Senescent Cells as Active Metabolic Reprogrammers of Tissue Homeostasis — Epoche C2
The claim, and what would settle it A cell that has undergone an essentially irreversible exit from the division cycle in response to damage — senescent, in the standard usage — is conventionally counted as a marker of tissue age, and the claim examined here is the stronger one: that such a cell alters the metabolism of the cells around it, in a specific direction, by exporting metabolites rather than merely by occupying space or leaking cytokines. The distinction is not rhetorical. If senescent cells are markers, the therapeutic target is whatever produced them; if they are active metabolic sources, the target may be a transporter or an enzyme in the senescent cell, and the cell itself need not be killed. What follows sets out what is established, what is inferred, what the quantitative structure of the inference is, and where the compressed version of this argument in the literature — including one attribution that is simply wrong, corrected below — outruns its evidence. The passive reading has a definite shape worth stating so that it can be tested rather than dismissed. It holds that cellular stress produces senescence, that senescent cells then accumulate because the immune clearance that normally removes them declines, and that tissue dysfunction follows from the accumulated burden — a chain in which the senescent cell contributes nothing beyond its own absence from the proliferative pool and a background of secreted material. The senescence-associated secretory phenotype (SASP), the programme of cytokine, chemokine, protease and growth-factor secretion that most senescent cells adopt, is on this reading a side effect of the arrested state rather than an instrument. The reading has one substantial virtue: senescence is genuinely beneficial in early life, arresting cells at risk of transformation and organising wound closure, so that its late-life costs look like the ordinary shape of antagonistic pleiotropy, a programme selected for its early benefit and unopposed in its late harm. Campisi and d'Adda di Fagagna set out this framing in 2007, and nothing below overturns it; what is at issue is whether the late harm is passive. What the clearance experiments established, and what they left open The strongest evidence against pure passivity is genetic ablation. Baker and colleagues in 2016 used the INK-ATTAC transgene, in which a fragment of the p16 Ink4a promoter drives an inducible caspase-8 construct so that administering a dimerising drug kills cells expressing p16 — the cyclin-dependent kinase inhibitor whose expression is the most widely used in vivo senescence marker. In naturally aged mice, not progeroid models, intermittent clearance begun in adulthood extended median lifespan in both genetic backgrounds tested and attenuated age-related deterioration in kidney, heart and adipose tissue. The logic is a straightforward intervention argument: removing the putative cause removed the effect while nothing else was changed, which no purely correlative accumulation model predicts. But note carefully what this does not establish, because the gap is where the present question lives. Killing a cell removes everything about it at once — its secretome, its metabolic activity, its physical occupancy, its consumption of shared nutrients. A clearance experiment therefore shows that senescent cells cause dysfunction; it says nothing about which of their properties does the causing. To get further one must either abolish a single property and leave the cell in place, or supply the property without the cell. That is the standard of evidence against which the metabolic-reprogramming claim should be judged, and it is a standard the field has met in fragments rather than as a whole. The metabolic phenotype of a senescent cell — and a correction Here the compressed literature, and the earlier version of this essay, made an error that needs stating rather than quietly repairing. The metabolic rewiring of senescent cells has been attributed to Basisty and colleagues (2020). That paper is real and important, but it is a quantitative mass-spectrometric proteomic atlas of the secretome — the proteins released into conditioned medium by human lung fibroblasts and renal epithelial cells driven into senescence by X-irradiation, by doxorubicin, or by oncogenic RAS. Its principal finding is that SASP composition is not a single programme: it depends strongly on both the cell type and the inducer, with only a modest shared core, and a subset of the identified proteins tracks with age in human plasma. It reports proteins, not fluxes; it contains no measurement of glycolytic rate, of oxygen consumption, or of fatty acid oxidation. It cannot support a claim about central carbon metabolism, and citing it for one is the kind of error that survives every check except reading the paper. The evidence for metabolic rewiring exists, but it comes from elsewhere, and — this is the substantive point — it does not describe a single direction of change. Three results, each measuring something different, illustrate the spread: Elevated glucose throughput in therapy-induced senescence. Dörr and colleagues (2013) studied lymphoma cells driven into senescence by chemotherapy and found markedly increased glucose utilisation, coupled to the energetic demands of coping with proteotoxic stress: senescent cells synthesise and degrade protein at high rates and must pay for it. The functional test is what makes the finding load-bearing rather than descriptive — blocking glucose utilisation, or blocking autophagy, was selectively lethal to the senescent cells and not to their proliferating counterparts. A dependency demonstrated by synthetic lethality is a dependency, not a correlation. Elevated pyruvate oxidation in oncogene-induced senescence. Kaplon and colleagues (2013) found the opposite sign in a different setting. In human cells driven into senescence by BRAF V600E , pyruvate dehydrogenase — the gatekeeper committing pyruvate to