Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2010 Jan 11.
Published in final edited form as: Genes Immun. 2009 Dec;10(Suppl 1):S49–S53. doi: 10.1038/gene.2009.91

No association of the IRS1 and PAX4 genes with type I diabetes

R Bergholdt 1, C Brorsson 1, B Boehm 2, G Morahan 3, F Pociot 1,4; the Type I Diabetes Genetics Consortium
PMCID: PMC2804261  NIHMSID: NIHMS163339  PMID: 19956100

Abstract

To reassess earlier suggested type I diabetes (T1D) associations of the insulin receptor substrate 1 (IRS1) and the paired domain 4 gene (PAX4) genes, the Type I Diabetes Genetics Consortium (T1DGC) evaluated single-nucleotide polymorphisms (SNPs) covering the two genomic regions. Sixteen SNPs were evaluated for IRS1 and 10 for PAX4. Both genes are biological candidate genes for T1D. Genotyping was performed in 2300 T1D families on both Illumina and Sequenom genotyping platforms. Data quality and concordance between the platforms were assessed for each SNP. Transmission disequilibrium testing neither show T1D association of SNPs in the two genes, nor did haplotype analysis. In conclusion, the earlier suggested associations of IRS1 and PAX4 to T1D were not supported, suggesting that they may have been false positive results. This highlights the importance of thorough quality control, selection of tagging SNPs, more than one genotyping platform in high throughput studies, and sufficient power to draw solid conclusions in genetic studies of human complex diseases.

Keywords: autoimmune disease, genetic susceptibility, single-nucleotide polymorphism, type I diabetes, IRS1, PAX4

Introduction

To reassess earlier suggested type I diabetes (T1D) associations of the insulin receptor substrate 1 (IRS1) and the paired domain 4 gene (PAX4) genes, the Type I Diabetes Genetics Consortium (T1DGC, http://www.t1dgc.org) evaluated single-nucleotide polymorphisms (SNPs) covering the two genomic regions. The design and sample size of the study, genotyping 2295 affected sib-pair (ASP) families from nine cohorts, should have adequate power to draw conclusions about the possible T1D association of these two candidate genes.

The IRS1 gene on chromosome 2q36 is the substrate of the insulin receptor tyrosine kinase participating in insulin signaling. The protein is expressed in a variety of insulin responsive cells and tissues. Binding of insulin to its receptor induces phosphorylation of the cytosolic substrates IRS1 and IRS2. Activation of IRS1 is a key initial step in the insulin-signaling pathway. Functional studies of variants in the IRS1 gene showed impaired insulin signaling through the PI3-kinase pathway1 and impaired insulin secretion.2 IRS1 can be considered as a biological candidate gene in type II diabetes (T2D) as well as T1D.

A SNP in the IRS1 gene (rs1801278, G972R) was initially shown associated with T2D in several studies.3–5 Although later studies questioned this association,6–8 a recent genome-wide association study found convincing association to T2D for SNPs adjacent to the IRS1 gene.9 In addition, a possible IRS1 gene association to T1D has not been clearly resolved. Two independent studies suggested association of the T2D-associated SNP (rs1801278) with T1D. The first study was an Italian case–control and family study10 and the second studied >700 US/UK families.11 Replication of these findings was not achieved either in a large study of >2000 families,12 or in a Danish meta-analysis combining T1D studies of IRS1.13 In addition, there have been no evidence from recent genome-wide association scans supporting association of T1D with a region near IRS1.7,14,15

The PAX4 gene on chromosome 7q32 belongs to a family of transcription factors containing a paired box domain involved in organogenesis of the embryo. The transcription factors PAX4 and PAX6 are presumed to trigger early events in cell differentiation, including mediating the differentiation of the endoderm-derived endocrine pancreas.16 In mice, PAX4 is essential for the differentiation of insulin-producing β cells in pancreas.17 Recent studies also suggested this gene to be crucial for mature β-cell expansion and survival.18 As with the case of IRS1, SNPs in PAX4 have been associated with T2D in both Japanese and African-American populations.19–23 A PAX4 missense SNP (rs712701) has been reported to be associated with T1D;24 however, subsequent studies were not able to replicate this finding in Finnish, Hungarian, UK, US, or Danish populations.25–28 Further, there was no evidence for an association of T1D with the region on chromosome 7q32 containing PAX4 in recent genome-wide association scans of T1D.7,14,15 It remains possible that significant genetic heterogeneity or population differences may exist and help explain the lack of replication related to both IRS1 and PAX4 with T1D.29

