Electrophysiological evidence for potassium accumulation between type I hair cells and calyx terminal in mammalian crista ampullaris

Two types of hair cells are present in mammalian vestibular sensory epithelia, called Type I and Type II hair cells, which differ in electrophysiological properties and innervation. Type II hair cells are contacted by several bouton nerve terminals, while Type I hair cells are contacted by a calyx nerve terminal that envelopes the entire basolateral membrane. Only Type I hair cells, moreover, express a low-voltage activated outward K+ current, called IK,L, which confers upon them a much lower input resistance at rest compared to Type II hair cells. As a consequence, in Type I hair cells large transducer currents would be necessary to change the cell membrane potential and to depolarize the cell enough to activate voltage-gated Ca2+ channels and related neurotransmitter release. How the calyx synapse operates remains in fact enigmatic. It has been speculated that K+ accumulation in the synaptic cleft cooperates with conventional (vesicular) synaptic transmission in sustaining afferent transmission by Type I hair cells. By combining the patch-clamp whole-cell configuration with the whole crista preparation, we have recorded the current and voltage responses of mouse semicircular canal Type I and Type II hair cells in situ. Depolarizing voltage steps elicited in Type II hair cells a large outward K+ current characterized by a substantial time-dependent inactivation, while the same voltage-protocol elicited in most Type I hair cells a large and sustained outward K+ current. However, in a notable percentage (51%) of Type I hair cells investigated, the outward K+ current showed a substantial time-dependent relaxation. In these cells, moreover, upon repolarization to –40 mV the instantaneous current was inward, reversing to outward slowly with time. A reasonable explanation for the above results is that during large outward K+ currents, K+ accumulates around Type I hair cells, thus shifting the K+ reversal potential (VrevK+) toward more positive values. The rightward shift of VrevK+ would produce both the outward current relaxation during depolarizing voltage steps and the instantaneous inward current upon repolarization at –40 mV. Since we never observed such effects when recording from Type II hair cells, we hypothesized that the presence of a residual nerve calyx was responsible for K+ accumulation around Type I hair cells. We also found that by using voltage protocols that increased extracellular K+ accumulation, IK,L deactivation was slowed down. Similar results, i.e. VrevK+ rightward shift and IK,L deactivation slowdown, were obtained by local perfusion of the preparation with an extracellular solution enriched in K+, thus corroborating our hypothesis about K+ accumulation. In conclusion, our results provide electrophysiological evidence for an increased K+ concentration in the synaptic cleft between Type I hair cell and its calyx ending during outward K+ current activation. The resulting depolarization might be aimed at reinforcing and prolonging Ca2+ channels activation and thus afferent transmission during slow head movements detected by vestibular organs.