| Type 2 diabetes is a complex genetic disease with heterogeneity among populations. In this study, we used the strategy of candidate gene and positional cloning to identify susceptible genes of type 2 diabetes in the Chinese population. We also evaluated the transferability of HapMap Chinese data among Shanghai population.1. Candidate gene association study on type 2 diabetes in ChineseFirst, we analyzed the association between diabetes and three candidate genes, peroxisome proliferator activated receptor delta (PPARD), activating transcription factor 6 (ATF6) and retinol binding protein 4 (RBP4). A total of 287 diabetic patients and 376 normal controls were recruited and genotyped for the single nucleotide polymorphisms (SNPs)–87 T>C of PPARD as well as Ala145Pro of ATF6. We found C allele carriers of PPARD–87 T>C had significantly higher plasma glucose levels at both fasting status (controls: 4.85±0.04 vs 4.99±0.04, P = 0.0078; cases: 6.89±0.12 vs 7.84±0.18, P < 0.0001) and 2 hours after glucose stimulation (cases: 13.53±0.23 vs 14.89±0.33, P < 0.0001). Meanwhile, poorer insulin sensitivity (controls: 3.29±0.08 vs 3.06±0.07, P=0.0365; cases: 1.51±0.03 vs 1.42±0.03, P = 0.0058) was detected among them. The Ala allele of ATF6 Ala145Pro was significantly more frequent in the probands of early onset type 2 diabetic pedigrees (77.6% vs 69.4%, P = 0.0417). The Pro allele carriers had significantly lower high-density lipoprotein cholesterol levels (1.33±0.02 vs 1.25±0.02, P = 0.0140). For RBP4, we sequenced exons and the putative promoter region in 32 subjects to identify variants. Taking account of the pairwise linkage disequilibrium (LD) and minor allele frequencies, five SNPs were further genotyped in 255 type 2 diabetes patients and 372 normal controls. We found a rare haplotype CAA formed by +5388 C>T, +8201 T>A and +8204 T>A was more frequent in diabetic patients (P=0.0343, empirical P=0.0659). In the normal controls, the SNP +5388 C>T was associated with serum C-peptide levels at both fasting status (1.80±0.05 vs 1.56±0.07, P=0.0162) and 2-hours after glucose stimulation (7.15±0.26 vs 5.71±0.38, P=0.0075). The non-coding SNPs were also found to be associated with circulating RBP4 concentrations in both groups (P<0.05). 2. Evaluation the transferability of HapMap Chinese data in a Shanghai populationThe HapMap project catalogs millions of common SNPs in the human genome in four major populations including northern Hans, with the aim to facilitate association studies on complex diseases. In this study, we used genotype data from International Type 2 Diabetes 1q Consortium to examine the transferability of HapMap Chinese data in Shanghai population. The HapMap Chinese data showed high consistence with Shanghai population on allele frequencies, LD coefficient r2 and haplotype frequencies (R=0.94, 0.97, 0.99 respectively, all P<0.0001). In addition, tagging SNP set selected from HapMap Chinese data also performed well in Shanghai population with the threshold 0.8 of r2 to cover 93.8% of the SNPs. The HapMap SNP data performed well in the Shanghai population and thus could be a powerful tool for the genetic studies on complex disease in southern Hans.3. Positional cloning study of type 2 diabetes susceptible gene on chromosome 1q23.3 in ChinesePreviously, we detected a linkage signal of type 2 diabetes on chromosome 1q21-q25. International Type 2 Diabetes 1q Consortium then did a fine mapping analysis and located one of the susceptible genes on chromosome 1q23.3. In this study, we did a positional cloning study based on the previous linkage and fine mapping studies. Considering both LD and potential SNP function, 94 SNP were selected from HapMap data and genotyped in 1892 diabetic patients and 1808 normal controls using matrix-assisted laser desorption inoization-time of flight mass spectrometry. After quality control of genotyping, a total of 3491 individuals and 73 SNPs were analyzed. We found rs12742393, located in intron2 of NOS1AP, was associated with type 2 diabetes, C allele was significantly higher in the cases (25.2% vs 21.6%, P = 0.0004, empirical P = 0.0342). The haplotype TC formed by rs4531272 - rs12742393 was also significantly more frequent in the cases (52.6% vs 49.5%, P = 0.0092, empirical P = 0.0424). In the normal controls, rs1415263 was associated with serum levels of total cholesterol (TT vs TC vs CC: 5.13±0.06 vs 4.99±0.03 vs 4.94±0.09, P < 0.05) and low-density lipoprotein cholesterol (TT vs TC vs CC: 3.45±0.05 vs 3.36±0.03 vs 3.24±0.04, P < 0.05). We also found rs12029454 was associated with serum level of low-density lipoprotein cholesterol (GG vs GA vs AA: 3.27±0.03 vs 3.39±0.03 vs 3.46±0.07, P < 0.05). |
PDF Full Text Request