Peroxisome proliferator–activated receptor coactivator-1α (PGC-1α) is a transcriptional coactivator implicated in transcriptional programs of hepatic gluconeogenesis, oxidative phosphorylation, and insulin release by β-cells. To study associations of the PGC-1α gene locus with carbohydrate metabolism and type 2 diabetes in humans, we identified several polymorphisms in the promoter region that were located in a haplotype block distinct from a second haplotype block containing part of intron 2 and extending beyond exon 13. Each block contained five common haplotypes. Oral glucose tolerance testing revealed associations of promoter haplotype combinations with 30- and 60-min postload plasma glucose levels, whereas haplotypes in both blocks were associated with indexes of β-cell function. The associations of promoter haplotypes are supported by functional studies showing that some polymorphisms are located in transcription factor binding sites and affect transactivation in an allele-specific manner. By comparing patients with type 2 diabetes and control subjects, we observed borderline significant differences of four-loci haplotype distributions in the downstream haplotype block. Moreover, the haplotype that was associated with the strongest insulin response to glucose conferred the lowest risk of type 2 diabetes (P < 0.01). Thus, the PGC-1α gene locus influences carbohydrate metabolism and contributes to type 2 diabetes in the population studied.

Insulin resistance, impaired compensatory upregulation of insulin secretion, and enhanced hepatic gluconeogenesis are the hallmarks of type 2 diabetes (1). Peroxisome proliferator–activated receptor coactivator-1α (PGC-1α) may play a pathogenic role in each component of this triad. PGC-1α interacts with hepatocyte nuclear factor (HNF)-4α, the glucocorticoid receptor, and the forkhead transcription factor FOXO 1 to execute transcriptional programs of insulin-regulated gluconeogenesis (24). In addition, PGC-1α enhances glucose uptake in muscle cells by the induction of GLUT4 expression via coactivation of the MADS box transcription enhancer factor (MEF) 2C (5). MEF2 family members, which are released from class II histone deacetylases in response to calcium signaling, transactivate the PGC-1α promoter (6,7). This autoregulatory mechanism may result in the stable induction of PGC-1α, which has been implicated in exercise-induced mitochondrial biogenesis and muscle fiber–type switching (8). Finally, elevated levels of PGC-1α in pancreatic β-cells have been shown to blunt the glucose-induced increase in cellular ATP levels and insulin exocytosis (9).

Studies in humans also support a role of PGC-1α in the pathogenesis of type 2 diabetes. Expression of PGC-1α and its target genes involved in oxidative phosphorylation were downregulated in muscle tissue of type 2 diabetic patients in comparison to control subjects (10). Moreover, skeletal muscle expression of PGC-1α and PGC-1β, another PGC-1 family member (11), was reduced in both diabetic subjects and family history–positive nondiabetic subjects compared with control subjects (12). Hence, genetic influences that affect PGC-1α and -1β expression in muscle may contribute to the disorder. Indeed, single nucleotide polymorphisms (SNPs) in the transcribed sequence of PGC-1α have been associated with type 2 diabetes in Danish, British, and Japanese but not in French populations (1316). To gain further insight into the role of PGC-1α in humans, we identified and characterized sequence substitutions across its gene locus, determined their haplotype arrangement, and studied their association with carbohydrate metabolism and type 2 diabetes.

For haplotype characterization, we studied 40- to 65-year-old male and 45- to 70-year-old female participants of the Salzburg Atherosclerosis Prevention Program in Subjects at High Individual Risk (SAPHIR). The overall study aim and recruitment procedures have been described (17). Associations of genotypes, haplotypes, and haplotype combinations with estimates of glucose metabolism were determined in a subset of 405 SAPHIR participants, who had fasting plasma glucose levels <110 mg/dl and underwent an oral glucose tolerance test (OGTT). After a 12-h overnight fast, subjects ingested a solution containing 75 g glucose, and plasma samples were obtained at 0, 30, 60, and 120 min for determination of glucose and insulin. Insulin sensitivity/resistance was estimated using the homeostasis model assessment (HOMA) insulin resistance index (HOMAIR) (18), the composite insulin sensitivity index [Si(comp)] (19), and the metabolic clearance rate and Si proposed by Stumvoll et al. (20). For metabolic clearance rate and Si calculations, average values of 60- and 120-min rather than 90-min glucose values were used. β-Cell function was assessed by the HOMA-based insulin release index (HOMASecr), first-phase and second-phase insulin release (20), the CIR30 (corrected insulin response at 30 min), the 30- to 0-min insulin ratio (I30/I0), the 30-min insulin-to-glucose ratio (I30/G30), and the insulinogenic index estimated as I30I0/G30 − G0I30G30) (21). A total of 506 unrelated patients with type 2 diabetes were recruited from diabetes outpatient clinics of the Landeskliniken Salzburg and the Hospital Hallein. Patients who were <73 years of age and <63 years of age at the time of diabetes diagnosis were included. SAPHIR participants who were not using hypoglycemic medications and had fasting blood glucose levels <110 mg/dl served as control subjects. Study populations comprised only white Europeans, mainly of Bavarian or Austrian-German descent living in the same geographic region. Participants provided informed consent, and the study was approved by the local ethics committee.

Laboratory analyses.

Plasma glucose, insulin, cholesterol, triglyceride, HDL cholesterol, and HbA1c levels were measured as described (22). Genomic DNA was extracted from peripheral white blood cells.

Identification of SNPs, genotyping, and haplotyping procedures.

SNPs in the promoter region of the human PGC-1α gene were identified by amplification of a 5,551-bp fragment from genomic DNA of six human subjects using 5′-CGCGGTACCCAATCCTAGAAGCGGTCGTCTTGG-3′ and 5′-CGCACGCGTTCCAGCTCCTGAATGACGCCAG-3′ as forward and reverse primers, respectively (NCBI genome annotation no. NT_006316). Restriction enzyme sites for MluI and KpnI, respectively, are in bold. The resulting PCR products were ligated into the pGL3 basic vector (Promega, Madison, WI). Individual clones and PCR products were sequenced from both directions by dye-terminator sequencing using an ABI Prism 310 genetic analyzer (Applied Biosystems, Foster City, CA). To delineate the phase of polymorphisms in 17 additional subjects, regions harboring SNPs were sequenced in PCR products and in one clone representing the sequence from one chromosome. Sequence differences between PCR products and clones thus provided the sequence of the second chromosome.

Four SNPs that unambiguously predicted phased haplotypes were typed in our populations. Typing procedures included restriction fragment–length polymorphism (RFLP) detection with or without artificially introduced restriction sites, using mismatch primers as well as real-time PCR. For the latter method, allele-specific hybridization probes conjugated with a fluorophore at the 5′-end were used with minor groove binder ligand and quencher moiety at the 3′-end (Applied Biosystems). For quality control, 30% of real-time PCR-based typings for SNP −1422 T/C were confirmed using an RFLP assay. Assay details are shown in Table 1 in the online appendix (available at http://diabetes.diabetesjournals.org).

