HLA-B Alleles Associated with the B15 Serologically Defined Antigens

HLA-B Alleles Associated with the B15 Serologically Defined Antigens

HLA-B Alleles Associated with the B15 Serologically Defined Antigens Noriko Steiner, Jennifer Ng, Janet Bush, Robert J. Hartzman, Leslie Johnston-Dow,...

568KB Sizes 0 Downloads 26 Views

HLA-B Alleles Associated with the B15 Serologically Defined Antigens Noriko Steiner, Jennifer Ng, Janet Bush, Robert J. Hartzman, Leslie Johnston-Dow, and Carolyn Katovich Hurley ABSTRACT: Cells expressing HLA molecules in the B15 family were identified by serologic typing in routine testing of volunteer donors of various ethnic backgrounds for a bone marrow registry. DNA sequencing was used to identify HLA-B15 alleles associated with each serologic type and to examine the diversity within the B15 antigen family. Alleles which appeared predominantly in each B15 serologic cluster included: B15 with no defined serologic subdivision (B*1501), B62 (B*1501), B63 (B*1516, B*1517), B75 (B*1502, B*1521), and B76/77 (B*1513). Other B*15 alleles were also found associated

INTRODUCTION The assignment of B*15 alleles to serologic specificities, B62, B63, B75, B76, and B77 (Table 1) [1–3] has been complicated by crossreactivity among the antigens encoded by these alleles and crossreactivity with antigens in the B70 (B71, B72) and B46 groups [2–12]. Thus, an allelic product such as that encoded by B*1508 might be assigned as B62 in one study [1] and as B75 in another study [2]. The difficulties in routinely assigning unique serologic types to B15 antigens creates problems particularly within volunteer bone marrow donor registries where patients search for an HLA match in a registry of potential donors typed by many different serology laboratories. Over 30 B*15 alleles encode antigens in the B15 family. Also included in the B*15 cluster due to their high sequence homology are alleles, B*1503, B*1509, From the Department of Microbiology and Immunology, Georgetown University Medical School, Washington, DC, USA (N.S., C.K.H.), Department of Pediatrics, Georgetown University Medical School, Washington, DC, USA (J.N., J.B.), Naval Medical Research Institute, Bethesda, MD, USA (J.N., R.J.H.) and PE-Applied Biosystems, Foster City, CA, USA (L.J-D.). Address reprint requests to: Carolyn Katovich Hurley, Ph.D., E404 Research Building, Georgetown University Medical Center, 3970 Reservoir Rd NW, Washington, DC 20007, USA. Received June 30, 1997; accepted July 29, 1997. Human Immunology 56, 84 –93 (1997) © American Society for Histocompatibility and Immunogenetics, 1997 Published by Elsevier Science Inc.

with the serotypes and some of these alleles (e.g., B*1501 and B*1516) were found in two or more serologic clusters illustrating the complexity of this family. The B15 unsplit and B75 groups were the most complex exhibiting 16 and 7 alleles, respectively, within each serotype. Five new B*15 alleles (B*1530, B*1531, B*1533, B*1534, B*1535) and 5 other new HLA-B alleles (B*38022, B*3910, B*4010, B*51012, and B*5108) were also identified. Human Immunology 56, 84 –93 (1997). © American Society for Histocompatibility and Immunogenetics, 1997. Published by Elsevier Science Inc.

B*1510, B*1518, which encode antigens in the B70 family [5, 8]. As observed for other HLA alleles, the B*15 alleles are likely to have arisen by gene conversionlike events [13] which have created many of the difficulties in serologic distinction of the encoded molecules. In this study, cells expressing molecules in the B15 family were identified by serologic typing in routine testing of volunteer donors for a bone marrow registry. DNA sequencing was used to identify HLA-B15 alleles associated with each serologic type and to examine the diversity within the B15 antigen family.

