Mitochondrial mutation spectrum in Chordates: damage versus replication signatures, causes, and dynamics

Abstract

To elucidate the primary factors shaping mitochondrial DNA (mtDNA) mutagenesis, we reconstructed a deep 192-component mtDNA mutational spectrum using 129,100 polymorphic synonymous mutations from the Cytb gene of 1040 chordate species. We then deconvoluted this spectrum into three key signatures: (i) symmetrical mutations, predominantly C>T and rare transversions, linked to pol-γ’s replication errors; (ii) asymmetrical C>T mutations, indicative of single-stranded DNA damage; and (iii) asymmetrical A>G mutations, also resulting from single-stranded DNA damage but particularly influenced by metabolic and age-specific mitochondrial environment. Interestingly, the two asymmetrical signatures collectively surpass pol-γ mutations, indicating that damage is the primary factor in mtDNA mutagenesis over replication. The diverse contribution of all three signatures across different species and human tissues provides insights into a range of physiological and metabolic traits. This mtDNA spectrum-aware methodology enhances phylogenetic inferences, deepens our understanding of species-specific mutational and selection processes, and offers insights into the dynamics of mutations in human mtDNA.

NeMu: A Comprehensive Pipeline for Accurate Reconstruction of Neutral Mutation Spectra from Evolutionary Data

“…We cannot recommend this method of sample construction to a higher taxonomic rank (e.g., the rank of taxonomic class or type) because of the much greater ambiguity in the classification of neutral (or nearly neutral) alignment positions with synonymous fixations and because of the synonymous substitution saturation 39,40 . To construct a mutational spectrum for higher taxonomic rank we recommend procedures described in our analysis of mtDNA mutation spectra of chordata species 41 . The manual method of sample construction is often exacerbated by unnoticed and not trivial difficulties: it is important for the user to make sure that the pairwise distance distribution between the aligned sequences is uniform (this means that it should not be multimodal), otherwise, the mutations on the longest tree branches, which correspond to outlayer branch distances, could cause a significant bias in the mutational spectrum 42 .…”