GLUT4 Gene Polymorphisms and Their Association with Type 2 Diabetes in South Indians

Abstract

Background and Objectives:

The GLUT4 gene, which encodes glucose transporter 4, is a candidate gene for type 2 diabetes mellitus (T2DM). The aim of this study was to screen the GLUT4 gene for polymorphisms and to study association of such polymorphisms with T2DM in an Asian Indian population in southern India.

Methods:

The GLUT4 gene was sequenced in 25 normal glucose tolerance (NGT) and 25 T2DM subjects, and the variants found were then genotyped by polymerase chain reaction–restriction fragment length polymorphism in a pilot study population of 552 NGT and 643 T2DM subjects, randomly selected from the Chennai Urban Rural Epidemiology Study. Two of the variants (rs5435 and the novel variant), which showed significantly higher minor allele frequency in T2DM compared with NGT individuals in the pilot study population, were then retested with an additional 465 NGT and 363 T2DM subjects, giving a final sample size of 1,017 NGT and 1,006 T2DM subjects.

Results:

Sequencing of the GLUT4 gene revealed three known polymorphisms (rs5418, rs5421, and rs5435) and one novel T→G variant in the 3′ untranslated region (UTR) at nucleotide position 6787483. The rs5418 and rs5421 polymorphisms did not show any association with diabetes. The rs5435 [Asn130Asn(C→T)] polymorphism was found to be associated with diabetes, with the odds ratio for the CT+TT genotype being 1.26 (95% confidence interval, 1.00–1.57; P=0.043) when the CC genotype was taken as reference. The frequency of the TG genotype of the novel 3′UTR T→G variant was significantly higher in diabetes subjects (1%) compared with NGT subjects (0.2%) (P=0.021). There was a significant difference in the proportion of the ACGT haplotype of the rs5418(A→G), rs5435(C→T), rs5421(C→G), and the T→G 3′UTR variant between the NGT (7.5%) and diabetes (5%) groups (P=0.003).

Conclusion:

The rs5435 (C→T) polymorphism of the GLUT4 gene is associated with type 2 diabetes in this south Indian population.

Introduction

T ype 2 diabetes mellitus (T2DM) is characterized at least in part by decreased insulin-stimulated glucose uptake in tissues such as muscle and adipose tissue. Defects in whole-body insulin-mediated glucose uptake in T2DM patients have been tracked to impairments in insulin signaling and translocation of the glucose transporter, GLUT4.^1
–3 The GLUT4 gene [also known as the SLC2A4 (solute carrier family 2, member 4) gene] encoding the GLUT4 protein is therefore a strong candidate gene for T2DM.

The GLUT4 gene has not been thoroughly investigated in Asian Indians, who have an increased susceptibility to diabetes, and currently there are over 50 million people with diabetes in India.⁴ It has been shown that plasma insulin levels, a surrogate marker of insulin resistance, were higher in Asian Indians compared with Europeans,⁵ and glucose clamp studies confirmed that Asian Indians have a greater degree of insulin resistance.⁶ Subsequently, several studies have confirmed these findings.^7
–9 These studies suggest that there could be a genetic predisposition to insulin resistance and diabetes in Asian Indians.

Hence, the objective of the present study was to screen the GLUT4 gene for mutations/polymorphisms and to study their association with T2DM in an Asian Indian population in southern India.

Research Design and Methods

Subjects

All the study subjects were unrelated and chosen from the Chennai Urban Rural Epidemiology Study (CURES). The methodology of the study has been published elsewhere.¹⁰ In Phase 1 of CURES, 26,001 subjects were recruited based on a systematic random sampling technique. Self-reported diabetes subjects were classified as “known diabetes subjects.” In Phase 2 of CURES, all known diabetes subjects (n=1,529) were invited to our center for detailed studies, of whom 1,382 responded. In Phase 3 of CURES, every 10^th individual from Phase 1 (n=2,600) was invited to undergo an oral glucose tolerance test (OGTT) using a 75-g oral glucose load (dissolved in 250 mL of water). Those who were confirmed by OGTT to have a 2-h plasma glucose value of ≥11.1 mmol/L (200 mg/dL) (based on World Health Organization Consulting Group criteria) were labeled as “newly detected diabetes subjects” (n=222). Subjects who were confirmed by OGTT to have a fasting plasma glucose level of <5.6 mmol/L (100 mg/dL) and a 2-h plasma glucose value of ≤7.8 mmol/L (140 mg/dL) were categorized as normal glucose tolerance (NGT) subjects¹¹ (n=1,736). The total number of diabetes subjects in the CURES population is 1,604 (1,382 known diabetes subjects+222 newly detected diabetes subjects). From the 1,604 diabetes subjects and the 17,36 NGT subjects, 643 and 552 subjects, respectively, were randomly selected for the present study, and a pilot study was carried out. Subsequently the sample size was increased to 1,006 T2DM subjects and 1,017 NGT subjects.

