Accurate estimates of HLA-A, B, DR and DQ phenotype, gene and haplotype frequencies (HF) in the normal population are of importance in, for example, disease susceptibility studies, platelet transfusion support and transplantation. HLA population genetics studies have been performed on numerous groups, however, no major studies have been carried out on the population of Wales. As part of the validation process for our routine HLA-A and B typing by PCR using sequence-specific primers (PCR-SSP) we examined 1,798 normal, unrelated Caucasoid blood donors living in Wales and recruited onto the Welsh Bone Marrow Donor Registry (WBMDR). Typing was performed by serology (HLA-A, B) and PCR-SSP at low resolution (HLA-A, B, DR, DQ) resulting in a particularly rigorous level of HLA specificity assignment. Four discrepancies were found between the HLA-A and B serological and PCR-SSP specificity assignments: (1) two instances of HLA-A2 by serology were undetected by PCR-SSP and were a new HLA-A2 allele - A*0224; (2) one example of HLA-B*15 by PCR-SSP failed to react by serology, and remained undetectable by serology in subsequent samples, and (3) one example of HLA-B45 by serology was identified as HLA-B*5002 by PCR-SSP. Hardy-Weinberg and homozygosity analysis showed that the goodness-of-fit was excellent (p > 0.05). for both phenotype distribution and the number of homozygotes identified, for all four loci. The phenotype and gene frequencies for the 18 HLA-A, 34 -B, 15 -DR and 8 -DQ specificities identified and two- and three-locus HF, linkage disequilibrium and related values for HLA-A/B, B/DR, DR/ DQ and HLA-A/B/DR and B/DR/ DQ were essentially typical of a northern European population. HLA-A2, B44, DR4 and DQ2 were the highest frequency phenotypes and HLA-A2403, A34, A74, B42, B75, B2708, B48, B67 and B703 occurred once only. There were no examples of: A36, A43, A69, A80, B46, B54, B59, B73, B76, B77, B7801, B8101 or DR18 specificities. DR17, DQ2 and Al, B8, DR17 were the highest frequency two- and three-locus haplotypes identified. Diverse HLA-A, B, DR phenotypes were identified in 87.0% (1,564) of subjects. When HLA-DQ was also considered, different four locus phenotypes were identified in 89.1% (1,602) of subjects. This frequency information will be beneficial as a high-quality reference control for disease susceptibility studies and in calculating the chances of identifying a bone marrow donor in a patient's extended family. This process was successful for the validation of our HLA-A and -B PCR-SSP typing procedure and the findings suggest an accurate level of specificity assignment of WBMDR panel donors who had previously been typed by serology alone.