To search for additional sequence variants in the transcribed region, we sequenced 12 alleles from the transcriptional start site to 300 bp beyond the more-downstream-located polyadenylation signal, but excluding the large intron 2. The +1302G/A, +1564G/A, and 1704A/G polymorphisms in exon 8 and the +2962A/G polymorphism in the 3′-untranslated region of the PGC-1α gene (Genbank accession no. NM_013261) were typed in our populations as described (22,23). Haplotypes were estimated from unphased genotypes using the expectation-maximization algorithm of Terwilliger and Ott (24). Four-loci genotype determinations in both the promoter and the transcribed region were available in 1,608 SAPHIR participants. In the 405 glucose-tolerant SAPHIR participants who underwent an OGGT, four-loci genotypes in the promoter or transcribed region were available in 361 or 398 subjects, respectively. In the 506 patients with type 2 diabetes and in the nondiabetic 1,480 SAPHIR participants with four-loci promoter genotype determinations, four-loci genotypes of the transcribed region were available in 494 and 1,478 subjects, respectively.

To identify haplotype boundaries, regions in intron 2 as well as upstream and downstream regions of PGC-1α that contained several SNPs (www.ncbi.nlm.nih.gov/SNP/) were sequenced in eight subjects to identify polymorphisms with high information content. Four SNPs at −22174, +16022, +55180, and +114481 relative to the translational start site were typed in >84 randomly selected SAPHIR participants.

Plasmid constructs.

PGC-1α promoter reporter plasmids encompassing SNPs at positions −1984, −1774, −1679 and −1422 were constructed by amplification of a 732-bp fragment from genomic DNA of patients with known haplotypes. The reporter constructs −3959 TK-Luc, −3818 TK-Luc, −3614 TK-Luc, and −1422 TK-Luc were generated by amplifying ∼250- to 350-bp fragments encompassing the respective sites using genomic DNA from heterozygous subjects. Primers, PCR product fragment lengths, and restriction enzymes used for cloning are listed in Table 2 in the online appendix. Fragments were cloned into a reporter vector harboring the firefly luciferase gene under the control of the TK promoter (25,26).

Primer sequences used to clone full-length HNF-1α into pcDNA6/V5-His (Invitrogen, Carlsbad, CA) were based on GenBank NM_012120. A 728-bp fragment was amplified from human liver cDNA using 5′-GAGAGCGGCCGCCAGGCCTCAACCAGCAGCCC-3′ and 5′-GAGAGCGGCCGCCCAGGTGCCGTGGTTACTGGGAG-3′ (stop codon underlined) as forward and reverse primers, respectively. For construction of a full-length MEF2C expression plasmid (GenBank accession no. NM_002397), we amplified a 1,438-bp fragment from human muscle cDNA using 5′-GAGAGCGGCCGCCATAATTTCAGGGACGAGAGAGAGAAG-3′ and 5′-GAGAGCGGCCGCTCATGTAGCCCATCCTTCAGAAAGTCG-3′ (stop codon underlined) as forward and reverse primers, respectively. NotI restriction enzyme sites introduced are in bold. All plasmid constructs were verified by sequencing.

Cell culture and transfection.

INS-1E cells, kindly provided by Claes Wollheim (Geneva, Switzerland); PAZ6 cells; and HepG2 cells were cultured as described (17,22). Near-confluent cells were transfected with constant amounts of reporter constructs and 500 ng/well of HNF-1α or MEF2C expression plasmids or their empty pcDNA6/V5-His plasmid backbone using Lipofectamine 2000 reagent. pRL-CMV (1 ng/well; Promega) was cotransfected as transfection control. Transfection medium was replaced with fresh culture medium 12 h later. Cells were harvested 24 h after transfection, and firefly and Renilla luciferase activities were measured in cell extracts using a Lucy2 luminometer (Rosys Anthos, Salzburg, Austria) and the Dual-Luciferase reporter assay system (Promega).

Computational analyses and statistics.

For prediction of putative transcription factor binding sites, we used MatInspector professional software (www.genomatix.de), which locates individual matches by comparing the sequences with weight matrix descriptions of binding sites (27). The matrix library is based on Transfac release 3.3 (28). We considered matrix and core similarities >80 and >85%, respectively.

Group differences of continuous variables were ascertained by ANOVA. We made logarithmic transformations if the equal variance and normality assumptions of ANOVA were rejected. Measurements were adjusted for the effects of age and BMI as indicated. Two-way ANOVA was used to compare reporter-gene activities from variant and wild-type alleles as well as transactivation from PGC-1α promoter alleles by MEF2C and HNF-1α. For post hoc comparisons of means, Tukey’s honest significant difference test was used.

Allele frequencies were estimated by gene counting. Agreement with Hardy-Weinberg expectations was tested using a χ2 goodness-of-fit test. The standardized pairwise linkage disequilibrium statistics (D′) were estimated (24). For effects of individual SNPs on OGGT parameters, we used ANOVA with adjustments for age, sex, and BMI. In addition, multivariate regression models with glucose and insulin levels or derived indexes as outcome variables and genotypes, age, sex, and BMI as independent variables were used. Genotypes were given codes of 0, 1, and 2; 0, 1, and 1; or 0, 0, and 1 in the additive, dominant, or recessive models, respectively. Differences in genotype or allele frequencies between patients with type 2 diabetes and control subjects were determined using a χ2 distribution with 2 or 1 degrees of freedom. Differences in four-loci haplotype distributions between patients and control subjects were ascertained by the log-likelihood ratio test with 15 degrees of freedom (29). For testing of associations between haplotypes and quantitative or binary traits, we used haplo.score software (http://www.mayo.edu/statgen/). This method is based on efficient score statistics, which provides both global and haplotype-specific tests and allows the exclusion of rare haplotypes as well as adjustments for nongenetic covariates (30).

We identified eight polymorphic sites between −1422 and −3959 (Table 3 in the online appendix and Fig. 1) by sequencing 12 promoter alleles up to −5551 relative to the translational start site. Of the 28 (= 256) possible combinations of these SNPs, only five haplotypes were observed in the 46 alleles in which phased genotypes were determined (Fig. 1B). Thus, SNPs at −3959, −3690, −1984, and −1774 as well as SNPs at −3614 and −1679 showed complete linkage disequilibrium. Four SNPs (at −3959, −3818, −1679, and −1422) that permitted unambiguous assignments of phased haplotypes were typed in 1,608 SAPHIR participants. All polymorphisms fulfilled Hardy-Weinberg expectations. Variant allele frequencies of promoter polymorphisms are shown in Table 3 in the online appendix. Standardized pairwise linkage disequilibrium statistics (D′) were highly significant for the promoter polymorphisms (Table 1), and the five haplotypes delineated by phased genotyping accounted for 97% of all haplotypes defined by the respective sites (Fig. 1B).

