Gene Polymorphism Association With Type 2 Diabetes And Related Gene-Gene And Gene-Environment Interactions In An Uyghur Population

Objective: To investigate the relationship among environmental factors, the gene polymorphism of CDKAL1, FTO, CDKN2A/2B, and type 2 diabetes mellitus(T2DM) in Uyghur, and further analyse the possible gene-gene and gene-environment interaction, which can provide theoretical basis for the prevention and treatment of DM in Uyghur. Methods: A case-control study was designed. 1,000 cases of newly hospitalized Uyghur with T2 DM were selected from March 2012 to September 2013 at the First Affiliated Hospital of Xinjiang Medical University and the First People’s Hospital of Kashi region. 1,010 Uyghur people without DM who did physical examination at the same period of time in those two hospitals were selected as control group. The relevant information was obtained through questionnaire survey, human body measurement and biochemical examination, reviewing medical records. Sequenom Mass ARRAY○R SNP technology was used to detect 23 gene loci polymorphism of CDKN2A/2B, CDKAL1, FTO. 22 gene loci in accordance with Hardy-Weinberg equilibrium were chose for further analysis. Chi-square was used to analyze the distribution of genotype and allele. Logistic regression analysis was used to analyze the relationship among environmental factors, genetic polymorphisms and T2 DM in Uyghur. The gene-gene, gene-environment interaction of three genes were analyzed by crossover analysis and GMDR. Results: 1) The multiple logistic regression result suggested that OR value was 0.462(P=0.018, 95% CI: 0.243~0.878) in the risk of T2 DM from the urban population than those from the rural areas. With the increase of the degree of culture, OR value was 0.682(P=0.005, 95% CI: 0.521~0.893) in the risk of T2 DM. With the increase in the monthly income, the risk of T2 DM elevated 2.885 times(P<0.001, 95% CI: 1.745~4.768). The risk of T2 DM in patients with hypertension, coronary heart disease, gout, and family history of DM raised 4.048 times(P=0.001, 95% CI: 2.356~6.956), 3.126 times(P=0.001, 95% CI: 1.559~6.270), 7.158 times(P=0.032, 95% CI: 1.187~43.168), 6.527 times(P<0.001, 95% CI: 3.703~11.506) than those without hypertension, coronary heart disease, gout, and family history of DM. With a history of smoking, the OR value was 0.446(P=0.009, 95% CI: 0.243~0.818) in the risk of T2 DM. The risk of T2 DM increased by 4.061 times when the staple food was rice(P<0.001, 95% CI: 1.913~8.622) than those was not rice. The risk of T2 DM increased by 1.503 times(P=0.026, 95% CI: 1.051~2.150) and 3.143 times(P<0.001, 95% CI: 1.712~5.770), as the frequency of cakes and vegetables intake per week decreased. With the increase of waist circumference, the risk of T2 DM increased 1.051 times(P=0.007, 95% CI: 1.014~1.091). With the increase of urine protein, the risk of T2 DM increased by 1.960 times(P=0.003, 95% CI: 1.260~3.050). 2) Compared with those in the female patients with T2 DM in Uyghur, the level of serum creatinine, uric acid, triglyceride, alanine aminotransferase and waistline in the male patients were higher(P<0.05), the high density lipoprotein level was lower(P<0.001). In T2 DM group, the proportion of male patients with coronary heart disease, dyslipidemia, drinking history, smoking history, working pressure, atherosclerosis were significantly higher than those of female patients(P<0.001). 3) The frequency of genotype and allele in locus rs10811661 of CDKN2A/2B gene between T2 DM group and control group were significantly different(P=0.027 and 0.007), and the OR value of allele C was 0.816(P=0.007, 95% CI: 0.705~0.945) in the risk of T2 DM compared to allele T. The frequency of allele T was 75.4% in this study, which was between that of Chinese(59.5%) and European(80.1%) in the Hap Map. The distribution of genotype and allele frequency in rs564398 were not significantly different between T2 DM group and control group(P>0.05). The OR value of the T2 DM risk were 0.818(P=0.037, 95% CI: 0.677~0.988) and 0.676(P=0.039, 95% CI: 0.466~0.980) in the co-dominant model of rs10811661, when genotype C/T, C/C compared to genotype T/T; in additive model, the OR value of the T2 DM risk was 0.820(P=0.007, 95% CI: 0.709~0.948) when each additional allele C; in the dominant model, compared to those without allele C, the OR value of T2 DM risk in the people with allele C was 0.794(P=0.012, 95% CI: 0.663~0.951). The genetic model was not statistically different after adjustment for covariates. There was no obvious linkage disequilibrium between rs10811661 and rs564398 in CDKN2A/2B gene(D`=0.267, r2=0.007). 