An extensive polymerase chain reaction-allele-specific polymorphism strategy for clinical ABO blood group genotyping that avoids potential errors caused by null, subgroup, and hybrid alleles.

BACKGROUND: ABO genotyping is complicated by the remarkable diversity at the ABO locus. Recombination or gene conversion between common alleles may lead to hybrids resulting in unexpected ABO phenotypes. Furthermore, numerous mutations associated with weak subgroups and nondeletional null alleles should be considered. All known ABO genotyping methods, however, risk incorrect phenotype predictions if any such alleles are present. STUDY DESIGN AND METHODS: An extensive set of allele-specific primers was designed to accomplish hybrid-proof multiplex polymerase chain reaction (PCR) amplification of DNA fragments for detection of ABO alleles. Results were compared with serologic findings and ABO genotypes defined by previously published PCR-restriction fragment length polymorphism/PCR-allele-specific polymorphism (ASP) methods or DNA sequencing.
RESULTS: Phenotypically well-characterized samples from blood donors with common blood groups and rare-subgroup families were analyzed. In addition to the commonly encountered alleles (A1, A1(467C>T), A2, B, O1, O1v, and O2), the new method can detect hybrid alleles thanks to long-range amplification across intron 6. Four of 12 PCR-ASP procedures are used to screen for multiple infrequent subgroup and null alleles. This concept allows for a low-resolution typing format in which the presence of, for example, a weak subgroup or cis-AB/B(A) is indicated but not further defined. In an optional high-resolution step, more detailed genotype information is obtained.
CONCLUSION: A new genotyping approach has been developed and evaluated that can correctly identify ABO alleles including nondeletional null alleles, subgroups, and hybrids resulting from recombinational crossing-over events between exons 6 and 7. This approach is clinically applicable and decreases the risk for erroneous ABO phenotype prediction compared to previously published methods.