Complete HLA-B44 subtyping in Spaniards. Additional evidence for heterogeneity in Caucasian populations

HLA-B44 is one of the most common HLA class I alleles in Caucasians. Exon 3 oligotyping and sequence analysis have define five B44 subtypes: B^*4402, B^*4403, B^*4404, B^*4405 and B^*4406, with variations in exons 2 and 3. We have developed a conventional DNA typing system by using a single B12-group specific amplification including exon2-intron2-exon3 in combination with 6 oligoprobes to define all B44 subtypes. 140 HLA-B44 positive unrelated Spanish Caucasians were typed. Family studies established 30 B44-bearing haplotypes. The distribution of B44 subtypes in our population was: B^*4402 32.5%, B^*4403 66.5%, B^*4404 0.5%, B^*4405 0.5%, B^*4406 not found. B^*4402 and B^*4403 represented the 99% of the B44 alleles, as described in Caucasians. However, these two major subtypes showed an inverted frequency when compared with other Caucasian populations, B^*4403 twice as frequently as B^*4402 in Spaniards. HLA-B44-associated chromosomes showed 20 different haplotypes (including HLA-A,-C,-DR,-DQ), although demonstrating clear separated haplotype composition between B^*4402 and B^*4403: B^*4402 associated to class I alleles A2 ( ) and Cw5 ( ), and B^*4403 associated to the class II allele DRB1^*0701 ( ). These findings, in addition to the validation of a complete B44 oligotyping system, revealed further evidence of antigen frequency differences among populations of the same ethnic origin.     

HLA-B27 subtypes in chinese patients with ankylosing spondylitis

The HLA-B27 allele has been extensively studied due to its strong association with ankylosing spondylitis (AS). In order to identify B27 alleles in Chinese patients with AS from the Shanghai area, we joined the AHS#5 of the 12 IHW and total of 68 B27 positive patients and 7 B27 positive normal persons have been investigated using polymerase chain reaction and Dig-ddTUP labeled oligonucleotides. Three primer pairs, E403 and E90as, E91As and E136as, E91Bs and E18as, were used to amplify codons 40-90 of HLA-B related alleles, codon 91 to 136 of HLA-B^*2701-B^*2706 and B^*2708 and codons 91-180 of B^*2707. A total of 11 probes were used to distinguish 8 B^*27 alleles from B^*2701 to B^*2708. 68 AS patients contain 69 B27 alleles because one patient is heterozygous B^*2704/B^*2705. A total of 4 alleles of B^*27 were detected in the AS-patient group. B^*2704 is the most common B^*27 allele in both AS patients and controls with similar frequency, 76.8% and 71.4%, respectively. We found a high proportion of B^*2705 in both AS patient (20.3%) and control (28.6%) groups. Although the control group is quite small we are still able to deduce that B^*2704 and probably also B^*2705 seem to be associated with AS in Shanghai area patients We also found one AS allele typed as B^*2707. Interestingly, for the first time we detected B^*2706 in an AS patient, which would argue against a protective effect of B^*2706 on AS susceptibility in Shanghai Chinese. The conclusion from this study is that the distribution of B^*27 alleles is not significantly different between AS patients and controls. Expanded numbers of AS patients and especially of healthy controls in different ethnic groups will be necessary to assess the contribution of different B27 subtypes to AS susceptibility.     

HLA-B35 alleles in 440 individuals from 16 different populations as assessed by DNA typing.

As part of the Anthropology Component, the distribution of HLA-B35 alleles (B^*3501 to 3513) was studied in 16 different populations by group specific amplification and SSOP hybridization. The results were as follows:

The predominant alleles in most Caucasian populations were 3501 > 3503 > 3502 > 3508. However, B^*3502 predominated in Jews, B^*3508 in Arabs, B^*3503 in Gypsies and Sardinians seem to have only B^*3501 and 3502. B^*3504, 3505, 3506 and 3509 were restricted to Amerindians, where there are still other new B35 variants to be characterized. In most individuals the different B35 alleles were found in phenotypic combinations with HLA-Cw4, suggesting that the B35, Cw4 haplotype may have existed before subtype diversification. A detailed analysis of HLA-B35 alleles in other populations might help to draw a precise picture of B35 evolution.     

