Definitive high resolution typing of HLA‐E allelic polymorphisms: Identifying potential errors in existing allele data

A set of robust PCR-SSP reactions were developed for each of the five polymorphic sites that define the five alleles of the HLA class Ib gene, HLA-E. This method was developed using 28 homozygous cell lines and further tested in a sample of African-Americans, a sample of Japanese, and a core panel of cell lines compiled for the 13th International Histocompatibility Workshop. Three alleles were found in each of these four sample groups, HLA-E*0101 (64.29, 50.00, 32.00 and 56.58%, respectively), *01031 (5.36, 20.65, 39.00 and 18.42%) and *01032 (30.35, 29.35, 29.00, and 25.00%). HLA-E*0102 was not detected in any of these samples nor in the cell line, LCL 722.221, in which this allele was originally described. HLA-E*0104 was not found either. This latter allele was originally reported in Japanese at a frequency of 1/22 (4.5%), which should have been high enough to have resulted in multiple occurrences of the *0104 allele in the samples tested in this study. We propose that the existence of the HLA-E*0102 and E*0104 alleles should be questioned.