Association Of Zinc Transporter8Autoantibodies With HLA-DR-DQ And IFIH1Gene Polymorphisms In Type1Diabetes

Part I Polymorphisms of HLA-DR-DQ genes in classical type1diabetes from Hunan Han populationObjective:To investigate the susceptible and protective HLA-DR-DQ genotypes and haplotypes in classical type1diabetes (T1DM) from Hunan Han population.Methods:All subjects including558classical T1DM patients and930healthy controls were genotyped at HLA-DRB1, DQA1and DQB1locus by PCR-direct sequencing. The HLA-DR-DQ haplotypes were constructed by the PHASE programme. Frequencies of genotypes and haplotypes between patients and controls were compared by chi square test.Results:①The frequencies of nine DR and DQ alleles in T1DM patients were were significantly higher than those in control subjects.The top three alleles with the highest odds ratio (OR) were DRB1*0301(21.37%vs.5.78%, OR=4.43, Note:cases vs. controls, Pc<0.001, the same as below), DRB1*0401(2.25%vs.0.60%, OR=3.84), DQBl*0201(23.46%vs.7.4%, OR=3.84). The frequencies of sixteen alleles were significantly reduced in TlDM patients. The top three alleles with the strongest protection were DRB1*1301(0.41%vs.2.45%, OR=0.16), DRB1*0406(0.51%vs.2.92%, OR=0.17), DQBl*0602(1.03%vs.4.79%, OR=0.21).②The frequencies of thirteen DR and DQ genotypes in TlDM patients were were significantly higher than those in control subjects. The top three genotypes with the highest OR were DQA1*03/03(39.76%vs.1.81%, OR=35.83), DQA1*03/05(21.67%vs.1.03%, OR=26.49), DRB1*0405/0701(1.84%vs.0.12%, OR=15.71). The frequencies of seven genotypes were significantly reduced in TlDM patients. The top three genotypes with the strongest protection were DRB1*0803/0803(0.61%vs.5.13%, OR=0.11), DQA1*0102/0102(0.60%vs.3.88%, OR=0.15) and DQA1*010X/0102(010X=0101or0104;0.80%vs.4.91%, OR=0.16).③The frequencies of nineteen DR-DQ haplotypes in TlDM patients were were significantly higher than those in control subjects.The top three haplotypes with the highest OR were DRB1*0901-DQA1*03(30.73%vs.0.22%, OR=203.07), DRB1*0301-DQA1*03(6.19%vs.0.07%, OR=90.77), DQA1*03-DQB1*0201(3.52%vs.0.07%, OR=51.82). The frequencies of twenty-two haplotypes were significantly reduced in TlDM patients. The top three haplotypes with the strongest protection were DRB1*0803-DQA1*03-DQB1*0601(O.12%vs.3.95％,OR=0.03),DRB1*0901-DQA1*0102(0.11%vs.1.96%,OR=0.06),DQA1*03-DQB1*0601(O.32%vs.4.57%,OR=0.07).④There are similarties and differences about HLA-DR-DQ risk markers among Chinese Han,Caucasian,Japanese and Korean.The common genes in these populations are susceptible DQB1*0303,DRB1*0405and protective DQB1*0301, DQB1*0601, DQB1*0602.One protective allele(DQB1*0502),four susceptible haplotypes(DRB1*0301-DQB1*0303, DRB1*0301-DQA1*03-DQB1*0201, DRB1*0301-DQA1*03-DQB1*0303,DRB1*0901-DQA1*05-DQB1*0201) and three protective haplotypes(DRB1*0406-DQB1*0302,DRB1*0803-DQA1*03-DQB1*0601,DRB1*1202-DQA1*060Y-DQB1*0301) have been identified for Chinese Han specific DR-DQ alleles or haplotypes that confer risk of TlDM.Conclusions:Eight Chinese-specific DR-DQ risk markers are identified and these results reveal significant genetic heterogeneity of the contribution of HLA-DR-DQ genes to susceptibility to TlDM. Part II Association of zinc transporter8autoantibodies with HLA-DR-DQ gene polymorphismsObjective:To explore the relationship between zinc transporter8autoantibodies (ZnT8A) and susceptible HLA-DR-DQ genes in the patients with classical type1diabetes (TlDM).Methods:Four hundred twenty two TlDM patients with ZnT8A, glutamic acid carboxylase antibody(GADA) or protein tyrosine phosphatase antibody (IA-2A) positive were selected. All autoantibodies were detected by radioligand assay. These patients were also genotyped at HLA-DRBl, DQAl and DQBl locus by PCR-direct sequencing. Association of ZnT8A with HLA-DR-DQ susceptible genes was analyzed by bivariate correlation test or chi square test.Results:①The prevalence of ZnT8A+was36.3%(153/422) in TIDM patients and the prevalence of the ZnT8A+alone was6.8%(28/412). The overlapping prevalence were25.1%(104/415),18.2%(75/412) for ZnT8A+and GADA+, ZnT8A+and IA-2A+, respectively.15%(62/412) of them were positive for three autoantibodies combined.②The levels of ZnT8A were positively correlated with suscepitable allele DQA1*05(Υ2=0.126, P=0.012) and genotype DQA1*03/05(Υ2=0.138, P=0.006). And ZnT8A levels were negatively associated with suscepitable genotypes DQB1*0303/0303(Υ2=-0.142, P=0.006) and DRB1*0901/0901(Υ2=-0.114,p=0.029).③The prevalence of ZnT8A+in the patients with DQA1*03/05(45.8%vs.32.7%)were significantly higher than those of non-carriers(P<0.05).The prevalence of ZnT8A+in the patients with DQB1*0303/0303(25.0%vs.38.5%),DRB1*0901/0901(26.6%vs.39.2%), DQA1*0102-DQB1*0601(12.5%vs.37.2%)and DRB1*0901-DQA1*03(29.6%vs.41.6%) were significantly lower than those of non-carrers,respectively(all P<0.05).