Hypothesis

With advancing age, persistent JNK1‑mediated phosphorylation of BCL‑2 at its unstructured loop shifts the BCL‑2/Beclin‑1 interaction from a simple inhibitory brake to a selective scaffold that preferentially recruits certain autophagy receptors (e.g., p62/SQSTM1 and NBR1) while excluding others (e.g., NDP52 and OPTN). This altered Beclin‑1 interactome re‑orders the hierarchical substrate selection process, sparing damaged mitochondria and photoreceptor outer‑segment proteins but promoting the clearance of soluble protein aggregates. The resulting imbalance leads to mitochondrial dysfunction, incomplete mitophagy, and the accumulation of lipofuscin‑forming retinal pigment epithelium (RPE) debris, thereby linking molecular rheostat dysfunction to age‑related retinal pathology.

Mechanistic Rationale

1. BCL‑2 phosphorylation creates distinct binding surfaces – Phosphorylation of BCL‑2’s loop reduces its affinity for Beclin‑1’s BH3 domain but simultaneously exposes a hydrophobic groove that can bind LC3‑interacting region (LIR) motifs of specific autophagy receptors. This bifunctional behavior converts BCL‑2 from a passive inhibitor into an active cargo‑sorting platform.
2. Receptor competition is skewed – In young cells, balanced BCL‑2/Beclin‑1 binding permits stochastic receptor access, allowing p62, NBR1, OPTN, and NDP52 to compete for ubiquitinated cargos. Age‑related hyper‑phosphorylation favors high‑affinity interaction with p62/NBR1 (which possess acidic LIRs) while sterically hindering OPTN/NDP52 (which require basic LIRs). Consequently, ubiquitinated protein aggregates are efficiently engulfed, whereas damaged mitochondria (relied on NDP52/OPTN) and photoreceptor-derived lipids (handled by LC3‑II‑dependent phagocytosis) are neglected.
3. Lipofuscin formation follows selective clearance – Undigested photoreceptor outer‑segment lipids and protein‑lipid complexes accumulate in lysosomes, where oxidative cross‑linking generates lipofuscin granules. Simultaneous clearance of soluble aggregates reduces proteotoxic stress but fails to mitigate lipid‑derived damage, reproducing the observed dissociation between macroautophagic flux and lipofuscin buildup.

Testable Predictions

• Prediction 1: In retinal extracts from aged mice, immunoprecipitated BCL‑2 will show increased phospho‑specific signal (p‑Ser/Thr) and co‑precipitate higher levels of p62/NBR1 but reduced OPTN/NDP52 compared with young tissue.
• Prediction 2: Expressing a phospho‑deficient BCL‑2 mutant (e.g., S70A/S87A) in aged RPE will restore OPTN/NDP52 binding to Beclin‑1, rescue mitophagy (measured by mt‑Keima), and decrease lipofuscin accumulation without altering bulk LC3‑II turnover.
• Prediction 3: Pharmacological inhibition of JNK1 (using SP600125) in aged retina will diminish BCL‑2 phosphorylation, normalize receptor competition, and reduce both mitochondrial damage and lipofuscin granules.

Experimental Approach

1. Proximity ligation assay (PLA) on retinal sections to quantify BCL‑2–Beclin‑1, BCL‑2–p62/NBR1, and BCL‑2–OPTN/NDP52 interactions across age groups.
2. Mito‑Keima flux assay in primary RPE cultures transfected with wild‑type or phospho‑mutant BCL‑2 to assess mitophagy efficiency.
3. Lipofuscin autofluorescence quantification and electron microscopy to correlate receptor selectivity with lipid‑laden lysosome formation.
4. Rescue experiments using JNK1 inhibitor or Bcl‑2 BH3 mimetic (ABT‑737) to test whether modulating BCL‑2 phosphorylation state rebalances substrate selection.

Falsifiability

If aged retinal tissue shows no alteration in BCL‑2 phosphorylation status, or if phospho‑deficient BCL‑2 fails to restore OPTN/NDP52 binding and mitophagy while lipofuscin remains unchanged, the hypothesis would be refuted. Conversely, confirmation of the predicted shifts in receptor selectivity and functional rescue would support the notion that age‑dependent BCL‑2 reprogramming—not merely autophagy inhibition—drives selective autophagic failure in retinal aging.