Biochemical measurements

Anthropometric measurements, including weight and height, were obtained using standardized techniques. The body mass index was calculated as the weight (in kilograms) divided by the square of height (in meters). Biochemical analyses were carried out on a autoanalyzer (Hitachi-912, Hitachi, Mannheim, Germany) using commercial kits (Roche Diagnostics, Mannheim). Fasting plasma glucose was estimated using the glucose oxidase–peroxidase method. Serum cholesterol was estimated using cholesterol oxidase–phenol 4-amino antipyrene peroxidase method. Serum triglyceride was estimated using the glycerol phosphatase oxidase–phenol 4-amino antipyrene peroxidase method. Serum insulin concentration was estimated using kits from Dako (Glostrup, Denmark).

Genotyping

DNA was isolated from whole blood using the phenol–chloroform method. Sequencing of the GLUT4 gene was carried out on 25 NGT and 25 T2DM subjects on a genetic analyzer (ABI 310, Applied Biosystems, Foster City, CA). The single nucleotide polymorphisms (SNPs) rs5418, rs5435, rs5421 and the novel 3′ untranslated region (UTR) variant were all genotyped by polymerase chain reaction (PCR)-restriction fragment length polymorphism (RFLP). Primers for sequencing and for PCR-RFLP were designed using Primer 3 software (available online at frodo.wi.mit.edu/primer3/), and the primer sequences are listed in Table 1. A final reaction volume of 15 μL of PCR mix contained 100 ng of genomic DNA, 5 pmol of each primer, PCR buffer with 1 mM MgCl₂, each deoxynucleotide triphosphate at 100 μM, and 0.5 U of Taq polymerase (Life Technologies, Carlsbad, CA). PCR was carried out on a Peltier Thermal Cycler (MJ Research Inc., Waltham, MA) using the following conditions: 95°C for 5 min, followed by 34 cycles of (95°C for 30 s, 60°C for 30 s, and 72°C for 30 s), and a final extension of 72°C for 9 min. PCR products were digested with 2 units of ScrFI (rs5418), BccI (rs5421), or BstFNI (rs5435 and novel 3′UTR variant) enzyme and electrophoresed on a 3% agarose gel. To assure that the genotyping was of adequate quality, we performed random duplicates in 20% of the samples. The assays were performed by a technician who was masked to the phenotype, and there was 99% concordance in the genotyping. Informed consent was obtained from all the subjects who participated in this study, and the study was approved by the institutional ethical committee.

Table 1.

List of Primers for Polymerase Chain Reaction and Direct Sequencing of the GLUT4 Gene

Region	Upstream primer (5′–3′)	Downstream primer (5′–3′)	PCR product (bp)
Promoter contiguous 1	CAGACACTCTCCACGCTCTG	AGCAATGCCCCAAAGTAAC	523
Promoter contiguous 2	CCAACGTGGGAGCTAAAAAT	AAGGACCCGGAGAGCAGT	537
Exon 1	GGGCTTCTCGCGTCTTTT	GAACCACCCCCAGCTGAC	352
Exon 2	CCCAGTATCTTCAGGCTCCA	GGACAGATTTCCTGCAGACC	312
Exon 3	CCAAACGAGGAGGACAGGT	GCAGACAGAGTGAGGCTGTG	337
Exon 4	TCTCAGTGGCTTGGAAGGTT	AGGTGTCCATTCCCATCTGA	314
Exon 5	TTGGCGCCTACTCAGGTACT	TGGAAGCCCACCATATTCTC	315
Exon 6	CAGAAGGGCTGAGTGACCT	CAGGTGCTAGCAAGATGAG	271
Exon 7	GCCTCATCTTGCTAGCACCT	GAGGCTGCTCCACACATACA	271
Exon 8	AGCAGCTCTCTGGCATCAAT	CCATTCCAGAGGTGAGCAGT	313
Exon 9	ACTGCTCACCTCTGGAATGG	CCTGGCTGGTCACAGAGACT	323
Exon 10	AGTTGAGGGCAAGGGAAGAT	AGGTGGGCTAGCTGTGTGAT	338
Exon 11	GAAGGCCTTTAGCTCCTGGT	GGGTTAAAGTGCTGCAGAGG	370
3′UTR	CAGAGACCCCTTCCTTTCCT	AGAGTCTGCGTGGCAAGAAT	434
SNPs studied
rs5418 (A→G)	CACCAACTCCAGCCAAACTC	CGGAGCCTATCTGTTGGAAG	381
rs5435 (C→T)	CATCTCTCAGTGGCTTGGAAGGTTC	ATGAGCATTTCATAGGAGGCAGCCG	206
rs5421 (C→G)	AGTTGAGGGCAAGGGAAGAT	AGGTGGGCTAGCTGTGTGAT	338
Novel T→G variant	CTTTAACCCTCTCTTCCCTATTATT	AACCACAACACATAAATAATCCACG	142

PCR, polymerase chain reaction; SNP, single nucleotide polymorphism; UTR, untranslated region.

