Lee RG, Rudler DL, Raven SA, et al. (2024) Quantitative subcellular reconstruction reveals a lipid mediated inter-organelle biogenesis network. Nature Cell Biology 26: 57–71.
www.nature.com/articles/s41556-023-01297-4
To survive, cells rely on collections of organelles, such as mitochondria, endoplasmic reticulum, peroxisomes, Golgi apparatus, and more. Each has specific critical functions. However, they don’t work in isolation. They interact by communicating signals and exchanging materials in an orderly manner. Amazingly, little is known about how they interact.
The laboratory of Aleksandra Filipovska used an innovative strategy to explore the interrelationships of several key organelles with mitochondria. They first established cell lines that lacked genes specific for the production of peroxisomes, Golgi, and ER with mitochondria. Then they combined observations by scanning electron microscopy with multi-omics profiling to examine the RNAs, proteins, lipids, and glycogens that were affected by the perturbations. In this way, they could carefully examine how dysfunction in one organelle affected the mitochondria.
Because mitochondria influence so many cell activities, it wasn’t surprising that the researchers found effects in many systems. However, they found that lipid transfers were the most affected component. This very complete paper documents the many metabolic and morphological interactions within cells and provides insights into how they might be involved in diseases.
Questions for Dr. Filipovska
Congratulations on a tour-de-force paper! Your strategy here seems to be an interesting variation on the systems biology: eliminate one component and see what happens. Has anyone else tried this strategy?
Thank you! This specific strategy of genome-wide screening coupled with FIB-SEM imaging and quantification has not been tried previously, to the best of our knowledge. One of the reasons we decided to pursue this strategy was to get an unbiased account of inter-organelle interactions and changes as opposed to traditional targeted analyses.
Your calculations on the number of genes that might be involved in inter-organelle dysfunctions and diseases large enough to be concerning. How do you read that?
We were quite surprised to see so many gene changes, and in addition to the genes reported in Lee et al., we have also analyzed many more that are involved in inter-organelle communication and metabolite exchanges. These numbers are starting to make a lot of sense to us now, given how many of these genes are involved in metabolic and biogenesis pathways that are shared across different organelles. For example, the glycerophospholipid pathway spans the endoplasmic reticulum, Golgi apparatus, mitochondria and peroxisomes. Systems biology methods are key to understanding changes across different organelles, providing a larger picture of metabolic processes and common dysfunctions that manifest in different diseases.
You gave a small hint about future experiments to look at tissues under different metabolic demands near the end of the Discussion. Could you expand on that a little?
We are particularly interested in understanding the tissue-specific defects in multi-systemic diseases, such as mitochondrial diseases where a common molecular defect can manifest in different pathologies depending on the affected tissue in a particular disease. This will enable us to design targeted therapeutics for specific organs. Our study also revealed that different molecular defects can cause mitochondrial dysfunction and that targeted treatments of mitochondrial dysfunction may help alleviate symptoms of other diseases that involve mitochondria.
How did you become interested in mitochondria in the first place?
As an undergraduate student, I became fascinated by the bacterial origin of mitochondria and how they retained a level of autonomy within eukaryotic cells, maintaining a very small genome that is essential for life. At the time, very little was known about the regulation of mitochondrial gene expression, despite the many diseases for which there were no cures or treatments that were caused by defects in the mitochondrial genome. I was inspired to understand how the small mitochondrial genome was regulated that could help identify much needed therapies for mitochondrial diseases that are devastating for the patients, often very young infants and children and their families. I started my career trying to modify the mitochondrial genome and find gene therapies, which of course at the time did not work, and the main reason for this was that there was so little known about the mitochondrial genome, transcriptome and proteome. My goal was to focus on understanding the regulation of the mitochondrial genome in my group to help us devise specific therapies for mitochondrial diseases.
Reference
Lee RG, Rudler DL, Raven SA, Peng L, Chopin A, Moh ES, McCubbin T, Siira SJ, Fagan SV, DeBono NJ, Stentenbach M, Brown J, Rackham FF, Li J, Simpson KJ, Marcellin E, Packer NH, Reid GE, Padman BS, Rackham O, Filipovska A (2024) Quantitative subcellular reconstruction reveals a lipid mediated inter-organelle biogenesis network. Nature Cell Biology 26: 57–71.
