| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
BRIEF REPORT |
Departments of Biochemistry (N.D.P., A.B.L., D.W.B.), Internal Medicine (D.W.B.), and Public Health Sciences (C.D.L., J.K.C.) and Center for Human Genomics (N.D.P., A.B.L., D.W.B.), Wake Forest University School of Medicine, Winston-Salem, North Carolina 27157, Department of Medicine (S.M.H.), University of Texas Health Sciences Center at San Antonio, San Antonio, Texas 78229; Department of Preventive Medicine and Biometrics (J.M.N.), University of Colorado Health Sciences Center, Denver, Colorado 80262; Department of Physiology and Biophysics (R.N.B.), Keck School of Medicine, University of Southern California, Los Angeles, California 90089; and Medical Genetics Institute (M.O.G., J.I.R.), Cedars-Sinai Medical Center, Los Angeles, California 90048
Address all correspondence and requests for reprints to: Donald W. Bowden, Ph.D., Department of Biochemistry, Wake Forest University School of Medicine, Winston-Salem, North Carolina 27157. E-mail: dbowden{at}wfubmc.edu.
| Abstract |
|---|
|
|
|---|
Objective: The objective of this study was to evaluate previously associated variants for association with measures of glucose homeostasis in Hispanic-Americans and African-Americans and determine the biological mechanism(s) through which these variants exert their effect.
Design: This study was the Insulin Resistance Atherosclerosis Family Study (IRAS-FS).
Setting: The IRAS-FS is a community-based study of Hispanic-Americans (San Antonio, TX, and San Luis Valley, CO) and African-Americans (Los Angeles, CA).
Participants: A total of 1040 Hispanic-American and 500 African-American individuals from the IRAS-FS formed the basis of this study.
Main Outcomes Measures(s): The primary glucose homeostasis phenotypes of interest in this study were derived from the frequently sampled iv glucose tolerance test and include insulin sensitivity, acute insulin response, and disposition index.
Results: In Hispanic-Americans, significant evidence of association was observed between single-nucleotide polymorphisms rs7903146 and rs112255372 with reduced insulin secretion as measured by acute insulin response and adjusted for the degree of insulin sensitivity (P = 0.032 and 0.036, respectively). Other quantitative measures, e.g. insulin sensitivity or disposition index, were not associated with the single nucleotide polymorphisms examined. In African-Americans there was no evidence of association observed.
Conclusions: These results suggest that transcription factor 7-like 2 variants could play a role in the pathogenesis of type 2 diabetes in the Hispanic-American population through a mechanism involving insulin secretion.
| Introduction |
|---|
|
|
|---|
TCF7L2 is located on 10q25.2, a region linked to type 2 diabetes (T2D) in Mexican-Americans (4) and an Icelandic population (5). Grant et al. (6) reported association of a microsatellite marker (DG10S478) in intron 3 of the TCF7L2 gene with T2D in the Icelandic population. Additionally, single nucleotide polymorphisms (SNPs) were genotyped to evaluate the correlation of alleles within the linkage disequilibrium (LD) block and found to be significantly associated with T2D. Specifically, the strongest evidence of association was found with the T allele of SNPs rs7903146 and rs12255372. These findings were replicated in Danish female and European-American cohorts (6). Subsequently several studies reported similar findings in predominantly European-derived populations (7, 8, 9, 10, 11, 12, 13, 14, 15).
The purpose of this study was to evaluate association of TCF7L2 SNPs with quantitative measures of glucose homeostasis in Hispanic- and African-American subjects from the Insulin Resistance Atherosclerosis Family Study (IRAS-FS). Quantitative trait analysis in appropriately phenotyped families is a powerful approach for gaining insights into complex biological diseases, e.g. T2D. Our extensive phenotypic data should facilitate assessment of the genetic impact of TCF7L2 SNPs on metabolic pathways (i.e. insulin sensitivity or insulin secretion) in which the protein product exerts its effects.
| Subjects and Methods |
|---|
|
|
|---|
Study design and recruitment have been described in detail (16). Briefly, the IRAS-FS is a multicenter study designed to identify the genetic determinants of glucose homeostasis. Large families of Hispanic-Americans (n = 1417; San Antonio, TX; San Luis Valley, CO) and African-Americans (n = 605; Los Angeles, CA) were recruited. A clinical examination included frequently sampled iv glucose tolerance test (FSIGT), anthropometric measurements, and blood collection. Specific to this report, measures of glucose homeostasis included those from the FSIGT (17, 18, 19) calculated by mathematical modeling methods (MINMOD) (20): insulin sensitivity (SI), acute insulin response (AIR), and disposition index (DI).
Genotyping
Five SNPs (rs7901695, rs7903146, rs7895340, rs11196205, and rs12255372) previously reported (6) to be correlated with marker DG10S478 were genotyped. Genotyping was performed using the Sequenom MassArray. Seventy blind duplicates were included to evaluate genotyping accuracy.
Statistical analysis
Mendelian inconsistencies were identified with PedCheck (21) and converted to missing. Maximum likelihood estimates of allele frequencies were computed using unrelated Hispanic- and African-American individuals (n = 229 and 58, respectively), tested for departures from Hardy-Weinberg proportions and used to compute LD statistics. Association between individual SNPs and each quantitative phenotype was performed in SOLAR (22) using variance component analysis. When necessary, quantitative traits were transformed to approximate the distributional assumptions of the test and minimize variance heterogeneity. For each phenotype, the two degree of freedom test of genotypic association was performed. Adjustments for multiple comparison tests were not performed due to the high correlation between the SNPs examined and the a priori hypotheses. Subjects with T2D were excluded from the analysis of glucose homeostasis traits.
| Results |
|---|
|
|
|---|
|
Association results are summarized in Tables 2
and 3
. In Hispanic-Americans (Table 2
), SNPs rs7903146 and rs12255372 were significantly associated (P = 0.032 and 0.036, respectively) with AIR. These associations were consistent with an additive model in which copy number of the minor allele T corresponded to the effect size as evidenced from the phenotypic means. Association with AIR in Hispanic-Americans is driven by the San Antonio subpopulation, whereas no associations were seen in the San Luis Valley population when analyzed separately (appendix Table 1).
|
|
| Discussion |
|---|
|
|
|---|
In Hispanic-Americans the SNPs genotyped were contained in two LD blocks. In comparison with European ancestry centre dEtude du Polymorphisme Humain (CEPH) samples genotyped by HapMap, block boundaries are disparate when using the confidence interval method with four SNPs (rs7901695-rs11196205) comprising one block; however, block boundaries are the same when using the solid spine method. Slight differences in block structure could be attributed to slightly increased minor allele frequencies (MAFs) in Hispanic-Americans possibly due to modest admixture.
Significant association was observed between AIR adjusted for age, gender, recruitment center, BMI, and SI and rs7903146 and rs12255372 (P = 0.036 and 0.032, respectively), which were strongly associated with T2D in the initial report (6). Exclusion of SI as a covariate resulted in marginally less significant associations (P = 0.055 and 0.047, respectively). These associations were driven by the San Antonio population and are not attributed to power issues; individual cohorts have approximately equivalent power or phenotypic differences for AIR (P = 0.034) but more likely due to increased MAF of variants in the San Luis Valley population or environmental contributions. The minor allele T for both SNPs, associated with T2D risk (6, 7, 8, 9, 10, 11, 12, 13, 14, 15), were associated with decreased AIR following an additive model. Notably, the exclusion of BMI as a covariate resulted in marginally more significant associations (P = 0.034 and 0.051, respectively), suggesting that adiposity is not a major contributor (data not shown). Although modestly associated with reduced AIR, the T alleles for rs7903146 and rs12255372 were not associated with future T2D in Hispanic-Americans, as assessed during follow-up studies of limited sample size (n = 65; data not shown). Such an association has been observed in studies of European-derived populations (23). The marginal associations observed with insulin secretion calculated from the FSIGT as compared with the oral glucose tolerance test (OGTT) (23), which was not performed in IRAS-FS, are consistent with emerging data suggesting variants in TCF7L2 effect the enteroinsular axis (23).
Evidence of association with only two of five SNPs could be attributed to low correlation (r2 < 0.54) between associated variants in this study and other SNPs. Lack of evidence for association with SI is unlikely a false-negative result, given the study has 80% power when examining a variant with a MAF = 0.20, to detect an SI difference as small as 15%. In addition, there was no evidence of association with fasting glucose observed (appendix Table 2). Evaluation of TCF7L2 SNPs for association with T2D resulted in a lack of association (data not shown), which could be attributed modest sample size (n = 186).
