HLA-C locus alleles in patients with psoriatic arthritis (PsA).

The aim of this investigation was to compare the frequency of alleles at the HLA-C locus in patients with psoriatic arthritis to that in the disease free population and to relate the HLA genes to disease phenotype in psoriatic arthritis (PsA). Ninety-four consecutive patients seen in the (PsA) clinic between April and July 1996 and 100 disease-free controls had HLA typing performed by both serologic and molecular techniques. Patients' disease was assessed according to a standard protocol. Fisher's exact test was used to compare the frequency of HLA alleles in the patients and controls. HLA-C determined by PCR-SSP decreased the frequency of non-identified, that is, "blank" (or null) alleles to 6% in patients and controls. HLA-Cw*0602 was present in 32 patients and 18 controls: allele frequency of 17% in the patients versus 9% in the controls (p < 0.01). Amongst patients with psoriatic arthritis, those who carried the HLA-Cw*0602 allele had a significantly earlier mean age of onset of their psoriasis (p = 0.003). This study confirms that molecular DNA techniques improve detection of C-locus alleles. The HLA-Cw*0602 is increased amongst patients with psoriatic arthritis compared to controls. The HLA-Cw*0602 is associated with an earlier age of onset of psoriasis.     

Polymorphism in MICA rather than HLA-B/C genes is associated with psoriatic arthritis in the Jewish population

The aim of this study was to examine whether the association of psoriatic arthritis (PsA) with human leukocyte antigen (HLA) class I genes is secondary to linkage disequilibrium with a nearby gene. We examined a sample of the Jewish population to investigate whether HLA-B/C and DR polymorphism is associated with susceptibility, or whether other closely related class I loci, such as the major histocompatibility complex class I chain-related gene A (MICA) and tumor necrosis factor (TNF), might play a role in disease development. Comparisons of different populations with different HLA profiles would be of value in identifying the candidate genes involved in PSA. Fifty-two patients with PsA and 73 random matched controls from a Jewish population were selected and DNA typed by polymerase chain reaction-single-strand oligonucleotide probe (PCR-SSOP) (HLA-C), PCR sequence-specific primers (PCR-SSP) (HLA-B, -DR), radioactive PCR (MICA-TM polymorphism in the transmembrane region), and PCR-RFLP (TNF). Some findings can be concluded from the study: (1) the frequency of HLA-B*5701, B*3801, B*39, B*27, Cw*0602, Cw*07, DRB1*0402, and DRB1*0701 were not found to be significantly increased in PsA; (2) no significant differences of TNFalpha promoter alleles at positions -308 and -238 were found between PsA and healthy controls; (3) the trinucleotide repeat polymorphism MICA-A9 was present at a higher frequency in PsA patients, (p(c) < 0.009, RR = 3.34, EF = 0.39); and (4) MICA-A9 polymorphism was found in linkage disequilibrium with HLA-B alleles (B*5701, B*3801) described to be associated with PsA in Caucasians. These results suggest that the MICA gene or other nearby gene(s) may be involved in the development of PsA, and it would thus appear that psoriasis vulgaris (PsV) and PsA are associated with different MHC susceptibility genes.     

Diversity of MICA and linkage disequilibrium with HLA-B in two North American populations

The MICA gene has a high degree of polymorphism. Allelic variation of MICA may influence binding of these ligands to the NK cell receptor NKG2D and may affect organ transplantation and/or disease pathogenesis. Knowledge of the population distribution of MICA alleles and their linkage disequilibrium (LD) with class I human leukocyte antigen (HLA) will enhance our understanding of the potential functional significance of the MICA polymorphism. In the present study, we characterized the MICA and HLA-B polymorphisms in two North American populations: European and African. The individual racial groups showed rather limited variation at the MICA locus, where the same set of three most common alleles, MICA*00201, *004, and *00801, account for 64 and 71% of the allele frequency in European-Americans and African-Americans, respectively. Other common alleles (allele frequency >5% in a population) include MICA*00901 and *010. MICA alleles showed strong linkage disequilibrium with HLA-B. Typically, a common MICA allele has strong LD with several HLA-B alleles, whereas most HLA-B alleles and their related serological groups are associated with a single MICA allele. The lack of evidence for an active diversification of the MICA gene after racial separation indicates an evolutionary history distinct from that of the classical HLA genes.     

