Journal of Physiology (1991), 434, pp. 369-398 369 With 16 figures Printed in Great Britain THE ACTIONS OF CALCIUM ON THE MECHANO-ELECTRICAL TRANSDUCER CURRENT OF TURTLE HAIR CELLS BY A. C. CRAWFORD, M. G. EVANS* AND R. FETTIPLACEt$ From the Physiological Laboratory, University of Cambridge, Cambridge CB2 3EG (Received 25 July 1990) SUMMARY 1. Mechano-electrical transducer currents evoked by deflections ofthe hair bundle were recorded in turtle isolated hair cells under whole-cell voltage clamp. The outcome of perfusing with solutions of reduced Ca2+ concentration was investigated. 2. The transducer current was roughly doubled by lowering the concentration of divalent cations from normal (2-2 mM-Mg2+, 2-8 mM-Ca2+) to 0 Mg2+, 0 5 mM-Ca2+. No significant effects on the current's kinetics or reversal potential, or on the current- displacement relationship, were noted. 3. If the Ca2+ concentration was lowered to 50 ,UM (with no Mg2+), there was about a threefold increase in the maximum current but other changes, including loss of adaptation and a decreased slope and negative shift in the current-displacement relationship, were also observed. As a result, more than half the peak transducer current became activated at the resting position of the hair bundle compared to about a tenth in the control solution. 4. The extra changes manifest during perfusion with 50 /LM-Ca2+ had also been seen when the cell was held at positive potentials near the Ca2+ equilibrium potential. This supports the view that some consequences of reduced external Ca2+ stem from a decline in its intracellular concentration. 5. With 20,tM-Ca2+, a standing inward current developed and the cell became unresponsive to mechanical stimuli, which may be explained by the transducer channels being fully activated at the resting position of the bundle. 6. The results are interpreted in terms of a dual action of Ca2+: an external block of the transducer channel which reduces the maximum current, and an intracellular effect on the position and slope of the current-displacement relationship; the latter effect can be modelled by internal Ca2+ stabilizing one of the closed states of the channel. 7. During perfusion with 1 gM-Ca2 , the holding current transiently increased but then returned to near its control level. There was a concomitant irreversible loss of sensitivity to hair bundle displacements which we suggest is due to rupture of the mechanical linkages to the transducer channel. * Present address: Department of Physiology, Medical School, University Walk, Bristol BS8 lTD. t Present address: Department of Neurophysiology, University of Wisconsin Medical School, 273 Medical Sciences Building, 1300 University Avenue, Madison, WI 53706, USA. $ Names printed in alphabetical order. MS 8682 370 A. C. CRAWFORD, M. C. EVANS AND R. FETTIPLACE 8. Following treatment with 1 #uM-Ca2", single-channel currents with an amplitude of -9 pA at -85 mV were sometimes visible in the whole-cell recording. The probability of such channels being open could be modulated by small deflections of the hair bundle which indicates that they may be the mechano-electrical transducer channels of conductance about 100 pS. 9. Open- and closed-time distributions for the channel were fitted by single exponentials, the mean open time at rest being approximately 1 ms. The mean open time was increased and the mean closed time decreased for movements of the hair bundle towards the kinocilium. 10. The Ca2+ concentration in the endolymph of the turtle's cochlear duct was measured with a Ca2+-sensitive microelectrode and found to be 65 /M. The signifi- cance of the composition of endolymph for mechano-electrical transduction by hair cells is discussed. INTRODUCTION The composition of endolymph, the fluid that bathes the transducing apparatus of hair cells in the inner ear, is unusual, since like an intracellular fluid it is rich in K+ but deficient in Na+ and other cations (Johnstone, Schmidt & Johnstone, 1963; Bosher & Warren, 1968). The divalent cation concentrations are kept particularly low, reported values being about 20 /SM for Ca2+ and 10/tM for Mg2+ (Bosher & Warren, 1978; Ikeda, Kusakari, Takasaka & Saito, 1987). The significance of the ionic composition has never been fully explained and the role ofCa21 ions is especially puzzling. It has been known for some time that in the absence of Ca2+ hair cells cease to function as mechano-electrical transducers (Sand, 1975; Jorgensen, 1983). As little as 20 /M can maintain transduction (Corey & Hudspeth, 1979; Ohmori, 1985), yet Ca2+ concentrations greater than 250 JSM diminish the flow of transducer current (Corey & Hudspeth, 1983). The mechanisms of these effects are unknown. It has recently been shown that an additional role for Ca2+ is to control adaptation of the transducer current during a maintained stimulus (Eatock, Corey & Hudspeth, 1987; Assad, Hacohen & Corey, 1989; Crawford, Evans & Fettiplace, 1989). The regulation probably occurs by Ca2+ entering the cytoplasm and resetting the range ofbundle displacements that are detected. Here we have extended these observations by recording transducer currents in isolated cells bathed in salines of reduced Ca2+ content. Most of our results are consistent with a dual action of Ca2+: a blocking of the transducer channel probably at its external surface, and an effect of intracellular Ca2+ on the gating process. An unexpected finding was that with 1 ,uM-Ca2+, almost all of the transducer channels vanished irretrievably, but occasionally a single channel remained long enough for its properties to be analysed. METHODS Preparation and recording techniques Turtles (Pseudemys scripta elegans) were decapitated and hair cells isolated from the basilar papilla by methods described fully in Art & Fettiplace (1987) and Crawford et al. (1989). Cells were plated out onto glass cover-slips coated with 2-5 mg/ml concanavalin A which immobilized the cell bodies but allowed the ciliary bundles to be manipulated. Displacements of the bundles, which had a maximum height of about 6 ,um, were produced by a glass stylus attached to a piezo-electric CALCIUM AiNTD HAIR CELL TRAiNSDUCTION3371 bimorph. The glass probe was placed near the tip of the bundle behind the tallest row of stereocilia, and if clean it adhered to the bundle and so could push or pull the bundle towards or away from the kinocilium. Since displacements of the hair bundle towards the kinocilium activate the transducer conductance (Shotwell. Jacobs & Hudspeth, 1981), these will sometimes be referred to as positive displacements, whereas bundle stimuli in the opposite direction will be denoted as TABLE 1. Composition of extracellular solutions containing various concentrations of calcium Solution NaCl KCI MgCl2 CaC12 Other A Normal 130 4 2-2 2-8 B Normal 0 Mg2+ 136 4 0 2-8 C 0 5 mM-Ca2l 136 4 0 0 5 D 50,uM-Ca2+ 136 4 0 0 05 E 20,uM-Ca2+ 136 4 0 0 F 1 /nM-Ca2+ 130 4 0 2-5 5 Na-HEDTA G Normal K+ 0 135 2 2 2-8 (5 CsCl) H K+, 0 5 mM-Caa2+ 0 135 2-2 0-5 I Endolymph 0 140 0 0 5 Concentrations are in mm. All solutions also contained 5 mM-Na-HEPES (or K-HEPES for G-I) and 4 mM-glucose. Solution A is artificial perilymph (Crawford & Fettiplace, 1980) and was usually used as the control solution. Ca2+ concentrations were measured with a calcium electrode. Solution E had no added Ca2+ but contained 14-28 ,UM, probably as contamination from NaCl. In some experiments with the K+-based solutions, a mm-CsCl was added to block the inward rectifier. negative. In some experiments where it was necessary to produce large negative displacements (i.e. away from the kinocilium) the glass probe was positioned on the kinocilial side of the bundle. A more detailed description of the stimulation equipment and its calibration is given in Crawford et al. (1989). The bundle could be stepped from one position to another in less than 100 ,us. Transducer currents were measured using the whole-cell patch-clamp technique (Hamill, Marty, Neher, Sakmann & Sigworth, 1981) with a List EPC-7 amplifier. Borosilicate recording electrodes (resistance 5-8 MQ) were usually filled with a solution containing (in mM): KCl, 125; MgCl2, 3; KHEPES, 5; K2EGTA, 5; Na2ATP, 2-5; pH adjusted to 7-2 with KOH. For experiments where the membrane potential was clamped to depolarized levels, the patch electrode was filled with a solution of composition (in mM): CsCl, 125; MgCl2, 3; NaHEPES, 5; Na2EGTA, 5; Na2ATP, 2-5; pH adjusted to 7 2 with NaOH. The time constant of the recording system, estimated from the product of the cell capacitance and the series resistance remaining after compensation, had a minimum value of 50-100 ,ts. Transducer currents were normally measured at a holding potential of -85 mV, 30-40 mV negative to where the voltage-sensitive currents are activated, so as to make it unlikely that variation in these currents, particularly the large Ca2+-activated K+ current, contributed to the effects observed. All membrane potentials were corrected for the junction potential (1-6 mV according to the solution) and for incomplete compensation of the series resistance. Experiments were performed at 18-23 'C. Solutions Experiments were normally begun in a control solution resembling perilymph which contained (in mm): NaCl, 130; KCI. 4; CaCl2, 2-8; MgCl2, 2-2; glucose, 4; NaHEPES, 5; pH adjusted to 7 6 with NaOH. The ionic environment of a cell was rapidly changed using a U-tube perfusion system (Krishtal & Pidoplichko, 1980; Fenwick, Marty & Neher, 1981). Under the best conditions the solution bathing a cell could be exchanged within a few hundred milliseconds, though positioning of the U-tube too close to a cell introduced turbulence and excess noise into the recording.