Polymorphisms in sodium-dependent vitamin C transporter genes and plasma, aqueous humor and lens nucleus ascorbate...

We have previously reported low concentrations of plasma ascorbate and low dietary vitamin C intake in the older Indian population and a strong inverse association of these with cataract. Little is known about ascorbate levels in aqueous humor and lens in populations habitually depleted of ascorbate and no studies in any setting have investigated whether genetic polymorphisms influence ascorbate levels in ocular tissues. Our objectives were to investigate relationships between ascorbate concentrations in plasma, aqueous humor and lens and whether these relationships are influenced by Single Nucleotide Polymorphisms (SNPs) in sodium-dependent vitamin C transporter genes (SLC23A1 and SLC23A2). We enrolled sixty patients (equal numbers of men and women, mean age 63 years) undergoing small incision cataract surgery in southern India. We measured ascorbate concentrations in plasma, aqueous humor and lens nucleus using high performance liquid chromatography. SLC23A1 SNPs (rs4257763, rs6596473) and SLC23A2 SNPs (rs1279683 and rs12479919) were genotyped using a TaqMan assay. Patients were interviewed for lifestyle factors which might influence ascorbate. Plasma vitamin C was normalized by a log10 transformation. Statistical analysis used linear regression with the slope of the within-subject associations estimated using beta (β) coefficients. The ascorbate concentrations (micromol/L) were: plasma ascorbate, median and inter-quartile range (IQR), 15.2 (7.8, 34.5), mean (SD) of aqueous humor ascorbate, 1074 (545) and lens nucleus ascorbate, 0.42 (0.16) (micromol/g lens nucleus wet weight). Minimum allele frequencies were: rs1279683 (0.28), rs12479919 (0.30), rs659647 (0.48). Decreasing concentrations of ocular ascorbate from the common to the rare genotype were observed for rs6596473 and rs12479919. The per allele difference in aqueous humor ascorbate for rs6596473 was -217 micromol/L, p < 0.04 and a per allele difference in lens nucleus ascorbate of -0.085 micromol/g, p < 0.02 for rs12479919. The β coefficients for the regression of log10 plasma ascorbate on aqueous humor ascorbate were higher for the GG genotype of rs6596473: GG, β = 1460 compared to carriage of the C allele, CG, β = 1059, CC, β = 1132, p interaction = 0.1. In conclusion we found that compared to studies in well-nourished populations, ascorbate concentrations in the plasma, aqueous humor and lens nucleus were low. We present novel findings that polymorphisms in SLC23A1/2 genes influenced ascorbate concentration in aqueous humor and lens nucleus.