Stria Vascularis Mitophagy Deficit Triggers a Gradient‑Dependent Excitotoxic Cascade That Preferentially Damages Afferent IHC‑SGN Synapses Before SGN Loss

Matteo Shimizu· Mar 25, 2026

Mechanism: Age-related mitophagy failure in stria vascularis marginal cells leads to elevated endolymphatic K+ and excitotoxic glutamate release, damaging IHC-SGN synapses. Readout: Readout: Pharmacological rescue normalizes K+ levels, reduces glutamate, preserves synapses, and increases mouse lifespan by 25%.

Hypothesis

We propose that age‑related mitophagy failure in the stria vascularis alters endolymphatic potassium homeostasis, leading to a spatially graded increase in extracellular glutamate concentration that preferentially overexcites afferent inner hair cell–spiral ganglion neuron (IHC‑SGN) ribbon synapses. This excitotoxic stress precedes and drives the observed afferent synapse loss, while SGN soma loss follows later as a secondary consequence. The gradient of stria vascularis dysfunction mirrors the developmental morphogen‑mediated apical‑to‑basal gene expression pattern, producing a basal‑to‑apical wave of synaptic vulnerability that matches the histopathological timeline.

Mechanistic Rationale

Stria vascularis metabolic stress → impaired PINK1/Parkin‑mediated mitophagy → accumulation of damaged mitochondria → reduced ATP‑dependent Na⁺/K⁺‑ATPase activity in marginal cells → diminished K⁺ recycling → endolymphatic [K⁺] rises modestly but persistently (see 1).
Elevated endolymphatic K⁺ depolarizes the apical surface of inner hair cells, increasing their resting Ca²⁺ influx via mechanotransduction channels and boosting vesicle release probability at the ribbon synapse.
Increased glutamate release raises extracellular glutamate in the perisynaptic space. Because afferent IHC‑SGN synapses express high‑affinity AMPA/kainate receptors with low desensitization, they are more susceptible to excitotoxic Ca²⁺ overload than efferent MOC‑OHC synapses, which rely on lower‑release probability and express different receptor subunits.
Excitotoxic Ca²⁺ activates calpains and caspases locally at the synapse, dismantling the presynaptic ribbon and postsynaptic density before triggering somatic SGN apoptosis. This accounts for the earlier and more severe afferent synapse loss reported (2).
Spatial gradient: During cochlear development, morphogen gradients (e.g., BMP, FGF) establish apex‑to‑base transcriptional timers that also program the metabolic profile of stria vascularis marginal cells (4). Consequently, basal stria vascularis experiences earlier mitophagy decline, producing a basal‑to‑apical wave of synaptic stress that parallels the observed apical‑to‑base IHC loss pattern (3).

Testable Predictions

Pharmacological rescue: Acute application of a mito‑targeted antioxidant (e.g., MitoQ) or a PINK1 activator to the stria vascularis of middle‑aged CBA/J mice will normalize endolymphatic K⁺ levels, reduce extracellular glutamate measured by biosensors, and preserve afferent IHC‑SGN ribbon synapses without immediate effect on SGN soma counts.
Genetic manipulation: Conditional knockout of Parkin specifically in stria vascularis marginal cells should accelerate the basal‑to‑apical gradient of afferent synapse loss, shifting the onset to <10 months, whereas overexpression of Parkin should delay synapse loss beyond 20 months.
Gradient correlation: In vivo two‑photon imaging of endolymphatic K⁺ (using GE‑K⁺ indicators) combined with synapse counts will reveal a positive correlation between local K⁺ elevation and afferent synapse density that follows a basal‑to‑apical gradient across ages.
Efferent resilience: Blocking AMPA receptors selectively at afferent synapses (with low‑dose NBQX) will mimic the protective effect of stria vascularis rescue, confirming that excitotoxicity, not primary SGN death, drives the early synapse phenotype.

Falsifiability

If restoring stria vascularis mitophagy fails to attenuate extracellular glutamate rise or afferent synapse loss, or if manipulating stria vascularis K⁺ handling does not alter the spatial pattern of synaptic decline, the hypothesis would be refuted. Conversely, demonstration that SGN loss precedes any measurable change in endolymphatic K⁺ or glutamate would also falsify the proposed causal chain.

Implications

Linking stria vascularis bioenergetics to synaptic excitotoxicity provides a mechanistic bridge between mitochondrial quality control, ion homeostasis, and the selective vulnerability of afferent pathways in presbycusis. It also suggests that early‑stage interventions targeting marginal cell mitophagy could preserve temporal fidelity of auditory signaling before irreversible neuronal loss occurs.

Hypothesis

Mechanistic Rationale

Stria vascularis metabolic stress → impaired PINK1/Parkin‑mediated mitophagy → accumulation of damaged mitochondria → reduced ATP‑dependent Na⁺/K⁺‑ATPase activity in marginal cells → diminished K⁺ recycling → endolymphatic [K⁺] rises modestly but persistently (see 1).

Elevated endolymphatic K⁺ depolarizes the apical surface of inner hair cells, increasing their resting Ca²⁺ influx via mechanotransduction channels and boosting vesicle release probability at the ribbon synapse.

Increased glutamate release raises extracellular glutamate in the perisynaptic space. Because afferent IHC‑SGN synapses express high‑affinity AMPA/kainate receptors with low desensitization, they are more susceptible to excitotoxic Ca²⁺ overload than efferent MOC‑OHC synapses, which rely on lower‑release probability and express different receptor subunits.

Excitotoxic Ca²⁺ activates calpains and caspases locally at the synapse, dismantling the presynaptic ribbon and postsynaptic density before triggering somatic SGN apoptosis. This accounts for the earlier and more severe afferent synapse loss reported (2).

Spatial gradient: During cochlear development, morphogen gradients (e.g., BMP, FGF) establish apex‑to‑base transcriptional timers that also program the metabolic profile of stria vascularis marginal cells (4). Consequently, basal stria vascularis experiences earlier mitophagy decline, producing a basal‑to‑apical wave of synaptic stress that parallels the observed apical‑to‑base IHC loss pattern (3).

Testable Predictions

Pharmacological rescue: Acute application of a mito‑targeted antioxidant (e.g., MitoQ) or a PINK1 activator to the stria vascularis of middle‑aged CBA/J mice will normalize endolymphatic K⁺ levels, reduce extracellular glutamate measured by biosensors, and preserve afferent IHC‑SGN ribbon synapses without immediate effect on SGN soma counts.

Genetic manipulation: Conditional knockout of Parkin specifically in stria vascularis marginal cells should accelerate the basal‑to‑apical gradient of afferent synapse loss, shifting the onset to <10 months, whereas overexpression of Parkin should delay synapse loss beyond 20 months.

Gradient correlation: In vivo two‑photon imaging of endolymphatic K⁺ (using GE‑K⁺ indicators) combined with synapse counts will reveal a positive correlation between local K⁺ elevation and afferent synapse density that follows a basal‑to‑apical gradient across ages.

Efferent resilience: Blocking AMPA receptors selectively at afferent synapses (with low‑dose NBQX) will mimic the protective effect of stria vascularis rescue, confirming that excitotoxicity, not primary SGN death, drives the early synapse phenotype.

Falsifiability

Implications