CMA works with remarkable precision. The cytosolic chaperone HSC70 patrols the cell interior, binding proteins that carry a KFERQ-like amino acid motif and delivering them to LAMP2A, a receptor embedded in the lysosomal membrane. When substrates arrive, LAMP2A molecules multimerize into a translocation complex that unfolds and threads each protein into the lysosomal lumen for degradation. This process is not merely housekeeping. CMA regulates the cell cycle, metabolism, cellular differentiation, immune activation and even circadian rhythms. It also declines with age in most tissues, and its malfunction has been linked to neurodegenerative disease, metabolic syndrome and atherosclerosis. What remained unknown until now was how this decline shapes the fate of senescent cells, the growth-arrested cells that accumulate in aging organs and drive inflammation and tissue dysfunction.

The team began with mouse ear fibroblasts, a reliable primary model for studying aging outside the confounding effects of replicative senescence. Using fibroblasts from transgenic KFERQ-Dendra reporter mice, in which a fluorescent CMA substrate lights up as puncta inside lysosomes, the researchers tracked CMA activity as cells entered senescence. In fibroblasts from young, four-month-old mice, senescence induction with the CDK4/6 inhibitor palbociclib triggered a robust upregulation of CMA, consistent with earlier work showing that the pathway is activated when cells remodel their proteome. Fibroblasts from twenty-three-month-old mice, however, told a different story. They displayed lower basal CMA activity and completely failed to boost the pathway in response to the senescence stimulus. The deficit was not due to reduced reporter expression but to impaired delivery and internalization of substrates into lysosomes, confirming that aged cells lose the ability to engage CMA precisely when they need it most.

To dissect the consequences, the researchers stably knocked down Lamp2a, the rate-limiting component of CMA, in untransformed NIH3T3 fibroblasts. Strikingly, these CMA-deficient cells acquired several senescence-like features even before any senescence stimulus: elevated senescence-associated beta-galactosidase staining, nuclear accumulation of the chromatin protein HMGB1, partial increases in p21 and the DNA damage marker gamma-H2A.X, and nuclear indentations. Yet they kept proliferating normally and arrested only after palbociclib treatment, indicating that CMA loss alone does not drive full conventional senescence. Principal component analysis of multiple readouts placed the knockdown cells in an intermediate cluster between healthy controls and senescent cells. The message was clear: CMA is dispensable for initiating senescence, but its loss warps the senescence program itself, producing cells with properties that resemble those of aged senescent cells.

Quantitative proteomics revealed just how deep that resemblance runs. Blocking CMA in young cells reproduced nearly half of the proteome changes seen in aged cells, with shared alterations concentrated in lipid metabolism, oxidative phosphorylation, protein translation, vesicular trafficking and extracellular matrix remodeling. When young control, young CMA-deficient and aged fibroblasts were all pushed into senescence, the CMA-deficient and aged groups showed forty-five percent concordance in their proteome remodeling, with close to four hundred proteins changing in parallel and only about two hundred differing between them despite their twenty-month age gap. By inhibiting lysosomal proteolysis and comparing degradation profiles, the team identified sixty-four proteins that are no longer degraded in lysosomes in aged cells, more than half of them bona fide CMA substrates. In other words, a substantial fraction of the aging proteome is simply a backlog of proteins that the failing CMA machinery can no longer clear.

Metabolomics added a second layer of evidence. Under basal conditions, CMA-deficient fibroblasts phenocopied many metabolic features of aged cells, including changes in energy production, carbohydrate and glycosylation pathways, lipid and membrane biosynthesis, and amino acid and methylation pathways. Upon senescence induction, young control cells mounted a characteristic metabolic shift, engaging fatty acid beta-oxidation, while CMA-deficient and aged cells failed to make that transition and instead showed unique changes in the tricarboxylic acid cycle, aromatic amino acid metabolism and nucleotide metabolism. An integrative analysis combining proteomic and metabolomic datasets confirmed that proteins normally degraded by CMA during senescence accounted for a large share of these metabolic differences, particularly the failure to engage fatty acid and one-carbon metabolism. Because the senescence-associated secretory phenotype, or SASP, depends on this metabolic reprogramming, the stage was set for a dramatic effect on what senescent cells secrete.