In this report, we present data from well-powered studies in ASP families with T1D. A panel of tagging SNPs for IRS1 and PAX4 has been genotyped to assess the association of both of these biological candidate genes with respect to their function in T1D susceptibility.

Results

SNPs were genotyped in a set of 11159 individuals comprising 5003 T1D affected individuals collected by the T1DGC. In total, the material comprised 2298 T1D nuclear families of which 2107 families were of European origin and 191 were Asian-Pacific. A total of 1942 individuals (of the 11159) were removed from the analyses because of missing phenotype information.

Genotyping of IRS1 and PAX4 SNPs was performed on both the Illumina and the Sequenom platforms (Table 1). Individuals and SNPs that had call rates below 90% were removed from analysis; genotyping success rate in remaining individuals was then determined. For IRS1 evaluation, a total of 7961 individuals were successfully genotyped on the Illumina platform (genotyping success rate for 16 SNPs was 99.88%) and had phenotype information making them eligible for analysis. For the Sequenom platform, a total of 7896 individuals had a genotyping success rate of 99.33% and were included in the final analyses. For PAX4, a total of 8005 individuals were successfully genotyped on the Illumina platform (genotyping success rate for 10 SNPs of 99.42%) and a total of 7878 individuals on the Sequenom platform (genotyping success rate 98.95%) were included for analyses.

Table 1.

Genotyped markers in PAX4 and IRS1

SNP Minor allele Illumina MAF Sequenom MAF Concordance rate Failed on Illumina but not on Sequenom Failed on Sequenom but not on Illumina
IRS1
 rs16822551 C 0.089 0.0902 99.94 397 391
 rs11683087 G 0.193 0.1926 99.95 402 349
 rs2251692 T 0.2616 NA — — —
 rs17208239 A 0.0631 0.0649 99.30 372 746
 rs17208470 T 0.0982 0.0978 99.93 398 392
 rs2435185 G 0.0259 0.0251 99.98 407 375
 rs4675095 T 0.0582 0.0583 99.99 403 377
 rs1801278 A 0.0653 0.0643 99.93 425 409
 rs1801123 G 0.0935 0.0931 99.95 413 385
 rs6725330 G 0.1142 0.1141 99.98 534 349
 rs6725556 C 0.0717 0.0713 99.95 375 370
 rs13018009 C 0.0237 0.0239 100 403 355
 rs4675096 A 0.0842 0.0837 99.97 398 381
 rs956115 C 0.0831 NA — — —
 rs957797 G 0.1677 0.1682 99.99 407 365
 rs13417106 C 0.0872 0.0968 98.14 403 416
PAX4
 rs3779536 T 0.0338 0.0325 99.43 411 369
 rs806213 T 0.0308 0.0298 100 391 466
 rs3735640 G 0.1254 0.1243 97.90 400 409
 rs806216 A 0.2113 0.2102 99.88 404 357
 rs3824006 A 0.1061 0.1054 99.93 409 394
 rs10229583 T 0.2359 0.2363 99.91 409 375
 rs712701 A 0.2395 0.2131 96.16 1269 853
 rs327518 T 0.3984 0.3976 99.92 398 429
 rs7801118 A 0.1741 0.1698 96.82 402 488
 rs806187 G 0.2747 0.276 99.98 403 358
Open in a new tab

Abbreviations: MAF, minor allele frequency; IRS1, insulin receptor substrate 1.

Single-nucleotide polymorphisms (SNPs) were genotyped on two platforms simultaneously. Genotyping was done using the Illumina technology (Illumina, San Diego, CA, USA) and the Sequenom technology (Sequenom, San Diego, CA, USA). Minor allele is indicated in the table, as well as the minor allele frequencies on the two genotyping platforms, Illumina and Sequenom. For each SNP, concordance rate for individuals successfully genotyped on both platforms is calculated, accompanied by the number of individuals genotyped on one platform but not the other.

