Two Voltage-Gated Ca2+ Channels in Hair Cells of Frog Semicircular Canals

Ca ions play an essential role in hair celi functioning. Indeed, localized changes in cytosolic Ca concentration, [Caji modulate the transduction process and regulate the amplitude of the receptor potential. The latter activates in turn Ca and K basolateral conductances, which ultimately control the transmitter release. Given the relevance of Ca in canal hair cell functioning, the present work is focused on the properties of Ca currents in whole-cell recorded hair cells that were mechanically isolated from the canals. Depolarization to -20 mV activated two Ca currents (4 mM [Ca]o): one current (A) had a mean activation time constant of 0.5 ms and an amplitude of 130 pA; the other one (B) had an activation time constant smaller than or similar to that of A, peaked at 42 pA and inactivated with a time constant of 7 ms. The activation threshold was -60 mV for A and -40 mV for B. The run-down of B was faster than that of A; the protease inhibitor calpastatin prevented the run-down of both currents. The A amplitude peaked at -20 mV, with a Vrev of +40 mV, whereas the B amplitude peaked at -30 mV and became progressively smaller for larger depolarizations, being undetectable for voltages >= 0 mV. These results strongly indicate that A is generated by a non-inactivating channel population, whereas B by an inactivating one. Preliminar pharmacological experiments suggest that the non-inactivating channels are of L-type, whereas the inactivating ones are of T-type. The small value of Vrev was due to an outward Cs current flowing through the non-inactivating channels at positive potentials. In some cells the A amplitude increased almost two fold upon repeating the stimulating protocol. This increase was probably due to the activation of silent non-inactivating channels through a run-up mechanism not involving a cGMP or cAMP dependent phosphorylation. Theoretical considerations demonstrated the impossibility to attain a control, by means of the patch pipette, over the temporal and spatial changes of [Ca]i produced by the Ca inflow. Thus, it was impossible to assess the Ca-dependency of the current run-up mechanisms. However, it was possible to demonstrate that B inactivation was Ca-dependent. Indeed, the inactivation disappeared reversibly when the current was carried by Ba instead of Ca. The recovery from inactivation was investigated by using the standard two-pulse protocol, where two depolarizations to -20 mV were separated by progressively longer interpulses at different holding potentials. It resulted that the recovery from inactivation required times of the order of 100 ms at -120 mV and 300 ms at -70 mV. If both channels control transmitter release, the non-inactivating one may generate the spontaneous activity, and its modulation may affect the response to slow and prolonged stimuli, whereas the inactivating channel may be important in sustaining fast transmitter release in response to strong stimuli.