Previous studies have identified allelic variants of the human glutathione transferase (GST) Pi gene and showed that the two different encoded proteins with isoleucine (GSTP1-1/I-105) or valine (GSTP1-1/V-105) at position 105, respectively, differ significantly in their catalytic activities with model substrates. Moreover, recent epidemiological studies have demonstrated that individuals differing in the expression of these allelic variants also differ in susceptibility to tumour formation in certain organs, including such in which polycyclic aromatic hydrocarbons (PAH) may be etiological factors. In the present study the catalytic efficiencies (kcat/Km) of these GSTP1-1 variants were determined with a number of stereoisomeric bay-region diol epoxides, known as the ultimate mutagenic and carcinogenic metabolites of PAH, including those from chrysene, benzo[a]pyrene and dibenz[a,h]anthracene. In addition, GSTP1-1 mutants in which amino residue 105 is alanine (GSTP1-1/A-105) or tryptophan (GSTP1-1/W-105) have been constructed and characterized. GSTP1-1/V-105 was found to be more active than GSTP1-1/I-105 in conjugation reactions with the bulky diol epoxides of PAH, being up to 3-fold as active towards the anti- and syn-diol epoxide enantiomers with R-absolute configuration at the benzylic oxiranyl carbon. Comparing the four enzyme variants, GSTP1-1/A-105 generally demonstrated the highest kcat/Km value and GSTP1-1/W-105 the lowest with the anti-diol epoxides. A close correlation was observed between the volume occupied by the amino acid residue at position 105 and the value of kcat/Km. With the syn-diol epoxides, such a correlation was observed with alanine, valine and isoleucine, whereas tryptophan was associated with increased kcat/Km values. The mutational replacement of isoleucine with alanine or tryptophan at position 105 did not alter the enantio selectivity of the GSTP1-1 variants compared with the naturally occurring allelic variants GSTP1-1/I-105 and GSTP1-1/V-105. Since the amino acid at position 105 forms part of the substrate binding site (H-site) the effect of increasing bulkiness is expected to cause restricted access of the diol epoxide and proper alignment of the two reactants for efficient glutathionylation. In conclusion, the present study indicates that individuals who are homozygous for the allele GSTP1* B (coding for GSTP1-1/V-105) display a higher susceptibility to malignancy because of other factors than a decreased catalytic efficiency of GSTP1-1/V-105 in the detoxication of carcinogenic diol epoxides of benzo[a]pyrene or structurally related PAH.

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 fjord-region analogues. GSTM1-1 demonstrated an enantioselectivity ranging from no preference (50%) to high preference (> or = 90%) 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 R-configuration 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 [Jernström 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.

Mammalian V79 cells stably expressing human glutathione transferase (GST) A1-1, M1-1, and P1-1 (the allelic variant with Val105 and Ala114) have been constructed and characterized. The cells have been used to study the capacity of individual GST isoenzymes in conjunction with GSH to detoxify diol epoxides from dibenzo[a,l]pyrene (DBPDE), the most carcinogenic polycyclic aromatic hydrocarbon (PAH) identified so far, and diol epoxides from benzo[a]pyrene (BPDE). The relationship between GSH-conjugation and DNA adduct-formation has been investigated as well as factors governing the accessibility of lipophilic diol epoxide substrates for the soluble GSTs in the cells. Relative to control cells, those expressing GSTA1-1 showed the highest rate (about 50-fold increase) to perform GSH-conjugation of (-)-anti-DBPDE (R-absolute configuration at the benzylic oxirane carbon in the fjord-region) followed by GSTM1-1 (25-fold increase) and GSTP1-1 (10-fold increase). GSTA1-1 was found to be strongly inhibited when expressed in cells (10% of fully functional protein). Taking this factor into account, the rates of conjugation found in the cells fairly well reflected the order of catalytic efficiencies (k(cat)/K(m)) obtained with the pure enzymes. Increased GSH conjugation of (-)-anti-DBPDE was associated with a reduction in DNA adduct formation. GSTA1-1 inhibited the formation of adducts more than 6-fold and GSTM1-1 and GSTP1-1 about 2-fold. With (+)-anti-BPDE, GSTP1-1-expressing cells demonstrated a substantially higher rate of GSH-conjugate formation than cells with GSTA1-1 and GSTM1-1 cells (33- and 10-fold increase, respectively). Relative to control cells, GSTM1-1 was found to inhibit DNA adduct formation of (+)-anti-BPDE most effectively followed by GSTP1-1 and GSTA1-1 (12-, 4-, and 3-fold, respectively). Values of k(cat)/K(m) and estimated oil/water partition coefficients of DBPDE and BPDE were used to calculate the concentration of free diol epoxides in solution and expected rates of GSH conjugate formation in cells, and these theoretical results were compared with the observed ones. With the highly reactive (+)-anti-BPDE, 1-2% of the expected activity was observed, whereas the corresponding values for the less reactive (-)-anti-DBPDE were up to 13%. The most obvious explanations for the low observed rate with (+)-anti-BPDE are rapid and competing reactions such as hydrolysis and/or more unspecific chemical and physical reactions with cellular constituents (proteins, lipids, nucleic acids, etc.). In addition, the difference between the theoretical and observed rates may also reflect participation of factors such as macromolecular crowding and reduced rates of diffusion, factors expected to further restrict the accessibility of GST and the diol epoxides in the intact cell.