Department of Endocrinology and Diabetes, Faculty of Medicine, Saitama Medical University, Saitama, Japan.
Diabetes Metab Res Rev. 2011 Nov;27(8):844-8. doi: 10.1002/dmrr.1234.
We previously reported the associations of human leukocyte antigen (HLA) (DRB1 and DQB1), INS, CTLA4, IL2RA, ERBB3 and CLEC16A with Japanese type 1 diabetes (T1D). In this study, we jointly analysed these loci in addition to IFIH1 and IL7R.
A maximum of 790 T1D patients and 953 control subjects were analysed. HLA was determined by sequencing-based typing. Seven non-HLA single nucleotide polymorphisms were genotyped using TaqMan assay.
HLA DRB10405, DRB10901 and DRB10802-DQB10302 haplotypes were positively associated with T1D, while the DRB115 haplotypes were negatively associated. Non-HLA single nucleotide polymorphisms, INS, IL2RA, ERBB3, CLEC16A and IL7R were associated with T1D. By a prediction model using the HLA loci alone (HLA model) or the non-HLA loci alone (non-HLA model), it was revealed that the cumulative effect of the non-HLA model was much weaker than that of the HLA model (average increase in odds ratio: 1.17 versus 3.14). Furthermore, the area under the receiver operating characteristic curve of the non-HLA model was also much smaller than that of the HLA model (0.65 versus 0.81, p<10(-11)). Finally, a patient-only analysis revealed the susceptible HLA haplotypes and the risk allele of INS to be negatively associated with slower onset of the disease. In addition, the DRB10901 haplotype and the risk alleles of ERBB3, CLEC16A and CTLA4 were positively associated with the co-occurrence of thyroid autoimmunity.
Although several non-HLA susceptibility genes in Japanese were confirmed trans-racially and appear to contribute to the heterogeneity of the clinical phenotypes, the cumulative effect on the ability to predict the development of T1D was weak.
我们之前报道了人类白细胞抗原（HLA）（DRB1 和 DQB1）、INS、CTLA4、IL2RA、ERBB3 和 CLEC16A 与日本 1 型糖尿病（T1D）之间的关联。在这项研究中，我们除了 IFIH1 和 IL7R 之外，还联合分析了这些基因座。
分析了最多 790 名 T1D 患者和 953 名对照者。HLA 通过测序分型确定。使用 TaqMan 测定法对 7 个非 HLA 单核苷酸多态性进行基因分型。
HLA-DRB10405、DRB10901 和 DRB10802-DQB10302 单体型与 T1D 呈正相关，而 DRB115 单体型与 T1D 呈负相关。非 HLA 单核苷酸多态性、INS、IL2RA、ERBB3、CLEC16A 和 IL7R 与 T1D 相关。通过仅使用 HLA 基因座（HLA 模型）或仅使用非 HLA 基因座（非 HLA 模型）的预测模型表明，非 HLA 模型的累积效应远弱于 HLA 模型（平均优势比增加：1.17 对 3.14）。此外，非 HLA 模型的受试者工作特征曲线下面积也远小于 HLA 模型（0.65 对 0.81，p<10(-11)）。最后，仅患者分析显示易感 HLA 单体型和 INS 的风险等位基因与疾病发病较慢呈负相关。此外，DRB10901 单体型和 ERBB3、CLEC16A 和 CTLA4 的风险等位基因与甲状腺自身免疫的同时发生呈正相关。
尽管在日本已确认了几个跨种族的非 HLA 易感基因，并且这些基因似乎对临床表型的异质性有贡献，但对预测 T1D 发生的累积效应较弱。
