Common Polymorphisms in the Adiponectin Gene ACDC Are Not Associated With Diabetes in Pima Indians
http://www.100md.com
糖尿病学杂志 2005年第1期
1 Clinical Diabetes and Nutrition Section, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Phoenix, Arizona
2 Department of Internal Medicine and Molecular Science, Osaka University, Graduate School of Medicine, Osaka, Japan
3 Institute Pasteur de Lille, Institute of Biology, Centre National de la Recherche Scientifique (CNRS), Lille, France
4 Diabetes Research Unit, Translational Genomics Research Institute, Phoenix, Arizona
ABSTRACT
Adiponectin is an abundant adipose tissue-derived protein with important metabolic effects. Plasma adiponectin levels are decreased in obese individuals, and low adiponectin levels predict insulin resistance and type 2 diabetes. Two variants in the adiponectin gene ACDC have been previously associated with plasma adiponectin levels, obesity, insulin resistance, and type 2 diabetes. To determine the role of genetic variation in ACDC in susceptibility to obesity and type 2 diabetes in Pima Indians, we screened the promoter, exons, and exon-intron boundaries of the gene to identify allelic variants. We identified 17 informative polymorphisms that comprised four common (minor allele frequency >15%) linkage disequilibrium clusters consisting of 1eC4 variants each. We genotyped one representative polymorphism from each cluster in 1,338 individuals and assessed genotypic association with type 2 diabetes, BMI, serum lipid levels, serum adiponectin levels, and measures of insulin sensitivity and secretion. None of the ACDC variants were associated with type 2 diabetes, BMI, or measures of insulin sensitivity or secretion. One variant, single nucleotide polymorphism (SNP)-12823, was associated with serum adiponectin levels (P = 0.002), but this association explained only 2% of the variance of serum adiponectin levels. Our findings suggest that these common ACDC polymorphisms do not play a major role in susceptibility to obesity or type 2 diabetes in this population.
Adiponectin is an adipose tissue-derived protein with important metabolic effects (1). Plasma adiponectin levels are decreased in individuals with obesity (2), insulin resistance (3), and type 2 diabetes (2), and low plasma adiponectin levels predict the development of insulin resistance (4) and type 2 diabetes (5) in Pima Indians.
Recent genome-wide scans in humans have mapped a susceptibility locus for type 2 diabetes and the metabolic syndrome to chromosome 3q27, where the gene encoding adiponectin, ACDC, is located (6,7). Two single nucleotide polymorphisms (SNPs) in ACDC were found to be associated with type 2 diabetes in Japanese individuals (8). One of the SNPs (TG in codon 276) was also associated with lower plasma adiponectin levels, although this relationship was limited to obese subjects (8). Another SNP (TG at nucleotide 94: SNP 45) was found to be associated with obesity, insulin resistance, and dyslipidemia in a German population (9). A haplotype defined by these two SNPs has also been associated with components of the insulin resistance syndrome in Caucasians (10).
The ACDC gene spans 16 kb, contains three exons, and yields a 4.5-kb mRNA transcript. We screened 8 kb of ACDC corresponding to 2.8 kb of promoter sequence, each exon, and flanking intronic sequence and identified 17 SNPs. These included five promoter, seven exonic, and five intronic SNPs. At the time of the present study, five SNPs (SNP-12823, SNP-12128, SNP 3187, SNP 3267, and SNP 3286) were confirmed in a public database (available at http://www.ncbi.nlm.nih.gov/SNP) and correspond to dbSNP entries rs860291, rs266730, rs1063537, rs2082940, and rs1063538, respectively.
The genotypic distribution among 100 individuals partitioned into high- and low-plasma adiponectin level extremes (following adjustment for percent body fat) allowed us to make certain assumptions regarding the amount of disequilibrium between SNPs. When genotypes at SNPs were in >90% concordance among the 100 individuals, we grouped them into clusters and only genotyped one representative SNP from each cluster. On this basis, the 17 SNPs identified in Pimas were divided into 8 clusters, each containing 1eC4 SNPs (Fig. 1). Clusters 3, 7, and 8 consisted of SNPs with a minor allele frequency <15% and, as the goal of this study was to evaluate relatively common variants, these SNPs were not evaluated further. Cluster 5 also had a minor allele that was relatively rare; however, we typed SNP 45 from this cluster because previous studies showed association with diabetes or related traits.
We genotyped SNPs representative of each nonrare linkage disequilibrium cluster corresponding to SNP-12823, SNP-11365, SNP 45, SNP 276, and SNP 3286 in 1,338 participants from a genome-wide linkage study (11). We first evaluated linkage disequilibrium, as quantified by D', between pairs of these SNPs, as shown in Table 1. Each of the SNPs was in statistically significant linkage disequilibrium with each of the others, but the extent of association varied somewhat, with D' ranging from 0.48 to 0.99. Furthermore, because the associated alleles often had different frequencies, the information associated with each SNP varied considerably.
We assessed the association of each SNP with type 2 diabetes and BMI. In a subgroup of 331 nondiabetic individuals who had undergone detailed physiologic studies (12), we also assessed the association with insulin sensitivity, measured by the hyperinsulinemic-euglycemic clamp and insulin secretion, as assessed by the acute insulin response to an intravenous glucose infusion. In the larger group of nondiabetic individuals, the insulin sensitivity index (ISI) and corrected insulin response (CIR) were analyzed. These are surrogate indexes of insulin sensitivity and insulin secretion, respectively, that are correlated with the more sophisticated measures (13) and that can be derived from an oral glucose tolerance test. ISI was calculated from the fasting serum insulin (I0) and fasting plasma glucose (G0) concentrations (ISI = 104/[I0G0]), while CIR was calculated from the 2-h postload insulin (I2) and glucose (G2) concentrations (CIR = I2/G2[G2B70 mg/dl]) as described (14,15). The association of these SNPs with serum adiponectin levels, as well as levels of triglycerides and HDL choleseterol, was also assessed among nondiabetic individuals with normal renal function (16).