We also ascertained haplotype structures across the transcribed region of PGC-1α. Our sequencing studies confirmed previously identified polymorphic sites in introns 1, 2, 4, and 5 as well as in exons 8, 9, and 13, but they failed to detect additional sequence variants. Frequencies of the variant alleles are also shown in Table 3 in the online appendix. Typing of four informative polymorphisms in exons 8 (at +1302, +1564, +1704) and exon 13 (at +2962) revealed strong linkage disequilibria (Table 1), and estimation of haplotype frequencies identified five common haplotypes (Fig. 1C) that accounted for 98% of haplotypes comprising the respective sites. However, at most, only weak linkage disequilibria were observed with the variant promoter sites (Table 1). Hence, two haplotype blocks, termed haplotype block 1 and 2, each characterized by five common haplotypes, were identified within PGC-1α. Based on additional typing studies of SNPs in introns 2 and 5, the boundary between the blocks was located in intron 2 between SNPs at +16022 and +55180, relative to the translational start site. Haplotype block 1 extended to ∼22 kb upstream of the translational start site, whereas haplotype block 2 extended <20 kb beyond the proximal polyA signal (Fig. 1A and Table 1).

Among the promoter polymorphisms typed, two sites displayed significant associations with oral glucose tolerance test parameters in glucose-tolerant subjects. Clinical characteristics of the study population are shown in Table 4 in the online appendix. The SNP at −3959 SNP was associated with differences in 30-min postload insulin levels (Fig. 2) and parameters reflecting insulin release, such as first-phase, second-phase, and 30-min insulin-to-glucose ratio (P < 0.02). The SNP at −1679 was associated with differences in 30- and 60-min postload glucose levels (Fig. 3) and the integrated area of glucose (P < 0.015, ANOVA). Linear regression models were most consistent with additive and variant-allele recessive effects at the −3959 and the 1679 sites, respectively. Neither polymorphism was associated with measures of insulin sensitivity such as HOMAIR, Si(comp), metabolic clearance rate, or Si (data not shown). Haplotype-specific analyses showed a consistent pattern of associations with parameters reflecting the insulin secretory response to glucose (Table 2), but not with 30- or 60-min glucose levels and integrated glucose areas. However, common four-loci promoter haplotype combinations showed associations with 30- and 60-min postload glucose levels and integrated postload glucose areas (Table 5 in the online appendix), a result compatible with the apparently recessive effect of the −1679 variant allele. None of the four SNPs in haplotype block 2 displayed an association with any of the OGTT parameters, but global and haplotype-specific statistics showed significant associations with calculated parameters of insulin response to glucose (Table 2). Haplotype block 2 combinations showed similar associations (data not shown).

Computational analyses of the 5′-flanking region of PGC-1α suggested that some polymorphic sites are located in cis regulatory elements that may affect interactions with various trans factors in an allele-specific manner, thereby modulating PGC-1α expression levels. Comparison of transfections with the −3959T TK-Luc and −3959C TK-Luc reporter constructs in INS1-E cells revealed lower reporter gene activity of plasmids containing the variant nucleotide (Fig. 2C), but no differences were noted in HepG2 and PAZ6 cells (data not shown). The inverse relationship of PGC-1α gene expression with indexes of insulin release in our study subjects is consistent with the proposed suppression of insulin release by PGC-1α (9). Transcription factors predicted to bind to the sequence harboring the polymorphic site at −3,959 are shown in Fig. 2D. We did not observe an effect of nucleotide substitutions at position −3818 on PGC-1α gene expression in HepG2, INS1-E, or PAZ6 cells (data not shown).

The polymorphic site at −1679 was associated with postload glucose levels and had little effect in transient expression studies (Fig. 1 in the online appendix), but it was in perfect linkage disequilibrium with the SNP at −3614, which is located within a potential HNF-1α binding site (Fig. 3). Because HNF-1α has been implicated in the regulation of hepatic gluconeogenesis (31), we performed transient transfection studies using the −3614A TK-Luc and −3614C TK-Luc reporter constructs as well as cotransfections with the HNF-1α expression plasmid in HepG2 cells. PGC-1α expression from the variant −3614C allele was significantly decreased in both the absence and presence of HNF-1α overexpression (Fig. 3C). Thus, the reduced promoter activity of the −3614 variant allele paralleled the lower 30- and 60-min postload glucose levels in subjects homozygous for the −1679 variant allele (Fig. 3A). Reporter gene activities of the variant and wild-type alleles were similar in PAZ6 or INS1-E cells (data not shown). Because overexpression of HNF-1α in HepG2 cells resulted in an approximately threefold induction of luciferase activity from both the −3614A TK-Luc and −3614C TK-Luc constructs (Fig. 3C), HNF-1α may participate in the regulation of PGC-1α expression.

The SNP at −1422T is located within a putative MEF2C binding site previously identified in the murine PGC-1α promoter (6,7). The SNP at −1774 is contained within another potential MEF2C binding site not conserved in the murine promoter. Contransfection studies with or without MEF2C expression plasmids revealed that constructs harboring the variant nucleotide at position −1422 increased reporter gene activity in HepG2 cells, in comparison to constructs harboring the wild-type allele (Fig. 1 in the online appendix).

To ascertain associations of the PGC-1α gene locus with type 2 diabetes, we determined genotype distributions in type 2 diabetic patients and control subjects. Clinical data of patients and control subjects are shown in Table 6 in the online appendix. None of the eight polymorphic sites displayed associations with type 2 diabetes in sex-specific or combined analyses in our population (data not shown). Four-loci haplotype distributions of haplotype block 2, but not haplotype block 1, showed borderline significant differences (Table 3). Score testing revealed that the common haplotype in block 2 that carries wild-type nucleotides at all four polymorphic sites was associated with the lowest risk of type 2 diabetes (P < 0.01). The same haplotype was associated with the strongest insulin secretory response to glucose (Table 2). Conversely, the common block 2 haplotype that contained variant nucleotides at the three distal polymorphic sites, and displayed the lowest scores for insulin secretory indexes, exhibited the highest score for type 2 diabetes among common haplotypes. Thus, effects of PGC-1α on the insulin response to glucose are likely to contribute to its association with type 2 diabetes.

Previous studies in animal models and humans strongly support a role of PGC-1α in the pathogenesis of type 2 diabetes. In this work, we have delineated SNPs within contiguous regions of the PGC-1α gene promoter and the transcribed sequence. The SNPs in the promoter region comprise a haplotype block that is distinct from a second haplotype block harboring SNPs downstream of intron 2. Each block contains five common haplotypes. Other studies that applied high marker densities over contiguous genome regions also demonstrated blocks of variable length. Such blocks typically contain only few haplotypes separated by sites at which recombinations occurred in the ancestors of current populations (32,33). In glucose-tolerant subjects, haplotype block 1 showed discrete associations with postload glucose levels, whereas both haplotype blocks were associated with indexes of the insulin secretory response to glucose. Moreover, a common specific haplotype in block 2 revealed associations with β-cell function and type 2 diabetes risk.

