 1980. Toxicity of metabolic benzo[a]pyrenediones to cultured cells and the dependence upon molecular oxygen. Cancer Res.  39, 3194–3199. Google Scholar Mackenzie, P. I., Owens, I. S., Burchell, B., Bock, K. W., Bairoch, A., Belanger, A., Fournel-Gigleux, S., Green, M., Hum, D. W., Iyanagi, T., et al. 1997. The UDP glucuronosyltransferase gene superfamily: Recommended nomenclature based on evolutionary divergence. Pharmacogenetics  7, 255–269. Google Scholar Perneger, T. V., Whelton, P. K., and Klag, M. J. 1994. Risk of kidney failure associated with the use of acetaminophen, and nonsteroidal use aspirin antiinflammatory drugs. N. Engl. J. Med.  331, 1675–1679. Google Scholar Ramesh, A., Inyang, F., Hood, D. B., Archibong, A. E., Knuckles, M. E., and Nyanda, A. M. 2001. Metabolism, bioavailability and toxicokinetics of benzo[a]pyrene in F344 rats following oral administration. Exp. Toxicol. Pathol.  53, 275–290. Google Scholar Roy Chowdhury, J. R., Wolkoff, A. W., Roy Chowdhury, N., and Arias, I. M. 2001. Hereditary jaundice and disorders of bilirubin metabolism. In The Metabolic and Molecular Bases of Inheritied Disease (C. R. Scriver, A. L. Beaudet, W. S. Sly, and D. Valle, Eds.), Vol. 2, pp. 3063–3101. McGraw-Hill, New York. Google Scholar Shamsuddin, A. K. M., Sinopololi, N. T., Hemminki, K., Boesch, R. R., and Harris, C. C. 1985. Detection of benzo[a]pyrene: DNA adducts in human white blood cells. Cancer Res.  45, 66–68. Google Scholar Tukey, R. H., and Strassburg, C. P. 2000. Human UDP-glucuronosyltransferases: Metabolism, expression and disease. Annu. Rev. Pharmacol. Toxicol.  40, 581–616. Google Scholar Vienneau, D. S., DeBoni, U., and Wells, P. G. 1995. Potential genoprotective role for UDP-glucuronosyltransferases in chemical carcinogenesis: Initiation of micronuclei by benzo[a]pyrene and benzo[e]pyrene in UGT-deficient rat skin fibroblasts. Cancer Res.  55, 1045–1051. Google Scholar Wells, P. G., Mackenzie, P. I., Roy Chowdhury, J., Guillemette, C., Gregory, P. A., Ishii, Y., Hansen, A. J., Kessler, F. K., Kim, P. M., Roy Chowdhury, N., et al. 2004. Glucuronidation and the UDP-glucuronosyltransferases in drug therapy and disease. Drug Metab. Dispos.,  in press. Google Scholar Wells, P. G., Obilo, F. C., and de Morais, S. M. F. 1989. Benzo[a]pyrene embryopathy in rats genetically deficient in bilirubin UDP-glucuronyltransferase. FASEB J.  3, A1025. Google Scholar Whitcomb, D. C., and Block, G. D. 1994. Association of acetaminophen hepatotoxicity with fasting and ethanol use. J. Am. Med. Assoc.  272, 1845–1850. Google Scholar Wood, A. W., Levin, W., Chang, R. L., Vagi, H., Thakker, D. R., Lehr, R. E., Jerina, D. M., and Conney, A. H. 1979. Bay-region activation of carcinogenic polycyclic hydrocarbons. In Polynuclear Aromatic Hydrocarbons (P. W. Jones and P. Leber, Eds.), pp. 531–540. Ann Arbor Science Publishers, Ann Arbor, MI. Google Scholar Author notes *Department of Pharmacology and †Faculty of Pharmacy, University of Toronto, Toronto, Ontario, Canada M5S 2S2 Society of Toxicology
TI - Human Interindividual Variation in Lymphocyte UDP-Glucuronosyltransferases as a Determinant of In Vitro Benzo[a]pyrene Covalent Binding and Cytotoxicity
JF - Toxicological Sciences
DO - 10.1093/toxsci/kfh010
DA - 2004-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/human-interindividual-variation-in-lymphocyte-udp-4VKY50w2Z5
SP - 32
EP - 40
VL - 78
IS - 1
DP - DeepDyve
ER -