As shown in Table 2, none of the SNPs were associated with type 2 diabetes, BMI, ISI, CIR, or directly measured insulin sensitivity or secretion. There was a modest assocation of the G allele at SNP-11365 with lower HDL cholesterol levels. We also found that the G allele of SNP-12823 was significantly associated with lower adiponectin levels. Further adjustment for BMI did not substantially change these results. While statistically significant, the effect of this polymorphism accounted for only 2% of the variance of serum adiponectin levels.
We identified six common ACDC haplotypes, none of which were significantly associated with type 2 diabetes, BMI, ISI, CIR, lipid levels, or directly measured insulin sensitivity or secretion (Table 3). We did observe an association between the A_G_T_C_C haplotype and serum adiponectin levels, but because this haplotype is the only one containing the A allele of SNP-12823, it essentially reflects the relationship between this variant and serum adiponectin previously found in the individual SNP analyses.
Adiponectin has been recently identified as the most abundant of the known adipose tissue-derived proteins (1) with insulin-sensitizing and anti-inflammatory properties (1,17). Recently, the G allele of SNP 276 (GT) was found to be associated with low plasma adiponectin levels in obese Japanese individuals (8). Moreover, individuals homozygous for the haplotype consisting of the G allele at position 276 and the T allele at position 45 had significantly lower adiponectin levels than those who did not carry the haplotype (10). In the present study, no association between SNPs 276 and 45 (or their haplotypes) and protein levels was found in Pima Indians, although another variant, SNP-12823, which is located within the ACDC promoter, was associated with serum adiponectin levels. However, it is worth noting that this SNP can only explain 2% of the variance in serum adiponectin levels. Thus, the present results suggest that none of these SNPs in ACDC have major effects on serum adiponectin concentrations.
In genome-wide linkage analyses in the present group of individuals we found that while 39% of the variance in serum adiponectin is potentially due to genetic factors (i.e., heritability = 0.39), there was no evidence for linkage on chromosome 3q27, where ACDC is located (16). Instead, significant linkage (logarithm of odds = 3.0 at 18 cM) was seen on chromosome 9p and modest evidence for linkage (logarithm of odds = 1.0 at 124 cM) was found on chromosome 3q13, 70 cM centromeric to ACDC. Comuzzie et al. (18) have identified two major loci influencing adiponectin expression in European Americans on chromosomes 5 and 14 and modest evidence for linkage (logarithm of odds = 1.3) on 3q27 near ACDC.
Although some studies have reported suggestive evidence for the linkage of obesity with 3q27 (7), genome-wide linkage analyses of BMI in Pima Indians did not show strong evidence for linkage to the ACDC region (11). Furthermore, in the present study we found no evidence for association between ACDC SNPs and obesity in Pima Indians. These findings are consistent with results obtained in a Japanese population showing no evidence for a significant association with BMI (8). However, the role of ACDC SNPs in mediating susceptibility to obesity has been inconclusive. One study has reported an association between the variant G allele of SNP 45 and obesity in a German population (9), while results derived from other populations have found that the T allele at this position was associated with a higher body weight (10). Although the factors underlying the disparities among these studies are not yet known, it is possible that different environmental exposures, different patterns of linkage disequilibrium among populations, and sampling variation may play a role in the reported findings. It is also worth noting that the variant allele of SNP 45 is relatively rare and as such may produce spurious associations.
Low plasma adiponectin levels predict the development of type 2 diabetes in Pimas (5), and this has been confirmed in a German population (19). Linkage of the region near ACDC with type 2 diabetes has been reported in French and Japanese populations (6,20) and with fasting insulin levels in European Americans (7), thereby suggesting that a locus in this region may influence insulin sensitivity and risk for diabetes. In the Pimas, however, there is little evidence for linkage with either type 2 diabetes (11) or measures of insulin sensitivity (12) in this region. Because we have also not detected an association between ACDC SNPs and insulin sensitivity or type 2 diabetes in Pima Indians, it is likely that this gene does not play a major role in conferring increased susceptibility to these disorders in this population.
RESEARCH DESIGN AND METHODS
Subjects who participated in this study included 1,338 Pima individuals (332 nuclear families and 112 extended pedigrees) who were previously selected for linkage studies (11) from participants in ongoing longitudinal studies of obesity and type 2 diabetes conducted in the Gila River Indian Community. Among the siblings in the linkage study, there were 1,080 (496 men and 584 women) who were typed for at least one of the ACDC polymorphisms. The prevalence of type 2 diabetes in these participants was 59%. The mean (±SD) age was 38.7 ± 13.7 years, mean BMI was 36.6 ± 8.2 kg/m2, and mean serum adiponectin was 5.46 ± 2.69 e/ml. A subset of 331 of these individuals had participated in detailed physiologic studies, including measurement of directly assessed insulin sensitivity at physiologic insulin concentrations by the hyperinsulinemic-euglycemic clamp and measurement of insulin secretion by the acute insulin response calculated at 3eC5 min after a 25-g intravenous glucose bolus (see Pratley et al. [12] for further details). In 639 individuals, insulin and glucose concentrations had been measured during an oral glucose tolerance test at their last nondiabetic examination, and these measures were used to calculate surrogate measures of insulin sensitivity and secretion, ISI, and CIR (13eC15). Serum adiponectin was measured in stored sera from most of these participants (16,21). Since adiponectin concentrations are strongly influenced by the presence of diabetes and renal function (21), these analyses were restricted to nondiabetic participants with normal renal function (n = 580). In those who had been examined since 1992 (n = 212), serum HDL cholesterol and triglycerides were measured. This study was approved by the institutional review board of the National Institute of Diabetes and Digestive and Kidney Diseases and the Tribal Council of the Gila River Indian Community. All subjects provided signed informed consent before participation.
