STRC Pharmacochaperone Virtual Screen E1659A

A small molecule docked into an allosteric pocket on WT STRC (3.2 Å from K1141 NZ, 9 Å from the mutation-responsive loop 1642-1651) can stabilise the WT-like fold conformation in E1659A and restore TMEM145-interface binding. Delivered as an intratympanic ear drop. VX-809 (lumacaftor) class: tertiary-fold stabiliser for a folded-but-non-binding missense mutation — the CFTR F508del-analogous opportunity for DFNB16.

The hidden assumption this breaks

Earth’s STRC-therapy toolbox is entirely protein/nucleic-acid scale: AAV, LNP, mRNA, prime editing, ASO. Nobody has asked the small-molecule question — because everyone assumes STRC mutations are “null” (true for Misha’s 98 kb deletion, false for E1659A). The computational evidence is unambiguous: E1659A protein is folded, expressed, and present on stereocilia — it just fails to bind. That is the exact class of mutation where small-molecule pharmacochaperones succeed (CFTR G551D → ivacaftor, $10 B revenue).

Mechanism

E1659A does NOT create a druggable cavity at the mutation site (pocket druggability only 0.53, volume score 0.1 — see STRC Pharmacochaperone Loop 1642-1651 Target for the refutation of the “fill the 49.8 ų cavity” framing). Instead, E→A propagates a 3 Å displacement of a proline-rich loop (L1642-P-G-G-F-G-P-G-N1651) across 10 contiguous high-pLDDT residues. K1141, the electrostatic partner of E1659 at 9.8 Å distance, physically moves 3.4 Å when E→A is introduced.

The pharmacochaperone strategy: bind to an adjacent druggable subpocket that engages K1141 NZ directly and the loop face laterally. A ligand there (a) replaces the lost E1659 carboxylate contribution via K1141 salt-bridge / H-bond, and (b) caps the loop against its LRR-face anchor in WT-like geometry. Result: folded-and-now-binding mutant.

Details of the pocket (druggability 0.86, combined V = 159 ų, docking box (7.7, −5.4, −41.5) Å): STRC Pharmacochaperone K1141 Fragment Pocket.

Pharmacophore hypothesis

Four constraints for a productive rescue ligand:

1. H-bond acceptor or carboxylate → K1141 NZ (≤3.5 Å, directional). Core salt-bridge replacement, ~3-4 kcal/mol budget.
2. Supporting H-bond donor / acceptor → D1140 or D1173 (LRR face).
3. Aromatic / hydrophobic cluster → F1646 (π-stacking with loop PHE).
4. Linker 6-10 Å between pharmacophores 1 and 3. MW 250-350, logP 2-4, HBA 3-5, HBD 0-2, rotatable bonds ≤5, TPSA 70-100 Ų (Salt & Plontke 2018 RWM PK envelope).

Candidate scaffolds for pharmacophore-guided triage before brute-force docking:

Scaffold	Example	Anchor	Rationale
Benzoic acids	probenecid, diclofenac	–COOH	FDA-approved chemistry
Aryl acetic acids	ibuprofen, naproxen	–CH₂COOH	Flexible arm, cavity-fill
Sulfonamides	furosemide	–SO₂NH	Multi-anchor
Phosphonates	alendronate	–PO₃²⁻	Strong ionic anchor
Amino-acid mimetics	L-dopa	α-COOH	Glu-surrogate logic
Indole carboxylics	indomethacin	–COOH	Rigid body + H-bond edge

DrugBank FDA-subset (~2.5 k) in one Vina run spans all six families.

Delivery: intratympanic ear drop

Small molecules (<500 Da, logP 1-4) cross the round window membrane passively. Intratympanic administration is already standard of care (dexamethasone, gentamicin for Ménière’s). Once a lead exists, formulation is routine.

Round-window PK envelope (Salt & Plontke 2018, Salt 2012, Hahn 2013):

Parameter	Favourable range
MW	200-500 Da
logP	1-4
TPSA	70-100 Ų
HBD	0-2
Formulation	hydrogel (chitosan-glycerophosphate) extends dwell time to weeks
Dose regimen	monthly in-office, pediatric-compatible

Pediatric compliance — monthly intratympanic during early intervention window is trivial.

