Relative to the whole body, the brain uses 10 times more energy by weight. Most of this energy is mediated by the ATP generated by mitochondria, which play many roles in brain development and function. In particular, mitochondrial energetics are critically important for neurotransmission, including neurotransmitter vesicle cycling, homeostasis of ionic gradients, axonal transport and other synaptic functions. Since pathobiological mechanisms underlying autism spectrum disorder (ASD), while heterogenous, also appear to converge on synaptic function, it stands to reason that mitochondrial energetics are important influencers of ASD-related phenotypes. Moreover, since mitochondrial energetics can be enhanced pharmacologically (not to mention by diet and exercise), the identification of a subset of individuals with ASD who have suboptimal mitochondrial energetics could lay the foundation for the development of a “personalized” therapeutic strategy.

Genetic variation in the more than 1,100 nuclear encoded genes that generate mitochondrial-localizing proteins and in the mitochondrial genome (encoding 13 proteins and a handful of non-coding RNAs) can indeed be associated with increased risk of ASD. While a small number of rare ASD-associated single gene mutations and copy number variations have been shown to cause mitochondrial dysfunction, a broader connection between common genetic variation, ASD risk and mitochondrial function has not been demonstrated. The goal of this project is to make just this connection.

In Aim 1, Stewart Anderson and collaborator Michael Gandal plan to build upon Gandal’s and others’ large-scale transcriptomic profiling studies in the human neocortex which identified mitochondrial dysfunction in ASD¹, to develop and validate a mitochondrial-polygenic function score (mitoPFS). Association of the mitoPFS will be made with ASD symptom domain severity scores in SPARK.

In Aim 2, whole-genome sequencing data from each of the 2,600 individuals with ASD enrolled in the Simons Simplex Collection (SSC) will be ranked using the mitoPFS. SSC-banked lymphoblastoid cell lines (LCLs) — immortalized B-lymphocytes that have been used in many studies of mitochondrial energetics — will be expanded from the 25 individuals with the highest and lowest mitoPFS. By assessing the mitochondrial respirometry of these lines, Anderson and Gandal will test the hypothesis that lower mitoPFS — within a sample wholly derived from individuals affected by ASD — will be associated with diminished activity in the oxidative phosphorylation pathway.

Success in this endeavor would be foundational for multiple important studies. Bolstered by the functionality of the mitoPFS, existing RNA sequencing data from these same lines could be used to link mitoPFS variants to gene expression variation on both the pathway and individual gene levels. As LCLs can be made into induced pluripotent stem cells (iPSCs), links between mitoPFS, gene expression and cellular function could be extended to ASD-relevant cell types such as cortical excitatory neurons, interneurons and microglia. Finally, these results would support the rationale for further study of the thousands of LCL lines banked from ASD-relevant individuals worldwide, and for studies that use blood-based or even non-invasive mitochondrial measures to prospectively justify a therapeutic intervention to enhance mitochondrial function in some ASD individuals.

1. Gandal M.J. et al. Science 359, 693-697 (2018) PubMed