Mitochondrial biology: what changed in 2025

Mitochondria have been on the cover of more journals in the past eighteen months than at any point since the 1990s — and for once, the hype is downstream of real findings, not the other way round. The 2023 update to the canonical “Hallmarks of Ageing” framework kept mitochondrial dysfunction as one of twelve interlocking processes that drive biological ageing, but it also flagged how tightly it sits in the middle of the network: nudging mitochondria moves nearly every other dial, from stem cell exhaustion to chronic inflammation (López-Otín et al., 2023, Cell).

The interesting thing about 2024–2025 is not that the field discovered something new about ATP. It is that several long-standing assumptions — that mitochondria stay inside the cell that made them, that mitophagy is roughly constant across tissues, that human trials would eventually match the mouse data — have all been complicated or partly overturned. This piece is a tour of what actually changed, with the receipts.

The basics: what mitochondria actually do beyond “powerhouse of the cell”

The textbook line about mitochondria being the powerhouse of the cell is true but increasingly insufficient. Mitochondria run oxidative phosphorylation, yes, but they also act as signalling hubs: they regulate calcium handling, set the threshold for programmed cell death by releasing cytochrome c, generate reactive oxygen species that double as second messengers, and host parts of haem, steroid and iron–sulphur cluster biosynthesis. They contain their own circular genome — mitochondrial DNA, or mtDNA — inherited maternally and present in hundreds to thousands of copies per cell.

That last detail matters more than it sounds, because mtDNA is replicated by its own machinery, repaired less effectively than nuclear DNA, and sits next to the electron transport chain where reactive oxygen species are made. The result is a mutation rate roughly ten to seventeen times that of the nuclear genome, and a steady accumulation of damaged copies across decades of life (Bae et al., 2024, Nature Genetics). When researchers talk about mitochondria as a driver of ageing, they are usually pointing at some combination of these layers — energetics, signalling and genome integrity — not just ATP.

Mitophagy and ageing — the imaging story in 2024–2025

Mitophagy is the selective autophagy of mitochondria: damaged organelles get tagged, engulfed and broken down in lysosomes so the cell can recycle the parts. The leading model says that mitophagy declines with age, allowing dysfunctional mitochondria to accumulate, and that this accumulation contributes to age-related tissue decline.

For years that model rested mostly on biochemical readouts in cultured cells. What changed is the imaging. Transgenic reporter mice carrying mitochondrially targeted pH-sensitive probes such as mt-Keima have made it possible to count mitophagy events directly in living tissue (Sun et al., 2017, Nature Protocols). In 2024, work using these tools showed that basal mitophagy is suppressed during cellular senescence and in naturally aged cells, and — importantly — that a small molecule targeting the cargo receptor p62 could partially restore it and reverse several senescence-associated phenotypes in human cells (Yamashita et al., 2024, Developmental Cell).

The picture is not uniform. Mitophagy decline appears tissue-specific and cell-type specific; astrocytes, for example, show a particularly steep decline that loads them up with damaged organelles under stress (Quintas-Granados et al., 2024, Mechanisms of Ageing and Development). A 2025 review in npj Aging put the synthesis bluntly: in senescent cells, mitochondrial fusion goes up and mitophagy goes down, reducing average organelle quality (Pereira et al., 2025, npj Aging). So the field’s confidence that mitophagy declines with age is reasonably high. The harder questions — how much of human ageing this explains, and whether boosting mitophagy in older humans translates to clinically meaningful outcomes — remain genuinely open.

Mitochondrial transfer — the surprising finding

If there was one direction-of-travel result in the past year, it was the realisation that mitochondria are not strictly cell-bound. Cells exchange whole organelles across thin membranous bridges called tunnelling nanotubes, and the implications are stranger than they first appear.

In a Cell paper from 2024, bone marrow stromal cells were shown to form nanotube connections with T cells and physically donate functional mitochondria into them — boosting the T cells’ metabolic capacity and antitumour activity (Baldwin et al., 2024, Cell). The flip side appeared a few months later in Nature: tumour cells exploit the same machinery in reverse, transferring mitochondria carrying mtDNA mutations into infiltrating T cells. Because those donated mitochondria arrive bundled with mitophagy-inhibiting molecules, the recipient T cell cannot get rid of them, gradually becomes homoplasmic for the mutant mtDNA, and slides into a senescent, metabolically broken state — a previously unknown mode of immune evasion (Ikeda et al., 2025, Nature).