MATERIALS AND METHODS Peripheral blood lymphocytes obtained from healthy unrelated volunteers from a U.S. bone marrow registry were typed using standard microcytotoxicity assays with commercial (One Lambda, Inc., Canoga Park, CA; Gen Trak, Inc., Plymouth Meeting, PA; GTI, Inc., Brookfield, WI) and local trays. Locus- and group-specific intron primers (5BIn1-ta, 5BIn1-cg, and 3BIn3-37) were used to amplify selectively both exons 2 and 3 of one or both alleles from the HLA-B locus for sequence analysis [14]. In some instances, M13 sequences (M13f: 59TGT AAA ACG ACG 0198-8859/97/$17.00 PII S0198-8859(97)00112-2

HLA-B15 Diversity

85

TABLE 1 Serologic Assignment of B*15 Allelesa Serologic Type

Associated Alleles

B62(15), Bw6

1501, 1504, 1505, 1506, 1507, 1508, 1515, 1520, 1524, 1525, 1527, 1532 1516, 1517 1502, 1508b, 1511c, 1521, 1530d, 1531d 1512, 1514, 1519 1513 1508b, 1511, 1521b, 1528, 1529, 1533d, 1534d, 1535d 1522, 1523, 1536, 1537 1510, 1518 1503 1509 1526

B63(15), Bw4 B75(15), Bw6 B76(15), Bw6 B77(15), Bw4 B15 Undefined B71(70), Bw6 B72(70), Bw6 B70, Bw6 Null a

Assigned by the WHO Nomenclature Committee [1] except where noted. [2]. c [3]. d Described in this study. b

GCC AGT and M13r: 59CAG GAA ACA GCT ATG ACC) were added to the primers. Single alleles that could not be readily isolated by PCR amplification were characterized by direct sequence analysis of both alleles simultaneously. Cloning using the PCRII vector (Invitrogen, San Diego, CA) was utilized to separate alleles in a few cases. After purification of the amplicon with Microcon 100 columns (Amicon, Inc., Beverly, MA), the DNA template (;100 ng) was used in an AmpliTaq DNA polymerase-FS (PE-Applied Biosystems Inc., Foster City, CA) sequencing reaction following the protocols (dye primer and dye terminator) described by the manufacturer. Sequencing primers utilized included M13f, M13r, the amplification primers, and forward and reverse primers located in intron 2 (Int2f: 59TTACCCGGTTTC ATTTTCAG; Int2r: 59GGATCTCGGACCCGGAG; B5.5 g/t 59 CAGGAGGGGGACGG(G/T)GCTGAC CG). The sequencing reaction was performed in a GeneAmp PCR Systems 9600 (Perkin-Elmer Corp., Foster City, CA) thermocycler. The sequencing product was isolated by ethanol precipitation and was electrophoresed on a Long Ranger gel (FMC BioProducts, Rockland, ME) using a 373A or 377 automated DNA sequencer (PEApplied Biosystems, Inc.). The sequences of both strands of exons 2 and 3 were analyzed using Factura, Sequence Navigator, Autoassembler, and a beta test version of MatchMaker software (PE-Applied Biosystems, Inc.). The sequences of previously undescribed alleles were obtained from hemizygous sequencing ladders obtained from both DNA strands from at least two different PCR amplifications. The GenBank accession numbers and WHO nomenclature assignments assigned to the novel alleles are listed in Table 2.

RESULTS Frequency of B15 in Unrelated Volunteers for a Bone Marrow Registry Serologic typing was utilized to identify individuals expressing HLA-B15 antigens from 18,967 consecutively typed volunteer donors for a U.S. bone marrow registry. Approximately 12% of these donors expressed B15 antigens as defined by serologic testing including B62 (10.3%), B63 (1.1%), B75 (0.2%), B76/77 (0.04%) and B15 (0.8%). The latter assignment, B15, includes those typings in which a B15 was assigned but no further serologic subdivision could be identified (B15 “unsplit”). The ethnic backgrounds of the B15 positive volunteers included Caucasian (73.2%), Hispanic (4.7%), African American (5.7%), Asian-Pacific Islander (5.6%) and Native American (8.2%). [The ethnic backgrounds of the remaining 2.6% of the individuals were not available.] Characterization of the Alleles in the B15 Unsplit Group Individuals (84) were selected who expressed a B15 serologic type which could not be further defined. The ethnic backgrounds of these individuals are described in Table 3 and Fig. 1. The HLA-B alleles carried by these individuals were identified by DNA sequence analysis of exons of 2 and 3 and the results are shown in Fig. 1. The majority of individuals (50/83) carried the B*1501 allele which is expected to serologically type as B62. Other B*15 alleles identified included B*1502, B*1504, B*1507, B*1515, B*1516, B*1517, B*1521, and B*1524. The serologic types assigned to each of these alleles by previous studies include B62, B63, and B75 (Table 1). A single sample carried 2 B*15 alleles, B*1516 and B*1503. Based on the serologic assignment

