How do different organelle-specific autophagy pathways coordinate during neurodegeneration?

Therapeutic Hypotheses: Coordination of Organelle-Specific Autophagy in Neurodegeneration

Hypothesis 1: Mitochondrial-ER Contact Sites as Coordination Hubs

Title: MFN2-PACS2 axis as a "mitophagy-ER-phagy sync switch" via MAM reorganization

Mechanism: MFN2 anchors mitochondria to ER at MAMs; upon mitochondrial stress, MFN2 mediates contact site remodeling that simultaneously positions mitophagy receptors (e.g., NDP52) near ER-sourced membranes while PACS2-regulated ER calcium microdomains trigger both organelle-specific autophagosome nucleation. Disrupting this axis collapses coordinated quality control.

Target gene/protein/pathway: MFN2 (mitochondrial fusion) + PACS2 (ER phosphoregulation); crosstalk via calcium/PI(4,5)P2 signaling at MAMs.

Supporting evidence with PMIDs:

• MFN2 physically interacts with LC3 via LIR motif; MFN2 knockdown impairs mitophagy (PMID: 31171695)
• PACS2 regulates ER-mitochondria tethering and calcium homeostasis (PMID: 25437556)
• MAM integrity is compromised in ALS/PD patient neurons (PMID: 31641032)
• ER contribute membranes to autophagosomes via WIPI2/PI3KC3 during selective autophagy (PMID: 25648100)

Confidence: 0.72

Hypothesis 2: TFEB/TFE3 Parallel Activation as Master Coordinator

Title: Simultaneous TFEB+TFE3 activation drives coordinated organelle clearance via CLEAR-box divergence

Mechanism: TFEB/TFE3 translocate to nucleus under mTORC1 inhibition or AMPK activation, binding CLEAR sequences in promoters of both shared (BECN1, GABARAP) and organelle-specific (PRKN/parkin, RETREG1/ FAM134B) genes. TFE3 preferentially drives ER-phagy gene programs; TFEB drives mitophagy genes, but both heterodimerize to co-regulate lysosomal biogenesis, creating a feedforward loop that simultaneously clears multiple damaged organelles.

Target gene/protein/pathway: TFEB/TFE3 nuclear translocation; upstream: mTORC1 (MTOR), LKB1/STK11-AMPK axis; downstream CLEAR network.

Supporting evidence with PMIDs:

• TFEB/TFE3 double KO in neurons causes severe neurodegeneration (PMID: 31801954)
• TFEB overexpression rescues mitochondrial and ER stress in PD models (PMID: 29311652)
• TFE3 drives reticulophagy via ER stress response (PMID: 29045917)
• CLEAR network encompasses >400 autophagy-lysosome genes (PMID: 26942069)
• TFE3 can compensate for TFEB loss in some contexts (PMID: 31501761)

Confidence: 0.78

Hypothesis 3: TBK1-OPTN-NDP52 Phospho-Cascade as Organelle-Spanning Autophagy Hub

Title: TBK1-mediated phosphorylation of multiple cargo receptors coordinates organelle turnover

Mechanism: TBK1 phosphorylates OPTN (Ser177) and NDP52 (Ser67) on ubiquitin-binding domains, enhancing affinity for ubiquitin-coated damaged organelles. OPTN primarily targets mitochondria; NDP52 can engage both mitochondria and Salmonella, but also ER-derived vesicles. TBK1 thus "broadcasts" autophagy investment to multiple organelles simultaneously; familial ALS mutations in TBK1 (loss-of-function) impair this multi-organelle response.

Target gene/protein/pathway: TBK1 (kinase); OPTN (cargo receptor); NDP52/CALCOCO2 (cargo receptor).

Supporting evidence with PMIDs:

• TBK1 phosphorylates OPTN Ser177, enhancing mitophagy (PMID: 24592263)
• NDP52 recruits autophagy machinery to damaged mitochondria independently of parkin (PMID: 25985789)
• TBK1 mutations cause ALS with impaired mitophagy (PMID: 24951150)
• OPTN also mediates ER-phagy under starvation (PMID: 32048902)
• TBK1 activity required for general selective autophagy (PMID: 25556504)

Confidence: 0.81

Hypothesis 4: p62 Phase Separation as Organelle-Agnostic Sequestration Platform

Title: p62 liquid-liquid phase separation nucleates cross-organelle protein aggregates for coordinated autophagy

Mechanism: p62 undergoes liquid-liquid phase separation (LLPS) upon phosphorylation (Ser403) and ubiquitination of bound cargo. p62 droplets can concentrate ubiquitinated proteins from multiple organelles (damaged mitochondria, ER fragments, protein aggregates) into a single autophagosomal capture event. This "mixed garbage collection" allows one phagophore to engulf multi-organelle cargo. Phosphorylated p62 also activates NRF2, providing transcriptional feedback.

