These results imply that the real disease susceptibility gene involved in the development of Behçet's disease is the HLA-B*51 allele itself and that the significant increase of the MIC-A009 allele in the patient group results secondarily from a strong linkage disequilibrium with HLA-B*51.

Recently a new family of non-classical MHC molecules, the MHC class I chain-related protein (MIC), encoded by genes located in the major histocompatability complex have been identified. On the basis of the location of MIC genes and the structure and expression of MIC molecules it has been postulated that MIC may be a susceptibility factor in Behçet's disease (BD). We investigated the association of the 16 described external domain alleles and the transmembrane triplet repeats of MIC-A with BD in a Middle Eastern population. DNA from ninety-five patients and 102 age- and sex-matched controls were analyzed by polymerase chain reaction using allele specific primers. Our results show an increase of MIC-A*009 in the BD patient group 44/95 (46%) compared with controls 24/102 (24%) (chi2=11.3, OR=2.8, P=0.00078). MIC-A*009 was also found to be strongly associated with HLA-B51 in the patients 39/44 (88%) when compared with controls 10/24 (42%) (chi2=4, P=0.04). MIC-A*009 was also found in linkage disequilibrium with HLA-B52, but only in controls. The A6 form of a MIC-A transmembrane triplet repeat was found to be significantly raised in the patients (80/95; 84%;) compared with controls (58/102, 57%) (chi2=17.5, OR=4, P=0.000028). Although the MIC-A associations described are highly significant, the association with HLA-B51 independently remains the most significant factor (chi2=56.8, P<10(-6)). The data suggests that as both MIC-A*009 and A6 are in strong linkage disequilibrium with HLA-B51, they are unlikely to be the susceptibility gene for BD but may be markers for additional risk factors.

Behçet's disease (BD) is known to be associated with human leukocyte antigen (HLA) B51 in many different ethnic groups. An increased incidence of HLA-B51 in the patient group has also been reported in a Japanese population. Recently, the B51 antigen has been identified to comprise 21 alleles, B*5101-B*5121. Further, not only HLA-B*5101 but also HLA-B*5108 were found to be relatively increased in the patient groups among Italian and Saudi Arabian populations. Therefore, we performed HLA-B*51 allele genotyping by the polymerase chain reaction-sequencing based typing (PCR-SBT) method in order to investigate whether there is any correlation of one particular B51-associated allele with Japanese BD. Ninety-six Japanese patients with BD and 132 healthy Japanese volunteers were enrolled in this study. As a result, the phenotype frequency of the B51 antigen was confirmed to be remarkably increased in the patient group as compared to the ethnically matched control group (59.4% in patients vs. 13.6% in controls; Pc=0.0000000000098, R.R.=9.3). In the B*51 allele genotyping, 56 out of 57 B51-positive patients were defined as B*5101 and the remaining one was B*5102. In contrast, all of 18 B51-positive normal controls were B*5101. None of the Japanese patients and healthy controls carried the HLA-B*5108 allele. This study revealed that B*51 allelic distribution in Japanese was different from those in Italian and Saudi Arabian populations, and that the significantly high incidence of the HLA-B51 antigen in the Japanese BD patient group was mostly caused by the significant increase of the HLA-B*5101 allele.