Glutathione Conjugation of Bay- and Fjord-Region Diol Epoxides of

In the present study, individual stereoisomers of the bay-region diol epoxides of ..... a Catalytic efficiency (kcat/Km) was calculated as described i...
1 downloads 0 Views 262KB Size
Download PDF
Chem. Res. Toxicol. 1997, 10, 1221-1227

1221

Articles Glutathione Conjugation of Bay- and Fjord-Region Diol Epoxides of Polycyclic Aromatic Hydrocarbons by Glutathione Transferases M1-1 and P1-1 Kathrin Sundberg,† Mikael Widersten,‡ Albrecht Seidel,§ Bengt Mannervik,‡ and Bengt Jernstro¨m*,† Institute of Environmental Medicine, Division of Biochemical Toxicology, Karolinska Institutet, Box 210, S-17177 Stockholm, Sweden, Department of Biochemistry, Uppsala University, Biomedical Center, Box 576, S-75123 Uppsala, Sweden, and Institute of Toxicology, University of Mainz, Obere Zahlbacher Strasse 67, D-55131 Mainz, Germany Received June 6, 1997X

Metabolism of polycyclic aromatic hydrocarbons in mammalian cells results in the formation of vicinal diol epoxides considered as ultimate carcinogens if the oxirane ring is located in a bay- or fjord-region of the parent compound. In the present study, individual stereoisomers of the bay-region diol epoxides of chrysene, dibenz[a,h]anthracene, and benzo[a]pyrene as well as of the fjord-region diol epoxides of benzo[c]phenanthrene, benzo[c]chrysene, and benzo[g]chrysene have been incubated with GSH in the presence of human glutathione transferases GSTM1-1 (a mu-class enzyme) and GSTP1-1 (a pi-class enzyme). As previously shown with GSTA1-1 (an alpha-class enzyme) both M1-1 and P1-1 demonstrate considerable activity toward a number of the diol epoxides studied, although a great variation in catalytic efficiency and enantioselectivity was observed. With GSTM1-1, the bay-region diol epoxides, in particular the syn-diastereomers were in most cases more efficiently conjugated with GSH than the fjordregion analogues. GSTM1-1 demonstrated an enantioselectivity ranging from no preference (50%) to high preference (g90%) for conjugation of the enantiomers with R-configuration at the benzylic position of the oxirane ring. With GSTP1-1, the enzyme demonstrated appreciable activity toward both bay- and fjord-region diol epoxides and, in most cases, a preference for the anti-diastereomers. In contrast to GSTM1-1 and as previously shown for GSTA1-1, GSTP1-1 showed an exclusive preference for conjugation of the enantiomers with Rconfiguration at the benzylic oxirane carbon. With both GSTM1-1 and GSTP1-1, the chemically most reactive diol epoxide, the (+)-syn-enantiomer of trans-7,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (BPDE), was the best substrate. As for GSTA1-1, no obvious correlation between chemical reactivity or lipophilicity of the compounds and catalytic efficiencies was observed. Molecular modeling of diol epoxides in the active sites of GSTP1-1 and -A1-1 is in agreement with the assumption, based on functional studies, that the H-site of GSTA1-1 [Jernstro¨m et al. (1996) Carcinogenesis 17, 1491-1498] can accommodate stereoisomers of different sizes. Further, modeling of the enantiomers of anti- and syn-BPDE in the active site of GSTP1-1 provides an explanation for the exclusive preference for the enantiomers with R-configuration at the benzylic oxirane carbon. These isomers could be snuggly fitted in the H-site close to the GSH sulfur, whereas those with opposite stereochemistry could not.

Introduction (PAH)1

Polycyclic aromatic hydrocarbons are common mutagenic and carcinogenic environmental pollutants, which require metabolic activation and subsequent adduct formation with nuclear DNA in order to be biologically active (2, 3). Several pure PAH have been studied in great detail using various mammalian systems, and the results clearly implicate sterically hindered bay- or fjord-region diol epoxides as the ultimate mutagenic and carcinogenic intermediates (4-6). * To whom correspondence should be addressed. Phone: ++46-87287576. Fax: ++46-8-334467. E-mail: [email protected]. † Karolinska Institutet. ‡ Uppsala University. § University of Mainz. X Abstract published in Advance ACS Abstracts, October 1, 1997.

S0893-228x(97)00099-4 CCC: $14.00

Studies of a number of bay-region diol epoxides in mammalian systems have shown that those associated with high mutagenic and tumorigenic activity are in general the anti-diastereomers, in particular the enantiomers with (R,S)-diol (S,R)-epoxide absolute configuration (4-6). For fjord-region diol epoxides the situation seems to be more complex. These diol epoxides demonstrate in many cases a relatively low solvolytic reactivity but are in general more biologically potent than bayregion diol epoxides (7-14). Moreover, both enantiomers of the syn- and anti-diastereomers can exhibit high mutagenic (7) and carcinogenic activity (8). Available data indicate that glutathione transferase (GST)-catalyzed conjugation of diol epoxides with GSH is important in cellular protection against these genotoxic © 1997 American Chemical Society