ACDC genomic screening and SNP genotyping.
ACDC was screened using denaturing high-performance liquid chromatography and amplicons, which yielded denaturing high-performance liquid chromatography heteroduplexes, which were validated by direct sequencing in the 50 individuals with the highest serum adiponectin levels and the 50 individuals with the lowest adiponectin levels as described (22).
We genotyped SNP-12823, SNP-11365, SNP 45, and SNP 276 using pyrosequencing according to the manufacturer’s recommendations (Pyrosequencing, Uppsala, Sweden) with minor modifications as described (23). SNP 3286 was genotyped by PCR restriction fragment-length polymorphism by first amplifying a PCR fragment of 397 bp and digesting it overnight at 371 with Nsp I (New England Biolabs, Beverly, MA). In the presence of the C allele, the PCR product was restricted into two fragments of 95 and 302 bp, while the T allele yielded an uncut fragment. Sequence information for all genotyping primers is available upon request.
Statistical analyses.
Haplotype frequencies for pairs of SNPs were estimated with a pedigree-based maximum likelihood method implemented in the ILINK program (24). The degree of linkage disequilibrium between alleles for each pair of loci was expressed as D' , which represents the proportion of the maximum possible allelic association given the allele frequencies and the direction of association. The likelihood ratio test was used to assess statistical significance of the observed allelic associations.
Association between alleles at each SNP and prevalence of diabetes was assessed by logistic regression with control for the effects of age, sex, birth year, and ethnicity (i.e., regardless of whether the participant was of full Pima heritage). The proportion of participants who were full Pima/Tohono O’odham was 79%; 94% of the remaining subjects were >50% Pima/Tohono O’odham. To account for familial relationships (i.e., sibship), these analyses were conducted using generalized estimating equations as previously described (23). The association of each SNP with type 2 diabetes was analyzed under an additive model in which the logarithm of the odds ratio is expressed as a function of the number of copies of the more common allele. Additional analyses were conducted under models in which the effect of the common allele was assumed to be dominant or recessive; as the conclusions of these models were identical to those of the additive models, only results for the additive models are presented here. For analysis of continuous variables (e.g., BMI, lipids, serum adiponectin concentrations, and measures of insulin sensitivity and secretion), linear regression analyses were conducted in an analogous fashion to the logistic models for type 2 diabetes. The generalized estimating equations were fit using PROC GENMOD of SAS (SAS Institute, Cary, NC).
A modification of the zero-recombinant haplotyping technique (25) was used to analyze the associations of haplotypes in the five genotyped SNPs in the region with the four different traits. First, the EH algorithm was used to estimate haplotype frequencies. There were six common haplotypes with frequencies >0.05, which accounted for 98% of the observed haplotypes among the five SNPs. Next, the MLINK program (23) was used to estimate the probability that each individual carried each possible haplotype given these frequencies, the genotypes of the individual, and the genotypes of other individuals in the pedigree. As only four SNPs were necessary to assign individuals to these haplotypes, SNP 3286 was omitted from the likelihood calculations. The probabilities derived from MLINK were then utilized in logistic or linear regression analyses to evaluate the association of each of the common haplotypes with the traits of interest. Since this approach requires the assumption of no recombination across the region, individuals whose genotypes produced obligate recombinations (n = 9) were deleted from these analyses. These analyses were conducted with six different common haplotypes; therefore, a corrected P value (Pcorr) was calculated using a Bonferroni correction [Pcorr = 1 eC (1eC P)] (5) to account for multiple comparisons.
ACKNOWLEDGMENTS
We thank Victoria M. Ossowski and Timothy Monette for technical assistance and Dr. Jonathan Krakoff for his advice. We are also grateful for the support and participation of the Gila River Indian Community.
CIR, corrected insulin response; ISI, insulin sensitivity index; SNP, single nucleotide polymorphism
REFERENCES
Scherer PE, Williams S, Fogliano M, Baldini G, Lodish HF: A novel serum protein similar to C1q, produced exclusively in adipocytes. J Biol Chem270 :26746 eC26749,1995
Weyer C, Funahashi T, Tanaka S, Hotta K, Matsuzawa Y, Pratley RE, Tataranni PA: Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia. J Clin Endocrinol Metab86 :1930 eC1935,2001
Arita Y, Kihara S, Ouchi N, Takahashi M, Maeda K, Miyagawa J, Hotta K, Shimomura I, Nakamura T, Miyaoka K, Kuriyama H, Nishida M, Yamashita S, Okubo K, Matsubara K, Muraguchi M, Ohmoto Y, Funahashi T, Matsuzawa Y: Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity. Biochem Biophys Res Commun257 :79 eC83,1999
Stefan N, Vozarova B, Funahashi T, Matsuzawa Y, Weyer C, Lindsay RS, Youngren JF, Havel PJ, Pratley RE, Bogardus C, Tataranni PA: Plasma adiponectin concentration is associated with skeletal muscle insulin receptor tyrosine phosphorylation, and low plasma concentration precedes a decrease in whole-body insulin sensitivity in humans. Diabetes51 :1884 eC1888,2002
Lindsay RS, Funahashi T, Hanson RL, Matsuzawa Y, Tanaka S, Tataranni PA, Knowler WC, Krakoff J: Adiponectin and development of type 2 diabetes in the Pima Indian population. Lancet360 :57 eC58,2002
