An in vitro preparation of the bullfrog sacculus was developed, in which themacular epithelium was arranged as a septum between two chambers. Hair cells arestimulated by moving the overlying otolithic membrane with a glass stimulus probe;the response is recorded as the transepithelial current with the epithelium voltage-clamped.The preparation complements the intracellular studies of individual haircells under way in this laboratory, and has the advantages of lower electrical noise,higher temporal resolution, greater stability over time, and separate control ofsolutions bathing apical and basal surfaces of the epithelium. Its major disadvantageis that stimulation and recording methods are somewhat less direct. Four projectswere carried out with this preparation:
The generation of the transepithelial "microphonic" current was analyzed interms of the passive electrical properties of the epithelium and time- and voltage-dependentconductances of the hair cell membrane. The summed receptor currentsare modified by the change in membrane potential of hair cells, by a voltage-dependentpotassium conductance, and by an adaptive shift of the responsive range ofhair bundles. A mathematical model that includes these effects predicts the observedwaveform of the microphonic current.
The ionic selectivity of the transduction channel was studied with thispreparation and with intracellular voltage-clamping of individual hair cells. Thetransduction channel is permeable to all alkali cations, to at least some of thedivalent alkaline earth metals, and to many small organic cations. The permeabilityto anions is not clear. The channel constitutes a large, water-filled pore, of at least0.65 nm diameter, which appears to contain a binding site for permeant organiccations. While a good fit to the current-voltage relation in Na+ saline can beobtained with a simple model that includes two energy barriers to permeation nearthe middle of the membrane, the distribution of barriers is not known with anyconfidence.
The hair cell response saturates if the mechanically sensitive hair bundle isdisplaced by more than about 0.4 μm. An adaptation, which follows such saturationand had been seen intracellularly, was shown to be an adaptive shift of the responsiverange such as to restore the sensitivity. The shape of the response curve is unchangedby adaptation. The adaptation constitutes a relaxation of the link between hairbundle displacement and bias on the transduction element. The position and shape ofthe response curve is changed by intracellular Ca++ and/or pH, and the rate of theshift is increased by increasing the Ca++ concentration.
Following a step displacement of the otolithic membrane, the microphoniccurrent approaches a new equilibrium value over several tens of microseconds. Thelatency of the response is less than 40 μs at 22°C. The kinetics of the approach toequilibrium are slower at lower temperatures, and depend on the position of the hairbundle. A three-state model for transduction channel gating is presented thatsupposes that changing the position of the hair bundle directly and continuouslychanges the free energies of states of the channel. The model can quantitativelypredict the observed kinetics of the current.
