Mitochondrial mutations in human cancer: Curation of translation

Abstract

As a genetic disease, cancer is caused by the activation of oncogenes and the inhibition of tumor suppressor genes via genetic and epigenetic mechanisms. Given the important role of energy metabolism in tumors, we analyzed the cancer-derived mutations occurring in the DNA of the mitochondrion. Mutations in the mitochondrial DNA (mtDNA) compared to nuclear DNA are 62% decreased relative to the coding length per chromosome. We find that the majority of these mutations affects highly conserved nucleotides - significantly exceeding the conservation of the mtDNA - and are devoid of single nucleotide polymorphisms (SNPs). Surprisingly, the leading resources for tumor genetics information universally use the standard genetic code for translation of nucleotide into amino acid sequences in their online resources. However, the nuclear and mitochondrial genetic codes differ for four codons and the usage of incomplete STOP codons. Hence, we analyze and curate the consequences for all mutations in the mtDNA and comprehensively reclassify missense, nonsense and synonymous mutations accordingly. In total, 10% of the mutations are incorrectly translated leading to significant changes in the distribution of mutation types with tripling of nonsense and 69% loss of nonstop extension mutations. Lastly, we provide a curated dataset of coding and non-coding mitochondrial mutations in cancer merged, standardized, duplicate-free and aggregated from two databases as a resource including orthogonal data on their high conservation and SNPs. This study generally highlights the need to universally regard the important differences between the standard and mitochondrial genetic code in life science research.

Keywords: Mitochondria; cancer; curation; genetic code; mutations.

Figure 1.

Cancer-derived mutations in the mitochondrial DNA. (A) The total number of coding mutations mapping to each chromosome, the length of each chromosome, the normalized mutation density relative to the length of the chromosome, the length of the coding sequences of each chromosomes and the normalized mutation density relative to the length of the coding sequences of each chromosome are shown (data: Table S2). (B) The distribution of mutation types in the entire COSMIC database (black) as well as in the mitochondrial mutations in COSMIC (blue) and in ICGC (red) before and after curation (light blue and light red, respectively) are depicted. (C) The bar graph depicts the evolutionary conservation of the entire mitochondrial chromosome (“all” = black) and the coding space on chrM (“CDS” = grey) in comparison to the loci on chrM affected by mutations annotated in the COSMIC or ICGC databases for coding (“CDS mut COSMIC” = blue; “CDS mut ICGC” = red) and non-coding sequences (“non-CDS mut COSMIC” = light blue; “non-CDS mut ICGC” = light red) as determined by the PhastCons scores aggregated in intervals of 0.1 increments. Statistical significance: PhastCons vs chrM all: COSMIC CDS: p = 1.5*10⁻²; ICGC CDS: p = 2.0*10⁻⁸; COSMIC non-CDS: p = 2.5*10⁻⁴; ICGC non-CDS: p = 3.8*10⁻⁹ / PhastCons vs chrM CDS: COSMIC CDS: p = 2.0*10⁻³; ICGC CDS: p = 1.1*10⁻¹⁰; COSMIC non-CDS: p = 6.4*10⁻⁵; ICGC non-CDS: p = 5.1*10⁻¹¹. (D) Violin plots depict the evolutionary conservation of the entire mitochondrial chromosome (“chrM all”) and the coding space on chrM (“chrM CDS”) in comparison to the loci on chrM affected by mutations annotated in the COSMIC or ICGC databases for coding (“chrM mut CDS”) and non-coding sequences (“chrM mut non-CDS”) as determined by the PhyloP scores. Bars indicate the median of each distribution. Statistical significance: PhyloP vs chrM CDS: COSMIC CDS: p = 1.1*10⁻¹ (not significant); ICGC CDS: p = 4.3*10⁻⁹; COSMIC non-CDS: p = 3.6*10⁻⁵; ICGC non-CDS: p = 2.7*10⁻¹⁷/ PhyloP vs chrM CDS: COSMIC CDS: p = 6.2*10⁻³; ICGC CDS: p = 1.1*10⁻¹¹; COSMIC non-CDS: p = 1.4*10⁻⁶; ICGC non-CDS: p = 5.8*10⁻²³. (E) The graph depicts the fraction of mutations coinciding with common SNPs in the entire COSMIC database (black) and in the mtDNA only as annotated by COSMIC (blue) as well as mitochondrial mutations mapping to SNPs annotated by dbSNP in COSMIC (green) or ICGC (red).