Hypothesis

Rapamycin‑mediated mTORC1 inhibition prolongs life by locking aged cells into a persistent, low‑activity state that mimics an embryonic stress‑response program. This state upregulates protective mechanisms (e.g., autophagy, xenobiotic detoxification, peroxisomal metabolism) that prevent the formation of new damage, yet it lacks the molecular machinery required to remove pre‑existing lesions such as senescent cells, protein aggregates, or mitochondrial dysfunction. Consequently, lifespan extension depends on continuous drug exposure; withdrawing rapamycin allows damage to reaccumulate to pretreatment levels.

Mechanistic rationale

1. Transcriptomic mimicry of fetal stress – Both caloric restriction (CR) and rapamycin suppress mTORC1, but RNA‑seq shows CR enriches pathways linked to inflammation and cholesterol homeostasis, whereas rapamycin preferentially induces xenobiotic metabolism and peroxisome function [1]. This pattern resembles the transcriptome of fetal tissues exposed to intermittent hypoxia, which prioritizes damage tolerance over repair.

2. Blocked damage clearance – Enhanced autophagy and proteasome activity under rapamycin increase turnover of nascent proteins but do not upregulate the specific cargo receptors (e.g., p62/SQSTM1, NBR1) needed for selective removal of ubiquitinated aggregates or senescent‑cell‑associated secretory phenotype (SASP) factors. Existing damage therefore persists, consistent with observations that rapamycin slows damage accrual rather than reversing it [2].

3. Dependence on continuous signaling – The fetal‑like state is maintained only while mTORC1 remains inhibited. Upon rapamycin washout, mTORC1 reactivation drives a shift back to an anabolic, growth‑promoting program that reactivates damage‑sensing pathways (e.g., NF‑κB, ROS production). Without ongoing suppression, the protective transcriptome collapses, and latent lesions become pathogenic again.

Testable predictions

• Prediction 1 (Damage rebound) – In aged mice treated with rapamycin for 6 months, withdrawal will lead to a statistically significant increase in senescent cell burden (p16^INK4a^‑positive cells) and insoluble protein aggregates within 4 weeks, reaching levels comparable to untreated controls. Persistence of low damage after washout would falsify the hypothesis.
• Prediction 2 (Transcriptomic overlap) – Bulk RNA‑seq of rapamycin‑treated aged muscle will show a higher enrichment score for fetal stress signatures (e.g., HIF‑1α targets, glycolytic shift) than for rejuvenation signatures (e.g., Yamanaka factor‑induced pluripotency network) when compared to CR‑treated tissue. A lack of such enrichment would challenge the mechanism.
• Prediction 3 (Additive effect with true repair) – Combining rapamycin with a senolytic that clears existing senescent cells will extend lifespan beyond rapamycin alone only if rapamycin remains present during the senolytic window. If lifespan extension persists after rapamycin cessation in the combined group, the hypothesis is refuted.

Experimental design

• Use male and female C57BL/6J mice aged 20 months. Groups: (a) vehicle, (b) rapamycin (14 ppm diet) continuously, (c) rapamycin withdrawn after 6 months, (d) rapamycin + senolytic (dasatinib + quercetin) administered during the last month of rapamycin treatment, (e) senolytic alone.
• Assess lifespan, frailty index, muscle grip strength, and histological markers (SA‑β‑gal, p62, ubiquitin) at baseline, 6 months, and 12 weeks post‑withdrawal.
• Perform RNA‑seq on gastrocnemius muscle at each timepoint; compute gene set enrichment analysis (GSEA) for fetal stress vs. rejuvenation gene sets.

If rapamycin’s benefits disappear upon cessation and are not accompanied by clearance of existing damage, the data will support the view that mTOR inhibition extends life by impersonating a harder life rather than by genuine rejuvenation.