In new research published in Nature, scientists have produced the first comprehensive atlas of the tiny energy producers in the human brain. By slicing a donated brain into more than seven hundred small cubes and measuring both the number of mitochondria and their energy output in each piece, they discovered that mitochondria vary widely across brain regions. They found that the parts of the brain that evolved most recently in our lineage not only contain more mitochondria but also house mitochondria tuned to work more efficiently. This resource, called MitoBrainMap, lays the groundwork for linking energy use in the brain to mood, cognition, and the development of neurological and psychiatric disorders.
Mitochondria are microscopic structures inside nearly every cell that transform nutrients into the energy that powers all cellular activity. In neurons and support cells in the brain, this energy fuels memory formation, visual processing, and emotion regulation. Despite their importance, little was known about how many mitochondria the brain contains, whether they are spread evenly through its many regions, or how they differ in key ways. To begin answering these questions, the research team set out to chart the distribution, density, and functional diversity of brain mitochondria at a resolution comparable to that of a standard magnetic resonance scan.
“There’s an emerging notion that energy is really important to health,” explained Martin Picard, an associate professor of behavioral medicine and director of the Mitochondrial Psychobiology Group at Columbia University. “But we don’t have a way to look at bioenergetics across the entire human brain.”
Picard led the study along with Michel Thiebaut de Schotten, research director at the University of Bordeaux. “My interest in this research topic was sparked by a longstanding desire to bridge the gap between neuroimaging and histological biology,” Thiebaut de Schotten told PsyPost. “Martin, who is a leading expert in mitochondrial research, has been a major source of inspiration. When we met, I was genuinely excited by the possibility of integrating our respective areas of expertise. Collaborating with him offered a unique opportunity to explore how mitochondrial biology could be linked with advanced MRI techniques.”
To build the atlas, the team obtained a frozen coronal slice of the right hemisphere from a neurotypical fifty-four-year-old man with no history of neurological or psychiatric conditions. Working in a subzero environment, they used a computer-controlled milling device to engrave a three-millimeter grid into the tissue and then picked out each cube by hand. In total, 703 cubes were collected, each roughly the size of a large grain of sand.
In three-quarters of these cubes, they measured two markers of mitochondrial quantity—the activity of a key enzyme and the amount of mitochondrial DNA—and three markers of energy-transformation capacity by testing the activity of three enzymes in the respiratory chain. To ensure robust results, each measure was repeated in duplicate and corrected for any batch differences.
In parallel, the researchers performed single-nucleus gene sequencing on samples drawn from four distinct brain regions: the cortex, the hippocampus, the putamen (a movement-control region), and the corpus callosum (the brain’s major communication pathway). This approach yielded data from more than 32,000 individual nuclei, allowing the team to relate mitochondrial measures to specific cell types.
Finally, they combined these laboratory data with MRI scans from nearly 2,000 healthy adults. Using statistical modeling, they linked common imaging signals to the mitochondrial features they had measured. By training the model on 80 percent of their tissue samples and testing it on the remaining 20 percent, they were able to predict mitochondrial density and energy output at the resolution of one cubic millimeter across the entire brain.
The results revealed striking regional and cellular patterns. Gray matter, which contains most of the brain’s cell bodies and connections, showed both higher mitochondrial density and greater energy-production capacity than white matter, which is made up of long projections that carry signals. Within gray matter, evolutionarily newer regions—such as parts of the frontal and temporal lobes involved in complex thought and language—harbored more mitochondria that were tuned for efficient energy production. An exception was the putamen, a deeply buried structure involved in movement control, which displayed exceptionally high mitochondrial markers, perhaps reflecting its dense network of projections and synapses.
“One of the most surprising and exciting findings was the link between mitochondria and brain evolution,” Thiebaut de Schotten explained. “We didn’t expect to see such a clear relationship, and it opens up fascinating new questions about how mitochondrial function may have shaped the development of the human brain over time and how it will interact with future evolution.”
Gene sequencing largely supported these biochemical findings. When the team examined expression levels of genes involved in mitochondrial energy production, they found higher expression in regions with greater enzyme activity. Although different cell types—neurons, support cells, blood-vessel cells, and others—showed subtle differences in mitochondrial gene activity, the strongest driver of variation was the brain region itself. In other words, a neuron in one part of the brain had a gene-activity profile more similar to its neighbors than to neurons in far-flung regions.
“One of the key takeaways from our study is that we now have a better understanding of how mitochondria—the energy processors of our cells—are distributed across the human brain on average,” Thiebaut de Schotten said. “This is important because mitochondria play a vital role in brain function and health. Our next goal is to map these mitochondrial patterns in individual brains, which could eventually help us understand how they relate to brain health, aging, and neurological conditions.”
To test how well their MRI-based model would generalize, the researchers applied it to a slice from the donor’s occipital lobe, which had not been used in model training. The predicted patterns of mitochondrial density and energy capacity closely matched the laboratory measurements, giving confidence that routine brain scans can serve as a window into cellular energy factories. When the model was extended to every cubic millimeter of the standard brain reference, it produced three-dimensional maps that align with known imaging measures of brain evolution and variability.
“This work helps us understand the energetic basis of brain function and brain health,” Picard told PsyPost. “Without energy, the brain is an inert fatty blob. But energized by mitochondria, the mind emerges and allows you to think, feel, and behave. We are, fundamentally, energetic processes. MitoBrainMap v1.0 helps us understand how energy flows to make this possible.”
Despite its strengths, the study has clear limitations. Because the atlas is based on a single human brain, it remains to be seen how mitochondrial patterns vary among individuals of different ages, sexes, and health conditions. The tissue-preparation method for gene sequencing involved aggressive mechanical disruption, which may have biased which cell types survived the process. Future research will need to include samples from multiple donors and refine sequencing protocols to capture a broader array of cell types.
“A major limitation of our study is that our model and conclusions are based on data from a single brain,” Thiebaut de Schotten noted. “While this provides valuable initial insights, studying more brains to understand how mitochondrial patterns vary across individuals is essential. Securing funding for this next phase, which is difficult in the current context, is a key priority, as it will allow us to explore individual differences and strengthen the broader applicability of our findings.”
“The long-term goal is to noninvasively quantify mitochondrial biology using only MRI for both research and health monitoring,” Picard said. “That would be amazing!”
“Our work is showcased at http://www.humanmitobrainmap.bcblab.com,” Thiebaut de Schotten added.
The study, “A human brain map of mitochondrial respiratory capacity and diversity,” was authored by Eugene V. Mosharov, Ayelet M. Rosenberg, Anna S. Monzel, Corey A. Osto, Linsey Stiles, Gorazd B. Rosoklija, Andrew J. Dwork, Snehal Bindra, Alex Junker, Ya Zhang, Masashi Fujita, Madeline B. Mariani, Mihran Bakalian, David Sulzer, Philip L. De Jager, Vilas Menon, Orian S. Shirihai, J. John Mann, Mark D. Underwood, Maura Boldrini, Michel Thiebaut de Schotten, and Martin Picard.