Vionnet N, Hani El H, Dupont S, Gallina S, Francke S, Dotte S, De Matos F, Durand E, Lepretre F, Lecoeur C, Gallina P, Zekiri L, Dina C, Froguel P: Genomewide search for type 2 diabetes-susceptibility genes in French whites: evidence for a novel susceptibility locus for early onset diabetes on chromosome 3q27-qter and independent replication of a type 2-diabetes locus on chromosome 1q21eCq24. Am J Hum Genet67 :1470 eC1480,2000
Kissebah AH, Sonnenberg GE, Myklebust J, Goldstein M, Broman K, James RG, Marks JA, Krakower GR, Jacob HJ, Weber J, Martin L, Blangero J, Comuzzie AG: Quantitative trait loci on chromosomes 3 and 17 influence phenotypes of the metabolic syndrome. Proc Natl Acad Sci U S A97 :14478 eC14483,2000
Hara K, Boutin P, Mori Y, Tobe K, Dina C, Yasuda K, Yamauchi T, Otabe S, Okada T, Eto K, Kadowaki H, Hagura R, Akanuma Y, Yazaki Y, Nagai R, Taniyama M, Matsubara K, Yoda M, Nakano Y, Tomita M, Kimura S, Ito C, Froguel P, Kadowaki T: Genetic variation in the gene encoding adiponectin is associated with an increased risk of type 2 diabetes in the Japanese population. Diabetes51 :536 eC540,2002
Stumvoll M, Tschritter O, Fritsche A, Staiger H, Renn W, Weisser M, Machicao F, Haring H: Association of the T-G polymorphism in adiponectin (exon 2) with obesity and insulin sensitivity: interaction with family history of type 2 diabetes. Diabetes51 :37 eC41,2002
Menzaghi C, Ercolino T, Di Paola R, Berg AH, Warram JH, Scherer PE, Trischitta V, Doria A: A haplotype at the adiponectin locus is associated with obesity and other features of the insulin resistance syndrome. Diabetes51 :2306 eC2312,2002
Hanson RL, Ehm MG, Pettitt DJ, Prochazka M, Thompson DB, Timberlake D, Foroud T, Kobes S, Baier L, Burns DK, Almasy L, Blangero J, Garvey WT, Bennett PH, Knowler WC: An autosomal genomic scan for loci linked to type II diabetes mellitus and body-mass index in Pima Indians. Am J Hum Genet63 :1130 eC1138,1998
Pratley RE, Thompson DB, Prochazka M, Baier L, Mott D, Ravussin E, Sakul H, Ehm MG, Burns DK, Foroud T, Garvey WT, Hanson RL, Knowler WC, Bennett PH, Bogardus C: An autosomal genomic scan for loci linked to prediabetic phenotypes in Pima Indians. J Clin Invest101 :1757 eC1764,1998
Hanson RL, Pratley RE, Bogardus C, Narayan KM, Roumain JM, Imperatore G, Fagot-Campagna A, Pettitt DJ, Bennett PH, Knowler WC: Evaluation of simple indices of insulin sensitivity and insulin secretion for use in epidemiologic studies. Am J Epidemiol151 :190 eC198,2000
Sluiter WJ, Erkelens DW, Terpstra P, Reitsma WD, Doorenbos H: Glucose tolerance and insulin release, a mathematical approach. II. Approximation of the peripheral insulin resistance after oral glucose loading. Diabetes25 :245 eC249,1976
Sluiter WJ, Erkelens Dw, Reitsma WC, Doorenbos H: Glucose tolerance and insulin release, a mathematical approach. I. Assay of the beta-cell response after oral glucose loading. Diabetes25 :241 eC244,1976
Lindsay RS, Funahashi T, Krakoff J, Matsuzawa Y, Tanaka S, Kobes S, Bennett PH, Tataranni PA, Knowler WC, Hanson RL: Genome-wide linkage analysis of serum adiponectin in the Pima Indian population. Diabetes52 :2419 eC2425,2003
Berg AH, Combs TP, Du X, Brownlee M, Scherer PE: The adipocyte-secreted protein Acrp30 enhances hepatic insulin action. Nat Med7 :947 eC953,2001
Comuzzie AG, Funahashi T, Sonnenberg G, Martin LJ, Jacob HJ, Black AE, Maas D, Takahashi M, Kihara S, Tanaka S, Matsuzawa Y, Blangero J, Cohen D, Kissebah A: The genetic basis of plasma variation in adiponectin: a global endophenotype for obesity and the metabolic syndrome. J Clin Endocrinol Metab86 :4321 eC4325,2001
Spranger J, Kroke A, Mohlig M, Bergmann MM, Ristow M, Boeing H, Pfeiffer AF: Adiponectin and protection against type 2 diabetes mellitus. Lancet361 :226 eC228,2003
Mori Y, Otabe S, Dina C, Yasuda K, Populaire C, Lecoeur C, Vatin V, Durand E, Hara K, Okada T, Tobe K, Boutin P, Kadowaki T, Froguel P: Genome-wide search for type 2 diabetes in Japanese affected sib-pairs confirms susceptibility genes on 3q, 15q, and 20q and identifies two new candidate Loci on 7p and 11p. Diabetes51 :1247 eC1255,2002
Looker HC, Krakoff J, Funahashi T, Matsuzawa Y, Tanaka S, Nelson RG, Knowler WC, Lindsay RS, Hanson RL: Adiponectin concentrations are influenced by renal function and duration of diabetes in Pima Indians with type 2 diabetes. J Clin Endocrinol Metab89 :4010 eC4017,2004
Wolford JK, Blunt D, Ballecer C, Prochazka M: High-throughput SNP detection by using DNA pooling and denaturing high performance liquid chromatography (DHPLC). Hum Genet107 :483 eC487,2000
Wolford JK, Gruber JD, Ossowski VM, Vozarova B, Antonio Tataranni P, Bogardus C, Hanson RL: A C-reactive protein promoter polymorphism is associated with type 2 diabetes mellitus in Pima Indians. Mol Genet Metab78 :136 eC144,2003
Lathrop GM, Lalouel JM: Easy calculations of lod scores and genetic risks on small computers. Am J Hum Genet36 :460 eC465,1984
O’Connell JR: Zero-recombinant haplotyping: applications to fine mapping using SNPs. Genet Epidemiol19 (Suppl. 1) :S64 eCS70,2000(Barbora Vozarova de Court)
2 Department of Internal Medicine and Molecular Science, Osaka University, Graduate School of Medicine, Osaka, Japan
3 Institute Pasteur de Lille, Institute of Biology, Centre National de la Recherche Scientifique (CNRS), Lille, France
4 Diabetes Research Unit, Translational Genomics Research Institute, Phoenix, Arizona
ABSTRACT
