CYP2C9 exon 4 mutations and warfarin dose phenotype in Asians

It is well established that functionally defective variant alleles of CYP2C9 have a major impact on anticoagulation-related outcomes during warfarin therapy.1A recent paper by Leung et al2 reported the discovery of 4 new single nucleotide polymorphisms (SNPs) in the coding region of theCYP2C9 gene: L208V, Q192P, H184P, and I181L, with allele frequencies of 0.09-0.75 that may be associated with a low warfarin maintenance dose in Hong Kong Chinese patients. This report is significant because Asian patients have long been recognized to require only about 50% of the average warfarin dose used in whites to elicit the same degree of anticoagulation.3 Asian patients in the Leung study who were heterozygous or homozygous for V208 exhibited 26%-34% reductions in their mean warfarin dose relative to “wild-type” subjects, although the differences were not statistically significant.2 Nonetheless, because the L208V mutation is located in a putative substrate-binding site of the enzyme, alterations in the catalytic efficiency of this variant could be anticipated, thereby providing a plausible metabolic explanation for the low warfarin dose phenotype in Asians.

Given our interests in genetic variability at this locus, we incorporated these SNPs into our own panel of assays. We have now screened for these variants in a total of 62 individuals of Asian descent: 52 Chinese and 10 Koreans. Subjects were healthy volunteers, none of whom were on anticoagulant therapy. Our screens were based on direct sequence analysis or a single-base extension assay. Two separate sets of primers located within intron 3 and intron 4 were used to amplify exon 4 of CYP2C9: forward 1: 5′-AAT ACA GTG TTT TAT ATC TAA AGT TTA ATA-3′; reverse 1: 5′-ATG CAA TTC AGA GCT TGA TC-3′; forward 2: 5′-TGT TAA GGG AAT TTG TAG G-3′; reverse 2: 5′-AAT TTT GGA TTT GTC AGA A-3′. All primers were checked against the entireCYP2C locus to ensure specificity and yielded single amplification products of the expected size. We detected noCYP2C9 exon 4 SNPs in these 2 Asian populations or in banked white DNA samples (n = 50). Moreover, Zarza et al4failed to identify these SNPs in 106 Spanish subjects, and Goldstein5 has not reported CYP2C9 exon 4 mutations from preliminary resequencing in 90 subjects from 3 racial groups.

In assessing variations in the experimental protocols that might underlie the interlaboratory differences, we re-examined the primers chosen for amplification of exon 4 by Leung et al.2 The first 10 bases of the forward primer and 8 bases of the reverse primer (5′ to 3′) appeared to have been included to incorporate a unique restriction endonuclease site into the resulting amplicon. Analysis of the remaining primer sequence revealed that the forward primer exhibits a 100% match to exon 4 sequences within CYP2C8,CYP2C9, CYP2C18, and CYP2C19. The reverse primer exhibits a 100% match to exon4/intron4 sequences withinCYP2C9 and CYP2C19, although the 3′ end of the primer shows considerable identity to sequences with CYP2C18and CYP2C19 and might well have resulted in some amplification at those loci as well. In any event, it is clear that the primers used by Leung et al2 are not adequate for SNP discovery, as they will result in a mixed template from multiple loci. These observations call into question the presence of the L208V, Q192P, H184P, and I181L variants of CYP2C9 and their association with a low warfarin dose phenotype in Asian patients.