1222 Chem. Res. Toxicol., Vol. 10, No. 11, 1997

compounds. The efficiency of the detoxication system is determined by the GSH level and, in particular, the amount and the nature of the GST isoenzymes present (15-22). Furthermore, lack of certain GST isoenzymes, as well known for isoenzymes belonging to GST mu- and theta-classes, or differences in the activity and distribution of allelic variants of GSTP1-1 have been implicated in increased risk of tumor formation following exposure to PAH-containing material (23-25). Previous results with purified GST isoenzymes from rat and human tissues and anti-diol epoxides from benzo[a]pyrene (BP), benz[a]anthracene, and chrysene have shown that isoenzymes of the pi- and mu-classes exhibit considerable activity toward these diastereomers and in particular toward the most mutagenic and tumorigenic enantiomers with (R,S)-diol (S,R)-epoxide absolute configuration (17-21). More recent studies have demonstrated that the fjord-region diol epoxides of benzo[c]phenanthrene are also substrates for human GSTs. Interestingly, with these diol epoxides GSTA1-1 is the most active isoenzyme (26). Furthermore, a pronounced preference for S-glutathionylation of the enantiomers with R-configuration at the benzylic position of the oxirane ring has been observed (26). To gain further insights into the substrate selectivity, we recently initiated a more systematic investigation of the activities of human GSTs toward a number of stereoisomeric PAH diol epoxides. In the first part of the project GSTA1-1 and GSH were incubated with several bayand fjord-region diol epoxides, including all possible enantiomers, in order to determine the extent of the conjugation reaction and to identify factors influencing the catalytic efficiency. It was concluded that GSTA1-1 is catalyzing GSH conjugate formation of all diol epoxides employed (bay-region diol epoxides of chrysene, and dibenz[a,h]anthracene as well as benzo[a]pyrene and fjordregion diol epoxides of benzo[c]phenanthrene, benzo[c]chrysene and benzo[g]chrysene) although a great variation in catalytic efficiency was observed. With all anti-diol epoxides and bay-region syn-diol epoxides, a significant preference for conjugation of the enantiomer with Rconfiguration at the benzylic position of the oxirane ring 1 Abbreviations: B[c]CDE, trans-9,10-dihydroxy-11,12-epoxy-9,10,11,12-tetrahydrobenzo[c]chrysene; (+)-anti-B[c]CDE, benzo[c]chrysene(9S,10R)-diol (11R,12S)-oxide; (-)-anti-B[c]CDE, benzo[c]chrysene(9R,10S)-diol (11S,12R)-oxide; (+)-syn-B[c]CDE, benzo[c]chrysene(9S,10R)-diol (11S,12R)-oxide; (-)-syn-B[c]CDE, benzo[c]chrysene(9R,10S)-diol (11R,12S)-oxide; B[g]CDE, trans-11,12-dihydroxy-13,14epoxy-11,12,13,14-tetrahydrobenzo[c]chrysene; (+)-anti-B[g]CDE, benzo[g]chrysene-(11S,12R)-diol (13R,14S)-oxide; (-)-anti-B[g]CDE, benzo[g]chrysene-(11R,12S)-diol (13S,14R)-oxide; (+)-syn-B[g]CDE, benzo[g]chrysene-(11S,12R)-diol (13S,14R)-oxide; (-)-syn-B[g]CDE, benzo[g]chrysene-(11R,12S)-diol (13R,14S)-oxide; BP, benzo[a]pyrene; BPDE, trans-7,8-dihydroxy-9,10-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene; (+)anti-BPDE, benzo[a]pyrene-(7R,8S)-diol (9S,10R)-oxide; (-)-antiBPDE, benzo[a]pyrene-(7S,8R)-diol (9R,10S)-oxide; (+)-syn-BPDE, benzo[a]pyrene-(7S,8R)-diol (9S,10R)-oxide; (-)-syn-BPDE, benzo[a]pyrene-(7R,8S)-diol (9R,10S)-oxide; B[c]PhDE, 3,4-dihydroxy-1,2-epoxy1,2,3,4-tetrahydrobenzo[c]phenanthrene; (+)-anti-B[c]PhDE, benzo[c]phenanthrene-(3R,4S)-diol (1S,2R)-oxide; (-)-anti-B[c]PhDE, benzo[c]phenanthrene-(3S,4R)-diol (1R,2S)-oxide; (+)-syn-B[c]PhDE, benzo[c]phenanthrene-(3R,4S)-diol (1R,2S)-oxide; (-)-syn-B[c]PhDE, benzo[c]phenanthrene-(3S,4R)-diol (1S,2R)-oxide; CDE, trans-1,2-dihydroxy3,4-epoxy-1,2,3,4-tetrahydrochrysene; (+)-anti-CDE, chrysene-(1R,2S)diol (3S,4R)-oxide; (-)-anti-CDE, chrysene-(1S,2R)-diol (3R,4S)-oxide; (+)-syn-CDE, chrysene-(1S,2R)-diol (3S,4R)-oxide; (-)-syn-CDE, chrysene-(1R,2S)-diol (3R,4S)-oxide; CDNB, 1-chloro-2,4-dinitrobenzene; DBADE, trans-3,4-dihydroxy-1,2-epoxy-1,2,3,4-tetrahydrodibenz[a,h]anthracene; (+)-anti-DBADE, dibenz[a,h]anthracene-(3S,4R)-diol (1R,2S)-oxide; (-)-anti-DBADE, dibenz[a,h]anthracene-(3R,4S)-diol (1S,2R)-oxide; (+)-syn-DBADE, dibenz[a,h]anthracene-(3R,4S)-diol (1R,2S)-oxide; (-)-syn-DBADE, dibenz[a,h]anthracene-(3S,4R)-diol (1S,2R)-oxide; GST, glutathione transferase; PAH, polycyclic aromatic hydrocarbons.