4) The frequencies of genotype and allele in CDKAL1 gene rs10498701, rs10946398, rs2305955, rs34525680, rs4710944, rs4712556, rs6456397, rs7741604, rs7754840, rs9368222 were not significantly different between T2 DM group and control group(P>0.05). After adjustment for the covariates, the risk of T2 DM in the people carrying genotype C/C was 1.476 times than those carrying genotype A/A+A/C in the recessive model of rs10946398(P=0.016, 95% CI: 1.076~2.024); the risk of T2 DM in the people carrying genotype C/C was 1.464 times than those carrying genotype G/G+C/G in the recessive model of rs7754840(P=0.017, 95% CI: 1.070~2.001). The risk of T2 DM in the people carrying genotype A/A was 1.422 times than those carrying genotype C/C in the co-dominant model of rs9368222(P=0.044, 95% CI: 1.009~2.004); the risk of T2 DM in the people carrying genotype A/A was 1.479 times than those carrying the genotype A/C+C/C in the recessive model of rs9368222(P=0.018, 95% CI: 1.069~2.045). Between rs10946398 and rs7754840(D`=0.997, r2=0.991), rs10946398 and rs9368222(D`=0.926, r2=0.809), rs9368222 and rs7754840(D`=0.930, r2=0.809) in CDKAL1 gene had strong linkage disequilibrium; the distribution of haplotype CCA, CCC, AGC, AGA were not significantly different between T2 DM group and control group(P > 0.05). 5) The distribution of genotype frequency in FTO rs7195539 between T2 DM group and control group was statistically different(P=0.045), and the allele in the distribution had no statistical difference between the two groups(P=0.470). The distribution of genotype and allele frequency in FTO rs8050136 between T2 DM group and control group were statistically different(P=0.018 and 0.005), the risk of T2 DM in people who carrying allele A was 1.220 times than those carrying allele C(P=0.005, 95% CI: 1.063~1.401), and the frequency of allele A was 29.7%, which was between that of Chinese(13.9%) and European(46.0%) in the Hap Map. The distribution of the genotype and allele in FTO rs9939609 between T2 DM group and control group were statistically different(P=0.006 and 0.001), the risk of T2 DM in people who carrying allele A was 1.255 times than those carrying allele T(P=0.001, 95% CI: 1.095~1.438), and the frequency of allele A was 31.8%, which was between that of Chinese(15.0%) and European(46.0%) in the Hap Map. The frequencies of genotype and allele in FTO rs1421091, rs17218700, rs2665272, rs2689264, rs7203521, rs9934504, rs9940629 between T2 DM group and control group had no significant difference(P>0.05). After adjustment for covariates, including BMI, the risk of T2 DM in the people carrying genotype A/A was 1.486 times than those carrying genotype C/C in the co-dominant model of rs8050136(P=0.043, 95% CI: 1.013~2.180); the risk of T2 DM in the people carrying genotype A/A was 1.470 times than those carrying the genotype A/C+C/C in the recessive model of rs8050136(P=0.041, 95% CI: 1.016~2.126).The risk of T2 DM in the people carrying genotype A/A was 1.641 times than those carrying genotype T/T in the co-dominant model of rs9939609(P=0.006, 95% CI: 1.156~2.330); as each additional alleles A, the risk of T2 DM increased by 1.192 times in additive model of rs9939609(P=0.029, 95% CI: 1.018~1.395); the risk of T2 DM in the people carrying genotype A/A was 1.627 times than those carrying the genotype T/T+A/T in the recessive model of rs9939609(P=0.004, 95% CI: 1.164~22.274). A strong linkage disequilibrium existed between FTO gene rs8050136 and rs9939609(D`=0.971, r2=0.831), and further to build the haplotype including CT, CA and AA, haplotype CT and AA in the distribution between the two groups were statistically different, the former P=0.003, OR=0.809(95% CI: 0.705~0.928), the latter P=0.006, OR=1.219(95% CI: 1.057~1.404). 