RECORDINGS FROM CAT TRAPEZOID BODY AND HRP LABELING OF GLOBULAR BUSHY CELL AXONS

1. Recordings were made from single nerve fibers in barbiturate-anesthetized cats in the midline trapezoid body, a location that permits selective sampling of efferent cells of the ventral cochlear nucleus. Single units were localized to either the dorsal or ventral components of the trapezoid body. The fibers were physiologically classified on the basis of their peristimulus time histograms (PSTH) and receptive-field properties. In addition, low characteristic frequency (CF) units were probed for rapid rate and phase shifts with increases in intensity. The projection patterns of some fibers were traced by iontophoresing horseradish peroxidase (HRP) into their axons. 2. HRP-labeled fibers most likely originated from globular bushy cells of the ventral cochlear nucleus in that they sent a large branch into the contralateral medial nucleus of the trapezoid body which terminated in a calyceal ending and an ipsilateral branch into the lateral nucleus of the trapezoid body. A thin branch, usually starting from the large branch, wound its way through the medial nucleus of the trapezoid body to its termination in the ventral nucleus of the trapezoid body. Additional branches from the parent axon could pass through medial periolivary groups throughout the rostrocaudal extent of the superior olivary complex. The parent fiber was traced as far as the ventral lateral lemniscus where it faded before reaching its termination. 3. The majority of units were recorded in the ventral component of the trapezoid body. Although the ventral component is comprised of both large and small diameter fibers, our sample was biased to the larger diameter fibers representing the activity of axons originating from globular bushy cells in the ventral cochlear nucleus. Ventral component units were not tonotopically arrayed and had CFs that spanned the audible range for cats. HRP labeling of ventral component axons revealed that the section of the axon traveling through the midline shifted its dorsal-ventral location. This pattern was compatible with the lack of tonotopy found in the ventral component. Recordings were also made from the dorsal component of the trapezoid body, which contained medium diameter axons. These axons originated from spherical bushy cells in the ventral cochlear nucleus. Dorsal component units were tonotopically arrayed and had CFs <7 kHz. 4. Cells were characterized by their PSTH at CF. Primary-like and phase-locked units constituted most of the dorsal component units. Primary-notch, onset-L, and chopper units were found almost exclusively in the domain of the ventral component. Primary-notch and onset-L units shared the shortest first-spike latency of any PSTH group, had the smallest standard deviation of first-spike latencies, and appeared to be a continuum of the same PSTH category, differing only in sustained rate. Chopper units also had precise first-spike latencies, but their mean latencies tended to be longer than primary-notch and onset-L units. Chopper units constituted only 11% of the recordings from the ventral component and were recorded from the ventral border of the ventral component. A first-order interval analysis indicated that chopper units could be subcategorized in the same way as units recorded directly from the cochlear nucleus. Chopper units probably represent the activity of stellate cells in the cochlear nucleus, indicating that these cells may be sampled from a restricted region of the trapezoid body. 5. The receptive fields of 22 ventral component neurons were mapped by the use of pure tones and correlated with PSTH responses. Inhibitory sidebands were found in 18 of these fibers. Sideband inhibition was associated with primary-like, primary-notch, and the only chopper unit tested. None of the five dorsal component units for which receptive fields were obtained had inhibitory sidebands. 6. The spontaneous and driven rate distributions for dorsal and ventral component units were very different from those of the auditory nerve. The majority of low CF units (<6 kHz) of the ventral component had spontaneous rates <10 spikes/s; high CF units of the ventral component had a wide range of spontaneous rates, up to 110 spikes/s. Dorsal component units rarely displayed spontaneous rates <10 spikes/s. Maximum driven rates of primary-like, chopper, and primary-notch units could be very high, on the order of 500-600 spikes/s. Driven rates >450 spikes/s came only from units with CFs >8 kHz.