Skip to main content

Deciphering and reversing the consequences of mitochondrial DNA damage

Article Category

Article available in the folowing languages:

Predicting health based on the fitness of mitochondria DNA

Damage to mitochondria DNA has serious consequences for human health. The RevMito project pinpointed nutritional pathways that influence this process and developed a machine-learning based predictor of likely health outcomes.


Mitochondrial DNA (mtDNA) encodes proteins that are important for the efficient conversion of food into energy, helping to power cells. At least one in 5 000 people are born with mtDNA mutations that lead to metabolic disease. If mtDNA is damaged, it often leads to bioenergetic problems, potentially affecting cell and tissue maintenance. Moreover, it is known that mtDNA mutations accumulate in several organs as people age. The EU-supported RevMito project was set up to discover how clinicians could improve the health of cells with mtDNA damage. “This European Research Council project enabled us to progress from understanding mtDNA damage using microbiology-based approaches, to pioneering bioinformatics. These can help clinicians determine whether a patient suffers from metabolic disease,” says Cory Dunn, project coordinator. The team’s work on the computational prediction of mtDNA variant pathogenicity is undergoing peer review, prior to publication, while the technology itself has been licensed to the project host, the University of Helsinki.

From yeast to hummingbirds, to patient diagnostics

The project started from the premise that because mtDNA is a hub of cellular metabolic activities, the nutritional status of cells would likely control the outcome of mtDNA damage. The team used fungi in a number of their experiments, because as Dunn explains: ”We share many proteins and pathways with yeast, so what we learn using this approach is often applicable to human health.” Additionally, Saccharomyces cerevisiae (the species of yeast used) cells divide quickly in the laboratory, offering rapid results. RevMito deleted selected yeast genes and then tested them under different conditions to reveal which might promote or inhibit the survival of cells depleted of mtDNA. They also tested the genetic profiles of the yeast’s cells using transcriptomics. Their findings provided support for the idea that the amount of glucose present influences the outcome of mtDNA damage. Next, to further understand the consequences of mtDNA variation, the team turned their attention to hummingbirds, as in flight they burn more energy, for their size and weight, than any other bird or mammal. RevMito developed and applied software-based methods to find mtDNA changes that may be linked to the aerobic power of hummingbirds. The team then collaborated with a structural biologist to better understand their potential impact on hummingbird bioenergetics.

Towards state-of-the-art diagnostics

A challenge facing researchers is that many people routinely carry unusual mtDNA variants that may or may not be linked to disease, and the genetic properties of mitochondrial genomes make it difficult for clinicians to link genotype to symptoms. RevMito, building on the computational approaches developed during their hummingbird study, has developed a machine language-based classifier that could help clinicians decide which mtDNA variants cause metabolic disease. “Predicting the pathogenicity of common genetic variants is a key aspect of genomic medicine. Further development of our findings is essential to take advantage of novel diagnostic methods,” says Dunn. As mitochondrial disease remains difficult to diagnose and treat, the RevMito laboratory continues to conduct fundamental studies into how mtDNA mutations can be classified and how the cellular outcome of mtDNA damage can be altered.


RevMito, mitochondria, mitochondrial DNA, disease, hummingbird, yeast, gene, mutations, metabolic, cell, mtDNA, bioenergetic

Discover other articles in the same domain of application