Search of a genomic sequence database for potential novel blood group antigens: Investigation into why some amino acid substitutions are not immunogenic.

Abstract

Polypeptide blood group antigens are typically identified through investigation of the antibodies they induce. Human genome sequence databases are a new tool to identify AA substitutions that potentially create blood group antigens. The Erythrogene genomic sequence database was searched for missense mutations not known to be blood group antigens in the extracellular domains of selected RBC proteins in European populations. Any mutations found with prevalence of 1%-90% and not known to have induced antibodies in transfusion practice were analyzed using protein structural analysis and epitope prediction programs to determine why they apparently lack immunogenicity. Thirteen missense mutations not known to create blood group antigens were identified in the extracellular domains of Kell, BCAM, and RhD proteins, but not in RhCE, Urea Transporter 1 (Kidd), Atypical Chemokine Receptor 1 (Duffy), glycophorin A or glycophorin B. While 11 of the 13 mutations had low prevalence (<1%), a Kell Ser726Pro substitution and a BCAM Val196Ile substitution had predicted phenotype prevalences of 43.2% and 5.7%, respectively. Ser726Pro had multiple properties of a linear B-cell epitope, but possible suboptimal protein location for B-cell receptor binding and limited T-cell epitope possibilities. Val196Ile was not predicted to be in a linear B-cell epitope. Multiple potential new blood group antigens of low prevalence were identified. Whether they are antigenic remains to be determined. Two higher prevalence variants in Kell and BCAM are unlikely antigens, otherwise their antibodies presumably would already have been identified. Possible reasons for their poor immunogenicity were identified.