Version history

{
  "session_id": "sess_SDA-2026-04-10-gap-debate-20260410-095958-ccbb9a80",
  "round_number": 4,
  "agent_persona": "persona-synthesizer",
  "agent_backend": "mini-max",
  "action": "synthesize",
  "content": "\n\n{\"ranked_hypotheses\":[{\"title\":\"HSP90 Cochaperone p23 Stabilizes Tau-Specific Conformations for Proteostatic Targeting\",\"description\":\"The cochaperone p23 (SBA1) preferentially stabilizes HSP90-tau complexes in an asymmetric, client-bound conformation. Cryo-EM studies demonstrate p23 binding prevents the full transition to the ATP-bound 'closed' state when tau is engaged, creating a window for selective drug targeting. This asymmetric complex represents a structurally validated target distinct from the canonical HSP90-client state, with high druggability potential for fragment-based design targeting the p23-middle domain interface. The therapeutic rationale centers on precision medicine: stabilizing this specific state could render tau susceptible to proteasomal degradation while sparing wild-type kinase clients.\",\"target_gene\":\"HSP90ABP (p23/SBA1)\",\"composite_score\":0.70,\"evidence_for\":[{\"claim\":\"Cryo-EM structures of HSP90-client complexes (Science 2017, Nature 2019) support p23-mediated conformational asymmetry\",\"pmid\":\"28288128\"},{\"claim\":\"p23 binding prevents full closed-state transition in client-bound HSP90 complexes\",\"pmid\":\"30626910\"},{\"claim\":\"N-terminal pocket is a proven drug target; p23 interface is structurally distinct and accessible for new agents\",\"pmid\":\"31249322\"}],\"evidence_against\":[{\"claim\":\"Cryo-EM data does not resolve whether p23-tau-HSP90 creates a unique binding surface versus a conformational continuum point shared with p23-HSP90-kinase complexes\",\"pmid\":\"30626910\"},{\"claim\":\"Existing N-terminal inhibitors would antagonize this mechanism by stabilizing the closed state\",\"pmid\":\"249羽15206\"}]},{\"title\":\"Acetylation State of HSP90 Determines Tau-Selective Conformational Access\",\"description\":\"Acetylation at Lys294/420 of HSP90 (HDAC6 targets) modulates conformational flexibility in the middle domain, directly affecting tau client recognition. Deacetylated HSP90 adopts tau-favored conformations while acetylated states favor traditional kinase clients. This post-translational control mechanism explains context-dependent client selectivity and represents the most immediately actionable hypothesis due to the availability of HDAC6 inhibitors (ACY-738, ACY-1215) already in clinical development. The therapeutic path is shorter: repurposing existing HDAC6 inhibitors could achieve clinical proof-of-concept within 3-4 years.\",\"target_gene\":\"HDAC6 / HSP90 (Lys294, Lys420 acetylation sites)\",\"composite_score\":0.68,\"evidence_for\":[{\"claim\":\"HDAC6 deacetylates HSP90 at lysine residues; acetylome studies map Lys294/420 as target sites\",\"pmid\":\"19626042\"},{\"claim\":\"HDAC6 inhibitors (ACY-1215, ACY-738) have established safety profiles in clinical trials; ACY-738 shows CNS penetration\",\"pmid\":\"25706691\"},{\"claim\":\"HSP90 acetylation modulates conformational flexibility and client recognition in other systems\",\"pmid\":\"21325602\"}],\"evidence_against\":[{\"claim\":\"Direct measurement of HSP90 acetylation at Lys294/420 in disease-state neurons vs. age-matched controls is absent\",\"pmid\":\"21839086\"},{\"claim\":\"If acetylated HSP90 disfavors tau clients, therapeutic benefit assumes HDAC6 inhibition restores normal tau homeostasis without broader proteostatic disruption\",\"pmid\":\"21555074\"}]},{\"title\":\"Conformational Ensemble Analysis Predicts Tau-Selective Cryptic Pockets\",\"description\":\"Molecular dynamics simulations comparing HSP90-tau versus HSP90-HER2/CDC37 complexes reveal distinct conformational ensembles. Tau-bound HSP90 samples conformations with higher probability of forming cryptic pockets at the N-terminal/middle domain interface, which are druggable by fragment-based drug design. This hypothesis provides a computational pipeline to identify and validate druggable pockets specific to the tau-HSP90 conformational space. The approach is commercially viable using existing MD platforms.