Studies on the involvement of the HLA complex in leprosy have consistently shown that this major histocompatibility region contains the major genes associated with susceptibility to the disease [39]. Many HLA class II molecules have been consistently found to be associated with different clinical phenotypes and disease per se. Studies of HLA class I initially did not consider the possibility that presentation by T-CD8+ lymphocytes could be important in leprosy. However, the positive association between HLA class I and leprosy reflects the role of these cells in the production of IFN-γ in response to M. leprae antigens [39].
Previous studies of different clinical forms of leprosy mainly included just the polar forms LL and TT. The consensus in the literature is that HLA-DR2 (subtypes HLA-DRB1*15 and HLA-DRB1*16) and HLA-DR3 are associated with the TT form [19-23,27,40] and HLA-DQB1*01 (subtypes HLA-DQB1*05 and HLA-DQB1*06) is associated with the LL form [25,26]. No previous studies included only the intermediate clinical forms of the disease (BT, BB, and BL). Therefore, the current study was proposed.
The results showed decreased frequencies of HLA-C*05 and HLA-DRB1*07 in B leprosy patients when compared to the healthy control individuals, suggesting that they are associated with protection against B leprosy. Analysis of HLA-B*49 showed a pc value that was nearly significant, suggesting a possible protective association. HLA-B*49 was found to be significantly associated with protection against leprosy per se in a Turkish population [41].
In a Brazilian population, HLA-C*05 is 1 of the 5 most frequent alleles at this locus (approximately 14.0% in our control group), and HLA-DRB1*07 is the most frequent (26.77% in our control group) [42]. In B leprosy patients, we observed decreased frequencies for these 2 alleles (5.94% for HLA-C*05 and 16.34% for HLA-DRB1*07). The protective effect of HLA-C*05 against in leprosy per se was recently described in a population in Rio de Janeiro, Brazil. However, in this study, the patients were not divided into groups according to their clinical forms; therefore, it was not possible to verify whether this allele was associated with B leprosy [43]. Rani et al. [24] reported a decreased frequency of HLA-DRB1*0701 in LL and BL patients compared to TT patients and healthy controls in a population from Northern India, and they assigned this allele a protective effect against multibacillary (MB) forms. HLA-DRB1*1501 and HLA-DQB1*0602 were associated with susceptibility in MB patients in Japan, and no allele was associated with protection in this population [34]. Silva et al. [32] reported that the HLA-DRB1*1601 allele was associated with susceptibility to B leprosy in a Brazilian population. In our study, no allele was associated with susceptibility.
Significant susceptibility and protective associations were observed in BB, BL, and BT, although the corrected p was higher than the significance level (5%). To confirm the associations in these clinical forms, the groups were compared 2 ways: BB, BL, and BT were compared with the healthy control group, and then each group (i.e., BB, BL, or BT) was compared to patients without the clinical form analyzed (see Additional file 2).
In the BB group, a protective association was observed for HLA-B*49 and HLA-C*05 when compared to the healthy controls. These same alleles were also found in B leprosy. In both analyses (compared to healthy controls and patient controls), HLA-B*58 and HLA-C*12 were associated with susceptibility to the BB form. HLA-C*12 was previously shown to be associated with susceptibility to leprosy per se in a population in Southeast Brazil [17]; however, no study has shown an association between HLA-B*58 and leprosy. Although our statistical analyses did not support susceptibility (pc ≥ 0.05), the finding of HLA-B*58 in our patient group suggests that this allele is a specific marker for clinical BB leprosy in our region, as the frequency of this allele in the general population is approximately 5.0%, whereas in BB patients, the frequency was 11.36% [42].
In BL patients, HLA-B*53 was associated with susceptibility when compared to the healthy controls; however, similar results were not observed when compared to the patient controls. The protective associations of HLA-DRB1*07 and HLA-DRB1*03 differed in the analyses with different control groups; HLA-DRB1*07 was significant when compared to the healthy controls, whereas HLA-DRB1*03 was significant when compared to the patient controls. Unlike the associations described above, HLA-DQB1*02 was associated with protection in BL patients when compared to both control groups, and a statistically significant association was maintained after Bonferroni correction.