Sundberg et al.

was noted. Interestingly, with the bay-region syn-diol epoxides this substrate selectivity was reversed, thus resulting in a preference for the enantiomer with the S-configuration. No obvious correlation between catalytic effiency and parameters such as chemical reactivity or lipophilicity of the diol epoxides was observed. Accordingly, stereochemical factors, including the size and the geometry of the aromatic ring system and the preferred conformation of the diol epoxide, seem to be the major determinants for the rate of catalysis by GSTA1-1. As an extension of this earlier work we now report the catalytic efficiency and enantioselectivity of GSTM1-1 and GSTP1-1, the former isoenzyme present in adult liver, breast, kidney, and heart and the latter present in most extrahepatic tissues (27), toward individual stereoisomers of the bay-region diol epoxides trans-1,2-dihydroxy-3,4-epoxy-1,2,3,4-tetrahydrochrysene (CDE), trans3,4-dihydroxy-1,2-epoxy-1,2,3,4-tetrahydrodibenz[a,h]anthracene (DBADE), and trans-7,8-dihydroxy-9,10epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene (BPDE) and the fjord-region diol epoxides trans-3,4-dihydroxy-1,2-epoxy1,2,3,4-tetrahydrobenzo[c]phenanthrene (B[c]PhDE), trans9,10-dihydroxy-11,12-epoxy-9,10,11,12-tetrahydrobenzo[c]chrysene (B[c]CDE), and trans-11,12-dihydroxy-13,14epoxy-11,12,13,14-tetrahydrobenzo[g]chrysene (B[g]CDE).

Materials and Methods Caution: Diol epoxides from polycyclic aromatic hydrocarbons are carcinogens, and thus experimental handling must be carried out under special safety conditions, e.g., those outlined in the NCI guidelines. Chemicals. Synthesis of the racemic and optically active syn- and anti-diol epoxides of chrysene, benzo[a]pyrene, dibenz[a,h]anthracene, benzo[c]phenanthrene, benzo[c]chrysene, and benzo[g]chrysene has been performed according to literature methods (1). The purity of the compounds used was in general g95% as determined by HPLC. Chromatographic standards of various diol epoxide derivatives were obtained as described previously (1, 25). Glutathione Transferases. Recombinant human GSTM1-1 (the allelic variant b with Asn at position 173) and GSTP1-1 (the allelic variant with Ile at position 105) were obtained by expression in Escherichia coli (28) and further purified by a modified affinity chromatography method (29). The purified enzyme (5-10 mg/mL dissolved in 10 mM Tris‚HCl, 10 mM dithiothreitol, and 0.02% sodium azide, pH 7.8) was used freshly prepared or after storage at 4 °C. Prior to use, the required amount of enzyme was freed of dithiothreitol and sodium azide by passing through a NAP-10 column (Pharmacia, Uppsala, Sweden) previously washed and equilibrated with 50 mM Tris‚ HCl, pH 7.5. The titer of active enzyme following storage at 4 °C was determined with 1-chloro-2,4-dinitrobenzene (CDNB) as substrate (30). Incubations. Prior to each experiment, the activity of GSTM1-1 or GSTP1-1 was checked with CDNB. The known specific activity of each GST toward CDNB (27, 31) was used to calculate the amount of active protein. Enzyme corresponding to 50-500 µg of active protein/mL was incubated at 37 °C with 40 and 80 µM diol epoxide (added in dimethyl sulfoxide, final concentration 5%) and 5 mM GSH in 50 mM Tris‚HCl buffer, pH 7.5 (final volume 100 µL), as previously described (1). After centrifugation at high speed, the samples were analyzed for GSH conjugates by HPLC as previously described (1). Molecular Modeling. Molecular models of GSTA1-1 and GSTP1-1 were constructed with InsightII (Biosym/MSI, San Diego, CA) using the atomic coordinates 1GUH (32) and 1GSS (33), respectively. Molecular models of the PAH diol epoxides were constructed with InsightII/Discovery 2.9.7. The syn- and anti-enantiomers of BPDE were manually docked into the H-site of GSTP1-1 (Ile 105). The formed GST-BPDE molecular models

Human GST and Conjugation of PAH Diol Epoxides

Chem. Res. Toxicol., Vol. 10, No. 11, 1997 1223

Figure 1. Structures and numbering system of the bay-region (panel A, upper row) and fjord-region (panel A, lower row) polycyclic aromatic hydrocarbons from which the diol epoxides used in this study were derived. The relative and the absolute stereochemistry of the syn- and anti-diol epoxides (the prefix syn indicates that the oxirane ring and the benzylic hydroxyl group are located on the same face of the molecule, whereas anti indicates location of these groups on opposite faces) are shown in panel B. Table 1. Catalytic Efficiency for Human GSTA1-1, GSTM1-1, and GSTP1-1 with Various Optically Active Bay- and Fjord-Region Diol Epoxides of Selected Polycyclic Aromatic Hydrocarbons compound (+)-anti-CDE (-)-anti-CDE (+)-syn-CDE (-)-syn-CDE (+)-anti-BPDE (-)-anti-BPDE (+)-syn-BPDE (-)-syn-BPDE (+)-anti-DBADE (-)-anti-DBADE (+)-syn-DBADE (-)-syn-DBADE (+)-anti-B[c]PhDE (-)-anti-B[c]PhDE (+)-syn-B[c]PhDE (-)-syn-B[c]PhDE (+)-anti-B[c]CDE (-)-anti-B[c]CDE (+)-syn-B[c]CDE (-)-syn-B[c]CDE (+)-anti-B[g]CDE (-)-anti-B[g]CDE (+)-syn-B[g]CDE (-)-syn-B[g]CDE

absolute configuration 1R,2S,3S,4Rc 1S,2R,3R,4S 1S,2R,3S,4R 1R,2S,3R,4S 7R,8S,9S,10R 7S,8R,9R,10S 7S,8R,9S,10R 7R,8S,9R,10S 1R,2S,3S,4R 1S,2R,3R,4S 1R,2S,3R,4S 1S,2R,3S,4R 1S,2R,3R,4S 1R,2S,3S,4R 1R,2S,3R,4S 1S,2R,3S,4R 9S,10R,11R,12S 9R,10S,11S,12R 9S,10R,11S,12R 9R,10S,11R,12S 11S,12R,13R,14S 11R,12S,13S,14R 11S,12R,13S,14R 11R,12S,13R,14S

