STRC Pharmacochaperone Phase 4g — Repurpose Screen
Rigid-receptor Vina docking of three repurposable chemical chaperones against the K1141 pocket on three STRC models (clinical Ultra-Mini × TMEM145 construct + WT decoy + E1659A mutant decoy). Driven by STRC Research Monitor routine papers flagging 4PBA as an FDA-approved in-vivo chaperone and IP-045 as a clinical indole-class hit.
Motivation
Phase 4c showed all five top-5 Phase 3C leads prefer WT over E1659A by 0.40-0.55 kcal/mol (mean +0.455) — falsifying aspect (a) of the pharmacochaperone mechanism (“carboxylate replaces E1659”). Aspect (b) “loop-capping fold-stabilizer” remained untested by rigid docking. Rather than continue de-novo screening from the same chemotype, test whether already-validated clinical/preclinical chaperones show the missing selectivity — a hit would shorten the clinical path to months (repurposing) instead of years (de-novo lead → FDA).
Roster
| Compound | SMILES | Provenance | Clinical status |
|---|---|---|---|
| 4-phenylbutyrate (4PBA) | OC(=O)CCCc1ccccc1 | hayase 2026 — DSCAML1 A2105T in-vivo oral rescue | FDA approved (Buphenyl, urea cycle disorder) |
| IP-045 | O=C(OC1=CC=CC=C1F)CCc2c[nH]c3ccccc23 | kaur 2026 — α-synuclein rotenone-PD rat in-vivo | Preclinical; indole class matches our indole-3-AA lead |
| TMAO | C[N+](C)(C)[O-] | Reference osmolyte | Dietary; generic protein stabiliser |
Method
Identical to STRC Pharmacochaperone Phase 4b Smoke Test — rigid receptor, Vina exhaustiveness 32, num_modes 9, 18 Å cubic box centered at K1141_Cα + 3 Å toward loop-1642-1651 centroid. Receptors re-used from Phase 4b (ultra × TMEM145) and Phase 4c (WT, E1659A mutant) PDBQT files — ensures exact comparability to prior gates. Ligand prep: RDKit ETKDGv3 (30 conformers) → MMFF94s minimise → obabel pH 7.4 → meeko PDBQT.
Result
| Compound | ΔG clinical | ΔG WT | ΔG E1659A | Selectivity (mut − WT) |
|---|---|---|---|---|
| 4-phenylbutyrate | −5.40 | −4.74 | −5.06 | −0.32 (mutant-preferring) |
| ip-045 | −6.08 | −6.46 | −6.19 | +0.27 (WT-preferring) |
| tmao | −2.40 | −2.53 | −2.05 | +0.48 (WT-preferring, too weak) |
Interpretation
4PBA is the first and only compound in the entire Phase 4 series with negative mut-wt selectivity. Reference values:
- Phase 4c top-5 leads (salicylic acid, nicotinic acid, cyclopropane-phenyl-COOH, indole-3-acetic acid, naphthalene-2-carboxylic): mean +0.455 WT-preferring (ALL failed selectivity gate).
- Diflunisal positive control: −0.04 flat.
- Negative controls (glucose, urea, acetamide): −0.14 near-flat.
- 4PBA: −0.32 mutant-preferring — a factor of 2.3× beyond diflunisal’s selectivity, opposite direction to all failed leads.
IP-045 is a strong binder (−6.08 on clinical construct, comparable to our top-5) but inherits the same WT-preference flaw as the indole-3-acetic-acid lead. Binding mode likely drives WT-like carboxylate-pocket engagement; adding the phenyl ester doesn’t shift selectivity. Not promising as repurposed hit.
TMAO is below the −5 kcal/mol Phase 4b binding gate — too weak to be a therapeutic pocket-binder even if osmolyte-class chaperones work via bulk stabilisation. Docking is the wrong assay for osmolytes; negative result here does not falsify their chaperone mechanism.
Caveats
- Rigid-receptor limitation persists. Phase 4c flagged that aspect (b) loop-capping fold-stabilizer mechanism is untested by rigid Vina. 4PBA’s −5.40 kcal/mol absolute affinity is modest — if its in-vivo rescue in hayase operates via a chaperone-scaffold mechanism rather than specific pocket binding, docking underestimates therapeutic potential. Phase 5 MD or MM-GBSA rescore needed before ranking this as confirmed positive.
- Single-pose best-of-9 analysis. Not ensemble-weighted. Standard for Phase 4 gates but worth noting.
- 4PBA is a small, flexible acid with modest pharmacophore presentation. The selectivity signal is real but the absolute effect size (−0.32) is within the noise floor where MM-GBSA re-scoring would significantly shift the ranking.
- Not an in-silico proof of therapeutic efficacy. 4PBA’s hayase result was oral administration → epileptic phenotype correction in KI mice — the STRC analogue is hair-cell surface trafficking, which depends on many factors docking cannot capture (ER quality control, unfolded protein response, cochlear pharmacokinetics).
What this opens
- Phase 4g-ext: expand to the broader FDA-approved chaperone tranche (TUDCA, sodium phenylbutyrate esters, glycerol phenylbutyrate (HPN-100, 4PBA prodrug), curcumin, trehalose) — at least 15-20 compounds already in clinical use with chaperone or protein-quality-control activity.
- Phase 4g-mmgbsa (recommended next step): re-score 4PBA on the E1659A mutant with the same STRC Pharmacochaperone Phase 4f Interface Rescue MM-GBSA pipeline. If MM-GBSA confirms mut-preference, 4PBA becomes a repurposing candidate that skips de-novo lead optimisation entirely.
- Phase 5 MD: prioritise 4PBA over synthetic top-5 for the first full MD run, given the FDA-approval lever on clinical translation.
Ranking delta
Hypothesis: STRC Pharmacochaperone Virtual Screen E1659A (#1, S-tier).
Move: no tier change. Stays S-tier.
Rationale: this result does NOT close Phase 4c’s FAIL on the de-novo top-5 — they still prefer WT, carboxylate sub-mechanism still flagged. But it opens a parallel path (repurposing) with a 4PBA hit that needs MM-GBSA confirmation. The evidence base strengthens marginally; the mechanism axis opens a new branch (generic chaperone vs specific pocket binder). Next step unchanged (Phase 4f MM-GBSA on Ultra-Mini × TMEM145) but adds: run 4f MM-GBSA on 4PBA × E1659A in parallel to confirm/refute the docking selectivity signal.
Other hypotheses touched: none (this is pharma-specific).
Connections
[part-of]index- STRC Pharmacochaperone Phase 4b Smoke Test
- STRC Pharmacochaperone Phase 4c WT Decoy (failure mode)
- STRC Pharmacochaperone Phase 4 Plan
[informed-by]paper 2026-04-21 hayase 4pba dscaml1 chaperone (routine-indexed)[informed-by]paper 2026-04-21 kaur ip045 synuclein chaperone (routine-indexed)[see-also]STRC Hypothesis Ranking