CTLA-4 49 A/G dimorphism and type 1 diabetes susceptibility: a French case-control study and segregation analysis. Evidence of a maternal effect.

Several studies have demonstrated an association of cytotoxic T lymphocyte-associated molecule 4 (CTLA-4) (IDDM 12) alanine 17 with type 1 diabetes, but we wished to study the parental effect of CTLA-4 49 A/G dimorphism in diabetic families. The CTLA-4 exon 1 polymorphism (49 A/G), HLA-DRB1 and insulin gene (INS) variable number tandem repeats (VNTR) were analysed in 134 type 1 diabetic patients vs. 273 control subjects. The segregation analysis for transmission was carried out in 70 informative diabetic families using the transmission distortion test (TDT). All genotyping was performed by PCR-RFLP. CTLA-4 49 G allele frequency was not increased in diabetic patients compared to controls (41 vs. 38%, not significant). The distribution of GG, AG and AA CTLA-4 genotypes was similar in the two groups (13, 57 and 30% vs. 11, 54 and 35%, respectively) and was independent of HLA-DRB1 or INS VNTR polymorphism. The CTLA-4 49 G allele showed weak distorted transmission to the diabetic offspring, whereas random transmission was observed in unaffected offspring. This distortion is attributable to a maternal effect (71% compared to the 50% expected ratio; tdt = 4.8; P < 0.03). The combined transmission of maternal CTLA-4 G with HLA-DRB1*03 (90%; tdt = 6.4; P < 0.01) and VNTR class I (80%; tdt = 5.4; P < 0.02) enhanced the susceptibility effect of each marker separately. We noted a slight CTLA-4 49 G and HLA-DRB1*04 distortion of transmission shared in paternal and maternal diabetic meiosis. In non-diabetic offspring, the CTLA-4 49 A allele confers a protective effect in the presence of maternal HLA-DRB1*03 and paternal HLA-DRB1*04 alleles. Despite the absence of a positive association of the CTLA-4 49 G allele with type 1 diabetes, our segregation analysis supports the hypothesis of a modulation by CTLA-4 49 G/A dimorphism of the susceptibility conferred by maternal HLA-DRB1*03 inheritance. This potential parental effect needs to be confirmed in a larger data set.