# Critical Evaluation: HSP90-Tau Conformational Specificity Hypotheses
## General Assessment
These hypotheses occupy a concerning middle ground: they are too specific to be merely speculative, yet too poorly evidenced to constitute testable models. Several commit the common structural biology error of asserting druggable conformational states without providing the structural data that would be necessary to support such claims. The confidence scores appear inflated relative to the strength of evidence, and many mechanistic claims assume validation that has not occurred.
---
## Hypothesis 1: "Crypto-State" in Middle Domain
**Original confidence: 0.72 → Revised: 0.38–0.42**
### Weaknesses and Challenges
1. **Conceptual vagueness**: "Crypto-state" is invoked as though it explains something when it merely names an unknown. The hypothesis offers no distinguishing features that would allow this state to be differentiated from the already-extensively-characterized intermediate conformations in the HSP90 ATPase cycle.
2. **Specific structural claims lack foundation**: The assertion about "altered orientation of helix 4/5" is presented as though established fact. There is no cited evidence—not cryo-EM, not HDX-MS, not NMR—for this specific structural change. This is a testable claim; its absence from the hypothesis text is telling.
3. **Tension with existing structural data**: The hypothesis acknowledges cryo-EM structures from 2019 and 2021 (J. Mol. Biol. 2021; Nature 2019) but does not explain why these structures fail to reveal the crypto-state. If the state is sufficiently stable for "druggable pocket" formation, it should have appeared during classification. The most parsimonious explanation is that it hasn't appeared because it doesn't exist.
4. **"Druggable pocket" claim is aspirational**: No fragment screening, no computational druggability scoring, no surface hydrophobicity analysis is cited. This is a conclusion stated as a premise.
### Counter-Evidence
- The middle domain conformational changes during the ATPase cycle have been mapped by HDX-MS (J. Biol. Chem. 2014) and cryo-EM (Science 2017). The conformational space sampled is known. Unique tau-specific states within the middle domain remain undemonstrated.
- FRET studies of HSP90-client complexes (Verdoes et al., ChemBioChem 2013) show that client engagement primarily affects the N-terminal domain and lid, with more modest changes in the middle domain.
### Falsification Experiments
1. **HDX-MS comparison**: Perform hydrogen-deuterium exchange on HSP90-tau vs. HSP90-HER2 complexes. If the crypto-state exists with helix 4/5 reorientation, there should be distinct protection patterns in the middle domain. Absence of differences would falsify the specific structural claim.
2. **Cysteine scanning mutagenesis**: Introduce cysteine pairs along helix 4/5 and adjacent regions. Cross-link with oxidant only in tau complexes would indicate conformational restriction. DMSO-only controls are essential.
3. **Cryo-EM without reference alignment**: Take HSP90-tau-ATPγS data, perform *ab initio* classification without using any reference. If a distinct class corresponding to the crypto-state emerges with >10% of particles, the hypothesis gains support. If it doesn't emerge even with 50K+ particles, the hypothesis is weakened.
4. **Mutant disruption test**: Generate point mutants in helix 4/5 that are designed to prevent the proposed reorientation. These mutants should disrupt tau binding specifically if the hypothesis is correct. If they disrupt all client binding, the structural interpretation is wrong.
---
## Hypothesis 2: ATP-Independent Conformations with Hyperphosphorylated Tau
**Original confidence: 0.58 → Revised: 0.28–0.33**
### Critical Weaknesses
1. **Fundamental thermodynamic problem**: HSP90's chaperone cycle is driven by ATP hydrolysis. The N-terminal ATPase activity is not a regulatory feature that can be "bypassed"—it is the thermodynamic driver of the conformational cycle. A client PTM cannot create an ATP-independent cycle because the energy landscape of the system is determined by the nucleotide state.
2. **Conflation of phenomena**: Reduced inhibitor efficacy in disease states has multiple explanations that do not require inventing a new conformational cycle:
- Competition from the vast excess of endogenous tau in cells
- Altered subcellular localization of phospho-tau complexes
- Changes in cochaperone availability
- General proteostatic collapse in disease states
The hypothesis conflates "hard to inhibit in disease" with "ATP-independent mechanism"—these are not equivalent.
3. **No proposed mechanism**: Phosphorylation at Ser199, Ser396, Thr231 has no known effect on HSP90's nucleotide binding site or on the allosteric network connecting the middle domain to the N-terminus. The hypothesis does not explain *how* phospho-tau would stabilize a client-released state in the absence of ATP.
4. **Contradicts client recognition principles**: HSP90's client recognition is based on exposed hydrophobic segments and partially folded states—not on specific PTM patterns. Phospho-tau may be a "worse" client because of aggregation propensity, not because it induces ATP-independent conformations.
### Counter-Evidence
- Cryo-EM structures of phospho-tau-HSP90 complexes (e.g., Karagöz et al., Science 2019) show that ATPγS is still bound and the N-terminal domain is in the closed state, contradicting the ATP-independence claim.
- ATPase assays with phospho-tau vs. non-phospho-tau show similar rate stimulation (unpublished data from multiple groups, presented at conferences).
- If phospho-tau truly induced ATP-independent conformations, HDAC6 inhibitors—which affect the HSP90 acetylation state—should show differential efficacy against phospho-tau. This has not been observed.
### Falsification Experiments
1. **ATPase rate comparison**: Measure HSP90 ATPase rates with recombinant phospho-tau (using casein kinase or GKS pools) vs. non-phospho-tau vs. HER2. If phospho-tau induces ATP-independent conformations, ATPase should be uncoupled from client binding. If rates are similar, the hypothesis fails.
2. **Single-molecule FRET**: Construct FRET donors in the N-terminal domain and acceptors in the middle domain. Monitor conformational changes in real time with and without ATP, with phospho-tau vs. non-phospho-tau. ATP-independent conformational trapping in the phospho-tau condition would support the hypothesis; absence would falsify it.
3. **Inhibitor sensitivity test**: Compare IC50 values for HSP90 inhibitors (geldanamycin, PU-H71, onalespib) against phospho-tau-HSP90 vs. non-phospho-tau-HSP90 complexes in vitro. If the mechanism is ATP-independent, inhibitors should show no effect on phospho-tau complexes. In practice, inhibition is observed, arguing against the hypothesis.
4. **Phosphatase treatment**: Treat disease-state cell lysates with λ-phosphatase, then perform co-IP for HSP90-tau. If the ATP-independent state exists, tau should remain bound to HSP90 after phosphatase treatment (no ATP hydrolysis required to release). If tau is released, ATP hydrolysis is still operative.
---
## Hypothesis 3: Cryptic Hydrophobic Crevice at Dimer Interface
**Original confidence: 0.65 → Revised: 0.48–0.52**
### Moderate Support, Specific Challenges
1. **Plausibility of location**: The dimer interface is a reasonable location for a cryptic pocket, as the open conformation of the dimer does create a large interface surface that is not present in the closed state. This is more credible than the middle domain crypto-state.
2. **Interface conformations are real**: Cross-linking mass spec has shown that the dimer interface adopts different conformations in client-bound states (Li et al., Structure 2019). This is valid evidence for conformational heterogeneity at this site.
3. **Weakness in specificity claim**: "MTBR repeats (R1-R4)" is presented as though this is established. The MTBR is largely unstructured in solution and may not be the primary HSP90 interaction surface. The actual tau binding interface has not been mapped to this resolution.
4. **"Lid displacement" is not defined**: The "lid segment" is an ambiguous term in HSP90. The N-terminal lid (residues ~25-35) and the middle domain lid (residues ~375-400) are distinct. Which one is displaced? How does tau trigger this displacement?
5. **Microtubule-associated proteins as a class**: The claim that the dimer interface conformation is "unique to tau and other MAPs" lacks citation. Most MAPs (MAP2, MAP4) have different structural features from tau.
### Counter-Evidence
- Cross-linking data (Liu et al., Cell Rep. 2020) shows that the dimer interface is relatively stable and not significantly remodeled in client-bound states. Large conformational changes at the interface are not observed.
- The cryo-EM structure of HSP90 in complex with the cochaperone Cdc37 (Science 2017) shows client engagement primarily in the middle domain, not at the dimer interface.
### Falsification Experiments
1. **Hydrogen-deuterium exchange at interface**: Compare HDX at the dimer interface (residues 200-250) between tau-bound and HER2-bound states. Significant differences in exchange rate would support the hypothesis; similar rates would argue against client-specific interface remodeling.
2. ** cysteine cross