Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Synaptic Inputs to Stellate Cells in the Ventral Cochlear Nucleus MICHAEL J. FERRAGAMO, NACE L. GOLDING, AND DONATA OERTEL Department of Neurophysiology, University of Wisconsin Medical School, Madison, Wisconsin 53706-1532 INTRODUCTION Auditory information is carried from the cochlea to the cochlear nuclear complex of the brain stem by auditory nerve fibers. In terminating on at least six types of principal cells, auditory nerve fibers feed information to at least six parallel ascending pathways in mammals. One pathway is mediated by T stellate cells of the ventral cochlear nucleus (VCN), stellate cells named for the trajectory of their axons through the trapezoid body (Oertel et al. 1990). T Stellate cells project to the contralateral inferior colliculus in mice and cats (mice: Oertel et al. 1990; Ryugo et al. 1981; cats: Adams 1979, 1983; Cant 1982; Oliver 1987; Osen 1972; Roth et al. 1978). Even the earliest studies revealed that stellate cells were of multiple types. Stellate (multipolar) cells were recognized by Osen (1969) to be of multiple classes based on somatic size. In contrast to the T stellate cells, the axons of which leave the VCN through the trapezoid body, the axons of D stellate cells have a dorsalward trajectory toward the intermediate acoustic stria. Stellate cells also differ in the synaptic density distributions (Cant 1981), dendritic architecture (Brawer et al. 1974; Oertel et al. 1990; Tolbert and Morest 1982), axonal destinations (Oertel et al. 1990; Smith and Rhode 1989), terminal vesicle shape (Smith and Rhode 1989), immunoreactivity (Adams and Mugnaini 1987; Wenthold et al. 1987), responses to sounds in vivo (Smith and Rhode 1989), and shock evoked synaptic responses in vitro (Oertel et al. 1990). T stellate cells have both their dendrites and terminal arbors aligned with isofrequency laminae in the tonotopically arranged caudal anterior VCN (AVCN) and posterior VCN (PVCN) (Oertel et al. 1990). Recordings have not been made from single, identified units in mice, but the arrangement of dendrites with respect to the tonotopy of the VCN suggests that T stellate cells are tuned narrowly. In cats, stellate cells that project through the trapezoid body respond to tones with steady firing as ‘‘choppers’’ (Smith and Rhode 1989). Choppers in cats, rats, and chinchillas are tuned sharply and respond to tones near the best frequency tonically at a steady rate (Rhode and Smith 1986; Wickesberg 1996). Inhibitory sidebands flank the excitatory response area. Although an initial spike signals the onset of the tone with temporal precision, the timing of subsequent spikes is independent of the phase of the sound. The ultrastructure of their terminals suggests that these stellate cells are excitatory (Smith and Rhode 1989). D stellate cells have long, sparsely branched dendrites that are spread across isofrequency lamina and terminate widely in the VCN with local collaterals (Oertel et al. 1990). The spread of dendrites and the properties of corresponding cells in cats indicate that D stellate cells are likely to be broadly tuned. Similar cells in cats and guinea pigs respond to sound as ‘‘onset-choppers’’, are inhibitory and glycinergic, and project to the contralateral cochlear nucleus (cats: Cant and Gaston 1982; Smith and Rhode 1989; guinea pigs: Schofield and Cant 1996; Wenthold 1987). As D stellate cells are inhibitory and glycinergic, they are labeled immunocytochemically with antibodies to glycine conjugates (Oertel and 0022-3077/98 $5.00 Copyright q 1998 The American Physiological Society 9k23 / 9k22$$de05 J294-7 12-08-97 08:35:37 neupa LP-Neurophys 51 Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 Ferragamo, Michael J., Nace L Golding, and Donata Oertel. Synaptic inputs to stellate cells in the ventral cochlear nucleus. J. Neurophysiol. 79: 51–63, 1998. Auditory information is carried from the cochlear nuclei to the inferior colliculi through six parallel ascending pathways, one of which is through stellate cells of the ventral cochlear nuclei (VCN) through the trapezoid body. To characterize and identify the synaptic influences on T stellate cells, intracellular recordings were made from anatomically identified stellate cells in parasagittal slices of murine cochlear nuclei. Shocks to the auditory nerve consistently evoked five types of synaptic responses in T stellate cells, which reflect sources intrinsic to the cochlear nuclear complex. 1) Monosynaptic excitatory postsynaptic potentials (EPSPs) that were blocked by 6,7-dinitroquinoxaline2,3-dione (DNQX), an antagonist of a-amino-3-hydroxy-5methyl-4-isoxazolepropionic acid receptors, probably reflected activation by auditory nerve fibers. Electrophysiological estimates indicate that about five auditory nerve fibers converge on one T stellate cell. 2) Disynaptic, glycinergic inhibitory postsynaptic potentials (IPSPs) arise through inhibitory interneurons in the VCN or in the dorsal cochlear nucleus (DCN). 3) Slow depolarizations, the source of which has not been identified, that lasted between 0.2 and 1 s and were blocked by DL-2-amino-5-phosphonovaleric acid (APV), the N-methyl-D-aspartate (NMDA) receptor antagonist. 4) Rapid, late glutamatergic EPSPs are polysynaptic and may arise from other T stellate cells. 5) Trains of late glycinergic IPSPs after single or repetitive shocks match the responses of D stellate cells, showing that D stellate cells are one source of glycinergic inhibition to T stellate cells. The source of late, polysynaptic EPSPs and IPSPs was assessed electrophysiologically and pharmacologically. Late synaptic responses in T stellate cells were enhanced by repetitive stimulation, indicating that the interneurons from which they arose should fire trains of action potentials in responses to trains of shocks. Late EPSPs and late IPSPs were blocked by APV and enhanced by the removal of Mg 2/ , indicating that the interneurons were driven at least in part through NMDA receptors. Bicuculline, a g-aminobutyric acid-A (GABAA ) receptor antagonist, enhanced the late PSPs, indicating that GABAergic inhibition suppresses both the glycinergic interneurons responsible for the trains of IPSPs in T-stellate cells and the interneuron responsible for late EPSPs in T stellate cells. The glycinergic interneurons that mediate the series of IPSPs are intrinsic to the ventral cochlear nucleus because long series of IPSPs were recorded from T stellate cells in slices in which the DCN was removed. These experiments indicate that T stellate cells are a potential source of late EPSPs and that D stellate cells are a potential source for trains of late IPSPs. 52 M. J. FERRAGAMO, N. L. GOLDING, AND D. OERTEL METHODS Tissue preparation The cochlear nuclei were obtained from 18- to 26-day–old CBA or ICR mice as described previously (Golding and Oertel 1996; Zhang and Oertel 1993). The dissection was performed in carbogen-infused saline of the following composition (in mM): 130 NaCl, 3 KCl, 1.2 K2HPO4 , 2.4 CaCl2 .H2O, 1.3 MgSO4 , 3 N-2hydroxyethylpiperazine-N *-2-ethanesulfonic acid, 20 NaHCO3 , and 10 glucose, pH 7.4, 317C. With a single parasagittal cut, the cochlear nuclei were removed from the brain stem with a tissue slicer (Frederick Haer, New Brunswick, ME) in a slice that was between 250 and 400 mm at its thickest point. The slice was immersed in oxygenated saline in a tissue chamber with a volume of 0.3 ml and continuously perfused at a rate of 10–12 ml/min (Oertel 1985). The temperature of the bath was maintained at 347C with a thermoregulator (UW-Madison Medical Electronic Shop) with feedback supplied by a temperature probe in the chamber (Physitemp, Clifton, NJ). The slice was allowed to ‘‘rest’’ for 60–90 min before recording. Pharmacological agents were dissolved in normal saline or saline in which MgSO4 was replaced with CaCl2 (0 Mg 2/ ) and introduced into the chamber without a break in the flow. Strychnine, picrotoxin bicuculline methiodide, DL-2-amino-5-phosphonovaleric acid (APV), and 6,7-dinitroquinoxaline-2,3-dione (DNQX) were obtained from Sigma (St. Louis, MO) and added to normal saline. Electrophysiological recording Recording electrodes were constructed of 1-mm–diam omega dot tubing (WPI) pulled (Sutter Instruments, San Francisco, CA) to impedances of 120–250 MV and filled with 1% biocytin (Sigma) in 2 M K / -acetate, pH 7.0. Intracellular potentials were amplified, low-pass filtered at 10 kHz (ICX2-700; Dagan, Minneapolis, MN), continually monitored audiovisually, and recorded on chart paper (Gould, Valley View, OH). Data acquisition, current injection, and shock triggering were all performed by a Digidata 1200A computer interface under control of pClamp software (Axon Instruments, Foster City, CA) in an IBM-compatible computer (Micron, Nampa, ID). All responses to current injection and synaptic responses °600-ms duration were sampled digitally at 25 kHz; synaptic responses exceeding 600 ms were sampled at 10 kHz. Shocks were delivered to the severed eighth nerve through a stimulating electrode constructed from an adjacent pair of insulated tungsten electrodes (Bak