Abstract
Mitochondria are essential cellular organelles that play critical roles in cancer development. Through the International Cancer Genome Consortium (ICGC), we performed a multidimensional characterization of mitochondrial genomes using whole-genome sequencing (WGS) data from ~2,700 patients across 37 cancer types. Our analysis presents a definitive mutational landscape, revealing similar mutational signatures across cancer types and suggesting powerful endogenous processes. We identified significant enrichment of truncating mutations in specific cancers and frequent somatic mitochondrial-to-nuclear DNA transfers (SMNTs) that disrupt therapeutic targets. This study establishes The Cancer Mitochondrial Atlas (TCMA), providing a foundation for translating mitochondrial biology into clinical applications.
Introduction
While the nuclear genome has been mapped extensively in oncology, the mitochondrial genome (mtDNA)—a 16.5 kb circular loop—has remained relatively under-explored. As the "powerhouse of the cell," mitochondria are central to energy metabolism, signaling, and apoptosis, all of which are frequently hijacked during tumorigenesis. This study, part of the Pan-Cancer Analysis of Whole Genomes (PCAWG) consortium, provides the most comprehensive map to date of mtDNA alterations across 37 different cancer types, fundamentally changing our understanding of mitochondrial roles in cancer.
Addressing the Multidimensional Gap
Previous mitochondrial studies were often limited by small sample sizes or focused on single dimensions, such as mutations or copy numbers in isolation. The primary challenge addressed here is the lack of a "multidimensional" portrait: understanding how mtDNA mutations, copy number variations, and gene expression interact with each other and the nuclear genome. Understanding this interplay is critical because mitochondrial dysfunction is a hallmark of metabolic reprogramming in cancer.
Methodology and Data Integration
The researchers utilized a sophisticated bioinformatic pipeline to extract mitochondrial data from high-depth WGS.
- High-Depth Sequencing: mtDNA was sequenced at an average of 9,959× coverage, enabling the detection of "heteroplasmic" mutations (variants in a small fraction of a cell's mitochondria) at frequencies as low as 1%.
- Bias Correction: A novel formula was developed to estimate mtDNA copy numbers, accounting for tumor purity and nuclear ploidy—factors that often skew results in standard analyses.
- Datasets: The study integrated WGS data from 2,658 cancer/matched-normal pairs (ICGC PCAWG) and RNA-sequencing data from 4,689 tumor samples (TCGA).
- Network Analysis: Weighted Gene Co-expression Network Analysis (WGCNA) was employed to link 13 mitochondrial protein-coding genes with thousands of nuclear genes.
Key Insights: A Unique Mutational Engine
In the nuclear genome, mutations are often driven by exogenous factors like tobacco smoke or UV light. In contrast, this study found that mitochondrial mutational signatures are remarkably monotonous across all cancer types. This suggests that mtDNA mutations are primarily driven by internal, endogenous processes, likely involving errors by mitochondrial DNA polymerase or unique repair mechanisms rather than external carcinogens.
Notably, the study identified 7,611 substitutions and 930 small indels, including the first-ever reported hypermutated mitochondrial case in a breast cancer sample.
Selection and Truncation Paradox
Typically, "truncating" mutations that disable a gene are eliminated by natural selection. However, this research found these mutations were significantly enriched in kidney, colorectal, and thyroid cancers. This suggests that the loss of specific mitochondrial functions may provide a selective survival advantage to these tumor types. Furthermore, tumors with high-frequency truncating mutations often exhibited higher mtDNA copy numbers, suggesting a "dosage compensation" mechanism where the cell attempts to maintain function by replicating the damaged genome.
Somatic Mitochondrial Nuclear Transfers (SMNTs)
One of the most striking findings was the frequency of "genomic jumping." Segments of mtDNA were found to migrate and integrate into the nuclear genome (SMNTs) in 2.1% of cases. This phenomenon was particularly prevalent in lung squamous cell carcinoma (14.6%) and skin cancers. Crucially, some of these transfers landed within known cancer-driving genes, such as ERBB2 (HER2), potentially altering their function and therapeutic susceptibility.
Clinical Implications and The Cancer Mitochondrial Atlas (TCMA)
The study revealed that mtDNA abundance varies significantly across cancers, ranging from a median of 90 copies in myeloid leukemia to 644 in ovarian cancer. These copy numbers correlate with clinical variables like patient age and tumor stage.
To facilitate further research, the authors introduced The Cancer Mitochondrial Atlas (TCMA), an open-access data portal. Co-expression analysis within this framework highlighted mitochondrial gene connections with clinically actionable nuclear genes like EGFR and PIK3CA, opening new avenues for targeted therapies that consider the mitochondrial state of the tumor.
Conclusion
This research provides a definitive foundation for understanding the interplay between the mitochondrial and nuclear genomes. While the findings are largely correlative, they highlight the necessity of functional validation in future "wet lab" experiments. By bridging the gap between mitochondrial biology and clinical oncology, this study sets the stage for the next frontier of mitochondrial-targeted cancer therapies.
Summary updated