Figure 2

Distribution and frequency of mtDNA hypermutation. D-loop–ND1 (nt16331–nt3729, NC012920.1), ND1–ATP6 (nt3646–nt9158, NC012920.1), ATP6–ND5 (nt8753–nt12831, NC012920.1), and ND5–tRNA-F (nt12540–nt645, NC012920.1) were amplified from two cervical intraepithelial neoplasia (CIN) and two oropharyngeal cancer samples along with a plasmid without mutation. The PCR products were subjected to next-generation sequencing and the output reads were sorted according to the number of C-to-T or G-to-A mutations per read (Tables 1–4). Those with four or more C-to-T/G-to-A mutations were extracted. (a) Alignment of reads with four or more C-to-T/G-to-A mutations. The rectangles indicate the positions of the extracted reads in the reference sequence of each sample. (b,c) Mutation matrices (b) and dinucleotide analysis (c) of the hypermutated reads, obtained from D-loop–ND1 and ND5–tRNA-F, amplified from CIN sample #1300. In (c), the contexts of Cs in the reference sequences that corresponded to the mutated Cs in the output reads were also analyzed to calculate the expected frequency of each context. The P-values were calculated using the chi-square test, comparing the numbers of TpC and non-TpCs, or CpC and non-CpCs. **P < 0.01 and ***P < 0.005. (d) The consensus sequences obtained from #1300, #2159, KZU33, and KZU45 were aligned with revised Cambridge Reference Sequence (NC_012920.1), and the discordant bases were summarized.