Additive effects of genetic variation in GCK and G6PC2 on insulin secretion and fasting glucose

Abstract

Objective: Glucokinase (GCK) and glucose-6-phosphatase catalytic subunit 2 (G6PC2) regulate the glucose-cycling step in pancreatic beta-cells and may regulate insulin secretion. GCK rs1799884 and G6PC2 rs560887 have been independently associated with fasting glucose, but their interaction on glucose-insulin relationships is not well characterized.

Research design and methods: We tested whether these variants are associated with diabetes-related quantitative traits in Mexican Americans from the BetaGene Study and attempted to replicate our findings in Finnish men from the METabolic Syndrome in Men (METSIM) Study.

Results: rs1799884 was not associated with any quantitative trait (corrected P > 0.1), whereas rs560887 was significantly associated with the oral glucose tolerance test 30-min incremental insulin response (30' Deltainsulin, corrected P = 0.021). We found no association between quantitative traits and the multiplicative interaction between rs1799884 and rs560887 (P > 0.26). However, the additive effect of these single nucleotide polymorphisms was associated with fasting glucose (corrected P = 0.03) and 30' Deltainsulin (corrected P = 0.027). This additive association was replicated in METSIM (fasting glucose, P = 3.5 x 10(-10) 30' Deltainsulin, P = 0.028). When we examined the relationship between fasting glucose and 30' Deltainsulin stratified by GCK and G6PC2, we noted divergent changes in these quantitative traits for GCK but parallel changes for G6PC2. We observed a similar pattern in METSIM.

Conclusions: Our data suggest that variation in GCK and G6PC2 have additive effects on both fasting glucose and insulin secretion.

FIG. 1.

Hypothesized effect of GCK rs1799884 and G6PC2 rs560887 genotype combinations on ATP production and insulin secretion in pancreatic β-cells. Upper portion of the figure shows the glucose-cycling step and the effect of specific alleles in GCK and G6PC2 on activity of each enzyme: thick arrow depicts increased activity, thin arrow depicts reduced activity. Lower portion shows the specific genotype combinations and the hypothesized effect on ATP production and subsequent insulin secretion.

FIG. 2.

The 30′ Δinsulin and fasting glucose stratified by GCK rs1799884 and G6PC2 rs560887 genotypes. A: Age- and sex-adjusted 30′ Δinsulin (means ± SE) stratified by GCK rs1799884 and G6PC2 rs560887 genotype combination. Genotype groups are ordered by hypothesized effect on glucose cycling in the pancreatic β-cell (see text for details). Dominant genetic models are assumed for both loci. 30′ Δinsulin changed linearly (P = 0.0016) with genotype group. B: Age- and sex-adjusted fasting plasma glucose (means ± SE) stratified by genotype. Details are as described for top panel. Fasting glucose was not linearly related to genotype group (P = 0.191).

FIG. 3.

The 30′ Δinsulin vs. fasting plasma glucose stratified by GCK rs1799884 and G6PC2 rs560887 genotype. A: Shows results from the BetaGene Study. Model-predicted age- and sex-adjusted 30′ Δinsulin (means ± SD) stratified by genotype is plotted against model- predicted age- and sex-adjusted fasting plasma glucose (means ± SD). Solid symbols represent G6PC2 rs560887 G/A or A/A genotypes and open symbols represent G6PC2 rs560887 G/G genotype. Circles represent GCK rs1799884 G/G genotype and squares represent GCK rs1799884 G/A and A/A genotypes. B: Shows results from the METSIM Study. Details are identical to that described for the top panel except that circles and squares represent GCK rs4607517 genotypes.