GSTM1-1a (60)d

2.08 ( 0.22 1.42 ( 0.03 3.53 ( 0.16 (66) 1.79 ( 0.04 10.5 ( 2.27 11.4 ( 2.03 (52) 23.2 ( 6.58 (80) 5.97 ( 1.63 1.81 ( 0.25 (90) 0.20 ( 0.07 4.85 ( 0.98 (77) 1.43 ( 0.61 0.04 ( 0.00 0.16 ( 0.02 (79) 0.51 ( 0.04 (81) 0.12 ( 0.01 0.13 ( 0.03 0.32 ( 0.05 (72) 0.57 ( 0.17 (86) 0.09 ( 0.02 0.39 ( 0.07 1.53 ( 0.03 (80) 3.24 ( 0.07 (96) 0.13 ( 0.02

kcat/Km (mM-1 s-1) GSTP1-1

GSTA1-1b

3.02 ( 0.35 (>99) nde 10.5 ( 0.29 (>99) nd 3.71 ( 0.46 (91) 0.38 ( 0.25 8.70 ( 0.72 (87) 1.27 ( 0.22 1.38 ( 0.36 (>99) nd 0.83 ( 0.36 (58) 0.60 ( 0.22 nd 0.47 ( 0.04 (>99) 0.10 ( 0.02 (>99) nd nd 2.29 ( 0.36 (>99) 1.06 ( 0.31 (>99) nd nd 1.80 ( 0.12 (>99) 1.11 ( 0.08 (>99) nd

1.6 (71) 0.7 1.4 2.7 (65) 3.2 (67) 1.6 7.6 14.0 (65) 2.9 (69) 1.3 3.4 5.9 (64) 0.6 1.4 (69) 4.8 (79) 1.2 1.3 3.8 (75) 7.1 (66) 3.7 0.7 2.3 (76) 6.5 (67) 3.3

a Catalytic efficiency (k /K ) was calculated as described in Results and Discussion. b Data taken from ref 1. c Configuration of the cat m GSH-attacked benzylic oxirane carbon is marked in bold. d Enantioselectivity in %. e nd ) not detectable.

were minimized as follows using the CVFF force field and a dielectricity constant set to 1 in all cases. The molecular assemblies were iteratively relaxed by steepest descent minimization by stepwise relaxing the hydrogens, the amino acid side chains, the backbones, and finally the R-carbons of the enzyme models. The relaxed assemblies were subsequently minimized first by steepest descent including charge terms followed by conjugate gradient minimization including charge and cross terms until the root mean square deviation of atoms was less than 1. These crudely minimized models were subjected to 1.94 ps of molecular dynamics at a temperature setting of 303 K. The resulting models were energetically minimized using conjugate gradient, including charge and cross terms until the systems reached a root mean square value of the atomic coordinates less than 0.008 (approximately 7500 iterations). The R-carbons of the minimized models were superimposed onto the 1GSS model in order to assess the deviation in the structures.

Table 2. Enantioselectivity of Human GSTs M1-1, P1-1, and A1-1 toward Bay- and Fjord-Region Diol Epoxide (DE) Stereoisomers compound bay region DE fjord region DE

diastereomer GSTM1-1 GSTP1-1 GSTA1-1 anti syn anti syn

Ra g 48 R g 66 R g 72 R g 81

R g 91 R g 99 R g 99 R g 99

R g 67 S g 64 R g 69 R g 66

a Denotes the absolute configuration of the benzylic position of the oxirane ring.

Results and Discussion In an extension of our ongoing work on the substrate enantioselectivity of human GSTs, we have incubated in the present study individual stereoisomers of a series of bay- and fjord-region diol epoxides (see Figure 1 for structures) with GSH in the presence or absence of

1224 Chem. Res. Toxicol., Vol. 10, No. 11, 1997

Sundberg et al.

Human GST and Conjugation of PAH Diol Epoxides

human glutathione transferase GSTM1-1 and an allelic variant of GSTP1-1 with Ile at position 105. Rate of conjugate formation and values of the catalytic efficiency (kcat/Km) were calculated as previously described (1, 26). The kcat/Km values obtained are compiled in Table 1. For comparison, the kcat/Km values obtained previously with GSTA1-1 (1) are also included in the table. The data indicate that both GSTM1-1 and GSTP1-1 demonstrate considerable activity toward a number of the diol epoxides studied, although a great variation in catalytic efficiency and the degree of enantioselectivity was observed. With GSTM1-1, the bay-region diol epoxides, in particular the syn-diastereomers were in most cases more efficiently conjugated with GSH than the fjord-region analogues. GSTM1-1 demonstrated an enantioselectivity ranging from no preference (50%) to high preference (g90%) for conjugation of the enantiomers with Rconfiguration at the benzylic position of the oxirane ring. With respect to GSTP1-1 (Table 1), the enzyme demonstrated an appreciable activity toward both bay- and fjord-region diol epoxides and in most cases a substrate preference for the anti-diastereomers. In contrast to GSTM1-1 and also observed for GSTA1-1 previously, GSTP1-1 exhibited in most cases an exclusive preference for S-glutathionylation of the enantiomers with R-configuration at the benzylic oxirane carbon. With both GSTM1-1 and GSTP1-1, the chemically most reactive diol epoxide investigated, i.e., (+)-syn-BPDE, was the best substrate. As previously shown for GSTA1-1 (1), no obvious correlation between chemical reactivity or lipophilicity of the compounds and catalytic efficiencies was observed with GSTs M1-1 and P1-1 (data not shown). The substrate preference of GSTs A1-1, M1-1, and P1-1 for the different optically active diol epoxide stereoisomers is summarized in Table 2. The crystal structures of human GSTA1-1 and GSTP1-1 have been determined (32, 33). Molecular modeling of diol epoxides in the active sites of GSTA1-1 and GSTP1-1 and the results of functional studies are in agreement with the assumption that the H-site of GSTA1-1 (1) can accommodate stereoisomers of different sizes. As shown previously, among the anti-diastereomers of the bayregion diol epoxides (CDE, BPDE, and DBADE) and the syn-diastereomers of the fjord-region diol epoxides GSTA1-1 displayed a moderate preference (about 2-fold) for the enantiomers with R-configuration at the benzylic oxirane carbon (1) (Table 1). However, with the syndiastereomers of the bay-region diol epoxides the preference was reversed. The shift in preference from the enantiomers with R-configuration to those with S-configuration may be due to the preferred orientation of the hydroxyl groups opposite the oxirane ring in bay- and fjord-region diol epoxides (pseudodiaxial and pseudodiequatorial, respectively). The results of molecular modeling of different diol epoxide stereoisomers in the active site of GSTA1-1 are compatible with this suggestion. Thus, in contrast to GSTA1-1, the orientation of the hydroxyl groups and thus the conformation of the diol

