Mitochondrial DNA (mtDNA) is a high copy number genome, creating a population of mtDNA molecules within a cell. As healthy cells age, mitochondrial genomes (mt-genomes) accumulate mutations that vary in frequency. Until recent technological advancements, the characterization of the mt-genome mutational landscape has been limited to high frequency variants. With the advent of ultra-sensitive sequencing technologies, surveying low frequency mutations is now possible, improving the investigation of mutational processes. In order to design efficient medical therapies for mitochondrial dysfunction, an understanding of the factors that impact mtDNA mutation is necessary. Here, I characterize the mtDNA mutational landscape across different mitochondrial haplotypes and tissues to discern the molecular and evolutionary processes underlying the origin and trajectory of mutation in mtDNA.
To begin, I investigate how mitochondrial haplotype shapes mutation in mtDNA. Importantly, over one thousand components needed for mitochondrial function are encoded on the nuclear genome; thus, harmony between these two genomes is vital. With this in mind, this work additionally studies how mito-nuclear ancestral mismatching impacts somatic mutation. To address these questions, I employ a panel of four mouse strains that are identical in their nuclear genomes, but differ in their mitochondrial haplotypes. I use ultra-sensitive Duplex Sequencing to generate maps of mutations with an unprecedented level of depth and accuracy across multiple tissues and mitochondrial haplotypes. From these maps, I confirm that the increase in mutation frequency with age is a constant molecular phenotype. Interestingly, comparison of rodent and primate mtDNA mutational spectra provide evidence of species-specific mutational signatures likely associated with distinct life history traits. My findings contrast the established notion that the non-coding region in the mt-genome has the highest mutation frequency. For example, I identify that the Origin of Replication (light strand) consistently has a higher mutation frequency across both tissues and mitochondrial haplotypes. Moreover, I highlight a mutational hotspot in MT-tRNA-Arg specifically in mice with mismatching mito-nuclear ancestry. Lastly, my work is the first to identify cases in which there is a preference for mutations that align mito-nuclear ancestry within the organism’s lifespan.
Secondly, I explore intra-individual variation by characterizing the tissue-specific mutational landscapes of two regions in the aging mouse brain. Employing Duplex Sequencing, I profile the cortex and cerebellum at three different time points in the mouse lifespan. These regions were chosen due to differences in their metabolic demand and distinct accumulation of deletions throughout aging. I identify mutations present in both the cortex and cerebellum, and note that a consistent feature between tissues is the inheritance of insertions and deletions in regions associated with mtDNA replication. Although the tissues have similar frequencies for shared mutations in young mice, these frequencies diverge with age likely as a result of genetic drift. Overall, the cortex has a higher average mutation frequency than the cerebellum, but does not differ in mutation rate. Examination of different mutation types shows that the cortex and cerebellum differ in mutational signatures associated with mtDNA replication and metabolic damage. Mutations associated with mtDNA replication error are more abundant in the cortex; yet, the cortex and cerebellum exhibit a similar mutation rate for this signature. Despite the cortex being more metabolically demanding, this tissue shows a decrease in metabolic damage with age, opposite the trend in the cerebellum, indicating that the tissues potentially differ in their regulation of metabolic damage.
Together, these findings identify robust properties of mtDNA, while pinpointing specific differences driven by tissue-type and mitochondrial haplotype for future considerations.