Our transient transfection studies in human cell lines identified several promoter SNPs that modulate promoter function, but it is likely that other functional sites in haplotype block 1 might exist that affect carbohydrate metabolism and/or type 2 diabetes risk. Nucleotides at −1422 and −1774 are located within MEF2 binding sites. Even though both sites are required for full transactivation, only the −1422 site displayed allele-specific promoter effects, but by itself it showed no associations with OGGT parameters or type 2 diabetes. Because MEF2 has been implicated in skeletal muscle GLUT4 expression (5,34), and transcriptional responses of the PGC-1α promoter to MEF2 may play an important role in programs controlling muscle fiber-type conversion (7,8) and cardiac energy production (6), the −1422 site may contribute to the pathogenesis of other disease phenotypes or to type 2 diabetes in other populations.

The −1679 polymorphic site as well as common haplotype combinations reflecting the integrated effects of promoter polymorphisms were associated with differences in 30- and 60-min postload glucose levels, and functionality of the perfectly linked polymorphic site at −3614 was established by transient transfection studies in HepG2 cells. Whether the association of the −3614/−1679 SNPs with decreased postload glucose levels reflects changes in glucose uptake or glucose production is not known, but both possibilities appear plausible. PGC-1α has been shown to participate in insulin-regulated hepatic gluconeogenesis by coactivating HNF-4α and FOXO1 (24). Moreover, hepatic gluconeogenesis contributes to glucose effectiveness ascertained in human studies (35,36). Conversely, PGC-1α enhances oxidative phosphorylation and glucose uptake in skeletal muscle (5,10,12). Hence, the variant allele or another functional site in linkage disequilibrium with it may have tissue-specific effects and enhance PGC-1α expression in skeletal muscle. Our cotransfection studies further suggest that HNF-1α contributes to the regulation of PGC1-α promoter activity. HNF-1α and HNF-4α, which is coactivated by PGC-1α, have been shown to activate the expression of each other (37). Mice lacking HNF-4α in liver exhibit higher PGC-1α mRNA levels than control mice (38). It is thus possible that HNF-1α is part of a regulatory loop that controls PGC-1α expression in liver.

The effects of the −3959 variant site were tissue-specific, since the less common allele reduced reporter gene activity in INS1-E cells but not in HepG2 or PAZ6 cells. An inverse association between PGC-1α expression and insulin secretion is supported by recent studies (9). The influence of this SNP was also apparent in a four-loci haplotype context because the haplotype harboring the −3959 variant nucleotide revealed the highest scores for insulin secretion indexes.

Haplotypes of block 2 revealed strong associations with derived estimates of the insulin secretory response to glucose. Associations of +1564G/A with lipid oxidation and early insulin secretion have been reported in Pima Indians (39). Even though we did not observe such an association, the results in Pima Indians are not inconsistent with our findings because haplotype block size, allele frequencies, and linkage disequilibrium relationships differ among populations (31). The absence of associations with individual polymorphisms in our population suggests that the block 2 SNPs studied are not major functional sites. Such site(s) may be located in introns and affect mRNA expression or splicing. Indeed, a PGC-1α isoform devoid of exon 8 and induced by exercise has been observed in rat skeletal muscle (40). Using RNase protection assays, we have recently observed distinct PGC-1α isoforms lacking specific interfaces in various human tissues (M. Xie, H.O., M. Roden, W.P., unpublished observations), but their pathophysiological significance awaits further analyses.

Given the role of PGC-1α in mitochondrial biogenesis and oxidative phosphorylation (10,12), the lack of associations with indexes of insulin resistance in either haplo-type block was surprising. Imprecision in the ascertainment of insulin resistance may have contributed (41). Furthermore, combined effects of both haplotype blocks could not be determined because of sample size constraints. Interestingly, muscle-specific overexpression of PGC-1α in transgenic mice was shown to reduce GLUT4 expression (42).

The borderline-significant different haplotype distributions in block 2 between patients with type 2 diabetes and control subjects suggest, but do not prove, that the PGC-1α gene locus contributes to the susceptibility for type 2 diabetes. Importantly, the haplotype that displayed the lowest score for type 2 diabetes was associated with the highest scores for indexes of β-cell function, thus providing physiological plausibility and internal consistency of results. Associations of the +1564 site with type 2 diabetes have been reported in a Danish population (13), but these results are at variance with studies in French and Japanese populations as well as with our findings. Nevertheless, two-loci haplotypes comprising the +1564 SNP displayed associations with diabetes in Japanese subjects. Among the three European populations, variant allele frequencies in subjects with type 2 diabetes were similar and ranged from 0.363 to 0.373, but they were slightly higher in French and Austrian than in Danish nondiabetic subjects (0.343 and 0.348 vs. 0.306). However, in all four populations, variant allele frequencies were higher in patients than in control subjects. It is therefore conceivable that the +1564 polymorphic site is not the main causal site, but is contained within haplotype block 2, that contributes to type 2 diabetes in several populations.

FIG. 1.

Polymorphic sites and haplotype blocks in PGC-1α. A: Linear map with arrows indicating translational initiation (ATG) and termination (TAA) sites as well as polyadenylation signals (AATAAA). ▪, exons; sqsw;, the 3′-untranslated region. Scales differ between the promoter and the transcribed regions. SNP identifiers are relative to the translational start site in the genomic sequence. SNPs at +75656, +75918, +76058, and +94580 correspond to positions +1302, +1564, +1704, and +2962 relative to the translational start site in the mRNA sequence. Haplotype blocks 1 and 2 with their boundaries are shown beneath the SNPs. B: Structures and frequencies of the five common haplotypes in haplotype block 1. C: Structures and frequencies of the five common haplotypes in haplotype block 2. SNPs typed in 1,608 subjects and used for calculations of frequency distributions are in bold.

FIG. 1.

Polymorphic sites and haplotype blocks in PGC-1α. A: Linear map with arrows indicating translational initiation (ATG) and termination (TAA) sites as well as polyadenylation signals (AATAAA). ▪, exons; sqsw;, the 3′-untranslated region. Scales differ between the promoter and the transcribed regions. SNP identifiers are relative to the translational start site in the genomic sequence. SNPs at +75656, +75918, +76058, and +94580 correspond to positions +1302, +1564, +1704, and +2962 relative to the translational start site in the mRNA sequence. Haplotype blocks 1 and 2 with their boundaries are shown beneath the SNPs. B: Structures and frequencies of the five common haplotypes in haplotype block 1. C: Structures and frequencies of the five common haplotypes in haplotype block 2. SNPs typed in 1,608 subjects and used for calculations of frequency distributions are in bold.

Close modal
FIG. 2.

