Genetic polymorphism of human drug metabolising enzymes : structural and functional studies

Sammanfattning: There is a pronounced interindividual variability in the levels and activity of many drug metabolising enzymes, which might cause differences in the sensitivity to and the toxicity of many clinically used drugs as well as environmental compounds such as nicotine and precarcinogens. In the present investigation the molecular genetic basis for some of the differences in the CYP2A6, CYP2D6, CYP2E1 and GSTM1 enzymes have been elucidated. Cytochrome P450 2D6 (CYP2D6) exhibits a marked interindividual and interethnic variability, with many Asian and black African populations have a generally reduced CYP2D6 activity compared to Caucasians. In Chinese, who are known to having reduced capacity for metabolism of CYP2D6 substrates, we have identified and characterised the CYP2D6'10 allele. This allele was shown to carry two missense mutations, P34S and S486T, and heterologous expression in COS-1 cells revealed that the P34S substitution yields an unstable enzyme that IS rapidly degraded. Indeed CYP2D6'10 was found to be the most common CYP2D6 allele among Chinese. Among black African populations, it has been shown that the presence of the CYP2D6'17 allele correlates well to reduced in vivo activity. The three missense mutations found in this allele, T107I, R296C and S486T, were introduced in all eight possible combinations into a CYP2D6 cDNA and were expressed in yeast. This revealed that a combination of the T107I and R296C substitutions yielded an enzyme with a 5-fold higher apparent Km for the substrates, whereas no effect was seen when the amino acid substitutions were introduced separately. In a Saudi Arabian population we found a low frequency of these reduced activity alleles and inactive alleles but a Egli frequency of the CYP2D6 gene duplication which is consistent with earlier phenotyping studies in this population. The expression of ethanol-inducible P450 2E1 (CYP2E1) is also known to exhibit striking interindividual variability. In order to elucidate possible genetic causes for this variation, we screened all exons for mutations using genomic DNA from almost 200 unrelated individuals. Two novel alleles were found, yielding enzymes with the amino add substitutions R76H and V3891, respectively. These alleles were however very rare, implicating a selective pressure for active CYP2E1 enzyme, most likely because of an important endogenous function. Sequencing and characterisation of the 5' flanking region of the CYP2E1 gene revealed however the presence of a polymorphic repeat region, with an allele frequency of 23 % in a Chinese population, but only 1 % in Swedes. This polymorphism might be of importance for interindividual differences in the regulation of the CYP2E1 gene. Cytochrome P450 2A6 (CYP2A6) is the major human nicotine C-oxidase. In fact, data has been presented where a correlation was seen between the number of defective CYP2A6'2 and CYP2A6'3 alleles in an individual and the risk of becoming a smoker as well as the number of cigarettes being smoked. The method used for CYP2A6 genotyping has, however, been found to give erroneous results with respect to the coumarin hydroxylase phenotype, a probe drug for the CYP2A6 enzyme. Improved genotyping methods were developed which could correctly predict the CYP2A6 phenotype. We could not, however, detect any CYP2A6'3 alleles in contrast to the 2-28 % previously reported and propose that a previously unknown common gene conversion event in the 3' flanking region of the CYP2A6 gene is the reason for this discrepancy. Furthermore, we found that the CYP2A6 gene deletion was due to an unequal cross-over event between the CYP2A6 gene and the related CYP2A7 gene. A PCR-based genotyping method was developed which showed a frequency of the deletion allele to be 15 % in a Chinese population, but only 1 % in Europeans. Finally, the molecular mechanism behind the ultrarapid glutathione S-transferase M1 (GSTM1) activity found in certain individuals was studied. Southern blotting and PCR analysis revealed a novel GSTM1 allele with a duplication of the active GSTM1 gene, which most likely occurred through an unequal cross-over event. A quantitative PCR method was developed which accurately quantified the number of GSTM1 genes. In summary, several novel cytochrome P450 and GSTM1 alleles have been identified and characterised. These are of putative importance for explanation of impaired drug metabolism and possibly for smoking behaviour and genetically determined sensitivity to carcinogens. These findings form a basis for future molecular epidemiological studies and for use in patients as a tool for individualised drug therapy resulting in less side effects and better efficacy.