{
"ranked_hypotheses": [
{
"title": "Metabolic Inflexibility Precedes Transcriptional Reprogramming (NAD+/SIRT3 Axis)",
"description": "Mitochondrial dysfunction represents the earliest and most fundamental irreversibility checkpoint, preceding and driving transcriptional lock-in through NAD+ depletion and SIRT3 inactivation. This hypothesis offers the highest commercial tractability due to existing NR/NMN safety profiles and Phase I/II trials in metabolic indications.",
"target_gene": "SIRT3/NAD+ salvage pathway, PGC-1α",
"dimension_scores": {
"evidence_strength": 0.68,
"novelty": 0.62,
"feasibility": 0.78,
"therapeutic_potential": 0.72,
"mechanistic_plausibility": 0.70,
"druggability": 0.82,
"safety_profile": 0.85,
"competitive_landscape": 0.75,
"data_availability": 0.65,
"reproducibility": 0.68
},
"composite_score": 0.735,
"evidence_for": [
{"claim": "Metabolic state determines macrophage inflammatory phenotype", "pmid": "27702813"},
{"claim": "Metabolic genes serve as early discriminators in AD trajectory mapping", "pmid": "30643258"},
{"claim": "TREM2-dependent metabolic reprogramming precedes DAM formation", "pmid": "33531068"}
],
"evidence_against": [
{"claim": "Causality vs correlation unresolved - metabolic changes may be consequence rather than driver", "pmid": "30643258"}
]
},
{
"title": "TREM2 Agonism Has Narrow Early-Window at DAM1→DAM2 Transition Checkpoint",
"description": "TREM2 agonism can only revert DAM1 microglia to homeostatic state but cannot rescue DAM2 microglia, which have undergone lipid-droplet accumulation and Apoe-dependent transcriptional rewiring. AL002 (Alector/AbbVie) Phase II trials provide competitive landscape context and first-in-class validation opportunity.",
"target_gene": "TREM2, SYK signaling axis",
"dimension_scores": {
"evidence_strength": 0.60,
"novelty": 0.58,
"feasibility": 0.72,
"therapeutic_potential": 0.78,
"mechanistic_plausibility": 0.65,
"druggability": 0.75,
"safety_profile": 0.62,
"competitive_landscape": 0.68,
"data_availability": 0.70,
"reproducibility": 0.60
},
"composite_score": 0.668,
"evidence_for": [
{"claim": "DAM分期 framework established with single-cell RNA-seq in 5xFAD mice", "pmid": "28678784"},
{"claim": "TREM2 loss-of-function blocks early DAM formation; DAM2 emerges independently", "pmid": "32349763"},
{"claim": "TREM2-dependent metabolic reprogramming precedes irreversible lipid accumulation", "pmid": "33531068"}
],
"evidence_against": [
{"claim": "DAM分期 framework oversimplifies - continuous gradients rather than discrete checkpoints", "pmid": "30770298"},
{"claim": "DAM2 may serve protective functions in amyloid clearance", "pmid": "28257655"},
{"claim": "TREM2-deficient mice show reduced amyloid burden in some contexts", "pmid": "28369781"}
]
},
{
"title": "BBB Integrity Loss Defines Absolute Therapeutic Window Closure",
"description": "BBB disruption beyond 40% permeability increase (Ktrans threshold) marks the point where systemically-delivered microglial reprogramming agents cannot reach effective concentrations. This hypothesis functions primarily as a companion diagnostic and delivery strategy that enables all other approaches.",
"target_gene": "MMP-9, Claudin-5, PDGFRβ (pericyte coverage)",
"dimension_scores": {
"evidence_strength": 0.62,
"novelty": 0.55,
"feasibility": 0.70,
"therapeutic_potential": 0.68,
"mechanistic_plausibility": 0.72,
"druggability": 0.58,
"safety_profile": 0.70,
"competitive_landscape": 0.72,
"data_availability": 0.68,
"reproducibility": 0.65
},
"composite_score": 0.660,
"evidence_for": [
{"claim": "BBB breakdown predicts cognitive decline in APOE4 carriers; pericyte loss is critical", "pmid": "25947343"},
{"claim": "Vascular contributions to neurodegeneration established", "pmid": "30061324"}
],
"evidence_against": [
{"claim": "DCE-MRI standardization across sites requires significant investment", "pmid": "ADNI-3 protocols"}
]
},
{
"title": "APOE4 Creates Accelerated, Compressed Reversibility Window",
"description": "APOE4 drives microglia toward a hyper-inflammatory state that prematurely exhausts the TREM2-TYROBP signaling axis, collapsing the therapeutic window by 40-60%. However, mechanism of compression remains underspecified and the deterministic narrative is challenged by heterogeneous progression rates in APOE4 carriers.",
"target_gene": "APOE/TREM2 axis, APOE-TREM2 physical interaction",
"dimension_scores": {
"evidence_strength": 0.52,
"novelty": 0.60,
"feasibility": 0.55,
"therapeutic_potential": 0.65,
"mechanistic_plausibility": 0.58,
"druggability": 0.52,
"safety_profile": 0.68,
"competitive_landscape": 0.62,
"data_availability": 0.60,
"reproducibility": 0.52
},
"composite_score": 0.584,
"evidence_for": [
{"claim": "APOE4 microglia have blunted TREM2-dependent clustering response", "pmid": "32873780"},
{"claim": "APOE4 triggers accelerated microglial aging signatures and earlier TYROBP activation", "pmid": "33478913"},
{"claim": "APOE4 modifies AD risk through microglial pathways", "pmid": "29100078"}
],
"evidence_against": [
{"claim": "APOE4 carriers show heterogeneous progression rates challenging deterministic model", "pmid": "9032702"},
{"claim": "APOE4 microglia can be functionally normalized ex vivo", "pmid": "29600276"},
{"claim": "Mechanism of compression not mechanistically specified - 40-60% figure is inferred not derived"}
]
},
{
"title": "Epigenetic Reprogramming Required for Late-Stage Interventions (OSKM)",
"description": "Beyond 12 months human equivalent, conventional approaches fail due to irreversible epigenetic changes (DNA methylation of P2ry12 promoter, H3K27ac accumulation at disease-specific enhancers). Partial Yamanaka factor reprogramming could reset the epigenetic clock but carries oncogenic risk requiring careful titration.",
"target_gene": "DNA methylation machinery (DNMTs), H3K27ac modifiers (p300/CBP, HDACs)",
"dimension_scores": {
"evidence_strength": 0.48,
"novelty": 0.75,
"feasibility": 0.42,
"therapeutic_potential": 0.70,
"mechanistic_plausibility": 0.60,
"druggability": 0.45,
"safety_profile": 0.35,
"competitive_landscape": 0.80,
"data_availability": 0.42,
"reproducibility": 0.45
},
"composite_score": 0.542,
"evidence_for": [
{"claim": "Partial epigenetic reprogramming restores visual function in aged glaucoma mice", "pmid": "32050043"},
{"claim": "Epigenomic changes mapped in AD mouse models; enhancer hyperacetylation identified", "pmid": "25504525"}
],
"evidence_against": [
{"claim": "Oncogenic risk of OSKM delivery significant and unresolved"},
{"claim": "Regulatory path for partial reprogramming undefined"},
{"claim": "Delivery to microglia via AAV9 requires demonstration in aged brains"}
]
},
{
"title": "TYROBP Network Hyperactivation Marks Point of No Return",
"description": "Chronic TYROBP activation drives a feedforward loop that locks microglia into neurodegenerative phenotype through sustained Syk and MAPK signaling. Once TYROBP network components exceed 2-fold upregulation, epigenetic silencing becomes irreversible. However, the mechanistic pathway from adaptor protein to irreversible state is unexplained.",
"target_gene": "TYROBP/SYK axis, MAPK/ERK signaling",
"dimension_scores": {
"evidence_strength": 0.42,
"novelty": 0.55,
"feasibility": 0.48,
"therapeutic_potential": 0.58,
"mechanistic_plausibility": 0.45,
"druggability": 0.40,
"safety_profile": 0.42,
"competitive_landscape": 0.55,
"data_availability": 0.38,
"reproducibility": 0.40
},
"composite_score": 0.463,
"evidence_for": [
{"claim": "TYROBP established as central hub in neurodegeneration-associated microglial network", "pmid": "26709853"}
],
"evidence_against": [
{"claim": "TYROBP is adaptor protein lacking enzymatic activity - mechanism to epigenetic silencing unspecified"},
{"claim": "2-fold threshold is arbitrary and biologically undefined"},
{"claim": "TYROBP knockout embryonically lethal in mice", "pmid": "21258326"},
{"claim": "SYK inhibitors tested in AD models with mixed results", "pmid": "33539912"}
]
},
{
"title": "CSF1R Inhibition Reversal Window Depends on Microglia Replacement Kinetics",
"description": "DISQUALIFIED: This hypothesis conflates replacement strategy with reprogramming strategy, representing a fundamental category error. CSF1R antagonism eliminates microglia and relies on precursor repopulation, which is distinct from converting existing disease-associated microglia to homeostatic state. The 30% precursor threshold is arbitrary and unvalidated.",
"target_gene": "CSF1R, nestin+ progenitor pool",
"dimension_scores": {
"evidence_strength": 0.40,
"novelty": 0.35,
"feasibility": 0.38,
"therapeutic_potential": 0.45,
"mechanistic_plausibility": 0.32,
"druggability": 0.50,
"safety_profile": 0.28,
"competitive_landscape": 0.42,
"data_availability": 0.45,
"reproducibility": 0.40
},
"composite_score": 0.395,
"evidence_for": [
{"claim": "Near-complete microglia depletion followed by repopulation rescues spatial memory", "pmid": "31653938"},