In the aggregate data of ASP families, all SNPs had genotypic distributions that were consistent with Hardy–Weinberg Equilibrium expectations (P>0.001, data not shown). Linkage disequilibrium within the IRS1 and PAX4 genes (Figure 1a and b) is consistent with that in the HapMap. The results of transmission disequilibrium test (TDT) for association of genotyped SNPs with T1D are also provided (Figure 1a and b). There was no significant evidence observed that supported an association for SNPs in any of the two genes with T1D (Table 2).

Figure 1.

Figure 1

Open in a new tab

(a) Linkage disequilibrium plots generated with the current genotyping data from the Illumina platform is shown including indication of resulting P-values from each platform for all SNPs for IRS1. Dotted lines represent P =0.1 and 0.05, respectively. Plots were drawn using the SNP Plotter algorithm (http://cbdb.nimh.nih.gov/~kristin/snp.plotter.html) in R (http://www.r-project.org). (b) Linkage disequilibrium plots generated with the current genotyping data from the Illumina platform is shown including indication of resulting P-values from each platform for all SNPs for PAX4. Dotted lines represents P =0.1 and 0.05, respectively. Plots were drawn using the SNP Plotter algorithm (http://cbdb.nimh.nih.gov/~kristin/snp.plotter.html) in R (http://www.r-project.org).

Table 2.

Results of TDT (transmission disequilibrium test) results as calculated in PLINK32

SNP A1:A2 Illumina
 Sequenom
T:U_TDT CHISQ_TDT P_TDT T:U_TDT CHISQ_TDT P_TDT
IRS1
 rs16822551 C:T 489:514 0.6231 0.4299 448:484 1.391 0.2383
 rs11683087 G:A 955:945 0.05263 0.8185 927:930 0.004847 0.9445
 rs2251692 T:C 1209:1103 4.86 0.02749 NA NA NA
 rs17208239 A:C 325:345 0.597 0.4397 287:338 4.162 0.0414
 rs17208470 T:G 548:544 0.01465 0.9037 510:518 0.06226 0.803
 rs2435185 G:A 153:118 4.52 0.0335 142:115 2.837 0.0921
 rs4675095 T:A 350:316 1.736 0.1877 325:310 0.3543 0.5517
 rs1801278 A:G 379:377 0.005291 0.942 355:353 0.00565 0.9401
 rs1801123 G:A 537:488 2.342 0.1259 499:481 0.3306 0.5653
 rs6725330 G:A 601:578 0.4487 0.503 595:585 0.08475 0.771
 rs6725556 C:T 396:396 0 1 369:398 1.096 0.295
 rs13018009 C:T 145:131 0.7101 0.3994 144:124 1.493 0.2218
 rs4675096 A:G 472:469 0.009564 0.9221 454:448 0.03991 0.8417
 rs956115 C:G 454:510 3.253 0.07129 NA NA NA
 rs957797 G:A 840:884 1.123 0.2893 808:863 1.81 0.1785
 rs13417106 C:T 482:476 0.03758 0.8463 505:490 0.2261 0.6344
PAX4
 rs3779536 T:G 200:214 0.4734 0.4914 181:194 0.4507 0.502
 rs806213 T:C 183:160 1.542 0.2143 162:149 0.5434 0.461
 rs3735640 G:C 677:665 0.1073 0.7432 639:634 0.01964 0.8886
 rs806216 A:C 1002:965 0.696 0.4041 960:941 0.1899 0.663
 rs3824006 A:G 579:594 0.1918 0.6614 536:569 0.9855 0.3208
 rs10229583 T:C 1082:1084 0.001847 0.9657 1026:1031 0.01215 0.9122
 rs712701 A:C 893:798 5.337 0.02088 969:969 0 1
 rs327518 T:C 1456:1475 0.1232 0.7256 1380:1419 0.5434 0.461
 rs7801118 A:G 879:878 0.0005692 0.981 782:833 1.611 0.2044
 rs806187 G:A 1222:1243 0.1789 0.6723 1164:1157 0.02111 0.8845
Open in a new tab

Number of transmissions vs non-transmissions (T:U_TDT), χ2- and P-values for each single-nucleotide polymorphism (SNP) on each genotyping platform are shown.

