Genetic variation of glucose transporter-1 (GLUT1) and albuminuria in 10,278 European Americans and African Americans: a case-control study in the Atherosclerosis Risk in Communities (ARIC) study

Abstract

Background: Evidence suggests glucose transporter-1 (GLUT1) genetic variation affects diabetic nephropathy and albuminuria. Our aim was to evaluate associations with albuminuria of six GLUT1 single nucleotide polymorphisms(SNPs), particularly XbaI and the previously associated Enhancer-2 (Enh2) SNP.

Methods: A two-stage case-control study was nested in a prospective cohort study of 2156 African Americans and 8122 European Americans with urinary albumin-to-creatinine ratio (ACR). Cases comprised albuminuria (N = 825; ≥ 30 μg/mg) and macroalbuminuria (N = 173; ≥ 300 μg/mg). ACR < 30 μg/mg classified controls (n = 9453). Logistic regression and odds ratios (OR) assessed associations. The evaluation phase (stage 1, n = 2938) tested associations of albuminuria (n = 305) with six GLUT1 SNPs: rs841839, rs3768043, rs2297977, Enh2(rs841847) XbaI (rs841853), and rs841858. Enh2 was examined separately in the replication phase (stage 2, n = 7340) and the total combined sample (n = 10,278), with all analyses stratified by race and type 2 diabetes.

Results: In European Americans, after adjusting for diabetes and other GLUT1 SNPs in stage 1, Enh2 risk genotype (TT) was more common in albuminuric cases (OR = 3.37, P = 0.090) whereas XbaI (OR = 0.94, p = 0.931) and remaining SNPs were not. In stage 1, the Enh2 association with albuminuria was significant among diabetic European Americans (OR = 2.36, P = 0.025). In African Americans, Enh2 homozygosity was rare (0.3%); XbaI was common (18.0% AA) and not associated with albuminuria. In stage 2 (n = 7,340), Enh2 risk genotype had increased but non-significant OR among diabetic European Americans (OR = 1.66, P = 0.192) and not non-diabetics (OR = 0.99, p = 0.953), not replicating stage 1. Combining stages 1 and 2, Enh2 was associated with albuminuria (OR 2.14 [1.20-3.80], P = 0.009) and macroalbuminuria (OR 2.69, [1.02-7.09], P = 0.045) in diabetic European Americans. The Enh2 association with macroalbuminuria among non-diabetic European Americans with fasting insulin (OR = 1.84, P = 0.210) was stronger at the highest insulin quartile (OR = 4.08, P = 0.040).

Conclusions: As demonstrated with type 1 diabetic nephropathy, the GLUT1 Enh2 risk genotype, instead of XbaI, may be associated with type 2 diabetic albuminuria among European Americans, though an association is not conclusive. The association among diabetic European Americans found in stage 1 was not replicated in stage 2; however, this risk association was evident after combining all diabetic European Americans from both stages. Additionally, our results suggest this association may extend to non-diabetics with high insulin concentrations. Rarity of the Enh2 risk genotype among African Americans precludes any definitive conclusions, although data suggest a risk-enhancing role.

Figure 1

GLUT1 is located on chromosome 1p34.2. The genomic structure of GLUT1 is illustrated with its 10 exons and three putative enhancers[5]. Genotyped polymorphisms located in intron 2 include the intron-2 SNP (rs841858), and the Enh2(rs841847) and XbaI (rs841853) SNPs which have been previously associated with diabetic nephropathy[5-7]. In the distal promoter region, the Promoter SNP (rs841839) was genotyped (located between putative enhancers 3 and 1). We also included two SNPs in intron 1, intron-1 SNP(rs3768043) and intron-2 SNP(rs2297977). No SNP was associated with albuminuria among all African Americans or among all European Americans, including XbaI (p = 0.08) and Enh2 (p = 0.08). None of the tagging SNPs were able to reach a Bonferroni adjusted level of statistical significance. ^†Bonferroni corrected level of statistical significance p = 0.01.

Figure 2

Macroalbuminuria, Enh2 risk genotypes and insulin concentrations among non-diabetic European Americans (n = 6629 (46 cases)). *P < 0.05. Among non-diabetic European Americans with fasting insulin, the risk of macroalbuminuria associated with the GLUT1 Enh2 TT risk genotype was 1.84, 95% CI: 0.71 - 4.77, P = 0.210, adjusting for age, gender, hypertension status, BMI, and GFR. For individuals in the upper quartile of insulin concentrations, the Enh2 TT genotype was associated with macroalbuminuria (OR 4.08, 95% CI: 1.06 - 15.61, p = 0.040) while it was not associated among those in the lower three quartiles (OR 1.00, 95% CI: 0.23 - 4.28, P = 0.995). There was no significant interaction between high concentrations of insulin and Enh2 TT genotypes on macroalbuminuria risk (P-interaction = 0.163).