Australasian Science: Australia's authority on science since 1938

Power Failures in Our Cells


In this fluorescent image of a cell, the mitochondrial network is labelled red and the DNA stained green. The bulk of the green staining is in the nucleus but you can see the green spots throughout the mitochondrial network, which represent mitochondrial nucleoids. These are thought to contain multiple mtDNA molecules plus much of the protein machinery needed for mtDNA replication and expression. Credit: Dr Ann Frazier

By David Thorburn

Severe defects in mitochondrial function affect at least one in every 5000 births, but mitochondrial disorders can reveal themselves at any age through a wide range of symptoms and as contributing factors to conditions as disparate as diabetes and Parkinson’s disease.

Anyone watching their waistline or thinking about athletic performance knows that dietary intake and energy expenditure go hand-in-hand. Fewer people appreciate that even at rest we need to constantly generate about 100 kCal of energy per hour. In electrical terms that is equivalent to about 100 W, the same amount used by a bright light globe.

We need this energy to allow our neurons to send messages, our heart to pump and our other organs to perform their roles. To do this we must be able to efficiently convert fuels such as fats, sugars and proteins into a small molecule called ATP, which is our chemical energy store.

Most of the ATP needed for cellular function and survival is generated in power plants within our cells called mitochondria. Each day we generate and consume about 65 kg of ATP, of which our hearts use about 6 kg and our brains about 12 kg.

Our development, growth and survival are thus heavily reliant on our mitochondrial power plants, but not everyone’s mitochondria work well enough. Severe defects in mitochondrial function are the most common group of inherited metabolic disorders, affecting at least one in every 5000 births. That means that at least 60 children born each year in Australia will develop a severe mitochondrial disease during their lifetime.

At the mildest end, the symptoms may only be exercise intolerance. However, every cell needs their mitochondria working properly, so mitochondrial diseases can affect any organ system, alone or in combination. Severe mitochondrial dysfunction can cause neurodegeneration, blindness, deafness, seizures, muscle weakness, heart, liver or kidney failure and early death. Only about half of all patients survive to adulthood.

Milder defects in mitochondrial function are increasingly being recognised as contributors to common conditions like diabetes and Parkinson’s disease. Mitochondrial disorders can present with almost any symptom at any age of onset, so they pose great challenges in diagnosis and treatment.

Most patients suffer from mitochondrial disease because a specific gene is not working properly. Some genetic disorders are relatively simple to diagnose. For example, every patient with cystic fibrosis has a mutation in the CFTR gene, so the diagnostic target is obvious.

In contrast, we know of more than 150 genes in which mutations can cause a mitochondrial disease. Most of these genes are located on the cell’s autosomes, and are present in two copies per cell, one inherited from the mother and one from the father. Others are on the X chromosome, so males have just one copy.

Thirty-seven genes are located in the mitochondrial DNA (mtDNA), which is something of an evolutionary relic. Instead of being in the cell’s nucleus with the other chromosomes, it is present in thousands of copies, all located within the mitochondria themselves. Mitochondrial DNA is inherited only from the mother, so mtDNA mutations are often associated with maternal inheritance.

Given this genetic complexity, how do we diagnose mitochondrial disorders? Genetic testing has been available in Australia for some DNA mutations for more than a decade. However, the complicated genetics and cost of DNA testing means that this has typically been limited to testing for a small number of mutations, mostly in the mtDNA. Relatively few patients are diagnosed quickly and easily by DNA testing of a blood sample, and the next step is usually a muscle biopsy.

Mitochondrial energy generation relies on five molecular machines called Complexes I–V. Each Complex contains between four and 44 different proteins that transport electrons and pump protons to drive ATP synthesis.

Muscle biopsies are tested to determine how well each of these complexes is working. This can confirm the diagnosis, and is helpful to guide genetic testing, but often there are still tens of candidate genes that could still be responsible for the patient’s symptoms. The cost of testing so many genes means that many patients and families remain on a diagnostic odyssey without a clear answer to the genetic cause of their symptoms.

