In silico analysis for predicting pathogenicity of five unclassified mitochondrial DNA mutations associated with mitochondrial cytopathies' phenotypes

Abstract Mitochondrial DNA (mtDNA) mutations have been assigned as a major cause of genetic disease. When a novel sequence variation is found, it is necessary to evaluate its functional impact, usually requiring functional molecular studies. Given the fact that this approach is difficult to put in practice in a routine basis, it is possible to take advantage of the in silico tools available and predict protein/RNA structure changes and therefore pathogenicity. Here, we describe the characterization of five undescribed mtDNA variants, upon detection of 23 unclassified alterations at Laboratory of Biochemical Genetics, from 2004 to 2014. Those five sequence variations are located in protein-coding genes, in five patients with a diverse range of mitochondrial respiratory chain disease phenotypes including encephalopathy, optic neuropathy, developmental delay, deafness and epilepsy. According to the prediction established by in silico analysis using tools to predict structure and function changes (ClustalW2® , PolyPhen-2® , SIFT® , MutationAssessor® , PredictProtein® , Provean® , I-TASSER® , Haplogrep® ), from the 23 variants analyzed, the five described are potentially pathogenic. This approach is inexpensive and compatible with a rapid first line response to clinical demanding, contributing to a more rationale genetic diagnosis concerning novel mutations and to clarify the mtDNA involvement in these pathologies.