None of this is fringe biology any more. Nature Metabolism published consensus nomenclature recommendations for mitochondrial transfer and transplantation in 2024, an editorial sign that the field expects this to be a research category, not a curiosity (Borcherding & Brestoff, 2024, Nature Metabolism). What it means for ageing specifically is less settled. There are early hints that intercellular organelle exchange contributes to tissue repair and possibly to age-related decline, but most of the strongest data are still in cancer and immunology.

The NAD+ booster trial picture — what RCTs in humans actually show vs marketing

This is the part of the field where the gap between the lab and the marketing copy is widest, so it is worth being careful.

NAD+ is a coenzyme central to oxidative metabolism and to the activity of enzymes (sirtuins, PARPs, CD38) that consume it. Tissue NAD+ levels fall with age in rodents and in some human tissues. From that starting point, a small industry has sprung up around precursors — nicotinamide riboside being the most-studied — sold on the premise that raising NAD+ should slow ageing. What do the human randomised controlled trials actually show?

What is reasonably well established

Nicotinamide riboside, taken orally, durably raises whole-blood NAD+ in a dose-dependent way and appears safe at the doses tested. A 2018 crossover trial in middle-aged and older adults established the basic pharmacology (Martens et al., 2018, Nature Communications), and a longer safety study in healthy overweight adults extended the picture to eight weeks at up to a gram per day (Conze et al., 2019, Scientific Reports).

What is more equivocal

Raising NAD+ is not the same as showing a clinical benefit. A 2023 review in Science Advances looked across the published human trials and concluded that, despite the volume of supportive preclinical literature, oral nicotinamide riboside has shown few clinically relevant effects in humans, and that the field has a tendency to over-state the importance of small or noisy findings (Freeberg et al., 2023, Science Advances). A 2024 trial in peripheral artery disease did report a meaningful improvement in six-minute walk distance with nicotinamide riboside, which is encouraging — but it is a single trial in a specific disease population, not evidence of a general anti-ageing effect (McDermott et al., 2024, Nature Communications).

The story for urolithin A — a gut-microbiome-derived metabolite of ellagitannins that activates mitophagy in cell and animal models — is similar in shape. The ATLAS trial in healthy middle-aged adults reported significant improvements in muscle strength and some endurance measures, though the pre-specified primary endpoint (peak power output) did not reach significance, a nuance that often gets lost in summaries (Singh et al., 2022, Cell Reports Medicine). A more recent randomised, placebo-controlled trial reported effects on the immune system — specifically, expansion of memory and naive T-cell populations in older adults — which is genuinely interesting and biologically coherent with a mitophagy mechanism, but again a single trial in a specific outcome (D’Amico et al., 2025, Nature Aging).

The honest summary

Across the boosters and mitophagy activators that have made it into human trials, the picture is: pharmacology works as advertised (you can move the biomarker), safety looks acceptable at the levels studied, and clinically meaningful, replicated outcome effects are still thin. That is a very different sentence from “NAD+ reverses ageing.” Readers should be sceptical of any source that compresses one into the other.

What is still open in 2025–2026

A few threads worth watching.

mtDNA mosaicism and clonal expansion. Sequencing single-cell-derived clones from human tissues has shown that mtDNA variants are surprisingly common in healthy individuals, that some are inherited from the fertilised egg and propagate through embryonic lineages, and that the age-related rise in detectable heteroplasmy in blood is at least partly driven by clonal expansion of haematopoietic stem cells carrying mutations as passengers (Bae et al., 2024, Nature Genetics; Hong et al., 2024, Nature Communications). How much of organism-level ageing this causes, versus reflects, is unresolved.