For IRS1 SNPs, no consistent association to T1D was shown (Table 2). Two IRS1 SNPs were significantly associated with T1D (P<0.05) in the Illumina data; however, one SNP was not genotyped on Sequenom and the other SNP did not attain statistical significance (P =0.09). It should be noted that the second SNP had equal ‘missingness’ values and high concordance rate with the Illumina data. One SNP in IRS1 was significantly associated (P =0.04) in Sequenom data, but not (P =0.43) in the Illumina data. This inconsistency may be due to a higher rate of ‘missingness’ because of difficulty in genotyping calls on the Sequenom platform.

In PAX4, the missense SNP, rs712701, exhibited substantial heterogeneity in the Illumina data (P =0.02), whereas random transmission was observed in the Sequenom data (Table 2). The concordance rate for all genotyped individuals for this SNP was lower than other SNPs, suggesting technical problems regarding genotyping of this SNP. All other SNPs in the PAX4 region did not exhibit a significant association with T1D. On the basis of these data, PAX4 is not associated with T1D, although this study shows that genotyping results of rs712701 should be interpreted with caution.

Discussion

We have evaluated the association of IRS1 and PAX4 SNPs with T1D in a large collection of ASP families. We report on aggregate data only in this study, but minor population differences exist. In IRS1, the SNP rs6725330 showed significant association to T1D in the HBDI collection (428 families; 921 individuals), particularly in the Illumina genotyping data (P =0.0005); however, this SNP did not reach statistical significance in other populations.

Analyses using both TDT and pedigree disequilibrium test methods were performed for two- and three-marker haplotypes for markers in both genes (excluding SNP rs712701 in PAX4). Calculations were performed using the TDTPHASE and PDTPHASE options in UNPHASED v2.404.30 There was no evidence of association with any two- or three-marker haplotype, consistent with the lack of significant evidence of association with SNP analyses.

In summary, the earlier suggested associations of IRS1 and PAX4 to T1D are not supported in this study. These data suggest that the original findings may have been false positive results because of insufficient sample size to detect effects of the magnitude reported. This study highlights the importance of thorough quality control, carefully chosen tagging SNPs, more than one genotyping platform in high throughput studies, and robustly designed experiments to provide enough power to draw solid conclusions in genetic studies of human complex diseases.

Materials and methods

Samples

The T1DGC has assembled a collection of ASPs families for the detection of T1D susceptibility genes (http://www.t1dgc.org). Samples used in this project consisted of 2295 ASP families from nine cohorts. The families in this study were collected from the following T1DGC networks and contributing sites: 9% Asia-Pacific, 29% European, 35% North American, and 27% UK/Sardinian. The majority of the subjects were Caucasian (81%) with 18% unknown ethnicity and 1% other (Asian, African American, Pacific Islander).

Genotyping

SNPs in IRS1 and PAX4 were genotyped on two platforms (Illumina and Sequenom). Details of the sample, quality control, and other aspects of the data can be found in this volume (Brown et al.31) Individuals and SNPs that had call rates below 90% were removed from analysis. For IRS1 on the Illumina platform, genotyping success rate for 16 SNPs was 99.88%; on the Sequenom platform, the genotyping success rate was 99.33%. For PAX4 on the Illumina platform, the genotyping success rate for 10 SNPs was 99.42%; on the Sequenom platform, the genotyping success rate was 98.95%. All 16 SNPs covering the IRS1 gene were successfully genotyped on the Illumina platform, whereas only 14 of these could be genotyped on the Sequenom platform (rs2251692 and rs956115 failed on Sequenom). All SNPs had minor allele frequencies >1%, and concordance rates between the two genotyping platforms were above 99.9% for most SNPs (Table 1).

