Although there is robust evidence for the role of mitochondrial dysfunction in some individuals with autism spectrum disorder (ASD), the genetic loci responsible and whether these are encoded in either the nuclear or mitochondrial DNA are uncertain. Genetic studies of mitochondrial potential must account for the products of two genomes with distinct properties and patterns of inheritance. Recent studies have suggested that either specific coinherited polymorphisms (known as ‘haplogroups’) or the inheritance of heteroplasmic variation in the underlying sequence is associated with ASD risk^1,2.

Here, Neal Sondheimer proposes to validate and extend these studies using data from the MSSNG cohort, a large collection that currently comprises whole-genome sequencing data from 2,757 families containing biological parents and at least one proband. In addition, Sondheimer will use innovative genetic analyses to probe the interaction between the inherited variants in mitochondrial and nuclear DNA. The central hypothesis of this proposal is that genetically determined impairments in mitochondrial function predispose individuals to ASD.

Sondheimer’s laboratory will test this hypothesis through three aims. First, they will determine whether specific haplogroups are enriched in ASD probands. Next, they will explore whether heteroplasmies are inherited from mother to proband and whether the total burden of heteroplasmy is higher or more pathogenic than what is seen in unaffected individuals. Finally, using a methodology previously developed in their lab for determining the ancestral distance between nuclear and mitochondrial inheritance, known as ‘divergent ancestry’³, Sondheimer’s group will test the relationship between mitonuclear incompatibility and ASD risk.

The proposed analysis of haplogroup and heteroplasmy was informed by the recent studies of mitochondrial DNA and autism risk^1,2; these studies analyzed whole-exome sequencing data from the Simons Simplex Collection (SSC) and single nucleotide polymorphism genotype data from the Autism Genetic Resource Exchange (AGRE), respectively. Evaluating these questions in the MSSNG data set will help confirm and extend this past work by using a larger study population, with deeper sequencing, distinct controls and different analytical and statistical methods. The extension of these ideas, through studies of divergent ancestry, will enable Sondheimer’s group to merge mitochondrial studies with the robust prior findings on the nuclear risk factors for ASD.

Findings from this study are expected to expand the understanding of the role of mitochondrial genetics in ASD. This is of significance for two main reasons. First, it will allow investigators using shared data sets from the MSSNG, SSC and AGRE cohorts to incorporate mitochondrial genetics into models of genetic risk for ASD. Second, the advancing recognition of the role of mitochondrial genetics may encourage new concepts in therapeutic research, including studies on the activation of mitochondriogenesis and the regulation of free radical formation and enhanced nutrition.

1. Wang Y. et al. PLoS Genet. 12, e1006391 (2016) PubMed
2. Chalkia D. et al. JAMA Psychiatry, 11, 1161-1168 (2017) PubMed
3. Crawford N. et al. J. Pediatr. 194, 40-46 (2018) PubMed