Receptor potentials of lizard cochlear hair cells with free‐standing stereocilia in response to tones.

Publisher Website
Google Scholar
Abstract
Intracellular potentials were recorded with micropipettes from hair cells with free-standing stereocilia in the cochleae of anesthetized alligator lizards. Wave forms of intracellular responses to click stimuli were classified into 3 types: hair cells, supporting cells, and untuned cells. The responses of hair cells to tonal stimuli was primarily studied. For most frequencies, f, and levels, P, of tone-burst stimuli, the response envelope of the receptor potential increases monotonically at the tone-burst onset, and decreases monotonically at tone-burst offset. Overshoot in the envelope of the response at the onset and offset of tone bursts is observed only for tone bursts of low f, high P, and short (.apprxeq. 1 ms) rise/fall time. The steady-state response to tones consists of a positive (depolarizing) DC component, V0, plus AC components, (e.g., a fundamental component, V1, 2nd harmonic, V2, and 3rd harmonic, V3). The magnitudes of AC and DC components are functions of f and P, and show 3 characteristics: frequency selectivity, nonlinearity and low pass filtering. The receptor potential is frequency selective. The frequency selectivity of V0 and V1 components was measured by means of isovoltage (iso-V0 and iso-V1) contours. Iso-V0 and iso-V1 contours are V-shaped: the maximum sensitivity occurs at a characteristic frequency (cf). The shapes of these contours near the cf depend on the values of V0 and V1 at which the contours were measured and are sharper for lower values of V0 and V1. The mean slopes of the low and high frequency sides of these contours are: -45.0 and +85.1 dB/decade for iso-V0 contours (n = 26), and -33.6 and + 103.8 dB/decade for iso-V1 contours (n = 28). The receptor potential has non-linear properties. The magnitudes and phase angles of V0, V1, V2 and V3 receptor-potential components were measured as a function of P for different f. The slopes of level functions (the dependence of log V0 and log /V1/ on log P) were measured at low levels for different f. For values of f differing from cf by more than half-octave, the slope for V0 is between 1 and 2 with a mean of 1.3; the slope for V1 is .apprx. 1, i.e., /V1/ increases approximately linearly with P. For frequencies near cf, the slopes for V0 and V1 are .apprx. 0.8 and 0.5, respectively, indicating the presence of a compressive nonlinearity. An effect of this frequency-dependent nonlinearity is to sharpen the frequency selectivity at lower response magnitudes for f near cf. The receptor potential shows low pass filtering. Whereas the maximum response magnitude of V0, (i.e., the saturation value measured at high P) does not depend on f, the maximum /V1/ decreases by .apprx. 20 dB/decade with increasing f. Values of /V1/ measured at a constant value of V0 also decline by .apprx. 20 dB/decade. Many of the results are consistent with the predictions of a 3-stage model of the generation of the receptor potential. The 1st stage, which represents the mechanical properties of the middle and inner ear, is a linear, time-invariant, band-pass filter; the 2nd stage, which represents mechanoelectric transduction in hair cells, is a zero-memory nonlinearity; the 3rd stage, which represents the electrical properties of hair cells, is a linear, time-invariant, low-pass filter. The results differ from the model predictions for low-level acoustic stimuli near cf, and indicate the presence of a frequency-dependent compressive nonlinear mechanism.