6) In the crossover analysis, between CDKN2A/2B(rs10811661), CDKAL1(rs10946398, rs7754840, rs9368222), and FTO(rs8050136, rs9939609), there were no significant interactions in the additive model and multiplication model(P>0.05). The interaction in multiplication model between CDKN2A/2B-rs10811661 and coronary heart disease was statistically different(P=0.035, ORint=2.130). The P value in interaction of CDKAL1-rs10946398, CDKAL1-rs7754840 and hypertension in the multiplication model were 0.023(ORint=2.312) and 0.016(ORint=2.432), respectively. FTO-rs8050136, FTO-rs9939609 and BMI in the additive model and multiplicative model were interactive, the P value of the additive model were 0.009(S=3.327) and 0.004(S=3.603), the P value of the multiplicative model were 0.038 and 0.032 respectively. 7) In the GMDR analysis, the third-order interaction of CDKN2A/2B, CDKAL1 and FTO: CDKAL1(rs7754840)-CDKN2A/2B(rs10811661)-FTO(rs9939609) was the best model, whose testing balance accuracy(0.5322) was maximum, and cross validation consistency was 8/10(P=0.011). The gene-environment interaction of CDKN2A/2B, CDKAL1, FTO gene indicated statistical differences(P=0.001) in the first-order, the second-order, the third-order, the fourth-order, the five-order and the six-order interaction. The best model was the fourth-order: hypertension-coronary heart disease-family history of DM-gout, whose testing balance accuracy(0.7108) and cross validation consistency(10/10) were the largest(P=0.001). The model including gene locus was the six-order interaction: CDKAL1(rs7754840)-CDKN2A/2B(rs10811661)-FTO(rs9939609)-hypertension-coronary heart disease-family history of DM, whose testing balance accuracy was 0.6882 and cross validation consistency was 5/10, meanwhile the model was statistically different(P=0.001). Conclusions: 1) In this study, the risk of T2 DM in urban population is lower than that of the rural population in Uyghur. Monthly income, hypertension, coronary heart disease, gout, family history of DM, the staple food of rice, waist circumference, urine protein may be independent risk factors for T2 DM in Uyghur. Degree of education, smoking history, eating more cakes and vegetables per week may be protection factors of T2 DM in Uyghur. 2) In this study, male patients with T2 DM have more obvious central obesity, and have higher risk of coronary heart disease, dyslipidemia, and have more unhealthy life habit than female patients in Uyghur. 3) Genotype T/T and allele T of CDKN2A/2B(rs10811661), genotype C/C of CDKAL1(rs10946398), genotype C/C of CDKAL1(rs7754840), genotype A/A of CDKAL1(rs9368222), genotype A/A and allele A of FTO(rs8050136), genotype A/A and allele A of FTO(rs9939609) may be associated with the risk of T2 DM in Uyghur. Among them, the correlation between genotype A/A of FTO(rs8050136 and rs9939609) and T2 DM exist independent of BMI in Uyghur. 4) Between CDKAL1 gene rs10946398 and rs10946398, rs7754840 and rs9368222, rs7754840 and rs9368222 have strong linkage disequilibrium, the haplotypes of those three loci may be not associated with the risk of T2 DM in Uyghur. Strong linkage disequilibrium exists between FTO gene rs8050136 and rs9939609, haplotype AA of FTO gene may increase and haplotype CT may reduce the risk of T2 DM in Uyghur. 5) The interaction of the three genes CDKN2A/2B(rs10811661), CDKAL1(rs10946398, rs7754840, rs9368222), FTO(rs8050136, rs9939609) are not found through crossover analysis. The crossover analysis indicates the interactions between CDKN2A/2B(rs10811661) and coronary heart disease, CDKAL1(rs10946398) and hypertension, CDKAL1(rs7754840) and hypertension, FTO(rs8050136) and BMI, FTO(rs9939609) and BMI, which may play a role in the pathogenesis of T2 DM in Uyghur. 6) GMDR analysis shows that gene-gene interaction may exist among CDKAL1(rs7754840)-CDKN2A/2B(rs10811661)-FTO(rs9939609), which may influence the pathogenesis of T2 DM in Uyghur. The six-order interaction of gene-environment ineraction: CDKAL1(rs7754840)-CDKN2A/2B(rs10811661)-FTO(rs9939609)-hypertension-coronary heart disease-family history of DM may play a role in the pathogenesis of T2 DM in Uyghur.