That effect materialized in the secretome. Using an ER-TurboID system that biotinylates proteins inside the endoplasmic reticulum, the researchers separated conventionally secreted proteins from those released through unconventional routes such as exosomes and lysosomal docking. CMA-deficient cells showed a deregulated basal secretome and a profoundly altered response to senescence: rather than completing a normal SASP, they failed to downregulate thirty-nine conventionally secreted proteins and upregulated forty-seven others involved in cell migration, protein folding and lysosomal enzymes, with pronounced enrichment of immune-related factors. Unconventionally secreted nuclear proteins appeared in the medium, a selective release not attributable to cell death. Functionally, the medium from CMA-deficient cells, even before senescence induction, drove paracrine senescence in healthy neighboring fibroblasts, raising SA-beta-gal positivity, p21 levels and nuclear HMGB1 loss to degrees comparable to medium from senescent control cells. The secreted metabolites were similarly skewed toward proinflammatory signaling and known regulators of macrophage function.

The most consequential discovery concerned macrophages, the professional scavengers that clear senescent cells through phagocytosis and efferocytosis. Bone marrow-derived macrophages from aged mice showed significantly reduced CMA activity, and exposure to the secretome of senescent fibroblasts suppressed CMA in young macrophages, with the strongest inhibition from CMA-deficient senescent cells. Macrophages engineered to lack LAMP2A displayed reduced phagocytosis and impaired efferocytosis of dying cells, and co-culture experiments showed senescent fibroblasts persisting in their presence. The mechanism traced to SIRPalpha, a phagocytosis-inhibitory receptor that is normally degraded via CMA after endocytic internalization. Without CMA, SIRPalpha accumulates on the macrophage surface, recruits the phosphatase SHP-1, and amplifies the CD47-SIRPalpha ‘do not eat me’ signal, suppressing the actin remodeling that engulfment requires. Aged macrophages showed the same elevated SIRPalpha and SHP-1 levels. In mice lacking LAMP2A specifically in macrophages, senescent cell burden rose in adipose tissue, liver, lung and brain, and full-thickness wounds healed more slowly, with lingering p21-positive cells despite abundant macrophages at the injury site.

The therapeutic implications were tested with CA77.1, a small-molecule CMA activator. Oral administration to eighteen-month-old mice for five months prevented the age-related rise in senescent cells across adipose tissue, liver and lung, dampened inflammatory and senescence gene expression, reduced fibrosis, and restored the phagocytic activity of aged macrophages to levels comparable to those of young animals. In a bleomycin model of pulmonary fibrosis, early CMA activation preserved body weight, cut lung fibrosis, lowered collagen and p16 expression, and normalized the expansion of lung macrophages, while late treatment was far less effective. Human data lent clinical weight: a transcriptional CMA score declined with age in lung cells and was significantly reduced in patients with idiopathic pulmonary fibrosis, whose isolated lysosomes showed unstable LAMP2 and reduced substrate uptake. Together, the results reframe senescent cell accumulation not as a failure of senescence itself but as a failure of its resolution, and they position CMA upregulation as a dual-action strategy, simultaneously improving the health of senescent cells’ secretions and re-arming the immune cells that must dispose of them.