Adiponectin is an abundant adipose tissue-derived protein with important metabolic effects. Plasma adiponectin levels are decreased in obese individuals, and low adiponectin levels predict insulin resistance and type 2 diabetes. Two variants in the adiponectin gene ACDC have been previously associated with plasma adiponectin levels, obesity, insulin resistance, and type 2 diabetes. To determine the role of genetic variation in ACDC in susceptibility to obesity and type 2 diabetes in Pima Indians, we screened the promoter, exons, and exon-intron boundaries of the gene to identify allelic variants. We identified 17 informative polymorphisms that comprised four common (minor allele frequency >15%) linkage disequilibrium clusters consisting of 1eC4 variants each. We genotyped one representative polymorphism from each cluster in 1,338 individuals and assessed genotypic association with type 2 diabetes, BMI, serum lipid levels, serum adiponectin levels, and measures of insulin sensitivity and secretion. None of the ACDC variants were associated with type 2 diabetes, BMI, or measures of insulin sensitivity or secretion. One variant, single nucleotide polymorphism (SNP)-12823, was associated with serum adiponectin levels (P = 0.002), but this association explained only 2% of the variance of serum adiponectin levels. Our findings suggest that these common ACDC polymorphisms do not play a major role in susceptibility to obesity or type 2 diabetes in this population.
Adiponectin is an adipose tissue-derived protein with important metabolic effects (1). Plasma adiponectin levels are decreased in individuals with obesity (2), insulin resistance (3), and type 2 diabetes (2), and low plasma adiponectin levels predict the development of insulin resistance (4) and type 2 diabetes (5) in Pima Indians.
Recent genome-wide scans in humans have mapped a susceptibility locus for type 2 diabetes and the metabolic syndrome to chromosome 3q27, where the gene encoding adiponectin, ACDC, is located (6,7). Two single nucleotide polymorphisms (SNPs) in ACDC were found to be associated with type 2 diabetes in Japanese individuals (8). One of the SNPs (TG in codon 276) was also associated with lower plasma adiponectin levels, although this relationship was limited to obese subjects (8). Another SNP (TG at nucleotide 94: SNP 45) was found to be associated with obesity, insulin resistance, and dyslipidemia in a German population (9). A haplotype defined by these two SNPs has also been associated with components of the insulin resistance syndrome in Caucasians (10).
The ACDC gene spans 16 kb, contains three exons, and yields a 4.5-kb mRNA transcript. We screened 8 kb of ACDC corresponding to 2.8 kb of promoter sequence, each exon, and flanking intronic sequence and identified 17 SNPs. These included five promoter, seven exonic, and five intronic SNPs. At the time of the present study, five SNPs (SNP-12823, SNP-12128, SNP 3187, SNP 3267, and SNP 3286) were confirmed in a public database (available at http://www.ncbi.nlm.nih.gov/SNP) and correspond to dbSNP entries rs860291, rs266730, rs1063537, rs2082940, and rs1063538, respectively.
The genotypic distribution among 100 individuals partitioned into high- and low-plasma adiponectin level extremes (following adjustment for percent body fat) allowed us to make certain assumptions regarding the amount of disequilibrium between SNPs. When genotypes at SNPs were in >90% concordance among the 100 individuals, we grouped them into clusters and only genotyped one representative SNP from each cluster. On this basis, the 17 SNPs identified in Pimas were divided into 8 clusters, each containing 1eC4 SNPs (Fig. 1). Clusters 3, 7, and 8 consisted of SNPs with a minor allele frequency <15% and, as the goal of this study was to evaluate relatively common variants, these SNPs were not evaluated further. Cluster 5 also had a minor allele that was relatively rare; however, we typed SNP 45 from this cluster because previous studies showed association with diabetes or related traits.
We genotyped SNPs representative of each nonrare linkage disequilibrium cluster corresponding to SNP-12823, SNP-11365, SNP 45, SNP 276, and SNP 3286 in 1,338 participants from a genome-wide linkage study (11). We first evaluated linkage disequilibrium, as quantified by D', between pairs of these SNPs, as shown in Table 1. Each of the SNPs was in statistically significant linkage disequilibrium with each of the others, but the extent of association varied somewhat, with D' ranging from 0.48 to 0.99. Furthermore, because the associated alleles often had different frequencies, the information associated with each SNP varied considerably.
We assessed the association of each SNP with type 2 diabetes and BMI. In a subgroup of 331 nondiabetic individuals who had undergone detailed physiologic studies (12), we also assessed the association with insulin sensitivity, measured by the hyperinsulinemic-euglycemic clamp and insulin secretion, as assessed by the acute insulin response to an intravenous glucose infusion. In the larger group of nondiabetic individuals, the insulin sensitivity index (ISI) and corrected insulin response (CIR) were analyzed. These are surrogate indexes of insulin sensitivity and insulin secretion, respectively, that are correlated with the more sophisticated measures (13) and that can be derived from an oral glucose tolerance test. ISI was calculated from the fasting serum insulin (I0) and fasting plasma glucose (G0) concentrations (ISI = 104/[I0G0]), while CIR was calculated from the 2-h postload insulin (I2) and glucose (G2) concentrations (CIR = I2/G2[G2B70 mg/dl]) as described (14,15). The association of these SNPs with serum adiponectin levels, as well as levels of triglycerides and HDL choleseterol, was also assessed among nondiabetic individuals with normal renal function (16).