Identification of HLA-B44 and HLA-B45 alleles using a nested arms-PCR method

The establishment of DNA based identification of HLA Class I specificities has made high resolution typing possible at the allelic level. With such methods, it is feasible to identify population variation in allelic frequencies. Working towards higher allelic resolution of the HLA Class I loci we have developed a panel of primers to identify all the known HLA-B^*44 alleles and B^*4501 using a nested ARMS-PCR approach. There are currently seven different B^*44 alleles identified: B^*4402, B^*44031, B^*44032, B^*4404, B^*4405, B^*4406 and B^*4407. The system was validated using DNAs from B lymphoblastoid cell lines which have been sequenced for their B^*44 alleles. Further typing has been performed on a panel of 29 cell lines from the IHW cell panel and some 37 samples collected as part of the anthropological study of the Orkney islanders, all of which were previously defined as being B^*44 by the 12th Workshop HLA Class I typing kit. In the Orcadian population B^*4402 was found in 28 individuals to be the most common B^*44 allele, with 18 of these individuals expressing the A^*02-Cw^*0501-B^*4402 haplotype. The remaining 9 B^*44 positive individuals were all found to express the A^*29-Cw^*1601-B^*44031 haplotype. Methods and results will be presented.     

B78Hen, a new Caucasian member within the B78 family

We have previously described an antigen B7801 (SNA) belonging into the B5 CREG, exclusively found in North and West African individuals. More recently, a new subtype B^*7802, differing from B^*7801 by two nucleotide substitutions within the class I heavy chain 1 domain has been found in a Caucasian individual (1).

Here we describe another B78 variant (B78Hen) that we observed in a Belgian German family. Serology displayed a pattern identical to the one of SNA. Careful analysis of the pedigree could rule out African ancestry. The exon 2 of B78Hen differed only by one nucleotide substitution from B^*7801 and B^*7802. In B78Hen, Asp at codon 74 of B^*7801 was substituted by Tyr, as in B^*7802. With respect to B^*7802, a silent mutation was observed in B78Hen at codon 23 (ATC to ATT). The ATT codon at that position is also found in most B51 subtypes, B^*52012 and B^*7801 and is absent in B^*51022, B^*5105, B^*7802 and B^*35.

Beside its phylogenetic implication, this finding has also practical consequences. The B78 antigens are not restricted to African subjects as it was previously thought. In view of its frequent confusion with B35 and B70 antigens, it must now be searched actively in Caucasian subjects.     

Analysis of endogenous peptides eluted from HLA-B27 variants: implications in the susceptibility to spondylarthropathies

Population studies suggest that some HLA-B27 subtypes (HLA-B^*2705, B^*2702) could be more strongly associated with the development of spondylarthropathies than others (B^*2703, B^*2706, B^*2709). Differences in the peptide binding groove could impose differences in the nature of peptides bound by these different alleles. We have eluted endogenous peptides from C1R-B^*2705 and B^*2703 transfectants. The B^*2705 HPLC profile was more complex than the B^*2703 one. Several B^*2705 and B^*2703 individual peaks were sequenced by Edman degradation and mass spectrometry. Some peptides were shared by both subtypes. One B^*2705 eluted peptide present in a major HPLC fraction was not found in the B^*2703 peptides. The corresponding synthetic peptide bound in vitro specifically to T2-B^*2705 and not to T2-B^*2703. This result emphasizes that even one amino-acid difference outside the major anchor binding pockets at position 59 between B^*2705 and B^*2703 could notably influence the endogenous peptides naturally presented. This could have consequences in terms of T cell repertoire selection and development of autoimmunity.     

Characterization of a new HLA-B allele (B3702) generated by an intronic recombination event

Routine serological HLA typing of a Syrian family revealed a Bw4-associated HLA-B blank antigen showing Mendelian segregation together with the haplotype A1, Cw2, DR11, DQ7. A more extensive serological analysis showed no conclusive typing reactions with sera towards HLA-B27 and HLA-B37 antigens. Full length cDNA was sequenced in both senses for two individuals. Comparison of the consensus sequence with all other HLA-B published sequences evidenced a new HLA-B allele, confirming, as serologically predicted, the greatest relationship to HLA-B37 and HLA-B27 genes. Based on the exon 2 sequence identity to B^*3701, this new allele has been designated B^*3702. Proteins derived from B^*3702 and B^*3701 differ in 9 amino acids at alpha 2 (7 residues), and in transmembrane and cytoplasmic domains. Exons 3 to 8 are identical between B^*3702 and a number of HLA-B27 subtypes (all except Oriental HLA-B27 subtypes B^*2704, B^*2706 and B^*2707). B^*3702 alpha 1 domain differs from those of B^*2701, B^*2702, B^*2703, B^*2705, and B^*2708 in 9, 11, 9, 7, and 11 residues, respectively. The mosaicism found in B^*3702 strongly suggests that this new allele could be derived by homologous recombination at the intron 2 region between the B^*3701 gene and some of the 5 non-oriental B27 subtypes.     

