Coincidence detection by binaural neurons in the chick brain stem.

1. Neurons in nucleus laminaris (NL) of birds are the first to receive binaural information and are presumed to play a role in encoding interaural time differences (ITDs). We studied extracellular single-unit responses of NL neurons in slices of the auditory brain stem of the chick. The afferents to NL were activated by electrical stimulation of nucleus magnocellularis (NM) or the auditory nerve. Changes in responses were measured as the delay between trains of bilateral stimuli (the simulated interaural time difference or S-ITD, n = 26) was varied and as the interstimulus interval and stimulus amplitude were varied (n = 61). 2. The probability of an action potential and the action-potential latency varied as a function of interstimulus interval. Most NL neurons showed a greater response probability and a shorter response latency to an interstimulus interval between 2.5 and 3.5 ms. The interstimulus interval that produced the minimum response latency was slightly longer than the interval that produced the maximum response probability. In contrast, NM neurons (n = 4) showed no preferred rate, instead, the probability of firing increased as the interstimulus interval increased. 3. Responses to bilateral stimulation showed that NL neurons can act as coincidence detectors. NL neurons responded most reliably when activated simultaneously by their two inputs and, at favorable S-ITDs, two subthreshold inputs combined to produce an action potential. 4. NL neurons also exhibited inhibition during bilateral stimulation. At unfavorable S-ITDs a subthreshold input combined with a suprathreshold input produced fewer action potentials than evoked by the suprathreshold input alone. 5. The latency of the bilateral response varied as a function of S-ITD. At S-ITDs near coincidence the latency of the bilateral response was shorter than the latency of either of the unilateral responses. Away from coincidence, the latency of the bilateral response was largely determined by the latency of the stronger unilateral response. When the unilateral responses were of similar strength, the earlier stimulus determined the latency of the bilateral response. 6. The range of S-ITDs producing a maximal response varied as a function of stimulus strength but was never less than approximately 300 microseconds. This is greater than the maximum possible ITD of sound calculated for the chick's head size. From these data we hypothesize that, in the chick, single units cannot uniquely encode ITDs, but rather ITDs may be coded by the proportion of maximally firing cells along an isofrequency band in NL.