Chem. Res. Toxicol., Vol. 10, No. 11, 1997 1225

epoxide moiety have little impact on the catalytic activity of either GSTM1-1 or GSTP1-1. In order to further analyze the catalytic differences of GSTP1-1 with the enantiomers of anti- and syn-BPDE, energy minimization calculations of these diol epoxides modeled in the H-site of the enzyme were performed. The deviation in the position of the protein backbone of the minimized structures as compared to the X-ray diffraction structure 1GSS (33) was less than 1 Å within regions of secondary structure and less than 2 Å within exposed loop regions, values comparable to those observed between different X-ray structures of the same protein. The BPDE enantiomers with R-configuration at the benzylic oxirane carbon fitted well in the H-site in the minimized models of GSTP1-1 (Figure 2); π-π interactions between the Tyr at position 109 and the aromatic ring system of the diol epoxide are suggested by the modeling (Figure 2, bottom). In addition, the hydroxyl group of Tyr 109 may contribute to catalysis by favorable interactions with the arene oxide. The distance of 4.4 Å in the minimized model is longer than that of a hydrogen bond, but small conformational changes may position these groups closer to each other during catalysis, thus facilitating ring opening. This suggestion is supported by a recent study by Ji et al. (34). It was directly shown that the hydroxyl group of Tyr 109 in GSTP1-1(Val 105) was hydrogenbonded to the 10-hydroxyl group in (9R,10R)-9-(S-glutathionyl)-10-hydroxy-9,10-dihydrophenanthrene indicating that Tyr 109 facilitates the ring opening of the corresponding arene oxide. The interactions between GSTP1-1 and the BPDE enantiomers with R-configuration at the benzylic oxirane carbon were not observed with the corresponding enantiomers with S-configuration, which may explain the difference in the determined catalytic efficiencies. Ile at position 105, which forms part of the H-site in GSTP1-1, is positioned very close to the polar part of the diol epoxide molecule and may restrict optimal interaction between the glutathione thiolate anion and the oxirane carbon. Modification of this position may well affect the catalytic property of the enzyme. Previous studies have identified an allelic variant of GSTP1-1 in which Ile 105 is replaced by Val (35). We are currently investigating how modification at position 105 affects the catalytic efficiency toward diol epoxides since epidemiological data indicate that individuals with the Val 105 allele might be at a higher risk of developing tumors in organs susceptible to PAH exposure (24, 25). As discussed previously (1) the binding site for GSH (the G-site) is essentially conserved among the GST isoenzymes studied so far. GSH is, by ionic interactions and hydrogen bonds, in a well-defined orientation of the tripeptide (32). The second binding site (the H-site) is principally hydrophobic and more flexible in order to accommodate the electrophilic and lipid soluble compounds which are substrates for GST (27). The combined structural features of the G- and H-sites have certain implications with regard to the enantioselectivity of bayand fjord-region diol epoxides. All the anti-diastereomers

Figure 2. Molecular model of (+)-anti-BPDE in the active site of human GSTP1-1 (Ile 105). (Top) Space-filling model of (+)-antiBPDE (blue) docked into the hydrophobic compartment of the active site of the GSTP1-1 dimer. The GSH derivative, one per subunit, is shown in red. One subunit is shown in green and the other in yellow. Hydrophobic amino acid residues of the yellow subunit are shown in orange. (Middle) Same model as in the top panel rotated 90° perpendicular to the symmetry axis of the protein dimer. (Bottom) Detailed view of the (+)-anti-BPDE docked into the active site. The diol epoxide was modeled so as to align the benzylic arene carbon in a position suited for attack by the GSH sulfur. Further, the modeled structure minimized clashes between the diol epoxide and the enzyme. The images were constructed with InsightII using the atomic coordinates 1GSS for GSTP1-1.