Statistical analysis

Data quality checks, using standard methods including the PedCheck program, are reported by Brown et al. (this volume). The linkage disequilibrium plots and TDT analyses for the haplotypes were computed using Haploview software (http://www.broad.mit.edu/mpg/haploview). TDT analyses were performed by using PLINK software.32 Analyses using both TDT and pedigree disequilibrium test methods were performed for two- and three-marker haplotypes for markers in both genes (excluding SNP rs712701 in PAX4). Calculations were done using the TDTPHASE and PDTPHASE options in UNPHASED v2.404.30

Acknowledgments

This research uses resources provided by the Type I Diabetes Genetics Consortium, a collaborative clinical study sponsored by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute of Allergy and Infectious Diseases (NIAID), National Human Genome Research Institute (NHGRI), National Institute of Child Health and Human Development (NICHD), and Juvenile Diabetes Research Foundation International (JDRF) and supported by U01 DK062418. We thank all the T1D families for their participation in the study. R Bergholdt was supported by a grant from the Danish Medical Research Council (271-05-0672). Genotyping was performed at the Broad Institute Center for Genotyping and Analysis is supported by grant U54 RR020278 from the National Center for Research Resources. BOB was support by the German Research Foundation (DFG: SFB 518 & GrK 1041).

Footnotes

Conflict of interest

The authors declare no conflict of interest.