Associations of −3959 T/C genotypes with OGTT glucose and insulin levels and effects on PGC-1α gene expression in INS1-E cells. Time-dependent values for plasma glucose (A) and insulin (B) in response to 75 g glucose in glucose-tolerant patients differing by −3959 T/C genotypes. Error bars indicate SE. For T/T, T/C, and C/C subjects, n was 238, 110, and 13, respectively. *P < 0.02, ANOVA, adjusted for age, sex, and BMI. C: INS1-E cells were transiently transfected with –3959T TK-Luc and –3959C TK-Luc. Results are means + SD and representative for one of two experiments, each performed in quadruplicate. **P < 0.001. D: Transcription factors and their binding matrices in relation to the PGC-1α promoter region encompassing the –959 T/C polymorphism. Variable nucleotides in the PGC-1α sequence and respective nucleotides in binding sequences of candidate transcription factors are in bold. Numbers refer to nucleotide positions relative to the transcriptional start site. The core sequences of binding sites are in upper case letters. Nucleotides underlined exhibit a high percentage of conservation. Binding matrices allowing for more than one nucleotide at a position are given in International Union of Pure and Applied Chemistry (IUPAC) standard codes for nucleic acids: w = A,T; r = A/G; s = C/G, and m = A,C.

FIG. 2.

Associations of −3959 T/C genotypes with OGTT glucose and insulin levels and effects on PGC-1α gene expression in INS1-E cells. Time-dependent values for plasma glucose (A) and insulin (B) in response to 75 g glucose in glucose-tolerant patients differing by −3959 T/C genotypes. Error bars indicate SE. For T/T, T/C, and C/C subjects, n was 238, 110, and 13, respectively. *P < 0.02, ANOVA, adjusted for age, sex, and BMI. C: INS1-E cells were transiently transfected with –3959T TK-Luc and –3959C TK-Luc. Results are means + SD and representative for one of two experiments, each performed in quadruplicate. **P < 0.001. D: Transcription factors and their binding matrices in relation to the PGC-1α promoter region encompassing the –959 T/C polymorphism. Variable nucleotides in the PGC-1α sequence and respective nucleotides in binding sequences of candidate transcription factors are in bold. Numbers refer to nucleotide positions relative to the transcriptional start site. The core sequences of binding sites are in upper case letters. Nucleotides underlined exhibit a high percentage of conservation. Binding matrices allowing for more than one nucleotide at a position are given in International Union of Pure and Applied Chemistry (IUPAC) standard codes for nucleic acids: w = A,T; r = A/G; s = C/G, and m = A,C.

Close modal
FIG. 3.

Associations of the −1679/−3614 alleles with OGGT glucose and insulin levels and effects on PGC-1α gene expression in HepG2 cells. Time-dependent values for plasma glucose (A) and insulin (B) in response to 75 g glucose in glucose-tolerant subjects by −1679 A/G genotypes. Error bars indicate SE. For A/A, G/A, and G/G subjects, n was 162, 161, and 38, respectively. *P < 0.015, ANOVA. C: HepG2 cells were transiently cotransfected with −3614A TK-Luc or −3614C TK-Luc and HNF-1α expression plasmid or the empty pcDNA6/V5-His backbone as indicated. **P < 0.001. D: HNF-1α binding matrix relative to the −3614 A/C polymorphic site. Variable nucleotides in the PGC-1α sequence and the respective nucleotide in the HNF-1α consensus binding sequence are in bold. IUPAC standard code for nucleic acids: w = A and m = A,C.

FIG. 3.

Associations of the −1679/−3614 alleles with OGGT glucose and insulin levels and effects on PGC-1α gene expression in HepG2 cells. Time-dependent values for plasma glucose (A) and insulin (B) in response to 75 g glucose in glucose-tolerant subjects by −1679 A/G genotypes. Error bars indicate SE. For A/A, G/A, and G/G subjects, n was 162, 161, and 38, respectively. *P < 0.015, ANOVA. C: HepG2 cells were transiently cotransfected with −3614A TK-Luc or −3614C TK-Luc and HNF-1α expression plasmid or the empty pcDNA6/V5-His backbone as indicated. **P < 0.001. D: HNF-1α binding matrix relative to the −3614 A/C polymorphic site. Variable nucleotides in the PGC-1α sequence and the respective nucleotide in the HNF-1α consensus binding sequence are in bold. IUPAC standard code for nucleic acids: w = A and m = A,C.

Close modal
TABLE 1

Standardized pairwise linkage disequlibrium coefficients (D′) of polymorphic sites at the PGC-1α gene locus

−3959 T/C−3818 T/G−1679 A/G−1422 T/C+16022 T/C+55180 A/G+64456 A/T+75656 G/A+75918 G/A+76058 A/G+94580 A/G+114481 T/C
−22174 T/C +0.887 −0.912 −0.512 −0.312 +0.333 −0.333 −0.307 +0.069 +0.074 +0.021 +0.093 +0.014 
−3959 T/C  −0.968 −0.984 −0.905 +0.684 −0.018 −0.080 +0.055 −0.230 −0.192 +0.033 +0.102 
−3818 T/G   −0.981 +0.955 +0.423 +0.348 +0.515 +0.094 +0.003 +0.051 −0.005 +0.097 
−1679 A/G    −0.951 −0.461 −0.305 −0.387 −0.385 −0.135 −0.170 −0.207 −0.479 
−1422 T/C     −0.741 +0.244 +0.105 +0.138 +0.103 +0.127 +0.121 +0.162 
+16022 T/C      +0.056 +0.043 +0.085 −0.262 −0.2211 −0.126 +0.054 
+55180 A/G       +1.000 +1.000 −0.940 −0.859 −0.693 +0.593 
+64456 A/T        +0.948 −0.938 −0.900 −0.725 +0.511 
+75656 G/A         −0.924 −0.921 −0.926 −0.112 
+75918 G/A          −0.970 +0.977 −0.549 
+76058 A/G           +0.980 −0.431 
+94580 A/G            −0.015 
−3959 T/C−3818 T/G−1679 A/G−1422 T/C+16022 T/C+55180 A/G+64456 A/T+75656 G/A+75918 G/A+76058 A/G+94580 A/G+114481 T/C
−22174 T/C +0.887 −0.912 −0.512 −0.312 +0.333 −0.333 −0.307 +0.069 +0.074 +0.021 +0.093 +0.014 
−3959 T/C  −0.968 −0.984 −0.905 +0.684 −0.018 −0.080 +0.055 −0.230 −0.192 +0.033 +0.102 
−3818 T/G   −0.981 +0.955 +0.423 +0.348 +0.515 +0.094 +0.003 +0.051 −0.005 +0.097 
−1679 A/G    −0.951 −0.461 −0.305 −0.387 −0.385 −0.135 −0.170 −0.207 −0.479 
−1422 T/C     −0.741 +0.244 +0.105 +0.138 +0.103 +0.127 +0.121 +0.162 
+16022 T/C      +0.056 +0.043 +0.085 −0.262 −0.2211 −0.126 +0.054 
+55180 A/G       +1.000 +1.000 −0.940 −0.859 −0.693 +0.593 
+64456 A/T        +0.948 −0.938 −0.900 −0.725 +0.511 
+75656 G/A         −0.924 −0.921 −0.926 −0.112 
+75918 G/A          −0.970 +0.977 −0.549 
+76058 A/G           +0.980 −0.431 
+94580 A/G            −0.015 