As shown in Table 2, none of the SNPs were associated with type 2 diabetes, BMI, ISI, CIR, or directly measured insulin sensitivity or secretion. There was a modest assocation of the G allele at SNP-11365 with lower HDL cholesterol levels. We also found that the G allele of SNP-12823 was significantly associated with lower adiponectin levels. Further adjustment for BMI did not substantially change these results. While statistically significant, the effect of this polymorphism accounted for only 2% of the variance of serum adiponectin levels.
We identified six common ACDC haplotypes, none of which were significantly associated with type 2 diabetes, BMI, ISI, CIR, lipid levels, or directly measured insulin sensitivity or secretion (Table 3). We did observe an association between the A_G_T_C_C haplotype and serum adiponectin levels, but because this haplotype is the only one containing the A allele of SNP-12823, it essentially reflects the relationship between this variant and serum adiponectin previously found in the individual SNP analyses.
Adiponectin has been recently identified as the most abundant of the known adipose tissue-derived proteins (1) with insulin-sensitizing and anti-inflammatory properties (1,17). Recently, the G allele of SNP 276 (GT) was found to be associated with low plasma adiponectin levels in obese Japanese individuals (8). Moreover, individuals homozygous for the haplotype consisting of the G allele at position 276 and the T allele at position 45 had significantly lower adiponectin levels than those who did not carry the haplotype (10). In the present study, no association between SNPs 276 and 45 (or their haplotypes) and protein levels was found in Pima Indians, although another variant, SNP-12823, which is located within the ACDC promoter, was associated with serum adiponectin levels. However, it is worth noting that this SNP can only explain 2% of the variance in serum adiponectin levels. Thus, the present results suggest that none of these SNPs in ACDC have major effects on serum adiponectin concentrations.
In genome-wide linkage analyses in the present group of individuals we found that while 39% of the variance in serum adiponectin is potentially due to genetic factors (i.e., heritability = 0.39), there was no evidence for linkage on chromosome 3q27, where ACDC is located (16). Instead, significant linkage (logarithm of odds = 3.0 at 18 cM) was seen on chromosome 9p and modest evidence for linkage (logarithm of odds = 1.0 at 124 cM) was found on chromosome 3q13, 70 cM centromeric to ACDC. Comuzzie et al. (18) have identified two major loci influencing adiponectin expression in European Americans on chromosomes 5 and 14 and modest evidence for linkage (logarithm of odds = 1.3) on 3q27 near ACDC.
Although some studies have reported suggestive evidence for the linkage of obesity with 3q27 (7), genome-wide linkage analyses of BMI in Pima Indians did not show strong evidence for linkage to the ACDC region (11). Furthermore, in the present study we found no evidence for association between ACDC SNPs and obesity in Pima Indians. These findings are consistent with results obtained in a Japanese population showing no evidence for a significant association with BMI (8). However, the role of ACDC SNPs in mediating susceptibility to obesity has been inconclusive. One study has reported an association between the variant G allele of SNP 45 and obesity in a German population (9), while results derived from other populations have found that the T allele at this position was associated with a higher body weight (10). Although the factors underlying the disparities among these studies are not yet known, it is possible that different environmental exposures, different patterns of linkage disequilibrium among populations, and sampling variation may play a role in the reported findings. It is also worth noting that the variant allele of SNP 45 is relatively rare and as such may produce spurious associations.
Low plasma adiponectin levels predict the development of type 2 diabetes in Pimas (5), and this has been confirmed in a German population (19). Linkage of the region near ACDC with type 2 diabetes has been reported in French and Japanese populations (6,20) and with fasting insulin levels in European Americans (7), thereby suggesting that a locus in this region may influence insulin sensitivity and risk for diabetes. In the Pimas, however, there is little evidence for linkage with either type 2 diabetes (11) or measures of insulin sensitivity (12) in this region. Because we have also not detected an association between ACDC SNPs and insulin sensitivity or type 2 diabetes in Pima Indians, it is likely that this gene does not play a major role in conferring increased susceptibility to these disorders in this population.
RESEARCH DESIGN AND METHODS
Subjects who participated in this study included 1,338 Pima individuals (332 nuclear families and 112 extended pedigrees) who were previously selected for linkage studies (11) from participants in ongoing longitudinal studies of obesity and type 2 diabetes conducted in the Gila River Indian Community. Among the siblings in the linkage study, there were 1,080 (496 men and 584 women) who were typed for at least one of the ACDC polymorphisms. The prevalence of type 2 diabetes in these participants was 59%. The mean (±SD) age was 38.7 ± 13.7 years, mean BMI was 36.6 ± 8.2 kg/m2, and mean serum adiponectin was 5.46 ± 2.69 e/ml. A subset of 331 of these individuals had participated in detailed physiologic studies, including measurement of directly assessed insulin sensitivity at physiologic insulin concentrations by the hyperinsulinemic-euglycemic clamp and measurement of insulin secretion by the acute insulin response calculated at 3eC5 min after a 25-g intravenous glucose bolus (see Pratley et al. [12] for further details). In 639 individuals, insulin and glucose concentrations had been measured during an oral glucose tolerance test at their last nondiabetic examination, and these measures were used to calculate surrogate measures of insulin sensitivity and secretion, ISI, and CIR (13eC15). Serum adiponectin was measured in stored sera from most of these participants (16,21). Since adiponectin concentrations are strongly influenced by the presence of diabetes and renal function (21), these analyses were restricted to nondiabetic participants with normal renal function (n = 580). In those who had been examined since 1992 (n = 212), serum HDL cholesterol and triglycerides were measured. This study was approved by the institutional review board of the National Institute of Diabetes and Digestive and Kidney Diseases and the Tribal Council of the Gila River Indian Community. All subjects provided signed informed consent before participation.
ACDC genomic screening and SNP genotyping.