Electronics, Rockville, MD), each with 9k23 / 9k22$$de05 J294-7 a 50-mm exposed tip. Stimulation voltage (0.1–100 V; 100-ms duration) was produced by an isolated DC source (S-100; Winston Electronics, Millbrae, CA) under control of a digitally triggered timer (A-65; Winston Electronics). Analysis The slope was measured of the linear portion of the rise of the excitatory postsynaptic potential (EPSP) from rest. When an EPSP was not detectable measurements of amplitude and slope were performed during a 0.5-ms window starting at the point when the resting potential was restored after the shock. K-means cluster analysis was performed by Statistica (Statistica, Rockville, MD). Histology Anatomic labeling was accomplished by iontophoretic injection of biocytin with depolarizing current steps (0.5–2 nA; 150–200 ms) at a rate of 2 Hz for roughly 2 min. At the termination of the experiment, the slice was fixed in 4% paraformaldehyde, 0.1 M phosphate buffer, pH 7.4, and stored at 47C for 24 h to 2 wk. For histological reconstruction, the tissue was embedded in a mixture of gelatin and albumin cross-linked with glutaraldehyde and sectioned at 60 mm on a vibratome. Sections were reacted with avidin conjugated to horseradish peroxidase (Vector ABC kit, Vector Laboratories, Burlingame, CA) and processed for horseradish peroxidase with Co 2/ and Ni 2/ intensification. Sections mounted on coated slides were conterstained with cresyl violet. RESULTS The present experiments are based on recordings from 54 T stellate cells and 3 D stellate cells. All were labeled with biocytin and identified morphologically according to the criteria of Oertel et al. (1990). The resting potentials of T stellate cells ranged from 051 to 068 mV [mean Å 056.7 { 4.9 (SD)]. Input resistance, measured from the magnitude of the response to injection of 00.1 nA current, ranged from 44 to 151 MV (mean Å 89.4 { 24.4). The average resting potential and input resistance for D stellate cells were 057 { 4.2 mV and 96.2 { 27.8 MV, respectively. T stellate cells The cell bodies of T stellate cells make up a large proportion of the large cells of the PVCN. Anatomical reconstructions of T stellate cells in this study resembled those reported previously (Oertel et al. 1990). Many lay medially, well away from the superficial granule cell domain on the lateral surface of the VCN. The dendrites of T stellate cells commonly were arranged parallel to the projection of auditory nerve fibers, putting the dendrites into the path of relatively few of the tonotopically arranged auditory nerve terminals, and ended in characteristic tufts that often came near but did not intermingle with the superficial granule cells. The axon of each cell was cut as it exited the VCN medially and entered the trapezoid body. All T stellate cells had local axonal collaterals restricted to roughly the same isofrequency lamina as its parent soma and dendrites. Because the region is populated primarily by other T stellate cells, it is likely that T stellate cells contact other T stellate cells. Many T stellate cells also had a collateral that projected to the fusiform cell layer of the dorsal cochlear nucleus (DCN). This projection maintained tonotopy and occupied a nar- 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 Wickesberg 1993; Wenthold 1987; Wickesberg et al. 1994). Such labeling stains õ0.1 of cells in the dorsocaudal AVCN and PVCN. In cats, too, corresponding cells represent a small fraction of stellate cells (Cant 1981). Although it is clear that T stellate cells transform the ‘‘primary-like’’ firing patterns of auditory nerve fibers to chopper patterns that are different, it is not at all clear what the brain accomplishes in making that transformation. To gain a better understanding of the integrative processes that contribute to this pathway, we have examined functionally and in detail the synaptic inputs to T stellate cells. We show that T stellate cells are subject to synaptically mediated feedforward excitation and inhibition that is under GABAA ergic control and the action of which continues over a time course one order of magnitude greater than that previously reported (Oertel et al. 1990). We suggest that T stellate cells integrate input from the auditory nerve with input from other T and D stellate cells. SYNAPTIC INPUTS TO VCN MULTIPOLAR CELLS 53 rowband (50–75 mm) spanning the fusiform cell layer from its most rostral to its most caudal end. Spontaneous EPSPs were recorded in 50 T stellate cells, in 9 of which spontaneous inhibitory postsynaptic potentials (IPSPs) also were detected. Spontaneous PSPs occurred infrequently and singly (not in bursts). The synaptic responses of T stellate cells to shocks of the auditory nerve have five components: monosynaptic EPSP, disynaptic IPSP, long, slow depolarization, late EPSPs, and trains of late IPSPs. The first two of these components are highly consistent. They have been described previously and therefore will be discussed only briefly (Oertel 1983; Oertel et al. 1990; Wickesberg and Oertel 1990; Wu and Oertel 1984, 1986). The later three components are more variable and have not been described before. Shocks to the auditory nerve evoked EPSPs the amplitude of which increased monotonically with the strength of the shock. The delay between beginning of the shock and the rise of the EPSPs was constant except in responses to the weakest shocks where delays were a little longer. The minimum latencies ranged between 0.48 and 0.92 ms (mean Å 0.70 { 0.13, n Å 53). These latencies indicate that the EPSPs were monosynaptic and therefore that they reflect direct input from the auditory nerve. A consistent feature of T stellate cells was that the EPSP grows stepwise with increasing shock strength, indicating that relatively few inputs converge on one cell. A superposition of traces selected at every other stimulation voltage reveals groups of similarly shaped EPSPs (Fig. 1). The accompanying plots of peak amplitudes also reveal the stepwise growth of EPSPs, indicating that relatively few inputs are recruited sequentially. The total number of steps, however, is obscured by the presence of an action potential in suprathreshold responses. The stepwise increases in amplitude were accompanied by stepwise increases in the slope of the rise of the EPSPs, revealing in addition recruitment of inputs in suprathreshold responses. Because it is possible that several auditory nerve fibers have similar thresholds, that contributions are too small to be resolved clearly and because it is conceivable that auditory nerve inputs might have been damaged in the preparation of slices, the number of resolved steps reflects a minimum number of auditory nerve inputs. The dotted lines in each plot indicate the mean of each cluster, determined by K-means cluster analysis. Analysis of variance was performed on the resultant clusters; these were found in each case to differ significantly (P õ 0.001). If each cluster with a mean greater than Ç0 represents a separate input, then the clusters in the plots of slope in Fig. 1, A and B, indicate that these cells received five and four inputs from the auditory nerve, respectively. Similar experiments were conducted on two additional cells in which six and five steps were resolved, contributing to an average of about five ( {0.8, n Å 4) resolvable steps. Long, slow depolarization A slow, long-lasting depolarization followed the early responses in 43 T stellate cells. The depolarization was generally observed only in responses to strong shocks. It followed 9k23 / 9k22$$de05 J294-7 FIG . 1. Number of converging auditory nerve fibers was estimated from the number of steps in the growth of responses with increasing shock strength for 2 cells. A and B, left: first response at every other stimulation voltage was selected and superimposed to demonstrate that there are distinct groups of similarly shaped excitatory postsynaptic potentials (EPSPs). Right: scatter plots of the maximal amplitude of each EPSP and of the initial slope of both the subthreshold (solid circle) and suprathreshold (open circle) responses. Dashed line, mean of each cluster determined by K-means cluster analysis. Total number of sub- and suprathreshold steps for the cells plotted in A and B are 5 and 4, respectively. Two (A) and 3 repetitions (B) were performed at each stimulation voltage. All were used in the statistical analysis: A slope, 6 clusters, F(5,58) Å 1,293.3, P õ 0.001; amplitude, 4 clusters, F(3,43) Å 335.9, P õ 0.001; B slope, 5 clusters, F(4,52) Å 1,378.2, P õ 0.001; amplitude, 4 clusters, F(3,38) Å 354.6, P õ 0.001. The bath contained 1 mM strychnine to avoid distortion of EPSPs by IPSPs. The cell in B was hyperpolarized with a 00.1-nA current pulse during synaptic stimulation to enhance resolution of subthreshold events. suprathreshold EPSPs and early IPSPs, becoming evident as the cell repolarized after the combination of the undershoot of the action potential and the early IPSP and lasting between 100 and 500 ms (Fig. 2A). The depolarizing hump could be suprathreshold, causing the T stellate to fire late action potentials in response to the shock. The appearance of the slow, long depolarization was variable in responses to single shocks, being more prominent in some responses than others in a single cell and being more consistent in some cells than others. Consistently, however, repetitive stimulation promoted its appearance. Figure 2B illustrates the responses of a T stellate cell the response of which to a single 5-V shock did not have a detectable long depolarizing hump. Repetitive stimulation revealed the presence of the slow depolarization and showed that its amplitude and duration increased with the rate of repetitive stimulation. Twenty shocks at 200/s evoked a long, suprathreshold, depolarization as well as late IPSPs. 