Mitochondria have been linked to multiple diseases and aging. Unlike other cellular organelles, mitochondria contain their own DNA, and that DNA can be damaged or mutated so that it no longer functions correctly. A cell might contain up to 1000 copies of the mitochondrial DNA (mtDNA), and the copies might be a mixture of mutated and wildtype mtDNAs, a state known as mtDNA heteroplasmy.
A disease is manifested only if there are too few wildtype copies to maintain normal function. Until recently, few if any treatment options were available.
The discovery of genome editing has initiated a new era of possibilities. However, translating it to living organisms is challenging. Now a group of researchers, led by Carlos Moraes at the University of Miami, applied genome editing to a mouse model of mitochondrial disease.1 These mice are heteroplasmic for a mutation in the gene for the alanine transfer RNA. Thus, the researchers used a cytosine base editor to insert a compensatory edit into the mtDNA. That edit restored the tRNA function by allowing it to reassume its native stem-loop structure. However, at higher doses, off-target editing of mtDNA occurred, emphasizing the need for more precise editing.
These exciting results indicate that genome editing may be a useful therapeutic strategy for treating the previously intractable mitochondrial disorders and that additional research efforts would accelerate the translation of these approaches to the clinics.
Discussion with Dr. Carlos T. Moraes
Your work is the first to use this technique in vivo. What are the challenges to expanding this strategy?
MtDNA base editing has been used mostly to create mtDNA mutations. There are very few mouse models of mtDNA mutations, and even the one we used could not be directly corrected with the approaches available when we started the project.
It is encouraging that your strategy worked in multiple mouse organs, but in a practical sense, is it really scalable to mitochondrial diseases in large organs, such as skeletal muscle?
As in any gene therapy approach, the bottle neck is the ability to deliver genes to the desired organ. Muscle is actually a good target, as AAV viruses have been shown to deliver genes to muscle after intravenous injections. The brain is a bit more difficult.
In your experiment, the disease mutation was well known. Is that true of other diseases or will potential application of your method have to await elucidation of those mutations? This is not meant as a criticism. It’s just a question about expanding its application in the future.
We are very good at detecting disease mutations these days. However, each mutation would require a specialized base editor, which could make the treatment expensive.
Do you have any thoughts on how to reduce the number of off-target edits?
Work around the world is focusing on this question. Some mutations in the base editors have already improved the ratio of on-target/off-target, but further experimentation will be required.
What mitochondrial diseases are the “low-hanging fruit” that might be cured or treated with gene editing?
There are basically two modalities of mtDNA gene editing: 1) mitochondrial nuclease. These cut specifically mutant mtDNA. Precision Biosciences is gearing up for clinical trials on the m.3243G mutation, targeting skeletal muscle. 2) Base editing, as reported in our publication. This maybe more appropriate for homoplasmic mutations, such as Leber hereditary optic neuropathy. These can be treated by AAV injection of the base editor gene into the eye.
How did you first get interested in studying mitochondria?
I have been studying mtDNA disease since my PhD, starting in 1987, at the time where the first mtDNA mutations in patients were described. 38 years and counting!
Reference
1Barrera-Paez JD, Bacman SR, Balla T, Van Booven D, Gannamedi DP, Stewart JB, Mok B, Liu DR, Lombard DB, Griswold AJ, Nedialkova DD, Moraes CT (2025) Correcting a pathogenic mitochondrial DNA mutation by base editing in mice. Science Translational Medicine 17: DOI: 10.1126/scitranslmed.adr0792.
The article “Jumping ‘Numts’ from Mitochondria Can Be Fast and Deadly” by Martin Picard, published online in Scientific American on January 3, 2025, sheds new light on a significant biological phenomenon with profound implications for both health and evolutionary biology.
Mitochondria have their own DNA, but over evolutionary time, they have transferred most of their genes to the host genome. Now research teams led by Martin Picard at Columbia University and Ryan Mills at the University of Michigan have found that this process is continuing to the present day, sometimes with profound implications for our health. 1, 2, 3
Mitochondrial DNA (mtDNA) fragments, known as nuclear mitochondrial DNA segments (numts), integrate into the nuclear genome at a surprisingly high rate, approximately once every 4,000 births.4 This process, called numtogenesis, does not just occur in the distant evolutionary past but continues actively throughout an individual’s life, particularly in brain cells where it might accelerate aging and reduce lifespan.
