Fig 1.
Schematic for isolation and sequencing of single-cell fibroblast clonal lineages.
Example of Donor 1 is shown. Boxed insert illustrates the design of the clone IDs. Biopsy number is indicated if adjacent biopsies were taken from the same site.
Fig 2.
Structural changes detected in skin fibroblast clones D1-L-F1 and D2-L-F.
(A) All genome changes detected in D1-L-F1 and D2-L-F clones. The tracks numbered from innermost are as follows: 1—structural changes. Green = deletions, black = duplications, blue = inversions and red = translocations. 2—deletions as detected by read-depth analyses. 3—amplifications as detected by read-depth analyses. 4—LOH events. 5—somatic SNVs, black dots are heterozygous SNVs and red dots are homozygous SNVs. (B) Schematic describing the chr19, chr20 translocation in D1-L-F1. Black rectangles depict region wherein translocation event was detected with a concomitant change in copy number.
Fig 3.
Somatic mutation load and spectra in the fibroblast clones.
(A) The number of somatic mutations detected in each clone and the rate of accumulation of mutations per year are provided. (B) The spectra of base changes in the clones. For each base change the reverse complements are also included.
Fig 4.
Analysis of mechanistic knowledge-based mutation signatures in the genomes of skin fibroblasts.
Similar analysis for the whole-genome sequenced melanoma (SKCM) cohort (dataset from [46]) is provided for comparison. (A) Fold enrichment of nCg →nTg and UV-signature mutations (yCn→yTn, nTt→nCt; y is either C or T, n is either A, T, G or C, in the trinucleotide context the mutated base is in capital). The black horizontal line denotes the level of no enrichment. (B) The minimum estimates of signature-specific mutation loads for each clone. For the melanoma cohort, the median of the minimum estimated mutation loads for each signature per genome in is shown. (C) Total CC→TT counts of tandem dinucleotide changes in each clone and the median of the total CC→TT counts per genome in the melanoma cohort.
Fig 5.
Somatic mutation loads vary with replication timing and transcription status.
(A) SNVs from both, hips and forearms are enriched in late replicating genomic regions and heterochromatic regions of the genome. Values on the horizontal axis were obtained by averaging values of the genomic feature into 5 equal bins. (B) UV-attributable mutations demonstrate a strand bias for the non-transcribed strand. * denotes p-value after false discovery rate (FDR) correction for multiple hypothesis testing < 0.05, *** denotes p-value after FDR correction < .01 (p-values are in S13 Table). There is no preference for non-transcribed strand for mutations in C and T which do not conform to any UV-signature motif.
Table 1.
Estimates of rates of accumulation of endogenous and UV-induced mutations per year.
Table 2.
Mutation signatures analyzed in this study.