86

N. Steiner et al.

TABLE 2 New Alleles Identified in the Study Cell Name

Ethnic Origin

Serologic Type

Alleles (Date Assigned by WHO)a,b

GN00104 GN00108 GN00110

Hispanic Hispanic African American

B15, B39 B15, B35 B15, B39

B*1530 (4/96), B*39062 B*1530 (4/96), B*3512 B*1531 (4/96), B*3910 (4/96)c

GN00103 GN00105 GN00106

Caucasian Caucasian Asian-Pacific Islander

B15, B8 B15, B7 B15, B51

B*1533 (4/96), B*0801 B*1534 (4/96), B*0702c B*1535 (4/96), B*51012 (4/96)b

GN00155 GN00160 GN00109

Asian-Pacific Islander Asian-Pacific Islander Caucasian

B75, B38 B77, B60 B15, B51

B*1521, B*38022 (4/97) B*1513, B*4010 (6/96) B*1501, B*5108 (4/96)

GenBank Accession No. for New Allelesc U49900, U49901 U52171, U52172 U52173, U52174 (B*15); U52175, U52176 (B*39) U49898, U49899 U49902, U49903 U52167, U52168 (B*15); U52169, U52170 (B*51) U90240, U90241 U93915, U93916 U52815, U52816

a

New alleles are bolded. Other HLA-B alleles in cells were defined by sequencing and were consistent with the serologic typings. Information on alleles previously reported is reviewed by Bodmer et al. [1]. c The 2 accession numbers indicate the sequences for exon 2 and exon 3, respectively. b

(B15, B70), the B*1516 allele was assigned to the B15 serotype. Another sample exhibited two possible allelic combinations, B*1524, B*5001 or B*1501, B*4901. In this individual, the alleles assigned were based on the serologic type (B15, B50) of the donor. There were 3 donors in which the B15 unsplit assignment was contributed by an allele other than B*15. These alleles fell within the 5C CREG which contains B15 [15]. One individual typed as B15, B40 was found to carry B*5101 in combination with B*4006 (B61(40)). Likewise, two B15,B77 donors were found to carry B*52011, B*1513 (B77) or B*3503, B*1516 (B77). Five new B*15 alleles were also identified (described below). In the course of the study, 3 donors were identified who received a B“X” assignment indicating serologic reactivity that could not be assigned to a specific type accompanied by another serologic type in the B15 family (B63 or B75). A second B*15 allele was associated with the B“X” in these individuals. Based on the expected

associations (B63 with B*1516; B75 with B*1502, B*1508) (Table 1), the B*15 alleles associated with B”X” included B*1510 (2 examples) and B*1513. B*1510 is usually associated with B71 while B*1513 is associated with B77. New Alleles within the B15 Unsplit Group DNA sequencing of exons 2 and 3 from 2 unrelated Hispanic individuals, GN00104 and GN00108, and a Caucasian identified a new allele, B*1530 (Table 2, Fig. 2). Overall, the B*1530 allele differs from the most closely related sequence, B*1501, by 2 nucleotide substitutions resulting in 2 amino acid differences in the a2 domain (codons 114 (D (B*1501) 3 N(B*1530)) and 116 (S 3 Y)). [Codon 1 specifies the first amino acid of the mature polypeptide.] DNA sequencing of exons 2 and 3 from individual, GN00110, identified a new allele, B*1531. Overall, the B*1531 allele differs from the most closely related sequence, B*1505, by 3 nucleotide