The significance of this finding lies in its potential to alter our understanding of genetic stability and disease, and of mitochondria themselves. By inserting into critical parts of our genome, Numts disrupt gene function,4 potentially leading to conditions, such as cancer or neurodegenerative diseases. The research also suggests that stress can amplify this DNA transfer, highlighting a new pathway through which environmental factors impact genetic health.
This discovery emphasizes the dynamic interplay between mitochondrial and nuclear genomes. It also reveals another way that mitochondria influence our lives: by changing the genome of our cells. Perhaps most importantly, it demonstrates the need for further investigation into how these interactions shape human health and longevity.
An interview with Martin Picard
- What made you begin your research into Numts?
We reviewed the literature on the many ways in which mitochondria influence gene expression and cellular behaviors. This highlighted a plurality of mechanisms.5 One of them is the transfer of mtDNA pieces to the nucleus, which could destabilize the genome, a hallmark of aging. This process is known to occur on the scale of millennia. So the initial thought was that this might be a mechanism of aging, just like the activation of “retrotransposons” that already live in our genomes cause aging.6 What if mitochondria—and particularly defective mitochondria in primary mitochondrial diseases or acquired mitochondrial defects—spit out pieces of DNA that disrupt the nucleus as a mechanism to speed up evolution but, in the process, end up driving the aging process? - How did you or others isolate this phenomenon?
We took a two-pronged approach. First, we used our Cellular Lifespan Study7 system where we can track the same cells over months in vitro, and we asked whether there were new Numts over this short time scale. The answer was a clear yes. It was particularly noted in cells that came from children with primary mitochondrial disorders, as predicted. Second, we had to check if this mattered to people: does it happen in the human body across a person’s lifetime? In blood, new Numts don’t accumulate. So we looked in the brains of about 1000 people who died of various causes. There, we found several new Numts that were “private” and unique to each person, meaning that there is almost no chance that they were inherited. So they must have arisen during a person’s lifetime. - If Numts are part of a larger process of the interaction between mtDNA and nuclear DNA, can you explain more about how the two genomes are designed how they interact with each other?
The mitochondria shipped pieces of their genome to the nuclear genome during evolution, explaining how it went from a full bacterial genome initially (with hundreds of genes) to only 37 now in our mitochondria. So as a result, the two genomes encode mitochondrial proteins and must be functionally coupled to make well-functioning, specialized mitochondria8 in each cell type. - Where do you see the research going from here, and how many labs, institutes or companies are working on these issues?
This study puts the process of Numt insertions, called Numtptogenesis, on the list of mechanisms of aging and possibly other diseases. In the brain, we found that people with more Numts in the prefrontal cortex, especially, died earlier than people with fewer new Numts. So maybe they regulate how long or how healthy we can live. There are many labs working on Numtogenesis as an evolutionary mechanism, and maybe our findings provide a rationale to studying this process in non-immune cells to understand its relevance to health. - Clearly, this is a frontier. What do you expect we will know more about in a year?
I hope we see researchers with whole-genome sequencing data on non-blood tissues explore and quantify the presence of Numts so we can see to what extent these results extend to other organs and health or disease phenotypes. For example, this may be relevant to mitochondrial diseases.
Comment by Alex Sercel
“The key advance of this work is measuring numtogenesis on a short timescale in tissue culture and in the somatic tissues of individuals over their lifespan instead of looking at the rate of Numt inheritance between generations. Genetic changes, such as Numts, are only observed in subsequent generations if they occur in the sperm or egg and are non-lethal to an individual’s progeny. Said differently, damaging numts are likely to be screened out by natural selection and are less likely to be seen in population-genetics studies.
“This work shows that Numts can occur more frequently than previously thought and may arise in all tissues of the body, specifically long-lived cells like neurons in the brain. The Numts that manifest in a person’s organs over the lifespan bear the potential to damage their cells and tissues because these genetic changes have not been subject to the selective pressures that filter out harmful mutations between generations,” Alex Sercel, PhD, MitoWorld’s Director of Scientific Affairs and a former Postdoctoral Scholar in the Picard Lab at Columbia University Irving Medical Center.