Target gene/protein/pathway: SQSTM1/p62 (scaffold); ULK1/FIP200 (phosphorylation at Ser403); Keap1 (NRF2 pathway).

Supporting evidence with PMIDs:

• p62 LLPS required for selective autophagy (PMID: 31439799, 31801953)
• p62 phosphorylated at Ser403 by casein kinase 2/TBK1 enhances aggregate clearance (PMID: 23842799)
• p62 body formation captures both mitochondria and ER in neuroprotection (PMID: 31506447)
• Keap1-p62 axis links autophagy to NRF2 antioxidant response (PMID: 27459026)
• p62 deletion causes mitochondrial and ER dysfunction in mice (PMID: 30626971)

Confidence: 0.75

Hypothesis 5: VPS34 Complex I Composition as Decision Point for Selective vs. Non-Selective Autophagy

Title: PIK3C3/VPS34 complex I subunit heterogeneity dictates organelle-specific vs. bulk autophagy

Mechanism: VPS34 forms complex I (with ATG14L) for omegasome/ER recruitment and phagophore initiation, but different regulatory subunits (UVRAG, BIF1, PLEKHM1) direct specificity. UVRAG-containing complexes can be recruited to damaged mitochondria or ER via interactions with cargo receptors (e.g., NRBF2). Adjusting complex composition or recruiting specific regulatory subunits could switch between mitophagy, ER-phagy, and general autophagy based on cellular need.

Target gene/protein/pathway: PIK3C3/VPS34, ATG14L, UVRAG, NRBF2 (complex I regulators); PI3P effector proteins (WIPI2, DFCP1).

Supporting evidence with PMIDs:

• NRBF2 recruits VPS34 to mitochondria-ER contact sites (PMID: 27840058)
• UVRAG mutations impair autophagy and cause neurodegeneration (PMID: 25985789)
• ATG14L required for ER-implicated autophagosome biogenesis (PMID: 19050071)
• PI3P at ER initiates both general and selective autophagy (PMID: 25648100)
• PLEKHM1 links VPS34 to RAB7 on late endosomes (PMID: 23009744)

Confidence: 0.65

Hypothesis 6: Calcium Microdomain Crosstalk Between ER and Mitochondria

Title: ER-localized IP3R1-mitochondria calcium flux synchronizes mitophagy and ER-phagy initiation

Mechanism: ER calcium release via IP3R1 at MAMs creates local calcium microdomains that activate mitochondria-localized calcium-dependent dehydrogenases (pyruvate dehydrogenase, α-KGDH). Severe calcium overload sensitizes mitochondria for mitophagy (via calcium-induced ROS and membrane potential collapse). Simultaneously, ER calcium depletion triggers ER stress-actors (IRE1α, PERK) that induce ER-phagy. Mitochondrial calcium uptake thus coordinates the "decision" to clear both organelles in parallel.

Target gene/protein/pathway: IP3R1 (ITPR1), VDAC1, MCU (mitochondrial calcium uniporter); downstream: ER stress sensors (ERN1/IRE1α, EIF2AK3/PERK).

Supporting evidence with PMIDs:

• ER-mitochondria calcium transfer drives mitophagy (PMID: 25895059)
• IRE1α activation induces ER-phagy via FAM134B (PMID: 28609667)
• PERK activation leads to reticulophagy (PMID: 29339433)
• VDAC1 oligomerization induced by calcium mediates mitophagy (PMID: 29162697)
• IP3R1 dysfunction in Huntington's disease impairs organelle crosstalk (PMID: 28666991)

Confidence: 0.70

Hypothesis 7: Metabolite Signaling Cross-Feeds Between Organelle Autophagy Programs

Title: NAD+/SARM1 axis provides metabolic feedback coupling mitophagy to ER-phagy

Mechanism: SARM1 (TIR domain enzyme) is activated by axonal injury and consumes NAD+ during Wallerian degeneration. Low NAD+ activates mitophagy (via PARP1 inhibition freeing SIRT1/SIRT3 deacetylases) while simultaneously triggering ER stress (via NAD+-dependent deacetylases affecting ER chaperones). This metabolic coupling ensures that mitochondrial stress that depletes NAD+ simultaneously signals for ER quality control to prevent ER stress from overwhelming an already compromised system.

Target gene/protein/pathway: SARM1 (enzyme), NAD+ salvage (NMN/NA salvage), PARP1, SIRT1/3.

Supporting evidence with PMIDs:

• SARM1 activation induces rapid axonal degeneration via NAD+ depletion (PMID: 30209461)
• SIRT3 deacetylates SOD2 and OPA1 to enhance mitophagy (PMID: 26109801)
• NAD+ restoration protects against neurodegenerative models (PMID: 28749327)
• ER stress and NAD+ depletion converge on mitochondrial dynamics (PMID: 31439712)
• NMN supplementation improves organelle quality in aging neurons (PMID: 30341063)

Confidence: 0.62

Summary Table

| # | Target | Confidence | Key readouts |
|---|--------|------------|--------------|
| 1 | MFN2/PACS2 | 0.72 | MAM integrity, dual-organelle autophagy flux |
| 2 | TFEB/TFE3 | 0.78 | Nuclear translocation, multi-organelle proteomics |
| 3 | TBK1-OPTN-NDP52 | 0.81 | Phospho-status, recruitment kinetics |
| 4 | p62 LLPS | 0.75 | Super-resolution droplet composition |
| 5 | VPS34 complexes | 0.65 | PI3P localization, complex interactome |
| 6 | Calcium signaling | 0.70 | Microdomain imaging, rescue with Opto channels |
| 7 | NAD+/SARM1 | 0.62 | Metabolomics, axonal organelle clearance assays | Note: Highest confidence targets (TBK1, TFEB/TFE3) are supported by human genetics (ALS mutations) and have established therapeutic modalities (small-molecule kinase inhibitors, AAV-mediated overexpression). Lower-confidence targets (VPS34 complexity, SARM1) represent emerging biology with mechanistic plausibility but less direct human genetic evidence.

Critical Evaluation of Coordination Hypotheses

Hypothesis 1: MFN2-PACS2 MAM Axis

Weak Links

• MFN2's role as a direct mitophagy receptor is contested: While PMID:31171695 demonstrates MFN2-LC3 interaction, most literature treats MFN2 primarily as a fusion regulator. Its LIR motif functionality varies by context and species.
• PACS2 connection to mitophagy is inferential: PACS2 regulates ER-mitochondria tethering, but evidence for direct involvement in mitophagy initiation is circumstantial.
• Causality vs. correlation of MAM disruption: PMID:31641032 shows MAM integrity is compromised in ALS/PD, but doesn't establish this drives coordination failure rather than reflecting a downstream effect.

Counter-Evidence

• MFN2 knockdown could impair mitophagy indirectly via fusion defects rather than receptor function—fused mitochondria are harder to segregate for autophagic clearance.
• PACS2 is primarily studied in ER quality control; its role in coordinating dual pathways remains speculative.

Falsifying Experiment

• Perform rescue experiments with MFN2 constructs lacking LIR motif but maintaining fusion function. If mitophagy-ER-phagy coordination still fails, the axis requires the receptor function; if coordination is maintained, the axis is downstream of mitochondrial dynamics.

Revised Confidence: 0.58 (down from 0.72)

Hypothesis 2: TFEB/TFE3 Parallel Activation

Weak Links

• Non-specific survival effect: TFEB/TFE3 double KO neurodegeneration (PMID:31801954) could reflect general lysosomal failure, not specifically the loss of "coordination."
• Promoter binding divergence is unproven: The assertion that TFEB/TFE3 preferentially drive mitophagy vs. ER-phagy genes lacks direct ChIP-seq evidence in neurons.
• Heterodimerization evidence is indirect: TFE3 compensating for TFEB loss suggests redundancy, not active coordination.

Counter-Evidence

• TFEB/TFE3 activation is triggered by general stress (mTORC1 inhibition, AMPK). This would induce coordinated response only if the stress simultaneously affects multiple organelles—which may not be the case in disease-specific contexts.
• The CLEAR network (PMID:26942069) includes general autophagy genes; organelle-specific targeting requires additional cargo receptor specificity beyond TFEB/TFE3.

Falsifying Experiment

• Use dCas9-KRAB to selectively repress TFEB or TFE3 binding at organelle-specific gene promoters (determined by neuron-specific ChIP-seq). If only one pathway is impaired, divergence exists; if both fail, coordination is downstream of both factors.

Revised Confidence: 0.64 (down from 0.78)

Hypothesis 3: TBK1-OPTN-NDP52 Phospho-Cascade

Weak Links

• Receptor specificity for ER is underexplored: OPTN for ER-phagy (PMID:32048902) is documented under starvation, but whether TBK1-phosphorylated OPTN engages ER membranes under disease conditions is unclear.
• NDP52 ER targeting is controversial: The cited Salmonella literature (PMID:25985789) involves cytosolic bacteria, not organelle-specific ER fragments. Direct NDP52 engagement of ER vesicles lacks validation.
• "Broadcast" model lacks kinetic evidence: No data showing TBK1 activation precedes coordinated multi-organelle recruitment in real-time.

Counter-Evidence

• TBK1 mutations cause ALS primarily through motor neuron-specific vulnerability; if TBK1 coordinated all organelle quality control, you'd expect broader tissue effects.
• Loss-of-function mutations should impair all selective autophagy equally if the model is correct—but some TBK1 mutations show tissue-specific phenotypes.

Falsifying Experiment

• Perform live-cell imaging of TBK1 activation (using FRET sensor) simultaneously with OPTN and NDP52 recruitment to mitochondria vs. ER. If TBK1 activation precedes both recruitments with similar kinetics, the model holds; if organelle-specific recruitment is sequential or TBK1-independent, the hub model fails.

Revised Confidence: 0.68 (down from 0.81)

Hypothesis 4: p62 Phase Separation

Weak Links

• Mixed-organelle capture is not demonstrated: PMID:31506447 shows p62 bodies "capture both mitochondria and ER," but doesn't prove a single droplet contains both organelles simultaneously—it could be adjacent separate droplets.
• LLPS specificity for damaged organelles: p62 LLPS is triggered by ubiquitylated cargo; damaged mitochondria and ER may have different ubiquitin chain types that partition into distinct droplets.
• Phase separation does not equal coordinated clearance: Concentrating cargo doesn't guarantee both are delivered to the same autophagosome.

Counter-Evidence

• Some studies suggest p62 bodies are organelle-specific, forming distinct mitochondrial vs. ER-associated puncta.
• p62's primary function may be aggregate clearance rather than dynamic organelle quality control.

Falsifying Experiment

• Use super-resolution STORM with 3D rendering to definitively show whether individual p62 droplets contain both Tom20+ and Sec61+ signals in the same aqueous compartment. Perform FRAP to confirm liquid-phase mixing. If organelles remain segregated within distinct p62 droplets, the coordination model fails.

Revised Confidence: 0.61 (down from 0.75)

Hypothesis 5: VPS34 Complex I Heterogeneity

Weak Links

• PI3P signaling is generic: PI3P at ER initiates autophagy but doesn't determine organelle specificity—cargo receptors do.
• NRBF2 recruitment to MAMs (PMID:27840058): Demonstrates localization but doesn't prove NRBF2 discriminates between mitophagy vs. ER-phagy substrates.
• Complex composition may be constitutive: VPS34 complexes may set basal autophagic capacity, not dynamic organelle-specific targeting.

Counter-Evidence

• VPS34 inhibitors (SAR405) block general autophagy; no selective effect on one pathway over another has been demonstrated.
• UVRAG mutations causing neurodegeneration (PMID:25985789) affect overall autophagic flux, not specific organelle clearance.

Falsifying Experiment

• Use rapid immunoprecipitation of endogenous VPS34 complexes followed by mass spec under different stresses. If complex subunits don't significantly rewire within 30 minutes of stress, specificity is set by other mechanisms; if complexes change composition dynamically, test whether swapping subunits redirects PI3P to the alternative organelle.

Revised Confidence: 0.52 (down from 0.65)

Hypothesis 6: Calcium Microdomain Crosstalk

Weak Links

• Temporal sequence is unclear: Does calcium flux first trigger mitophagy, then ER-phagy, or simultaneously? The mechanism for parallel initiation is vague.
• IRE1α/PERK activation doesn't exclusively drive ER-phagy: These sensors trigger integrated stress response including apoptosis; attributing ER-phagy specifically to their activation is reductive.
• VDAC1 oligomerization (PMID:29162697): While demonstrated, VDAC1 is on the OMM and doesn't directly interface with ER-phagy machinery.

Counter-Evidence

• Calcium-induced mitophagy (PMID:25895059) involves mitochondrial permeability transition pore opening—a lethal signal. Whether this represents coordinated quality control vs. pathology is debated.
• IP3R1 dysfunction in HD (PMID:28666991) causes metabolic deficits; the "ER-phagy" consequence is inferred, not measured.

Falsifying Experiment

• Use targeted optogenetics to uncouple mitochondrial calcium uptake from ER calcium release temporally. If ER-phagy can be triggered by IP3R1 activation without mitochondrial calcium uptake (and vice versa), the coupling model fails. Measure each pathway independently with high temporal resolution.

Revised Confidence: 0.58 (down from 0.70)

Hypothesis 7: NAD+/SARM1 Axis

Weak Links

• SARM1 is injury-activated, not disease-relevant: SARM1's primary role is acute axonal degeneration after injury (PMID:30209461). Whether this mechanism applies to chronic neurodegeneration is unclear.
• NAD+ depletion is a general stress signal: Low NAD+ triggers many pathways; attributing specific coordination to SARM1 vs. PARP1 or other NAD+-consuming enzymes is difficult.
• SIRT3 (mitochondrial) vs. general ER NAD+ sensing: SIRT3 is mitochondrial; ER NAD+ sensing mechanisms are less characterized.

Counter-Evidence

• NMN supplementation (PMID:30341063) improves organelle quality, but this may reflect general bioenergetic support, not specific coordination.
• SARM1 knockout is neuroprotective in injury models but doesn't prevent all neurodegenerative pathology.

Falsifying Experiment

• Measure organelle-specific autophagy flux in SARM1 KO vs. PARP1 KO vs. dual KO neurons. If SARM1 deletion specifically impairs both pathways while PARP1 deletion only affects one, the model holds. If NAD+ restoration bypasses SARM1 requirement, SARM1 is upstream but not specific to coordination.

Revised Confidence: 0.49 (down from 0.62)

Revised Summary

| # | Target | Original | Revised | Major Issue |
|---|--------|----------|---------|-------------|
| 1 | MFN2/PACS2 | 0.72 | 0.58 | Direct coordination mechanism unsupported |
| 2 | TFEB/TFE3 | 0.78 | 0.64 | General lysosomal effect, not specific coordination |
| 3 | TBK1-OPTN-NDP52 | 0.81 | 0.68 | ER-targeting of receptors insufficiently validated |
| 4 | p62 LLPS | 0.75 | 0.61 | Organelle co-localization in droplets unproven |
| 5 | VPS34 complexes | 0.65 | 0.52 | PI3P specificity insufficient for coordination |
| 6 | Calcium signaling | 0.70 | 0.58 | Temporal coupling to parallel pathways weak |
| 7 | NAD+/SARM1 | 0.62 | 0.49 | Injury-specific mechanism doesn't fit chronic disease |

Overarching Methodological Concerns

Recommended Priority Experiments

Feasibility Assessment: Organelle-Specific Autophagy Coordination in Neurodegeneration

Executive Summary

Based on critical evaluation of the proposed mechanisms, I identify three priority targets for therapeutic development in coordination of organelle-specific autophagy. The remaining hypotheses, while mechanistically plausible, present significant translational barriers related to target tractability, assay development, or disease relevance.

Priority 1: TBK1-OPTN-NDP52 Phospho-Cascade (Revised Confidence: 0.68)

Druggability Assessment

| Aspect | Rating | Rationale |
|--------|--------|-----------|
| Target Class | Excellent | Serine/threonine kinase with established medicinal chemistry precedent |
| Active Site Tractability | High | ATP-competitive inhibitors widely achievable; structural data available (PDB: 5JPA, 6NXK) |
| Allosteric Potential | Moderate | Protein-protein interactions between receptors and LC3 may be harder to drug |
| Blood-Brain Barrier Penetration | Achievable | Kinase inhibitors can achieve CNS exposure with appropriate physiochemical properties |

Existing Precedents:

• Amgen/Biogen TBK1 inhibitors in oncology/immunology pipelines (e.g., BIIB080/ASG-1ME)
• OPTN mutations (E478G) identified in ALS/FTD — loss-of-function alleles define mechanism
• NDP52/CALCOCO2 — less tractable as PPI target, but downstream of TBK1

Biomarkers & Model Systems

| Category | Recommendations |
|----------|-----------------|
| Pharmacodynamic | pTBK1 (S172), pOPTN (S177), pNDP52 (S67) by phospho-specific ELISA; LC3-II flux in iPSC neurons |
| Organelle-specific readouts | mt-Keima (mitophagy), ER-phyto (reticulophagy), dual-luciferase reporters for coordination |
| Patient stratification | TBK1 LOF variants, GBA1+TBK1 polygenic risk in PD; C9orf72+TBK1 in ALS/FTD |
| Model systems | iPSC-derived cortical/motor neurons from TBK1-mutant ALS patients;睲嚟 CRISPR isogenic lines |

Validation Gaps: Direct evidence that TBK1 phosphorylates NDP52 on ER-derived vesicles in neurons is absent. Recommend orthogonal validation before committing to clinical development.

Clinical Development Constraints

| Factor | Assessment |
|--------|------------|
| Indication selection | TBK1 mutations cause ALS/FTD — pursue ALS with TBK1 mutation as genetically defined cohort (rare: ~1-2% of ALS) |
| Regulatory path | Orphan designation applicable; accelerated approval possible with biomarker endpoint |
| Competitive timeline | Amgen TBK1 inhibitor (BIIB080) in Phase 1 for ALS (NCT05683578) — differentiation needed |
| Combination potential | Synergistic with autophagosome-lysosome fusion enhancers (e.g., PIKFYVE inhibitors) |

Key Development Risk: TBK1 has pleiotropic functions (NF-κB, interferon signaling). Full inhibition may cause immunosuppression; partial inhibition or allele-specific targeting may be required.

Safety Profile

| Risk | Mitigation Strategy |
|------|---------------------|
| Immune dysregulation | CNS-restricted delivery (AAV9, intrathecal); intermittent dosing |
| Off-target kinase inhibition | Selectivity profiling against 400+ kinases; limit systemic exposure |
| Heterozygosity concerns | TBK1 haploinsufficiency appears tolerated (patients are heterozygous); design for partial inhibition |

Timeline & Cost Realism

| Phase | Estimate |
|-------|----------|
| Lead optimization + CNS PK/PD | 18-24 months, $3-5M |
| IND-enabling studies (GLP tox, PK)
| 12-18 months, $4-6M |
| Phase 1 (CNS-penetrant dose escalation) | 24 months, $8-12M |
| Phase 2/3 (genetically defined cohort) | 36-48 months, $30-50M |
| Total to approval | 7-9 years, $60-100M |

Major uncertainty: Biomarker-driven development (phospho-OPTN as surrogate) could accelerate, but will require FDA dialogue on regulatory acceptance.

Priority 2: TFEB/TFE3 Parallel Activation (Revised Confidence: 0.64)

Druggability Assessment

| Aspect | Rating | Rationale |
|--------|--------|-----------|
| Target Class | Moderate | Transcription factors historically difficult to drug; nuclear localization is indirect control |
| Upstream tractability | High | mTORC1 inhibitors (rapalogs, everolimus) approved; mTORC1-independent TFEB activation achievable via MG53 or LKB1-AMPK pathway |
| Gene therapy approach | Mature | AAV-mediated TFEB/TFE3 expression viable; multiple CNS gene therapy precedents (SPINRAZA, Zolgensma) |
| BBB penetration | Variable | mTOR inhibitors have variable CNS penetration; AAV9 crosses BBB in non-human primates |

Recommended Modality: AAV-mediated TFEB overexpression in neurons (Proof-of-concept via intracranial delivery) OR small-molecule TFEB/TFE3 nuclear translocation agonists (e.g., trehalose, disaccharide derivatives).

