It has been shown recently that mitochondrial (mtDNA) somatic variants are numerous enough to trace cellular lineages in our body. Here we extend this statement and demonstrate that mtDNA variants can be interpreted not only as neutral markers of cell divisions but the relative frequency of different mtDNA substitutions (i.e. mtDNA mutational spectrum) can inform us about important biological properties such as cell longevity. Analysing 7611 somatic mtDNA mutations from 37 types of human cancers and more than 2000 somatic mtDNA mutations from 25 healthy human tissues we observed that mtDNA mutational spectrum is associated with cell turnover rate: the ratio of T>C to G>A is increasing with cell longevity. To extend this logic we considered that, if universal, the discovered mutation bias may drive the differences in mtDNA mutational spectrum between mammalian species with short-(‘mice’) and long-(‘elephants’) lived oocytes. Based on presumably neutral polymorphisms in MT-CYB we reconstructed mutational spectra for 424 mammalian species and obtained that the fraction of T>C positively correlated with the species-specific generation length, which is a good proxy for oocyte longevity. Next, comparing complete mitochondrial genomes of 650 mammalian species we confirmed that exactly the same process shapes the nucleotide content of the most neutral sites in the whole mitochondrial genomes of short-(high T, low C) versus long-(low T, high C) lived mammals. Altogether analysing mtDNA mutations in time interval from dozens of years (somatic mutations) through the hundreds of thousands of years (within species polymorphisms) to millions of years (between species substitutions) we demonstrated that T>C/G>A positively correlates with cellular and organismal longevity. We hypothesize that the discovered mtDNA signature presents a chemical damage which is associated with the level of oxidative metabolism which, in turn, correlates with cellular and organismal longevity. The described properties of mtDNA mutational spectrum shed light on mtDNA replication, mtDNA evolution of mammals and can be used as a marker of cell longevity in single-cell analyses of heterogeneous samples.
It has been recently shown, that because of high mitochondrial DNA (mtDNA) mutation rate and high mtDNA copy number per cell somatic variants in mtDNA are informative to trace different cellular lineages in our body (Ludwig et al. 2019). Here we hypothesize further that mitochondrial variants can be interpreted not just as neutral markers of cellular lineages but the relative frequency of different mtDNA substitutions (i.e. mtDNA mutational spectrum) may also contain a functional signature of cellular properties such as the level of metabolism or turnover rate.
Indeed there are several properties of mtDNA mutagenesis which make it especially predisposed to sense cell-specific metabolism. First, during asynchronous mtDNA replication parental heavy chain is spending significant amount of time in single-stranded state during which (i) spontaneous deamination of adenine to guanine is increasing as a linear function of a time (hereafter we will follow widely accepted in mitochondrial community light chain notation according to which A>G on heavy chain is called T>C on light chain), (ii) spontaneous deamination of cytosine to thymine (G>A, light chain notation) is less sensitive to the time being single-stranded and (iii) all other rare transitions and transversions are not sensitive at all (Faith and Pollock 2003). These different time dynamics of substitutions will lead to unique signatures (ratios of substitutions) corresponding to various time intervals of mtDNA replication. Second, mitochondria is tightly involved in the determination of the cell specific level of oxidative metabolism and thus mutagens, associated with it are expected to affect mtDNA directly and strongly (Ericson et al. 2012). Third, cancer studies have shown that mtDNA is apparently affected by very strong endogenous mutagen because it completely overwhelms effects of other expected exogenous mutagens such as tobacco smoke in lung cancers or ultraviolet light in melanomas (Yuan et al. 2017; Ju et al. 2014). All these properties (time dependent dynamics, tight involvement into the determination of the oxidative metabolism, low sensitivity to exogenous mutagens) make mtDNA mutational spectrum a good potential sensor of cellular metabolic processes.
In our work we analysed several cellular phenotypes, related to the level of metabolism such as the number of mtDNA copies, expression level of mitochondrial genes and the turnover rate of cells from different tissues. Correlating mtDNA mutational spectrum (i.e. probabilities of different substitutions) with all these cellular phenotypes we found the only robust relationship with cell turnover rate. This might show that cell turnover rate is indeed one of the most important cellular phenotypes, affecting mutagenesis (Tomasetti and Vogelstein 2015; Tomasetti et al. 2017). Interestingly, this metric due to its simplicity allowed us to extend our analyses to other mammalian species. Since in all mammals mtDNA is inherited exclusively through oocytes, and the division of oocytes is arrested after the birth, the species-specific generation length well enough approximates the longevity of oocytes from different mammalian species (i.e. time which mtDNA spent in dormant oocyte) and thus might correlate with specific mtDNA signature of longevity.
Altogether, using a collection of 7611 somatic mtDNA mutations from 37 types of human cancers (TCGA and ICGC), more than 2000 somatic mtDNA mutations from 25 human healthy tissues (GTEx), 39112 polymorphic synonymous substitutions in four-fold degenerate MT-CYB sites from 424 mammalian species and nucleotide content in whole mitochondrial genomes of 650 mammalian species we observed one universal trend: ratio of T>C to G>A positively correlates with cellular and organismal longevity. Out of several potential explanations of the observed associations we prefer one according to which T>C/G>A is sensitive to the level of oxidative metabolism which is expected to be higher in cancer cells in early versus advanced stages, in slow-dividing versus fast-dividing tissues and in long-versus short-lived species. The discovered mtDNA signature of cellular and organismal longevity helps to understand better the mechanism of mtDNA replication, shed light on evolution of mammalian mtDNA and opens a possibility to mark cell longevity in heterogeneous tissues.
The rest at the source: https://www.biorxiv....1/589168v1.full
