Mitochondrial DNA Mosaics May Serve as a Molecular Clock of Human Aging

Deep inside nearly every cell in the human body, thousands of small circular genomes are quietly keeping time. A review published in Aging Cell makes the case that these mitochondrial genomes, long studied as a source of age-related cellular damage, can be read as something far more precise: a molecular clock of aging that is written, cell by cell, into a faithfully copied and segregated genome. The argument rests on a deceptively simple observation. Mitochondrial DNA mutations arise continuously throughout life as rare, individually invisible events. Yet within many aging cells they undergo clonal expansion, multiplying until they crowd out the healthy copies and cripple the cell’s power supply. The result is a tissue mosaic, in which scattered single cells are respiratory-deficient while their neighbors function normally, and the pattern of that mosaic tracks chronological age with striking fidelity.

The mitochondrial genome occupies a peculiar position among the body’s genetic material. Unlike the two copies of each nuclear chromosome, mitochondria carry hundreds to thousands of copies of their own compact genome per cell, encoding thirteen proteins of the oxidative phosphorylation machinery along with the RNA molecules needed to build them. Because this multi-copy system complements defective genomes with functional ones, a newly arising mutation is initially diluted to near-undetectable levels and produces no biochemical consequence. The biologically relevant quantity, the review emphasizes, is not the tissue average but the per-cell distribution, and in particular the fraction of cells in which a mutant clone has crossed the functional threshold for that particular mutation. This threshold behavior explains a long-standing paradox: bulk measurements of heteroplasmy, the coexistence of mutant and wild-type mitochondrial genomes, often appear too low to matter, even though individual cells can be nearly homoplasmic for a damaging variant and frankly respiratory-deficient.

Two mechanistically distinct classes of somatic lesion drive the clock, and they behave differently. Point mutations, dominated by a G-to-A mutational signature indicative of replication errors rather than oxidative damage, accumulate across essentially all tissues and constitute the bulk of the cryptic, cell-unique burden. Large-scale deletions are rarer at the molecular level but expand with particular efficiency in long-lived post-mitotic cells such as skeletal muscle fibers and the dopaminergic neurons of the substantia nigra, where a single deleted species can come to dominate a cell and abolish respiration. Classic studies showed that these neurons accumulate high deletion loads during normal aging, with deletion burden significantly higher in respiratory-deficient neurons than in healthy ones. In muscle, decades of debate about how deletions expand have been resolved by models of density-dependent stochastic survival that require no replicative advantage at all, while droplet digital PCR has linked clonal deletion load directly to focal respiratory failure, fiber by fiber.

Proliferative tissues convert the same mutational process into a visible clonal history. The human colonic crypt is maintained by a handful of basal stem cells whose descendants migrate upward, so a mutation fixed in a stem cell is propagated through the entire crypt. Comprehensive analysis of aging colorectal epithelium showed that clonally expanded point mutations arise from early to mid-life and drive a substantial burden of respiratory-chain dysfunction by old age. Because adjacent respiratory-deficient cells are clonally derived, naturally occurring mitochondrial mutations act as endogenous lineage labels, and their distribution has even been used to quantify human intestinal stem-cell numbers and crypt fission dynamics. The convergence of proliferative and post-mitotic tissues on the same outcome, age-dependent, clonal, threshold-limited respiratory deficiency, forms the empirical core of the mitochondrial clock concept.

Causality, long contested, was decisively addressed by the mitochondrial mutator mouse, which expresses a proofreading-deficient polymerase and accumulates a markedly increased mutation load, developing a premature-aging phenotype. Independent lines confirmed that mutation accumulation drives apoptosis in aging tissues, and heterozygous models that accumulate clonally expanded point mutations without gross structural instability reproduce a respiratory-deficiency phenotype closely paralleling aged human colon. A multi-tissue duplex-sequencing survey cataloged more than 89,000 independent somatic mitochondrial mutations in aging mice, revealing tissue-specific accumulation rates that did not track mitochondrial content and confirming that the dominant aging signature reflects replication errors rather than reactive oxygen species, a finding with sobering implications for antioxidant-based anti-aging strategies.

