The study, led by investigators at Shanxi Medical University, tackled a long-standing puzzle in cardiac aging research. Scientists have known that mitochondria, the power plants of the cell, deteriorate as hearts grow old. But the precise mechanism linking that deterioration to the broader decline of the aging myocardium has remained elusive. Because natural aging in mice takes many months and carries high mortality, the team used doxorubicin, a chemotherapy drug well established to produce cardiac damage that closely mimics natural cardiac aging, including mitochondrial dysfunction, enlargement of individual cardiomyocytes, and weakened contraction. They treated young male C57BL/6J mice with repeated low doses of the drug, building up a cumulative regimen that reliably induced senescence in heart tissue.

The first major finding concerned a process called mitophagy, the specialized form of autophagy that clears out damaged mitochondria. Under normal conditions, worn-out mitochondria are tagged, engulfed by autophagosomes, and delivered to lysosomes for degradation, keeping the mitochondrial network healthy. In the aging hearts, electron microscopy revealed a striking accumulation of mitophagosomes—structures that had captured damaged mitochondria but were never processed further. Levels of the autophagy marker LC3B-II and the cargo receptor p62 were simultaneously elevated, a classic signature of a blocked degradative pathway. When the researchers applied chloroquine, a drug that halts lysosomal degradation, to young cells, mitochondrial–LC3B colocalization rose sharply; in aged cells, the drug produced no further increase, indicating that the degradative arm of mitophagy was already saturated. In other words, the aging heart was not failing to initiate mitophagy—it was failing to finish it, creating a traffic jam of half-processed cellular garbage.

The consequences of this congestion were visible throughout the senescent cardiomyocytes. Live-cell imaging showed mitochondria that were shorter, rounder, and more fragmented than those in healthy cells, with fewer branch points and reduced network integrity. Flow cytometry using the MitoSOX probe revealed a significant surge in mitochondrial superoxide, a reactive oxygen species generated by dysfunctional mitochondria. Energy production faltered, oxidative stress mounted, and the senescence program advanced: senescence-associated beta-galactosidase activity climbed, and the proteins p16 and apolipoprotein E—well-established markers of cellular aging—rose in both the mouse hearts and in HL-1 cardiomyocytes treated with doxorubicin in culture.

Enter the therapy. The team isolated mitochondria from murine bone marrow mesenchymal stem cells, a source chosen for its regenerative potential and shared mesodermal origin with cardiomyocytes, and delivered them through the tail vein at 0.5 milligrams per kilogram every two days for six injections. Fluorescent tracing with MitoTracker-labeled mitochondria, monitored by whole-body imaging over 24 hours, confirmed that the organelles reached the cardiac region, and time-lapse microscopy captured them entering cardiomyocytes in culture. The results were dramatic. Echocardiography showed significant recovery of left ventricular fractional shortening and ejection fraction, along with reduced end-systolic volume. Wheat germ agglutinin staining revealed smaller, less hypertrophied cardiomyocytes, senescence staining dropped markedly, and electron microscopy showed fewer mitophagosomes, more intact mitochondria, and restored cristae density. Superoxide levels fell, and the senescence markers p16 and APOE declined in both tissue and cultured cells.

To find the molecular switch behind this rejuvenation, the researchers turned to multi-omics. RNA sequencing of aged versus young hearts identified 27 differentially expressed mitochondrial genes, and one stood out: Bnip3, which encodes an atypical BCL-2 family protein embedded in the mitochondrial outer membrane. BNIP3 normally acts as a mitophagy receptor, recruiting the autophagosome marker LC3B through its LC3-interacting region to initiate mitochondrial clearance. In the aging heart, BNIP3 was dramatically overexpressed, and its expression correlated far more strongly with the autophagy marker Map1lc3b than did PINK1, the canonical ubiquitin-dependent mitophagy pathway, whose Parkin protein levels were unchanged. Crucially, the pattern was conserved across species: hearts from 80-week-old naturally aged mice showed elevated BNIP3, and single-cell RNA sequencing of human heart datasets from GSE156703 and the Heart Cell Atlas revealed significantly higher BNIP3 expression in aged human cardiomyocytes.

Functional experiments confirmed that BNIP3 was not merely a bystander but a driver of senescence. When the researchers forced BNIP3 overexpression in HL-1 cardiomyocytes, the cells developed the full senescent phenotype—elevated p16 and APOE, increased beta-galactosidase staining, and rising levels of LC3B-II and p62 that signaled autophagic congestion. Conversely, when BNIP3 was overexpressed in cells that had received mitochondrial transplantation, the therapy’s benefits vanished: senescence staining returned to aged levels, LC3B–mitochondria colocalization climbed back up, and the accumulated autophagy proteins reappeared. This reversal experiment demonstrated that downregulation of BNIP3 is essential for mitochondrial transplantation to work.

The upstream regulator proved to be an unexpected player. Screening transcription factors altered in the aging heart, the team identified HIF-3α, the least-studied member of the hypoxia-inducible factor family, long regarded mainly as a transcriptional repressor. Here it acted as an activator. Bioinformatic prediction with JASPAR revealed a high-affinity HIF-3α binding site in the Bnip3 promoter, dual-luciferase reporter assays confirmed that HIF-3α could drive Bnip3 transcription, and chromatin immunoprecipitation followed by quantitative PCR pinned down the binding at a specific promoter site. Overexpressing HIF-3α raised BNIP3 levels; silencing it with siRNA lowered them. HIF-1α, the family’s famous member, showed no significant changes, indicating a distinct, non-redundant pathway. HIF-3α was also elevated in aged mouse hearts and aged human cardiomyocytes, and clinical data from heart failure samples in prior literature align with the axis’s upregulation.

Perhaps the most provocative discovery concerns what HIF-3α actually senses: energy. When the researchers measured ATP, they found that mitochondrial transplantation markedly boosted energy levels in aged cardiac tissue. To test whether ATP fluctuations drive the axis, they treated cardiomyocytes with CCCP, a chemical that collapses mitochondrial membrane potential and depletes ATP. ATP levels fell, and HIF-3α rose in response. The emerging model is elegant: as the aging heart’s mitochondria falter and ATP production drops, the energy crisis stabilizes HIF-3α, which switches on BNIP3, floods the cell with mitophagy receptors, and jams the mitophagy flux. Transplanting fresh mitochondria restores ATP, HIF-3α and BNIP3 subside, and the disposal system flows freely again.

The findings arrive as mitochondrial transplantation itself moves into early clinical exploration. Phase I trials in pediatric ischemia-reperfusion injury have already demonstrated preserved cardiomyocyte viability, and systematic reviews support the approach’s potential to restore myocardial bioenergetics, with encouraging safety data showing no immune responses after autologous transplantation. Significant questions remain before aging hearts could be treated this way: the precise mechanism by which HIF-3α senses low ATP is still unknown, whether transplanted mitochondria physically fuse with the recipient’s mitochondrial network has not been directly visualized, and no clinical safety evaluations exist for this application. Delivery routes also pose challenges, from the invasiveness of intramyocardial injection to the low cardiac targeting efficiency of intravenous administration. Still, by identifying the HIF-3α–BNIP3 axis as a druggable control point for mitophagy flux, the study offers something the field has lacked—a concrete molecular explanation for why aging hearts clog their own cleanup machinery, and a demonstration that the clog can be cleared.