TABLE 3 Ethnic Background of B*15 Positive Donors Characterized by DNA Sequence Analysis Ethnic Background

B15 Unsplita

B62

B63

B75

B76/77

Caucasian Hispanic African American Asian-Pacific Islander Native American Other/Undefined

51 5 10 8 9 1

30 2 1 3 4 4

15 2 7 1 3 1

2 2 1 25 0 1

1 0 1 6 0 0

1

Total

84

44

29

31

8

5

1

1

10

6

4

1

4

1

Number of cells with 2 B*15 allelesb a b

Donors carry B15 antigens which were not split by serologic typing. Some cells are represented more than once in this row.

B70 Unsplit

B71

B72

3 1 1

HLA-B15 Diversity

FIGURE 1 Graph showing the alleles identified in B15 unsplit serologic group and the ethnic background of the allele source. Cau, Caucasian; AfAm, African American; AsPI, Asian-Pacific Islander; His, Hispanic; AmInd, American Indian; Other, other and undefined. Alleles found in presumed homozygous individuals are counted twice.

substitutions resulting in 2 amino acid differences. These amino acid differences are located in the a1 (codon 63 (E 3 N)) and a2 (codon 113 (H 3 Y)) domains. DNA sequencing of exons 2 and 3 from individual, GN00103, identified a new allele, B*1533. Overall, the B*1533 allele differs from the most closely related sequence, B*1501, by 3 nucleotide substitutions resulting in 2 amino acid differences in the a2 domain (codons 131 (S 3 R) and 138 (T 3 K)) and a silent substitution at codon 135. DNA sequencing of exons 2 and 3 from individual, GN00105, identified a new allele, B*1534. Overall, the B*1534 allele differs from the most closely related sequence, B*1501, by 1 nucleotide substitution resulting in 1 amino acid difference. This amino acid difference is located in the a1 domain (codon 103 (V 3 L)). DNA sequencing of exons 2 and 3 from individual, GN00106, identified a new allele, B*1535. Overall, the B*1535 allele differs from the most closely related sequence, B*1501, by 1 nucleotide substitution resulting

87

in 1 amino acid difference. This amino acid difference is located in the a2 domain (codon 97 (R 3 T)). All of these differences with the exception of the B*1533 allele differences are found in other HLA-B alleles including B*15 alleles and are likely generated by intraallelic gene conversion events [13]. The new sequence motif found in B*1533 is found in C*0501, C*0802, and A*0218 and may represent an interlocus conversion event. (Previously reported class I sequences are described in reviews by Arnett and Parham [16] and Bodmer et al. [1].) Characterization of the Alleles in the B62 Group Individuals (44) were selected who had been serologically typed as B62 at the same time period as the individuals expressing an unsplit B15 type were characterized. Some of these individuals from various ethnic backgrounds (Table 3) were also selected to carry a second HLA-B specificity that was the same as that carried by the cells typed as unsplit B15 (e.g., B15,B7 and B62,B7). Although B62 has been assigned to a number of allelic products (Table 1), the majority of individuals (43/44) carried the B*1501 allele (Fig. 3). One Asian-Pacific Islander donor carried B*1507 associated with B62. Both alleles had been assigned to B62 in previous studies

88

FIGURE 2 DNA sequence of exons 2 and 3 of the new B*15 alleles. Previously reported class I sequences are described in reviews by Arnett and Parham [16] and Bodmer et al. [1].