References

  • 1.Almind K, Inoue G, Pedersen O, Kahn CR. A common amino acid polymorphism in IRS-1 causes impaired insulin signaling: evidence from transfection studies. J Clin Invest. 1996;97:2569–2575. doi: 10.1172/JCI118705. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Stumvoll M, Fritsche A, Volk A, Stefan N, Madaus A, Maerker E, et al. The Gly972Arg polymorphism in the insulin receptor substrate-1 gene contributes to the variation in insulin secretion in normal glucose-tolerant humans. Diabetes. 2001;50:882–885. doi: 10.2337/diabetes.50.4.882. [DOI] [PubMed] [Google Scholar]
  • 3.Almind K, Bjorbaek C, Vestergaard H, Hansen T, Echwald S, Pedersen O. Amino acid polymorphisms of insulin receptor substrate-1 in. Lancet. 1993;342:828–832. doi: 10.1016/0140-6736(93)92694-o. [DOI] [PubMed] [Google Scholar]
  • 4.Hitman GA, Hawrami K, McCarthy MI, Viswanathan M, Snehalatha C, Ramachandran A, et al. Insulin receptor substrate-1 gene mutations in NIDDM; Implications for the study of polygenic disease. Diabetologia. 1995;38:481–486. doi: 10.1007/BF00410287. [DOI] [PubMed] [Google Scholar]
  • 5.Jellema A, Zeegers MPA, Feskens EJM, Dagnelie PC, Mensink RP. GlY972Arg variant in the insulin receptor substrate-1 gene and association with Type 2 diabetes: a meta-analysis of 27 studies. Diabetologia. 2003;46:990–995. doi: 10.1007/s00125-003-1126-4. [DOI] [PubMed] [Google Scholar]
  • 6.Florez JC, Sjogren M, Burtt N, Orho-Melander M, Schayer S, Sun M, et al. Association testing in 9000 people fails to confirm the association of the insulin receptor substrate-1 G972R polymorphism with type 2 diabetes. Diabetes. 2004;53:3313–3318. doi: 10.2337/diabetes.53.12.3313. [DOI] [PubMed] [Google Scholar]
  • 7.Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature. 2007;447:661–678. doi: 10.1038/nature05911. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Zeggini E, Parkinson J, Halford S, Owen KR, Frayling TM, Walker M, et al. Association studies of insulin receptor substrate 1 gene (IRS1) variants in type 2 diabetes samples enriched for family history and early age of onset. Diabetes. 2004;53:3319–3322. doi: 10.2337/diabetes.53.12.3319. [DOI] [PubMed] [Google Scholar]
  • 9.Rung J, Cauchi S, Albrechtsen A, Shen L, Rocheleau G, Cavalcanti-Proenca C, et al. Genetic variant near IRS1 is associated with type 2 diabetes, insulin resistance and hyperinsulinemia. Nat Genet. 2009;41:1110–1115. doi: 10.1038/ng.443. [DOI] [PubMed] [Google Scholar]
  • 10.Federici M, Petrone A, Porzio O, Bizzarri C, Lauro D, D’Alfonso R, et al. The Gly972fwdarwArg IRS-1 variant is associated with type 1 diabetes in continental Italy. Diabetes. 2003;52:887–890. doi: 10.2337/diabetes.52.3.887. [DOI] [PubMed] [Google Scholar]
  • 11.Morrison VA, Onengut-Gumuscu S, Concannon P. A functional variant of IRS1 is associated with type 1 diabetes in families from the US and UK. Mol Genet Metab. 2004;81:291–294. doi: 10.1016/j.ymgme.2003.10.018. [DOI] [PubMed] [Google Scholar]
  • 12.Eftychi C, Howson JM, Barratt BJ, Vella A, Payne F, Smyth DJ, et al. Analysis of the type 2 diabetes-associated single nucleotide polymorphisms in the genes IRS1, KCNJ11, and PPARG2 in type 1 diabetes. Diabetes. 2004;53:870–873. doi: 10.2337/diabetes.53.3.870. [DOI] [PubMed] [Google Scholar]
  • 13.Johansen A, Jensen DP, Bergholdt R, Mortensen HB, Pociot F, Nerup J, et al. IRS1, KCNJ11, PPARgamma2 and HNF-1alpha: do amino acid polymorphisms in these candidate genes support a shared aetiology between type 1 and type 2 diabetes? Diabetes Obes Metab. 2006;8:75–82. doi: 10.1111/j.1463-1326.2005.00471.x. [DOI] [PubMed] [Google Scholar]
  • 14.Cooper JD, Smyth DJ, Smiles AM, Plagnol V, Walker NM, Allen JE, et al. Meta-analysis of genome-wide association study data identifies additional type 1 diabetes risk loci. Nat Genet. 2008;40:1399–1401. doi: 10.1038/ng.249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Hakonarson H, Grant SFA, Bradfield JP, Marchand L, Kim CE, Glessner JT, et al. A genome-wide association study identifies KIAA0350 as a type 1 diabetes gene. Nature. 2007;448:591–594. doi: 10.1038/nature06010. [DOI] [PubMed] [Google Scholar]
  • 16.Mansouri A, St-Onge L, Gruss P. Role of Pax genes in endoderm-derived organs. Trends Endocrinol Metab. 1999;10:164–167. doi: 10.1016/s1043-2760(98)00133-7. [DOI] [PubMed] [Google Scholar]
