STRC Research

STRC Mini-STRC Single-Vector Hypothesis

9 min read

STRC Mini-STRC Single-Vector Hypothesis

Core Idea

STRC coding sequence (5,325 bp) exceeds AAV capacity (~4,700 bp). If the N-terminal half is intrinsically disordered and functionally dispensable, removing it creates a truncated protein that fits in a single AAV.

Evidence Chain

1. N-terminal is disordered

2. Removing it improves folding

ConstructResiduespTMDisorder
Full STRC1-17750.6316%
Mini-STRC (conservative)616-17750.817%
Shorter mini-STRC700-17750.864%
C-term only1075-17750.876%

3. Truncation doesn’t break interactions

All AF3 interaction jobs: truncated constructs perform same as full-length (both negative). This is not a truncation artifact: full STRC is also negative. STRC interactions require membrane/glycosylation context.

PartnerFull STRC ipTMMini-STRC ipTMUltra-Mini ipTMVerdict
TMEM145 (full)0.470.430.43No regression; 23/41 contacts in GOLD zone
TMEM145 GOLD (pruned)——0.68Derstroff-style confirmation (published 0.91)
Homodimer0.240.200.28-0.30Ultra-Mini clearer dimer signal; 94% C2, aa 1579-1581 self-contacts
Piezo2 CED—0.30—No interaction (expected)
Otoancorin—0.29—No direct contact
Tectorin ZP—0.24—No binding
TMC1—0.20—No interaction
NFAT+Calcineurin (control)—0.73—✓ Method validated

4. It fits in AAV

See STRC AAV Vector Design for full vector layout.

  • Between ITRs: 4,103 bp / 4,700 bp = 597 bp headroom
  • With WPRE3: 4,350 bp = 350 bp headroom

5. Signal peptide solved

IgK SP validated by SignalP 6.0 with 99.97% confidence. See STRC Signal Peptide Validation.

6. GPI-anchor preserved

C-terminal intact in all constructs. Omega site candidates S1758/S1761/S1763. See STRC GPI-Anchor Analysis.

Shorter mini-STRC (residues 700-1775)

  • pTM 0.86, 4% disordered
  • 1,472 bp AAV headroom
  • Cut point at natural domain boundary (pLDDT dip at res 685-694)
  • Retains LRR domain + C-terminal functional core

Backup: C-term only (1075-1775), pTM 0.87, 2,597 bp headroom. More aggressive.

Micro-Dystrophin Precedent

Same approach used for dystrophin (DMD). Full gene 11,000 bp, micro-dystrophin ~3,600 bp. Sarepta SRP-9001 FDA-approved 2023.

Experimental Questions (lab required)

  • Does mini-STRC localize to stereocilia tips?
  • Does it form horizontal top connectors?
  • Does hair cell function recover (ABR/DPOAE)?
  • Does loss of 9/14 glycosylation sites affect trafficking?

Completed Computational Validations

  • ✅ AF3: IgK-SP + mini-STRC → pTM 0.85 (SP doesn’t break fold)
  • ✅ AF3: truncation sweep 650/680/700/720 → 700 and 720 optimal (see STRC AF3 Truncation Boundary Sweep)
  • ✅ SignalP 6.0: IgK SP 99.97% confidence
  • ✅ NetGPI 1.1: GPI-Anchored, omega S1749
  • ✅ DeepLoc 2.1: Extracellular + Lipid anchor 72.1%
  • ✅ Codon optimization v1: GC 54.5%, CAI 0.711
  • ✅ ESMFold: third independent disorder method, avg pLDDT 0.80 in ordered zone (res 750+), three-method convergence at res 691 (see STRC ESMFold Disorder Validation)
  • ✅ Conservation: BioPython pairwise alignment across 9 mammals, N-term 91.5% / C-term 92.4% — reframed as conserved spacer/stalk (see STRC Cross-Species Conservation Analysis)
  • ✅ Transduction model v2: Gamma-Poisson (negative binomial), calibrated to both Omichi 2020 data points, 2.8-4.7x advantage (see STRC Gamma-Poisson Transduction Model)

Derstroff et al. 2026 — Independent Validation (April 2026)

Paper: “TMEM145 is a principal component of outer hair cell stereocilia” (Neuron, 2026). Holt is co-author.

