Hypothesis

Rapamycin extends lifespan by simultaneously triggering two opposing genomic effects: (1) autophagy‑dependent reduction of mitochondrial ROS that lowers oxidative DNA damage in non‑dividing cells, and (2) chronic mTORC1 inhibition that diminishes de novo dNTP synthesis, increasing replication stress and copy‑number variation (CNV) formation in proliferative compartments. The net longevity outcome depends on the tissue‑specific balance between these protective and mutagenic forces.

Mechanistic Basis

• ROS scavenging arm: Rapamycin activates ULK1‑dependent autophagy, clearing damaged mitochondria and decreasing superoxide production. Lower ROS reduces 8‑oxoguanine lesions, thereby diminishing the substrate for error‑prone repair that can generate CNVs 2. This mirrors caloric restriction’s direct enhancement of base excision repair and mitochondrial DNA stability.
• dNTP depletion arm: mTORC1 stimulates S6K1‑mediated phosphorylation of CAD, the rate‑limiting enzyme for de novo pyrimidine biosynthesis. Inhibition of mTORC1 lowers CAD activity, shrinking dNTP pools 6. In S‑phase, reduced dNTP availability stalls replication forks, elevating fork collapse and mis‑segregation events that produce CNVs 4.
• Tissue specificity: Post‑mitotic tissues (neurons, muscle) rely more on ROS mitigation than on high dNTP fluxes, so the protective arm dominates. In contrast, stem‑cell niches (intestinal crypts, hematopoietic progenitors) experience high replication demand; here the dNTP shortage tip the balance toward increased CNV burden.

Testable Predictions

1. ROS and oxidative damage: Rapamycin‑treated mice will show decreased mitochondrial ROS and 8‑oxoguanine levels in brain and muscle relative to controls, comparable to caloric restriction.
2. dNTP pools: Same tissues will exhibit reduced CAD activity and lower intracellular dNTP concentrations in rapamycin‑treated animals.
3. CNV burden: Single‑cell whole‑genome sequencing will reveal (a) no increase—or a modest decrease—in CNV frequency in post‑mitotic tissues, and (b) a significant rise in CNV burden in intestinal crypts and hematopoietic stem/progenitor cells after chronic rapamycin exposure.
4. Genetic interaction: In a mouse model with heterozygous loss of CAD (reducing dNTP synthesis), rapamycin’s lifespan‑extending effect will be attenuated or abolished, whereas overexpression of CAD will rescue the proliferative‑tissue CNV increase without affecting ROS levels.
5. Autophagy dependence: Knock‑out of autophagy genes (e.g., Atg7) in specific tissues will abolish the ROS‑protective benefit and eliminate any lifespan extension, confirming the autophagy‑mediated arm.

Experimental Design

• Animal cohorts: Wild‑type C57BL/6 mice assigned to ad libitum, rapamycin (14 ppm diet), or 30 % caloric restriction groups; plus tissue‑specific CAD‑overexpression and CAD‑heterozygous lines.
• Readouts:
  • MitoROS measured by MitoSOX fluorescence in fresh tissue homogenates.
  • 8‑oxoguanine quantified via ELISA or LC‑MS/MS.
  • dNTP concentrations assessed by mass‑spectrometry‑based metabolomics.
  • CAD activity assayed in lysates.
  • Autophagy flux monitored by LC3‑II/I ratio and p62 turnover after chloroquine block.
  • CNV burden determined by single‑cell DNA sequencing (10x Genomics) from isolated neurons, cardiomyocytes, intestinal epithelial cells, and Lin⁻Sca1⁺cKit⁺ hematopoietic stem cells after 6 months and 12 months of treatment.
  • Survival curves and frailty indices recorded.

Potential Outcomes and Falsifiability

• Support: Observing reduced ROS/8‑oxoguanine and unchanged or decreased CNVs in non‑dividing tissues, alongside lowered dNTP pools and elevated CNVs in proliferative tissues, would validate the dual‑mechanism model.
• Refutation: If rapamycin fails to lower ROS or 8‑oxoguanine in brain/muscle, or if dNTP levels remain unchanged, the protective arm is unsupported. Conversely, if CNV burden does not rise in stem‑cell compartments despite demonstrable dNTP depletion, the replication‑stress link is challenged. Lack of lifespan extension in CAD‑heterozygous mice despite intact autophagy would further falsify the hypothesis.

This framework moves beyond the "impersonation" concept by specifying how a single pharmacological intervention can simultaneously guard against and provoke distinct forms of genome instability, offering a clear, falsifiable roadmap for future work.