Mitochondrial energy generation disorders: genes and mechanisms

Thorburn DR1,2,3

  1. Murdoch Childrens Research Institute, Parkville, VIC 3052.
  2. Victorian Clinical Genetics Services, Parkville, VIC 3052.
  3. Department of Paediatrics, University of Melbourne, Melbourne, VIC 3052.

Inherited disorders of oxidative phosphorylation (OXPHOS) cause a clinically and genetically heterogeneous range of mitochondrial diseases. Mutations causing these disorders have been identified in 35 of the 37 genes encoded by mitochondrial DNA (mtDNA) and over 250 nuclear genes, with many still to be discovered. Apart from primary defects affecting OXPHOS subunits, assembly factors, electron carriers, and mtDNA maintenance or expression, many patients with mitochondrial disease have secondary defects related to enzyme cofactors, metabolite transport or other aspects of mitochondrial biogenesis, homeostasis and quality control. Several examples of recent collaborative studies illustrate the challenges of elucidating disease mechanisms. Firstly, we identified four families with Leigh syndrome with bi-allelic mutations that destabilized the MRPS34 protein, located in the foot of the small 28S mitoribosome subunit. Ribosome profiling and quantitative proteomic analyses of patient fibroblasts showed this led to loss of most of the 30 proteins of the 28S subunit, relative sparing of the large subunit, failure to assemble actively translating mitoribosomes and loss of many subunits of OXPHOS complexes I and IV. Secondly, we identified large deletions and gene conversions in a highly repetitive chromosomal region refractory to conventional analyses. The resulting ATAD3B/ATAD3A gene fusions cause lethal pontocerebellar hypoplasia and lead to mtDNA aggregation in patient fibroblasts, with multiple indicators of altered cholesterol metabolism. ATAD3 appears to be a key protein linking formation of mitochondrial nucleoids in cholesterol-rich mitochondrial inner membrane domains to cellular cholesterol metabolism. A final example is the SURF1 gene, in which patients typically have bi-allelic knockout mutations and suffer severe neurological and other symptoms but mouse knockouts show no clear phenotype. We used gene editing to generate human embryonic stem cell SURF1 knockouts. These differentiate well to cardiomyocytes and show marked abnormalities of OXPHOS function and calcium signaling as well as decreased contractility in 3D organoids.