Role of HLA class II alleles in susceptibility to and protection from localized cutaneous leishmaniasis

Abstract

Localized cutaneous leishmaniasis (LCL) is the prevalent form of leishmaniasis in Mexico. It is limited to the skin; reversible upon treatment and the host cellular immune response is intact. Several genes that influence the expression of LCL have been described in the mouse. In humans, we, as well as others, have demonstrated that HLA-DQ3 antigens seem to play some role in host susceptibility. We therefore analyzed at the DNA level, the class II loci of the same patients that were previously studied by serology. The purpose of this study was to assess the contribution of HLA DR, DQ, and DP genes in the protection and/or the susceptibility to LCL. Sixty-five patients with LCL from Comalcalco, state of Tabasco, were recruited and 100 healthy controls were included for comparison. All were Mexican Mestizos. DRB1, DQA1, DQB1, DPA1, and DPB1 alleles were typed using two different methods: PCR-SSO and PCR-SSP. Results indicate that class II genes are relevant for the expression of LCL and several loci contribute independently and sinergically. DRB1*0407 participates in susceptibility with an etiological fraction (EF) of 20% and an odds ratio (OR) of 2.92. Two additional susceptibility genes were found. These are located to the DP locus: DPA1*0401 (OR = 10.07; EF=7%) and DPB1*0101 (OR = 5.99 EF = 13%). Resistance was found associated to DPB1*0401, thus *0401 “motif” could be an ideal candidate for the development of a vaccine. DR2 (DRB1*1500+DRB1*1600) has also a significant p for protection, suggesting that the sequence common to this group of antigens may anchor parasite peptides which trigger a protective response.

Introduction

Leishmaniasis is a parasitic disease caused by a trypanosomatid protozoan belonging to the genus Leishmania [1]. The disease is transmitted by a sandfly, which bites a vertebrate host, and the parasite infects cells of the phagocytic mononuclear lineage, where it lives and multiplies. Leishmanial infections are characterized by a broad spectrum of diseases that may affect the skin, or may disseminate to mucocutaneous or visceral tissues. The clinical manifestation of the disease varies (localized cutaneous [LCL], diffuse cutaneous [DCL], mucocutaneous [MCL], or visceral [VL]), depending partly of the parasite species and of the host immune response. The etiologic agents of LCL are Leishmania tropica, L. mexicana, L. major, and L. aethiopica, MCL is caused by L. braziliensis, whereas L. donovani produces VL or kala-azar, a chronic and often fatal disease [2].

LCL is characterized by the presence of a strong cellular immune response against the parasite that can be detected in vivo and in vitro. The in vivo response is measured by a delayed type hypersensitivity (DTH) skin reaction to whole killed parasite preparations before healing occurs and, once positive, it remains so for many years 3, 4. The resolution of the disease is correlated with a Th1 response 5, 6. In contrast, DCL is a progressive, anergic, nonulcerative form of the disease, which is accompanied by a defective cellular immune response, correlating with the development of a Th2 type response 4, 6. In Mexico, the most common form of the disease is LCL, although there are a few cases of DCL coexisting in the same endemic area; the species causing the two forms seems to be Leishmania mexicana [7].

Genetic analysis of phenotypes of the disease using inbred, recombinant inbred and backcross mice strains, as well as cutaneous species of Leishmania (L. major, L. mexicana) have permitted the mapping of susceptibility genes into five different genomic regions, named Scl-2, Scl-1, Th2 “cluster”, Nramp and major histocompatibility complex (MHC) 8, 9, 10, 11, 12, 13. These regions have also been mapped in the human genome and the analysis of candidate genes are under study. Using congenic mice strains, it has been demonstrated that the mouse H-2 genes participate in the regulation of the latter phases of all of leishmanial infections, but the most dramatic effect is seen with L. donovani [9]. The use of anti-I-A and anti-I-E antibodies in vivo (DQ and DR in humans, respectively) [14], and the analysis of I-E transgenic mice [15], demonstrated that class II polymorphism is responsible of the differences in the disease phenotype, since the I-E^d/k haplotypes are associated with nonhealing responses. In humans, several studies have been done using serological methods to look for antigens participating in disease expression, the analysis of HLA class I and class II molecules revealed a decrease in the frequency of HLA-Cw7 in New Guinea LCL patients [16]. In distinct Latin American populations different associations have been shown; a significant decrease of DR2 in Brazilian MCL patients, as well as an increased frequency of HLA-DQ3 [17]. In a recent study with Mediterranean VL patients, DR2 was also found as a tendency towards protection although Bonferoni correction showed no significance [18]. In Venezuelan families the HLA-Bw22 DRw11-DQw7 haplotype was found significantly more frequent in affected compared to healthy siblings [19]. In a previous study done by some of us with the same Mexican Mestizo patients using class I and class II serological typing, DQ3 was found significantly increased and DPw4 was shown to be significantly decreased in the LCL patients [20]. Thus, a susceptibility gene associated with DQ region and a possible putative gene in or close to DPB1 locus was suggested. The goal of the present study was to define the contribution of DRB1, DQA1, DQB1, DPA1 and DPB1 loci in Mexican patients with LCL.