Mary Ellexson

CHARACTERIZATION OF AN HLA-A26 SEROLOGIC VARIANT

Currently there are over 300 HLA class I alleles and the number continues to grow (Bodmer et al. 1997). HLA class I molecules are the most polymorphic molecules in humans, and polymorphic positions within are primarily located within the antigen binding groove of the molecule which is encoded by exons 2 and 3 of the class I gene (Parham 1995). Concentration of diversity within the antigen binding groove allows the presentation of a vast array of antigenic peptides to CD8+ T cells (Rammensee et al. 1995). While the vicissitude of HLA genes assists in the survival of species through the pressures of natural selection, it creates a significant barrier to the transplantation of organs, tissue, and bone marrow. HLA class I typing for transplantation has traditionally relied upon alloantisera to distinguish among many different class I molecules. Antibodies contained within alloantisera recognize epitopes unique to individual or familial class I molecules, with antibody epitopes typically located in solvent accessible regions atop the alpha helices of the HLA class I heavy chain (Parham et al. 1992). It is therefore common that new polymorphisms detected via serology are either located atop helices or located within solvent accessible loops connecting beta sheets on the floor of the groove. For example, HLA-A*2603, A*2604, A*2605, A*2606, and A*2607 all differ from A*2601 by a mutagenic event in an a-helix and can be distinguished from A*2601 by serologic methods (Arnett 1996; Ishikawa 1994; Maruya 1996; Miyashita 1995). The exception to this pattern of A26 a-helical polymorphism altering serologic recognition is A*2602, where a difference at 116 in the F specificity pocket alters the peptides bound such that antibodies detect a conformational change in the class I heavy chain (Ishikawa 1994). Here, the nucleotide sequence of an A26 serologic variant of Korean ethnicity is reported. Serologic typing of the HLA-A locus for this cell, 034-SEA-HK, indicated HLA-A11, -A10v (A26 short), prompting high-resolution sequence-based typing (SBT). Given that this A26 variant was detected during serologic typing, and that most members of the A26 family differ from A*2601 by a single mutagenic event, we hypothesized that a small cluster or a single amino acid substitution atop the alpha-1 or alpha-2 a-helices would separate this Korean A26 allele from A*2601. Exons 2 and 3 of HLA-A loci were amplified from genomic DNA using HLA-A locus-specific primers as previously described by Cereb and co-workers (1996) yielding a product of 940 base pairs (bp). The first round polymerase chain reaction (PCR) product was then diluted 1:100 and used as template for nested PCR reactions which amplify exons 2 and 3 separately. Bidirectional sequencing of each exon followed using a CY5 dye-labeled ±21mer M13 Universal primer and an AutoLoad Kit (Amersham Pharmacia Biotech, Piscataway, N.J.). Sequencing reactions were loaded onto a 6% Page Plus (Amresco, Solon, Ohio) gel and run on a Pharmacia ALFexpress automated DNA sequencer. Data were analyzed using Pharmacia HLA SequiTyper software (version 2.0) and a novel polymorphism was detected at nucleotide 453 of exon 2. As the SBT strategy produces heterozygous sequence, it was not certain whether the A11 or A26 allele possessed the new polymorphism. To clarify, genomic DNA was amplified as previously described (Cereb 1996) using HLA class I locus-specific primers and cloned into the blunt-end TA vector (Invitrogen, Carlsbad, Calif.) according to the manufacturers instructions. Ten white colonies were picked and grown overnight in 10 ml of LB media containing 50 mg/ml of ampicillin. Plasmid DNA were isolated using M.E. Ellexson ? R.L. Peck ? W.H. Hildebrand ( ) Department of Microbiology and Immunology, University of Oklahoma Health Sciences Center, P.O. Box 26901, Oklahoma City, OK 73190, USA

Polymorphism at codons 114, 116, 145, and 163 muddle the typing of HLA-B*1304.

Genetic exchanges often muddle the typing of HLA class I molecules, this is also the case for HLA-B*1304. Serologic and molecular DNA class I typing report a B15/B55 type for cell 847, whereas DNA sequencing finds B*5501/B*1304. HLA-B*1304 differs by no more than four amino acids from other HLA-B13 molecules, a comparative analysis of the B13 and B15 families was therefore performed to determine why serologic and molecular DNA approaches report a B15 type for B*1304. Comparisons demonstrate that limited differences individuate the B15 and B13 molecules such that the genetic recombination of codons 145 and 163 in the class I heavy chains alpha 2 alpha helix prompt B*1304 to exhibit a B15X21 pattern of serologic cross-reactivity. Molecular DNA class I typing approaches are also swayed by genetic recombinations to type B*1304 as a B15 molecule: B15-like nucleotide sequences encoding residues 114, 116, and 145, lead B*1304 to exhibit a B15 PCR amplification pattern. Thus, genetic exchanges encoding key amino acids in the class I heavy chain lead molecular and serologic typing approaches to categorize HLA-B*1304 as a member of the B15 family.

Molecular characterization of HLA-A*6803.

More than 200HLA class I and class II typing laboratories participate in the International Cell Exchange (ICE). Laboratories participating in the typing of HLA class I molecules as part of the exchange are sent four unknown samples each month upon which class I typing is completed. The participating laboratories then communicate the class I type identified, the data from all participants is combined, and a report assessing class I typing consistency from laboratory to laboratory is published. Laboratories participating in the ICE have historically determined a class I type using alloantibodies: however, the realization that serologic typing cannot discriminate among all class I subtypes has led to the implementation of molecular class I HLA typing methods. In the autumn of 1995, serologic typing indicated that cell 859 expressedHLA-A32 and -A68 [with no allelic subtypes assigned; Fig. 1 (Lau 1995)], while the consensus HLA-A type assigned by the various DNA-based methods was A*3201/A*68012. Because two of 14 DNA labs assigned anA68variant and four DNA labs failed to assign a type other thanA*3201 (data not shown), we cloned and sequenced the two HLA-A alleles from cell 859 under the hypothesis that a newHLA-Avariant was present in cell 859 (Domena et al. 1993; Ellexson 1996). Indeed, a new HLAA68 allele elusive to both serologic and DNA typing techniques was identified. A similarity search found this new HLA-A68 allele to be most homologous to A*68012, from which it differs at a single nucleotide (C instead of G) at position 282. In accordance with this high level of similarity to other A68 alleles, the WHO nomenclature committee (Bodmer et al. 1995) assigned the name A*6803 to the newHLA-A allele. The single nucleotide substitution which differentiates A*6803 from A*68012is a coding substitution; A*6803has a histidine at position 70 in theα1 domain of the class I heavy chain, whileA*68012 has a glutamine (Table 1). Because the side chain of amino acid 70 is positioned to affect the stereochemistry of specificity pockets B and C (Saper et al. 1991), the substitution of a positively charged histidine (A*6803) for an uncharged glutamine (all other A68s) may alter the peptides bound by these A68 molecules; no other pair of class I molecules differs only at amino acid 70 (Parham 1995), such that comparisons of peptides eluted fromA*68012 and A*6803 will help to elucidate the effect of polymorphisms at amino acid 70.