Abstract:Background: Mitochondria play essential roles in tumorigenesis; however, little is known about the contribution of mitochondrial DNA (mtDNA) to esophageal squamous cell carcinoma (ESCC). Whole-genome sequencing (WGS) is by far the most efficient technology to fully characterize the molecular features of mtDNA; however, due to the high redundancy and heterogeneity of mtDNA in regular WGS data, methods for mtDNA analysis are far from satisfactory. Methods: Here, we developed a likelihood-based method dMTLV to identify low-heteroplasmic mtDNA variants. In addition, we described fNUMT, which can simultaneously detect non-reference nuclear sequences of mitochondrial origin (non-ref NUMTs) and their derived artifacts. Using these new methods, we explored the contribution of mtDNA to ESCC utilizing the multi-omics data of 663 paired tumor-normal samples. Results: dMTLV outperformed the existing methods in sensitivity without sacrificing specificity. The verification using Nanopore long-read sequencing data showed that fNUMT has superior specificity and more accurate breakpoint identification than the current methods. Leveraging the new method, we identified a significant association between the ESCC overall survival and the ratio of mtDNA copy number of paired tumor-normal samples, which could be potentially explained by the differential expression of genes enriched in pathways related to metabolism, DNA damage repair, and cell cycle checkpoint. Additionally, we observed that the expression of CBWD1 was downregulated by the non-ref NUMTs inserted into its intron region, which might provide precursor conditions for the tumor cells to adapt to a hypoxic environment. Moreover, we identified a strong positive relationship between the number of mtDNA truncating mutations and the contribution of signatures linked to tumorigenesis and treatment response. Conclusions: Our new frameworks promote the characterization of mtDNA features, which enables the elucidation of the landscapes and roles of mtDNA in ESCC essential for extending the current understanding of ESCC etiology. dMTLV and fNUMT are freely available from https://github.com/sunnyzxh/dMTLV and https://github.com/sunnyzxh/fNUMT , respectively.

Approach, Application, and Bioethics of mtDNA Sequencing in Cancer.

Nuclear Genes and Mitochondrial Genes Associated with Mitochondrial Diseases

Prospective of Mitochondrial DNA Variations in Cancer on Genomic Medicine

Research Progress in Mitochondrial Gene Editing Technology.

Functional Role of Mitochondrial DNA in Cancer Progression

Advancing Variant Phenotyping in Myelodysplastic Syndromes via Computational Genomics of Mitochondrial Enzyme Complexes

Single-Cell Technologies for Studying the Evolution and Function of Mitochondrial DNA Heteroplasmy in Cancer

Mitochondrial DNA, Mitochondria, Disease and Stem Cells

Association Between Mitochondrial DNA Mutations and Cancer in Human

Mitochondrial molecular genetics and human disease

Leveraging new methods for comprehensive characterization of mitochondrial DNA in esophageal squamous cell carcinoma

Current Approaches to Reduce or Eliminate Mitochondrial DNA Mutations

Our current understanding of the biological impact of endometrial cancer mtDNA genome mutations and their potential use as a biomarker

Mitochondrial DNA in Lung Cancer

Mitochondrial DNA variation and cancer

Bioinformatics Tools and Databases to Assess the Pathogenicity of Mitochondrial DNA Variants in the Field of Next Generation Sequencing

Potential of Mitochondrial Ribosomal Genes as Cancer Biomarkers Demonstrated by Bioinformatics Results

How Far Can Mitochondrial DNA Drive the Disease?

Enhanced mitochondrial genome analysis: bioinformatic and long-read sequencing advances and their diagnostic implications

Research progress of sequencing method for animal mitochondrial genome

Mitochondria transcription and cancer