References
- Xue L, Moreira JD, Smith KK, Fetterman JL (2023) The Mighty NUMT: Mitochondrial DNA flexing its code in the nuclear genome. Biomolecules 13(5):753. doi: 3390/biom13050753
- Zhou W, Karan KR, Gu W, et al. (2024) Somatic nuclear mitochondrial DNA insertions are prevalent in the human brain and accumulate over time in fibroblasts. PLoS Biol 22(8): e3002723.
https://doi.org/10.1371/journal.pbio.3002723 - Picard M (2025) Jumping ‘Numts’ from mitochondria can be fast and deadly. Scientific American.
https://www.scientificamerican.com/article/jumping-numts-from-mitochondria-can-be-fast-and-deadly/ - Wei W, Schon KR, Elgar G, et al.(2022) Nuclear-embedded mitochondrial DNA sequences in 66,083 human genomes. Nature 611: 105–114.
https://doi.org/10.1038/s41586-022-05288-7 - Picard M, Shirihai OS (2022) Mitochondrial signal transduction. Cell Metabolism 34: 1620–1653.
https://doi.org/10.1016/j.cmet.2022.10.008 - Gorbunova V, Seluanov A, Mita P et al.(2021) The role of retrotransposable elements in ageing and age-associated diseases. Nature 596: 43–53.
https://doi.org/10.1038/s41586-021-03542-y - Sturm G, Monzel AS, Karan KR, et al.(2022) A multi-omics longitudinal aging dataset in primary human fibroblasts with mitochondrial perturbations. Sci Data 9: 751.
https://doi.org/10.1038/s41597-022-01852-y - Monzel AS, Enríquez JA, Picard M (2023) Multifaceted mitochondria: moving mitochondrial science beyond function and dysfunction. Nat Metab5: 546–562.
https://doi.org/10.1038/s42255-023-00783-1
Mitochondria have long been recognized for the production of energy within cells. But there is more than one type of energy. In a recent review, Jacobs et al. (2024)1 described the results of their studies and that of others that show how mitochondria produce heat as well as chemical energy. Like many motors, mitochondria produce heat at the same time as performing ‘work’. Furthermore, they note that the heat produced by mitochondria might have been important in the evolution of eukaryotes and warm-blooded animals (e.g., birds, mammals).
Jacobs et al. used heat-sensitive dyes and fluorescent proteins to measure the temperature of mitochondria. They found that the temperature of the mitochondria was about 15°C higher than the environment.
They speculate that the heat energy had profound effects on life on Earth. First, eukaryotes resulted from an endosymbiotic relationship between a bacterium and another prokaryotic organism. The bacterium evolved into a mitochondrion. For some time, the assumption has been that the proto-mitochondrion provided additional chemical energy as ATP. However, Jacobs et al. point out that the most likely partner was an archaean that lived near a hydrothermal vent, as proposed originally by Dunn (2017)2. By entering into a partnership, the proto-mitochondrion provided heat that allowed the hybrid organism to move away from the warm waters of the hydrothermal vent. The most compelling aspect is that the transfer of heat does not require the evolution of additional mechanisms for transfer that ATP does. Second, the heat from mitochondria might have allowed warm-blooded animals to evolve.
While still just a hypothesis, the involvement of heat in evolution is an intriguing possibility that suggests lots of additional experiments.
1Jacobs HT, Rustin P, Bénit P, Davidi D, Terzioglu M (2024) Mitochondria: Great balls of fire. The FEBS Journal 291: 5327-5341. https://doi.org/10.1111/febs.17316
2Dunn CD (2017) Some liked it hot: A hypothesis regarding establishment of the proto-mitochondrial endosymbiont during eukaryogenesis. Journal of Molecular Evolution 85: 99-106. https://doi.org/10.1007/s00239-017-9809-5
We appreciate the willingness of lead author Howy Jacobs of Tampere University, Finland, to answer some questions about the significance of heat from mitochondria.For some time, biologists had assumed that energy was the main bartering chip provided by mitochondria to the new eukaryotes. Could heat be an equal driver, as Dunn suggests?