(Table 1). Ten samples carried two B*15 alleles; 8 were B*1501 homozygotes. The assignments of the other two donors were based on the predicted association of B*1502 with B75 (B62,B75 typed donors carrying B*1501,B*1502 or B*1507,B*1502 alleles). Role of Second Expressed HLA-B Antigen in Serologic Assignment of B62 Due to the fact that 50 of the 84 individuals expressing an unsplit B15 antigen carried B*1501 and 43 of the 44 B62 individuals also carried B*1501, the possibility that the second expressed HLA-B allele may affect the serologic typing was explored. The HLA-B antigens expressed by individuals expressing an unsplit B15 type were compared to those from individuals expressing B62 (Fig. 4). No differences were observed in the general distribution of HLA-B types. Further characterization of the alleles encoding the second expressed HLA-B antigen suggested that the identity of the second antigen did not appear to affect the assignment of the serologic type (Table 4). For example, B*0702 was associated with B*1501 in both B15 unsplit samples (9) and B62 samples (7); B*3501 was associated with B*1501 in both B15 unsplit samples (4) and B62 samples (3).

N. Steiner et al.

Characterization of the Alleles in the B63 Group Only 3 alleles, B*1513, B*1516, and B*1517, were found in the 29 samples serologically typed as B63 (Fig. 3, Table 3). In previous studies, B*1516 and B*1517 had been assigned to the B63 group while B*1513 had been assigned to B77. Six samples carried 2 B*15 alleles— B63,B75 (B*1516,B*1503); B63,BX (B*1516,B*1510); B63,B70 (B*1516,B*1510); B63,B70 (B*1516,B*1503); B63,B70 (B*1517,B*1503); and B63,B77 (B*1513). The B*15 allele associated with B63 was assigned based on Table 1. Characterization of the Alleles in the B75 Group Seven alleles were found in the 31 individuals typing as B75 (Table 3, Fig. 3). Thirteen individuals carried the B*1502 allele and 10 carried the B*1521 allele. While B*1502 is assigned to B75, the serologic type of B*1521 is not defined by the WHO Nomenclature Committee (Table 1). The other alleles observed included B*1501, B*1508, B*1511, B*1515, and B*1535. Three of these alleles had previously been assigned to the B62 group (Table 1). Four samples carried two B*15 alleles— B75,B62 (B*1502,*1507); B75,B62 (B*1502,*1501); B75,BX (B*1508,*1510); and B75,BX (B*1502,*1513). The assignment of specific alleles to a serologic type was based on Table 1. Two samples gave two potential allele combinations: B75,B61 (B*1501,*4002 or B*1530,*4003) and B75,B62 (B*1502,*1501 or B*1515,*1525). Sequencing of the cloned products was

HLA-B15 Diversity

FIGURE 3 Graph showing the alleles identified in B62, B63, B75, and B76/77 serologic groups and the ethnic background of the allele source. Cau, Caucasian; AfAm, African American; AsPI, Asian-Pacific Islander; His, Hispanic; AmInd, American Indian; Other, other and undefined. Alleles found in presumed homozygous individuals are counted twice.

used to determine the correct assignments (indicated by the underline). Characterization of the Alleles in the B76/77 Group Serologic typing was not able to distinguish B76 from B77. HLA-B alleles from all of the individuals (8) expressing these types were characterized by DNA sequencing (Fig. 3, Table 3). Six of these individuals carried B*1513 which encodes the B77 specificity [1]. The sequence of both alleles for one of these samples, serologically typed as B77,B50 was ambiguous since both B*1513,B*5001 and B*1502,B*4901 give identical sequencing ladders. In this case, the allele assignment was based on the serologic types. One of these individuals, typed as B77,B63 appeared to be a B*1513 homozygote. All of these individuals were from Filipino extraction, the same ethnic background as the original source of the B*1513 sequence [2]. The two other B77

89

positive individuals included one Caucasian expressing B*1501 (together with B*0801) and an African American expressing B*1516 (together with B*3503). These alleles are normally associated with B62 and B63, respectively. Other Alleles Encoding Cross-Reactive Antigens B15 antigens are crossreactive with B70 (B71 and B72) and B46 antigens. Of the alleles assigned to the B70 serologic types, B*1503, B*1509, B*1510, and B*1518, none were observed associated with a B15 serologic type. Two of the B70-associated alleles, B*1503 and B*1510, were observed associated with B70, B72 and B“X” serologic assignments in 6 individuals (Table 3). The B*4601 allele was also not found associated with B15 serotypes. Three Asian-Pacific Islanders carried B*4601 in addition to B75-associated alleles; however, B46 was not detected by serology in two of these individuals. Other New Alleles Identified in the Study Forty four non-B*15 alleles representing 28 serologic antigens were also sequenced in the study. Five individuals carrying B15-associated types were found to carry a second HLA-B antigen which was encoded by a new allele. These alleles, B*38022, B*3910, B*4010,

