There are three main alleles in the ABO blood group system, A, B, and O. The former two alleles encode glycosyltransferases resulting in the wild-type A and B phenotypes, whereas the latter allele does not encode a functional enzyme owing to a frameshift polymorphism in the majority of cases. Thus the group O phenotype is the absence of A or B sugars. More than 15 years ago the O2 allele was described; this allele did not feature the usual crippling 261delG polymorphism, which up to that point was the hallmark of an allele encoding group O, but instead had several other nucleotide polymorphisms that reduced or eliminated the activity of its resulting protein. The classification of this type of allele as encoding group O has been called into question of late as some individuals with an O2 allele appear to have a weak A phenotype. Others with the same allele do not demonstrate any A antigens on their RBCs but might be involved in reverse typing discrepancies. Even within the same pedigree these alleles do not necessarily produce a consistent phenotype. This paper will summarize the detailed biochemical and population-based evidence both for and against the O2 allele’s ability to create A antigens or the absence of anti-A in plasma. Immunohematology 2008;24:138–147.
