Differences in the Contribution of the CTLA4 Gene to Susceptibility to Fulminant and Type 1A Diabetes in Japanese Patients

We examined 55 patients with fulminant type 1 diabetes (49% female, median age at onset 35.0 years), 91 patients with classic type 1A diabetes (57% female, median age at onset 17.0 years), and 369 healthy control subjects. Diagnostic criteria for fulminant type 1 diabetes are described elsewhere (1). The criteria for the recruitment of type 1A diabetic patients were presence of diabetic ketosis at onset, duration of hyperglycemic symptoms <3 months prior to initiation of insulin therapy, and positive for at least one of the anti-islet autoantibodies. This study was approved by the appropriate ethics committees, and informed consent was obtained from all subjects.

Genotyping of two single nucleotide polymorphisms in the CTLA4 gene, +49G>A (rs231775) and CT60 (rs3087243), was performed as reported previously (5). Serum concentration of sCTLA4 was measured by enzyme-linked immunosorbent assay using human soluble CTLA4 (sCTLA4) kit (MedSystems Diagnostics, Vienna, Austria), according to the manufacturer's protocol. Sera from type 1 diabetic patients were obtained at disease onset and stored at −20°C until use.

The significance of differences in the distribution of genotypes between case and control subjects was determined by χ² test or Fisher's exact probability test. Comparisons of the sCTLA4 levels were made by ANOVA with phenotypic group alone and ANOVA with phenotypic group and CTLA4 genotype. P < 0.05 was considered to be statistically significant.

The +49G>A variant was associated with classic type 1A diabetes but not with fulminant type 1 diabetes (Table 1). In contrast, the contribution of CT60 to disease is distinct from that of +49G>A. The frequency of the CT60AA genotype in fulminant type 1 diabetic patients was significantly higher than in control subjects (P = 0.021) and type 1A diabetic patients (P = 0.031). CT60GG was associated with type 1A diabetes (P = 0.008). Because of the strong association of HLA-DR4 in both patient groups (1), the effect of CTLA4 on type 1 diabetes susceptibility relative to HLA-DR4 was also examined. Among DR4-positive individuals, the frequency of the CT60AA genotype was significantly increased in patients with fulminant type 1 diabetes (P = 0.005). However, stratification of patients by the presence or absence of HLA-DR4 did not affect the association between the +49GG genotype and type 1A diabetes (Table 1).

It has been reported that the CT60G allele might be associated with lower production of sCTLA4 mRNA (3). We therefore determined serum sCTLA4 levels. The mean sCTLA4 levels in fulminant type 1 diabetic patients (0.56 ± 0.24 ng/ml [mean ± SD], n = 36) was significantly lower than those in type 1A diabetic patients (0.94 ± 0.87 ng/ml, n = 45) and control subjects (0.89 ± 0.76 ng/ml, n = 23) (P = 0.043). A mixed-model ANOVA using phenotypic group (fulminant type 1 diabetes, type 1A diabetes, and control) and CT60 genotype (GG and GA+AA) as factorial fixed effects revealed no differences in sCTLA4 levels between CT60 genotypes (P = 0.76) or phenotype/genotype interactions (P = 0.40).

CTLA4, which delivers inhibitory signals to T-cell activation, is expressed on the surface of activated T-cells and regulatory T-cells, and the lack of CTLA4 results in uncontrolled T-cell–mediated lymphoproliferative disease (6). Furthermore, CTLA4 also has a significant biological role in attenuating T-cell responses in the context of an inflammatory environment, such as infection with a pathogen (7). We showed that CTLA4 is associated with an increased risk of fulminant type 1 diabetes and that its contribution is distinct from classic type 1A diabetes. As reported previously (5,8), a significant association between classic type 1A diabetes and +49GG and CT60GG genotype was also found in the present study. However, the CT60AA genotype contributes to the susceptibility to fulminant type 1 diabetes. Moreover, it is implicated that HLA-DR4 influences the association of fulminant type 1 diabetes with the CT60AA genotype.

In this study, we also revealed that serum sCTLA4 level in fulminant type 1 diabetic patients were significantly lower than those in type 1A diabetic patients and control subjects. Although it remains unknown how sCTLA4 regulates T-cell activation, recombinant sCTLA4 inhibits T-cell proliferation in vitro. Furthermore, sCTLA4 is constitutively expressed in nonstimulated T-cells, and its expression is downregulated after T-cell activation (9). Therefore, the decreased levels of sCTLA4 might indicate a lower potential of T-cell inhibition in fulminant type 1 diabetes, which might be caused by functional defects leading to reduced production of sCTLA4 or ongoing activation of the immune system eventually leading to decreased levels of sCTLA4.

In conclusion, the present study implicates that CTLA4 confers susceptibility to fulminant type 1 diabetes. Furthermore, the different contributions of CTLA4 to susceptibility to fulminant and classic type 1A diabetes indicate that the underlying immune process–primed β-cell injury might be distinct between these subtypes of type 1 diabetes.

This study was partly supported by a grant from the Ministry of Education, Culture, Science, Sports and Technology of Japan.

We thank Shinobu Mitsui for excellent technical assistance.