ACDC was screened using denaturing high-performance liquid chromatography and amplicons, which yielded denaturing high-performance liquid chromatography heteroduplexes, which were validated by direct sequencing in the 50 individuals with the highest serum adiponectin levels and the 50 individuals with the lowest adiponectin levels as described (22).
We genotyped SNP-12823, SNP-11365, SNP 45, and SNP 276 using pyrosequencing according to the manufacturer’s recommendations (Pyrosequencing, Uppsala, Sweden) with minor modifications as described (23). SNP 3286 was genotyped by PCR restriction fragment-length polymorphism by first amplifying a PCR fragment of 397 bp and digesting it overnight at 371 with Nsp I (New England Biolabs, Beverly, MA). In the presence of the C allele, the PCR product was restricted into two fragments of 95 and 302 bp, while the T allele yielded an uncut fragment. Sequence information for all genotyping primers is available upon request.
Statistical analyses.
Haplotype frequencies for pairs of SNPs were estimated with a pedigree-based maximum likelihood method implemented in the ILINK program (24). The degree of linkage disequilibrium between alleles for each pair of loci was expressed as D' , which represents the proportion of the maximum possible allelic association given the allele frequencies and the direction of association. The likelihood ratio test was used to assess statistical significance of the observed allelic associations.
Association between alleles at each SNP and prevalence of diabetes was assessed by logistic regression with control for the effects of age, sex, birth year, and ethnicity (i.e., regardless of whether the participant was of full Pima heritage). The proportion of participants who were full Pima/Tohono O’odham was 79%; 94% of the remaining subjects were >50% Pima/Tohono O’odham. To account for familial relationships (i.e., sibship), these analyses were conducted using generalized estimating equations as previously described (23). The association of each SNP with type 2 diabetes was analyzed under an additive model in which the logarithm of the odds ratio is expressed as a function of the number of copies of the more common allele. Additional analyses were conducted under models in which the effect of the common allele was assumed to be dominant or recessive; as the conclusions of these models were identical to those of the additive models, only results for the additive models are presented here. For analysis of continuous variables (e.g., BMI, lipids, serum adiponectin concentrations, and measures of insulin sensitivity and secretion), linear regression analyses were conducted in an analogous fashion to the logistic models for type 2 diabetes. The generalized estimating equations were fit using PROC GENMOD of SAS (SAS Institute, Cary, NC).
A modification of the zero-recombinant haplotyping technique (25) was used to analyze the associations of haplotypes in the five genotyped SNPs in the region with the four different traits. First, the EH algorithm was used to estimate haplotype frequencies. There were six common haplotypes with frequencies >0.05, which accounted for 98% of the observed haplotypes among the five SNPs. Next, the MLINK program (23) was used to estimate the probability that each individual carried each possible haplotype given these frequencies, the genotypes of the individual, and the genotypes of other individuals in the pedigree. As only four SNPs were necessary to assign individuals to these haplotypes, SNP 3286 was omitted from the likelihood calculations. The probabilities derived from MLINK were then utilized in logistic or linear regression analyses to evaluate the association of each of the common haplotypes with the traits of interest. Since this approach requires the assumption of no recombination across the region, individuals whose genotypes produced obligate recombinations (n = 9) were deleted from these analyses. These analyses were conducted with six different common haplotypes; therefore, a corrected P value (Pcorr) was calculated using a Bonferroni correction [Pcorr = 1 eC (1eC P)] (5) to account for multiple comparisons.
ACKNOWLEDGMENTS
We thank Victoria M. Ossowski and Timothy Monette for technical assistance and Dr. Jonathan Krakoff for his advice. We are also grateful for the support and participation of the Gila River Indian Community.
CIR, corrected insulin response; ISI, insulin sensitivity index; SNP, single nucleotide polymorphism
REFERENCES
Scherer PE, Williams S, Fogliano M, Baldini G, Lodish HF: A novel serum protein similar to C1q, produced exclusively in adipocytes. J Biol Chem270 :26746 eC26749,1995
Weyer C, Funahashi T, Tanaka S, Hotta K, Matsuzawa Y, Pratley RE, Tataranni PA: Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia. J Clin Endocrinol Metab86 :1930 eC1935,2001
Arita Y, Kihara S, Ouchi N, Takahashi M, Maeda K, Miyagawa J, Hotta K, Shimomura I, Nakamura T, Miyaoka K, Kuriyama H, Nishida M, Yamashita S, Okubo K, Matsubara K, Muraguchi M, Ohmoto Y, Funahashi T, Matsuzawa Y: Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity. Biochem Biophys Res Commun257 :79 eC83,1999
Stefan N, Vozarova B, Funahashi T, Matsuzawa Y, Weyer C, Lindsay RS, Youngren JF, Havel PJ, Pratley RE, Bogardus C, Tataranni PA: Plasma adiponectin concentration is associated with skeletal muscle insulin receptor tyrosine phosphorylation, and low plasma concentration precedes a decrease in whole-body insulin sensitivity in humans. Diabetes51 :1884 eC1888,2002