Statistical analysis

SPSS Windows version 15.0 (SPSS, Inc., Chicago, IL) was used for statistical analysis. One-way analysis of variance or Student's t test, as appropriate, was used to compare groups for continuous variables. The χ ² test or Fisher's exact test, as appropriate, was used to compare the proportions of genotypes or alleles. Logistic regression analysis was performed using diabetes as dependent variable and the genotypes as independent variables. P values of <0.05 were considered statistically significant. Power was calculated using the free online (ps-power-and-sample-size-calculation.software.informer.com/) tool, PS-Power and sample size calculation, version 3.0.14. The linkage disequilibrium (LD) pattern between the SNPs was analyzed using the software HAPLOVIEW12. The HAPLOPOP program¹³ was used to determine the frequencies of the four-locus haplotypes. A Bonferroni correction was carried out for multiple testing comparisons in haplotype analysis.

Results

Sequencing of the GLUT4 gene

The entire coding region (11 exons) including the intron–exon junctions, promoter, and 3′UTR of the SLC2A4 or the GLUT4 gene was sequenced in 25 NGT and 25 T2DM subjects, and this revealed three known polymorphisms and one novel variant as listed below: (1) rs5418 (A→G) in 5′UTR, (2) rs5435 [Asn130Asn (AAC→AAT)] in exon 4, (3) rs5421 (C→G) in intron 9, and (4) the novel T→G variant in 3′UTR at nucleotide position 6787483.

As a pilot study, all the variants were genotyped by PCR-RFLP in 552 NGT subjects and 643 T2DM subjects, and their association with type 2 diabetes was tested. The genotypic distribution of all the four variants in both the NGT and T2DM groups was in Hardy–Weinberg equilibrium.

Subject characteristic

The diabetes subjects were older (49±10 years) compared with the NGT subjects (46±11 years). Body mass index (NGT, 23.7±4.6 kg/m²; diabetes, 25.1±4.3 kg/m²; P<0.001), fasting plasma glucose (NGT, 4.7±0.5 mmol/L; diabetes, 9.0±3.8 mmol/L; P<0.001), total cholesterol (NGT, 4.78±1.0 mmol/L; diabetes, 5.24±1.0 mmol/L; P<0.001), and serum triglycerides (NGT, 1.35±0.73 mmol/L; diabetes, 2.02±1.42 mmol/L; P<0.001) were all significantly higher in the T2DM subjects.

rs5418 and rs5421 polymorphisms in the GLUT4 gene

Table 2 shows the genotypic and allelic frequency of the rs5418 and rs5421 variants. The minor allele “G” of rs5418 and rs5421 polymorphisms had a respective prevalence of 41% and 13% among NGT subjects and 44% and 11% among T2DM subjects and was not significantly different between the two groups (P=0.15 and P=0.19, respectively). The genotypic frequencies of both the SNPs did not show any significant difference between the NGT and T2DM groups when analyzed according to codominant, dominant, and recessive models (Table 2). A comparison of the biochemical parameters of the NGT and T2DM subjects stratified based on these two SNPs did not show any significant differences among the genotypes in any of the groups.

Table 2.

Genotype and Allele Frequency of the rs5418 and rs5421 Polymorphisms in the GLUT4 Gene

			P value
	NGT subjects (n=552)	T2DM (n=643)	Codominant model ^a	Dominant model ^b	Recessive model ^c
rs5418 (A→G)
AA	195 (35.3%)	197 (30.6%)	0.22	0.10	0.61
AG	266 (48.2%)	332 (51.6%)
GG	91 (16.5%)	114 (17.8%)
MAF “G” allele	448 (41%)	560 (44%)		0.15
rs5421 (C→G)
CC	417 (75.5%)	508 (79.0%)	0.36	0.17	0.93
CG	130 (23.5%)	130 (20.2%)
GG	5 (1%)	5 (0.8%)
MAF “G” allele	140 (13%)	140 (11%)		0.19

For rs5418 (A→G), AA versus AG versus GG; for rs5421 (C→G), CC versus CG versus GG.

For rs5418 (A→G), AA versus AG+GG; for rs5421 (C→G), CC versus CG+GG.

For rs5418 (A→G), AA+AG versus GG; for rs5421 (C→G), CC+CG versus GG.

MAF, minor allele frequency; NGT, normal glucose tolerance; T2DM, type 2 diabetes mellitus.

Asn130Asn [rs5435(C→T)] polymorphism in the GLUT4 gene

The frequency of the “T” allele of the Asn130Asn [rs5435(C→T)] variant was significantly higher among the T2DM subjects compared with the NGT subjects in the pilot study population of 552 NGT and 643 T2DM subjects. This difference in the allele frequency remained significant even after increasing the sample size of subjects with NGT (from 552 to 1,017) and those with T2DM (from 643 to 1,006). The “T” allele showed a frequency of 0.30 in NGT and 0.33 in T2DM subjects, and this difference was statistically significant (P=0.046) (Table 3).

Table 3.

Genotype and Allele Frequency of the rs5435 (Asn130Asn) Polymorphism in the GLUT4 Gene

			P value
	NGT subjects (n=1,017)	T2DM subjects (n=1,006)	Codominant model ^a	Dominant model ^b	Recessive model^c
Genotype frequency
CC	486 (48%)	432 (43%)	0.08	0.032	0.44
CT	442 (43%)	475 (47%)
TT	89 (9%)	99 (10%)
Allele frequency
“C” allele	1,414 (70%)	1,339 (67%)		0.046
“T” allele	620 (30%)	673 (33%)

Logistic regression (dominant model)	NGT subjects (n=1,017)	T2DM subjects (n=1,006)	Odds ratio^*	95% CI	P value
CC	486 (48%)	432 (43%)	Reference	—	—
CT+TT	531 (52%)	574 (57%)	1.26	1.00–1.57	0.043

CC versus CT versus TT.

CC versus CT+TT.

CC+CT versus TT.

Adjusted for age, sex, and body mass index.

CI, confidence interval; NGT, normal glucose tolerance; T2DM, type 2 diabetes mellitus.

Because of the complex inheritance pattern of T2DM, different genetic models were tested. According to the dominant model, the genotypic distribution was significantly different between the NGT and T2DM groups, with the frequency of the CT+TT genotype being 0.52 among NGT subjects and 0.57 among T2DM subjects (P=0.032) (Table 3). Logistic regression analysis was performed according to the dominant model taking T2DM as the dependent variable and the genotypes as the independent variables (CC coded as 0 and CT+TT coded as 1). After adjusting for age, sex, and body mass index, the odds ratio for diabetes for the CT+TT genotype was 1.26 (95% confidence interval, 1.00–1.57; P=0.043) when the CC genotype was taken as reference (Table 3). Using the empirical genotype relative risks, the power to detect the difference between the CT+TT genotype and CC genotype at the 5% level of significance was 0.6.

Because the rs5435 (Asn130Asn) SNP showed an association with diabetes under a dominant model, the biochemical parameters were also analyzed according to the same model (CC vs. CT+TT). Fasting serum insulin levels were significantly higher in the NGT subjects with the CT+TT genotype (mean±SD, 9.3±6.0 μIU/mL) compared with those with the CC genotype (8.1±5.4 μIU/mL) (P=0.014), and the association remained significant even after adjusting for age and sex (P=0.019).

T→G novel variant in the 3′UTR (at nucleotide position 6787483)

In the pilot study, there was a significant difference in the frequency of the TG genotype between the NGT and the T2DM subjects. The sample size was further increased to 1,017 NGT and 1,006 T2DM subjects, and the association of this variant with T2DM was tested. The frequency of the TG genotype of the novel T→G variant in 3′UTR was significantly higher among T2DM subjects (1%) compared with NGT subjects (0.2%) (P=0.021) (Table 4). None of the subjects studied was found to carry the GG (rare homozygous) genotype. The allele-level test was not performed because of the extremely low frequency of the minor allele (G) in both NGT subjects (0.1%) and T2DM subjects (0.5%). The biochemical parameters of the TT and TG genotype of the T→G novel variant in the 3′UTR were not compared because of the very low frequency of the TG genotype. Using the empirical genotype relative risks, the power to detect the difference between the TG genotype and TT genotype at a 5% level of significance was 0.53.

Table 4.

Genotype and Allele Frequency of the T→G 3′ Untranslated Region Variant at Nucleotide Position 6787483 in the GLUT4 Gene

Genotype	NGT subjects (n=1,017)	T2DM subjects (n=1,006)	Fisher's exact test P value
TT	1,015 (99.8%)	996 (99%)	0.021
TG	2 (0.2%)	10 (1%)
GG	—	—
MAF “G” allele	2 (0.1%)	10 (0.5%)

MAF, minor allele frequency; NGT, normal glucose tolerance; T2DM, type 2 diabetes mellitus.

LD estimation between GLUT4 SNPs and haplotype analysis

LD estimation between these SNPs in the pilot study population showed that the pairwise LD between any of these SNPs was not high (r ² values in Table A1 in the Appendix). In order to perform haplotype-based analyses, four locus haplotypes were constructed, and analyses were restricted to the haplotypes that have a frequency of at least 0.01 in either cases or controls (Table 5).

Table 5.

Comparison of Haplotype Frequencies of the GLUT4 Single Nucleotide Polymorphisms in Normal Glucose Tolerance and Type 2 Diabetes Subjects

Haplotypes for SNP rs5418/rs5435/rs5421/3′UTR variant	NGT (n=552)	Type 2 diabetes (n=643)	Observed value of test statistics (z)	P value
ACCT	0.493	0.485	−0.451	0.652
GTCT	0.264	0.269	0.318	0.748
GCCT	0.092	0.099	0.671	0.509
ACGT	0.075	0.05	−2.928	0.0033^*
ATCT	0.024	0.034	1.672	0.094
GCGT	0.021	0.028	1.272	0.204

Significant difference.

NGT, normal glucose tolerance; SNP, single nucleotide polymorphism; UTR, untranslated region.

There was a significant difference in the proportion of the ACGT haplotype of the rs5418 (A→G), rs5435 (C→T), rs5421 (C→G) and the T→G variant in the 3′UTR between the NGT (7.5%) and T2DM (5%) groups (nominal P value=0.003). Because six tests were performed corresponding to the six haplotypes satisfying the selection criterion, a multiple correction was done using the Bonferroni test. The association finding remained significant as the nominal P value was lower than the Bonferroni threshold of 0.05/6, that is, 0.0083. We also performed a power computation to evaluate whether our sample size had sufficient power to detect the observed difference in the proportion of the haplotype ACGT in the two groups and found that the power was approximately 0.68 at the 5% level of significance.

Discussion

This study assumes significance for two reasons: (1) The entire coding region along with the promoter and 3′UTR region of the GLUT4 gene has been thoroughly sequenced for the first time in this Asian Indian population, and (2) a case-control association study has been performed on all the four variants found by sequencing, thus examining the association of GLUT4 variants with T2DM in this population.

Association studies between the GLUT4 locus and T2DM have been performed in different ethnic groups using different technologies,^14

–18 and the results have so far been negative. Studies using RFLP revealed no association between T2DM patients and GLUT4 when compared with appropriate control subjects in Italian whites,¹⁵ English whites,¹⁷ African Americans,¹⁴ Chinese,¹⁹ Japanese,¹⁹ whites,¹⁹ Asian Indians,¹⁹ and American black populations.¹⁹ A study using single-stranded conformation polymorphism after PCR amplification also revealed no differences between T2DM patients and control subjects in Welsh whites.¹⁶ In another study, four GLUT4 variants, rs2654185, rs5412, rs5418, and rs5435, selected from the SNP database were genotyped by tetraprimer amplification refractory mutation system PCR and PCR-RFLP in subjects belonging to Dravidian ethnicity from Kerala, South India; this study also found no evidence of association of these SNPs with diabetes.²⁰ Two studies sequenced the entire protein coding region, including the intron–exon junctions,^18,21 of the GLUT4 gene and failed to find any genetic variant that was associated with T2DM. However, these findings do not exclude the possibility of association of genetic variants in the GLUT4 gene with T2DM as most of these studies were conducted on small numbers.

In the present study, all the variants detected during sequencing (rs5418, rs5421, rs5435, and the novel T→G 3′UTR variant) were genotyped in a pilot study population of 552 NGT and 643 T2DM subjects. The rs5418 and rs5421 polymorphisms did not show any association with T2DM and hence were not genotyped further. The rs5435 showed significant difference in the allele frequency, and the novel T→G 3′UTR variant showed significant difference in the frequency of the TG genotype in the pilot study population, and hence these variants were genotyped on larger samples (1,017 NGT and 1,006 T2DM subjects).

The rs5435 (Asn130Asn) is a common polymorphism that has been noted in individuals of different racial groups.^{14,16,18,20

–23} In a Welsh population of 86 T2DM and 76 control subjects, no significant difference in the allelic or genotypic frequency of this polymorphism was observed.¹⁶ In whites, there was no significant difference in the frequency of this polymorphism between a group of 30 diabetes patients (0.25) and 17 control patients (0.32).²³ In a study on a south Indian population from Kerala, the minor allele frequency of this rs5435 (C→T) polymorphism was found to be 0.29 in cases and 0.32 in controls and showed no significant association with T2DM.²⁰

In this study, we report that the Asn130Asn polymorphism is associated with T2DM. The significantly higher fasting serum insulin levels in the CT+TT genotype compared with the TT genotype further strengthens the role of the Asn130Asn polymorphism in T2DM. Although it is a silent polymorphism, the Asn130Asn appears to alter the risk of T2DM in this study population. This observation is not surprising because a recent study has shown that silent SNPs can affect in vivo protein folding and, consequently, function.²⁴ This study showed that substrate specificity of P-glycoprotein, the product of the multidrug resistance 1 gene, is altered by SNPs presumed to be synonymous and silent.²⁴ In addition, the Asn130Asn SNP need not be functionally relevant by itself, but can be in LD with polymorphisms in introns that would destabilize pre-mRNA and result in reduced mRNA levels.²⁵ As intronic regions of the GLUT4 gene have not yet been systematically screened for SNPs, this scenario remains a possibility. Moreover, a silent polymorphism could also affect translation efficiency by changing the codon preference. Another mechanism could be an effect on splice enhancer regions within the exon and/or activation of a cryptic splice site.^26,27 We have earlier reported that another silent polymorphism, the Thr394Thr polymorphism of the peroxisome proliferator-activated receptor-γ coactivator-1α gene is associated with T2DM,²⁸ metabolic syndrome,²⁹ and total, visceral, and subcutaneous body fat³⁰ in our population.

In addition to the three known SNPs found in this study, a novel T→G variant in the 3′UTR at nucleotide position 6787483 was also discovered during sequencing. Although the frequency of the TG genotype was significantly higher in T2DM subjects, the overall frequency of this variant was very low, and hence this variant could represent only a very small subset of the T2DM population. Because this is the first report on this variant, the allele or genotype frequency from other populations is not available for comparison. The evidence that an SNP in the 3′UTR is associated with a complex genetic trait such as diabetes is not surprising as similar findings with Crohn's disease,³¹ asthma,³² and schizophrenia³³ have been reported.

The 3′UTRs regulate gene expression through the modulation of mRNA stability.^34
–36 The 3′UTR may modulate mRNA stability through the binding with specific proteins, which occurs mostly but not exclusively at AU-rich regions.^34,37 Moreover, the role of a 3′UTR SNP in T2DM diabetes has already been demonstrated in the protein-tyrosine phosphatase 1B gene.³⁸ It has been shown that the minor allele “G” of the 1484insG polymorphism is likely to play a role in protein-tyrosine phosphatase 1B mRNA stability, causing protein-tyrosine phosphatase 1B overexpression and plays a role in insulin resistance.³⁸

This is the first study to our knowledge, to perform a haplotype analysis of the previously reported rs5418, rs5435, and rs5421 variants. Hence, data with respect to the frequency of the various haplotypes in other populations are not available for comparison. The ACGT haplotype of rs5418, rs5435, rs5421, and T→G variant in the 3′UTR, which has shown protection against type 2 diabetes, consists of the “C” allele of the Asn130Asn [rs5435 (C→T)] SNP. This is quite possible as individual analysis of the Asn130Asn SNP showed the “T” allele of Asn130Asn (C→T) SNP to be significantly susceptible to T2DM. Inclusion of the 3′UTR variant for haplotype analysis in spite of the low prevalence of the minor allele “G” should not be a major consideration as the haplotype that showed significance in this study (ACGT) consisted of the major allele “T” of the T→G variant in the 3′UTR.

The major finding of this study is that the Asn130Asn [rs5435(AAC→AAT)] polymorphism in the GLUT4 gene was associated with T2DM. The other important finding is that the ACGT haplotype of the rs5418 (A→G), rs5435 (C→T), rs5421 (C→G), and the T→G variant in the 3′UTR showed significant protection against diabetes. This study also shows the presence of a novel T→G variant in the 3′UTR at nucleotide position 6787483 and its frequency were very low in both NGT and T2DM subjects. One of the limitations of this study is that because it is a cross-sectional one, no cause–effect relationship between the variants and T2DM could be established. A larger study is required to confirm the role of these GLUT4 variants in T2DM.

Footnotes

Acknowledgments

This study was supported by a grant from the Department of Biotechnology, Government of India. We also thank the Chennai Wellingdon Corporate Foundation, Chennai, India, for their financial support for the CURES field study. This is the 107^th publication from the CURES study (CURES 107).

Author Disclosure Statement

No competing financial interests exist.

Appendix

Table A1.

Linkage Disequilibrium Among the GLUT4 Polymorphisms

		r² values
Locus 1	Locus 2	NGT	Type 2 diabetes
rs5418	rs5435	0.498	0.462
rs5418	rs5421	0.001	0.01
rs5418	3′ UTR variant^*	0.002	0.011
rs5435	rs5421	0.002	0.005
rs5435	3′ UTR variant^*	0.002	0.014
rs5421	3′ UTR variant^*	0	0.007

Nucleotide position 6787483.

NGT, normal glucose tolerance; UTR, untranslated region.

References

Zierath

, Krook

, Wallberg-Henriksson

. Insulin action and insulin resistance in human skeletal muscle. Diabetologia, 2000; 43:821–835.

Dresner

, Laurent

, Marcucci

, Griffin

, Dufour

, Cline

, Slezak

, Andersen

, Hundal

, Rothman

, Petersen

, Shulman

. Effects of free fatty acids on glucose transport and IRS-1-associated phosphatidylinositol 3-kinase activity. J Clin Invest, 1999; 103:253–259.

Groop

, Bonadonna

, DelPrato

, Ratheiser

, Zyck

, Ferrannini

, DeFronzo

. Glucose, free fatty acid metabolism in non-insulin-dependent diabetes mellitus. Evidence for multiple sites of insulin resistance. J Clin Invest, 1989; 84:205–213.

Wild

, Roglic

, Green

, Sicree

, King

. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care, 2004; 27:1047–1053.

Mohan

, Sharp

, Cloke

, Burrin

, Schumer

, Kohner

. Serum immunoreactive insulin responses to a glucose load in Asian Indian and European type 2 (non-insulin-dependent) diabetic patients and control subjects. Diabetologia, 1986; 29:235–237.

Sharp

, Mohan

, Levy

, Mather

, Kohner

. Insulin resistance in patients of Asian Indian and European origin with non-insulin dependent diabetes. Horm Metab Res, 1987; 19:84–85.

McKeigue

, Pierpoint

, Ferrie

, Marmot

. Relationship of glucose intolerance and hyperinsulinaemia to body fat pattern in south Asians and Europeans. Diabetologia, 1992; 35:785–791.

Simmons

, Powell

. Metabolic and clinical characteristics of south Asians and Europeans in Coventry. Diabet Med, 1993; 10:751–758.

Misra

, Vikram

. Insulin resistance syndrome [metabolic syndrome] and Asian Indians. Curr Sci, 2002; 83:1483–1496.

10.

Deepa

, Pradeepa

, Rema

, Anjana

, Deepa

, Shanthirani

, Mohan

. The Chennai Urban Rural Epidemiology Study (CURES)—study design and methodology (urban component) (CURES-1) J Assoc Physicians India, 2003; 51:863–870.

11.

Alberti

, Zimmet

. Definition diagnosis, classification of diabetes mellitus, its complications. Part 1: Diagnosis and classification of diabetes mellitus, provisional report of a WHO Consultation. Diabet Med, 1998; 15:539–553.

12.

Barrett

, Fry

, Maller

, Daly

. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics, 2005; 21:263–265.

13.

Majumdar

, Majumder

. HAPLOPOP: A Computer Program to Obtain Maximum-Likelihood Estimates of Haplotype Frequencies from Genotype Data on a Set of Unrelated Individuals via the EM Algorithm. Technical Report AHGU-1/1999. Calcutta: Indian Statistical Institute, 1999.

14.

Matsutani

, Koranyi

, Cox

, Permutt

. Polymorphisms of GLUT2, GLUT4 genes. Use in evaluation of genetic susceptibility to NIDDM in blacks. Diabetes, 1990; 39:1534–1542.

15.

Baroni

, Oelbaum

, Pozzilli

, Stocks

, Li

, Fiore

, Galton

. Polymorphisms at the GLUT1 (HepG2) and GLUT4 (muscle/adipocyte) glucose transporter genes and non-insulin-dependent diabetes mellitus (NIDDM) Hum Genet, 1992; 88:557–561.

16.

O'Rahilly

, Krook

, Morgan

, Rees

, Flier

, Moller

. Insulin receptor and insulin-responsive glucose transporter (GLUT 4) mutations and polymorphisms in a Welsh type 2 (non-insulin-dependent) diabetic population. Diabetologia, 1992; 35:486–489.

17.

Oelbaum

. Analysis of three glucose transporter genes in a Caucasian population: no associations with non-insulin-dependent diabetes and obesity. Clin Genet, 1992; 42:260–266.

18.

Bjørbaek

, Echwald

, Hubricht

, Vestergaard

, Hansen

, Zierath

, Pedersen

. Genetic variants in promoters and coding regions of the muscle glycogen synthase and the insulin-responsive GLUT4 genes in NIDDM. Diabetes, 1994; 43:976–983.

19.

Pontiroli

, Capra

, Veglia

, Ferrari

, Xiang

, Bell

, Baroni

, Galton

, Weaver

, Hitman

, Kopelman

, Mohan

, Viswanathan

. Genetic contribution of polymorphism of the GLUT1 and GLUT4 genes to the susceptibility to type 2 (non-insulin-dependent) diabetes mellitus in different populations. Acta Diabetol, 1996; 33:193–197.

20.

Nair

, Sugunan

, Kumar

, Anilkumar

. Case-control analysis of SNPs in GLUT4, RBP4 and STRA6: association of SNPs in STRA6 with type 2 diabetes in a South Indian population. PLoS One, 2010; 5:e11444.

21.

Kusari

, Verma

, Buse

, Henry

, Olefsky

. Analysis of the gene sequences of the insulin receptor and the insulin-sensitive glucose transporter (GLUT-4) in patients with common-type non-insulin-dependent diabetes mellitus. J Clin Invest, 1991; 88:1323–1330.

22.

Muraoka

, Sakura

, Kim

, Kishimoto

, Akanuma

, Buse

, Yasuda

, Seino

, Bell

, Yazaki

, Kasuga

, Kadowaki

. Polymorphism in exon 4a of the human GLUT4/muscle-fat facilitative glucose transporter gene detected by SSCP. Nucleic Acids Res, 1991; 19:4313.

23.

Buse

, Yasuda

, Lay

, Seo

, Olson

, Pessin

, Karam

, Seino

, Bell

. Human GLUT4/muscle-fat glucose-transporter gene. Characterization and genetic variation. Diabetes, 1992; 41:1436–1445.

24.

Kimchi-Sarfaty

, Oh

, Kim

, Sauna

, Calcagno

, Ambudkar

, Gottesman

. A silent polymorphism in the MDR1 gene changes substrate specificity. Science, 2007; 315:525–528.

25.

Baier

, Permana

, Yang

, Pratley

, Hanson

, Shen

, Mott

, Knowler

, Cox

, Horikawa

, Oda

, Bell

, Bogardus

. A calpain-10 gene polymorphism is associated with reduced muscle mRNA levels and insulin resistance. J Clin Invest, 2000; 106:69–73.

26.

Alhopuro

, Katajisto

, Lehtonen

, Ylisaukko-oja

, Näätsaari

, Karhu

, Westerman

, Wilson

, de Rooij

, Vogel

, Moeslein

, Tomlinson

, Aaltonen

, Mäkelä

, Launonen

. Mutation analysis of three genes encoding novel LKB1-interacting proteins, BRG1, STRADa, and MO25a, in Peutz–Jeghers syndrome. Br J Cancer, 2005; 92:1126–1129.

27.

Fernandez-Cadenas

, Andreu

, Gamez

, Gonzalo

, Martin

, Rubio

, Arenas

. Splicing mosaic of the myophosphorylase gene due to a silent mutation in McArdle disease. Neurology, 2003; 61:1432–1434.