Subject of Research: Mitochondrial transplantation therapy for cardiac aging via restoration of mitophagy flux through the HIF-3α–BNIP3 pathway

Article Title: Mitochondrial Transplantation Rejuvenates Aging Heart by Restoring Mitophagy Flux via the HIF‐3α‐BNIP3 Axis

Article References: Mitochondrial Transplantation Rejuvenates Aging Heart by Restoring Mitophagy Flux via the HIF‐3α‐BNIP3 Axis. (n.d.). https://doi.org/10.1111/acel.70720

Image Credits: AI Generated

DOI: 10.1111/acel.70720

Keywords: mitochondrial transplantation, cardiac aging, mitophagy, BNIP3, HIF-3α, senescence, mesenchymal stem cells, doxorubicin, ATP, autophagy flux, cardiomyocytes, Aging Cell

The technical challenge at the heart of the study is that autophagy is notoriously difficult to measure in living humans. Macroautophagy is a multi-step process in which cellular cargo is sequestered into double-membraned vesicles called autophagosomes and delivered to lysosomes for degradation. Common laboratory readouts, such as the abundance of autophagosomes or levels of the autophagosome-associated protein LC3/Atg8-II, are static snapshots that cannot distinguish a cell cranking out autophagosomes from one whose degradation machinery has stalled. The gold-standard approach, pharmacologically blocking lysosomal degradation and watching cargo accumulate, is not feasible in vivo in people. The research team therefore built a multi-pronged strategy using accessible human cells that retain molecular signatures of their donors’ ages.

The team worked with three complementary cell systems drawn from the well-characterized SHOCK healthy aging cohort, whose participants self-report freedom from chronic disease. Low-passage primary dermal fibroblasts from the skin preserve transcriptional signatures of aging. From those same fibroblasts, the researchers generated induced neurons, or iNs, by directly converting the cells with the neuronal transcription factors NGN2 and ASCL1, creating subject-matched pairs of skin and nerve cells. Finally, they isolated peripheral blood mononuclear cells, or PBMCs, freshly drawn from participants, a minimally invasive window into autophagy in circulating T, B, and NK cells. Measuring all three cell types from overlapping groups of donors allowed the investigators to ask whether autophagy is coordinated across a single individual’s body or regulated independently in each tissue.

The first surprise came from transcription. In fibroblasts from older donors, bulk RNA sequencing revealed broad upregulation of autophagy-related gene programs: modules for macroautophagy, autophagosome assembly, and autophagosome maturation were significantly enriched with age, alongside classic senescence-associated signals such as TNFα, IL6-JAK-STAT3, and p53 signaling. Individual genes behaved heterogeneously, with some, including ATG5 and ATG7, decreasing while most macroautophagy genes rose or peaked in mid-to-late adulthood. In iNs, the pattern was weaker and more scattered; autophagy-related pathways were enriched at the pathway level, but curated gene modules were not significantly changed. If transcription were destiny, one would expect both cell types to show boosted autophagic activity. They did not.

Functional measurements told a very different story. Using chloroquine to block lysosomal degradation for two hours and quantifying the resulting accumulation of LC3B-positive structures, the researchers calculated autophagy flux, the true throughput of the pathway. A critical methodological finding emerged immediately: static LC3B-positive foci did not correlate with flux in any cell type, confirming that autophagosome abundance alone is an unreliable proxy for autophagic activity. Flux, not abundance, is what matters, and flux behaved in ways the transcriptomic data could not have predicted.

In fibroblasts, autophagy flux was sex-specific. Young male donors’ fibroblasts trended toward higher flux than young female donors’, but with age the pattern reversed: flux declined significantly in cells from males, while remaining stable in cells from females, so that older donors of both sexes converged. A multiple linear regression with an age-by-sex interaction confirmed that age was significantly associated with declining flux in males, with the interaction term trending toward significance. In induced neurons, the pattern flipped again. Neuronal flux was stable with age in cells from males but increased with age in cells from females, a significant age-by-sex interaction. When the researchers compared subject-matched fibroblasts and iNs from the same donors, flux in one cell type did not predict flux in the other, demonstrating that autophagy is not coordinately regulated across tissues within an individual.

PBMCs added a third pattern. Across twenty-three participants spanning ages twenty-four to eighty, autophagy flux in fresh immune cells showed no significant association with chronological age and no sex differences. What changed instead was variability: flux became markedly more heterogeneous with advancing age, with a trend toward higher values in individuals around seventy and older. This echoes recent independent reports that autophagy flux in freshly isolated PBMCs can rise with age. The researchers verified that the findings were not artifacts of age-related shifts in immune cell composition, either in whole blood or after isolation.

