Stereocilia biomechanics — validated parameters

Source agent: Domain 3 (Sonnet 4.6), 2026-04-23. Consumer: hydrogel_phase1_self_assembly.py + h09/h26/h02 bundle-mechanics gates.

Parameter table

ParameterLiterature valueSourceConditionsStatusUsed in
Mouse OHC bundle stiffness (apical, P13-15, +HTC)5.12 ± 0.46 pN/nmdulon-2019-htc-bundle-mechanics (Sci Adv, FM-AFM)Mouse apical✅ primaryh09 Phase 1 bundle gate
Mouse OHC bundle stiffness (apical, P13-15, −HTC Strc/Tecta null)2.05 ± 0.15 pN/nmdulon-2019-htc-bundle-mechanicsMouse apical✅ primaryHTC knockout baseline
HTC contribution to bundle stiffness3.07 pN/nm (~60% of WT)dulon-2019-htc-bundle-mechanicsDerived✅ primaryh09 therapeutic target
Rat OHC bundle stiffness (1 kHz, P7-10)2.5 pN/nm2019-tobin-elife-tiplink-stiffness-gradients (Tobin eLife)Rat apical✅ primaryCF-dependent baseline
Rat OHC bundle stiffness (4 kHz, P7-10)8.6 pN/nm2019-tobin-elife-tiplink-stiffness-gradientsRat✅ primarymid-cochlea
IHC bundle stiffness (rat, 1-4 kHz)1.7 → 3.8 pN/nm2019-tobin-elife-tiplink-stiffness-gradientsRat✅ primaryreference
Tip-link tension (rat OHC 1→4 kHz)5 → 34 pN2019-tobin-elife-tiplink-stiffness-gradientsRat✅ primarygating context
HTC links per mature bundle (mouse OHC)~19dulon-2019-htc-bundle-mechanics (derived from 20-stereocilium bundle)Mouse⚠ derivedh09 HTC density
HTC per-link stiffness (derived)~0.16 pN/nm3.07 pN/nm aggregate / 19 linkscalculation⚠ derivedalternative to 7.5 in model
Aggregate HTC stiffness (Kozlov model)20 pN/nmKozlov 2011 Nature PMC3150833Bullfrog sacculus, finite-element fit✅ primary (non-mammalian)upper-bound reference
HTC developmental stiffness trajectory (P9→P15, +HTC)0.92 → 5.4 pN/nmdulon-2019-htc-bundle-mechanicsMouse✅ primary

STRC × TMEM145 Kd — NOT MEASURED ANYWHERE

SourceFinding
2026-04-17-derstroff-tmem145-ohc-stereocilia (Neuron 2026)Main text: CoIP, nanoSPD, AF3 ipTM 0.79. SI contains only Fig S11 (CoIP band quantification). NO SPR, NO BLI.
NatComms 2025 TMEM145 paper (Reimann lab, parallel group)CoIP + pull-down only. NO biophysical Kd.
verpy-2011-strc-tm-morphology (J Comp Neurol)Localization/developmental only. No stiffness, no binding data.

Verdict: no published experimental Kd for STRC × TMEM145 exists. The model’s 10 nM claim in Phase 4h competition math is unsupported. AF3 ipTM is structural plausibility, not affinity.

Red flags in current h09 models

Model constantValueProblem
K_HTC_PN_PER_NM = 7.5cites “Tobin 2019”Wrong citation. Tobin 2019 measures K_HB = K_GS + K_SP only, NO HTC term. Real per-link value is 0.16 pN/nm (Dulon derived); aggregate bullfrog model is 20 pN/nm (Kozlov). 7.5 is between but not directly from any paper.
WT_BUNDLE_STIFFNESS_PN_PER_NM = 1400Wrong by 150–560×. Real values 2.5–8.6 pN/nm (Dulon/Tobin). Adult basal extrapolation might reach ~20–80 pN/nm. 1400 likely a unit error. Also: this constant is DEAD CODE — gate3_bundle_stiffness() doesn’t reference it.
HTC_PER_STEREOCILIUM = 6Unverified. Dulon’s ~19 links for ~20 stereocilia suggests ~1 link per adjacent pair (2 per stereocilium counting both sides). “6” may count diagonal links to non-adjacent rows or use a different architectural interpretation. Tsuprun & Santi 2002 (J Histochem Cytochem) would resolve — not yet retrieved.
HTC_SPACING_NM = 8.0Unclear what it means. Inter-stereocilium center-to-center spacing is 250–450 nm. 8 nm matches a single peptide span (the connector molecule), not inter-stereocilium gap. Model comment should clarify.
STRC_NORMAL_OHC_M = 1 μM (Phase 4h)No paper quantifies STRC copy number per OHC. Flag as estimated.

Structural claim that IS backed

“≥60% bundle stiffness restoration claim for h09 is biologically coherent.” Dulon 2019 directly shows HTC removal causes ~60% bundle stiffness loss (5.12 → 2.05 pN/nm). A synthetic HTC-substitute that recovers 60% of 5.12 (≈3 pN/nm) would bring a null-STRC bundle from 2.05 → 5.05 pN/nm = essentially WT. This is the realistic therapeutic target, not the 1400 pN/nm straw figure.

Connections