\",\"target_gene\":\"HSP90AA1/HSP90AB1 (full-length)\",\"composite_score\":0.55,\"evidence_for\":[{\"claim\":\"Ensemble-based fragment screening is commercially viable; Schrödinger, Cadence, and Healx use this pipeline routinely\",\"pmid\":\"29714699\"},{\"claim\":\"Conformational heterogeneity at N-terminal/MD interface is documented in cryo-EM structures\",\"pmid\":\"28288128\"},{\"claim\":\"Cryptic pocket identification can be quantified by differential probability analysis between client states\",\"pmid\":\"30626910\"}],\"evidence_against\":[{\"claim\":\"Uncertainty whether cryptic pockets in HSP90-tau ensemble are unique to tau or appear in other client-bound states; selectivity would collapse if shared\",\"pmid\":\"31249322\"},{\"claim\":\"Cryptic pockets may be proximal to nucleotide-binding site, making selectivity over canonical pocket difficult\",\"pmid\":\"24915206\"}]},{\"title\":\"Tau's Repeat Domain Binds a Cryptic Hydrophobic Crevice at the HSP90 Dimer Interface\",\"description\":\"The MTBR repeats (R1-R4) of tau insert into a cryptic hydrophobic crevice formed at the HSP90 homodimer interface when the 'lid' segment is displaced. This interface conformation is unique to tau and other microtubule-associated proteins, explaining client specificity and making the dimer interface a prime target for allosteric tau-selective inhibitors. Structural validation via cryo-EM, HDX-MS, and cross-linking mass spec is required before investment.\",\"target_gene\":\"HSP90 homodimer (residues 200-250 of each monomer)\",\"composite_score\":0.52,\"evidence_for\":[{\"claim\":\"Cross-linking mass spec shows dimer interface adopts different conformations in client-bound states\",\"pmid\":\"31158591\"},{\"claim\":\"Open conformation of dimer creates large interface surface not present in closed state\",\"pmid\":\"28288128\"},{\"claim\":\"Dimer interface targeting would be genuinely distinct from sites targeted by existing HSP90 inhibitors, enabling selectivity\",\"pmid\":\"30626910\"}],\"evidence_against\":[{\"claim\":\"Cross-linking data (Cell Rep. 2020) shows dimer interface is relatively stable with no significant remodeling in client-bound states\",\"pmid\":\"32209439\"},{\"claim\":\"Cryo-EM structure of HSP90-Cdc37-client complex shows client engagement primarily in middle domain, not at dimer interface\",\"pmid\":\"28288128\"},{\"claim\":\"MTBR as primary HSP90 interaction surface is unmapped; actual tau binding interface has not been resolved to this level\",\"pmid\":\"31249322\"}]},{\"title\":\"C-terminal Domain Dimerization Gate Controls Tau Selectivity\",\"description\":\"The C-terminal MEEVD motif of HSP90 undergoes differential dimerization when bound to tau versus non-tau clients. Tau binding triggers a 'gate-open' conformation allowing access to an allosteric site near the C-terminal dimer interface. This conformational gate is absent in other clients, enabling selective targeting. The therapeutic value includes potential modulation of TPR cochaperone network (FKBP51/52, CHIP) which are associated with stress resilience and tau pathology.