SNPs at −22174, +16022, +55180, +64456, and +114481 were determined in >84 subjects. All other SNPs were determined in 1,608 subjects. Polymorphisms are identified by their nucleotide positions in the genomic sequence relative to the translational start site. SNPs at +16022 and +55180 are located in intron 2, and SNP at +64456 is located in intron 5. SNPs at +75656, +75918, +76058, and +94580 in the coding region correspond to positions +1302, +1564, +1704, and +2962, respectively, in the mRNA sequence relative to the translational start site (23). +/− linkage among sites. D′ > 0.90 (); D′ > 0.70 ()

TABLE 2

PGC-1α haplotype scores for indexes of insulin secretion

IndexesBlock 1 haplotypes
P*Block 2 haplotypes
P*
0000000101010010100000000001001101111000
Frequencies 0.063 0.28 0.14 0.31 0.18 — 0.328 0.096 0.061 0.333 0.166 — 
HOMAsecr −0.89 −2.27 0.68 0.76 1.48 0.131 1.10 −0.04 1.15 −1.55 −0.32 0.426 
First phase −2.19 −1.48 0.18 0.65 2.09 0.091 2.31 0.63 −0.85 −1.36 −1.43 0.091 
Second phase −1.99 −1.47 −0.07 0.41 2.38 0.079 2.27 0.85 −0.53 −1.63 −1.43 0.090 
I30/I0 −1.26 0.15 −0.61 0.45 0.78 0.634 2.76 −0.34 −1.63 −0.81 −1.25 0.022 
I30/G30 −1.96 −1.45 0.01 0.57 2.13 0.126 3.14§ 0.45 −0.66 −2.07 −1.60 0.005 
CIR30 −1.44 −1.50 −0.09 0.83 1.64 0.290 2.89§ 0.73 −0.60 −1.98 −1.63 0.016 
ΔI30G30 −1.73 −1.80 −0.05 0.75 2.09 0.103 2.19 0.16 −1.56 −1.69 −0.68 0.069 
IndexesBlock 1 haplotypes
P*Block 2 haplotypes
P*
0000000101010010100000000001001101111000
Frequencies 0.063 0.28 0.14 0.31 0.18 — 0.328 0.096 0.061 0.333 0.166 — 
HOMAsecr −0.89 −2.27 0.68 0.76 1.48 0.131 1.10 −0.04 1.15 −1.55 −0.32 0.426 
First phase −2.19 −1.48 0.18 0.65 2.09 0.091 2.31 0.63 −0.85 −1.36 −1.43 0.091 
Second phase −1.99 −1.47 −0.07 0.41 2.38 0.079 2.27 0.85 −0.53 −1.63 −1.43 0.090 
I30/I0 −1.26 0.15 −0.61 0.45 0.78 0.634 2.76 −0.34 −1.63 −0.81 −1.25 0.022 
I30/G30 −1.96 −1.45 0.01 0.57 2.13 0.126 3.14§ 0.45 −0.66 −2.07 −1.60 0.005 
CIR30 −1.44 −1.50 −0.09 0.83 1.64 0.290 2.89§ 0.73 −0.60 −1.98 −1.63 0.016 
ΔI30G30 −1.73 −1.80 −0.05 0.75 2.09 0.103 2.19 0.16 −1.56 −1.69 −0.68 0.069 
*

Global statistics for score comparisons of haplotypes with estimated frequencies >0.01;

P < 0.05,

P < 0.01, or

§

P < 0.005 for haplotype-specific statistics. Scores of haplotypes, which were assigned numbers in the order of the respective polymorphism along the PGC-1α gene, were estimated by score statistics (30). Block 1 haplotypes: 0, more common allele (−3959T, −3818T, −1679A, −1422T); 1, less common allele (−3959C, −3818G, −1679G, −1422C). Block 2 haplotypes: 0, more common allele (+1302G, +1564G, +1704A, +2962A); 1, less common allele (+1302A, +1564A, +1704G, +2962G).

TABLE 3

Associations of PGC-1α haplotype blocks with type 2 diabetes

Block 1 haplotypes
Block 2 haplotypes
Frequencies
ScoreP*Frequencies
ScoreP*
Type 2 diabetesControlType 2 diabetesControl
No. of subjects 506 1,480   494 1,478   
0000 0.0606 0.0620 0.406 NS 0.2960 0.3190 −2.623 0.009 
0001 0.2594 0.2624 −0.489 NS 0.1143 0.0961 0.632 NS 
0010 0.3073 0.3167 0.014 NS 0.0000 0.0005 — — 
0011 0.0049 0.0087 −1.270 NS 0.0509 0.0531 0.256 NS 
0100 0.0014 0.0008   0.0057 0.0000 2.785 0.006 
0101 0.1558 0.1522 0.915 NS 0.0000 0.0046 −0.947 NS 
0110 0.0060 0.0019 1.848 0.065 0.0022 0.0028 −0.607 NS 
0111 0.0000 0.0000 —  0.3507 0.3348 1.614 NS 
1000 0.1897 0.1862 −0.542 NS 0.1699 0.1805 0.339 NS 
1001 0.0061 0.0061 −0.275 NS 0.0013 0.0014 −0.540 NS 
1010 0.0000 0.0000 — — 0.0000 0.0001 — — 
1011 0.0000 0.0004 — — 0.0031 0.0011 1.423 NS 
1100 0.0071 0.0015 2.189 0.029 0.0018 0.0011 0.764 NS 
1101 0.0016 0.0010 −0.021 NS 0.0010 0.0000 — — 
1110 0.0000 0.0000 — — 0.0001 0.0000 — — 
1111 0.0000 0.0000 — — 0.0029 0.0048 0.218 NS 
Block 1 haplotypes
Block 2 haplotypes
Frequencies
ScoreP*Frequencies
ScoreP*
Type 2 diabetesControlType 2 diabetesControl
No. of subjects 506 1,480   494 1,478   
0000 0.0606 0.0620 0.406 NS 0.2960 0.3190 −2.623 0.009 
0001 0.2594 0.2624 −0.489 NS 0.1143 0.0961 0.632 NS 
0010 0.3073 0.3167 0.014 NS 0.0000 0.0005 — — 
0011 0.0049 0.0087 −1.270 NS 0.0509 0.0531 0.256 NS 
0100 0.0014 0.0008   0.0057 0.0000 2.785 0.006 
0101 0.1558 0.1522 0.915 NS 0.0000 0.0046 −0.947 NS 
0110 0.0060 0.0019 1.848 0.065 0.0022 0.0028 −0.607 NS 
0111 0.0000 0.0000 —  0.3507 0.3348 1.614 NS 
1000 0.1897 0.1862 −0.542 NS 0.1699 0.1805 0.339 NS 
1001 0.0061 0.0061 −0.275 NS 0.0013 0.0014 −0.540 NS 
1010 0.0000 0.0000 — — 0.0000 0.0001 — — 
1011 0.0000 0.0004 — — 0.0031 0.0011 1.423 NS 
1100 0.0071 0.0015 2.189 0.029 0.0018 0.0011 0.764 NS 
1101 0.0016 0.0010 −0.021 NS 0.0010 0.0000 — — 
1110 0.0000 0.0000 — — 0.0001 0.0000 — — 
1111 0.0000 0.0000 — — 0.0029 0.0048 0.218 NS 
*