{"claim": "Nestin+ progenitors established as microglia source during repopulation", "pmid": "26063358"}
],
"evidence_against": [
{"claim": "CATEGORY ERROR: Replacement ≠ reprogramming - fundamentally different therapeutic paradigms"},
{"claim": "30% precursor threshold is arbitrary logical construct, not empirically derived"},
{"claim": "Microglia repopulation in aged brains often yields disease-associated cells", "pmid": "29429962"},
{"claim": "PLX3397 systemic administration causes substantial immunosuppression in humans"}
]
}
],
"knowledge_edges": [
{"source_id": "H1", "source_type": "hypothesis", "target_id": "TREM2", "target_type": "gene", "relation": "primary_target"},
{"source_id": "H2", "source_type": "hypothesis", "target_id": "APOE", "target_type": "gene", "relation": "primary_target"},
{"source_id": "H3", "source_type": "hypothesis", "target_id": "CSF1R", "target_type": "gene", "relation": "primary_target"},
{"source_id": "H4", "source_type": "hypothesis", "target_id": "TYROBP", "target_type": "gene", "relation": "primary_target"},
{"source_id": "H5", "source_type": "hypothesis", "target_id": "SIRT3", "target_type": "gene", "relation": "primary_target"},
{"source_id": "H5", "source_type": "hypothesis", "target_id": "NAMPT", "target_type": "gene", "relation": "primary_target"},
{"source_id": "H6", "source_type": "hypothesis", "target_id": "DNMT3A", "target_type": "gene", "relation": "primary_target"},
{"source_id": "H7", "source_type": "hypothesis", "target_id": "MMP9", "target_type": "gene", "relation": "primary_target"},
{"source_id": "H1", "source_type": "hypothesis", "target_id": "H4", "target_type": "hypothesis", "relation": "signaling_axis_overlap"},
{"source_id": "H1", "source_type": "hypothesis", "target_id": "H2", "target_type": "hypothesis", "relation": "mechanistic_convergence"},
{"source_id": "H5", "source_type": "hypothesis", "target_id": "H1", "target_type": "hypothesis", "relation": "temporal_precedence"},
{"source_id": "H7", "source_type": "hypothesis", "target_id": "H1", "target_type": "hypothesis", "relation": "delivery_constraint"},
{"source_id": "H7", "source_type": "hypothesis", "target_id": "H2", "target_type": "hypothesis", "relation": "delivery_constraint"},
{"source_id": "H7", "source_type": "hypothesis", "target_id": "H5", "target_type": "hypothesis", "relation": "delivery_constraint"},
{"source_id": "H1", "source_type": "hypothesis", "target_id": "AL002", "target_type": "clinical_program", "relation": "competitive_landscape"},
{"source_id": "H5", "source_type": "hypothesis", "target_id": "NR", "target_type": "clinical_compound", "relation": "repurposing_opportunity"},
{"source_id": "H5", "source_type": "hypothesis", "target_id": "NMN", "target_type": "clinical_compound", "relation": "repurposing_opportunity"},
{"source_id": "H2", "source_type": "hypothesis", "target_id": "APOE4", "target_type": "risk_factor", "relation": "genotype_modifier"},
{"source_id": "H1", "source_type": "hypothesis", "target_id": "DAM1", "target_type": "cellular_state", "relation": "target_cell_state"},
{"source_id": "H1", "source_type": "hypothesis", "target_id": "DAM2", "target_type": "cellular_state", "relation": "irreversible_state"}
],
"synthesis_summary": "The Agora debate converges on a hierarchical model where the therapeutic window for microglial reprogramming is bounded by multiple sequential checkpoints: metabolic inflexibility (H5, composite 0.735) represents the earliest and most tractable intervention point, followed by the TREM2-dependent DAM1→DAM2 transition (H1, 0.668) which is now de-risked by AL002 Phase II trials but requires companion biomarker development for patient stratification. BBB integrity (H7, 0.660) functions as a universal delivery constraint that should guide trial design across all systemically-administered approaches. APOE4 genotype (H2, 0.584) modulates window duration by 40-60% but mechanistic underspecification limits immediate therapeutic targeting. TYROBP network (H4) and OSKM epigenetic (H6) represent high-risk/high-reward long-term strategies requiring substantial mechanistic clarification before development investment. CSF1R targeting (H3) is disqualified as a category error—replacement therapy via precursor repopulation is fundamentally distinct from reprogramming existing disease-associated microglia, and the arbitrary 30% precursor threshold lacks empirical support."
}