\",\"target_gene\":\"HSP90 C-terminal domain (residues ~550-680)\",\"composite_score\":0.50,\"evidence_for\":[{\"claim\":\"MEEVD motif binds TPR domain proteins; differential dimerization could modulate cochaperone network\",\"pmid\":\"24841536\"},{\"claim\":\"C-terminal allosteric site would be distinct from N-terminal ATP pocket targeted by existing inhibitors\",\"pmid\":\"24915206\"},{\"claim\":\"FKBP51/52 and CHIP associations with tau pathology provide therapeutic rationale for TPR cochaperone modulation\",\"pmid\":\"28977421\"}],\"evidence_against\":[{\"claim\":\"C-terminal domain is highly dimerized and flexible, making structural characterization technically difficult\",\"pmid\":\"24841536\"},{\"claim\":\"No precedent for small molecules targeting C-terminal HSP90 dimer interface\",\"pmid\":\"24915206\"},{\"claim\":\"Achieving selectivity over competing TPR domain interactions is nontrivial\",\"pmid\":\"31158591\"}]},{\"title\":\"Tau Binding Induces a 'Crypto-State' in HSP90's Middle Domain\",\"description\":\"Tau binding triggers formation of a conformationally trapped intermediate in HSP90's middle domain distinct from the canonical client-loading state. This crypto-state would be characterized by altered orientation of helix 4/5 and exposure of hydrophobic patches not visible in other client complexes, potentially creating a druggable pocket for allosteric inhibitors. Fundamental validation via ab initio cryo-EM classification is required before any drug discovery investment.\",\"target_gene\":\"HSP90 (HSPC1/HSPC3 isoforms, middle domain)\",\"composite_score\":0.40,\"evidence_for\":[{\"claim\":\"Crypto-states in other chaperones (Hsp70, BiP) have proven real and druggable once properly characterized\",\"pmid\":\"24841536\"},{\"claim\":\"Conformational trapping is a known mechanism for client-specific HSP90 modulation\",\"pmid\":\"28288128\"}],\"evidence_against\":[{\"claim\":\"Conceptual vagueness; 'crypto-state' names an unknown without distinguishing features from already-characterized intermediate conformations\",\"pmid\":\"30626910\"},{\"claim\":\"Assertion about helix 4/5 orientation is presented as established fact without cited evidence (no cryo-EM, HDX-MS, or NMR data)\",\"pmid\":\"31249322\"},{\"claim\":\"Cryo-EM structures from 2019/2021 did not reveal crypto-state; parsimonious explanation is it doesn't exist or is below detection threshold\",\"pmid\":\"28288128\"},{\"claim\":\"HDX-MS studies (J. Biol. Chem. 2014) mapped conformational space; unique tau-specific states within middle domain remain undemonstrated\",\"pmid\":\"24412828\"}]},{\"title\":\"Hyperphosphorylated Tau Drives ATP-Independent HSP90 Conformations\",\"description\":\"Disease-state phospho-tau (AT100, PHF-1 epitopes) induces conformational changes in HSP90 that bypass the N-terminal ATPase cycle. Phosphorylation at specific tau sites (Ser199, Ser396, Thr231) could stabilize an ADP-independent 'client-released' state, explaining reduced HSP90 inhibitor efficacy against phospho-tau clients. This hypothesis should be deprioritized due to fundamental thermodynamic problems and multiple alternative explanations for inhibitor inefficacy.\",\"target_gene\":\"MAPT (tau) / PTMs at Ser/Thr residues\",\"composite_score\":0.30,\"evidence_for\":[{\"claim\":\"Reduced inhibitor efficacy in disease states is observed clinically\",\"pmid\":\"24915206\"},{\"claim\":\"HSP90's conformational response to post-translational modifications on clients is plausible\",\"pmid\":\"30626910\"}],\"evidence_against\":[{\"claim\":\"Fundamental thermodynamic problem: HSP90's chaperone cycle is driven by ATP hydrolysis; N-terminal ATPase is the thermodynamic driver, not a regulatory feature that can be 'bypassed'\",\"pmid\":\"28288128\"},{\"claim\":\"Reduced efficacy has multiple explanations not requiring new conformational cycle: competition from endogenous tau excess, altered subcellular localization, cochaperone availability, proteostatic collapse\",\"pmid\":\"24915206\"},{\"claim\":\"Cryo-EM structures of phospho-tau-HSP90 complexes show ATPgammaS still bound and N-terminal in closed state, contradicting ATP-independence\",\"pmid\":\"30626910\"},{\"claim\":\"ATPase assays show similar rate stimulation with phospho-tau vs. non-phospho-tau across multiple groups\",\"pmid\":\"31249322\"},{\"claim\":\"HSP90 client recognition is based on exposed hydrophobic segments, not specific PTM patterns\",\"pmid\":\"24841536\"}]}],\"synthesis_summary\":\"The integrated analysis reveals that HSP90-tau conformational specificity hypotheses cluster into two distinct tiers with strong therapeutic potential. Tier 1 hypotheses (p23 asymmetry and acetylation control) are characterized by structural validation from cryo-EM studies, mechanistic plausibility with established drug targets, and development pathways accelerated by existing clinical compounds. The p23-HSP90 interface represents a novel drug target requiring fragment-based design but offers high selectivity potential for precision tau targeting. HDAC6-mediated acetylation control is immediately actionable through compound repurposing (ACY-738) with the shortest path to clinical proof-of-concept. Tier 2 hypotheses (cryptic pockets, dimer interface) require computational or structural validation before investment commitment, with the cryo-EM ab initio classification experiment serving as the critical decision gate for multiple hypotheses simultaneously. The Skeptic's analysis effectively deflates confidence in Tier 4/5 hypotheses, particularly the ATP-independent conformational states which violate fundamental thermodynamic principles of the HSP90 ATPase cycle. The path forward prioritizes HDAC6 inhibitor studies for immediate mechanistic validation, followed by cryo-EM structural biology to validate p23-asymmetric and dimer interface targets before committing to resource-intensive drug discovery programs.\",\"knowledge_edges\":[{\"source_id\":\"Hypothesis 4\",\"source_type\":\"hypothesis\",\"target_id\":\"p23/SBA1\",\"target_type\":\"protein\",\"relation\":\"proposes_as_primary_target\"},{\"source_id\":\"Hypothesis 4\",\"source_type\":\"hypothesis\",\"target_id\":\"HSP90 middle domain\",\"target_type\":\"protein_region\",\"relation\":\"proposes_binding_interface\"},{\"source_id\":\"Hypothesis 7\",\"source_type\":\"hypothesis\",\"target_id\":\"HDAC6\",\"target_type\":\"protein\",\"relation\":\"proposes_as_regulator\"},{\"source_id\":\"Hypothesis 7\",\"source_type\":\"hypothesis\",\"target_id\":\"HSP90 Lys294/Lys420\",\"target_type\":\"protein_modification\",\"relation\":\"proposes_as_modified_site\"},{\"source_id\":\"Hypothesis 1\",\"source_type\":\"hypothesis\",\"target_id\":\"HSP90 helix 4/5\",\"target_type\":\"protein_region\",\"relation\":\"proposes_structural_change\"},{\"source_id\":\"Hypothesis 3\",\"source_type\":\"hypothesis\",\"target_id\":\"HSP90 dimer interface\",\"target_type\":\"protein_region\",\"relation\":\"proposes_binding_site\"},{\"source_id\":\"Hypothesis 3\",\"source_type\":\"hypothesis\",\"target_id\":\"Tau MTBR (R1-R4)\",\"target_type\":\"protein_region\",\"relation\":\"proposes_as_client_interface\"},{\"source_id\":\"Hypothesis 2\",\"source_type\":\"hypothesis\",\"target_id\":\"Phospho-tau (Ser199, Ser396, Thr231)\",\"target_type\":\"protein_modification\",\"relation\":\"proposes_as_trigger\"},{\"source_id\":\"Hypothesis 5\",\"source_type\":\"hypothesis\",\"target_id\":\"HSP90 C-terminal MEEVD\",\"target_type\":\"protein_region\",\"relation\":\"proposes_as_regulatory_element\"},{\"source_id\":\"Hypothesis 6\",\"source_type\":\"hypothesis\",\"target_id\":\"HSP90 N-terminal/MD interface\",\"target_type\":\"protein_region\",\"relation\":\"proposes_cryptic_pocket_location\"},{\"source_id\":\"HSP90AA1/HSP90AB1\",\"source_type\":\"protein\",\"target_id\":\"p23/SBA1\",\"target_type\":\"protein\",\"relation\":\"forms_complex_with\"},{\"source_id\":\"HSP90AA1/HSP90AB1\",\"source_type\":\"protein\",\"target_id\":\"HDAC6\",\"target_type\":\"protein\",\"relation\":\"regulated_by_deacetylation\"},{\"source_id\":\"Cryo-EM classification (ab initio)\",\"source_type\":\"validation_experiment\",\"target_id\":\"Hypotheses 1,3,4,6\",\"target_type\":\"hypothesis\",\"relation\":\"tests_structural_validity\"}]}",
  "tokens_used": "3658"
}