Individual proteins converge to form the distinctive shape of mitochondria. Design by Bang Wong, Broad Institute / Protein structure images were downloaded from the Joint Center for Structural Genomics website
Within cells, tiny biological machines called mitochondria convert energy from food into a form the body can use. These powerhouses are crucial to metabolism in everything from humans to yeast and are one of the most well-studied “organelles” in cells, yet until recently only half of the estimated 1,500 proteins that make up the machine had been uncovered. The lack of a complete parts list hampered research into the basic workings of this organelle and severely restricted genetic studies of a variety of diseases.
Through a combination of experimental and computational techniques, a team of scientists based at the Broad Institute of MIT and Harvard has now identified nearly all the proteins that comprise mitochondria — the most comprehensive catalog yet of an organelle’s parts. Led by Broad associate member Vamsi Mootha, who is also an assistant professor of systems biology at Harvard Medical School and an assistant professor of medicine in the Center for Human Genetic Research at Massachusetts General Hospital, the team identified nearly 1,100 of the mitochondria’s proteins, roughly a quarter of them previously uncharacterized. Using the catalog, dubbed "MitoCarta," the team was then able to pinpoint a mitochondrial gene that, when mutated, can give rise to a severe metabolic disease that strikes infants, evidence that the tool may have utility both in the lab and in the clinic. The work appears in the July 11 issue of Cell.
The cell's mitochondrial powerhouse is a unique structure that contains its own genome, a reminder of its origins as an ancient bacterium engulfed by a host cell over a billion years ago. Thirteen proteins are encoded within mitochondria's DNA, but most of the organelle’s building blocks reside in the nuclear genome, many well-hidden until recently.
Dysfunctioning mitochondria are implicated in over 50 diseases, such as neurodegeneration, cancer, and diabetes, although the genetic risk factors are difficult to find without a full parts list for the organelle. Moreover, many aspects of normal mitochondrial function remain unknown. As Mootha explained, "For years, a fundamental question has gone largely unanswered — what proteins function in mitochondria?"
Several techniques can be used to find the organelle's genes and proteins, but none are comprehensive. To create the most complete catalog of mitochondrial proteins to date, Mootha gathered experts in experimental biology, proteomics, and computational biology, in order to tackle the problem from many angles. "One of the neatest things about working here at the Broad is that people with totally different skill sets can come together and do what none of us could do alone," said team member Sarah Calvo, an MIT graduate student in computational biology at Harvard and MIT and co-first author of the new work.
"The technologies and analytical methods for measuring proteins on a large scale are really transforming what we can learn about human biology," said Steve Carr, a co-author and director of the Proteomics Platform at the Broad. In the first step of this work, David Pagliarini, a postdoctoral fellow at Massachusetts General Hospital and co-first author of the study, painstakingly isolated mitochondria from 14 different mouse tissues. The organelles were then analyzed by Carr's team of scientists, led by Betty Chang, using a technique known as mass spectrometry (MS) that can systematically identify proteins found in tissues.
Because the proteomic method tends to overlook scarce or variable proteins, the scientists then integrated the MS data with six other genome-wide methods, an approach that revealed hundreds of genes likely to be mitochondrial. The team next wanted to verify these findings in real human cells. By tagging candidates with a fluorescent protein that glows when seen under a microscope, the scientists were able to view which proteins were actually located inside the mitochondria.
The combined approach produced a list of 1098 genes and their expression across 14 different mouse tissues. Roughly a third of these genes in this "MitoCarta" set had not been associated with mitochondria before, and a quarter of the genes had no known function before this study. The catalog — and its hundreds of novel genes — provided the team with a unique opportunity to explore biological pathways at work in mitochondria. “It can take several graduate students’ careers to figure out what a protein does,” said Calvo. "We didn’t want to spend five years on each one, so we took a more global approach."
To further investigate mitochondrial biology, the team chose to focus on a central piece of machinery in the organelle, involving a large assembly of proteins known as "complex I" or CI. Deficiency in CI is a common mitochondrial disease with a variety of effects, including liver disease and possibly some forms of Parkinson’s disease. A handful of CI proteins had been described, but many cases of complex I deficiency had no genetic mutations in those proteins, suggesting others remained undiscovered.
One clever way to identify genes with a given function is to examine their evolutionary history. Genes that are involved in similar pathways are typically gained and lost together. As an evolving species loses the need for complex I, it also discards the genes that encode those proteins. By comparing the genomes of 42 species, from yeast to mammals, the researchers homed in on 19 candidate genes suspected to play a role in complex I.
After disabling four of those genes in human cells, the team found one — C8orf38 — especially interesting. In addition to reducing CI activity when disabled, the gene was found to be mutated in two CI-deficient infant siblings, providing evidence that it is a human CI disease gene.
Prior to this study, researchers had no knowledge of what C8orf38’s function was in the cell. In light of the findings in cells and in patients, the new knowledge on C8orf38 and its role in CI biology is a powerful validation of the research team's approach, said Pagliarini.
The new catalog of mitochondrial proteins may prove useful for scientists searching for other genes associated with mitochondrial disorders, like C8orf38 or perhaps others that underlie more common diseases. Rather than scanning the entire genome for genes of interest, scientists can more quickly identify mitochondrial genes likely to be important. The MitoCarta compendium can also serve as a framework for understanding how mitochondria are made. "Piecing together functioning organelles is a complicated and poorly understood cellular task," said Pagliarini. "But having a high quality parts list is a great first step in understanding that process."
Other Broad Institute scientists involved with the work include Sunil Sheth and Shao-En Ong. Clinical work on the C8orf38 gene was a collaboration with researchers at the University of Melbourne and Royal Children’s Hospital in Australia.