  • 17.Sosa-Pineda B, Chowdhury K, Torres M, Oliver G, Gruss P. The Pax4 gene is essential for differentiation of insulin-producing beta cells in the mammalian pancreas. Nature. 1997;386:399–402. doi: 10.1038/386399a0. [DOI] [PubMed] [Google Scholar]
  • 18.Brun T, Gauthier BR. A focus on the role of Pax4 in mature pancreatic islet beta-cell expansion and survival in health and disease. J Mol Endocrinol. 2008;40:37–45. doi: 10.1677/JME-07-0134. [DOI] [PubMed] [Google Scholar]
  • 19.Kanatsuka A, Tokuyama Y, Nozaki O, Matsui K, Egashira T. Beta-cell dysfunction in late-onset diabetic subjects carrying homozygous mutation in transcription factors NeuroD1 and Pax4. Metabolism. 2002;51:1161–1165. doi: 10.1053/meta.2002.34707. [DOI] [PubMed] [Google Scholar]
  • 20.Mauvais-Jarvis F, Sobngwi E, Porcher R, Riveline JP, Kevor-kian JP, Vaisse C, et al. Ketosis-prone type 2 diabetes in patients of sub-Saharan African origin: clinical pathophysiology and natural history of beta-cell dysfunction and insulin resistance. Diabetes. 2004;53:645–653. doi: 10.2337/diabetes.53.3.645. [DOI] [PubMed] [Google Scholar]
  • 21.Shimajiri Y, Sanke T, Furuta H, Hanabusa T, Nakagawa T, Fujitani Y, et al. A missense mutation of Pax4 gene (R121W) is associated with type 2 diabetes in Japanese. Diabetes. 2001;50:2864–2869. doi: 10.2337/diabetes.50.12.2864. [DOI] [PubMed] [Google Scholar]
  • 22.Shimajiri Y, Shimabukuro M, Tomoyose T, Yogi H, Komiya I, Takasu N. PAX4 mutation (R121W) as a prodiabetic variant in Okinawans. Biochem Biophys Res Commun. 2003;302:342–344. doi: 10.1016/s0006-291x(03)00176-1. [DOI] [PubMed] [Google Scholar]
  • 23.Tokuyama Y, Matsui K, Ishizuka T, Egashira T, Kanatsuka A. The Arg121Trp variant in PAX4 gene is associated with beta-cell dysfunction in Japanese subjects with type 2 diabetes mellitus. Metabolism. 2006;55:213–216. doi: 10.1016/j.metabol.2005.08.014. [DOI] [PubMed] [Google Scholar]
  • 24.Biason-Lauber A, Boehm B, Lang-Muritano M, Gauthier BR, Brun T, Wollheim CB, et al. Association of childhood type 1 diabetes mellitus with a variant of PAX4: possible link to beta cell regenerative capacity. Diabetolog. 2005;48:900–905. doi: 10.1007/s00125-005-1723-5. [DOI] [PubMed] [Google Scholar]
  • 25.Geng DG, Liu SY, Steck A, Eisenbarth G, Rewers M, She JX. Comment on: Biason-Lauber A, Boehm B, Lang-Muritano M et al. (2005) Association of childhood type 1 diabetes mellitus with a variant of PAX4: possible link to beta cell regenerative capacity. Diabetologia 48: 900–905. Diabetologia. 2006;49:215–216. doi: 10.1007/s00125-005-0064-8. [DOI] [PubMed] [Google Scholar]
  • 26.Gylvin T, Bergholdt R, Nerup J, Pociot F. Reply to: Biason-Lauber A, Boehm B, Lang-Muritano M et al. (2005) Association of childhood type 1 diabetes mellitus with a variant of PAX4: possible link to beta cell regenerative capacity. Diabetologia 48: 900–905. Diabetolog. 2005;48:2183–2184. doi: 10.1007/s00125-005-1920-2. [DOI] [PubMed] [Google Scholar]
  • 27.Hermann R, Mantere J, Lipponen K, Veijola R, Soltesz G, Otonkoski T, et al. Lack of association of PAX4 gene with type 1 diabetes in the Finnish and Hungarian populations. Diabetes. 2005;54:2816–2819. doi: 10.2337/diabetes.54.9.2816. [DOI] [PubMed] [Google Scholar]
  • 28.Martin RJ, Savage DA, Carson DJ, Maxwell AP, Patterson CC. The PAX4 gene variant A1168C is not associated with early onset Type 1 diabetes in a UK population. Diab Med. 2006;23:927–928. doi: 10.1111/j.1464-5491.2006.01869.x. [DOI] [PubMed] [Google Scholar]
  • 29.Barrett J, Clayton D, Concannon P, Akolkar B, Erlich HA, Julier C, et al. Genome-wide association study and meta-analysis find that over 40 loci affect risk of type 1 diabetes. Nat Genet. 2009;41:703–707. doi: 10.1038/ng.381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Dudbridge F. Pedigree disequilibrium tests for multilocus haplotypes. Genet Epidemiol. 2003;25:115–121. doi: 10.1002/gepi.10252. [DOI] [PubMed] [Google Scholar]
  • 31.Brown WM, Pierce JJ, Hilner JE, Perdue LH, Lohman K, Lu L, et al. the Type I Diabetes Genetics Consortium. Overview of the Rapid Response data. Genes Immun. 2009;10(Suppl 1):S5–S15. doi: 10.1038/gene.2009.85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–575. doi: 10.1086/519795. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES