Hypothesis

The CDKN2A/B locus functions not as a static aging driver but as a dynamic rheostat whose alternative splice products—p16INK4a and p19ARF—are differentially regulated by cellular metabolic state to balance tumor suppression against tissue regenerative capacity. When nutrient‑sensing pathways (e.g., mTORC1) are chronically active, the locus favors p16INK4a expression, promoting senescence and limiting proliferation. Conversely, intermittent suppression of mTORC1 or elevation of NAD+ shifts splicing toward p19ARF, which attenuates senescence while preserving anti‑oncogenic activity. Thus, age‑related increase in total CDKN2A transcription reflects a stress‑induced bias toward the p16INK4a isoform, not an immutable program.

Mechanistic Basis

1. Splice‑site regulation by metabolic signals – mTORC1‑dependent phosphorylation of SR proteins enhances inclusion of exon 2 (p16INK4a) while skipping exon 1β (p19ARF). NAD+‑activated SIRT1 deacetylates hnRNPA1, favoring exon 1β inclusion. This creates a metabolic switch that modulates isoform ratios without altering promoter activity.
2. Feedback via SASP – Secreted IL‑6 from senescent cells activates JAK/STAT signaling in neighboring progenitors, increasing mTORC1 activity and reinforcing p16INK4a splicing, establishing a bistable loop that can be broken by rapamycin or NAD+ boosters.
3. Isoform‑specific functions – p16INK4a enforces G1 arrest via CDK4/6 inhibition, whereas p19ARF stabilizes p53 independently of CDK inhibition, promoting apoptosis of severely damaged cells while allowing transient cell‑cycle exit for repair.

Testable Predictions

• Prediction 1: In young, mTORC1‑inhibited (rapamycin‑treated) mice, the p19ARF/p16INK4a ratio in liver and muscle will be significantly higher than in aged controls, correlating with improved regenerative markers (e.g., Ki‑67, Sox9) without increased tumorigenesis.
• Prediction 2: CRISPR‑mediated disruption of the mTORC1‑responsive splice enhancer in exon 2 will reduce age‑associated p16INK4a elevation, extend healthspan, and not elevate spontaneous tumor incidence beyond that seen with wild‑type alleles.
• Prediction 3: Exogenous NAD+ supplementation will increase p19ARF expression in human fibroblast cultures exposed to sub‑lethal oxidative stress, decreasing senescence‑associated β‑galactosidase activity while maintaining p53‑dependent DNA‑damage response.

Potential Experiments

• Perform isoform‑specific qPCR and Western blot on tissues from mice treated with rapamycin, NAD+ precursors (NR/NMIT), or vehicle across lifespan; assess tumor histopathology at 24 months.
• Generate a knock‑in mouse line bearing point mutations in the predicted SR‑protein binding site of exon 2; monitor splicing ratios, senescence biomarkers, and cancer onset.
• Treat primary human fibroblasts with nicotinamide riboside and measure changes in splice‑factor phosphorylation (SRPK1, AKT) and isoform ratios via RNA‑seq; functional readouts include colony formation and γ‑H2AX foci.

If these experiments confirm that metabolic cues can shift the CDKN2A/B output toward a less senescent, still tumor‑protective isoform, it would support the idea that aging at this locus is a tunable trade‑off rather than a fixed, selected program, opening therapeutic avenues that negotiate with, rather than override, evolution’s cancer‑suppression logic.