1226 Chem. Res. Toxicol., Vol. 10, No. 11, 1997

are structurally very similar with regard to the preferred conformation of the diol epoxide part of the molecule in solution (e.g., pseudodiequatorial orientation of the hydroxyl groups). As shown previously (1) and in this study, GSTA1-1 and GSTM1-1 catalyze S-glutathionylation at the benzylic oxirane carbon with both R- and S-configurations although the S-configuration is associated with a lower reaction rate. This implies that the accommodation of these enantiomers in the H-site is restricted relative to their enantiomers with R-configuration. Assuming that the enzymatic reaction always takes place by trans-addition of GSH and only from one side in the active site of the enzyme, the enantiomers have to flip relative to each other in order to form the proper prereaction complex between the thiolate anion of GSH and the benzylic oxirane carbon. For GSTA1-1, we previously suggested that the enantiomers have to flip around an axis determined by the benzylic oxirane carbon and the hydroxylated benzylic carbon (1). The results from molecular modeling with GSTA1-1 and various diol epoxide isomers are compatible with this view. If and to what extent this model is applicable to GSTM1-1 is not known. It should be mentioned, however, that modeling of GSTA1-1 does not show any obvious interaction between the arene oxide and any polar amino acid residue although the catalytic activity is high toward several of the diol epoxide substrates. This is in contrast to what is expected for GSTM1-1. With this enzyme a tyrosine residue in the H-site is expected to facilitate ring opening by hydrogen bonding to the arene oxide (36). It is possible that the amino acid at position 208 in the H-site in GSTA1-1 substantially contributes to the high activity expressed by this isoenzyme (37). It is generally accepted that exposure to PAH is an etiological factor in human carcinogenesis. Furthermore, it is believed that the balance between the enzymatic processes is involved in the activation of PAH (e.g., CYP1A1) to their ultimate carcinogens (i.e., diol epoxides) and those involved in detoxification (e.g., GST in conjunction with GSH). In fact, available epidemiological information indicates that individuals lacking certain GST isoenzymes (i.e., isoenzymes belonging to GST mu- and theta-classes) or differ in the activity and distribution of allelic variants of GSTP1-1 may be at a higher risk of developing tumors in organs susceptible to PAH exposure (23-25, 38). The direct information on the protective role of individual GST isoenzymes against carcinogenic diol epoxides in biological systems is surprisingly scarce. However, studies in isolated cells have demonstrated that the intracellular level of GSH plays a crucial role on DNA binding or cytotoxicity of BP or its metabolites including racemic anti-BPDE (15, 39-41). Furthermore, it has been demonstrated in a cocultivation assay that mammalian cells containing GST isoenzymes active toward (+)-anti-BPDE inhibit the formation of mutations in recipient cells more effectively than cells devoid of this capacity (20). As has been shown for GSTP1-1, transient expression of recombinant GST in COS cells confers resistance to racemic anti-BPDE (21), and more recently it was found that V79 cells expressing human GSTP1-1 were significantly more resistant to mutagenicity of the bay-region anti- and syn-BPDE and the fjord-region anti2 A. Seidel, T. Friedberg, B. Lo ¨ llmann, A. Schwierzok, M. Funk, H. Frank, F. Oesch, and H. R. Glatt, manuscript submitted for publication.

Sundberg et al.

B[c]PhDE and anti-dibenzo[a,l]pyrene-11,12-diol 13,14epoxide enantiomers with R-configuration at the benzylic oxiranyl carbon relative to the corresponding enantiomers with S-configuration.2 Finally, Townsend et al. also using cultured mammalian cells transfected with human GSTP1-1 demonstrated a large reduction in DNA binding of (+)-anti-BPDE (42) and a significant protection against mutagenicity (22). Accordingly, the data taken together strongly implicate that GST-catalyzed conjugation of harmful diol epoxides with GSH plays an important role in counteracting the accumulation of these reactive intermediates, thus preventing or reducing the extent of DNA binding and, as a consequence, lowering the incidence of mutations and subsequent tumor initiation.

Acknowledgment. This study was supported by grants from the Swedish Tobacco Co., the National Board for Laboratory Animals, the Swedish Cancer Society, and the Deutsche Forschungsgemeinschaft (SFB 302).

References (1) Jernstro¨m, B., Funk, M., Frank, H., Mannervik, B., and Seidel, A. (1996) Glutathione transferase A1-1 catalyzed conjugation of bay- and fjord-region diol epoxides of polycyclic aromatic hydrocarbons with glutathione. Carcinogenesis 17, 1491-1498. (2) Sims, P., and Grover, P. L. (1974) Epoxides in polycyclic aromatic hydrocarbon metabolism and carcinogenesis. Adv. Cancer Res. 20, 165-275. (3) Thakker, D. R., Yagi, H., Levin, W., Wood, A. W., Conney, A. H., and Jerina, D. M. (1985) Polycyclic aromatic hydrocarbons: metabolic activation to ultimate carcinogens. In Bioactivation of Foreign Compounds (Anders, M. W., Ed.) pp 177-242, Academic Press, New York. (4) Dipple, A. (1985) Polycyclic aromatic hydrocarbon carcinogenesis: an introduction. In Polycyclic Hydrocarbons and Carcinogenesis (Harvey, R. G., Ed.) pp 1-17, ACS Symposium Series 283, American Chemical Society, Washington, DC. (5) Jerina, D. M., Chadha, A., Cheh, A. M., Schurdak, M. E., Wood, A. W., and Sayer, J. M. (1991) In Biological Reactive Intermediates IV (Witmer, C. M., Snyder, R., Jollow, D. J., Kalf, G. S., Kocsis, J. J., and Sipes, I. G., Eds.) pp 533-553, Plenum Publishing Corp., New York. (6) Harvey, R. G. (1991) In Polycyclic Aromatic Hydrocarbons: Chemistry and Carcinogenicity (Harvey, R. G., Ed.) pp 13651373, Cambridge University Press, Cambridge. (7) Wood, A. W., Chang, R. L., Levin, W., Thakker, D. R., Yagi, H., Sayer, J. M., Jerina, D. M., and Conney, A. H. (1984) Mutagenicity of the enantiomers of the diastereomeric bay-region benzo[c]phenanthrene 3,4-diol-1,2-epoxides in bacterial and mammalian cells. Cancer Res. 44, 2320-2324. (8) Levin, W., Chang, R. L., Wood, A. W., Thakker, D. R., Yagi, H., Jerina, D. M., and Conney, A. H. (1986) Tumorigenicity of optical isomers of the diastereomeric bay-region 3,4-diol-1,2-epoxides of benzo[c]phenanthrene in murine tumor models. Cancer Res. 46, 2257-2261. (9) Glatt, H., Pie´e, A., Pauly, K., Steinbrecher, T., Schrode, R., Oesch, F., and Seidel, A. (1991) Fjord- and bay-region diol epoxides investigated for stability, SOS induction in Escherichia coli, and mutagenicity in Salmonella typhimurium and mammalian cells. Cancer Res. 51, 1659-1667. (10) Phillips, D. H., Hewer, A., Seidel, A., Steinbrecher, T., Schrode, R., Oesch, F., and Glatt, H. (1991) Relationship between mutagenicity and DNA adduct formation in mammalian cells for fjordand bay-region diol epoxides of polycyclic aromatic hydrocarbons. Chem.-Biol. Interact. 80, 177-186. (11) Gill, H. S., Kole, P. L., Wiley, J. C., Li, K.-M., Higginbotham, S., Rogan, E. G., and Cavalieri, E. L. (1994) Synthesis and tumorinitiating activity in mouse skin of dibenzo[a,l]pyrene syn- and anti-fjord-region diolepoxides. Carcinogenesis 15, 2455-2460. (12) Luch, A., Glatt, H., Platt, K. L., Oesch, F., and Seidel, A. (1994) Synthesis and mutagenicity of the diastereomeric fjord-region 11,12-dihydrodiol 13,14-epoxides of dibenzo[a,l]pyrene. Carcinogenesis 15, 2507-2516. (13) Amin, S., Krzeminski, J., Rivenson, A., Kurtzke, C., Hecht, S. S., and El-Bayoumy, K. (1995) Mammary carcinogenicity in female