Subject of Research: How age-related decline of chaperone-mediated autophagy alters senescent cells and impairs macrophage-mediated immune clearance

Article Title: Decline of chaperone-mediated autophagy in aging impairs macrophage clearance of senescent cells

Article References: Sereda, R., Lindenau, K., Diaz, A., Liu, Z., Santiago-Fernández, O., Khawaja, R. R., Cutler, R., Calyeca, J., McCabe, M., Durand, S., Aprahamian, F., Vanegas, N. D. P., Chen, H., Vilicich, F., Chavda, B., Botbol, Y., Kroemer, G., Finkel, T., Gavathiotis, E., … Cuervo, A. M. (2026). Decline of chaperone-mediated autophagy in aging impairs macrophage clearance of senescent cells. Nature Aging. https://doi.org/10.1038/s43587-026-01240-w

Image Credits: AI Generated

DOI: 10.1038/s43587-026-01240-w

Keywords: chaperone-mediated autophagy, cellular senescence, macrophages, aging, LAMP2A, SASP, efferocytosis, proteomics, metabolomics, pulmonary fibrosis, CA77.1, lysosomes

The research centers on chaperone-mediated autophagy, or CMA, a highly selective waste-disposal process that study leader Ana Maria Cuervo, M.D., Ph.D., has spent much of her career elucidating. Unlike other forms of autophagy that engulf bulk cellular debris, CMA relies on specialized chaperone proteins that recognize specific damaged or unnecessary proteins, unfold them, and deliver them across the lysosomal membrane for digestion and recycling. This precision matters enormously for cellular health. Cuervo, a distinguished professor of developmental and molecular biology and of medicine at Einstein and co-director of its Institute for Geroscience, previously discovered that CMA activity declines with age, and her group has since linked that decline to the accumulation of cellular garbage in neurons and other cell types, contributing to Alzheimer’s disease and other neurodegenerative conditions as well as vascular and metabolic disorders such as atherosclerosis and diabetes.

In the new work, Cuervo and her collaborators asked whether declining CMA might also explain one of aging’s most persistent mysteries: why senescent cells, which younger bodies clear away efficiently, linger and multiply in older tissues. Most of the experiments were conducted in mice, but the team also analyzed human lung tissue, finding evidence that the mechanisms they uncovered in animals may operate in people as well. The answer, it turns out, involves a double failure. Aging degrades CMA activity inside the senescent cells themselves, making them more toxic and harder to recognize, while simultaneously weakening the macrophages, the immune cells responsible for engulfing and removing them.

To appreciate the significance of this finding, it helps to understand that zombie cells are not always villains. During wound healing, for example, some injured cells deliberately enter a senescent state and release signaling substances that recruit other cells to the site of injury, orchestrating tissue repair. In a healthy young body, this is a carefully timed operation: the senescent cells do their job, and then macrophages, essential immune cells that fight infection, clear dead cells, and regulate inflammation, arrive to sweep them away. When the researchers genetically engineered mice whose macrophages lacked CMA, they found that these animals accumulated more zombie cells at wound sites and healed more slowly than control mice, demonstrating that CMA within macrophages is essential for clearing senescent cells that would otherwise interfere with tissue repair.

The team then zoomed in on the process of zombification itself. Scientists can create senescent cells in the laboratory by exposing them to drugs that halt cell division, and when the researchers did this with fibroblasts, cells found in connective tissue, taken from young and old mice, they observed striking differences. Fibroblasts from young mice ramped up their CMA activity as they became senescent, apparently as part of a controlled transition. Fibroblasts from old mice, which already had low CMA activity, failed to increase it at all. That failure had cascading consequences. Certain proteins inside the old senescent cells were never properly broken down through CMA, and the cells consequently secreted undigested toxic substances into their surroundings. Those secretions did two kinds of damage: they pushed nearby healthy cells toward senescence themselves, spreading the zombie population, and they interfered with macrophages’ ability to recognize and remove the senescent cells, allowing them to hide in plain sight.