Peptide binding to the differentially disease-associated HLA-B2704 and B2706 and to mutants mimicking their polimorphism

Peptide binding to the disease-associated antigen HLA-B27 and its modulation by subtype polimorphism was addressed in this study. The effect of subtype changes was analyzed using a quantitative stabilization assay in which the surface expression of HLA-B27 on RMA-S cells was measured as a function of the concentration of peptides naturally presented by B^*2705, B^*2702 and their analogues. Binding to B^*2704, B^*2706 and to mutants mimicking the changes between these subtypes and B^*2705 was analyzed. Bulky aliphatic (Leu), aromatic (Phe or Tyr) or basic (Arg,, Lys) C-terminal residues contribute similarly to binding to B^*2705. For B^*2704 aliphatic C-terminal residues are the most suitable, but aromatic and even basic residues can also be accomodated. B^*2706 has strong preference for bulky aliphatic C-terminal residues and moderate suitability for aromatic ones. The effects of individual changes in the subtypes account only partially for the binding properties of B^*2704 and B^*2706, suggesting interactive effects of the changes in determining the peptide specificity of these subtypes that are not fully accounted for by simple additive effects of the individual changes.     

HLA-B40, B18, B27, and B37 allele discrimination using group-specific amplification and SSCP method

We developed a system for discriminating HLA-B40, B18, B27, and B37 alleles using a two-step PCR method followed by SSCP analysis. Fragments (0.8 kb) including exon 2, intron 2, and exon 3 were amplified in the first PCR. We used two sets of primers, one specific for HLA-B60-related alleles and the other specific for HLA-B6l-related, B18, B27, and B37 alleles. No amplifications of other class I genes or pseudogenes were observed. In the second PCR, exon 2 and exon 3 were amplified separately, using diluents of the first PCR products as templates. HLA-B6l-related, B18, B27, B37, and B60-related alleles were clearly discriminated in the SSCP analysis of the second PCR products. In a population study in which B6l alleles were analyzed, B^*4003 was detected in two Japanese individuals in addition to two B6l alleles previously reported to occur in Japanese, B^*4002 and B^*4006. The relative frequencies of B^*4002, B^*4006, and B^*4003 in Japanese were 58, 35, and 6%, respectively. The individuals having B^*4003 are the first non-South Americans in whom this allele has been detected. The SSCP banding patterns of 18 HLA-B60-positive Japanese population samples were identical to those of a B^*40012 sample for both exon 2 and exon 3. We also demonstrated that the B37 allele occurring in some Japanese is B^*3701.     

Nucleotide sequences of novel HLA-B35 subtypes

In recent reports we described novel hybridization patterns (HP) corresponding to 22 potentially new HLA-B locus alleles in a panel of 547 subjects studied by PCR-SSOP. Three of them correspond to new subtypes of B35. To confirm the hybridization results we have isolated DNA from PBL and performed PCR, DNA cloning and nucleotide sequencing. One of the alleles, locally called B-3505v was found in three individuals: two Hispanic, one Caucasoid. It differs from B^*3505 by 3 nucleotide substitutions that lead to changes in residues 94 (Ile > Thr), 95 (Ile > Leu) and 103 (Val > Leu). B-3505v differs from B^*3501 in residues 97 and 103. Another allele called B-3508v, was found in 7 individuals, (6 of 122 Toba Indians, 1 of 18 Pilaga Indians). It differs from B^*3509 in two silent nucleotide substitutions (codons 135 and 138) and in one substitution in residue 156 (Arg > Leu). The new allele has a hybrid sequence between B^* 3508 and B^*4801. A third subtype, locally called B-3504v, observed in two Hispanic individuals, is identical to B^*3512. B^*3512 differs from B^*3504 by 3 nucleotides and one amino acid. Substitutions in residue 95 contribute to the structure of specificity pocket F, 97 to C and E, and 156 to pockets D and E. Therefore it is possible that some of the new alleles may have different peptide binding profiles. Since differences in residue 156 have been shown to affect allorecognition and mediate GvHD, identification of such variants may have important implications in transplantation and perhaps in studies of immune responses to peptides and pathogens.     

