Nuclear genetic control of mitochondrial function and its contribution to human disease: insights at biobank scale

Abstract

Humans rely on two genomes: the diploid nuclear genome, which encodes ~20,000 proteins, and the mitochondrial genome (mtDNA), which encodes 13 proteins and exists in many copies per cell. Crosstalk is essential: all machinery used to replicate mtDNA is nuclear-encoded, and the oxidative phosphorylation (OXPHOS) pathway requires components from both genomes. Hundreds of studies have linked OXPHOS decline to many age-related diseases including type 2 diabetes and Alzheimer’s disease. Despite these associations, the genetic basis for variation in mitochondrial physiology across humans and subsequent role in disease remains poorly understood. Large biobanks with sequenced genomes and disease diagnoses, which came online during my graduate research, provide an opportunity to address these gaps. The overarching goal of my thesis is to elucidate the nuclear genetic basis for variation in human mitochondrial function and how this may influence disease and aging: In Chapter 1, I describe MitoCarta3.0, an updated inventory of mitochondrial proteins. I identify 1123 nuclear DNA-encoded genes which produce proteins localizing to the mitochondrion and place them in a novel ontology of 149 “MitoPathways”. In Chapter 2, I perform several genome-wide association studies (GWAS) testing for nuclear genetic variants that control mtDNA copy number (mtCN) and heteroplasmic mtDNA mutations after quantifying these across >250,000 individuals. I uncover a vast inventory of such nuclear variants, finding that common human nuclear variants can influence the abundance of mtDNA mutations for the first time and identifying mechanisms underlying this phenomenon. In Chapters 3 and 4, I focus on genetic links between mitochondrial function and common disease. In Chapter 3, I explore the enrichment of genetic signal implicating each cellular organelle in 24 common age-related diseases and longevity, finding that DNA-binding proteins are enriched for age-related disease heritability. In Chapter 4, I use plasma metabolomics from patients with rare mitochondrial disease to create a measure of OXPHOS dysfunction we call “MitoScore,” use GWAS to define its common nuclear genetic basis, and explore its genetic correlation with age-related diseases. Together, these human genetic results have broad implications for understanding both the mechanisms influencing human mitochondrial biology and the links between OXPHOS dysfunction and age-related disease.