Catechol O-methyltransferase pharmacogenomics: human liver genotype-phenotype correlation and proximal promoter studies

Abstract

Objectives: Catechol O-methyltransferase (COMT) is expressed as both soluble (S) and membrane-bound (MB) isoforms, with S-COMT predominantly expressed in the liver. A common nonsynonymous single nucleotide polymorphism (SNP), 472G > A (108/158Val > Met, S/MB), has been associated with variation in levels of COMT enzyme activity and thermal stability. We set out to test the hypothesis that additional COMT polymorphisms might also be associated with phenotypic variation.

Methods: We phenotyped 268 liver biopsy samples for S-COMT activity and thermal stability, resequenced a portion of the gene that had not been resequenced earlier, and genotyped DNA from these same samples for 16 COMT polymorphisms.

Results: There was a significant association between the two COMT phenotypes and genotype at the codon 108 SNP. A haplotype-based approach was then used to assess the possible association of other polymorphisms with phenotype. Specifically, the codon 108 SNP explained 20.4% of variance in enzyme activity (P < 10), and 59% of variance in thermal stability (P < 10). Haplotypes that included SNPs at cDNA nucleotides 408 and 472 explained additional variance in enzyme activity (up to 24.4%), and the addition to the haplotype of a SNP at intron 2 (51) explained a total of 27.5% of the variance. However, no SNPs beyond that at the nucleotide 472G > A polymorphism were associated with variation in thermal stability. We also observed a three-fold variation in the ability of reporter gene constructs for 'proximal promoter' haplotypes to drive transcription.

Conclusion: The common COMT 108Val > Met polymorphism is associated with human liver S-COMT activity and thermal stability, but additional COMT SNPs also contribute to variation in activity.

Figure 1

Human COMT polymorphisms. Black rectangles represent coding exons, and open rectangles represent portions of exons that encode UTR sequences. The cross-hatched area encodes protein in MB-COMT but not in S-COMT. “CA” indicates Caucasian-American and “AA” African-American subjects. (A) Polymorphisms observed during our COMT resequencing studies are shown. Arrows show polymorphism locations, with frequencies indicated by the color of the arrows. The “box” labeled “New Resequencing Data” contains the novel resequencing data obtained during the present study. (B) Polymorphisms selected for inclusion in the genotyping panel. Arrows indicates locations of those polymorphisms, and are “color-coded” based on their frequencies in CA subjects. The two “boxed” polymorphisms were not within the area resequenced, but were included because they had been reported to be of possible functional significance during previous studies [20].

Figure 2

Linkage disequilibrium for COMT polymorphisms in DNA samples from AA (panel A) and CA (panel B) subjects based on both our original [10] and the present resequencing data.

Figure 3

Human liver S-COMT activity and thermal stability. (A) Frequency distribution histogram of hepatic biopsy S-COMT activity in samples from 268 Caucasian women. (B) Frequency distribution histogram of hepatic biopsy S-COMT thermal stability in the same samples.

Figure 4

Human liver S-COMT activity and thermal stability. (A) Scattergram of S-COMT activity on the X-axis and S-COMT thermal stability on the Y-axis (expressed as an H/C ratio). (B) Average values ± SD for each genotype shown in (A).

Figure 5

Sequential selection of SNPs for their association with human hepatic S-COMT activity, using haplotype analyses, adjusted for sub-haplotype effects.

Figure 6

Human COMT “proximal promoter” (intron 2) reporter gene studies. Activities of luciferase reporter gene constructs that included different COMT proximal promoter haplotypes were assayed after the transfection of HepG2 cells are expressed as a percentage of the value for the most common haplotype (construct #1). Each bar represents the average of 6 independent assays (mean ± SEM); * p < 0.05; ** p< 0.01 when compared with the value for construct #1.