Rani et al. [24] also reported a protective association for HLA-DQB1*02 as well as HLA-DRB1*07 in MB leprosy patients in India. Motta et al. [44] studied 89 leprosy patients (70 with MB and 19 with paucibacillary [PB]) and 112 healthy controls in Argentina (province of Chaco), and confirmed HLA-DQB1*02 as a protective factor against the MB form. In this study, they also showed that HLA-DRB1*04 was associated with protection against the PB form. However, in our study, this association was not found in B leprosy patients, suggesting that HLA-DQB1*02 is specific to the BL form or MB patients.
In the BT group, a protective association was found for HLA-B*40 when compared to both control groups, whereas HLA-B*49 and HLA-C*05 were only found to be protective when compared to the healthy control group. HLA-B*49 and HLA-C*05 were also associated with protection against B leprosy. HLA-B*53 and HLA-C*16 were associated with susceptibility only when compared to the healthy controls, whereas HLA-A*33 and HLA-DRB1*03 were associated with susceptibility when compared to both control groups. Kim et al. [28] showed that HLA-A*33 was shown associated with susceptibility to leprosy per se in Korea.
Studies [22,40,45] of different populations have demonstrated an association between HLA-DRB1*03 and predisposition to TT. In our study, the association of this allele with BT suggests that it has an effect on susceptibility to this clinical form, as this same allele was found to be associated with protection against the BL form.
These results suggest that HLA-DRB1*03 is associated with polarization to the B form and manifestation of this clinical form. Until now, this is a controversial point in the diagnosis of leprosy [3,4]. To better understand and confirm this result, the frequency of this allele in the polar forms of leprosy (LL and TT) should be determined, and these results should be replicated in other independent populations.
Due the small number of patients with ENL, they could not be studied. Therefore, only patients with reversal reactions were included, and a significant association with HLA-B*15 was found. Although this result lost significance after statistical correction, this finding must be considered, as the frequency of this allele differs considerably between patients with reversal reactions and non-reactional patients. In patients with reversal reactions, the frequency of HLA-B*15 was approximately 29.0%, whereas in the non-reaction group, the frequency was 13.0%.
In the literature, there are few studies describing the association between HLA and leprosy reactions. Ottenhoff found an association between HLA-DR3 and reactions in an Ethiopian population; however, this study involved a specific group of patients, Mitsuda-positive BT leprosy patients with reversal reactions. HLA-DR3 was shown to be involved in the acquisition of cellular immunity during the reactional state, but not in predisposition to this form of the disease [46]. In our study, HLA-DR3 was not associated with reactions. In addition, a negative association with BL was observed, which reinforces the hypothesis that HLA-DR3 is associated with BT leprosy and polarization in B leprosy.
The haplotype analysis showed that haplotypes A*02-B*07-C*07-DRB1*15-DQB1*06 and A*02-B*40-C*03-DRB1*13-DQB1*06 were more frequent in B patients, and the difference between the frequencies in B leprosy patients and the controls was significant, suggesting a possible role in susceptibility to B leprosy. In Indian populations [15,16], MB leprosy was associated with haplotypes A*1102-B*4006-Cw*1502 and A*0203-B*4016-Cw*0703. Although the haplotypes are different, the B*40 allele found in these Indian populations was also detected in our population. However, it is not possible to say whether this allele genotype is the same as that observed in our population. In the Rio de Janeiro population, the haplotypes associated with susceptibility to leprosy per se were A*02-B*35-C*04, A*02-B*15-C*03, A*30-B*42-C*17, A*03-B*35-C*04, A*26-B*38-C*12, and A*24-B*15-C*03, which differ from those found in our population [43].
Interestingly, the class II alleles present in both haplotypes found in our population, were previously shown to be associated with the clinical forms TT (DRB1*15) and LL (DQB1*06). This observation caused us to examine these class II alleles in B patients in greater detail. When analyzed independently, the results showed that haplotype HLA-DRB1*16/DQB1*05 was associated with susceptibility.
Analysis of all possible haplotypes together (DRB1*15/DQB1*05, DRB1*15/DQB1*06, DRB1*16/DQB1*05, DRB1*16/DQB1*06, DRB1*03/DQB1*05, and DRB1*03/DQB1*06) showed a greater association with susceptibility. Thus, we hypothesized that the presence of alleles associated with the polar clinical forms LL and TT may induce the manifestation of B leprosy by triggering the intermediate immune response that is characteristic of this form of the disease.