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 Monosynaptic EPSPs reflect the convergence of few inputs from the auditory nerve 54 M. J. FERRAGAMO, N. L. GOLDING, AND D. OERTEL Late EPSPs reveal the existence of excitatory interneurons A single shock commonly evoked occasional unitary EPSPs °500 ms after a shock to the auditory nerve. Figure 3 ( ● ) illustrates examples of late EPSPs recorded in six T stellate cells. Even relatively weak, subthreshold shocks could activate late EPSPs (Fig. 3A). Late EPSPs occurred between tens and hundreds of milliseconds in 44 of 54 T stellate cells; in some cells, the presence of IPSPs obscured late EPSPs. The long latency of late EPSPs indicates that they are polysynaptic and, therefore, that there exist excitatory interneurons that contact T stellate cells in slices of the cochlear nuclei. Responses such as those illustrated in Fig. 3A show that these excitatory interneurons are activated by shocks to the auditory nerve. The existence of excitatory interneurons was confirmed FIG . 4. An inhibitory postsynaptic potential (IPSP) with a threshold lower than that of the monosynaptic EPSP was observed consistently after auditory nerve stimulation. The 1.5- to 2-ms latency indicated that the IPSP was disynaptic. with another series of experiments. The finding that excitatory interneurons can be activated polysynaptically with shocks indicates that their dendrites lie in the slice and suggests that they might be activated chemically through excitatory synaptic receptors. Figure 3B shows that the application of glutamate does indeed increase the frequency of EPSPs. This experiment was done in the presence of strychnine and bicuculline to eliminate inhibition. Disynaptic IPSPs in responses to shocks to the auditory nerve confirm the existence of inhibitory interneurons IPSPs were evoked in 54 of 54 labeled T stellate cells by weak shocks to the auditory nerve. As described previously, the fact that their latencies were between 1.2 and 2 ms in most cells indicates that they are disynaptic (Wu and Oertel 1986). Disynaptic IPSPs recorded from one T stellate cell are shown in Fig. 4. The threshold of IPSPs in this cell, as in most, was slightly lower than that of EPSPs. The result that thresholds of EPSPs and IPSPs were not identical indicates that different populations of auditory nerve fibers mediate EPSPs and the IPSPs with the lowest thresholds. Trains of late IPSPs FIG . 3. T stellate cells received input from excitatory interneurons that could be activated electrically and pharmacologically. A: examples of EPSPs ( ● ) recorded from 6 different cells that occurred hundreds of milliseconds after the stimulus. Resting membrane potential is indicated ( left). Stimulation voltages (top to bottom) were /7, 2, 5, 7.5, 5, and 5. B: spontaneous EPSPs were not frequent and may have been masked by inhibition when the slice was bathed in normal saline. Blocking all inhibition with 1 mM strychnine (STR) and 10 mM bicuculline (BIC) increased the frequency of spontaneous EPSPs. Addition of 1 mM glutamate (GLU) to the bath excited intact neurons, including excitatory interneurons, in the slice and resulted in the summing of many EPSPs occasionally evoking a discharge (bottom). 9k23 / 9k22$$de05 J294-7 Strong shocks to the auditory nerve also evoked trains of late IPSPs that in some cases lasted for ú500 ms. In 49 of 54 T-stellate cells, shocks evoked trains of late IPSPs. For an individual cell, the duration of the trains of IPSPs varied on a trial-to-trial basis at a single shock strength. The cell the responses of which are shown in Fig. 5 ( left) responded to shocks of 65 V with trains of IPSPs in 4 of 5 trials; the duration of the trains of IPSPs varied between 100 and 400 ms. The probability of evoking a train of IPSPs increased with increasing shock. Examples of responses from the same cell to a series of shocks increasing in 10-V increments show that generally the longest trains of IPSPs are evoked with the strongest shocks strength (Fig. 5, middle). As in all 49 cells, the threshold and the range of stimulus strengths over which the trains of IPSPs grew was higher than the threshold of mono- and disynaptic PSPs. Examples of trains of IPSPs from six other cells are shown in Fig. 5, right. The beginning of the trains coincided with other synaptic inputs and could 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 FIG . 2. A: in 1 cell, monosynaptic EPSPs were followed by a long, slow depolarization, which sometimes elicited discharges. Shock occurred at 0 ms, and its strength (in V) is indicated above each trace. Spikes are digitally truncated. B: responses from another cell illustrate that the appearance of late synaptic events was enhanced by repetitive stimulation. SYNAPTIC INPUTS TO VCN MULTIPOLAR CELLS 55 FIG . 5. Characteristics of trains of IPSPs after auditory nerve stimulation. Left: appearance of trains of IPSPs was variable from trial to trial. Five successive responses to identical shocks show that the trains of IPSPs occurred in an all-or-none fashion and that the duration of trains of IPSPs was variable. Middle: in the same cell, the threshold of the trains of IPSPs was higher than the threshold of the monosynaptic EPSP. Right: examples of trains of IPSPs recorded from 6 cells show that there is considerable variability from 1 cell to another. In some cells, there was a mixture of both late excitation and late inhibition. dorsal to the granule cell region at the VCN-DCN border. All T stellate cells recorded in such slices exhibited late IPSPs, indicating that they received inhibition from neurons in the VCN (Fig. 7; n Å 6/6 in 4 slices). The disynaptic IPSP (Fig. 7A) and trains of IPSPs (Fig. 7B) in VCN slices were identical to those observed in slices of the entire cochlear nuclear complex. In each case, removal of the DCN was verified histologically (Fig. 7C). In addition to this VCN source of glycinergic inhibition, a DCN contribution has been demonstrated to arise from the tuberculoventral cells of the DCN (Wickesberg and Oertel 1990), indicating that disynaptic IPSPs probably have multiple components. The long trains of IPSPs, on the other hand, are unlikely to represent summed components from multiple sources. They are so unusual that they serve as physiological tags of those interneurons. Their presence in the isolated VCN shows that the interneurons which mediate trains of IPSPs lie in the VCN. Monosynaptic and late EPSPs are glutamatergic In the presence of Mg 2/ , all synaptic responses in 5 of 5 cells to stimulation of the auditory nerve were blocked by 10 mM DNQX, a blocker of a-amino-3-hydroxy-5-methyl-4isoxazolepropionic acid (AMPA) receptors (Fig. 8A) (Honoré et al. 1988). This finding confirms the conclusion that input from the auditory nerve is glutamatergic (Raman et al. 1994; Wenthold 1985; Wickesberg and Oertel 1988; Zhang and Trussell 1994). It also indicates that late responses are consequences of the activation of AMPA receptors. Long, slow depolarization in T stellate cells is mediated through NMDA receptors FIG . 6. Repetitive stimulation lowered the threshold of trains of IPSPs in a T stellate cell. Weak shocks evoked trains of IPSPs when they were presented repetitively but not when they were presented singly. 9k23 / 9k22$$de05 J294-7 Application of 100 mM APV, an antagonist of NMDA receptors, abolished the long, slow depolarization in 6/8 cells tested (Fig. 8A). In the two cells in which the slow depolarization was not abolished, it was a reduced and shortened. The slow depolarization after repetitive stimulation also was blocked by APV (Fig. 8B). A shorter-lasting depolarization that was not studied further remained in the presence of APV. Currents through NMDA receptors are known to be af- 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 not be resolved. As other synaptic responses subsided, the trains of regularly occurring IPSPs emerged, lasting between 50 and 600 ms. The long trains of late IPSPs had three features that are consistent with their arising from one or few inhibitory interneurons. First, the IPSPs were rapid and almost stereotyped in their shape. Second, they occurred in an all-or-none fashion as if an inhibitory interneuron was either activated or not activated. Third, the latest IPSPs occurred with remarkable regularity, the interval between them increasing with time. The long trains of late IPSPs behave as if they were generated by interneurons that fired long and regularly in responses to shocks. Repetitive shocks evoked trains of late IPSPs in T stellate cells more effectively than single shocks. The shock strength required to evoke late IPSPs was consistently greater than that required to evoke a monosynaptic EPSP. Shocks too weak to evoke trains of IPSPs singly often activated them when applied in rapid succession and the rate of occurrence and the duration of the train of IPSPs increased with the frequency of shocks (Fig. 6). It is known that there are several populations of glycinergic and GABAergic neurons in the cochlear nuclei (Adams and Mugnaini 1987; Mugnaini 1985; Oertel and Wickesberg 1993; Osen et al. 1990). To determine whether the interneurons that generate the long trains of IPSPs lie in the VCN or in the DCN, recordings were made from T stellate cells in slices in which the DCN was removed with a cut just 56 M. J. FERRAGAMO, N. L. GOLDING, AND D. OERTEL FIG . 7. Recordings from T stellate cells in slices of the isolated ventral cochlear nucleus (VCN) show that glycinergic interneurons reside in the VCN. A: responses of 1 cell to a series of shocks to the auditory nerve show that the disynaptic IPSP was present. B: trains of IPSPs recorded from 3 cells, each located in a different VCN slice, were identical to those observed in slices of the entire cochlear nuclear complex. Stimulation voltages (top to bottom) were /65, 50, and 65. C: these experiments were repeated for 6 cells, the location of which is indicated ( ● ) in 4 slices in which the absence of the DCN was assessed histologically. Remaining granule cell border between the VCN and dorsal cochlear nucleus (DCN) is indicated in 3 slices. Disynaptic and late IPSPs are glycinergic Consistent with earlier findings, all early and late IPSPs were blocked by 0.5 or 1 mM strychnine, indicating that they were glycinergic (Wu and Oertel 1986). The existence of the trains of IPSPs raises the question how effective they are in blocking excitation. This question was addressed by comparing responses in the absence and presence of strychnine (Fig. 9). In slices, the balance of excitation and inhibition could tilt either toward excitation or inhibition. Most commonly, shocks strong enough to evoke trains of IPSPs evoked a suprathreshold monosynaptic EPSP even when the shocks occur in rapid succession, indicating that the synchronous excitation was more potent than the inhibition (Fig. 6). With weaker shocks, however, IPSPs prevented subsequent EPSPs from reaching threshold (Fig. 9). Application of 1 mM strychnine blocked all IPSPs and tipped the balance toward excitation, so that each shock evoked a spike. GABAergic inhibition of T stellate cells is subtle Applications of picrotoxin and bicuculline, blockers of GABAA receptors, also were made to T stellate cells to test whether GABAergic inhibition played a role. Although generally all IPSPs were eliminated by strychnine, these experiments revealed the existence of GABAergic inhibition. The clearest manifestations were polysynaptic and will be discussed below. If there was a direct effect on GABAA receptors of T stellate cells, that effect was subtle. We could not demonstrate blocking of visible IPSPs, but we cannot exclude the possibility that small, slow IPSPs, such as those that might be generated in distal dendrites, were blocked. Pharmacological manipulations to characterize interneurons that impinge on T stellate cells The late synaptic responses of T stellate cells, those that are mediated by interneurons, serve as assays of the activity FIG . 8. Excitation was mediated by a-amino-3-hydroxy-5-methyl-4isoxazolepropionic acid (AMPA) and N-methyl-D-aspartate (NMDA) receptors. A: DL-2-amino-5-phosphonovaleric acid (APV) abolished the slow depolarization, indicating that the slow depolarization was mediated by NMDA receptors. 6,7-dinitroquinoxaline-2,3-dione (DNQX) eliminated the rapid, monosynaptic EPSP, indicating that it was mediated by AMPA receptors. B: in a different cell, APV partially blocked the slow depolarization and associated action potentials after repeated shocks (200 Hz; 95 ms). C: hyperpolarization ( 00.05 and 00.1 nA; 300 ms) during synaptic stimulation ( /1 V) shortened the depolarization, showing that some NMDA currents were intrinsic to T stellate cells. 9k23 / 9k22$$de05 J294-7 FIG . 9. Trains of IPSPs inhibited the monosynaptic EPSPs evoked with weak shocks. EPSPs were reduced and prevented from reaching threshold by IPSPs in response to the later shocks of a repetitive stimulus (100 Hz; 90 ms). Removal of glycinergic IPSPs with strychnine enabled a one-toone discharge to each shock as well as firing after the shock train. 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 fected by a voltage-dependent block by Mg 2/ (Nowak et al. 1984). The enhancement of late PSPs, however, raises the possibility that the depolarization arises from interneurons. To test whether the action of NMDA receptors was intrinsic to the T stellate cell from which the depolarization was measured or on excitatory interneurons, the voltage sensitivity of the response was measured in the presence of Mg 2/ (Fig. 8C). The duration of the synaptically evoked late depolarization shortened as the cell was hyperpolarized, indicating that NMDA receptors were intrinsic to the recorded T stellate cell and not on an excitatory interneuron. SYNAPTIC INPUTS TO VCN MULTIPOLAR CELLS 57 repetitive shocks evoked long-lasting excitation that resulted from the slow depolarization and from late EPSPs. Interestingly, bicuculline enhanced that excitation indicating that GABAA ergic inhibition affected excitatory interneurons and perhaps also the recorded T stellate cell. D stellate cells of the interneurons and reveal how pharmacological manipulations affect interneurons. To test the possibility that NMDA receptors mediate long-lasting excitation in the excitatory and inhibitory interneurons, their action was unblocked by removing extracellular Mg 2/ and their action was blocked by 100 mM APV. Experiments from two cells illustrated in Fig. 10 show that the firing of both excitatory and inhibitory interneurons is influenced by NMDA receptors. In the absence of Mg 2/ , even weak shocks evoked trains of IPSPs that were blocked by APV (Fig. 10A). A similar experiment in another cell shows that excitatory interneurons also are influenced by NMDA receptors. In a cell that did not respond with late EPSPs or late IPSPs to single shocks at low voltages in normal saline, the removal of Mg 2/ caused an increase in late excitation that was blocked by APV (Fig. 10B). Late inhibition was present but not as dramatic as in some other cells. The experiments illustrated in Fig. 10 were chosen to illustrate that both excitation and inhibition were affected by NMDA receptors and illustrate extremes in the range of responses that were recorded in six experiments. The contributions of GABAA ergic inhibition to the late PSPs also was examined. One experiment is illustrated in Fig. 11. Under normal conditions, this cell revealed a slow depolarization that lasted Ç200 ms in responses to single shocks and Ç500 ms in responses to repetitive shocks as well as late IPSPs and late EPSPs. Bicuculline (10 mM) had no effect on the monosynaptic spike but reversibly enhanced late IPSPs. In the presence of bicuculline, single and repetitive shocks evoked trains of late IPSPs that overcame the long, slow depolarization and caused the cell to hyperpolarize. The slow depolarization and late EPSPs were evident after the end of the train of late IPSPs. Strychnine eliminated all visible IPSPs. In the presence of strychnine, single and 9k23 / 9k22$$de05 J294-7 Synaptic responses Figure 12 displays responses of a D stellate cell to shocks delivered to the auditory nerve. Weak shocks evoked a sub- FIG . 11. GABAergic inhibition was subtle in T stellate cells. Bicuculline (BIC) promoted IPSPs (f ), enhanced the long depolarization, and promoted late EPSPs (*) in response to both single shocks and repetitive stimuli (100 Hz; 90 ms). – – – , resting potential. Strychnine (STR) abolished all IPSPs. 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 2/ FIG . 