Oral Presentation 39th Annual Lorne Genome Conference 2018

Whole Genome Sequencing for the Detection of Pathogenic Mitochondrial Mutations (#42)

Clare Puttick 1 , Kishore R Kumar 1 2 , Ryan Davis 1 2 , Marcel E Dinger 1 3 , Carolyn M Sue 2 , Mark J Cowley 1 3 4
  1. Kinghorn Centre for Clinical Genomics, Garvan Institute, Darlinghurst, NSW, Australia
  2. Department of Neurogenetics, Kolling Institute of Medical Research, Royal North Shore Hospital and University of Sydney, St Leonards, NSW, Australia
  3. St Vincent’s Clinical School, University of New South Wales, Darlinghurst, NSW, Australia
  4. Children’s Cancer Institute, Randwick, NSW, Australia

Mitochondrial diseases (MD) represent one of the most common groups of inherited metabolic disorders, however due to their wide-ranging phenotypic spectrum they are notoriously difficult to diagnose. There are hundreds of mitochondria per cell, and pathogenic mutations can be present in 0-100% of their genomes, known as heteroplasmy. We performed Whole Genome Sequencing (WGS) on 250 patients with known or suspected MD. The aim was to evaluate the ability of WGS to detect mitochondrial mutations in blood. This could potentially avoid invasive muscle biopsies and simplify the diagnostic paradigm for MD.

In a 30-45x WGS genome we observed 4000-8000x coverage in the mitochondrial genome. We developed a dedicated bioinformatic pipeline, mity, to comprehensively identify SNVs and INDELs down to 0.3% heteroplasmy. mity identified ~39 mitochondrial variants per patient with a heteroplasmy of >1%, 30% of which had a heteroplasmy <5%. We identified 71/79 known pathogenic mitochondrial variants identified prior to WGS, three of which had heteroplasmy < 1%. Pyrosequencing confirmed the remaining eight variants were absent in blood. Of the 67 m.3243A>G variants identified in blood, the heteroplasmy identified by mity and pyrosequencing was highly correlated.

Interestingly, there was no relationship between the Nijmegen Clinical Criteria Score for MD and pathogenic variant heteroplasmy, or the likelihood of a WGS diagnoses. In fact, 9 patients scoring ‘not fulfilling’ were diagnosed with pathogenic mutations with heteroplasmy up to 100%. We identified a number of individuals with complex clinical presentation, with more than one pathogenic mutation, which would have been missed using a more targeted approach. Finally, we diagnosed 4 previously undiagnosed patients with a pathogenic mitochondrial mutation.

mity’s detection of low heteroplasmy mitochondrial variants, combined with an assessment of the nuclear genome, resulted in an overall diagnostic rate of 56%, further supporting the utility of WGS as a diagnostic test for MD.