90

N. Steiner et al.

FIGURE 4 Graph showing the identities of the second HLA-B antigens found in B15 unsplit and B62 positive cells. Antigens are defined by serology.

B*51012, and B*5108, are described in Table 2 and Fig. 5. DNA sequencing of exons 2 and 3 from individual, GN00155, identified a new allele, B*38022 which differs from the most closely related sequence, B*3802, by 1 nucleotide resulting in a silent substitution at codon 134. DNA sequencing of exons 2 and 3 from individual, GN00110, identified a new allele, B*3910 which differs from the most closely related sequence, B*39011, by 2

nucleotide substitutions resulting in 1 amino acid difference in the a1 domain (codon 67 (C (B*39011) 3 Y) and by a silent substitution at codon 58. DNA sequencing of exons 2 and 3 from individual, GN00160, identified a new allele, B*4010 which differs from the most closely related sequences, B*40011 and B*40012, by 4 nucleotide substitutions resulting in 3 amino acid differences in the a1 domain (codons 9 (H 3 Y), 24 (T 3 A), and 32 (L 3 Q)). Allele B*51012, found in GN00106, carries a silent substitution at codon 23. Allele B*5108, found in GN00109, differs from

TABLE 4 Sequence of the Second HLA-B Allele in Selected B*1501 Cells Serologic Type of Second Antigen

B15 Unsplit Associated Alleles (No. Identified)

B7

B*0702 (9)

B8 B65

B*0801 (13) B*1401 (1), B*1402 (3)

B35

B*3501 (4)

B51 B52

B*5101 (3), B*5108 (1) B*52011 (1)

B62 Associated Alleles (No. Identified) B*0702 (7), B*8101 (1) B*0801 (12) B*1401 (1), B*1402 (5) B*3501 (3), B*3503 (1) B*5101 (3) B*52011 (1)

HLA-B15 Diversity

FIGURE 5 DNA sequence of new non-B*15 alleles found in association in cells expressing B15 antigens.

B*51012 at codons 152 (E 3 V) and 156 (L 3 D) in the a2 domain. RESULTS Several studies suggest that allele level matching of HLA-DRB1 between donor and recipient in allogeneic bone marrow transplantation may improve outcome [17]. Furthermore, the isolation of cytotoxic T cells with specificity for allele level mismatches of HLA class I antigens in patients receiving a bone marrow transplant [18, 19] suggests that class I allele level matches may also be desirable in selecting the optimal bone marrow donor. To facilitate the selection of allele matched donors, this study has focused on defining the HLA-B alleles which encode antigens in the complex HLA-B15 antigen family. Such information will guide the initial selection of serologically matched donors with the ultimate aim of identifying an allele level match upon more extensive HLA typing. While, as a whole, many of the previous assignments of alleles to serologic types were observed (e.g., B62 encoded by B*1501; B63 encoded by B*1516; B75 encoded by B*1502), inconsistency in serologic assign-

91

ment was also apparent in this study. For example, B*1501 was found associated with 4 serotypes, B62, B75, B76/77, and B15 unsplit while B*1516 was associated with B63, B76/77, and B15 unsplit groups. Much of the inconsistency in assignment is likely due to the crossreactivity caused by sequence similarities among these HLA molecules. The increased use of DNA-based HLA typing for the characterization of new volunteer donors for bone marrow registries will alleviate this complexity [20]. One of the more interesting aspects of the study was the group of volunteer donors in which the B15 antigen could not be clearly defined (B15 unsplit). This group encompassed individuals of all ethnic backgrounds suggesting that B*15 alleles found in any population may be serologically complex. The B*1501 allele was the predominant allele in these individuals even though many other individuals with the same allele were serologically typed as B62. The inability to define the specific B15 split did not appear to be related to the second HLA-B antigen expressed by these individuals but could be due to day to day variation in the serologic typing procedure or reagents, the condition of the cell preparations, or the contribution of polymorphic peptides to the serologic type [21, 22]. Seventeen other alleles were found encoding B15 unsplit antigens; 12 had previously been de-

