STRC × TMEM145 interactions — literature status

Source agents: Domain 3 (Sonnet 4.6), cross-checked against Domain 1/2, 2026-04-23. Consumer: hydrogel_phase4e_cochlear_pkpd.py (KD_TMEM145_M), hydrogel_phase4h_endogenous_strc_competition.py (competition math), h03 AAV dosing, h26 avidity design.

What is known

Evidence classSourceFinding
Coimmunoprecipitation2026-04-17-derstroff-tmem145-ohc-stereocilia Fig S11STRC × TMEM145 interact in native OHC (CoIP positive)
Pull-downSame + NatComms 2025 (Reimann lab)Confirmed pair
nanoSPD localizationDerstroff 2026TMEM145 and STRC co-localize at OHC bundle
AF3 ipTMDerstroff 2026 + internal Phase 3 (h09)0.71–0.79 (structural plausibility, not affinity)
GeneticVerpy 2008 Nature + DFNB16 clinicalSTRC loss → OHC bundle disruption (functional interaction)

What is NOT known

There is zero published SPR, BLI, ITC, or fluorescence measurement of STRC × TMEM145 binding affinity.

  • Derstroff 2026 SI → only Fig S11 (CoIP band quantification). Nothing quantitative.
  • NatComms 2025 parallel paper → CoIP and pull-down only.
  • Verpy 2011 → no binding data.
  • No other paper in literature reports biophysical Kd.

Red flag in current h09 Phase 4 models

Model constantValueProblem
KD_TMEM145_M = 100e-9 (Phase 4e)100 nMDerived from AF3 ipTM 0.68 via heuristic mapping. AF3 ipTM is NOT calibrated to affinity.
WT STRC × TMEM145 Kd (Phase 4h assumption)10 nMCompletely unsourced. No SPR or BLI value exists in the literature. The entire E1659A competition scan in Phase 4h is anchored to this number.

Implication

All dose-occupancy calculations in Phase 4e and all competition scans in Phase 4h inherit this uncertainty. Sensitivity analysis should explicitly include Kd scenarios ranging from 1 nM (strong) to 10 μM (weak) — current Phase 4h already spans this but the WT STRC baseline is fixed at 10 nM without justification.

Actionable next steps

  1. Request SPR/BLI on STRC × TMEM145 from the Holt, Oliver, or Reimann labs — all three have the reagents and could plausibly add the measurement to a revision. This is the single highest-leverage experiment to validate h03, h09, and h26 at once.
  2. AF3 ipTM → Kd calibration: run AF3 on pairs with known experimental Kd (positive controls from SKEMPI or PDBbind) to derive an empirical mapping, then reapply to STRC × TMEM145. This is a computational stopgap.
  3. Include Kd sensitivity in all dose calculations: no single-value claims; report dose × (Kd ∈ {1, 10, 100, 1000, 10000 nM}) matrix.

Cross-hypothesis impact

This gap is not h09-specific:

  • h03 (Mini-STRC AAV): transgene product needs to engage TMEM145. Dose-expression model depends on Kd.
  • h26 (Engineered Homodimer): avidity design explicitly targets increasing effective Kd against TMEM145. Baseline reference needed.
  • h09 (Hydrogel): every Phase 4 number with “occupancy” in it depends on this.

Connections