Computational proof path

Phases 0, 1, 1B, 2B, 3A, 3B, and 3C complete. See child notes:

• STRC Pharmacochaperone Loop 1642-1651 Target — Phase 1B loop-displacement structural finding
• STRC Pharmacochaperone K1141 Fragment Pocket — Phase 2B 159 ų druggable pocket, docking box
• STRC Pharmacophore Model K1141 Pocket — Phase 3A 65-feature pharmacophore with anchor triangle + design rules
• STRC Pharmacochaperone Virtual Screen Ranked Leads — Phase 3B/3C 29-compound screen + ranked top-5 shortlist (salicylic acid, nicotinic acid, cyclopropane-phenyl-COOH, indole-3-acetic acid, naphthalene-2-carboxylic acid; diflunisal as known-TTR-pharmacochaperone positive comparator)

Remaining pipeline:

• Phase 4 — proper docking with AutoDock Vina + GNINA CNN rescore against the identified box (18 × 18 × 18 Å at (7.7, −5.4, −41.5)). Expands library to DSi-Poised (2 k) + DrugBank FDA (2.5 k) + Enamine fragments (40 k).
• Phase 5 — MD validation — GROMACS 2024 (ff19SB + GAFF2 + TIP3P, 0.15 M NaCl) 50 ns × 3 replicates on top 20. Stability: ligand RMSD <3 Å, K1141 contact >60% frames, loop-1642-1651 RMSD back <2 Å in MUT. MM-PBSA ΔG_bind ≤ −6 kcal/mol → advance.
• Phase 6 — FEP+ — OpenFE absolute FEP+. Target ΔΔG_rescue ≤ −4 kcal/mol (restores ≥50% of 8.4 kcal/mol lost binding). Off-target panel: hERG, CYP3A4, HSA.

Deliverable after Phase 3: 5-compound wet-lab shortlist (documented in STRC Pharmacochaperone Virtual Screen Ranked Leads) ready for HEK-STRC rescue assay with Jeffrey Holt.

Kill criteria and go/no-go gates

Gate	Advance if	Kill if
Phase 0	Pocket scan finds loop-facing druggable subpocket	No subpocket with druggability ≥0.4 within 15 Å of loop → pharmacochaperone hypothesis dead
Phase 3	≥20 of 4.5 k have Vina ≤ −7.5 and pharmacophore match	<5 hits → expand to ZINC22 15 M tier
Phase 4	≥3 of top 20 have stable MD pose + MM-PBSA ≤ −6	all drift / MM-PBSA > −5 → re-examine pocket definition
Phase 5	≥1 of top 5 has FEP+ ΔΔG_rescue ≤ −4	none → hypothesis fails computational test; pivot to STRC Calcium Oscillation Acoustic Therapy or STRC In Situ SpyCatcher Assembly
Wet (Phase 6)	Lead restores ≥30% TMEM145 binding in explant	<10% → hypothesis wrong, publish negative result

Phase 0 gate: PASSED — subpocket druggability 0.86 @ 9.3 Å from loop. See STRC Pharmacochaperone K1141 Fragment Pocket.

Phase 3 gate: PASSED — 10 carboxylate candidates dock with zero clashes and productive K1141-NZ / F1646 engagement; top-5 are fragment-sized, commercially cheap, clinically established. See STRC Pharmacochaperone Virtual Screen Ranked Leads.

Pharmacochaperone precedent class

• VX-770 / ivacaftor (CFTR G551D) — gating defect rescue, $10 B revenue. Closest mechanistic precedent for direct interface rescue but our class is NOT this.
• VX-809 / lumacaftor (CFTR F508del) — tertiary-fold stabiliser caps misfolding NBD1 surface. This is our actual class — caps a loop against its domain anchor.
• Migalastat (GLA α-galactosidase A) — chaperone for Fabry disease. Oral, monthly dosing. Approved 2018. MW 199 Da.
• Tafamidis (TTR stabiliser, ATTR amyloidosis) — binds a hydrophobic cleft, prevents tetramer dissociation. MW 308 Da. Approved 2011.

Four pharmacochaperones approved since 2012. Our target is structurally tractable and the clinical path is mapped by this class.

Resources and validation partners

Compute: local 32-core for Vina/Smina/GNINA; rented A100 for GROMACS (~$500-1000fullscreen);H100forFEP+($2000-5000 top 20). Software: open-source stack sufficient (Vina, Smina, GNINA, GROMACS 2024, AmberTools 24, OpenFE, RDKit, PyMOL).