92

scribed and most were associated with other serologic types. The ability to select serologically typed donors as potential allele matches for a patient will vary depending on the diversity of alleles found in each serologic group. B62, B63, and B76/77 are relatively homogeneous regarding their allele composition in this population of U.S. volunteer donors while B75 and B15 unsplit are more diverse. Within a serologic group, the ethnic background of the potential donor might also guide selection. For example, within the B76/B77 group, the predominant allele, B*1513, was found in only Asian-Pacific Islanders. The two other B76/77 positive individuals were of a different ethnic background (Caucasian and African American) and carried alleles different from B*1513 (B*1501 and B*1516). Since the extent of HLA-B diversity is not yet well understood, DNA sequencing of exons 2 and 3 was employed to identify the alleles within the B15 samples. The utility of this sequencing approach was proven by the identification of 10 new alleles which may have been missed if another technique with less resolution had been used to define alleles. Individual HLA-B alleles from approximately half of the samples could be isolated for sequence analysis by group-specific amplification utilizing primer sets which detect an intron polymorphism. The remainder of the HLA-B alleles could not be individually isolated and were identified primarily by sequence analysis of both alleles simultaneously followed by analysis using allele-calling software. Although the majority of the HLA-B alleles could be identified with this approach, 3 ambiguous allele combinations were identified in which the polymorphisms detected could be contributed by 2 different allele combinations. These combinations were found in 4 out of 190 individuals studied (2%). In two instances, the serologic typing was utilized to predict the correct combination but, in the remaining samples, the alleles fell within the same serologic group and the sequencing of cloned alleles was required to clarify the typing. It is also possible that ambiguous allele combinations may also be present in the samples with clearly defined types; however, the alternative alleles involved in these combinations have not yet been identified. While direct sequencing is not applicable to large volume typing as required for a bone marrow registry, it is useful to define the relatively unexplored diversity of the HLA-B locus. ACKNOWLEDGMENTS

This research has been supported by funding from the Office of Naval Research N00014-94-1-0049 and N00014-95-1-0053 to the C.W. Bill Young Marrow Donor Recruitment and Research Program. The views expressed in this article are those of the authors and do not reflect the official policy or position

N. Steiner et al.

of the Department of the Navy, the Department of Defense, or the U.S. government. The authors would like to thank Lori Tull for assistance with the EBV transformation and Dr. S. Y. Yang for the HLA-B primer sequences.