Additional studies have evaluated TCF7L2 variants in Hispanic-Americans (8, 24, 25). In comparison with the study by Florez et al. (8), MAFs for SNPs rs7903146 and rs12255372 were slightly higher (0.27 and 0.24, respectively; n = 497), but direct comparison of association analysis results is not possible due to their combining of ethnicities for analysis. Lehman et al. (24) evaluated TCF7L2 variants in a T2D cohort containing a nondiabetic subset (n = 367) but that did not have FSIGT data. Similar to results presented here, they observed modest attributable risk, compared with the magnitude of associations observed in European-derived populations. Using an iv glucose tolerance test, Watanabe et al. (25) assessed the impact of previously associated TCF7L2 variants (rs7901695 and rs12255372; n < 537) on measures of glucose homeostasis and observed no evidence of association with AIR. In the same study, rs12255372 was associated with fasting plasma insulin concentration from an OGTT modified by percent body fat, an effect not directly assessed in IRAS-FS due to absence of OGTT data. Overall the results reported here are from a sample that is comparable in size with each of these prior studies combined.
In parallel with the association studies in Hispanic-Americans, we also evaluated these variants in a similarly phenotyped African-American population. The SNPs genotyped formed two LD blocks with variable intrablock SNP correlations. The same set of quantitative traits was evaluated and no evidence of association was observed. This lack of association could be attributed to an ungenotyped variant(s), which is tagged by this SNP subset in Hispanic-Americans due to LD and MAF differences. Alternatively, these results could be attributed to ethnic-specific effects because previous associations have been limited to predominantly European-derived populations. This report is contrary to the findings of Munoz et al. (26), which reported association of rs12255372 with a reduction in DI and AIRG adjusted for SI but consistent with the lack of association with insulin secretion observed by Elbein et al. (27), both in African-Americans. One potential source of variation between the conflicting report could be attributed to age of the recruited subjects. IRAS-FS had a mean recruitment age of 42.9 ± 14, whereas Munoz et al. recruited individuals aged 7–57 yr for a mean age of 31 ± 12 yr. Because T2D has a later age of onset, the hallmarks of the disease, insulin resistance and defective insulin secretion, may not be pronounced in younger, genetically predisposed individuals. This rationale could explain their loss of significance after adjustment for age (26).
Previous studies in other ethnicities examining the influence of TCF7L2 on measures of glucose metabolism are also inconsistent. Methods chosen to derive measures of glucose homeostasis and sample sizes differ significantly between studies. Two recent studies derived quantitative measures of glucose homeostasis from the OGTT (8, 13). Florez et al. (8) evaluated two correlated measures of insulin secretion (fasting insulin to fasting glucose ratio and corrected insulin response) and two correlated measures of insulin sensitivity (reciprocal of insulin resistance and the insulin sensitivity index) in a multiethnic cohort (n = 554). This study observed the T allele of SNPs rs7901695 and rs12255372 to be associated with decreased insulin secretion and increased insulin sensitivity. Saxena et al. (13) evaluated measures of insulin resistance and insulin secretion in a northern European-derived sample (n = 995 and 721, respectively) and found a decrease in insulin secretion associated with the T allele of rs7903146. The authors concluded that TCF7L2 variants influence first-phase insulin response and overall capacity of β-cells to secrete insulin. Complimentary to this approach, Damcott et al. (12) used the iv glucose tolerance test in a sample of European-American subjects (n = 48) and observed evidence of association with SI, AIR, and DI, although the magnitude of the effect was reduced after adjustment for BMI.
This study represents an attempt to assess the contribution of TCF7L2 variants to alterations in quantitative measures of glucose homeostasis in large Hispanic- and African-American cohorts. Whereas this study does not elucidate a causal variant, it supports a probable mechanism whereby variants contribute to T2D susceptibility by decreasing insulin secretion. Further investigations of TCF7L2 across additional ethnicities are warranted to evaluate possible ethnic-specific effects. In addition, functional studies are needed to confirm associations observed with insulin secretion.
| Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online October 30, 2007
Abbreviations: AIR, Acute insulin response; BMI, body mass index; DI, disposition index; FSIGT, frequently sampled iv glucose tolerance test; IRAS-FS, Insulin Resistance Atherosclerosis Family Study; LD, linkage disequilibrium; MAF, minor allele frequency; OGTT, oral glucose tolerance test; SI, insulin sensitivity; SNP, single nucleotide polymorphism; TCF7L2, transcription factor 7-like 2 locus; T2D, type 2 diabetes.
Received June 1, 2007.
Accepted October 24, 2007.
| References |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |