Hypothesis

Accumulated p16INK4a in aged cells does not merely correlate with declining autophagy; it actively suppresses autophagic flux by serving as a scaffold for epigenetic repressors that silence key autophagy genes. This creates a feed‑forward circuit where impaired autophagy permits p16INK4a buildup, and p16INK4a‑mediated chromatin repression locks autophagy in a low‑activity state.

Mechanistic Model

1. p16INK4a as a chromatin adaptor – p16INK4a contains a nuclear localization signal and interacts with histone deacetylases (HDAC1/2) and Polycomb Repressive Complex 2 (PRC2) components (EZH2, SUZ12) as shown in cancer contexts (see its role in transcriptional repression [https://pmc.ncbi.nlm.nih.gov/articles/PMC6743916/]). In senescent fibroblasts, we propose that p16INK4a recruits these complexes to promoters of autophagy‑essential genes such as Becn1, Map1lc3b, and Atg5.
2. Establishment of a repressive chromatin state – HDAC activity removes acetyl groups, while PRC2 deposits H3K27me3, leading to compacted nucleosomes and reduced transcription factor binding (e.g., FOXO3, TFEB). This epigenetic silencing persists even if upstream signals (AMPK activation, mTOR inhibition) are restored.
3. Feedback amplification – Reduced autophagic degradation of p16INK4a (via p62‑lysosomal pathway [https://pmc.ncbi.nlm.nih.gov/articles/PMC7370706/]) allows its nuclear concentration to rise, reinforcing repression. Concurrently, mTORC1 hyperactivity in senescent cells [https://doi.org/10.1016/j.exger.2014.11.004] further blocks autophagy initiation, deepening the suppression.
4. Escape from quiescence – In muscle stem cells, loss of repression at the INK4a/ARF locus drives senescence [https://doi.org/10.4161/15384101.2014.965072]; our model adds that the ensuing p16INK4a surge actively locks down autophagy, preventing the metabolic reset needed for stem cell re‑activation.

Testable Predictions

• Prediction 1: In aged mouse tissues, nuclear p16INK4a occupancy at autophagy gene promoters will be positively correlated with H3K27me3 enrichment and negatively correlated with RNA Pol II occupancy.
• Prediction 2: Genetic knock‑down of p16INK4a (or expression of a p16INK4a mutant lacking the HDAC/PRC2‑binding domain) will increase histone acetylation (H3K27ac) at Becn1/Map1lc3b promoters and restore autophagic flux measured by LC3‑II turnover and p62 degradation, independent of mTOR activity.
• Prediction 3: Pharmacological inhibition of HDACs (e.g., with vorinostat) or EZH2 (e.g., with GSK126) will partially rescue autophagy in p16INK4a‑high senescent cells, even without clearing p16INK4a protein.
• Prediction 4: Senolytic treatment that eliminates p16INK4a‑positive cells will increase autophagic flux in the remaining tissue, and this increase will be attenuated if autophagy genes are simultaneously silenced by CRISPRi.

Potential Experiments

• ChIP‑seq / CUT&RUN for p16INK4a, HDAC1, EZH2, H3K27me3, and H3K27ac in young vs. old mouse liver and muscle, focusing on autophagy gene loci.
• Autophagy flux assays (mCherry‑GFP‑LC3, bafilomycin A1 chase) in primary senescent fibroblasts treated with siRNA against p16INK4a or expressing a ΔHDAC‑binding p16INK4a mutant.
• Rescue studies using HDAC/EZH2 inhibitors in aged mice, measuring LC3‑II/p62 dynamics and functional outcomes (e.g., grip strength, hepatic glucose tolerance).
• Senolytic + autophagy read‑out (e.g., navitoclax followed by Lysotracker staining) to test whether p16INK4a‑cell clearance alone restores flux, and whether combining senolytics with epigenetic drugs yields synergistic effects.

If these experiments show that p16INK4a directly imposes a repressive chromatin program on autophagy genes, the hypothesis of active suppression will be validated. Conversely, if p16INK4a loss or chromatin drug treatment fails to restore autophagy despite reducing p16INK4a levels, the model would be falsified, favoring a passive decline or alternative regulatory mechanisms.