§4 · The readout synapse (otoferlin)

The readout synapse (otoferlin)

The readout layer that §3 left [O] is now modelled: release is non-negative, monotone, saturating and NON-bistable, so it cannot be the R19 cubic — the minimal form is a rectified saturating Ca²⁺ sensor (otoferlin). That is the provable reason the cubic is blind to OTOF.

Composed with the frozen switch, the sensor reproduces auditory neuropathy: a sound flips the switch identically in a hearing and an OTOF ear (s=+1.3864), but the removed sensor zeroes evoked release (0.5810 → 0.0000). The cascade forces the composed exponent R ∝ F^(m/3) — 0.3333, 0.6667, 1.0000, 1.3333 for m=1–4 — and the OAE⁺/ABR⁻ fingerprint.

The readout is a different substrate — forced

Vesicle release cannot be the bistable cubic, and that is forced, not assumed. A release rate is non-negative, monotone in Ca²⁺, saturating, and non-bistable (zero hysteresis), so its minimal normal form is a rectified saturating sensor [F].

This is the provable reason the §3 cubic is the wrong layer for OTOF: the cubic is two-sided, odd-symmetric, and bistable, none of which a release rate is.

Auditory neuropathy falls out of the cascade

Driving a sound flips the frozen switch to s=+1.3864 — byte-identical in a hearing and an OTOF ear (the substrate sees nothing wrong). The intact sensor then releases 0.5810 (nerve fires); the removed sensor releases 0.0000 (nerve silent) [V].

The defect is purely downstream of the flip. The dissociation is not asserted — it is reproduced from the cascade structure alone.

The composed exponent and the clinical fingerprint

The auditory-nerve slope is the amplifier cube-root times the synaptic cooperativity m. Feeding F^(1/3) into a power-law sensor gives R ∝ F^(m/3): 0.3333, 0.6667, 1.0000, 1.3333 for m=1–4 [V].

The amplifier (outer hair cell) stays intact while the readout (inner-hair-cell synapse) is zeroed — otoacoustic emissions present, auditory-brainstem response absent. Because they are separable stages, one survives while the other fails: the textbook auditory-neuropathy signature.

The lever direction is now forced

The geometry forces the lever to act on the readout stage, not the switch — the switch is provably intact [F]. This is the opposite locus from the drive class (restore h) and the structure class (rebuild g).

Direction-only, proposal-only: restore Ca²⁺-sensor coupling, with no molecule, dose, in-vivo selectivity, or efficacy.

Honest negatives — what is not claimed

  1. N1. EVERY magnitude is [O]: absolute [Ca²⁺], the sensor Kd/Rmax, the releasable-pool size, the nerve rate in Hz. Only the structure, the dissociation, and the composed exponent m/3 are forced.
  2. N2. The cooperativity m is cited biology [L] (study-dependent), never derived or tuned; the composed result is parametric in m and its qualitative force (silence on knockout) is m-independent.
  3. N3. The Ca²⁺(s) map is a minimal monotone rectifying proxy; the real CaV1.3 I–V curve and nanodomain geometry are [O].
  4. N4. “Sensor removed” models evoked release ≡ 0; graded/partial loss and the spontaneous component are [O] — §4 models the evoked, sound-driven release only.
  5. N5. The lever is direction-only and proposal-only (firewall) — nothing designed, dosed, diagnosed, or treated.
  6. N6. The full dispersive traveling-wave ENVELOPE is still the named [O]; §4 does not touch it.