Meanwhile, the macrophages of older animals were fighting their own battle. The researchers found that CMA activity in macrophages from aged mice was significantly lower than in those from young animals, and this deficit reduced the immune cells’ capacity to engulf and eliminate senescent cells. In other words, aging simultaneously makes zombie cells more numerous, more poisonous, and better camouflaged, while disarming the very immune soldiers tasked with destroying them. It is a vicious circle in which each failing reinforces the others, and it helps explain why senescent cell burden rises so steeply in later life and why that burden correlates so strongly with chronic inflammation and age-related disease.

These findings carry an important caution for the booming field of senolytics, drugs designed to kill senescent cells outright. Since scientists first recognized that zombie cells help drive the aging process, pharmaceutical researchers have pursued compounds to eliminate them, but the new results suggest that testing such drugs on cells taken from young animals and artificially made senescent in the laboratory may not predict how they will perform in older bodies, where the biology of senescence is fundamentally altered. The Einstein team points toward a different approach: rather than killing zombie cells directly, use a compound that reactivates CMA, reducing the harmful effects of their secretions while simultaneously restoring macrophages’ ability to recognize and engulf them. In essence, the goal shifts from assassination to rehabilitation, coaxing the aging body’s own clearance machinery back into working order.

Encouragingly, the researchers had already developed such a tool. A small-molecule CMA activator called CA77.1, created in earlier work by the Cuervo laboratory, was put to the test in aged mice. After five months of daily oral treatment, the animals showed reduced buildup of zombie cells in several organs, along with diminished signs of inflammation and fibrosis. In a separate experiment, macrophages isolated from aged mice and treated with CA77.1 recovered their ability to engulf particles to levels comparable to macrophages from young mice, evidence that the drug can functionally rejuvenate aging immune cells rather than merely slowing further decline.

The final step was to ask whether any of this could matter for human disease, and the team chose a particularly grim candidate: idiopathic pulmonary fibrosis, or IPF, a condition in which scar tissue progressively invades the lungs, making breathing increasingly difficult. Median survival after diagnosis is only three to five years, and the disease is tightly connected to cellular senescence. When the researchers examined lung samples from IPF patients, they found markedly reduced CMA activity, raising the possibility that a CMA-boosting drug might help. Testing in a mouse model of the disease delivered promising results: when treatment with CA77.1 was begun early after lung injury, it reduced the severity of lung fibrosis as well as markers of cellular senescence and inflammation.

Our research connects two major drivers of aging, declining CMA and cellular senescence, and shows for the first time how their interaction allows senescent cells to evade clearance by the immune system in old organisms, Cuervo said. We have also found that instead of trying to kill zombie cells, we may be able to restore their interaction with the immune system so that the body can clear them naturally. The next challenge, she noted, is determining whether this approach can eventually be developed into a safe treatment for age-related diseases in people. The paper, first authored by Rebecca Sereda, who conducted the work as a graduate student in Cuervo’s laboratory co-mentored by Susmita Kaushik, drew on researchers from two additional Einstein laboratories and three external collaborating groups at Ohio State University, the Centre de Recherche des Cordeliers and Gustave Roussy, and the University of Pittsburgh. It was supported by the NIH National Institute on Aging, the Hevolution Foundation, the Freedom Together Foundation, and the Rainwaters Foundation, and it grew out of an Einstein Hevolution Partnership Award. Albert Einstein College of Medicine holds a portfolio of intellectual property related to the research and is seeking licensing partners to develop and commercialize the technology. If future studies confirm that boosting CMA is safe and effective in humans, the work could mark a turning point in geroscience, replacing the blunt instrument of cell killing with a subtler strategy: teaching an aging immune system to remember how to clean up after itself.

Subject of Research: Age-related decline in chaperone-mediated autophagy and its role in senescent cell accumulation

Article Title: Einstein researchers discover why “zombie cells” accumulate with age

Article References: Einstein researchers discover why “zombie cells” accumulate with age. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: senescent cells, zombie cells, chaperone-mediated autophagy, aging, macrophages, lung fibrosis, CA77.1, inflammation, geroscience, senolytics, Nature Aging, immunosenescence