Two new class I alleles (HLA-Cw15052 and B0706) constitute a novel haplotype in Caucasians.

HLA-Cw^*15 alleles have been previously described in association with the HLA-B specificities B51, B61 and B73. In this study, we have characterized a novel haplotype constituted by the new alleles Cw^*15052 and B^*0706.

The two variants of Cw^*1505 differ in a synonymous change in nucleotide 24 of exon 1 coding for the signal peptide. The new B^*0706 allele differs from B^*0702 and B^*0705 in single changes at exons 3 and 5, respectively, and could be considered an intermediate evolutionary step between them.

Even though strong linkage disequilibrium between the HLA antigens Cw7 and B7 has been observed, ‘HLA-C blank’-B7 haplotypes are not rare either. Cw^*15 and B7 have been found to be associated in several Caucasian individuals. Therefore, it is likely that Cw^*15 accounts for a part of the aforementioned ‘HLA-C blank’-B7 haplotypes.

With the identification of Cw^*15052, a previously proposed pathway for the evolution of Cw^*15 alleles can be completed and will be discussed.     

Conservation of genes encoding HLA-B5 and B35 cross-reactive group antigens in various races

HLA-B5 and B35 CREG antigens include HLA-B35, B51, B52, B53, and B78. Recent studies suggest that the genes encoding the HLA-B5, B35 CREG, and HLA-B58 antigens share a common ancestor. We sequenced the exons of the genes encoding HLA-B51, B53, and B58 from American black individuals and the gene HLA-B52 encoding from an Arabic individual, and compared them with previously reported sequences of HLA-B51 (B^*5101) and HLA-B52 (B^*5201) from Japanese, HLA-B53 (B^*5301) from an Algerian, and HLA-B58 (B^*5801) from a Sardinian. The sequences of the genes from the American black and Arabic individuals were identical to those from the other races. These findings support the hypothesis that these antigens have evolved prior to divergence of the major ethnic groups.     

HLA-C allele frequencies and linkage disequilibrium in 604 random Caucasian donors by high resolution PCR-SSP.

Sequence-specific PCR (PCR-SSP) methods have been successfully used to type HLA-C to a medium resolution level through the use of group-specific amplifications. We have developed high resolution PCR-SSP utilising 48 allele and group-specific PCR reactions to define HLA-C to the allelic level in order to find out the true level of polymorphism in a random Caucasian population and hence define the linkage disequilibrium between HLA-A, HLA-B and HLA-C.

604 consecutive samples from random individuals consisting of 552 cadaver donors and 52 random blood donors were typed by high resolution PCR-SSP. The PCR conditions and parameters are described in Tissue Antigens 95, vol. 46 pg 355. Cw^*0101, ^*0201, ^*0302, ^*0402/3, ^*0703, ^*0801/3, ^*1201,^* 1301, ^*1401,^* 1403 ^*1501/3/4 were not detected in our population although we have detected Cw^*0302, ^*1403, ^*0801 and ^*1504 in other populations. Cw7 is the most common Caucasian HLA-C antigen with frequencies of 30, 28 and 3.8% for Cw^*0701,^* 0702 and ^*0704 respectively. Cw^*0701 is in linkage disequilibrium with B8, B49 and B18, Cw^*0702 is associated with B7 and B39 whilst Cw^*0704 is associated with B44 and B^*1518. Complete frequencies, linkage disequilibria, statistics and full information on the primer mixes will be given.     

Identification of the gene encoding a novel HLA-B39 subtype Two amino acid substitutions on the β-sheet out of the peptide-binding floor form a novel serological epitope

Serologic analysis suggests the existence of a novel HLA-B39 subtype (HLA-B39N) in the Japanese population. To identify this novel allele, a gene encoding HLA-B39N was cloned and the exons were sequenced. A gene encoding HLA-B39N (B^*3904) and B^*39011 differs by two nucleotide substitutions at codons 11 and 12 whereas B^*3904 and B^*39013 differ by three nucleotide substitutions at codons 11, 12, and 312. One nucleotide difference at codon 11 produces a change from serine in B^*3901 to alanine in B^*3904 whereas another difference at codon 12 changes valine in B^*3901 to methionine in B^*3904. The residues 11 and 12 are located on the β-sheet out of the peptide-binding floor and are completely buried in the molecule. These results suggest that the substitutions at these residues alter the conformation of other residues forming epitopes of alloantibodies. Analysis of HLA-B^*3901 genes in the Japanese population showed that both B^*39011 and B^*39013 were observed in the Japanese population. The present study suggests that B^*3904 may have evolved from B^*39011 rather than B^*39013.     

Two new alleles of HLA-B22 group in Japanese

Five different B22 split antigens, B54, B55.1, B55.2, B56 and B22N, have been reported in Japanese. Genomic DNAs were analyzed by PCR-SSCP and sequencing methods for Japanese samples serologically positive for each split antigen. In the SSCP analysis of exon 2, B55.1, B55.2 and B56 showed identical patterns, whereas B54 and B22N showed distinct patterns. In the SSCP analysis of exon 3, B54 and B55.1 showed identical patterns, whereas B55.2, B56 and B22N showed distinct patterns. Sequence analysis further showed that B54, B55.1 and B56 were encoded by B^*5401, B^*5502 and B^*5601, respectively. The other two rare antigens, B55.2 and B22N, were found to be encoded by new alleles. The exon 2 amino acid sequences encoded by B55.2 and B22N alleles were identical to those of B55- and B56-group alleles, although that of B22N included a synonymous substitution. The exon 3 sequence of B55.2 was unique among HLA-B alleles. The exon 3 sequence of B22N was identical to those of B^*1501 and B^*4601, but distinct from those of other B22-group alleles.     

False serological assignment of HLA B5 CREG alleles in African subjects as revealed by a high resolution DNA typing

In view of the growing importance of HLA class I antigens in matching for transplantation and of the difficulties to type several alleles by serology in African subjects, mostly those belonging to the B5 CREG, we have compared results obtained by a novel high resolution DNA typing with those of serology in 240 samples from Senegalese and Zairean individuals. High resolution PCR-SSO typing was designed by using a single set of B5 CREG-specific primers targetting codons M12 and L163/R239 and a combination of 52 SSO probes allowing extensive definition of 54 allelic variants of : B51 (N=5), B52 (N=2), B53, B35 (N=9), B78 (N=2), B70 (N=4), B57 (N=3), B58 (N=3), B44 (N=4), B45, B49, B50, B62 (N=7), B63 (N=2), B75 (N=2), B76 (N=3), B77, B56 (N=2) and B46. False serological assignment was demonstrated in 26% of the 90 Zairean cells and 17% of the 150 Senegalese cells. In 50 Caucasian cells and in 106 controls from the 12th IHWS, only 3% discrepancy was found. 88% of the discrepancies concerned B^*5201 which was erroneously typed as B51, B52 or B53; B^*7801 as B35 or B70; B^*5201 as B5102, B53 or B51 and B^*1503 as B7801 or B35. Splitting of HLA B15, B70 and B17 could be achieved with the present DNA typing protocol. It was also found that the actual frequencies of some African alleles was underestimated by the 11th IHWS (e.g.: B^*5301 (15%) vs B53 (4,2%); B^*7801 (16%) vs B7801 (5,2%). In view of its high degree of complexity, we conclude that accurate B5 CREG definition requires high resolution DNA typing.     

HLA-B44 subtyping in a Spanish population: further evidence of Caucasian population diversity

HLA-B44 is the most frequent HLA-B allele in Caucasian populations. Several B44 subtypes, B*4402-B*4406, have been identified in individuals with this ethnic origin. Mismatches among B44 subtypes have been described as major targets for allogeneic responses in bone marrow transplantation. We have developed a PCR-SSO method, based on a B12- specific DNA amplification of exon 2 through exon 3 and subsequent non radioactive hybridization with eight probes, which allow us to discriminate all B12 homozygous combinations. We applied this method to determine the frequency of B44 subtypes in a Spanish population, as well as their HLA-A.-C.-DRB1,-DRB3/DRB4/DRB5.-DQA1 and -DQB1 associated haplotypes. A total of 141 healthy unrelated Spanish individuals and 31 B44-bearing haplotypes were investigated. Four B44 alleles were identified, B*4402 (33%), B*4403 (66%), B*4404 (0.7%), and B*4405 (0.7%). Haplotype analysis showed a clear differentiated distribution pattern for the two major B44 subtypes. B*4402 is associated with Cw5 (11/13) and A2 antigens (10/13). In contrast, B*4403 is mainly found together with DRB1*0701 (14/16). An inverted B*4402/B*4403 frequency in comparison with other European and North American Caucasian populations, revealed the existence of an extended haplotype diversity between populations of the same ethnic origin. Apart from anthropological studies, high resolution typing for HLA class I antigens presenting molecular polymorphism will be of great relevance in unrelated bone marrow transplantation.     