HLA-C locus alleles distribution in patients from northern Poland with psoriatic arthritis--preliminary report

The aim of the study was to compare the frequency of human leucocyte antigen-C (HLA-C) locus alleles in patients with psoriatic arthritis and in healthy controls in the same ethnic group in Poland, and to correlate them with age of onset of psoriatic skin changes and joints symptoms. HLA-C locus alleles of 41 patients and 80 controls were determined by a polymerase chain reaction (PCR) low-resolution method. The Cw*06 allele occurred more frequently (P adjusted for multiple comparison = 0.004) in patients with psoriatic arthritis than in controls. Patients who carried the HLA-Cw*06 allele had a significantly earlier mean age of onset of both psoriasis (P = 0.01) and arthritis (P = 0.008) compared with Cw*06-negative patients. Our results confirm the association between Cw*06 allele and psoriatic arthritis in the northern Poland population and suggest that the HLA-Cw*06 may determine not only the disease susceptibility, but also the age of onset of psoriatic arthritis.     

Role of HLA-B and C alleles in development of psoriasis in patients from North India

Abstract: The association of HLA antigens with susceptibility for development of psoriasis vulgaris has been studied mostly using serologic methods. A number of different HLA class I antigens have been reported to be associated with predisposition to develop this disease. While Cw6 is the allele most commonly thought to confer risk for psoriasis, a number of HLA-B locus alleles have also been thought to be involved and, recently, in at least three studies, a specific amino acid in the HLA-C heavy chain has been implicated. With the recent development of molecular methods for typing for HLA class I alleles, it has become possible to re-examine this association by performing high resolution typing. In the present study we have investigated 38 psoriasis patients and 84 ethnically and geographically matched controls from North India. They were typed for HLA-B and HLA-C. The results showed the Cw*0602 was the main allele that was increased in this group of patients. The previously reported association with alanine-73 was not found to be significant. Cw*0602 was found in 71% of the patients. B*5701 and B*3701 were also increased but appeared to be secondary to linkage disequilibrium with Cw*0602.     

Association study between HLA-A, -B, -C, -DRB1 alleles and Psoriatic arthritis in southern France

Psoriatic arthritis (PsA) is a type of inflammatory arthritis associated with psoriasis. HLA association studies performed in northern Europe, comparing patients with control populations, have shown that the highest risk for PsA is carried by HLA-C*06, HLA-B*57 and HLA-B*27 alleles.This retrospective association study compared HLA-A, -B, -C, and -DR alleles of 500 patients from southern France, who fulfilled the CASPAR criteria for Psoriatic Arthritis (PsA), with 2346 controls from healthy blood donors, using the chi-square test.We classified PsA patients into three different subgroups according to disease: purely axial, purely peripheral and combined axial and peripheral. The ‘axial’ subgroup was associated with HLA-B*27 (OR = 16.3, p = 2.7 × 10^?28) and its haplotypes: HLA- B*27-C*01 (OR = 12.4, p = 1.7 × 10^?12) and HLA-B*27-C*02 (OR = 8.7, p = 10 × 10^?9). The ‘axial and peripheral’ and the ‘peripheral’ subgroups were associated with HLA-C*06 (respectively OR = 1.5, p = 3.6 × 10^?10 and OR = 2.4, p = 3.6 × 10^?12) and its haplotypes HLA-C*06-B*13 (respectively OR = 2.4, p = 1.2 × 10^?6 and OR = 2.8, p = 6.4 × 10^?11).This association study on a southern French PsA cohort identifies HLA-C*06 as a marker for peripheral PsA and HLA-B*27 as a marker for purely axial PsA.     

The MICA‐A4 triplet repeats polymorphism in the transmembrane region confers additional risk for development of psoriatic arthritis in the Croatian population

The aim of this study was to investigate possible differences in the frequencies of alleles at the HLA loci and at microsatellite loci within the HLA region among patients suffering from psoriatic arthritis (PsA) and healthy controls. Fifty‐eight Croatian PsA patients (28 male and 30 female) and 157 healthy unrelated controls were typed for HLA alleles (A, B, Cw and DRB1) by the polymerase chain reaction–sequence‐specific primers (PCR‐SSP) method, while microsatellite alleles (D6S265, D6S273, MHC class I chain‐related gene (MICA) and MIB) were analysed by electrophoresis in an ALFexpress sequencer (Pharmacia Biotech, Uppsala, Sweden). The findings from this study were: (1) the frequencies of B*39 and B*57 were significantly increased in PsA patients; (2) differences in the frequencies of B*13 and B*27 were not statistically significant after correction; (3) the B*0702, B*18, and B*38 alleles were decreased in patients only before correction; (4) none of the alleles at other HLA loci tested were associated with PsA in Croatia; (5) polymorphism at D6S265, D6S273, and MIB microsatellites in patients did not show any statistically significant differences when compared to controls; (6) the increase in the MICA‐A4 allele frequency in PsA patients was independent of the B*39 and B*57 alleles.     

MICA polymorphism in a population from north Morocco, Metalsa Berbers, using sequence-based typing

The MICA gene encodes a family of nonclassical major histocompatibility complex class I molecules. Data on MICA polymorphism in different populations are still limited. In the present study, MICA allele frequencies (af) were assessed in 82 unrelated healthy individuals from a Moroccan Berber population named Metalsa (ME) by means of sequence-based typing of exons 2, 3, 4, and 5. In consideration of the linkage disequilibrium existing between MICA and human leukocyte antigen (HLA) class I alleles, MICA/HLA-B, MICA/HLA-Cw, and MICA/HLA-A haplotype frequencies (hf) were estimated. A wide allelic distribution including 16 different MICA alleles was found in ME. The most common MICA alleles were MICA*00801 (af = 0.268), *004 (0.232), *00902 (0.140), *00901 (0.085), and *00901 (0.073). The most common MICA/HLA-B haplotypes were MICA*004-B*4403 and MICA*009-B*50 (hf = 0.113 for both these haplotypes). Some known MICA and HLA-B associations were confirmed in this population. Noteworthy was the high frequency of MICA*009 (af = 0.226); the high frequency of B*50 found in ME (af = 0.114) permitted us to evidence the associations of MICA*00902 with B*5001 (hf = 0.068) or *5002 (hf = 0.045), whereas MICA*00901 was mainly associated with B*5101 (hf = 0.038), which corresponds to the previously described association MICA*009/A6-HLA-B*51. This study extends the previous knowledge on MICA polymorphism to a North African white population and may have implications for disease associations and transplantation.     

MICA polymorphisms and haplotypes with HLA-B and HLA-DRB1 in Koreans

Abstract
Major histocompatibility complex (MHC) class I chain-related gene A ( MICA ) is located within the human MHC, centromeric to HLA-B and telomeric to HLA-DRB1 . The location of MICA in the MHC indicates the presence of linkage disequilibrium with human leukocyte antigen (HLA). Like HLA, MICA is highly polymorphic; however, the information available for MICA polymorphisms is not as comprehensive as that for HLA polymorphisms. We estimated the allelic frequencies of MICA and haplotypes with HLA-B and HLA-DRB1 at high-resolution in a population of 139 unrelated Korean individuals by applying the newly developed method of sequence-based typing (SBT). A total of 17 MICA alleles were identified. The most frequent allele was MICA*010 (19.4%), followed by alleles *00201 (17.6%), *00801 (14.7%), *01201 (9.4%), *004 (8.3%) and *049 (7.9%). The most common two- and three-locus haplotypes were HLA-B*1501-MICA*010 (10.4%), MICA*010-HLA-DRB1*0406 (5.8%) and HLA-B*1501-MICA*010-HLA-DRB1*0406 (5.8%). This is the first study to provide such high-resolution information on the distribution of haplotypes comprising MICA , HLA-B and HLA-DRB1 in Korean individuals, a level of resolution made possible by use of the SBT method. The results of this study should help determine the mechanisms underlying diseases associated with MICA polymorphisms in Korean individuals.     

The MICA-A4 triplet repeats polymorphism in the transmembrane region confers additional risk for development of psoriatic arthritis in the Croatian population.

The aim of this study was to investigate possible differences in the frequencies of alleles at the HLA loci and at microsatellite loci within the HLA region among patients suffering from psoriatic arthritis (PsA) and healthy controls. Fifty-eight Croatian PsA patients (28 male and 30 female) and 157 healthy unrelated controls were typed for HLA alleles (A, B, Cw and DRB1) by the polymerase chain reaction-sequence-specific primers (PCR-SSP) method, while microsatellite alleles (D6S265, D6S273, MHC class I chain-related gene (MICA) and MIB) were analysed by electrophoresis in an ALFexpress sequencer (Pharmacia Biotech, Uppsala, Sweden). The findings from this study were: (1) the frequencies of B*39 and B*57 were significantly increased in PsA patients; (2) differences in the frequencies of B*13 and B*27 were not statistically significant after correction; (3) the B*0702, B*18, and B*38 alleles were decreased in patients only before correction; (4) none of the alleles at other HLA loci tested were associated with PsA in Croatia; (5) polymorphism at D6S265, D6S273, and MIB microsatellites in patients did not show any statistically significant differences when compared to controls; (6) the increase in the MICA-A4 allele frequency in PsA patients was independent of the B*39 and B*57 alleles.     