Human GST and Conjugation of PAH Diol Epoxides

(14)

(15)

(16)

(17)

(18)

(19)

(20)

(21) (22)

(23)

(24)

(25)

(26)

(27)

(28)

CD rats of fjord region diol epoxides of benzo[c]phenanthrene, benzo[g]chrysene, and dibenzo[a,l]pyrene. Carcinogenesis 16, 1971-1974. Amin, S., Desai, D., Dai, W., Harvey, R. G., and Hecht, S. S. (1995) Tumorigenicity in newborn mice of fjord region and other sterically hindered diol epoxides of benzo[g]chrysene, dibenzo[a,l]pyrene (dibenzo[def,p]chrysene), 4H-cyclopenta[def]chrysene and fluoranthene. Carcinogenesis 16, 2813-2817. Ho, D., and Fahl, W. E. (1984) Modification of glutathione levels in C3H/10T1/2 cells and its relationship to benzo[a]pyrene anti7,8-dihydrodiol 9,10-epoxide-induced cytotoxicity. J. Biol. Chem. 259, 11231-11235. Jernstro¨m, B., Martinez, M., Meyer, D. J., and Ketterer, B. (1985) Glutathione conjugation of the carcinogenic and mutagenic electrophile (()-7β,8R-dihydroxy-9R,10R-oxy-7,8,9,10-tetrahydrobenzo[a]pyrene catalyzed by purified rat liver glutathione transferases. Carcinogenesis 6, 37-47. Robertson, I. G., Jenssen, H., Mannervik, B., and Jernstro¨m, B. (1986) Glutathione transferases in rat lung: the presence of transferase 7-7, highly efficient in the conjugation of glutathione with the carcinogenic (+)-7β,8R-dihydroxy-9R,10R-oxy-7,8,9,10tetrahydrobenzo[a]pyrene. Carcinogenesis 7, 295-299. Robertson, I. G., Guthenberg, C., Mannervik, B., and Jernstro¨m, B. (1986) Differences in stereoselectivity and catalytic efficiency of three human glutathione transferases in the conjugation of glutathione with 7β,8R-dihydroxy-9R,10R-oxy-7,8,9,10-tetrahydrobenzo[a]pyrene. Cancer Res. 46, 2220-2224. Robertson, I. G., and Jernstro¨m, B. (1986) The enzymatic conjugation of glutathione with bay-region diol epoxides of benzo[a]pyrene, benz[a]anthracene and chrysene. Carcinogenesis 7, 16331636. Romert, L., Dock, L., Jenssen, D., and Jernstro¨m, B. (1989) Effects of glutathione transferase activity on benzo[a]pyrene 7,8-dihydrodiol metabolism and mutagenesis studied in a mammalian cell co-cultivation assay. Carcinogenesis 10, 1701-1707. Puchalski, R. B., and Fahl, W. E. (1990) Expression of recombinant glutathione S-transferase π, Ya, or Yb1 confers resistence to alkylating agents. Proc. Natl. Acad. Sci. U.S.A. 87, 2443-2447. Townsend, A., Fields, W. R., Karper, A. J., Doehmer, J., and Morrow, C. S. (1996) Chemoprotective functions of glutathione S-transferases. International Conference on Glutathione and Glutathione-Linked Enzymes in Human Cancer and Other Diseases, Abstract OP-42, Hilton Head, SC. Strange, R. C. (1996) Glutathione S-transferases and cancer susceptibility. In Glutathione S-Transferases, Structure, Function, and Clinical Implications (Vermeulen, N. P. E., Mulder, G. J., Nieuwenhuyse, N., Peters, W. H. M., and Van Bladeren, P. J., Eds.) pp 239-247, Taylor and Francis, London. Kadlubar, F. (1996) Polymorphisms involved in the bioactivation and detoxification of aromatic and heterocyclic amine carcinogens. 7th North American ISSX Meeting, Abstract No. 10, San Diego, CA. Harries, L. W., Stubbins, M. J., Forman, D., Howard, G. C. W., and Wolf, C. R. (1997) Identification of genetic polymorphisms at the glutathione S-transferase pi locus and association with the susceptibility to bladder, testicular and prostate cancer. Carcinogenesis 18, 641-644. Jernstro¨m, B., Seidel, A., Funk, M., Oesch, F., and Mannervik, B. (1992) Glutathione conjugation of trans-3,4-dihydroxy 1,2-epoxy 1,2,3,4-tetrahydrobenzo[c]phenanthrene isomers by human glutathione transferases. Carcinogenesis 13, 1549-1555. Mannervik, B., and Widersten, M. (1995) Human glutathione transferases: Classification, tissue distribution, structure and functional properties. In Advances in Drug Metabolism in Man (Pacifici, G. M., Fracchia, G. N., Eds.) pp 408-459, European Commission, Luxembourg. Stenberg, G., Bjo¨rnestedt, R., and Mannervik, B. (1992) Heterologous expression of recombinant human glutathione transferase A1-1 from a hepatoma cell line. Protein Express. Purif. 3, 80-84.