10. Removal of Mg from the extracellular saline promoted trains of IPSPs, late EPSPs, and the long, slow depolarization. A: a weak shock ( /1 V) evoked few IPSPs and EPSPs when the slice was bathed in normal saline. When Mg 2/ was removed, trains of IPSPs were evoked consistently with each trial. They subsequently were abolished by APV, indicating that they were mediated by NMDA receptors. B: in a different cell, the balance favored excitation when Mg 2/ was removed, although IPSPs were promoted as well. Late EPSPs and IPSPs were blocked completely, and the long depolarization was blocked partially by APV. These manipulations were reversed. The cell bodies of each of the labeled D stellate cells lay just beneath the superficial granule cells, resembling D stellate cells described before (Oertel et al. 1990). The dendrites radiated from the cell body, independently of the tonotopic organization, and spanned extensively the dorsoventral and rostrocaudal axis of the VCN. Collaterals of D stellate cells were intermingled with the large cells of the PVCN and also invaded the granule cell domains. Terminals were observed in regions densely populated by T stellate cells (Oertel et al. 1990; this study). In two cases where the axon was not severed, it could be traced to the deep layer of the DCN before its exit through the intermediate acoustic stria. The recordings in T stellate cells predict the existence of inhibitory interneurons that respond to strong shocks with long-lasting trains of action potentials. Those long trains depend on the activation of NMDA receptors and are countered by the activation of GABAA receptors. In the present series of experiments, recordings were made from three anatomically identified D stellate cells in which some of these tests were made. The results support the hypothesis that D stellate cells are the source of the long trains of IPSPs but does not prove it. 58 M. J. FERRAGAMO, N. L. GOLDING, AND D. OERTEL DISCUSSION FIG . 12. Responses to shocks in 1 D stellate cell. Strong shocks evoked an initial burst of spikes that was followed by subthreshold synaptic input that could last for hundreds of milliseconds. IPSPs (f ) occasionally occurred in short bursts. Late EPSPs ( ● ) often resulted in discharges. threshold depolarization of 30- to 40-ms duration. Stronger shocks elicited suprathreshold initial depolarizations and a long, slow depolarization the amplitude and duration of which grew with shock strength. The step-like manner in which the cell was depolarized suggests that it might be excited polysynaptically through excitatory interneurons. Strong shocks produced long trains of spikes that lasted hundreds of milliseconds. Inhibition was present but relatively inconspicuous in comparison with T stellate cells. The long-lasting depolarization in responses to strong shocks occasionally was interrupted by a burst of IPSPs. The pattern of activity, a long-lasting burst of spikes with increasing spike intervals, mirrored that of the IPSPs observed in T stellate cells. D stellate cells received glycinergic inhibition. Figs. 13A and 14 show that 1 mM strychnine eliminated the occasional IPSPs but that glycinergic inhibition in D stellate cells did not prominently affect synaptic responses. To test whether NMDA receptors mediate the long-lasting firing of D stellate cells in responses to shocks, 100 mM APV was applied to the bath. Figure 13A shows that APV reversibly eliminated the long-lasting depolarization. The remaining early excitation was blocked reversibly by DNQX (n Å 2/2). To determine whether NMDA receptors were intrinsic to the recorded cell or on excitatory interneurons, the voltage dependence of the late response was examined (Fig. 13B). The long, late depolarization that caused the D stellate cell to fire for Ç200 ms was shortened to Ç100 ms when the cell was hyperpolarized, this is consistent with synaptic excitation mediated by NMDA receptors that were intrinsic to the D stellate cell. GABAergic inhibition plays a prominent role in the synaptic responses of D stellate cells. Even in the absence of visible IPSPs, picrotoxin, a blocker of GABAA receptors, enhanced the firing of the cell (Fig. 14; n Å 1/1). Both the frequency and duration of firing were augmented in the presence of picrotoxin. 9k23 / 9k22$$de05 J294-7 FIG . 13. A: late excitation in D stellate cells was abolished by APV. Strychnine (STR) reversibly blocked all IPSPs but had little effect on the firing of this cell. DNQX reversibly blocked the monosynaptic EPSP mediating the first few spikes. B: hyperpolarization during stimulation ( /10 V) shortened the depolarization and associated activity, showing that some NMDA currents were intrinsic to the D stellate cell. Strychnine was included to isolate excitatory inputs. 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 T stellate cells form one of the parallel ascending auditory pathways from the ventral cochlear nucleus to the inferior colliculi. In considering the role of these neurons in the auditory pathway, the significance of the pattern of convergence of auditory nerve fiber inputs and the interaction of those synaptic inputs with the intrinsic electrical properties to generate the chopper responses to tones has been appreciated (Banks and Sachs 1991; Molnar and Pfeiffer 1968; Oertel 1983; Wang and Sachs 1995; White et al. 1994; Wu and Oertel 1984). The present study indicates that neuronal circuits that provide long-lasting excitatory and inhibitory feedforward interactions also contribute significantly to the responses of T stellate cells to activation of auditory nerve fibers. The new results raise the question what is the source of the additional inputs. T stellate cells are known to receive glycinergic inhibition from tuberculoventral cells (Wickes- SYNAPTIC INPUTS TO VCN MULTIPOLAR CELLS FIG . 14. Blocking the GABAA receptor with picrotoxin (PIC) prolonged the firing in one D stellate cell, whereas strychnine (STR) had little effect. Innervation of T stellate cells by auditory nerve fibers Auditory nerve fibers are of two types. Type I fibers are large and myelinated and comprise 95% of the total while type II fibers are small, unmyelinated and comprise only Ç5% of the total (cats: Kiang et al. 1982; mice: Ehret 1979). On the basis of extracellular injections of auditory nerve fibers in mice, type I auditory nerve fibers have been observed to terminate on both D and T stellate cells (M. W. Garb and D. Oertel, unpublished observations). Most neurons in the multipolar cell area of the PVCN (probably T stellate cells) are contacted heavily at the cell body unlike cells that project through the trapezoid body in cats. In cats type I fibers innervate all of the large cells, including those that correspond to T and D stellate cells (Liberman 1991, 1993). The anatomic findings are consistent with what is known about responses to activation of auditory nerve fibers in vivo and in vitro. The short-latency, sharply timed responses to the onset of tones indicate that chopper and onsetchopper units receive input from the large, myelinated auditory nerve fibers (Blackburn and Sachs 1989; Rhode and Smith 1986; Smith and Rhode 1989). In slices from mice, both D and T stellate cells respond to shocks of the auditory nerve with EPSPs (Oertel et al. 1990; Wu and Oertel 1986). As thresholds for EPSPs are low and latencies are õ1 ms, the input is probably from myelinated auditory nerve fibers. 9k23 / 9k22$$de05 J294-7 Anatomic and electrophysiological evidence indicates that few auditory nerve fibers innervate a T stellate cell. The orientation of the dendrites of T stellate cells parallel to the path of auditory nerve fibers and spanning a small proportion of the tonotopic axis indicates that T stellate cell dendrites are positioned to receive input from a limited group of fibers. The result that the amplitude of responses to shocks of the auditory nerve grow in three or four discrete jumps with shock strength indicates that the number of fibers innervating one T stellate cell in a mouse is small, perhaps as small as three or four (Fig. 1). As any of the jumps in amplitude could have resulted from the recruitment of more than one fiber and as it is possible that inputs might have been cut or damaged, this estimate represents a minimum. This conclusion is in contrast with the results of similar experiments in octopus cells, in which such subthreshold jumps cannot be detected (Golding et al. 1995). This result also indicates that models of choppers, based on what is known in cats, that require the integration of many inputs might be oversimplified (Banks and Sachs 1991; Molnar and Pfeiffer 1968; Wang and Sachs 1995). It is intriguing that the NMDA-receptor–mediated slow depolarizations were generated with shock strengths greater than those required to produce apparently maximal monosynaptic EPSPs. This finding suggests that different sources of glutamatergic input may activate different populations of receptors. It raises the possibility that type I auditory nerve fibers act primarily through AMPA receptors, as they are known to do in other vertebrate cochlear nuclei (Raman et al. 1994; Zhang and Trussell 1994) whereas other sources of excitation, alone or in combination, are required to activate NMDA receptors. It is conceivable that type II auditory nerve fibers contribute to the long, slow depolarization. Small, unmyelinated fibers would be expected to have higher thresholds for shocks than larger, myelinated fibers and their responses would be expected to be later. Sources of polysynaptic excitation The late EPSPs observed in T stellate cells indicate that T stellate cells receive excitatory input from excitatory interneurons in the slices. In being separated from their natural synaptic inputs, isolated axons cannot contribute to polysynaptic responses. Monosynaptic responses have latencies between 0.5 (synaptic delay) and Ç3 ms (2.5-ms conduction delay for an unmyelinated fiber of 0.5-mm plus 0.5-ms synaptic delay). Therefore EPSPs the latencies of which are ú3 ms are polysynaptic and must be