Key confirmation for mini-STRC:

  • STRC binds TMEM145 via C-terminal armadillo repeats (aa ~700-1780)
  • AF3 pruning: isolated GOLD domain + ARM repeats → ipTM 0.91 (high confidence)
  • N-terminal STRC (aa 1-720) shows NO interaction with TMEM145 (ipTM < 0.23)
  • This means our truncation (removing aa 1-615/700) does NOT remove the TMEM145 binding interface
  • Confirmed by coIP in HEK cells: GOLD domain deletion abolishes STRC binding

Architecture revealed:

Tubby (cytosolic C-term of TMEM145) → TMEM145 (7TM, GOST family) → STRC (extracellular GOLD domain)

Our earlier AF3 results were correct:

  • We got ipTM 0.47 (full STRC + TMEM145) and 0.43 (mini-STRC + TMEM145)
  • Low confidence was expected: AF3 can’t model membrane/glycosylation context
  • Derstroff et al. independently confirmed this and solved it with pruning + coIP

What this means:

  • ✅ Mini-STRC preserves the TMEM145 binding site
  • ✅ The disordered N-terminal we remove is NOT involved in TMEM145 binding
  • ✅ DFNB16 patients have intact TMEM145 (only STRC is mutated)
  • ✅ Single-vector mini-STRC therapy should work with endogenous TMEM145

See Derstroff et al 2026 TMEM145 Paper and 2026-04-17-derstroff-tmem145-ohc-stereocilia for full analysis.

TMEM145 reinterpretation (Apr 2026): Derstroff shows TMEM145 is an anchor, not just a binding partner. Without TMEM145, stereocilin is lost from OHCs entirely. Mini-STRC must preserve the anchoring interface specifically — not just the binding affinity. ipTM 0.43 confirms interaction; anchoring residues need explicit mapping within 700–1775.

Clinical Outcome Projections (April 2026)

Motivated by 2026-04-16-tsai-slc26a4-postnatal-gene-therapy (postnatal SLC26A4 gene therapy) and 2026-04-16-lesperance-otoferlin-gene-therapy (OTOF for ANSD).

Therapeutic window — therapeutic_window_model.py

ODE model of OHC stereocilia bundle degradation in STRC-null inner ear, calibrated to Verpy 2011 Strc-/- mouse data, scaled 10x to human.

  • Intervention deadline for functional hearing (<55 dB): ~30 months postnatal
  • Intervention deadline for responder status (≥20 dB improvement): ~30 months postnatal
  • Standard UNHS → treatment pathway: ~5 months
  • Margin: 25 months. Compatible with universal newborn hearing screening.

ABR outcome projection — abr_transfer_model.py

Log-linear model calibrated to OHC-count vs threshold correlations (Bredberg 1968; Schuknecht & Gacek 1993) and OTOF clinical trial data.

  • Mini-STRC at clinical titer (67% transduction): 33 dB predicted threshold
  • Improvement over DFNB16 baseline (85 dB): 52 dB
  • Category: Mild — conversational speech without aids
  • Minimum transduction for responder status: 13%
  • Minimum transduction for functional hearing: 24%

Combined clinical statement

Mini-STRC (Anc80L65, 3.75×10¹² GC/mL) must be delivered before ~30 months postnatal to achieve functional hearing (<55 dB). At clinical titer, 67% OHC transduction produces a predicted 33 dB ABR threshold — mild hearing loss, conversational speech without aids.

The HTC bundle-mechanics model (stereocilia_bundle_mechanics.py) gives a converging estimate (~26–29 dB) via a different physical model. See STRC Stereocilia Bundle Mechanics Model.

Recent Papers (April 2026)

2026-04-17-liang-pcdh15-cryo-em-tip-link — PCDH15 dimer architecture

PCDH15 (tip link, same mechanosensory system as STRC) forms right-handed double-helix dimers. Dimerization required for function. Raises the question: does STRC self-assemble? AF3-Multimer on STRC homodimer is now a required step before lab validation.

2026-04-17-liu-mechanoluminescent-sono-optogenetics — implant-free sono-optogenetics

Mechanoluminescent nanoparticles convert focused ultrasound → tunable visible light. Closes the light-delivery problem for sonogenetic approaches. Lateral relevance to mini-STRC (the sonogenetic and mini-STRC tracks may converge — sound-activated promoter driving mini-STRC). See Sonogenetic STRC Computational Proof.

