Hypothesis

**Crocetin acutely re-programs the autophagy cargo hierarchy in aged neurons by transiently lowering HIF-1α activity, which decreases BNIP3/NIX-mediated mitophagy and concurrently lifts the repression on ER-phagy (FAM134B) and aggrephagy (p62/SQSTM1) pathways. This shift protects neurons by clearing dysfunctional ER and protein aggregates before mitochondrial turnover becomes energetically costly. Chronic crocetin dosing, however, induces P-glycoprotein (P-gp) expression at the blood-brain barrier, reducing brain crocetin levels and reversing the HIF-1α suppression; the autophagy hierarchy reverts to a mitophagy-dominant state, contributing to the loss of long-term neuroprotection. Formulating crocetin with γ-cyclodextrin should bypass P-gp efflux, sustaining the HIF-1α-mediated hierarchy shift and preserving cognitive benefit in aged mice.

Testable Predictions

1. Acute treatment (single intraperitoneal dose, 10 mg/kg) in 24-month-old mice will:

   • Decrease hippocampal HIF-1α protein and its transcriptional targets (VEGF, GLUT1, BNIP3, NIX) by ~30% measured by Western blot/qPCR.
   • Reduce mitophagy flux (mito-Keima red/green ratio) by ~25% without altering bulk autophagy (LC3-II turnover with bafilomycin A1).
   • Increase ER-phagy flux (FAM134B-GFP colocalization with LC3) and aggrephagy flux (p62-Selenium-GFP clearance) each by ~20-35%.
   • Improve performance in the Y-maze spontaneous alternation test (% alternation ↑15%).

2. Chronic treatment (daily oral gavage, 10 mg/kg for 4 weeks) will:

   • Elevate brain endothelial P-gp (MDR1) expression 20-25%.
   • Lower crocetin brain concentration (LC-MS/MS) from ~31 nM to <15 nM despite dosing.
   • Restore HIF-1α and BNIP3/NIX levels to baseline, mitophagy flux returns to pretreatment, while ER-phagy/aggrephagy fluxes decline.
   • Cognitive advantage seen after acute dosing disappears (Y-maze performance returns to aged control).

3. γ-Cyclodextrin-encapsulated crocetin administered chronically will:

   • Maintain brain crocetin >30 nM, prevent P-gp up-regulation (no significant change vs. vehicle).
   • Sustain the acute hierarchy shift (low mitophagy, high ER-phagy/aggrephagy) throughout the 4-week period.
   • Produce persistent improvement in spatial memory (Morris water maze escape latency ↓20% vs. aged control).

Falsification Criteria

• If acute crocetin does not suppress HIF-1α targets or alter the relative flux of mitophagy versus ER-phagy/aggrephagy (differences <5% and not statistically significant), the hierarchy-reprogramming claim is false.
• If chronic crocetin fails to increase P-gp expression or decrease brain crocetin levels, the proposed self-limiting efflux mechanism is invalid.
• If γ-cyclodextrin formulation does not preserve brain crocetin concentrations or prevent the chronic reversal of the autophagy hierarchy, the formulation’s role in overcoming P-gp efflux is unsupported.

Experimental Approach (brief)

• Use mitochondrial (mito-Keima), ER-phagy (FAM134B-mCherry-GFP), and aggrephagy (p62-Selenium-GFP) reporters delivered via AAV to hippocampal neurons of aged mice.
• Measure LC3-II turnover with and without bafilomycin A1 to distinguish bulk vs. selective autophagy.
• Quantify HIF-1α, BNIP3, NIX, FAM134B, p62, and MDR1 by Western blot and immunostaining.
• Determine crocetin and γ-cyclodextrin-crocetin brain levels by LC-MS/MS.
• Behavioral assays: Y-maze, Morris water maze, novel object recognition.

By linking HIF-1α-driven selective autophagy to a pharmacodynamic feedback loop via P-gp, this hypothesis explains why acute saffron-derived neuroprotection wanes with long-term use and offers a concrete strategy—efflux-avoidant formulation—to lock the autophagy hierarchy in a neuroprotective state.