Well, heat IS a form of energy! What we are proposing is that heat produced by mitochondria (or by aerobic bacteria) is a crucial factor in eukaryote biology, that has been largely overlooked. We propose that it has a crucial role not only in evolution but also in many aspects of cell biology, metabolism, immunity, physiology and disease. Dunn’s idea is, of course, really only speculation at this point, but is now supported by the evidence that the ancient host cell that engulfed the mitochondrial ancestor was a moderate thermophile. Note also that it is not completely accurate to regard heat and ATP as alternative forms of energy produced by mitochondria: the energy conserved in the form of ATP and exported to the rest of the cell can also be converted partly or even entirely into heat. For example, the pumping action of the heart depends heavily on mitochondrial ATP production, but only a part of the stored energy drives the contractile action of the heart muscle, the rest being converted to heat which is then carried away in the bloodstream to maintain body temperature. Note that ATP may not be the only ‘energy-rich’ molecule produced by mitochondria that could be considered as a heat store.
Are you planning to measure the temperature of chloroplasts?
We are not planning to do this ourselves but at some point someone needs to do it. But it is going to require a different technology than the ones we have been using, which are based on fluorescent dyes and proteins. The green pigments naturally present in the photosynthetic system would just obliterate any signal. But the existing technologies should suffice to measure the intracellular temperature of aerobic bacteria and also to test our idea that the bacterial cell wall functions as an insulating layer.
Mitochondrial heating seems to open an entirely new way of looking at biology. There are so many manifestations that seem to be covered by it.
I am sure that the topics covered in our short review are far from exhaustive. Here are a few more than could be considered (but undoubtedly there are many others). We have suggested that ‘heat delivery’ could be a mechanism for killing pathogens or infected cells. But maybe it also plays a wider role in programmed cell death, which is a major process in animal and plant development, tumour suppression and stress management. Heat (or its absence) may also play a pivotal role in many diseases, notably neurodegeneration, where the accumulation of protein aggregates is an obligatory step in pathology. And if mitochondria do contain a store of heat-buffering molecules as suggested above, they might also function in some contexts to absorb excess heat coming from the environment.
What is the relationship between mitochondrial heating in brown adipose tissue and other tissues? Are there simply more mitochondria in brown adipose?
Although intensively studied, there are still outstanding questions as to how brown adipocyte mitochondria are ‘repurposed’ to deliver heat rather than ATP. One major mechanism is clearly the expression and activation of the uncoupler protein UCP1, which provides a proton channel in the inner mitochondrial membrane, thus dissipating the proton gradient that is normally driving ATP synthesis, instead releasing all the energy of substrate oxidation as heat. But the normal functioning of the respiratory chain in other tissues lacking UCP1 also generates heat, since only about half of the energy yield from biological oxidation is conserved by proton pumping against the gradient. The rest is converted to heat.
Do you have any speculation about the control systems for this heat?
Our FEBS Journal paper makes some suggestions as to the possible nature of heat sensors inside mitochondria (and the rest of the cell). But for now this is purely in the realm of speculation. The narrow temperature range tolerated by mammalian cells suggests that mitochondrial heat output must be finely regulated, almost certainly by multiple, i.e. redundant systems operating in parallel, as for almost all of the important processes in biology. Such redundancy invariably makes it hard to identify the relevant machinery using the standard tools of genetics.
Is it possible to determine the efficiency of mitochondria?
I don’t much care for this term, since it implies that heat production is somehow a wasteful by-product of mitochondrial metabolism. Thermodynamically it is appropriate, but biologically not! Rather I think we should be thinking about the overall energy transactions taking place, that result in the generation (or absorption) of heat, the production of heat- storage molecules (thus far not identified) and of ‘dual-use’ molecules, predominantly ATP, as a store of useful energy to drive biochemical, mechanical or electrical processes, as well as heat production. Yes, we can start to measure all these, and theoretical considerations can help, but to do this fully we need to identify all the cell’s heat storage systems and derive a more accurate picture of heat flow within the cell, including the temperature gradients within and between cellular compartments – not just mitochondria. Recent findings that mitochondria even within a single cell type can be functionally diverse further complicate the picture. So we are a long way from the end goal of comprehensively profiling energy transactions inside cells, let alone tissues and organs.
How did you get interested in mitochondria?
As so often in science, by studying something else and stumbling into mitochondria by accident. Details on request if you are really interested!