Association study of MICA and MICB in Alzheimer's disease

Following the replication of the association of the human leucocyte antigen (HLA) allele, HLA-B*07, with Alzheimer's disease (AD) in the cohort of the Oxford Project to Investigate Memory and Ageing (OPTIMA) in a previous study, we examined whether that association could be due to linkage disequilibrium with MICA or MICB alleles. We found a possible association of MICA*00801 heterozygotes with AD in subjects positive for the ɛ4 allele of apolipoprotein E. This finding was supported by Hardy–Weinberg analysis, by stratified association analysis and by interaction analysis, but did not survive correction for multiple testing. In any case, these results do not explain our previously reported association of HLA-B*07 with AD.     

HLA-C locus and genetic susceptibility to psoriatic arthritis in Romanian population

We determined the distribution of human leukocyte antigen-C (HLA-C) allelic groups in a cohort of psoriatic arthritis (PsA) patients and a control population of Romanian ethnicity. A nominal association of HLA-C*06 with susceptibility to PsA was observed [P = 0.014, p(corr) > 0.05, odds ratio (OR) 2.1, 95% confidence interval (CI) 1.08-4.46]. When subanalyzing data according to PsA clinical phenotypes, association was noticed between HLA-C*06 and PsA with psoriasis onset before 40 years (p(corr) = 0.013, OR 3.7, 95% CI 1.58-9). This first report from Romania confirmed the association of HLA-C*06 with type I psoriasis in PsA patients. Other study findings, such as the relationship between HLA-C*06 and spondylitis or the protective effect of HLA-C*07 for the polyarthritis clinical phenotype of PsA, are of preliminary character and require verification.     

Distribution of MICA alleles and haplotypes associated with HLA in the Korean population

The MICA (MHC class I chain-related gene A) is a polymorphic gene located 46 kb centromeric of the HLA-B gene, and is preferentially expressed in epithelial cells and intestinal mucosa. The MICA gene, similar to human leukocyte antigen (HLA) class I, displays a high degree of genetic polymorphism in exons 2, 3, 4, and 5, amounting to 54 alleles. In this study, we investigated the polymorphisms at exons coding for extracellular domains (exons 2, 3, and 4), and the GCT repeat polymorphism at the transmembrane (exon 5) of MICA in 199 unrelated healthy Koreans. Eight alleles were observed in the Korean population, with allele frequencies for MICA*010, MICA*00201, MICA*027, MICA*004, MICA*012, MICA*00801, MICA*00901, and MICA*00701 being 18.3%, 17.8%, 13.6%, 12.3%, 11.1%, 10.8%, 10.6%, and 3.3%, respectively. Strong linkage disequilibria were also observed between the MICA and HLA-B gene-MICA*00201-B58, MICA*004-B44, MICA*00701-B27, MICA*00801-B60, MICA*00901-B51, MICA*010-B62, MICA*012-B54, and MICA*027-B61. In the analysis of the haplotypes of HLA class I genes (HLA-A, B, and C) and the MICA, the most common haplotype was MICA*004-A33-B44-Cw*07, followed by MICA*00201-A2-B58-Cw*0302 and MICA*012-A2-B54-Cw*0102. The MICA null haplotype might be identified in the HLA-B48 homozygous individual. These results will provide an understanding of the role of MICA in transplantation, disease association, and population analyses in Koreans.     

A candidate gene for psoriasis near HLA-C, HCR (Pg8), is highly polymorphic with a disease-associated susceptibility allele

A susceptibility gene for psoriasis, a chronic skin disorder, resides in chromosome 6p near the HLA-C locus. Sequencing of the region has allowed the identification of a new gene, HCR. We found that HCR is highly polymorphic with at least 12 coding variants. An association study of the new HCR polymorphisms and the previously suggested susceptibility alleles HLA-Cw*0602 and corneodesmosin allele 5 (CD*5) with psoriasis revealed a specific HCR variant associated with psoriasis susceptibility. However, the HLA-Cw*0602 allele was rarer in controls and associated with a stronger relative risk. Association analysis did not support CD*5 as a psoriasis susceptibility allele in our sample of patients (n = 100) and population-matched controls (n = 93) from an isolated population. We found HCR to be overexpressed in keratinocytes of psoriatic lesions compared with paired samples of healthy skin. Our results suggest a potential role for HCR in the pathogenesis of psoriasis.     

