In APOE4 carriers, the composition of astrocyte-secreted APOE-containing lipid particles becomes enriched in gangliosides (GM1/GM3 ratio elevation) and depleted in phosphatidylinositol (PI). We propose that this altered lipid particle composition reduces neuronal membrane PI(4,5)P2 pools via impaired PI transfer and phospholipid flippase activity. The resulting PI(4,5)P2 deficit simultaneously creates GM1-enriched membrane microdomains that serve as heterogeneous nucleation sites for amyloid-β42 aggregation, while disrupting phosphoinositide-dependent synaptic scaffolding (PSD-95, Homer1) and glutamate receptor trafficking. This dual mechanism explains why APOE4 is the strongest genetic risk factor: it creates a membrane environment that both accelerates amyloidogenesis and impairs synaptic resilience. Testable predictions include: (1) APOE4 astrocytes secrete lipid particles with <40% of normal PI content; (2) neurons exposed to APOE4 lipid particles show reduced synaptic PI(4,5)P2 and enhanced Aβ42 fibril nucleation rates; (3) restoring PI(4,5)P2 via DAGKα overexpression or PI delivery prevents both amyloid nucleation and synaptic deficits in APOE4-targeted models.
Curated pathway from expert analysis
flowchart TD
A["APOE4 Expression<br/>Poorly Lipidated Particles"]
B["Neuronal PI(4,5)P2<br/>Loss"]
C["Ganglioside-Mediated<br/>Amyloid Nucleation"]
D["Phosphoinositide<br/>Synaptic Failure"]
E["APOE4-P2 signaling<br/>as Therapeutic Target"]
F["Phosphoinositide<br/>Restoration"]
A --> B
B --> C
C --> D
D --> E
E --> F
style A fill:#b71c1c,stroke:#ef9a9a,color:#ef9a9a
style F fill:#1b5e20,stroke:#a5d6a7,color:#a5d6a7No linked papers recorded for this hypothesis yet.
Median TPM across 13 brain regions for APOE from GTEx v10.
No clinical trials data linked to this hypothesis yet.
No curated ClinVar variants loaded for this hypothesis.
Run scripts/backfill_clinvar_variants.py to fetch P/LP/VUS variants.
No DepMap CRISPR Chronos data found for APOE.
Run python3 scripts/backfill_hypothesis_depmap.py to populate.
| Prediction | Predicted | Observed | Status | Conf |
|---|---|---|---|---|
| IF DAGKα is overexpressed specifically in neurons of APOE4/4 targeted replacement mice (via AAV9-Synapsin-DAGKα) to restore neuronal PI(4,5)P2 synthesis, THEN amyloid plaque load (measured by Aβ42 ELI | APOE4/4 mice receiving neuronal DAGKα overexpression will show: (1) ≥50% reduction in cortical Aβ42 concentration vs vehicle-treated APOE4/4 controls; (2) PSD-9 | — no observation — | pending | 0.65 |
| IF human iPSC-derived astrocytes carrying APOE4/4 are cultured and their secreted lipid particles are applied to wild-type primary neurons, THEN neuronal membrane PI(4,5)P2 levels will be reduced by ≥ | Neuronal PI(4,5)P2 signal (measured by PI(4,5)P2-specific antibody or Bodipy-PI(4,5)P2 probe) will be significantly lower in APOE4 lipid particle-treated neuron | — no observation — | pending | 0.72 |