How clonal expansion actually proceeds remains an active frontier. Because mitochondrial DNA turns over continuously and is randomly partitioned, any variant’s frequency performs a random walk, and some drift toward fixation by chance even in non-dividing cells. Recent single-cell work, combining precise mitochondrial base editing with ultra-high-throughput heteroplasmy tracking, showed that in dividing cells selection rather than simple drift shapes population heteroplasmy, with the direction of selection on a given variant depending wholly on the cellular environment. Single-cell multi-omic profiling of human immune cells carrying pathogenic variants revealed dynamic purifying selection that varies across cell types. The contemporary synthesis is that drift and selection operate in different tissues and contexts, and that many apparent somatic variants are pre-existing heteroplasmies carried as passengers during clonal expansion.

The clock, it turns out, is not a single dial but several. Multi-tissue profiling across dozens of human tissues from hundreds of individuals revealed that mitochondrial clonal mosaicism develops with two distinct, tissue-dependent aging signatures. In proliferative tissues with constant turnover, aging is marked by accelerated accumulation of sporadic mutations and their clonal expansion, with implications for tumorigenesis, since pan-cancer data show that truncating mitochondrial mutations are positively selected in kidney, colorectal, and thyroid cancers. In post-mitotic, high-energy tissues such as heart and brain, mutations accumulate preferentially at deterministic, recurrently mutated hotspots. Any practical biological-age estimator built on heteroplasmy would therefore need separate, tissue-specific calibration, combining a point-mutation term for blood and renewing epithelia with a deletion term for muscle and brain, an integration the authors explicitly flag as an open calibration problem rather than a solved feature.

Population-scale evidence has moved heteroplasmy from tissue-biology curiosity to clinical trait. Sequencing across roughly 300,000 individuals showed that heteroplasmic variants accumulate sharply after about age seventy and that clock rate is itself partly heritable under nuclear genetic control. In nearly 195,000 UK Biobank participants, heteroplasmy was associated with an approximately 1.5-fold increased risk of all-cause mortality, along with cancer incidence and cancer-specific mortality. A 2025 analysis of more than 369,000 participants linked accumulated heteroplasmic mutations to chronic kidney disease severity and acute kidney injury through suppressed purine metabolism amenable to metabolic rescue. A two-step mechanism, drawn from genome-wide analysis of blood in roughly 750,000 people, ties the picture together: cells first acquire low-level cryptic mutations, which then become detectable when the clones carrying them proliferate, unifying somatic mitochondrial accumulation with clonal hematopoiesis and explaining why the clock appears to accelerate in late life.

Can the clock be slowed? The review maps three classes of intervention. Caloric restriction slows cryptic mutation accumulation across single cells, and exercise improves mitochondrial capacity in aging muscle. The mitophagy activator urolithin A has improved muscle strength and endurance in randomized trials in middle-aged and older adults and enhanced cardiac mitochondrial quality in preclinical and human studies, potentially biasing the surviving mitochondrial pool toward functional genomes even without lowering the mutation rate. Most ambitiously, CRISPR-free mitochondrial base editors and TALE-linked deaminases now allow programmable editing of the mitochondrial genome in principle, making the intracellular mitochondrial population a manipulable substrate, though delivery and off-target concerns keep such approaches at proof-of-concept for somatic aging. Positioned alongside epigenetic methylation clocks, the mitochondrial clock offers something orthogonal: a high-copy, single-cell-traceable genome whose damage is mechanistically tied to bioenergetic failure, inflammaging through cGAS-STING signaling, senescence, and hard clinical outcomes. The authors argue that the path forward lies less in discovering new phenomena than in the unglamorous work of calibration, standardization, and longitudinal validation that could turn a compelling biological signal into a dependable, blood-readable instrument for measuring how fast an individual is aging.

Subject of Research: Clonal mosaicism of mitochondrial DNA heteroplasmy as a tissue-specific molecular clock of biological aging.

Article Title: Clonal Mosaicism of Mitochondrial DNA Heteroplasmy as a Molecular Clock of Aging

Article References: Chang, R., Tsai, A. P., Wang, B., Tsui, K.-H., Pang, C.-Y., & Li, C.-J. (2026). Clonal Mosaicism of Mitochondrial DNA Heteroplasmy as a Molecular Clock of Aging. Aging Cell, 25(9), Article e70718. https://doi.org/10.1111/acel.70718

Image Credits: AI Generated

DOI: 10.1111/acel.70718

Keywords: mitochondrial DNA, heteroplasmy, clonal mosaicism, molecular clock, aging, somatic mutations, clonal expansion, mitophagy, inflammaging, biological age, single-cell sequencing, mtDNA deletions