Wet partners:

• Primary: Jeffrey Holt lab, Boston Children’s — OHC explants, ABR, DPOAE, AFM, stereocilin IHC. Warm from Derstroff 2026.
• Secondary: Shu lab — in-vivo STRC KO mouse, explant culture. Reviewed mini-STRC construct as “theoretically reasonable.”
• Chemistry: Enamine on-demand ($500-2000 / hit, 4-6 wk). Molport for vendor aggregation.

Budget: $3-6kcomputetotop-3in-silicohits;$5-15 k first wet data (Enamine synthesis + Holt explant assay). Total: **$10-20k,3-4months\ast \ast forpublishableresult.AAVprogramcomparator:$500 k-$2 M, 18 months.

Risks and mitigations (condensed)

Risk	Mitigation
AF3 pocket mis-modeled	Dual-scan Job 2 + Job 4 CIFs; wet validation as truth
Pocket too solvent-exposed	Accept ΔΔG_rescue ≤ −4 kcal/mol as partial rescue; 50% binding recovery may suffice clinically
Intratympanic PK too fast	Hydrogel-release formulation (chitosan-glycerophosphate) extends dwell time to weeks
Off-target (Na/K-ATPase, prestin)	Phase 5 selectivity panel; OHC cytotoxicity assay
Pediatric cochlear toxicity	ABR-monitored dose-escalation; precedent from intratympanic steroids
Hit is natural product	Acceptable for POC; repositioning path still viable
Mouse → human species gap	Target residues E1659, K1141 conserved in primates — see STRC Cross-Species Conservation Analysis

Why this is the single highest-leverage idea

1. Smallest cost: $10-20kall-invs.$500 k-$2 M for AAV.
2. Fastest proof: 3-4 months vs. 18 months.
3. Clinical path is mapped: 4 pharmacochaperones approved since 2012; intratympanic delivery is standard.
4. Pediatric-compatible: monthly ear drop, not gene therapy surgery.
5. Additive with other hypotheses: can combine with STRC Calcium Oscillation Acoustic Therapy (quantity × affinity rescue).

Files / Models

Full phase-by-phase results in child notes:

• STRC Pharmacochaperone Loop 1642-1651 Target — Phase 1B structural analysis
• STRC Pharmacochaperone K1141 Fragment Pocket — Phase 2B cavity scan + docking box
• STRC Pharmacophore Model K1141 Pocket — Phase 3A pharmacophore extraction
• STRC Pharmacochaperone Virtual Screen Ranked Leads — Phase 3B/3C screen + top-5

Scripts + raw outputs in ~/STRC/models/:

• pharmacochaperone_target_prep.py / .json — Phase 0 target definition
• pharmacochaperone_phase1b_highconf_realign.py / _results.json — Phase 1B alignment
• pharmacochaperone_phase2b_subpockets.py / _results.json — Phase 2B pocket scan
• pharmacochaperone_phase2b_top_subpocket.pdb — dummy-atom pocket for PyMOL
• pharmacochaperone_phase3a_pharmacophore.py / .json — Phase 3A pharmacophore target
• pharmacochaperone_phase3b_virtual_screen.py / .json — Phase 3B 2D screen, 29 compounds
• pharmacochaperone_phase3c_shape_fit.py / .json — Phase 3C 3D shape fit, top 10

Connections

• [see-also] STRC Pharmacochaperone Loop 1642-1651 Target — Phase 1B structural target
• [see-also] STRC Pharmacochaperone K1141 Fragment Pocket — Phase 2B druggable pocket
• [see-also] STRC Pharmacophore Model K1141 Pocket — Phase 3A pharmacophore target
• [see-also] STRC Pharmacochaperone Virtual Screen Ranked Leads — Phase 3B/3C ranked candidates
• [see-also] STRC Electrostatic Analysis E1659A — 8.4 kcal/mol binding-energy gap this rescue targets
• [see-also] STRC E1659A Conservation and Reclassification — reclassification from “null” to “misfolded fully-expressed” that makes this strategy viable
• [see-also] STRC TMEM145 GOLD Domain Interaction — atomistic interface context (ARM repeats 1603-1770)
• [see-also] STRC Calcium Oscillation Acoustic Therapy — complementary hypothesis (quantity × affinity rescue)
• [see-also] Jeffrey Holt — primary wet-validation partner
• [applies] Misha — his E1659A is the specific mutation targeted