Haplotype-specific statistics of haplotypes with estimated frequencies >0.001; P values obtained by log-likelihood ratio tests for block 1 and block 2 haplotype frequency distributions were 0.8106 and 0.0772, respectively. Block 1 haplotypes: 0, more common allele (−3959T, −3818T, −1679A, −1422T); 1, less common allele (−3959C, −3818G, −1679G, −1422C). Block 2 haplotypes: 0, more common allele (+1302G, +1564G, +1704A, +2962A); 1, less common allele (+1302A, +1564A, +1704G, +2962G).

Additional information for this article can be found in an online appendix at http://diabetes.diabetesjournals.org.

This study was supported by Jubilaeumsfondsprojekt no. 9364 and no. 10678 from the Oesterreichische Nationalbank, and by grants from the Medizinische Forschungsgesellschaft Salzburg and the Verein für Medizinische Fortbildung, Salzburg.

We acknowledge the critical reading of the manuscript by Dr. S.M. Soyal.

1
Fajans SS, Bell GI, Polonsky KS: Molecular mechanisms and clinical pathophysiology of maturity-onset diabetes of the young.
N Engl J Med
345
:
971
–980,
2001
2
Yoon JC, Puigserver P, Chen G, Donovan J, Wu Z, Rhee J, Adelmant G, Stafford J, Kahn CR, Granner DK, Newgard CB, Spiegelman BM: Control of hepatic gluconeogenesis through the transcriptional coactivator PGC-1.
Nature
413
:
131
–138,
2001
3
Herzig S, Long F, Jhala US, Hedrick S, Quinn R, Bauer A, Rudolph D, Schutz G, Yoon C, Puigserver P, Spiegelman B, Montminy M: CREB regulates hepatic gluconeogenesis through the coactivator PGC-1.
Nature
413
:
179
–183,
2001
4
Puigserver P, Rhee J, Donovan J, Walkey CJ, Yoon JC, Oriente F, Kitamura Y, Altomonte J, Dong H, Accili D, Spiegelman BM: Insulin-regulated hepatic gluconeogenesis through FOXO1-PGC-1alpha interaction.
Nature
423
:
550
–555,
2003
5
Michael LF, Wu Z, Cheatham RB, Puigserver P, Adelmant G, Lehman JJ, Kelly DP, Spiegelman BM: Restoration of insulin-sensitive glucose transporter (GLUT4) gene expression in muscle cells by the transcriptional coactivator PGC-1.
Proc Natl Acad Sci U S A
98
:
3820
–3825,
2001
6
Czuryt MP, McAnally J, Fishman GI, Olson EN: Regulation of peroxisome proliferator-activated receptor γ coactivator 1α (PGC-1α) and mitochondrial function by MEF2 and HDAC5.
Proc Natl Acad Sci U S A
100
:
1711
–1716,
2003
7
Handschin C, Rhee J, Lin J, Tarr PT, Spiegelman BM: An autoregulatory loop controls peroxisome proliferator-activated receptor gamma coactivator 1 alpha expression in muscle.
Proc Natl Acad Sci U S A
100
:
7111
–7116,
2003
8
Lin J, Wu H, Tarr PT, Zhang CY, Wu Z, Boss O, Michael LF, Puigserver P, Isotani E, Olson EN, Lowell BB, Bassel-Duby R, Spiegelman BM: Transcriptional co-activator PGC-1 alpha drives the formation of slow-twitch muscle fibres.
Nature
418
:
797
–801,
2002
9
Yoon JC, Xu G, Deeney JT, Yang SN, Rhee J, Puigserver P, Levens A, Yang R, Zhang CY, Lowell BB, Berggren PO, Newgard CB, Bonner-Weir S, Weir G, Spiegelman BM: Suppression of beta cell energy metabolism and insulin release by PGC-1alpha.
Dev Cell
5
:
73
–83,
2003
10
Mootha VK, Lindgren CM, Eriksson KF, Subramanian A, Sihag S, Lehar J, Puigserver P, Carlsson E, Ridderstrale M, Laurila E, Houstis N, Daly MJ, Patterson N, Mesirov JP, Golub TR, Tamayo P, Spiegelman BM, Lander ES, Hischhorn JN, Altshuler D, Groop LC: PGC-1alpha-responsive genes involved in oxidative phsophorylation are coordinately downregulated in human diabetes.
Nat Genet
34
:
267
–273,
2003
11
Lin J, Puigserver P, Donovan J, Tarr P, Spiegelman BM: Peroxisome proliferator-activated receptor gamma coactivator 1beta (PGC-1beta), a novel PGC-1-related transcription coactivator associated with host cell factor.
J Biol Chem
277
:
1645
–1648,
2002
12
Patti ME, Butte AJ, Crunkhorn S, Cusi K, Berria R, Kashyap S, Miyazaki Y, Kohane I, Costello M, Saccone R, Landaker EJ, Goldfine AB, Mun E, DeFronzo R, Finlayson J, Kahn CR, Mandarino LJ: Coordinated reduction of genes of oxidative metabolism in humans with insulin resistance and diabetes: potential role of PGC1 and NRF1.
Proc Natl Acad Sci U S A
100
:
8466
–8471,
2003
13
Ek J, Andersen G, Urhammer SA, Gaede PH, Drivsholm T, Borch-Johnsen K, Hansen T, Pedersen O: Mutation analysis of peroxisome proliferator-activated receptor-gamma coactivator-1 (PGC-1) and relationships of identified amino acid polymorphisms to type II diabetes mellitus.
Diabetologia
44
:
2220
–2226,
2001
14
Hara K, Tobel K, Okada T, Kadowaki H, Akanuma Y, Ito C, Kimura S, Kadowaki T: A genetic variation in the PGC-1 gene could confer insulin resistance and susceptibility to type II diabetes.
Diabetologia
45
:
740
–743,
2002
15
Andersen G, Hansen T, Gharani N, Frayling TM, Owen KR, Sampson M, Ellard S, Walker M, Hitman GA, Hattersley AT, McCArthy MI, Pedersen O: A common Gly482Ser polymorphism of PGC-1 is associated with type 2 diabetes in two European populations (Abstract).
Diabetes
51 (Suppl. 2)
:
A49
,
2002
16
Lacquemant C, Chikri M, Boutin P, Samson C, Froguel P: No association between the G482S polymorphism of the proliferator-activated receptor-γ coactivator-1 (PGC-1) gene and type II diabetes in French Caucasians.
Diabetologia
45
:
602
–603,
2002
17
Esterbauer H, Schneitler C, Oberkofler H, Ebenbichler C, Paulweber B, Sandhofer F, Ladurner G, Hell E, Strosberg AD, Patsch JR, Krempler F, Patsch W: A common polymorphism in the promoter of UCP2 is associated with decreased risk of obesity in middle-aged humans.