28.

Vimaleswaran

, Radha

, Ghosh

, Majumder

, Deepa

, Babu

, Rao

, Mohan

. Peroxisome proliferator-activated receptor-gamma co-activator-1alpha (PGC-1alpha) gene polymorphisms and their relationship to Type 2 diabetes in Asian Indians. Diabet Med, 2005; 22:1516–1521.

29.

Vimaleswaran

, Radha

, Deepa

, Mohan

. Absence of association of metabolic syndrome with PPARGC1A, PPARG and UCP1 gene polymorphisms in Asian Indians. Metab Syndr Relat Disord, 2007; 5:153–162.

30.

Vimaleswaran

, Radha

, Anjana

, Deepa

, Ghosh

, Majumder

, Rao

, Mohan

. Effects of polymorphisms in the PGC-1α gene on body fat distribution in Asian Indians. Int J Obes (Lond), 2006; 30:884–891.

31.

Rioux

, Daly

, Silverberg

, Lindblad

, Steinhart

, Cohen

, Delmonte

, Kocher

, Miller

, Guschwan

, Kulbokas

, O'Leary

, Winchester

, Dewar

, Green

, Stone

, Chow

, Cohen

, Langelier

, Lapointe

, Gaudet

, Faith

, Branco

, Bull

, McLeod

, Griffiths

, Bitton

, Greenberg

, Lander

, Siminovitch

, Hudson

. Genetic variation in the 5q31 cytokine gene cluster confers susceptibility to Crohn's disease. Nat Genet, 2001; 29:223–228.

32.

Van Eerdewegh

, Little

, Dupuis

, Del Mastro

, Falls

, Simon

, Torrey

, Pandit

, McKenny

, Braunschweiger

, Walsh

, Liu

, Hayward

, Folz

, Manning

, Bawa

, Saracino

, Thackston

, Benchekroun

, Capparell

, Wang

, Adair

, Feng

, Dubois

, FitzGerald

, Huang

, Gibson

, Allen

, Pedan

, Danzig

, Umland

, Egan

, Cuss

, Rorke

, Clough

, Holloway

, Holgate

, Keith

. Association of the ADAM33 gene with asthma and bronchial hyper-responsiveness. Nature, 2002; 18:426–430.

33.

Chumakov

, Blumenfeld

, Guerassimenko

, Cavarec

, Palicio

, Abderrahim

, Bougueleret

, Barry

, Tanaka

, La Rosa

, Puech

, Tahri

, Cohen-Akenine

, Delabrosse

, Lissarrague

, Picard

, Maurice

, Essioux

, Millasseau

, Grel

, Debailleul

, Simon

, Caterina

, Dufaure

, Malekzadeh

, Belova

, Luan

, Bouillot

, Sambucy

, Primas

, Saumier

, Boubkiri

, Martin-Saumier

, Nasroune

, Peixoto

, Delaye

, Pinchot

, Bastucci

, Guillou

, Chevillon

, Sainz-Fuertes

, Meguenni

, Aurich-Costa

, Cherif

, Gimalac

, Van Duijn

, Gauvreau

, Ouellette

, Fortier

, Raelson

, Sherbatich

, Riazanskaia

, Rogaev

, Raeymaekers

, Aerssens

, Konings

, Luyten

, Macciardi

, Sham

, Straub

, Weinberger

, Cohen

. Genetic and physiological data implicating the new human gene G72 and the gene for d-amino acid oxidase in schizophrenia. Proc Natl Acad Sci U S A, 2002; 99:13675–13680.

34.

Day

, Tuite

. Post-transcriptional gene regulatory mechanisms in eukaryotes: an overview. J Endocrinol, 1998; 157:361–371.

35.

Xia

, Sherer

, Cohen

, Major

, Xi

, Norman

, Knowler

, Bogardus

, Prochazka

. A common variant in PPP1R3 associated with insulin resistance and type 2 diabetes. Diabetes, 1998; 47:1519–1524.

36.

Frittitta

, Erccolino

, Bozzali

, Argiolas

, Graci

, Santagati

, Spampinato

, Di paola

, Cisternino

, Tassi

, Vigneri

, Pizzuti

, Trischitta

. A cluster of 3 single nucleotide polymorphisms in the 3′-untranslated region of human glycoprotein PC-1 gene stabilizes PC-mRNA and associates with increased PC-1 protein content and insulin resistance-related abnormalities. Diabetes, 2001; 50:1952–1955.

37.

Conne

, Stutz

, Vassalli

. The 3′ untranslated region of messenger RNA: a molecular “hotspot” for pathology? Nat Med, 2000; 6:637–641.

38.

Di Paola

, Frittitta

, Miscio

, Bozzali

, Barrata

, Centra

, Spampinato

, Santagati

, Ercolino

, Cisternino

, Soccio

, Mastroianno

, Tassi

, Almgren

, Pizzuti

, Vigneri

, Trischitta

. A variation in 3′ UTR of hPTP1B increases specific gene expression and associates with insulin resistance. Am J Hum Genet, 2002; 70:806–812.