Mitochondrial DNA mosaicism in human normal somatic cells

Abstract

Genomic alterations accumulate in the somatic cells throughout an individual’s lifetime1-5. Recent sequencing studies have documented widespread mutations in the nuclear genome and the frequent clonal competition of normal cells carrying mutations6-11. However, the landscape of mitochondrial DNA (mtDNA) heteroplasmy in normal human tissues is poorly understood. In this study, we investigated the whole genome sequences of 2,096 clones established from single cells obtained from 31 donors. We identified 6,451 mtDNA variants with >0.3% heteroplasmy in the clones. Approximately 6% of these variants originated from fertilized eggs, suggesting pervasive mtDNA heteroplasmy in the first cell of our life, despite the purification bottleneck in the maternal germline12,13. Using the landscape of variants in the fertilized egg, sporadic mutations (88%) and hotspot variants (6%), we determined two fundamental components that shape mtDNA heteroplasmy in somatic cells: the absolute mutation rate in mtDNA (1.01x10-7 per bp) and the frequency of mtDNA turnovers (22-43 turnovers per year varying across cell types). We found that a dominant mutational signature exhibiting extreme replication strand bias was mainly responsible for sporadic mutations, and the expansion of truncating mutations was substantially repressed in the somatic lineages. Cross-comparison between DNA and RNA sequences further confirmed that mtDNA localized in the stems of tRNA and rRNA had an impact on the regulation of mtDNA transcripts. Despite extensive inter-clonal and inter-individual heterogeneity in the mtDNA copy number per cell, the cellular transcriptional profiles were surprisingly uniform. Taken together, our findings provide comprehensive insights into the origin and dynamics of heteroplasmy levels, and the functional consequences of mtDNA variants in human somatic cells.

Quantifying cell divisions along evolutionary lineages in cancer

“…Our inferred colonic stem cell division rate aligns with a recent estimate of one division per 16 days based on mtDNA variants 34 , and is slower than another estimate of one division per 7 days based on visually identifying dividing cells at the bottom of the colonic crypt 35 . The latter estimate could be problematic because stem cells are not the only dividing cells in that location 36 .…”