Hypothesis

In aged cells, autophagic flux is actively blocked not only by canonical signaling inhibitors (mTORC1, Rubicon, EP300, STK4) but also by the sequestration of the ULK1 initiation complex into inhibitory biomolecular condensates formed by age‑upregulated RNA‑binding proteins. This condensate‑mediated trap creates a reversible, mTORC1‑independent layer of suppression that explains why autophagy can be re‑activated only when both signaling and condensate barriers are overcome.

Mechanistic Basis

• Persistent mTORC1 hyperactivation keeps ULK1 phosphorylated at inhibitory sites, preventing complex assembly [1].
• Rubicon accumulation blocks VPS34 lipid kinase activity, hindering phagophore nucleation [2].
• Cytoplasmic EP300 and STK4/MST1 further acetylate or phosphorylate core autophagy proteins, adding layers of inhibition [4][5].
• With age, cytoplasmic levels of low‑complexity RNA‑binding proteins such as TIA1 and G3BP1 rise, promoting liquid‑liquid phase separation (LLPS) that captures ULK1, ATG13 and FIP200 into dynamic granules [7].
• These condensates sequester the ULK1 complex away from phosphatases and upstream activators, rendering mTORC1 inhibition insufficient to restore initiation.
• The condensates are stable under nutrient‑poor conditions, providing a mechanistic basis for the observed persistence of autophagy suppression even when mTORC1 is pharmacologically inhibited.

Testable Predictions

1. Condensate colocalization – In fibroblasts from old mice (or human donors >65 y), endogenous ULK1 will show increased puncta that co‑localize with TIA1/G3BP1, a signal absent in young cells. This can be quantified by immunofluorescence and Pearson’s coefficient.
2. FRAP mobility – Fluorescence recovery after photobleaching of GFP‑ULK1 will be significantly slower in aged cells, indicating reduced diffusion due to condensate binding; treatment with 1,6‑hexanediol or PRMT5 inhibitor (which dissolves LLPS) will restore rapid recovery.
3. Functional rescue – Disrupting condensates (1,6‑hexanediol 0.5 % for 10 min or genetic knock‑down of TIA1) will increase LC3‑II turnover and p62 degradation in aged fibroblasts even when mTORC1 remains active (verified by phospho‑S6K staining). Conversely, forcing condensate formation in young cells (over‑expressing TIA1 low‑complexity domain) will suppress autophagy despite nutrient starvation.
4. In vivo relevance – Administering a BBB‑permeant LLPS modulator (e.g., APTES) to aged mice will improve lysosomal flux in brain and muscle, reducing age‑related phenotypes such as frailty or α‑synuclein accumulation, without altering mTORC1 activity.

Potential Challenges

• Condensate disruptors may have off‑target effects on other RNA‑protein granules; controls using non‑hydrolytic analogs and rescue with condensate‑resistant ULK1 mutants are needed.
• Tissue‑specific differences in Rubicon (e.g., adipose) could modulate condensate formation; experiments should be performed in multiple tissues to map context‑dependency.
• The hypothesis predicts that mTORC1 inhibition alone will not fully rescue autophagy in aged cells if condensates persist; a combined approach (rapamycin + LLPS modulator) should be synergistic, a testable interaction.