REFERENCES 1. Bodmer JG, Marsh SGE, Albert ED, Bodmer WF, Bontrop RE, Charron D, et al. Nomenclature for factors of the HLA system, 1996. Tissue Antigens 49:297, 1997. 2. Hildebrand WH, Domena JD, Shen SY, Lau M, Terasaki PI, Bunce M, Marsh SGE, Guttridge MG, Bias WB, Parham P: HLA-B15: A widespread and diverse family of HLA-B alleles. Tissue Antigens 43:209, 1994. 3. Lin L, Tokunaga K, Tanaka H, Nakajima F, Imanishi T, Kashiwase K, Bannai M, Mizuno S, Akaza T, Tadokoro K, et al. Further molecular diversity in the HLA-B15 group. Tissue Antigens 47:265, 1996. 4. Choo SY, Fan L, Hansen JA: Allelic variations clustered in the antigen binding groove of HLA-Bw62 molecules. Tissue Antigens 37:108, 1993. 5. Domena JD, Little AM, Madrigal AJ, Hildebrand WH, Johnston-Dow L, Du Toit E, Bias WB, Parham P: Structural heterogeneity in HLA-B70, a high-frequency antigen of black populations. Tissue Antigens 42:509, 1993. 6. Lienert K, McCluskey J, Bennett G, Fowler C, Russ G: HLA class I variation in Australian Aborigines: characterization of allele B*1521. Tissue Antigens 45:12, 1995. 7. Martinez-Laso J, Layrisse Z, Gomez-Casado E, Montoya F, Gue´dez Y, Dominguez E, Varela P, Arnaiz-Villena A: A new HLA-B15 allele (B*1522) found in bari-motilones Amerindians in Venezuela: Comparison of its intron 2 sequence with those of B*1501 and B*3504. Immunogenetics 43:108, 1996. 8. Rodriguez SG, Bei M, Inamdar A, Stewart D, Johnson AH, Hurley CK: Molecular and serological characterization of HLA-B71 in association with different class I haplotypes or in different ethnic groups. Tissue Antigens 47:58, 1996. 9. Chandanayingyong D, Adams EJ, Arnett KL, Hildebrand WH, Parham P, Lau M: The presence of HLA-A*2403 and HLA-B*1512 on the same haplotype in a Thai family. Tissue Antigens 47:426, 1996. 10. Ellexson ME, Zhang GZ, Stewart D, Lau M, Teresi G, Terasaki P, Roe B, Hildebrand W: Nucleotide sequence analysis of HLA-B*1523 and B*8101—Dominant a-helical motifs produce complex serologic recognition patterns for the HLA-B“DT” and HLA-B“NM5” antigens. Hum Immunol 44:103, 1995. 11. Domena JD, Little AM, Arnett KL, Adams EJ, Marsh SG, Parham P: A small test of a sequence-based typing method: definition of the B*1520 allele. Tissue Antigens 44: 217, 1994. 12. Little AM, Parham P: The HLA-Bw75 subtype of B15:

HLA-B15 Diversity

13. 14.

15.

16. 17.

Molecular characterization and comparison with crossreacting antigens. Tissue Antigens 38:186, 1991. Parham P, Ohta T: Population biology of antigen presentation by MHC Class I molecules. Science 272:67, 1996. Cereb N, Yang SY: Dimorphic primers derived from intron 1 for use in the molecular typing of HLA-B alleles. Tissue Antigens 50:74, 1997. Rodey GE, Neylan JF, Whelchel JD, Revels KW, Bray RA: Epitope specificity of HLA Class I alloantibodies. I. Frequency analysis of antibodies to private versus public specificities in potential transplant recipients. Human Immun 39:272, 1994. Arnett KL, Parham P: HLA Class I nucleotide sequences, 1995. Tissue Antigens 46:217, 1995. Petersdorf EW, Longton GM, Anasetti C, Martin PJ, Mickelson EM, Smith AG, Hansen JA: The significance of HLA-DRB1 matching on clinical outcome after HLA-A, B, DR identical unrelated donor marrow transplantation. Blood 86:1606, 1995.

93

18. Fleischhauer K, Kernan NA, O’Reilly RJ, Dupont B, Yang SY: Bone marrow-allograft rejection by T lymphocytes recognizing a single amino acid difference in HLAB44. N Engl J Med 323:1818, 1990. 19. Keever CA, Leong N, Cunningham I, Copelan EA, Avalos BR, Klein J, Kapoor N, Adams PW, Orosz CG, Tutschka PJ, et al. HLA-B44-directed cytotoxic T cells associated with acute graft-versus-host disease following unrelated bone marrow transplantation. Bone Marrow Transplant 14:137, 1994. 20. Hurley CK: Acquisition and use of DNA-based HLA typing data in bone marrow registries. Tissue Antigens 49:323, 1997. 21. Catipovic B, Dal Porto J, Mage M, Johansen TE, Schneck JP: Major histocompatibility complex conformational epitopes are peptide specific. J Exp Med 176:1611, 1992. 22. Bluestone JA, Kaliyaperumal A, Jameson S, Miller S, Dick R, II: Peptide-induced changes in class I heavy chains alter allorecognition. J Immunol 151:3943, 1993.