④The prevalence of ZnT8A+alone in patients with susceptible DQA1*03/060Y(75.00%vs.16.67%).DQB1*0301/0301(44.44%vs.17.14%),DRB1*1401-DQA1*010X(80.00%vs.17.12%) were significantly higher than those of non-carriers,respectively(all P<0.05).The prevalence of ZnT8A+alone in patients with protective DQA1*03/05(6.98%vs.16.67%). DRB1*0901-DQA1*03-DQB1*0303(4.88%vs.25.35％)were significantly lower than those of non-carriers(Both P<0.05).Conclusions:ZnT8-specific humoral autoimmunity is linked to DRB1*1401-DQA1*010X and DRB1*0901-DQA1*03-DQB1*0303haplotypes.These data provide further evidence that production of major autoantibody specificities are associated with HLA class II genes. Part III Analysis of IFIH1gene rs1990760polymorphism in type1diabetes from Hunan Han populationObjective:To investigate whether IFIH1gene rs1990760polymorphism has a role in genetic predisposition of type1diabetes (T1DM) and latent autoimmune diabetes in adults (LADA) in Hunan Han population.Methods:All subjects including930T1DM patients,434LADA patients and1010healthy controls were genotyped for IFIH1rs1990760polymorphism using the Custom TaqMan SNP genotyping assay. The frequencies of alleles and genotypes of rs1990760were compared between cases and controls and genotype-phenotype correlation were carried out.Results:①Allele C (79.8%vs.80.7%) and T (20.2%vs.19.3%) of IFIH1gene rs1990760polymorphism were similar in T1DM and control subjects, respectively (P>0.05). At the genotypic level, the frequencies of the CC (63.1%vs.65.0%), CT(33.5%vs.31.3%) and TT (3.5%vs.3.7%) genotypes were similar in cases and control subjects, respectively (all P>0.05).②There was no significant difference of allele C/T frequencies of the rs1990760polymorphism between LADA and control groups (80.6%vs.80.7%;19.4%vs.19.3%, respectively, both P>0.05). Similarly, no significant differences in the genotypic (CC, CT, TT) frequencies were found (all P>0.05).③The results did not show significant difference among genotypes in each subset including age of onset, body mass index, waist to hip, HbAlc, islet beta cell function and metablism components from TlDM and LADA patients (all P>0.05).④The prevalence and titer of ZnT8A, GADA and IA-2A among the CC, CT and TT genotypes of IFIHl gene rs1990760polymorphism in TlDM patients were not statistically significant (all P>0.05).⑤The frequency of allele C of rs1990760polymorphism in patients with susceptible allele DQB1*0303were lower than those of non-carrier. However, it was boundary statistical differences (76.76%vs.81.42%, P=0.048). The alleles and genotypes of IFIH1gene rs1990760polymorphism had no interaction with high risk or low risk HLA-DRBl,DQAl,DQBl genotypes (all P>0.05).Conclusions:These results suggest no contribution from IFIHl gene rs1990760polymorphism to the genetic predisposition of TIDM and LADA in Hunan Han population. There is no association of rs1990760polymorphism with clinical features, islet-associated autoimmunity as well as HLA-DR-DQ risk markers in TlDM. Part IV Association between IFIH1gene rs1990760polymorphism and type1diabetes:a Meta-analysisObjective:To explore comprehensively the relationship between IFIH1gene rs1990760polymorphism and susceptibility to type1diabetes (TlDM).Methods:References were retrieved through the computerized PubMed and Web of Knowledge.The number of allele or genotype of rs1990760polymorphism were taken in case-control studies.The unpublished data of our study was also included. The pools ORs with95%CI were calculated to assess the association strength. Either the fixed or random effect model was applied to conducted Meta-analysis in additive model, homozygote comparison, heterozygote comparison, dominant and recessive genetic models. Publication bias and sensitivity were evaluated at the same time. All analyses were conducted with Stata11.0software.Results:①A total of thirteen case-control studies were enrolled, including19932cases of TlDM and28606controls.The subjects enrolled in ten studied were European descent, others were Asian population.②The results of Meta-analyses showed the A allele was associated with the risk of TlDM in a additive model [A vs G:OR(95%CI)=1.14(1.08～1.22), P=0.000]; the AA+AG genotype was associated with the risk of TlDM in a dominant genetic model [AA+AG vs. GG: OR(95%CI)=1.25(1.16～1.34), P=0.000]; the AA genotype was associated with the risk of T1DM in homozygote comparison and recessive genetic models[AA vs. GG:OR(95%CI)=1.36(1.26～1.47), P=0.000; AA vs. AG+GG:OR(95%CI)=1.159(1.06～1.27), P=0.001]; the AG genotype was associated with the risk of T1DM in heterozygote comparison[AG vs. GG:OR(95%CI)=1.15(1.07～1.24), P=0.000].③After stratification according to European descent and Asian ethnicity, the variant allele and genotype of rs1990760was associated with the risk of T1DM from European descent in all genetic models [A vs. G: OR(95%CI)=1.18(1.11-1.26), P=0.000]. Conversely, there was no difference of those in Asian population [A vs. G:OR(95%CI)=0.97(0.86-1.08),P=0.921].④Sensitivity analysis showed that the aboved combined results could not be effected by individual study, no publication bias was found, the conclusions were reliable.Conclusions:The meta-analysis suggests that the IFIH1gene rs1990760polymorphism is a genetic risk factor for T1DM of European descent, and the A allele elevates the risk of T1DM. But it is not a definite risk of T1DM in Asian population.