Hypothesis

Senescent cells do not merely broadcast damage; they actively impose a reversible proliferation brake on adjacent stromal cells by triggering EZH2‑dependent H3K27me3 accumulation. This brake limits clonal expansion of damaged tissue, and its removal—by senolytics or natural clearance—releases the brake, leading to uncontrolled proliferation, genomic instability, or fibrotic remodeling.

Mechanistic Basis

Recent work shows EZH2 loss initiates senescence entry before global H3K27me3 erosion, providing a temporal window for DNA repair 1. Senescence then drives spatially reorganized H3K27me3 domains that selectively activate SASP while repressing proliferation genes 2. We propose that, in parallel, senescent cells secrete factors (e.g., TGF‑β, Wnt antagonists) that transiently upregulate EZH2 in neighboring fibroblasts and epithelial cells. Elevated EZH2 catalyzes H3K27me3 deposition at promoters of cyclin‑dependent kinases and Myc, imposing a reversible G1 arrest. This creates a bistable epigenetic switch: senescent cells maintain the brake via continuous signaling; once they are cleared, EZH2 activity falls, H3K27me3 marks are diluted through DNA replication, and neighbors resume proliferation.

The SASP composition dictates the strength and duration of the brake. In oncogene‑induced senescence, a SASP rich in IL‑6 and CCL2 favors a strong EZH2 induction in neighbors, whereas replicative senescence yields a milder signal 2. Loss of the brake after senescent‑cell removal could explain why some tissues develop hyperplasia or fibrosis following senolytic treatment, as observed in lung and liver models 3.

Predictions & Tests

1. EZH2 elevation in neighbors – Co‑culture of senescent human fibroblasts (induced by irradiation) with naïve fibroblasts should increase EZH2 protein and H3K27me3 levels in the naïve population within 12‑24 h. Blocking SASP with neutralizing antibodies (anti‑IL‑6, anti‑TGF‑β) should abolish this effect.
2. Reversibility of the brake – Removing senescent cells from the co‑culture (using senolytic navitoclax) should lead to a decline in EZH2/H3K27me3 in neighbors and a concomitant rise in EdU incorporation and γH2AX foci, indicating proliferation coupled with DNA stress.
3. In vivo validation – In p16‑3MR mice, inducible clearance of senescent cells after hepatic CCl4 injury should reduce EZH2 expression and H3K27me3 in peri‑portal hepatocytes, accompanied by increased PCNA staining and collagen deposition compared with mice retaining senescent cells.
4. Falsification – If senescent‑cell clearance does not alter EZH2 levels or H3K27me3 in adjacent tissues, or if proliferation rates remain unchanged, the hypothesis is falsified.

Potential Implications

This reframes senolytics not merely as damage removers but as modulators of a tissue‑wide epigenetic rheostat. Timing or dosing regimens could be adjusted to preserve the proliferative brake during early repair phases while removing senescent cells later to prevent chronic SASP‑driven inflammation. Moreover, combining senolytics with transient EZH2 inhibitors might prevent the post‑clearance proliferative surge that underlies hyperplasia or fibrosis, offering a combinatorial strategy to improve regenerative outcomes.