Hypothesis: AMPK-Mediated Phosphorylation of Beclin-1 at Ser15 Tunes BCL-2 Interaction in a Tissue-Specific Manner

Core Idea

We propose that AMPK directly phosphorylates Beclin-1 at Ser15, a site distinct from the Thr108 Mst1 site, to weaken the BCL-2/Beclin-1 interaction preferentially in cardiomyocytes under metabolic stress, thereby promoting protective autophagy without triggering apoptosis. In neurons, the same phosphorylation is counteracted by phosphatase PP2A, preserving BCL-2 binding and limiting autophagy to avoid excitotoxic cell death.

Mechanistic Rationale

• AMPK activation occurs during energy deprivation and is known to phosphorylate many autophagy regulators (e.g., ULK1).
• Structural modeling shows Ser15 lies within the BH3-like domain of Beclin-1, adjacent to the hydrophobic groove that engages BCL-2; phosphorylation introduces a negative charge that disrupts this interface.
• In heart tissue, AMPK activity rises with age-related metabolic decline, predicting increased Ser15‑P and reduced BCL-2 binding.
• In brain, constitutive PP2A activity dephosphorylates Ser15, maintaining BCL-2 suppression of autophagy despite AMPK activation.

Testable Predictions

1. Phospho‑specific detection – We'll generate a p‑Ser15-Beclin-1 antibody; Western blot of aged mouse heart and brain will show elevated signal in heart but not brain.
2. Genetic manipulation – Express a non‑phosphorylatable Beclin-1 S15A mutant in cardiomyocytes; this should blunt autophagy flux (measured by LC3‑II turnover) and increase apoptosis (caspase‑3 activation) under starvation, whereas the phosphomimetic S15D will enhance autophagy and reduce cell death.
3. Tissue‑specific phosphatase inhibition – Treat neuronal cultures with a PP2A inhibitor (e.g., okadaic acid low dose) and observe increased p‑Ser15-Beclin-1, reduced BCL-2 co‑immunoprecipitation, and heightened autophagy without apoptosis.
4. In vivo longevity – Cross Becn1^S15D knock‑in mice with a cardiac‑specific driver; expect extended median lifespan compared with wild‑type, while neuronal‑specific S15D shows no lifespan benefit or possible deleterious effects due to excessive autophagy.

Novel Insight Beyond Existing Data

Current models focus on Thr108 phosphorylation (Mst1) and BCL‑2 phosphorylation (JNK1) as the main switches. By positioning Ser15 as an AMPK‑responsive rheostat that directly alters the BH3‑like domain’s affinity for BCL-2, we link cellular energy sensing to the autophagy‑apoptosis decision node in a tissue‑dependent way. This explains why systemic BH3 mimetics can have divergent outcomes: they may unleash autophagy in tissues where AMPK‑Ser15‑P already weakens BCL-2 binding (heart) but cause apoptosis where the interaction remains strong (brain).

Experimental Approach (brief)

• Generate phospho‑specific antibody (pSer15-Beclin-1).
• Perform co‑IP of BCL-2 and Beclin-1 from heart/brain lysates of young vs old mice, with/without AMPK activator (AICAR).
• Measure autophagic flux (mCherry‑GFP‑LC3) and apoptosis (Annexin V/PI) in primary cardiomyocytes and neurons expressing WT, S15A, S15D Beclin-1.
• Assess lifespan and healthspan in tissue‑specific knock‑in mice.

If predictions hold, the hypothesis will redefine the BCL-2/Beclin-1 axis as a nutrient‑sensing switch whose output is sculpted by tissue‑specific phosphatase activity, offering a route to selectively boost autophagy in ageing heart while sparing brain.