Allelic drop-out and preferential amplification in single cells and human blastomeres: implications for preimplantation diagnosis of sex and cystic fibrosis.

Previously the diagnosis of sex and cystic fibrosis status has been studied on single cells using the polymerase chain reaction (PCR). It has been suggested that allelic drop-out (PCR failure of one allele) and/or preferential amplification (hypo-amplification of one allele) may contribute to poor reliability and misdiagnosis, although this remains controversial as some reports suggest that allelic drop-out does not occur. We investigated an improved method of diagnosing sex and cystic fibrosis in single cells using a new technology (fluorescent PCR) to determine the base level of PCR artefacts (allelic drop-out and preferential amplification) which, in combination with improved sensitivity, should improve PCR reliability and accuracy. Fluorescent PCR gives high reliability (approximately 97%) and accuracy rates (approximately 97%) in somatic cells for both sex and cystic fibrosis diagnosis and its lower detection threshold allows allelic drop-out and preferential amplification to be easily distinguished. We also achieved high reliability and accuracy in diagnosing cystic fibrosis in human blastomeres. This study confirms earlier reports of both allelic drop-out and preferential amplification in single cell analysis. We demonstrate that both allelic drop-out and preferential amplification occur in somatic cells and suggest these are separate phenomena. Preferential amplification appeared common in single cell PCR while allelic drop-out apparently occurred at random in each allele. Preferential amplification was mainly amplification of the larger allele. We suggest that some inaccuracy/misdiagnosis may be due to both preferential amplification as well as allelic drop-out. Other findings were variability in drop-out between PCR and that amplification of signals from human blastomeres may be linked to embryo quality. We suggest that allelic drop-out is dependent on the number of cells within the sample.