Association of MICA polymorphism with HLA-B51 and disease severity in Korean patients with Behcet's disease

The HLA-B51 allele is known to be associated with Behcet's disease (BD) in many ethnic group. However, it has not yet been clarified whether the HLA-B51 gene itself is the pathogenic gene related to BD or whether it is some other gene in linkage disequlibrium with HLA-B51. Recently, the Triplet repeat (GCT/AGC) polymorphism in transmembrane region of the MHC class I chain-related A (MICA) gene was identified. To investigate the association of MICA with BD, we studied the MICA polymorphism in 108 Korean BD patients and 204 healthy controls in relation to the presence of HLA-B51 and clinical manifestations. The triplet repeat polymorphism was determined by polymerase chain reaction (PCR)-denaturing polyacrylamide gel electrophoresis (PAGE). The phenotype frequency of the MICA*A6 allele (relative risk, RR=2.15, p=0.002) and HLA-B51(RR=1.87, p=0.022) were significantly increased in the Korean patients with BD. A strong linkage disequilibrium was observed between the MICA*A6 and HLA-B51 in both the patients with BD and control subjects. Stratification analysis showed that MICA*A6 homozygosity was strongly associated with BD in the HLA-B51-negative population, and HLA-B51 was also associated with MICA*A6-negative population. In conclusion, MICA*A6 rather than HLA-B51 was strongly associated with Korean patients with BD, and the MICA*A6 allele is a useful susceptibility marker of BD, especially in the HLA-B5-negative     

中国北京男男同性恋艾滋病感染者HLA 分型及与疾病进展分析

目的:分析中国北京男男同性恋艾滋病感染者HLA-A、HLA-B、HLA-C 基因频率以及HLA 基因分型与疾病进展关系。方法:序列特异性引物的PCR 扩增(PCR-sequence specific primer typing,PCR-SSP)对310 名随机中国北京男男同性恋艾滋病感染者HLA-A、HLA-B、HLA-C 基因进行分型,采用HIV Molecular Immunology Database 中HLA Analysis Tools 方法计算等位基因频率。结果:在北京男男同性恋艾滋病感染者队列中,HLA-A*1101(34.52%)基因频率最高,其次为HLA-A*0201(31.94%)、HLA-C*0102(27.10%)以及HLA-A*2402(26.45%)。根据艾滋病感染者1 年内CD4 细胞计数变化来判断感染者是否为快速进展者,结果发现快速进展者134 例,非快速进展者176 例。快速进展者中存在低水平的HLA-B*4403(快速进展者1.12%,非快速进展者为4.27%,P = 0.0276),HLA-B*1511(非快速进展者5.98%,快速进展者2.25%,P =0.0282),HLA-B*5701(非快速进展者2.56%,快速进展者0.37%,P =0.0491)表达,而HLA-C*0304 则在快速进展者中高表达(非快速进展者4.56%,快速进展者8.96%,P =0.0319)。结论:本研究获得了中国北京男男同性恋艾滋病感染者HLA-A、HLA-B、HLA-C 基因的等位基因分布频率数据,并发现HLA-B*4403、HLA-B*1511、HLA-B*5701 与延缓艾滋病疾病进展有关,而HLA-C*0304 则与加速艾滋病疾病进展相关。     

Linkage disequilibrium analysis of familial psoriasis: identification of multiple disease-associated MHC haplotypes

Although psoriasis vulgaris (PsV) is strongly associated with certain human leukocyte antigens, the pathogenetic nature of these associations remains elusive. The objectives of this study were: (i) to determine whether HLA loci directly determine susceptibility or merely serve as markers for the susceptibility allele; and (ii) to identify additional disease-associated haplotypes. By applying maximum likelihood linkage disequilibrium analysis (LDA) in cases vs. controls, we found the susceptibility gene to be more strongly associated with specific HLA haplotypes than with their component alleles. Stronger linkage disequilibrium between PsV and HLA alleles was detected at HLA-C and HLA-B than at DRB1 and DQB1. Parametric linkage analysis accounting for marker-trait disequilibrium in psoriasis vulgaris pedigrees yielded most significant results for a locus close to HLA-B and -C. Furthermore, we found that susceptibility is linked to at least three different ancestral HLA haplotypes; among them, HLA-Cw7-B8-DRB1*0301-DQB1*02 is linked to PsV for the first time. These results identify a major PsV susceptibility locus in the immediate vicinity of, but distinct from HLA-B or HLA-C, and suggest that multiple disease alleles have arisen during human evolution.     

