Objective: To explore the relationship between the 3 islet autoantibodies, i.e., glutamic acid decarboxylase antibody (GADA), protein-tyrosine-phosphatase-2 autoantibody (IA-2A), and insulin autoantibody (IAA), and human leukocyte HLA-DQ genotypes in the patients with type 1 diabetes mellitus (T1DM).
Methods: Peripheral blood samples were collected from 495 T1DM patients and 376 healthy controls. Radioligand assay was used to detect the levels of GADA and IA-2A. Seventy-one of the 495 T1DM patients with insulin treatment less than 14 days underwent radioligand assay to measure the IAA level. 338 TIDM patients, including 187 antibodies-positive and 151 antibodies-negative patients, and 278 healthy controls underwent PCR and sequencing so as to examine the genetic polymorphism of HLA-DQ.
Results: (1) The frequencies of DQA1*03-DQB1 *0303, DQA1*05-DQB1*0201, and DQA1*03-DQB1*0401 haplotypes in the T1DM patients were 32.6%, 14.1% and 10.2% respectively, all significantly higher than those in the controls (21.9%, 3.5%, and 2.9% respectively, all P < 0.01), however, the frequency of DQA1*0102-DQB1*0602 haplotype in the T1DM patients was 1.7%, significantly lower than that in the controls (5.3%, P < 0.05). (2) The GADA positive rates of in the T;DM patients with DQA1*05-DQB1*0201 or DQA1*03-DQB1*0401 haplotype were 55.8% and 65.5% respectively, both significantly higher than those of the T1DM patients without DQA1*05-DQB1*0201 or DQA1*03-DQB1*0401 haplotype (41.0% and 40.3%, respectively P < 0.05 or P < 0.01), and the IA-2A positive rate of the T1DM patients with DQA1*03-DQB1*0303 haplotype was 27.0%, significantly higher than that of the T1DM patients without DQA1*03-DQB1*0303 haplotype (7.9%, P < 0.01), there were no significant differences in GADA and IA-2A positivity between the T1DM patients with and without DQA1*03-DQB1*0302 haplotype (48.5% vs 43.9% and 24.2% vs 15.4%, both P > 0.05). The GADA positivity rate of the T1DM patients with DQA1*0102-DQB1 *0602 haplotype was 16.7%, significantly lower than those patients without DQA1*0102-DQB1 *0602 haplotype (45.9%, P < 0.05). There was no significant difference in IAA level between those patients with and without susceptible DQ alleles and haplotypes (P > 0.05).
Conclusion: GADA is associated with HLA-DQA1*05-DQB1*0201 and DQA1*03-DQB1*0401 haplotypes, and IA-2A is associated with DQA1*03-DQB1 *0303 haplotype.