Detection of Low Levels of the Mitochondrial tRNALeu(UUR) 3243A>G Mutation in Blood Derived from Patients with Diabetes

Abstract

Mutations in the human mitochondrial genome have been suspected to play a significant role in the etiological development of mitochondrial diabetes. Detection of the 3243A>G mutation in the mitochondrial transfer RNALeu(UUR) gene (MTTL1), especially at low heteroplasmy levels, is highly desirable since it facilitates the diagnosis and subsequent management of the disease. The proportions of mutant mitochondrial DNA (mtDNA) can vary between tissues and are usually significantly higher in muscle than in blood, but muscle biopsies from patients with diabetes are rarely available. Here, we describe a technique that can not only determine the presence of MTTL1 3243A>G, but can also estimate the percentage of mutant DNA. The technique is based on the use of the WAVE system for the high-performance liquid chromatography (HPLC)-mediated analysis of mutation-specific restriction fragments derived from mutant PCR amplicons. PCR amplicon restriction fragment analysis by HPLC (PARFAH) can also be used for the detection of other mutations. This PARFAH analytical approach led to the discovery of the 3243A>G mutation in blood samples from a series of patients who had initially been reported to lack the mutation, even though matrilineal relatives had been shown to harbor the mutation associated with maternally inherited diabetes and deafness (MIDD) or mitochondrial myopathy encephalopathy lactic acidosis stroke-like episodes (MELAS) phenotypes. We have established that the PARFAH method can reliably detect as little as 1% mutant DNA in a sample, which would normally be missed by commonly used gel electrophoresis or sequencing methods. The PARFAH method not only provides a sensitive, high-throughput, and cost-effective strategy for the detection of low levels of mtDNA mutations in peripheral tissues, but also facilitates the estimation of the percentage of mutant DNA in the sample. The fact that samples can be readily obtained from peripheral tissues in many cases will avoid the need for invasive muscle biopsies. Our ability to detect low levels of mtDNA mutations in blood samples of carriers will allow us to reassess the prevalence of the MTTL1 3243A>G mutation in patients with diabetes.