The mitochondrial unfolded protein response. UPRmt is a retrograde stress signal: when proteostasis inside mitochondria fails, the nucleus is told to upregulate chaperones and proteases. In C. elegans and mice, modest activation of UPRmt extends lifespan; in human cells, the same pathway can also drive senescence depending on context. A 2024 review captures the current ambivalence: UPRmt is clearly conserved and clearly important, but whether transient activation is universally pro-longevity or sometimes the opposite is an active debate (Anand et al., 2024, Frontiers in Cell and Developmental Biology).

The translational gap. The hardest, most persistent problem in this field is that a lot of the strongest mechanistic claims come from worms and mice, where interventions can extend lifespan by tens of per cent. Humans live decades, have heterogeneous diets and microbiomes, and run trials that last months. The gap between “moves a biomarker in an eight-week RCT” and “extends human healthspan” is wide, and any honest editorial reading of 2025 mitochondrial biology has to sit inside that gap rather than pretending it has been closed.

What changed in the past year is not that the gap closed. It is that the field’s tools — in vivo mitophagy imaging, single-cell mtDNA sequencing, mitochondrial transfer assays — got sharp enough that the next decade of human studies can ask better questions. That is real progress. It is also, deliberately, less dramatic than the supplement aisle would have you believe.

Sources

• López-Otín, C., Blasco, M.A., Partridge, L., Serrano, M., Kroemer, G. (2023). Hallmarks of aging: An expanding universe. Cell, 186(2), 243–278.
• Bae, J.H., et al. (2024). Mitochondrial DNA mosaicism in normal human somatic cells. Nature Genetics.
• Sun, N., Malide, D., Liu, J., Rovira, I.I., Combs, C.A., Finkel, T. (2017). A fluorescence-based imaging method to measure in vitro and in vivo mitophagy using mt-Keima. Nature Protocols, 12(8), 1576–1587.
• Yamashita, S.-I., et al. (2024). Suppressed basal mitophagy drives cellular ageing phenotypes that can be reversed by a p62-targeting small molecule. Developmental Cell.
• Quintas-Granados, L.I., et al. (2024). Accumulation of damaged mitochondria in ageing astrocytes due to mitophagy dysfunction. Mechanisms of Ageing and Development.
• Pereira, B.I., et al. (2025). Mitochondrial dysfunction in cellular senescence: a bridge to neurodegenerative disease. npj Aging.
• Baldwin, J.G., et al. (2024). Intercellular nanotube-mediated mitochondrial transfer enhances T cell metabolic fitness and antitumour efficacy. Cell.
• Ikeda, H., et al. (2025). Immune evasion through mitochondrial transfer in the tumour microenvironment. Nature.
• Borcherding, N. & Brestoff, J.R. (2024). Recommendations for mitochondria transfer and transplantation nomenclature and characterisation. Nature Metabolism.
• Martens, C.R., et al. (2018). Chronic nicotinamide riboside supplementation is well-tolerated and elevates NAD+ in healthy middle-aged and older adults. Nature Communications, 9, 1286.
• Conze, D., Brenner, C., Kruger, C.L. (2019). Safety and metabolism of long-term administration of NIAGEN in a randomised, double-blind, placebo-controlled clinical trial of healthy overweight adults. Scientific Reports, 9, 9772.
• Freeberg, K.A., et al. (2023). What is really known about the effects of nicotinamide riboside supplementation in humans. Science Advances.
• McDermott, M.M., et al. (2024). Nicotinamide riboside for peripheral artery disease: the NICE randomised clinical trial. Nature Communications.
• Singh, A., et al. (2022). Urolithin A improves muscle strength, exercise performance, and biomarkers of mitochondrial health in a randomised trial in middle-aged adults. Cell Reports Medicine, 3(5), 100633.
• D’Amico, D., et al. (2025). Effect of the mitophagy inducer urolithin A on age-related immune decline: a randomised, placebo-controlled trial. Nature Aging.
• Hong, Y.S., et al. (2024). Mitochondrial heteroplasmy improves risk prediction for myeloid neoplasms. Nature Communications.
• Anand, A., et al. (2024). Mitochondrial unfolded protein response (UPRmt): what we know thus far. Frontiers in Cell and Developmental Biology.