Biomarkers & Model Systems

| Category | Recommendations |
|----------|-----------------|
| Pharmacodynamic | Nuclear:cytoplasmic TFEB ratio (IF); CLEAR network gene expression (RT-PCR panel: 20-30 genes); LAMP1, CTSB upregulation |
| Coordination readouts | Multi-organelle proteomics (TMT labeling); simultaneous mitophagy/reticulophagy flux in same neurons |
| Patient stratification | mTORC1 hyperactivation (e.g., TSC mutations) in neurodegeneration? Less established; focus on AD/PD without genetic stratification |
| Model systems | 3D brain organoids; AAV-mediated TFEB in mouse neurodegenerative models (MPTP, α-syn PFF) |

Critical Validation Needed: Direct ChIP-seq demonstrating TFEB/TFE3 divergence toward mitophagy vs. reticulophagy genes in neurons. Current evidence is correlative.

Clinical Development Constraints

| Factor | Assessment |
|--------|------------|
| Indication selection | Broad potential in AD, PD, ALS; but no genetically defined subgroup. Consider AD with evidence of autophagysome-lysosome dysfunction (CSF CTSD elevation) |
| Regulatory path | Standard development path; no accelerated pathway without genetic anchor |
| Competition | mTOR inhibitors in AD trials (everolimus, sirolimus); TFEB agonists in preclinical development |
| Combination potential | Synergistic with lysosomal enzyme replacement (for lysosomal storage disorders overlapping with neurodegeneration) |

Development Risk: TFEB/TFE3 activation drives lysosomal biogenesis broadly. Cannot selectively enhance "coordination" without affecting general autophagy. Therapeutic window depends on disease-specific autophagic failure.

Safety Profile

| Risk | Mitigation Strategy |
|------|---------------------|
| Autophagy过度 | mTOR inhibition causes immunosuppression, metabolic effects; TFEB overexpression may increase lysosomal storage disease risk |
| Tumorigenesis | Autophagy inhibition is anti-tumor; TFEB activation theoretically promotes tumor survival — monitor for malignancy signals |
| Lysosomal membrane permeabilization | Excessive lysosomal biogenesis may destabilize membranes; dose-titration critical |

Timeline & Cost Realism

| Phase | Estimate |
|-------|----------|
| Lead identification (agonists) or AAV construct optimization | 24-30 months, $5-8M |
| IND-enabling studies | 12-18 months, $4-6M |
| Phase 1 (safety, PK, target engagement) | 18-24 months, $10-15M |
| Phase 2 (efficacy in AD/PD) | 36-48 months, $40-60M |
| Total to approval | 7-10 years, $80-150M |

Note: AAV approach may shorten Phase 1 (single-dose escalation) but faces manufacturing cost ($1-3M per patient for AAV9).

Priority 3: p62 Phase Separation (Revised Confidence: 0.61)

Druggability Assessment

| Aspect | Rating | Rationale |
|--------|--------|-----------|
| Target Class | Challenging | Intrinsically disordered scaffold protein; phase separation not traditional drug target |
| Kinase upstream | Moderate | TBK1, CK2 inhibitors could modulate p62 phosphorylation (S403) — indirect approach |
| PPI disruption | Difficult | p62-Ubiquitin and p62-LC3 interfaces are large, flat surfaces |
| BBB penetration | Unknown | No CNS data for p62 modulators |

Recommended Modality: CK2 or TBK1 inhibitors to modulate p62 phosphorylation state (indirect) OR peptidomimetics targeting p62 multimerization interfaces.

Low confidence in direct targeting. Prioritize validating phase separation as coordination mechanism before committing to drug discovery.

Biomarkers & Model Systems

| Category | Recommendations |
|----------|-----------------|
| Pharmacodynamic | p62 Ser403 by phospho-specific antibodies; p62 body formation (fluorescence recovery after photobleaching, FRAP); NRF2 target genes (NQO1, HMOX1) |
| Coordination readouts | Super-resolution STORM to validate hetero-organellar p62 droplets — this is the key experiment |
| Model systems | Primary neurons from p62 KO mice; patient iPSC neurons with SQSTM1 variants |

Validation Gaps: Whether individual p62 droplets truly contain both mitochondria and ER markers has not been definitively shown. This must be established before target prioritization.

Clinical Development Constraints

| Factor | Assessment |
|--------|------------|
| Indication selection | SQSTM1 mutations cause ALS/FTD and Paget's disease — smallest genetically defined subgroup |
| Regulatory path | Orphan designation applicable; biomarker-driven development requires extensive FDA negotiation |
| Competitive landscape | No direct competitors; TBK1/CK2 inhibitors could be repurposed |

Major Risk: Phase separation as therapeutic target is unprecedented. Even if coordination mechanism is validated, pharmacologic modulation of LLPS is uncharted territory.

