Folate-homocysteine interrelations: Potential new markers of folate status

We report a transient drop in plasma Hey and Cys following a single oral do se of PteGlu. The thiol change was concomitant with both the peak plasma 5C H(3)H(4)PteGlu1 level (by HPLC) and the maximum plasma Lactobacillus casei activity which reflects absorption of unmodified PteGlu. The significant re ciprocal association of Hey with radioassay RBC folate (r = -0.28, 99% CI - 0.48, -0.05, P = 0.0016), serum folate (r = -0.37, 99% CI -0.56, -16, P = 0 .0001), and vitamin B-12 (r = -0.42, 99% CI -0.59, -21, P = 0.0001) is show n and reflects the long-term nutritional effect of B vitamins on this impor tant, potentially atherogenic thiol. These are now well-established associa tions. We extend the potential for investigation of folate metabolism in he alth and disease by evaluating a range of new folate indices which are base d on erythrocyte coenzymes. These have been looked at independently and in association with established parameters. Erythrocyte methylfolates (mono- t o hexaglutamate-5CH(3)H(4)PteGlu(1-6)), formylfolates (tri-to pentaglutamat e-5CHOH(4)PteGlu(3-5)),formiminotetrahydrofolate (formiminoH(4)PteGlu(1)), unsubstituted tetrahydrofolate (H(4)PteGlu(1)), andpara-aminobenzoylglutama te (P-ABG) have been measured by HPLC with fluorescence detection. A positi ve linear association exists between (i) H(4)PteGlu(1) and radioassay RBC f olate (r = 0.50, 99% CI 0.07, 0.77, P = 0.0036), and (ii) H(4)PteGlu(1) and tetraglutamates of both formyland methylfolate (r = 0.52, 99% CI 0.10, 0.7 8, P = 0.0022, and r = 0.56, 99% CI 0.15, 0.80, P = 0.0009, respectively). Since, in addition, a reciprocal linear association exists between Hey and tetraglutamyl formylfolate (r = -0.41, 99% CI -0.73, 0.05, P = 0.0206), ery throcyte tetraglutamates may be a good reflection of the bodies' active coe nzyme pools. Pentaglutamyl formylfolate, the longest oligo-gamma-glutamyl c hain form of this coenzyme may be a good indicator of folate depletion. The abundance of this coenzyme both increases with increasing Hcy (r = 0.55, 9 9% CI 0.13, 0.80, P = 0.0014) and increases as H(4)PteGlu(1), the principle folate congener, decreases (r = -0.59, 99% CI -0.82, -0.20, P = 0.0004). F urthermore, the apparent equilibrium between substrate (5CH(3)H(4)PteGlu(1) ) and product (H(4)PteGlu(1)) of methionine synthase is significantly assoc iated with the abundance of 5CHOH(4)PteGlu(5) (r = -0.53, 99% CI -0.79, -0. 11, P = 0.0018). This suggests that low methionine synthase activity for wh atever reason (metabolic or dietary) may lead to an increase in the relativ e abundance of 5CHOH(4)PteGlu(5). Like 5CHOH(4)PteGlu(5), evidence is given that 5CH(3)H(4)PteGlu(6), the longest oligo-gamma-glutamyl chain form of t his particular coenzyme pool, may also be a good indicator of folate deplet ion. This is shown by a change in the relative proportion of erythrocyte me thylfolate polyglutamates following supplementation with 400 mu g/day PteGl u. Short-chain polyglutamates of methylfolate (5CH(3)H(4)PteGlu(1)-->5CH(3) H(4)PteGlu(4)) increase in proportion to the total methylfolate pool, while long-chain polyglutamates of methylfolate (5CH(3)H(4)PteGlu(5) and particu larly 5CH(3)H(4)PteGlu(6)) decrease in their relative abundance. (C) 1999Ac ademic Press.