Nat Genet
28
:
178
–183,
2001
18
Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC: Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man.
Diabetologia
28
:
412
–419,
1985
19
Matsuda M, DeFronzo R: Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp.
Diabetes Care
22
:
1462
–1470,
1999
20
Stumvoll M, Mitrakou A, Pimenta W, Jenssen T: Use of the oral glucose tolerance test to assess insulin release and insulin sensitivity.
Diabetes Care
23
:
295
–301,
2000
21
Hanson RL, Pratley RE, Bogardus C, Narayan KMV, Roumain JML, 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 Epidemiol
151
:
190
–198,
2000
22
Krempler F, Esterbauer H, Weitgasser R, Ebenbichler C, Patsch JR, Miller K, Xie M, Linnemayr V, Oberkofler H, Patsch W: A functional polymorphism in the promoter of UCP2 enhances obesity risk but reduces type 2 diabetes risk in obese middle-aged humans.
Diabetes
51
:
3331
–3335,
2002
23
Oberkofler H, Hölzl B, Esterbauer H, Xie M, Iglseder B, Krempler F, Paulweber B, Patsch W: Peroxisome proliferator-activated receptor-γ gene locus: associations with hypertension middle-aged men.
Hypertension
41
:
368
–372,
2003
24
Terwilliger JD, Ott J:
Handbook of Human Genetic Linkage.
Baltimore, MD, Johns Hopkins University,
1994
25
Oberkofler H, Esterbauer H, Linnemayr V, Strosberg AD, Krempler F, Patsch W: Peroxisome proliferator-activated receptor (PPAR) gamma coactivator-1 recruitment regulates PPAR subtype specificity.
J Biol Chem
277
:
16750
–16757,
2002
26
Oberkofler H, Schraml E, Krempler F, Patsch W: Potentiation of liver X receptor transcriptional activity by peroxisome-proliferator-activated receptor gamma co-activator 1 alpha.
Biochem J
371
:
89
–96,
2003
27
Quandt K, Frech K, Karas H, Wingender E, Werner T: MatInd and MatInspector: new fast and versatile tools for detection of consensus matches in nucleotide sequence data.
Nucleic Acid Res
23
:
4878
–4884,
1995
28
Heinemeyer T, Chen X, Karas H, Kel AE, Kel OV, Liebich I, Meinhardt T, Reuter I, Schacherer F, Wingender E: Expanding the TRANSFAC database towards an expert system of regulatory molecular mechanisms.
Nucleic Acid Res
27
:
318
–322,
1999
29
Mander AP, Clayton DG: Haplotype frequency estimation using an EM algorithm and log-linear modeling.
Stata Tech Bull
55
:
10
–12,
2000
30
Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA: Score tests for association between traits and haplotypes when linkage phase is ambiguous.
Am J Hum Genet
70
:
425
–434,
2002
31
Chiu KC, Chuang L-M, Ryu JM, Tsai GP, Saad MF: The I27L amino acid polymorphism of hepatic nuclear factor-1-alpha is associated with insulin resistance.
J Clin Endocrinol Metab
85
:
2718
–2183,
2000
32
Reich DE, Cargill M, Bolk S, Ireland J, Sabeti PC, Richter DJ, Lavery T, Kouyoumjian R, Farhadian SF, Ward R, Lander ES: Linkage disequilibrium in the human genome.
Nature
411
:
199204
,
2001
33
Daly MJ, Rioux JD, Schaffner SF, Hudson TJ, Lander ES: High-resolution haplotype structure in the human genome.
Nat Genet
29
:
229
–232,
2001
34
Thai MV, Guruswamy S, Cao KT, Pessin JE, Olson AL: Myocyte enhancer factor 2 (MEF2)-binding is required for GLUT4 gene expression in transgenic mice: regulation of MEF2 DNA binding activity in insulin-deficient diabetes.
J Biol Chem
273
:
14285
–14292,
1998
35
Nagasaka S, Tokuyama K, Kusaka I, Hayashi H, Rokkaku K, Nakamura T, Kawakami A, Higashiyama M, Ishikawa S, Saito T: Endogenous glucose production and glucose effectiveness in type 2 diabetic subjects derived from stable-labeled minimal model approach.
Diabetes
48
:
1054
–1060,
1999
36
Henriksen JE, Levin K, Thye-Ronn P, Alford F, Hother-Nielsen O, Holst JJ, Beck-Nielsen H: Glucose-mediated glucose disposal in insulin-resistant normoglycemic relatives of type 2 diabetic patients.
Diabetes
49
:
1209
–1218,
2000
37
Shih DQ, Stoffel M: Dissecting the transcriptional network of pancreatic islets during development and differentiation.
Proc Natl Acad Sci U S A
98
:
14189
–14191,
2001
38
Rhee J, Inoue Y, Yoon JC, Puigserver P, Fan M, Gonzalez FJ, Spiegelman BM: Regulation of hepatic fasting response by PPARgamma coactivator-1alpha (PGC-1): requirement for hepatocyte nuclear factor 4alpha in gluconeogenesis.
Proc Natl Acad Sci U S A
100
:
4012
–4017,
2003
39
Muller YL, Bogardus C, Pedersen O, Baier L: A Gly482Ser missense mutation in the peroxisome proliferator-activated receptor γ coactivator-1 is associated with altered lipid oxidation and early insulin secretion in Pima Indians.
Diabetes
52
:
895
–898,
2003
40
Baar K, Wende AR, Jones TE, Marison M, Nolte LA, Chen M, Kelly DP, Holloszy JO: Adaptations of skeletal muscle to exercise: rapid increase in the transcriptional coactivator PGC-1.
FASEB J
16
:
1879
–1886,
2002
41
Bergman RN, Zaccaro DJ, Watanabe RM, Haffner SM, Saad MF, Norris JM, Wagenknecht LE, Hokanson JE, Rotter JI, Rich SS: Minimal model –based insulin sensitivity has greater heritability and a different genetic basis than homeostasis model assessment or fasting insulin.
Diabetes
52
:
2168
–2174,
2003
42
Miura S, Kai Y, Ono M, Ezaki O: Overexpression of peroxisome proliferator-activated receptor γ coactivator-1α downregulates GLUT4 mRNA in skeletal muscles.
J Biol Chem
278
:
31385
–31390,
2003

Supplementary data