Defining the nuclear genetic architecture of a common maternally inherited mitochondrial disorder

Abstract

AbstractMaternally inherited mitochondrial diseases are caused by pathogenic mitochondrial (mt)DNA variants. Affecting individuals at any age, they are often multi-systemic and manifest extreme clinical variability. We have limited understanding of the cause of this heterogeneity, which makes disease diagnosis and prognosis exceptionally challenging. This is clearly demonstrated by disease caused by m.3243A>G, the most common pathogenic mtDNA variant. m.3243A>G can cause a severe syndrome characterised by mitochondrial encephalomyopathy lactic acidosis and stroke-like episodes (MELAS), but individuals who carry m.3243A>G may be asymptomatic or manifest with any number of a range of phenotypes. There is strong evidence for the presence of nuclear factors that modify phenotype; we set out to characterise the nature of this nuclear involvement using genetic linkage analysis.We assembled a multi-centre cohort of well-characterised patients and their maternal relatives, comprising 76 pedigrees, and characterised the nuclear genetic landscape of m.3243A>G- related disease phenotypes using non-parametric genetic linkage analysis. We considered eight of the most common m.3243A>G-related phenotypes, accounted for known risk factors using logistic regression, and determined empirical significance using simulation to identify regions of the nuclear genome most likely to contain disease modifying variants.We identified significant genetic linkage to encephalopathy on chromosome 7q22, and suggestive regions for encephalopathy, stroke-like episodes and psychiatric involvement on chromosomes 1, 5, 6, 11 and 13. These findings suggest that these neurological features are likely to be influenced by a small number of nuclear factors with a relatively large effect size. In contrast, no linkage regions were identified for cerebellar ataxia, migraine, diabetes mellitus, hearing impairment or chronic progressive external ophthalmoplegia.The genetic architecture of the nuclear factors influencing disease related to m.3243A>G differs between phenotypes. Severe and cardinal neurological features of MELAS are likely to be strongly influenced by a small number of nuclear genes, whereas the nuclear influence over other phenotypic presentations is more likely to be polygenic and complex in nature, composed of a larger number of factors that each exert a small effect. These results will inform strategies for future studies to identify the genes and pathways that influence clinical heterogeneity in m.3243A>G-related disease, with the ultimate aim of better understanding disease development and progression.