Lindsay RS, Funahashi T, Hanson RL, Matsuzawa Y, Tanaka S, Tataranni PA, Knowler WC, Krakoff J: Adiponectin and development of type 2 diabetes in the Pima Indian population. Lancet360 :57 eC58,2002
Vionnet N, Hani El H, Dupont S, Gallina S, Francke S, Dotte S, De Matos F, Durand E, Lepretre F, Lecoeur C, Gallina P, Zekiri L, Dina C, Froguel P: Genomewide search for type 2 diabetes-susceptibility genes in French whites: evidence for a novel susceptibility locus for early onset diabetes on chromosome 3q27-qter and independent replication of a type 2-diabetes locus on chromosome 1q21eCq24. Am J Hum Genet67 :1470 eC1480,2000
Kissebah AH, Sonnenberg GE, Myklebust J, Goldstein M, Broman K, James RG, Marks JA, Krakower GR, Jacob HJ, Weber J, Martin L, Blangero J, Comuzzie AG: Quantitative trait loci on chromosomes 3 and 17 influence phenotypes of the metabolic syndrome. Proc Natl Acad Sci U S A97 :14478 eC14483,2000
Hara K, Boutin P, Mori Y, Tobe K, Dina C, Yasuda K, Yamauchi T, Otabe S, Okada T, Eto K, Kadowaki H, Hagura R, Akanuma Y, Yazaki Y, Nagai R, Taniyama M, Matsubara K, Yoda M, Nakano Y, Tomita M, Kimura S, Ito C, Froguel P, Kadowaki T: Genetic variation in the gene encoding adiponectin is associated with an increased risk of type 2 diabetes in the Japanese population. Diabetes51 :536 eC540,2002
Stumvoll M, Tschritter O, Fritsche A, Staiger H, Renn W, Weisser M, Machicao F, Haring H: Association of the T-G polymorphism in adiponectin (exon 2) with obesity and insulin sensitivity: interaction with family history of type 2 diabetes. Diabetes51 :37 eC41,2002
Menzaghi C, Ercolino T, Di Paola R, Berg AH, Warram JH, Scherer PE, Trischitta V, Doria A: A haplotype at the adiponectin locus is associated with obesity and other features of the insulin resistance syndrome. Diabetes51 :2306 eC2312,2002
Hanson RL, Ehm MG, Pettitt DJ, Prochazka M, Thompson DB, Timberlake D, Foroud T, Kobes S, Baier L, Burns DK, Almasy L, Blangero J, Garvey WT, Bennett PH, Knowler WC: An autosomal genomic scan for loci linked to type II diabetes mellitus and body-mass index in Pima Indians. Am J Hum Genet63 :1130 eC1138,1998
Pratley RE, Thompson DB, Prochazka M, Baier L, Mott D, Ravussin E, Sakul H, Ehm MG, Burns DK, Foroud T, Garvey WT, Hanson RL, Knowler WC, Bennett PH, Bogardus C: An autosomal genomic scan for loci linked to prediabetic phenotypes in Pima Indians. J Clin Invest101 :1757 eC1764,1998
Hanson RL, Pratley RE, Bogardus C, Narayan KM, Roumain JM, Imperatore G, Fagot-Campagna A, Pettitt DJ, Bennett PH, Knowler WC: Evaluation of simple indices of insulin sensitivity and insulin secretion for use in epidemiologic studies. Am J Epidemiol151 :190 eC198,2000
Sluiter WJ, Erkelens DW, Terpstra P, Reitsma WD, Doorenbos H: Glucose tolerance and insulin release, a mathematical approach. II. Approximation of the peripheral insulin resistance after oral glucose loading. Diabetes25 :245 eC249,1976
Sluiter WJ, Erkelens Dw, Reitsma WC, Doorenbos H: Glucose tolerance and insulin release, a mathematical approach. I. Assay of the beta-cell response after oral glucose loading. Diabetes25 :241 eC244,1976
Lindsay RS, Funahashi T, Krakoff J, Matsuzawa Y, Tanaka S, Kobes S, Bennett PH, Tataranni PA, Knowler WC, Hanson RL: Genome-wide linkage analysis of serum adiponectin in the Pima Indian population. Diabetes52 :2419 eC2425,2003
Berg AH, Combs TP, Du X, Brownlee M, Scherer PE: The adipocyte-secreted protein Acrp30 enhances hepatic insulin action. Nat Med7 :947 eC953,2001
Comuzzie AG, Funahashi T, Sonnenberg G, Martin LJ, Jacob HJ, Black AE, Maas D, Takahashi M, Kihara S, Tanaka S, Matsuzawa Y, Blangero J, Cohen D, Kissebah A: The genetic basis of plasma variation in adiponectin: a global endophenotype for obesity and the metabolic syndrome. J Clin Endocrinol Metab86 :4321 eC4325,2001
Spranger J, Kroke A, Mohlig M, Bergmann MM, Ristow M, Boeing H, Pfeiffer AF: Adiponectin and protection against type 2 diabetes mellitus. Lancet361 :226 eC228,2003
Mori Y, Otabe S, Dina C, Yasuda K, Populaire C, Lecoeur C, Vatin V, Durand E, Hara K, Okada T, Tobe K, Boutin P, Kadowaki T, Froguel P: Genome-wide search for type 2 diabetes in Japanese affected sib-pairs confirms susceptibility genes on 3q, 15q, and 20q and identifies two new candidate Loci on 7p and 11p. Diabetes51 :1247 eC1255,2002
Looker HC, Krakoff J, Funahashi T, Matsuzawa Y, Tanaka S, Nelson RG, Knowler WC, Lindsay RS, Hanson RL: Adiponectin concentrations are influenced by renal function and duration of diabetes in Pima Indians with type 2 diabetes. J Clin Endocrinol Metab89 :4010 eC4017,2004
Wolford JK, Blunt D, Ballecer C, Prochazka M: High-throughput SNP detection by using DNA pooling and denaturing high performance liquid chromatography (DHPLC). Hum Genet107 :483 eC487,2000
Wolford JK, Gruber JD, Ossowski VM, Vozarova B, Antonio Tataranni P, Bogardus C, Hanson RL: A C-reactive protein promoter polymorphism is associated with type 2 diabetes mellitus in Pima Indians. Mol Genet Metab78 :136 eC144,2003
Lathrop GM, Lalouel JM: Easy calculations of lod scores and genetic risks on small computers. Am J Hum Genet36 :460 eC465,1984
O’Connell JR: Zero-recombinant haplotyping: applications to fine mapping using SNPs. Genet Epidemiol19 (Suppl. 1) :S64 eCS70,2000(Barbora Vozarova de Court)