generated by excitatory interneurons. Two other experimental observations confirm this conclusion. As cut axons have not been observed to fire spontaneously, the presence of spontaneous EPSPs is an indication of the existence of excitatory interneurons. Furthermore, the activation of EPSPs with the application of glutamate indicates that the dendrites of excitatory interneurons are accessible from the bath. T stellate cells are excitatory neurons known to terminate in the vicinity of T stellate cells. T stellate cells terminate locally in the multipolar cell area of the PVCN (Oertel et al. 1990; this study). This area is occupied by T stellate cells and occasional D stellate and bushy cells, some or all of which are therefore presumably their targets. The ultra- 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 berg and Oertel 1990). The present experiments show that, in addition, T stellate cells are a possible source of feedforward excitation and D stellate cells are a possible source of feed-forward inhibition. The finding that T stellate cells are influenced by GABAergic neurons is particularly intriguing. Golgi cells in the superficial granule cell domain are the only known source of GABA intrinsic to the VCN. They do not receive input from the large, myelinated type I auditory nerve fibers but may be innervated by the small, unmyelinated, type II auditory nerve fibers (Ferragamo et al. 1997). These experiments thus raise the possibility that T stellate cells are influenced by neurons in the superficial granule cell layer and that they are influenced directly by acoustic input from the large, myelinated type I auditory nerve fibers and also indirectly by the small, unmyelinated, type II auditory nerve fibers through Golgi cells. 59 60 M. J. FERRAGAMO, N. L. GOLDING, AND D. OERTEL Sources of glycinergic cochlear nuclear inhibition Glycinergic inhibition is recorded consistently in T stellate cells spontaneously and in responses to shocks of the auditory nerve as prominent, rapid IPSPs. The latencies of IPSPs indicate that they are polysynaptic and arise through interneurons that are intrinsic to the slice. All distinct IPSPs in T stellate cells, as in other cells of the VCN, are blocked by strychnine, indicating that they are glycinergic (Wu and Oertel 1986). An ability to label glycinergic interneurons with antibodies to glycine conjugates allows the population of glycinergic neurons to be identified (Oertel and Wickesberg 1993). Three groups of cells account for immunopositive labeling: in the DCN, tuberculoventral cells (Osen et al. 1990; Saint Marie et al. 1991; Wenthold et al. 1987; Wickesberg et al. 1994) and cartwheel cells (Osen et al. 1990; Saint Marie et al. 1991; Wenthold et al. 1987), and in the VCN, multipolar cells (Schofield and Cant 1996; Wenthold 1987), which correspond to D stellate cells (Oertel et al. 1990). Tuberculoventral cells have been shown to provide disynaptic, glycinergic inhibition to T stellate cells in responses to shocks of the auditory nerve (Wickesberg and Oertel 1990). Although there is no doubt that tuberculoventral cells contribute to the disynaptic IPSPs, several experimental 9k23 / 9k22$$de05 J294-7 findings show that they do not mediate the long trains of IPSPs. First, tuberculoventral cells do not fire for prolonged periods when activated through eighth nerve inputs (Golding and Oertel 1997; Zhang and Oertel 1993). Second, long trains of IPSPs are preserved in slices in which the DCN was removed from the slice (Fig. 7). Considerable experimental evidence indicates that D stellate cells are the source of the trains of IPSPs. First, it is the only class of glycine-immunopositive neurons in the VCN. Furthermore, pharmacological manipulations produce parallel changes in the firing of D stellate cells and the appearance of IPSPs in T stellate cells. 1) Late IPSPs in T stellate cells were evoked by strong shocks that lasted for hundreds of milliseconds. D stellate cells fire for long periods in responses to strong shocks. 2) Both the trains of IPSPs of T stellate cells and the late firing of D stellate cells were blocked by APV. 3) Both the trains of IPSPs in T stellate cells and late firing of D stellate cells were promoted by application of GABAA antagonists. The results that D stellate cells contact T stellate cells and that they respond to weak shocks with single spikes monosynaptically indicate that they contribute to the disynaptic IPSP. GABAA ergic influence Markers of GABAergic neurotransmission in the cochlear nucleus reveal the presence of both cell bodies and terminals that could be GABAergic. Antibodies to GABA conjugates and to glutamate decarboxylase (GAD) generally label neurons that are functionally GABAergic. Occasionally GAD and GABA are associated with neurons that are functionally glycinergic; cartwheel cells of the DCN, for example, are labeled for GABA and GAD yet seem to be glycinergic (Golding and Oertel 1997; Golding et al. 1996). Functionally GABAergic neurons and their terminals are labeled consistently for GABA and GAD, however, indicating that the source of GABAergic input in T stellate cells would be expected to be labeled. GABAergic input could arise from neurons intrinsic to the cochlear nuclei or from sites external to the nucleus, such as the superior olivary nucleus (Saint Marie et al. 1989). Only GABAergic neurons in the cochlear nuclei can function in polysynaptic circuits in slices as was observed in the present study, however, isolated terminals of extrinsic sources cannot be activated synaptically. Labeling for GAD and GABA is associated strongly with regions that contain granule cells, the molecular and fusiform cell layers of the DCN and the superficial granule cell domain of the VCN. In cats and guinea pigs, antibodies to GABA conjugates and to GAD, a biosynthetic enzyme, have been shown to label specific groups of cells and terminals (GABA: Kolston et al. 1992; Osen et al. 1990; Wenthold et al. 1986; GAD: Adams and Mugnaini 1987; Moore and Moore 1987; Mugnaini 1985; Saint Marie et al. 1989). In the DCN, the majority of cell bodies and puncta that were labeled with antibodies against GABA and GAD lie in the superficial and fusiform cell layers (Adams and Mugnaini 1987; Kolston et al. 1992; Moore and Moore 1987; Mugnaini 1985; Osen et al. 1990; Saint Marie et al. 1989; Wenthold et al. 1986). Labeled neurons are cartwheel, stellate, and Golgi cells. As none of these neurons make direct or indirect connections with the VCN, it is unlikely that cartwheel, 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 structure of T stellate cell terminals and functional studies of the inputs to the inferior colliculi is consistent with their being excitatory (Oliver 1984, 1987; Smith and Rhode 1989). The present experiments provide functional evidence in support of the conclusion that T stellate cells mediate late EPSPs. If T stellate cells are excited by other T stellate cells, then disynaptic EPSPs that reflect the firing of other stellate cells should be observed under similar conditions as stellate cell firing. The present experiments reflect the parallel nature of T stellate cell firing and late EPSPs under five experimental conditions. 1) Stellate cells consistently are brought to threshold Ç1 ms after shocks to the auditory nerve. Disynaptic EPSPs with latencies of Ç1.6 ms are observed but in the presence of monosynaptic EPSPs and disynaptic IPSPs the early disynaptic EPSPs are sometimes difficult to resolve. 2) Strong shocks evoke a long, slow depolarization in T stellate cells that causes T stellate cells to fire hundreds of milliseconds after a strong shock to the auditory nerve. Strong shocks also evoke very late EPSPs in T stellate cells. 3) APV reduces late firing and late EPSPs in T stellate cells. 4) The removal of extracellular Mg 2/ enhances firing as well as late EPSPs. 