STRC AC1-CREB Monotonic Sound Response
Sustained above-threshold sound tonically phosphorylates CREB(Ser133) via AC1-stimulated cAMP → PKA → CREB, driving CRE-responsive STRC transcription with a monotonic dose-response (silence → 45 dB = 1.82× STRC protein, 48× CREB-P fold). This is the mechanism by which chronic audible sound can up-regulate Misha’s remaining STRC allele.
Why this replaced the CaMKII-CaN frequency-decoder model
Phase 1 of the hypothesis proposed Dolmetsch-style frequency decoding: sub-Hz FM → Ca²⁺ oscillations → CaMKII (fast) vs CaN (slow) competition → RBM24 phos state → STRC splicing shift. The 8-variable ODE gave silence 398 > sound 343 STRC protein — wrong direction. Root cause: at sub-µM OHC apical [Ca²⁺], CaN saturates faster than CaMKII due to k_on/k_off ratio, so CaN wins and RBM24-P drops with sound.
Dolmetsch 1998 decoder is biologically real in T-cells but does not apply to OHC apex at sub-µM [Ca²⁺]. “Published mechanism ≠ universal mechanism” — context-dependent on intracellular [Ca²⁺] range.
Pivoted to AC1-cAMP-CREB (Wu 2011 on AC1-CaM kinetics; Masada 2012 on AC1 in sensory cells; Gonzalez & Montminy 1989 on CREB-P decay). AC1 activation is strictly Ca²⁺-monotonic via Ca₄·CaM → no directionality problem.
The cascade
Ca_apex → Ca₄·CaM·AC1 → cAMP → PKA holoenzyme dissociation
→ CREB-P(Ser133) → CRE promoter → STRC_mRNA → STRC_protein
7-variable ODE (down from 8 — RBM24 layer removed since transcription, not splicing, is the regulated step). Recalibrated for dynamic range: silence [Ca²⁺] = 25 nM (well below AC1 K_Ca of 150 nM), K_PKA raised to 300 nM (physiological holoenzyme dissociation), PDE4 basal 10 nM/s.
Dose-response across SPL (20-hour steady state, constant sound)
| SPL (dB) | Ca (nM) | AC1 | cAMP (nM) | PKA | CREB-P | mRNA | Protein |
|---|---|---|---|---|---|---|---|
| 30 silence | 25 | 0.013 | 18 | 0.004 | 0.015 | 1.34 | 109 |
| 45 | 498 | 0.31 | 347 | 0.57 | 0.70 | 4.88 | 198 |
| 60 | 1034 | 0.33 | 364 | 0.60 | 0.70 | 4.89 | 198 |
| 75 | 1758 | 0.33 | 368 | 0.60 | 0.71 | 4.89 | 198 |
| 100 | 2645 | 0.33 | 369 | 0.60 | 0.71 | 4.89 | 198 |
Key findings
- Monotonic dose-response (0.5% tolerance) — directionality correct, fixes Phase 1 failure
- Silence → 45 dB: 1.82× STRC protein, 48× CREB-P fold change
- Cascade saturates above 45 dB (hormone-class signalling is supposed to saturate)
- AM modulation ≤1 Hz does NOT help: CREB cycles on ~2 min timescale (dephosphorylation t½), too slow to track sub-Hz AM envelopes
- The cochlea is a long-time-constant transcriptional integrator in this regime, not a frequency decoder
Implication — acoustic protocol simplifies
| Phase 1 claim (falsified) | Phase 1B claim (supported) |
|---|---|
| Sub-Hz FM decoded by CaMKII:CaN → RBM24 → STRC splicing | Sustained above-threshold sound → AC1-cAMP-PKA → CREB → STRC transcription |
| Carrier FM pattern is the active ingredient | LAeq dose above 45 dB is the active ingredient |
| Content matters (FM shape) | Content doesn’t matter (any broadband > 45 dB) |
| Weeks to see effect | 6-24 h pCREB; weeks for protein accumulation |
Touch Grass reframes from “FM-modulated therapy” to “chronic sub-loud acoustic dose” — ≥30 min/day LAeq 45-60 dB. NIOSH <85 dB 8-hr equivalent keeps under NIHL threshold.
Wet validation priority
- pCREB (Ser133) IHC on cochlear sections ± 60 dB broadband exposure (2 h), C57BL/6 adult mouse — predicted: 3-5× OHC nuclear pCREB signal with sound (EASY readout — 48× fold model change, strong expected signal)
- STRC mRNA qPCR from cochlear explants ± sound exposure — predicted: 2-4× mRNA increase at 6-24 h post-exposure
- Gate: both positive → chronic 4-wk exposure → ABR + STRC IHC in Rbm24 hypomorph or Strc heterozygous background
- Kill: if pCREB does not rise with sound → not AC1-CREB-mediated; pivot to direct mechano-CRE or other pathway
Files / Models
~/STRC/models/ca_oscillation_rbm24_ode.py— Phase 1 (CaMKII-CaN, falsified; retained for audit)~/STRC/models/ca_oscillation_rbm24_results.json— Phase 1 numbers (wrong direction)~/STRC/models/ca_oscillation_ac1_creb_pivot.py— Phase 1B AC1-CREB 7-var ODE~/STRC/models/ca_oscillation_ac1_creb_results.json— load-bearing numbers~/STRC/models/ca_oscillation_ac1_creb.png— 6-panel figure
Connections
[part-of]STRC Calcium Oscillation Acoustic Therapy — parent hypothesis this finding updates[see-also]STRC AC1-CREB Parameter Robustness — Phase 2 sensitivity sweep confirming this mechanism is not parameter-fragile[see-also]STRC AC1-CREB Alternative Hypothesis — pre-computation motivation for the pivot[contradicts]Sonogenetic STRC Computational Proof — the Phase 1 CaMKII-CaN frequency-decoder model inside that note is superseded by the AC1-CREB pivot here[applies]Touch Grass — sound delivery vehicle; protocol now chronic LAeq, not FM-carrier[applies]Misha — therapy target (his E1659A allele)