Long-extension PCR to detect deleted mitochondrial DNA molecules is compromized by technical artefacts.

Long-extension PCR (LX-PCR), followed by Southern hybridization to probes for two different regions of the mitochondrial genome, was used to evaluate the presence of deleted mtDNA molecules in heart muscle samples from alcoholic cardiomyopathy patients compared with age-matched controls. Two different primer pairs capable of amplifying the entire genome, as well as a variety of other primer pairs predicted to amplify the genome in large, overlapping fragments, were tested. Products indicating the presence of a variety of subgenomic, deleted molecules were detected in variable amounts from patient and control myocardial samples alike. Most of these hybridized with a probe for the 16S/ND1 region, but not with a probe for the ND4/ND5 region that is commonly deleted. Dilution of a given template DNA in which deleted products were prominent resulted in the disappearance of the subgenomic bands in favour of the full-length, undeleted product. Therefore, the appearance and amount of such products is subject to template concentration or quality. The results indicate that the application of LX-PCR to the detection and quantitation of deleted mtDNAs is inherently unreliable, and findings using this technique should be treated with caution unless supported by an independent method.