that is sufficient to stabilize the mutant enzyme to the extent that homozygotes would behave identically to controls is not known.

Theoretically, a decreased activity of MTHFR would decrease the rate of formation of methyltetrahydrofolate and consequently homocysteine remethylation. However, this might also be expected to redirect some of the one-carbon flux into other pathways of folate metabolism such as nucleotide metabolism, which could have a positive effect on nucleotide synthesis. Decreased MTHFR activity may explain a recent epidemiological study that reported that homozygotes of poor folate status had a reduced cancer risk compared to control subjects with poor folate status (Ma et al., 1997).

Folate deficiency per se would be expected to adversely affect all metabolic cycles of one-carbon metabolism. However, a metabolic defect in one enzyme may adversely affect one metabolic cycle but may promote another metabolic cycle. Metabolic defects in a single enzyme also greatly complicate the interpretation of normal measures of status. Methyltetrahydrofolate is a very poor substrate for folylpolyglutamate synthetase and has to be demethylated via the methionine synthase reaction before it can be converted to polyglutamates and retained by tissues (Cichowicz and Shane, 1987; Shane, 1989). If folate status is such that MTHFR levels decrease, the rate of formation of methyltetrahydrofolate should be reduced; this would promote folate accumulation by tissues and consequently decrease plasma folate concentrations. A lower plasma folate concentration in such a case would not represent poorer folate status, merely more effective folate accumulation by tissues. Some studies have shown lower plasma folate and increased erythrocyte folate in subjects homozygous for the T⁶⁷⁷ mutation (van der Put et al., 1996). Folate accumulation by fibroblasts from patients with severe defects in MTHFR is normal or increased above normal (Foo et al., 1982). Other studies, however, have reported low plasma and erythrocyte folate in homozygous subjects (Molloy et al., 1997), which is more difficult to explain in terms of our current understanding of folate metabolism. It is possible that the turnover and catabolism of folate is more rapid with nonmethylfolate. Although this could explain this last observation, there is only very limited data to support such a mechanism.

Currently, there is quite good evidence suggesting that the polymorphism has an adverse effect on homocysteine concentrations in subjects with relatively poor folate status (Selhub et al., 1993). However, there is no substantial evidence suggesting that this effect is not corrected by consuming the Recommended Dietary Allowance for folate that is presented in this report.