5) Strychnine and bicuculline enhance firing as well as late EPSPs in T stellate cells. In summary, although the results of the present experiments are consistent with the conclusion that T stellate cells excite one another, it does not rule out the possibility that other, hitherto unknown, cells contribute to the excitation. The only other known excitatory neurons that terminate in the vicinity of T stellate cells are granule cells. The dendrites of T stellate cells end in bushy branches, some of which often come near, but never penetrate, the layer of superficial granule cells that overlies them. It is conceivable, therefore, that granule cells could provide polysynaptic excitation. SYNAPTIC INPUTS TO VCN MULTIPOLAR CELLS 61 the same group of auditory nerve fibers innervates tuberculoventral cells which, in turn, provide delayed, glycinergic inhibition (Wickesberg and Oertel 1988, 1990). D stellate cells contribute to the disynaptic IPSP and at high shock strengths can provide trains of late IPSPs to T stellate cells. D Stellate cells are driven by type I auditory nerve fibers (Oertel et al. 1990; this study), and they receive GABAergic inhibition, of which Golgi cells are a likely source (Mugnaini 1985). Golgi cells lie in the granule cell domain, away from the terminals of type I auditory nerve fibers. The finding that they are activated by shocks to the auditory nerve more slowly than that to T or D stellate cells in the vicinity suggests that they are activated by type II auditory nerve fibers (Benson et al. 1996; Ferragamo et al. 1997). superficial stellate or Golgi cells of the DCN contribute to GABAergic inhibition in T stellate cells of the VCN. GABAergic input to T stellate cells of the VCN could arise from Golgi cells in the superficial granule cell domain either mono- or disynaptically. Labeled cell bodies identified as Golgi cells were observed to be associated with the superficial granule cell layer (Mugnaini 1985). These neurons terminate locally in the superficial granule cell layer with very dense terminal arbors that abut the underlying large cell area (Ferragamo et al. 1997). The dendrites of D stellate cells lie just beneath the superficial granule cell domain, poised to be contacted by Golgi cells proximally and distally, indicating that D stellate cells could mediate GABAergic responses. Furthermore, some of the branches of the distal dendrites of T stellate cells approach the superficial granule cell domain. If Golgi cells contact T stellate cells directly, those contacts can only be on distal dendrites. In contrast with glycinergic IPSPs, GABAergic IPSPs were not prominent in T or D stellate cells; IPSPs that remained in the presence of strychnine were small and inconspicuous, if present. There are four possible reasons for this observation: the synaptic currents associated with GABAergic inputs were relatively slower and weaker, they were generated relatively far from the somatic recording site, they were mediated through an excitatory interneuron, or there were presynaptic GABAergic receptors present. Proposed neuronal connections The present considerations have provided evidence for the connections that are summarized in Fig. 15. We propose that T stellate cells receive excitatory, glutamatergic input from a small number of type I auditory nerve fibers (monosynaptic EPSPs) as well as through collaterals of other T stellate cells (late EPSPs) (Oertel et al. 1990). The topographic arrangement of tuberculoventral cells indicates that roughly 9k23 / 9k22$$de05 J294-7 Implications for acoustic processing T stellate and D stellate cells, identified in vitro in mice correspond to cells in vivo in cats as choppers and onsetchoppers, respectively (Oertel et al. 1990; Smith and Rhode 1989). Although response patterns to tones have not been measured in mice, it is likely that all mammals have units with common characteristics. Chopper and onset-chopper units with similar characteristics have been made in cats (Blackburn and Sachs 1989; Rhode and Smith 1986; Smith and Rhode 1989) and rats (W. S. Rhode, personal communication). Choppers have been recorded in chinchillas (Wickesberg 1996) and gerbils (Frisina et al. 1990). Choppers fire regularly in response to short tone bursts, are tuned narrowly with prominent inhibitory sidebands, and have dynamic ranges that average 30 dB but rarely exceed 40 dB (Evans and Nelson 1973; Rhode and Greenberg 1994; Rhode and Smith 1986; Shofner and Young 1985). Onset choppers fire only at the beginning of sound pulses with precisely timed action potentials and have dynamic ranges that average 60 dB but can span 90 dB (Rhode and Smith 1986). The present results suggest that choppers excite other choppers tuned to similar frequencies. In responses to tones, auditory nerve fibers excite choppers most strongly at the beginning of the pulse when the firing rates of auditory nerve fibers are highest (Rhode and Smith 1986). Presumably other similarly tuned choppers boost excitation after the initial transient and account for the ability of choppers to respond with steady firing rates when their primary afferent inputs have a strong transient. This circuit raises the questions whether the mutual excitation in choppers could be self-sustaining and how chopper responses are terminated. Probably, in vivo as in vitro, the excitation is too weak to be self-sustaining; inhibition from tuberculoventral cells could terminate responses (Wickesberg and Oertel 1990). Our results also suggest that onset-choppers inhibit choppers. The possibility that onset-choppers inhibit choppers has been proposed before (Smith and Rhode 1989) and is consistent with what is known about their responses in vivo. The widely tuned onset-choppers could provide choppers with inhibitory sidebands. Near the characteristic frequency excitation from auditory nerve fibers is strong and can overcome inhibition by onset-choppers. At the edges of the response area such a model predicts that excitation by the auditory nerve would provide an onset transient that is cut short by inhibition from onset-choppers. As predicted, chop- 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 FIG . 15. Summary of the proposed connections to T stellate cells (T St). A small number of auditory nerve fibers excite T stellate cells monosynaptically. Those auditory nerve fibers also excite tuberculoventral cells (TV), which inhibit T stellate cells, contributing to the disynaptic IPSP. A different population of type I auditory nerve fibers excites D stellate cells (D St), which contribute to the disynaptic glycinergic IPSPs and occasionally provide long trains of late IPSPs. Golgi cells (Go) are driven through type II auditory nerve fibers and inhibit D stellate cells. T stellate cells excite one another. Glutamatergic inputs are indicated with clear symbols, glycinergic inputs are indicated with stippled symbols and GABAergic inputs are shown with black symbols. 62 M. J. FERRAGAMO, N. L. GOLDING, AND D. OERTEL pers do respond with onset transients away from the characteristic frequency (W. S. Rhode, personal communication). The suggestion that Golgi cells inhibit D stellate cells, onset-choppers, is also consistent with what is known about their responses to sound. GABAergic inhibition could contribute to the cessation of firing after the two or three chopping responses at the onset. Consistent with this role, application of bicuculline converted a phasic pattern to a tonic one in response to tone bursts in the PVCN (Palombi and Caspary 1992). It also has been suggested that GABAergic inhibition modifies the gain of activity in the ventral cochlear nucleus (Caspary et al. 1994; Evans and Zhao 1993, 1997). These findings together with the present results suggest that a role of type II auditory nerve inputs might be in regulating the gain of the circuits of the cochlear nuclei through Golgi cells. Received 11 April 1997; accepted in final form 5 August 1997. REFERENCES ADAMS, J. C. Ascending projections to the inferior colliculus. J. Comp. Neurol. 183: 519–538, 1979. ADAMS, J. C. Multipolar cells in the ventral cochlear nucleus project to the dorsal cochlear nucleus and the inferior colliculus. Neurosci. Lett. 37: 205–208, 1983. ADAMS, J. C. AND MUGNAINI, E. Patterns of glutamate decarboxylase immunostaining in the feline cochlear nucleus complex studied with silver enhancement and electron microscopy. J. Comp. Neurol. 262: 375–401, 1987. ALTSCHULER, R. A., BETZ, H., PARAKK AL, M. H., REEKS, K. A., AND WENTHOLD, R. J. Identification of glycinergic synapses in the cochlear nucleus through immunocytochemical localization of the postsynaptic receptor. Brain Res. 369: 316–320, 1986. BANKS, M. I. AND SACHS, M. B. Regularity analysis in a compartmental model of chopper units in the anteroventral cochlear nucleus. J. Neurophysiol. 65: 606–629, 1991. BENSON, T. E., BERGLUND, A. M., AND BROWN, M. C. Synaptic input to cochlear nucleus dendrites that receive medial olivocochlear synapses. J. Comp. Neurol. 365: 27–41, 1996. BERGLAND, A. M. AND BROWN, M. C. Central trajectories of type II spiral ganglion cells from various cochlear regions in mice. Hear. Res. 75: 121–130, 1994. BLACKBURN, C. C. AND SACHS, M. B. Classification of unit types in the anteroventral cochlear nucleus: PST histograms and regularity analysis. J. Neurophysiol. 62: 1303–1329, 1989. BRAWER, J. R., MOREST, D. K., AND KANE, E. C. The neuronal architecture of the cochlear nucleus of the cat. J. Comp. Neurol. 155: 251–300, 1974. BROWN, M. C., BERGLUND, A. M., KIANG, N.Y.S., AND RYUGO, D. K. Central trajectories of type II spiral ganglion neurons. J. Comp. Neurol. 278: 581–590, 1988. CANT, N. B. The fine structure of two types of stellate cells in the anterior division of the anteroventral cochlear nucleus of the cat. Neuroscience 6: 2643–2655, 1981. CANT, N. B. Identification of cell types in the anteroventral cochlear nucleus that project to the inferior colliculus. Neurosci. Lett. 332: 241–246, 1982. CANT, N. B. AND GASTON, K. C. Pathways connecting the right and left cochlear nuclei. J. Comp. Neurol. 