Computational and experimental approaches for evaluating the genetic basis of mitochondrial disorders

Mitochondria are responsible for many fundamental biological pathways and metabolic processes, including aerobic ATP production by the mitochondrial respiratory chain. In humans, mitochondrial dysfunction can lead to severe disorders of energy metabolism, which are collectively referred to as mitochondrial disorders and affect approximately 1:5,000 individuals. These disorders are clinically heterogeneous and can affect multiple organ systems, often within a single individual. Symptoms can include myopathy, exercise intolerance, hearing loss, blindness, stroke, seizures, diabetes, and GI dysmotility. Mutations in over 150 genes in the mitochondrial DNA (mtDNA) and nuclear genome are known to cause mitochondrial diseases and an additional ~1,000 nuclear-encoded mitochondrial proteins have the potential to underlie mitochondrial disorders but have not yet been linked to human disease. As a result, determining a molecular diagnosis for patients with suspected mitochondrial disorders remains a challenge.;To improve our understanding of the genetic basis of mitochondrial disorders, we have developed a "Mito-Exome" sequencing approach that targets the mtDNA and the exons of nearly 1600 nuclear-encoded genes implicated in mitochondrial function, mitochondrial disease, or monogenic disorders with phenotypic overlap. We have applied this method to a diverse set of 102 patients referred to Massachusetts General Hospital (MGH) with clinical and/or biochemical suspicion of mitochondrial disease. In addition to identifying pathogenic mutations in genes previously known to cause mitochondrial disease (e.g. NDUFV1, POLG2), our study has implicated several new or unexpected genes, including WFS1, HSD17B4, DPYD, and ATP5A1. However, the genetic basis of disease remains unclear for the majority of our patients. Our study demonstrates the strengths and limitations of next-generation sequencing approaches for the molecular diagnosis of suspected mitochondrial disorders.;To infer the function of uncharacterized mitochondrial proteins implicated in our genetic studies, we have also developed a computational genomic approach utilizing protein homology and bacterial gene neighborhoods. This approach has implicated an uncharacterized mitochondrial protein, C6orf57, in a role in the mitochondrial respiratory chain. The computational and genomic approaches described in this dissertation seek to improve our understanding of mitochondrial biology and its relationship to human health and disease.