Pre- and post-synaptic effects of temperature at the posterior canal cytoneural junction in the isolated frog labyrinth

The effect of temperature on resting synaptic activity at the cytoneural junction of the isolated frog labyrinth was studied by intracellularly recording mEPSPs from the posterior nerve. Continuous recordings were obtained from single units during slow temperature ramps (10°C change in bath temperature) in the range from 8-30°C. Spontaneous mEPSP frequencies were generally high (>100/s) and single events extensively overlapped on one another, thus preventing direct counting. The mean rate of occurrence and size of the events were therefore estimated by means of signal processing techniques based on the estimation of mEPSP waveform from the autoregressive (minimum phase) fit to the autocorrelation of the noisy recording generated by the summated potentials. For mEPSP rates <300/s the event waveform was used to build a Wiener filter; deconvolution by this filtering procedure sharpened the events and shortened their duration, thereby eliminating extensive overlapping. For frequencies >300/s mEPSP amplitude and rate of occurrence were computed by noise analysis, from the variance and skew of membrane potential fluctuations. The frequency of mEPSP increased on increasing the temperature and vice versa (Q10 = 1.11-2.56). Statistical analyses of peak amplitudes and time intervals between successive events were performed in eight Wiener-filtered units, before and after the 10°C temperature ramp: (1) mEPSP amplitude distributions were continuous, unimodal and well fitted by single log nornal distribution functions; and (2) mEPSP time intervals were always distributed along single exponentials. Peak mEPSP amplitudes decreased on increasing the temperature, yielding log normal distributions markedly shifted to the left. The elementary mEPSP waveform was further analysed by fitting a modified gamma distribution function to the average of 20-50 unitary events with a fully resolved time course. Such fits confirmed the reliability of the autoregressive fits and yielded the following shape parameters: duration, maximum amplitude and time to the peak. All three parameters proved to decrease on increasing the temperature and vice versa. These results point to a relevant post-synaptic sensitivity to temperature. The mean rates of transmitter release also proved to be sensitive to temperature, but the basic pre-synaptic mechanisms appeared to be unaffected.