Remaining Computational Steps

  • MD simulation for fold stability at 37C (GROMACS/OpenMM)
  • Commercial codon optimization (GenSmart/IDT) for higher CAI
  • NetNGlyc for glycosylation site confidence
  • AF3 with membrane context and glycan modifications
  • Systematic exon-deletion sweep (29 AF3 jobs) to find structurally dispensable exons
  • HADDOCK/ClusPro docking as orthogonal validation of AF3 interaction predictions
  • CpG depletion from current 156 (4.9× genome avg) — 87% reducible at 3% CAI cost. ✅ Resolved 2026-04-21 — see STRC CpG Depletion Mini-STRC. 100% CpG elimination (156 → 0) at 3.5% CAI cost (1.000 → 0.965); GC 68.6 → 63.8%. Clinical CDS = cpg_depletion_mini_strc_max.fasta. Full-vector CpG audit still pending (UTRs + ITRs).
  • AF3-Multimer STRC homodimer (new from PCDH15 finding) — if STRC self-assembles like PCDH15, oligomerization domain must be in 700–1775. Check first. ✅ Resolved 2026-04-21 — Ultra-Mini homodimer AF3 job: ipTM 0.28-0.30, 94% C2 symmetry, 98% solvent exposure, self-contacts at aa 1579-1581 in ARM deep zone. Falsifies AF3-artifact hypothesis; weak real dimerization surface centered in the Ultra-Mini zone (not in the removed LRR region). See STRC Homodimer Interface From CIF.
  • Map TMEM145 anchoring interface (new from Derstroff) — extract interface residues from AF3 model, verify all fall within 700–1775. ✅ Resolved 2026-04-20 — strc_tmem145_interface_from_cif.json: interface at aa 1603-1770, fully inside 700-1775. ✅ Extended 2026-04-21 — STRC Mini-STRC Truncation Interface Validation: C-term-only Ultra-Mini (1075-1775) also preserves binding pocket at sub-Å RMSD. Opens 2.6 kb AAV headroom.
  • New: AF3 Ultra-Mini (1075-1775) × TMEM145 — direct multimer test for C-term-only construct. If ipTM > 0.4, promote Ultra-Mini to clinical candidate (OHC-specific Prestin promoter becomes feasible). ✅ Resolved 2026-04-21 — all three gating jobs returned:
    • GOLD pruned (Ultra-Mini × TMEM145 GOLD domain): ipTM 0.68 across 5 models, PAE 2.26 Å, 21/21 residues in zone — Derstroff-style confirmation
    • Homodimer (Ultra-Mini × Ultra-Mini): ipTM 0.28-0.30, 94% C2, weak real dimerization at aa 1579-1581
    • Full-length TMEM145 (Ultra-Mini × full TMEM145): ipTM 0.43, 23/41 contacts in GOLD zone 1603-1749, four of six canonical ARM clusters reproduced; matches full-STRC precedent (0.47) — no regression from truncation. See STRC Ultra-Mini Full-Length TMEM145 AF3
    • Delivery score 4 → 5 officially passed. Ultra-Mini promoted to clinical candidate. Next step: gBlock order + pAAV B8-IgK-Ultra-Mini-WPRE3-bGH clone + Phase 4 HEK coIP.
  • AF3 with membrane context for interaction re-testing → partially answered by Derstroff et al. pruning approach

Models

  • ~/STRC/models/therapeutic_window_model.py → therapeutic_window_results.json
  • ~/STRC/models/abr_transfer_model.py → abr_transfer_results.json
  • ~/STRC/models/rescue_threshold_model.py → rescue_threshold_results.json

Website

Published on strc.egor.lol, Mini-STRC hypothesis page. 10 sections including pLDDT chart, SignalP plots, 3D viewers, AF3 results table.

Deploy: cd ~/Sites/site-strc-egor-lol && npm run build && CLOUDFLARE_API_TOKEN=... npx wrangler pages deploy dist --project-name strc-egor-lol --branch main --commit-dirty=true

Files

  • Site source: ~/Sites/site-strc-egor-lol/src/components/MiniSTRC.astro
  • AF3 results: ~/DeepResearch/strc/af3-results/ (job-a through job-h, job8-nfatc1)
  • CIF models: ~/Sites/site-strc-egor-lol/public/models/
  • Protein FASTA: /tmp/strc_constructs.fasta
  • Codon-optimized: /tmp/strc_final_optimized.fasta
  • Desktop summary: ~/Desktop/2026-03-19-*-af3-jobs.md

Connections

Graph View

Backlinks