Antagonistic effects of perilymphatic calcium and magnesium on the activity of single cochlear afferent neurons.

The dependence of the spontaneous and sound driven activity of single cochlear nerve fibers on the calcium and magnesium content of the perilymph was studied by perfusion of the perilymphatic space. It was possible to study these effects under steady-state conditions by continuously perfusing scala tympani at low rates while simultaneously recording from units in the chinchilla auditory nerve. Preparations were stable for many hours. As previously reported [Robertson and Johnstone (1979) Pflügers Arch. 380, 7-12], perfusion with solutions containing elevated concentrations of magnesium reduces both the spontaneous and driven activity. When calcium was eliminated from the perfusate, activity was completely abolished for stimuli with sound pressure levels below 100 dB. During partial blocks, a relatively frequency-independent threshold elevation was seen for frequencies well below the characteristic frequency (CF) of the unit, with greater elevations closer to CF. When the threshold elevation at CF was 30-40 dB, the width of the 'tip' portion of the tuning curve was reduced, resembling that of naturally-occurring units with low spontaneous rates of discharge. These effects are similar to that of raising the criterion for response during threshold measurement and are probably related to a frequency-dependent nonlinearity exhibited by the motion of the basilar membrane. The dynamic range for the growth of average rate with level was increased and saturation was shifted to higher stimulus levels during elevated magnesium perfusion. Raising the calcium content of the perfusate increased both spontaneous and driven rates, even in the saturated portion of the rate-intensity plot. Under these conditions, the response of the unit may more directly correspond to the intracellular potential of the presynaptic hair cell. It is argued that the primary site of divalent cation interaction is in the control of transmitter release. Inner hair cells of the mammalian cochlea apparently do not release transmitter in the absence of a calcium influx. The size of the pool of 'readily-available' transmitter appears to be influenced by divalent cations. Even though this synapse is probably specialized for the transmission of auditory signals, the mechanism of synaptic transmission is probably not fundamentally different from that of other well-characterized synapses.