Epitope mapping of human monoclonal antibodies to HLA-B27 by using natural and mutated antigenic variants

The epitopes defined by three human monoclonal antibodies (mAbs) (Tr3B6, TrCG10, TrBH12) against HLA-B27 have been mapped by flow cytometry. For this purpose we used murine transfected cells expressing at their surface hybrid antigens between HLA-B7 and -B27 and, in addition, Epstein-Barr virus cells lines expressing the six HLA-B27 alleles B^*2701 to B^*2706. The results indicated that the mAbs are domain specific. TrBH12 recognizes the first external (alpha-1) domain. Residues critical for the TrBH12 epitope are located in the alpha-1 helix and include the polypeptide stretch 63–76 plus a critical amino acid at position 77. Tr3B6 binds the second external (alpha-2) domain, and one mutation (VAL¹⁵²→GLU¹⁵²) destroyed its epitope. TrCG10 also binds the alpha-2 domain.     

Role of strong anchor residues in effective binding of 10-MER and 11-MER peptides to HLA-A2402 molecules.

Analysis of the structural requirements for the interaction of antigenic peptides with HLA-A24 molecules are very important for studies of T cell recognition of various antigens, because HLA-A24 (A^*2402) is most common HLA-A allele in the world, especially in Oriental population. In order to precisely investigate the interaction of peptides with HLA-A24 molecules beyond previous analysis of self-peptides eluted from HLA-A24 molecules, we examined the A^*2402 interaction of 172 chemically synthesized 8-mer to 11-mer peptides carrying two residues (Try and Phe) at P2 and four residues (Phe, Trp, Leu and Ile) at their C-terminus by the use of stabilization assay. The results were statistically analyzed to assess the influence of anchor residues on peptide binding. The length of peptides (9- to 11-mer) did not affect A^*2402 binding except 8-mer peptides. Peptides possessing the aromatic residues at their C-terminus bound to A^*2402 molecules stronger than those bearing the aliphatic hydrophobic residues. These results indicate that two aromatic hydrophobic anchor residues permit the binding of longer peptides to A^*2402 molecules. Compared to our recent studies of B^*3501 and B^*5101 binding peptides, the present study suggested that both B and F pockets of A^*2402 molecules might be large and deep because these pockets favored bulky aromatic residues.     

Heterogeneity of HLA-A2 in Asia-Oceanic populations

HLA-A2 is one of the most common yet most diversified HLA antigens with 17 subtypes so far identified at the molecular level. A2 subtyping may have significant impact on clinical medicines. We developed a PCR/SSO-based comprehensive typing protocol for HLA-A and investigated the distribution of A2 alleles in regional ethnic groups. A2 was detected with high frequencies in most study populations. A total of 480 A2+ samples were identified and subtyped. The gene frequencies of A2 ranged from 34% in Chinese, 29% in Australian Caucasoids, 21% in Polynesians, 14% in Javanese and 13% in Australian Aborigines. However, in Melanesians and Micronesians A2 was absent. Six A2 alleles were found in the present experiments including A^*0201, 0203, 0205, 0206, 0207 and 0210. In Aborigines all the A2+ donors were typed as 0201. In Caucasoids A^*0201 accounted for 95% of A2+ samples though other three subtypes A^*0203, 0205 and 0207 were also detected. Extraordinary A2 heterogeneity was observed in Asia-Pacific populations where A^*0201 has become a minority. In Chinese all the six A2 alleles were discovered with A^*0201, 0203, 0206 and 0207 as the four major ones. In Javanese A2 was equally divided into A^*0201, 0203 and 0206 while in Polynesians A2 is overwhelmingly dominated by the oriental A^*0206 (71%). Our study also showed that comprehensive DNA matching for A2 would eliminate most A mismatches in the unrelated-donor transplantation in study populations.