Hypothesis

The neuroprotective efficacy of crocin/crocetin in the aged brain is limited not by BBB permeability per se, but by age‑related decline in hepatic and intestinal esterase activity that reduces the conversion of crocin to crocetin, thereby lowering systemic crocetin availability. Once formed, crocetin directly binds the ferrous iron (Fe²⁺) in the active site of prolyl‑hydroxylase domain protein 2 (PHD2), inhibiting its hydroxylase activity and stabilizing HIF‑1α under mild hypoxic conditions. This HIF‑1α stabilization drives VEGF‑mediated BBB reinforcement and mitochondrial adaptations that underlie crocetin’s observed neuroprotection. In young animals, sufficient esterase activity yields adequate crocetin to engage this pathway; in aged animals, reduced esterase activity diminishes crocetin formation, uncoupling the HIF‑1α‑mediated protective cascade despite unchanged BBB permeability.

Novel Mechanistic Insight

Crocetin’s conjugated polyene system presents a planar, electron‑rich scaffold capable of chelating transition metals. Structural modeling shows that the central C‑C double bond array can coordinate the Fe²⁺ ion of PHD2 in a bidentate fashion, displacing the essential 2‑oxoglutarate co‑substrate and mimicking known iron‑chelating HIF stabilizers (e.g., dimethyloxalylglycine). This direct enzyme inhibition would occur independently of cellular oxygen tension, providing a mechanistic link between crocetin’s mitochondrial oxygen‑enhancement effects and HIF‑1α pathway activation that has not been previously considered.

Testable Predictions

1. Esterase dependence – In young vs. aged mice, oral crocin (50 mg/kg) will produce significantly lower plasma and brain crocetin Cmax and AUC in aged cohorts. Co‑administration of a broad‑spectrum esterase inhibitor (e.g., BNPP) in young mice will recapitulate the aged PK profile, while esterase gene overexpression (CES1/CES2) in aged mice will rescue crocetin exposure.
2. PHD2 inhibition – Recombinant PHD2 activity assays will show dose‑dependent inhibition by crocetin (IC₅₀ ≈ 5‑10 µM) that is reversible by excess Fe²⁺ or 2‑oxoglutarate, confirming direct metal chelation. Cellular HIF‑1α stabilization (measured by Western blot and HRE‑luciferase reporter) will occur in neuronal and astrocytic lines treated with crocetin under normoxia, an effect abolished by PHD2 knockdown.
3. HIF‑1α‑mediated BBB protection – In aged mice subjected to chronic hypoperfusion, crocetin treatment (delivered via γ‑cyclodextrin to bypass absorption limits) will increase brain VEGF and tight‑junction protein (claudin‑5, occludin) expression only when esterase activity is pharmacologically enhanced or when crocetin is administered directly (bypassing the hydrolysis step). HIF‑1α antagonism (using echinomycin) will abolish these BBB‑protective effects.
4. Behavioral correlation – Improved spatial memory in the Morris water maze will correlate with brain crocetin levels and HIF‑1α target expression across age groups, but not with safranal levels, isolating the crocetin‑specific pathway.

Experimental Approach

• PK: LC‑MS/MS quantification of crocetin in plasma and brain at multiple time points post‑oral crocin (±γ‑cyclodextrin) in young (3 mo) and aged (18 mo) mice, with and without esterase modulator.
• Enzyme assay: Purified PHD2 activity measured via hydroxylation of a HIF‑1α peptide substrate; determine crocetin IC₅₀ and kinetic mode (competitive vs. non‑competitive) via Lineweaver‑Burk plots.
• Cellular assays: SH‑SY5Y neuroblastoma and primary astrocytes treated with crocetin (0‑20 µM); assess HIF‑1α accumulation, VEGF secretion, and ROS/mitochondrial membrane potential.
• In vivo validation: Aged mice receive crocetin (±esterase activator) or vehicle; evaluate BBB integrity (Evans blue extravasation), VEGF/tight‑junction protein levels, and cognitive performance.

Falsifiability

If crocetin does not inhibit PHD2 directly (no measurable IC₅₀, no Fe²⁺‑reversible effect), or if brain crocetin levels and HIF‑1α activation remain unchanged despite manipulated esterase activity, the hypothesis would be refuted. Conversely, confirming direct PHD2 chelation coupled with age‑dependent esterase control of crocetin exposure would substantiate the proposed mechanism.