The most provocative results connected these cellular measurements to whole-body physiology. SHOCK participants underwent standardized fitness assessments, including maximal treadmill testing for VO2max, grip strength, walking speed, blood pressure, and the Short Physical Performance Battery. Across the full cohort, the correlations between flux and fitness varied by cell type: higher flux in fibroblasts and PBMCs sometimes tracked with more youthful cardiovascular measures, such as higher maximum heart rate and lower resting heart rate, while higher flux in iNs correlated with weaker grip strength and slower walking speed. But when the analysis focused on participants aged seventy and older, a consistent picture emerged: higher autophagy flux in all three cell types was associated with reduced physical function. Higher PBMC flux accompanied lower VO2max, slower walking speed, and longer chair-stand times; higher fibroblast flux accompanied reduced resting VO2; and higher iN flux accompanied dramatically reduced maximal grip strength. The authors caution that these subgroup correlations rest on small samples, but the direction was strikingly uniform.

Why would elevated recycling signal decline rather than resilience in the oldest adults? The authors propose that the meaning of high flux depends on context. In younger people, high flux may reflect abundant autophagic capacity, a well-oiled maintenance system. In the very old, elevated flux may instead represent increased autophagic demand, a compensatory response to accumulating cellular stress and damage, or a struggle to keep pace with degradation needs. This reframing carries real consequences for interventions. Strategies aimed at simply boosting autophagy, whether pharmacological or dietary, may need to account for biological age and tissue context, because pushing flux higher in an eighty-year-old could be counterproductive if the elevation already reflects stress rather than capacity.

The study’s final act tested whether the system is modifiable. In a pilot within the STRONG fall-prevention exercise cohort, five participants aged seventy-seven to eighty-eight completed twelve weeks of mild exercise, one-hour weekly sessions focused on balance, strength, and posture. At baseline, four of the five showed PBMC autophagy flux more than one standard deviation above the SHOCK cohort mean, consistent with the link between elevated flux and reduced function. After the program, flux decreased in four of five participants, falling back within one standard deviation of the younger cohort average, while SPPB scores improved. Immune cell composition was unchanged, suggesting the effect was not a shift in cell populations. Intriguingly, the single participant whose flux rose was the youngest and most physically capable at baseline. The sample is tiny and the timepoints limited, but the implication is bold: autophagy flux in human immune cells is not a fixed consequence of aging but a dynamic, responsive parameter, one that mild exercise can nudge back toward a youthful setpoint while physical performance improves. Larger studies with frailty indices, cognitive measures, and single-cell resolution will be needed to confirm when rising flux is beneficial and when it is a warning light, but this work makes clear that the old dogma of universal autophagic decline cannot survive direct measurement.

Subject of Research: Sex- and cell-type-specific remodeling of autophagy flux during human aging and its relationship to physical function

Article Title: Autophagy Flux Is Remodeled Sex‐ and Cell Type‐Specifically During Human Aging, and Is Linked to Reduced Physical Function in Older Adults

Article References: Moreno, T. M., Heimler, S. R., Moran, R. J., Barkai, H. S., Scandalis, L., Traxler, L., Neil, A., Dozier, S., Bergstrom, J., Colas, A. R., Ranade, S. S., Bang, A. G., Mertens, J., Wing, D., Molina, A. J., & Kumsta, C. (2026). Autophagy Flux Is Remodeled Sex‐ and Cell Type‐Specifically During Human Aging, and Is Linked to Reduced Physical Function in Older Adults. Aging Cell, 25(10), Article e70734. https://doi.org/10.1111/acel.70734

Image Credits: AI Generated

DOI: 10.1111/acel.70734

Keywords: autophagy, aging, autophagy flux, induced neurons, fibroblasts, PBMCs, physical function, exercise, chloroquine assay, LC3B, sex differences, healthy aging