Interaction between common folate polymorphisms and B-vitamin nutritional status modulates homocysteine and risk for a thrombotic event

We have assessed the relationship between homocysteine, its thiol metabolites, specific folate coenzymes, and vitamin B-12 according to the two main functionally relevant genotype-genotype categories that maintain the balance between homocysteine transsulphuration to cysteine, and homocysteine remethylation via folate dependent methionine biosynthesis, namely 2756A --> G-MS/66A --> G-MSR and 677C --> T-MTHFR/1298A --> C-MTHFR. We examined 152 individuals who were being treated for either thromboembolic (TE) or non-thromboembolic (non-TE) events. chi(2) test for linear trend in odds ratio provides reasonable evidence for an altered risk of thromboembolism within the range of compound MS/MSR genotypes encountered (wt/wt --> recessive/recessive) (p less than or equal to 0.05), but not within the same range of MTHFR/MTHFR genotypes. Logistic regression analysis of the risk for a TE event gave OR = 0.49 (95% CI, 0.26-0.92; p = 0.026) for 2756A --> G-MS, OR = 1.08 (95% CI, 0.65-1.78) for 66A --> G-MSR, OR = 1.19 (95% CI, 0.69-2.06) for 677C --> T-MTHFR and OR = 0.98 (95% CI, 0.52-1.85) for 1298A --> C-MTHFR. When genotypes were examined individually, one-way ANOVA showed only 677C --> T-MTHFR (p = 0.005 [TE]) and 2756A --> G-MS (p = 0.005 [non-TE] and p = 0.0006 [all subjects]) influence homocysteine. One-way ANOVA also showed that MTHFR/MTHFR compound genotype significantly influences TE homocysteine distribution (p = 0.044), but no other variable. In MS/MSR, homocysteine distribution is not significantly affected in TE subjects, but approaches significance in non-TE individuals (p = 0.062). However, the increased power obtained when all subjects are analysed demonstrates a significant influence of MS/MSR upon homocysteine distribution (p = 0.008). Other significant influences of MS/MSR were on total cellular 5-methyl-H(4)folate in non-TE subjects (p = 0.042) and vitamin B-12 in TE subjects (p = 0.018). Given the central role of vitamin B-12 in MS/MSR activity, 5-methyl-H(4)folate and homocysteine were also looked at by vitamin B-12 quartile, independent of genotype: Vitamin B-12 quartile significantly affected homocysteine distribution in TE (p = 0.013) but not non-TE individuals, with no effect on 5-methyl-H(4)folate distributions. Similarly, the prevalence of clinical phenotypes (p = 0.013) and of 'high risk' 2756A --> G-MS wildtypes (p = 0.039) was associated with the disposition of homocysteine/B-12 in TE but not non-TE subjects. Overall, results indicate compound MS/MSR genotype is associated with risk for a TE event. This may be related to variation in activity of the functional enzymes coded for by polymorphic forms of compound MS/MSR, resulting in altered catalytic cycling of methylcobalamin/cob(I)alamin, which in turn influences Hcy (and total 5-methyl-H(4)folate). The effect on vitamin B-12 is greater in TE than non-TE subjects. The compound MTHFR/MTHFR genotype also influences variation in Hcy in TE subjects, but seemingly without the same level of mediation by vitamin B-12. These results are consistent with accepted paradigms and offer a plausible explanation for the effect and interaction of specific SNPs in the TE phenotype. The biological implications of the limited number of MTHFR/MTHFR mutant alleles that can coexist, usually no more than two, may be explained by the serious consequences to folate status that these genotype combinations precipitate. We show that lowering of all folate 1-C pools occurs in the rare ct/cc compound genotype, except for the 5,10-methenyl-H(4)folate pool, which expands. 5,10-methenyl-H(4)folate is the immediate product of 5,10-methylene-H(4)folate, which is likely diverted away from methionine biosynthesis via the aberrant MTHFR enzyme. Consequences for the methylation cycle may be severe, and in most cases lethal for the developing embryo, where methylation is required for dozens of critical processes, but particularly for maintaining DNA methylation patterns that are now known to regulate the expression of half the complement of human genes via CpG islands located in the 5' promotor region, or within the first few exons of the gene. (C) 2003 Elsevier Science (USA). All rights reserved.