The game changer in genetic diagnosis has been the emergence of new technologies called “next generation” or “massively parallel” DNA sequencing. Instead of having to decide on the most likely handful of genes that we can afford to sequence, we can now sequence panels of 100 genes, 1000 genes or all 20,000 different genes at once.

This has become possible because the cost of sequencing all our genes has fallen from a few million dollars as recently as 2007 to less than $2000 in 2014. In theory this means we should be able to sequence blood first and sift through the sequencing data to pull out the genetic cause without needing to do a muscle biopsy.

This approach is starting to be used more widely, and should be the reality for many patients within a few years’ time. It has been made possible by the extraordinary fall in DNA sequencing costs in the past 6 years and advances in bioinformatic analyses of the data. However, it is still transitioning from being a research tool into routine diagnostic use.

The power of massively parallel sequencing is demonstrated by a study we conducted on 42 infants with deficient activity of one or more of the mitochondrial enzyme complexes. The first part of this study was chosen by the US National Institutes of Health as one of 12 “genome advances of the month” in 2012.

In conjunction with Prof Vamsi Mootha’s group at the Broad Institute of MIT and Harvard, we sequenced more than 1000 genes encoding all the known mitochondrial proteins in each patient. We made genetic diagnoses in ten children with mutations in mitochondrial DNA or seven different nuclear genes previously linked to mitochondrial disease.

We also identified 15 novel candidate genes not previously linked to mitochondrial disease. In that study and subsequent publications we have shown that at least seven of these candidate genes are true novel disease genes. Indeed, half of the patients in whom we have now confirmed a molecular diagnosis have mutations in genes not previously known to cause mitochondrial disease.

The novel disease genes encode proteins with roles in a wide range of different basic functions. Some encode protein subunits of Complexes I–V. Others are needed for proper assembly and function of the Complexes, such as making iron–sulfur clusters or modifying membrane lipids.

Internationally, new mitochondrial disease genes are being identified on at least a monthly basis. This shows that we still have plenty to learn about which genes can cause mitochondrial disease, and that these approaches will identify many more such genes in the next few years.

We don’t yet have effective treatments for most patients, so what are the incentives to develop this technology? First, a diagnosis can end the diagnostic odyssey and provide dignity in diagnosis. In some cases an accurate diagnosis does guide treatment options, and in others it can aid access to additional benefits and support. It can also enable patients to proactively manage their disease progression and quality of life. Identification of the causative genetic abnormality offers the opportunity for informed family planning using prenatal genetic diagnosis and other specialised IVF-based methodologies. Finally it is important to understand the true incidence and impact of mitochondrial disorders, both of which are likely to be highly underestimated.

Treatments for mitochondrial disease remain very unsatisfactory. In 2012, our Cochrane review of more than 700 papers on the treatment of mitochondrial disease concluded: “There is currently no clear evidence supporting the use of any intervention in mitochondrial disorders”. This overstates the reality to some degree, since such reviews require hard evidence in the form of blinded placebo-controlled cross-over studies.

Some symptoms such as seizures, heart disease and acidosis can be managed by standard therapies, at least in the short- to medium-term. Modified diets, vitamin cocktails and modest exercise can also be helpful, and some conditions appear to respond to specific treatments such as arginine to decrease stroke-like events. However, patients remain prone to episodic deterioration, and the severity of the disease continues to progress in most patients.

Encouragingly, an increasing number of new approaches are showing promise in pre-clinical and early phase trials. These approaches mostly aim to boost cellular processes such as mitochondrial biogenesis, quality control, signalling and redox balance via drugs such as rapamycin, bezafibrate, resveratrol, EPI-743 and nicotinamide analogues.

Establishing new genomic technologies in routine diagnostic practice will allow early and accurate genetic diagnosis, early intervention, and trials of new preventative treatments.

David Thorburn is a NHMRC Principal Research Fellow and Head of Mitochondrial Research at the Murdoch Childrens Research Institute.