Chem. Res. Toxicol., Vol. 10, No. 11, 1997 1227 (29) Cameron, A. D., Sinning, I., L'Hermit, G., Olin, B., Board, P. G., Mannervik, B., and Jones, T. A. (1995) Structural analysis of human alpha-class glutathione transferase A1-1 in the apo-form and in complexes with ethacrynic acid and its glutathione conjugate. Structure 3, 717-727. (30) Habig, W. H., and Jakoby, W. B. (1981) Assays for differentiation of glutathione S-transferases. Methods Enzymol. 77, 398-405. (31) Mannervik, B., and Danielson, U. H. (1988) Glutathione transferases - structure and catalytic activity. C. R. C. Crit. Rev. Biochem. 23, 283-337. (32) Sinning, I., Kleywegt, G. J., Cowan, S. W., Reinemar, P., Dirr, H. W., Huber, R., Gilliland, G. L., Armstrong, R. N., Ji, X., Board, P. G., Olin, B., Mannervik, B., and Jones, T. A. (1993) Structure determination and refinement of human alpha class glutathione transferase A1-1, and a comparison with the mu and pi class enzymes. J. Mol. Biol. 232, 192-212. (33) Reinemer, P., Dirr, H. W., Ladenstein, R., Huber, R., Lo Bello, M., Federici, G., and Parker, M. W. (1992) Three-dimensional structure of class pi glutathione S-transferase from human placenta in complex with S-hexylglutathione at 2.8Å resolution. J. Mol. Biol. 227, 214-226. (34) Ji, X., O′Donnell, R., Gilliland, G. L., Tordova, M., Armstrong, R. N., Parsons, J. F., Zimniak, P., and Hayden, J. B. (1996) A direct comparison of xenobiotic binding site between class mu and pi glutathione S-transferases. International Conference on Glutathione and Glutathione-Linked Enzymes in Human Cancer and Other Diseases, Abstract PP-6, Hilton Head, SC. (35) Ahmad, H., Wilson, D. E., Fritz, R. R., Singh, S. V., Medh, R. D., Nagle, G. T., Awasthi, Y. C., and Kurosky, A. (1990) Primary and secondary structural analyses of glutathione S-transferase pi from human placenta. Arch. Biochem. Biophys. 278, 398-408. (36) Ji, X., Johnson, W. W., Sesay, M. A., Dickert, L., Prasad, S. M., Ammon, H. L., Armstrong, R. N., and Gilliland, G. L. (1994) Structure and function of the xenobiotic substrate binding site of a glutathione S-transferase as revealed by X-ray crystallographic analysis of product complexes with the diastereomers of 9-(Sglutathionyl)-10-hydroxy-9,10-dihydrophenanthrene. Biochemistry 33, 1043-1052. (37) Widersten, M., Bjo¨rnestedt, R., and Mannervik, B (1994) Contribution of amino acid residue 208 in the hydrophobic binding site to the catalytic mechanism of human glutathione transferase A11. Biochemistry 33, 11717-11723. (38) Smith, G., Stanley, L. A., Strange, R. C., and Wolf, C. R. (1995) Metabolic polymorphisms and cancer susceptibility. Cancer Surveys 25, 27-65. (39) Burke, M. D., Vadi, H., Jernstro¨m, B., and Orrenius, S. (1977) Metabolism of benzo[a]pyrene with isolated hepatocytes and the formation and degradation of DNA-binding derivatives. J. Biol. Chem. 252, 6424-6431. (40) Shen, A. L., Fahl, W. E., and Jefcoate, C. R. (1980) Metabolism of benzo[a]pyrene by isolated hepatocytes and factors affecting covalent binding of benzo[a]pyrene metabolites in hepatocytes and microsomal systems. Arch. Biochem. Biophys. 204, 511-523. (41) Jernstro¨m, B., Babson, J. R., Molde´us, P., Holmgren, A., and Reed, D. J. (1982) Glutathione conjugation and DNA-binding of (()trans-7,8-dihydroxy, 7,8-dihydrobenzo[a]pyrene and (()-7b,8adihydroxy-9a,10a-epoxy-7,8,9,10-tetrahydrobenzo[a]pyrene in isolated rat hepatocytes. Carcinogenesis 3, 861-866. (42) Fields, W. R., and Townsend, A. J. (1995) Transfected human and murine glutathione S-transferases protect against DNA modification by [3H]benzo[a]pyrene diol-epoxide or [3H]aflatoxin B1 in V79 cells. ISSX Proceedings, 1995 International ISSX-Workshop on Glutathione S-Transferases, Noordwijkerhout, The Netherlands, Vol. 7, p 28, ISSX, Bethesda, MD.

TX970099W