Nomenclature for factors of the HLA system, update May 2009

Major histocompatibility complex class I chain-related gene A ( MICA ) is located 46 kb centromeric to HLA-B locus and encodes a stress-inducible protein. MICA allelic variation is thought to be associated with disease susceptibility and immune response to transplants. In this study, polymerase chain reaction sequence-based typing (PCR-SBT) method for MICA alleles has been established. Genomic DNAs from 100 healthy Chinese Han individuals were typed for MICA alleles by this method. The microsatellite polymorphism in the exon 5 of MICA gene, MICA * Del allele, and human leukocyte antigen-B alleles was also detected by the polymerase chain reaction–GeneScan, polymerase chain reaction sequence-specific primer, and PCR-SBT methods, respectively. Fourteen MICA alleles were found in the population, with MICA * 00801/04 having the highest frequency of 27.0%. MICA * A4 , * A5 , * A5.1 , * A6 , and * A9 microsatellites were identified. Two samples with HLA-B * 4801 were MICA * Del positive, with the frequency of 1.0%. The data showed that the new PCR-SBT method for MICA alleles was reliable and Chinese Han population was distinct in distribution of MICA alleles.     

HLA-DRB1*04 alleles in psoriatic arthritis: comparison with rheumatoid arthritis and healthy controls.

Psoriatic arthritis (PsA) is an inflammatory arthritis associated with psoriasis usually seronegative for rheumatoid factor. An increased frequency of HLA-DR4 has been noted in PsA, particularly among patients with a rheumatoid arthritis like (RA) arthritis. The aim of the current investigation was to compare HLA-DRB1*04 alleles in patients with PsA, patients with RA, and healthy controls. Sample size calculations based on the frequency of HLA-DR4 suggested that 90 individuals in each patient group would be sufficient to address our question. Therefore, 90 HLA-DRB1*04 positive patients from each patient group underwent high resolution molecular typing and were included in this study. Although HLA-DRB1*0401 was the most frequent allele in all groups, its frequency among the PsA patients was lower than that of RA patients and controls. HLA-DRB1*0402 was higher among patients with PsA. Patients with RA were more likely to have more than one shared epitope allele than either PsA or the healthy control group. HLA-DQB1 alleles did not contribute further information. We suggest that the differences in the class II HLA epitope(s) may also be related to interaction specificity with another molecule functioning in the immune response to a putative arthritogenic antigen and result in differences in disease expression.     

HLA class-I and class-II allele frequencies and two-locus haplotypes in Melanesians of Vanuatu and New Caledonia

HLA class-I and class-II allele frequencies and two-locus haplotypes were examined in 367 unrelated Melanesians living on the islands of Vanuatu and New Caledonia. Diversity at all HLA class-I and class-II loci was relatively limited. In class-I loci, three HLA-A allelic groups (HLA-A*24, HLA-A*34 and HLA-A*11), seven HLA-B alleles or allelic groups (HLA-B*1506, HLA-B*5602, HLA-B*13, HLA-B*5601, HLA-B*4001, HLA-B*4002 and HLA-B*2704) and four HLA-C alleles or allelic groups (HLA-Cw*04, HLA-Cw*01, HLA-Cw*0702 and HLA-Cw*15) constituted more than 90% of the alleles observed. In the class-II loci, four HLA-DRB1 alleles (HLA-DRB1*15, HLA-DRB1*11, HLA-DRB1*04 and HLA-DRB1*16), three HLA-DRB3-5 alleles (HLA-DRB3*02, HLA-DRB4*01 and HLA-DRB5*01/02) and five HLA-DQB1 alleles (HLA-DQB1*0301, HLA-DQB1*04, HLA-DQB1*05, HLA-DQB1*0601 and HLA-DQB1*0602) constituted over 93, 97 and 98% of the alleles observed, respectively. Homozygosity showed significant departures from expected levels for neutrality based on allele frequency (i.e. excess diversity) at the HLA-B, HLA-Cw, HLA-DQB1 and HLA-DRB3/5 loci on some islands. The locus with the strongest departure from neutrality was HLA-DQB1, homozygosity being significantly lower than expected on all islands except New Caledonia. No consistent pattern was demonstrated for any HLA locus in relation to malaria endemicity.