Safety Profile

| Risk | Mitigation Strategy |
|------|---------------------|
| Loss of aggregate clearance | p62 deletion causes neurodegeneration in mice — therapeutic window may be narrow |
| NRF2 pathway effects | p62 activates NRF2; excessive p62 activity could cause oxidative stress from NRF2 hyperactivation |
| Off-target kinase effects | If using CK2/TBK1 inhibitors, kinase selectivity critical |

Timeline & Cost Realism

| Phase | Estimate |
|-------|----------|
| Mechanism validation + assay development | 24-36 months, $4-6M (high attrition risk) |
| Lead optimization (if target validated) | 24 months, $5-7M |
| IND-enabling + Phase 1 | 24-30 months, $15-20M |
| Total to proof-of-concept | 5-7 years, $30-40M |

Overall assessment: This hypothesis should be de-risked experimentally (STORM validation) before significant investment. If droplets are organelle-segregated, this target should be deprioritized.

Lower-Priority Hypotheses: Summary Dismission

| Hypothesis | Revised Confidence | Primary Barrier |
|------------|-------------------|-----------------|
| MFN2/PACS2 | 0.58 | Mechanism requires LIR-independent coordination; MFN2 is multi-functional (fusion) |
| VPS34 complexes | 0.52 | PI3P insufficient for specificity; VPS34 inhibitors block bulk autophagy |
| Calcium signaling | 0.58 | Temporal coupling unproven; calcium induces cell death, not coordinated QC |
| NAD+/SARM1 | 0.49 | Injury-specific mechanism; chronic neurodegeneration context unclear |

Consolidated Strategic Recommendations

Tier 1: Advance to IND-Enabling (Next 6-12 Months)

| Target | Confidence | Action | Investment |
|--------|------------|--------|------------|
| TBK1 | 0.68 | License or partner on existing kinase inhibitor; assess CNS exposure; run PK/PD in ALS iPSC neurons | $5-8M |
| TFEB/TFE3 | 0.64 | AAV-TFEB construct for direct CNS delivery; GLP tox for AAV9 in NHPs | $8-12M |

Tier 2: De-risk Mechanistically (12-24 Months)

| Target | Confidence | Action | Investment |
|--------|------------|--------|------------|
| p62 LLPS | 0.61 | STORM validation of hetero-organellar droplets; if confirmed, initiate assay development | $2-4M |
| MFN2/PACS2 | 0.58 | LIR motif rescue experiment to establish mechanism; if confirmed, assess MFN2 as indirect target | $1-2M |

Tier 3: Monitor (No Active Investment)

| Target | Confidence | Reason |
|--------|------------|--------|
| VPS34 complexes | 0.52 | Insufficient specificity; await dynamic complex composition data |
| Calcium signaling | 0.58 | Unproven temporal coordination; mPTP induction suggests toxicity, not QC |
| NAD+/SARM1 | 0.49 | Injury-specific; chronic neurodegeneration context insufficient |

Cross-Cutting Constraints for All Programs

| Factor | Recommendation |
|--------|----------------|
| Biomarker development | Multi-organelle autophagy flux assays in CSF (mitochondrial DNA, ER proteins) for patient stratification |
| Clinical endpoints | Biomarker-based surrogate endpoints (autophagy flux, lysosomal function) to enable accelerated approval |
| Regulatory engagement | Pre-IND meeting with FDA for all programs; discuss biomarker-driven approval pathways |
| Combination strategies | TBK1 + TFEB activation may be synergistic; consider staggered development |
| Patient stratification | Genetics-first approach (TBK1, SQSTM1, C9orf72) most likely to show efficacy |

Final Assessment

The field is moving from descriptive biology (organelle autophagy defects in neurodegeneration) toward mechanistic understanding (coordination between pathways). The TBK1-OPTN-NDP52 axis represents the strongest combination of human genetics, target tractability, and assay feasibility. TFEB/TFE3 offers a broader therapeutic approach but lacks genetic specificity. p62 phase separation is scientifically interesting but requires validation before investment.

Recommended portfolio: 60% TBK1 program, 30% TFEB/TFE3 program, 10% p62 de-risking—with reassessment at 18 months based on emerging validation data.