212: 313–326, 1982. CASPARY, D. M., BACKOFF, P. M., FINLAYSON, P. G., AND PALOMBI, P. S. Inhibitory inputs modulate discharge rate within frequency receptive 9k23 / 9k22$$de05 J294-7 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 We are grateful to I. Siggelkow, J.A. Ekleberry, and J. Meister for flawless histological processing and to P. Heinritz for administrative support. We also are indebted to J. M. Wotton for advice on statistics and to colleagues of the Friday morning ‘‘Hearing and Donuts’’ for comments on this work. This study was supported by National Institute of Deafness and Other Communications Disorders Grant DC-00176. Address for reprint requests: D. Oertel, Dept. of Neurophysiology, University of Wisconsin Medical School, 1300 University Ave., Madison, WI 53706-1532. fields of anteroventral cochlear nucleus neurons. J. Neurophysiol. 72: 2124–2133, 1994. EHRET, G. Quantitative analysis of nerve fibre densities in the cochlea of the house mouse (Mus musculus). J. Comp. Neurol. 183: 73–88, 1979. EVANS, E. F. AND NELSON, P. G. The responses of single neurones in the cochlear nucleus of the cat as a function of their location and the anaesthetic state. Exp. Brain Res. 17: 402–427, 1973. EVANS, E. F. AND ZHAO W. Varieties of inhibition in the processing and control of processing in the mammalian cochlear nucleus. Prog. Brain Res. 97: 117–126, 1993. EVANS, E. F. AND ZHAO W. Onset units in guinea pig ventral cochlear nucleus: neuropharmacological studies (Abstract). Assoc. Res. Otolaryngol. 20: 116, 1997. FERRAGAMO, M. J., GOLDING, N. L., GARDNER, S. M., AND OERTEL, D. Golgi cells in the superficial granule cell domain over the VCN (Abstract). Assoc. Res. Otolaryngol. 20: 44, 1997. FERRAGAMO, M. J., GOLDING, N. L., AND OERTEL, D. A possible ventral cochlear nucleus source of inhibition upon T-stellate cells. Soc. Neurosci. Abstr. 22: 647, 1996. FRISINA, R. D., SMITH, R. L., AND CHAMBERLAIN, S. C. Encoding of amplitude modulation in the gerbil cochlear nucleus. I. A hierarchy of enhancement. Hear. Res. 44: 99–122, 1990. GOLDING, N. L. AND OERTEL, D. Context-dependent action of glycinergic and GABAergic inputs in the dorsal cochlear nucleus. J. Neurosci. 16: 2208–2219, 1996. GOLDING, N. L. AND OERTEL, D. Physiological identification of the targets of cartwheel cells in the dorsal cochlear nucleus J. Neurophysiol. 78: 248–260, 1997. GOLDING, N. L., ROBERTSON, D., AND OERTEL, D. Recordings from slices indicate that octopus cells of the cochlear nucleus detect coincident firing of auditory nerve fibers with temporal precision. J. Neurosci. 15: 3138– 3153, 1995. GREENBERG, S. AND RHODE, W. S. Periodicity coding in cochlear nerve and ventral cochlear nucleus. In: Auditory Processing of Complex Sounds, edited by W. A. Yost and C. S. Watson. Hillsdale, NJ: Erlbaum, 1987, p. 225–236. HONORÉ, T., DAVIES, S. N., DREJER, J., FLETCHER, E. J., JACOBSEN, P., LODGE, D., AND NIELSEN, F. E. Quinoxalinediones: potent competitive non-NMDA glutamate receptor antagonists. Science 241: 701–703, 1988. KIANG, N.Y.S., RHO, J. M., NORTHROP, C. C., LIBERMAN, M. C., AND RYUGO, D. K. Hair-cell innervation by spiral ganglion cells in adult cats. Science 217: 175–177, 1982. KOLSTON, J., OSEN, K. K., HACKNEY, C. M., OTTERSEN, O. P., AND STORMMATHISEN, J. An atlas of glycine- and GABA-like immunoreactivity and colocalization in the cochlear nuclear complex of the guinea pig. Anat. Embryol. (Berl.) 186: 443–465, 1992. LIBERMAN, M. C. Central projections of auditory-nerve fibers of differing spontaneous rate. I. Anteroventral cochlear nucleus. J. Comp. Neurol. 313: 240–258, 1991. LIBERMAN, M. C. Central projections of auditory nerve fibers of differing spontaneous rate. II. Posteroventral and dorsal cochlear nuclei. J. Comp. Neurol. 327: 17–36, 1993. MOLNAR, C. E. AND PFEIFFER, R. R. Interpretation of spontaneous spike discharge patterns of neurons in the cochlear nucleus. Proc. IEEE 56: 993–1004, 1968. MOORE, J. K. AND MOORE, R. Y. Glutamic acid decarboxylase-like immunoreactivity in brainstem auditory nuclei of the rat. J. Comp. Neurol. 260: 157–174, 1987. MUGNAINI, E. GABA neurons in the superficial layers of the rat dorsal cochlear nucleus: light and electron microscopic immunocytochemistry. J. Comp. Neurol. 235: 61–81, 1985. NOWAK, L., BREGETOVSKI, P., ASCHER, P., HERBET, P., AND PROCHIANTZ, A. Magnesium gates glutamate-activated channels in mouse central neurones. Nature 307: 462–465, 1984. OERTEL, D. Synaptic responses and electrical properties of cells in brain slices of the mouse anteroventral cochlear nucleus. J. Neurosci. 3: 2043– 2053, 1983. OERTEL, D. Use of brain slices in the study of the auditory system: spatial and temporal summation of synaptic inputs in cells in the anteroventral cochlear nucleus of the mouse. J. Acoust. Soc. Am. 78: 328–333, 1985. OERTEL, D. AND WICKESBERG, R. E. Glycinergic inhibition in the cochlear nuclei: evidence for tuberculoventral neurons being glycinergic. In: The Mammalian Cochlear Nuclei: Organization and Function, edited by SYNAPTIC INPUTS TO VCN MULTIPOLAR CELLS 9k23 / 9k22$$de05 J294-7 SMITH, P. H. AND RHODE, W. S. Structural and functional properties distinguish two types of multipolar cells in the ventral cochlear nucleus. J. Comp. Neurol. 282: 595–616, 1989. TOLBERT, L. P. AND MOREST, D. K. The neuronal architecture of the anteroventral cochlear nucleus of the cat in the region of the cochlear nerve root: electron microscopy. Neuroscience 7: 3053–3067, 1982. WANG, X. AND SACHS, M. B. Neural encoding of single-formant stimuli in the cat. II. Responses of anteroventral cochlear nucleus units. J. Neurophysiol. 71: 59–78, 1994. WANG, X. AND SACHS, M. B. Transformation of temporal discharge patterns in a ventral cochlear nucleus stellate cell model: implications for physiological mechanisms. J. Neurophysiol. 73: 1600–1616, 1995. WENTHOLD, R. J. Glutamate and aspartate as neurotransmitters of the auditory nerve. In: Auditory Biochemistry, edited by D. G. Drescher. Springfield, IL: Charles C. Thomas, 1985, pp. 125–140. WENTHOLD, R. J. Evidence for a glycinergic pathway connecting the two cochlear nuclei: an immunocytochemical and retrograde transport study. Brain Res. 415: 183–187, 1987. WENTHOLD, R. J., HUIE, D., ALTSCHULER, R. A., AND REEKS, K. A. Glycine immunoreactivity localized in the cochlear nucleus and superior olivary complex. Neuroscience 22: 897–912, 1987. WENTHOLD, R. J. AND HUNTER, C. Immunocytochemistry of glycine and glycine receptors in the central auditory system. In: Glycine Neurotransmission, edited by O. P. Ottersen and J. Storm-Mathisen. New York: Wiley, 1990, p. 392–416. WENTHOLD, R. J., ZEMPEL, J. M., PARAKK AL, M. H., REEKS, K. A., AND ALTSCHULER, R. A. Immunocytochemical localization of GABA in the cochlear nucleus of the guinea pig. Brain Res. 380: 7–18, 1986. WHITE, J. A., YOUNG, E. D., AND MANIS, P. B. The electrotonic structure of regular-spiking neurons in the ventral cochlear nucleus may determine their response properties. J. Neurophysiol. 71: 1774–1786, 1994. WICKESBERG, R. E. Rapid inhibition in the cochlear nuclear complex of the chinchilla. J. Acoust. Soc. Am. 100: 1691–1702, 1996. WICKESBERG, R. E. AND OERTEL, D. Tonotopic projection from the dorsal to the anteroventral cochlear nucleus of mice. J. Comp. Neurol. 268: 389–399, 1988. WICKESBERG, R. E. AND OERTEL, D. Delayed, frequency-specific inhibition in the cochlear nuclei of mice: a mechanism for monaural echo suppression. J. Neurosci. 10: 1762–1768, 1990. WICKESBERG, R. E., WHITLON, D., AND OERTEL, D. In vitro modulation of somatic glycine-like immunoreactivity in presumed glycinergic neurons. J. Comp. Neurol. 339: 311–327, 1994. WU, S. H. AND OERTEL, D. Intracellular injection with horseradish peroxidase of physiologically characterized stellate and bushy cells in slices of mouse anteroventral cochlear nucleus. J. Neurosci. 4: 1577–1588, 1984. WU, S. H. AND OERTEL, D. Inhibitory circuitry in the ventral cochlear nucleus is probably mediated by glycine. J. Neurosci. 6: 2691–2706, 1986. ZHANG, S. AND OERTEL, D. Cartwheel and superficial stellate cells of the dorsal cochlear nucleus of mice: intracellular recordings in slices. J. Neurophysiol. 69: 1384–1397, 1993. ZHANG, S. AND TRUSSELL, T. O. A characterization of excitatory postsynaptic potentials in the avian nucleus magnocellularis. J. Neurophysiol. 72: 705–718, 1994. 12-08-97 08:35:37 neupa LP-Neurophys Downloaded from http://jn.physiology.org/ by 10.220.33.4 on April 29, 2017 M. A. Merchan, J. J. Juiz, D. A. Godfrey, and E. Mugnaini. New York: Plenum, 1993, p. 225–238. OERTEL, D. AND WU, S. H. Morphology and physiology of cells in slice preparations of the dorsal cochlear nucleus of mice. J. Comp. Neurol. 283: 228–247, 1989. OERTEL, D., WU, S. H., GARB, M. W., AND DIZACK, C. Morphology and physiology of cells in slice preparations of the posteroventral cochlear nucleus of mice. J. Comp. Neurol. 295: 136–154, 1990. OLIVER, D. L. Dorsal cochlear nucleus projections to the inferior colliculus in the cat: a light and electron microscopic study. J. Comp. Neurol. 224: 155–172, 1984. OLIVER, D. L. Projections to the inferior colliculus from the anteroventral cochlear nucleus in the cat: possible substrates for binaural interaction. J. Comp. Neurol. 264: 24–46, 1987. OSEN, K. K. Cytoarchitecture of the cochlear nuclei in the cat. J. Comp. Neurol. 136: 453–484, 1969. OSEN, K. K. Projection of the cochlear nuclei on the inferior colliculus in the cat. J. Comp. Neurol. 144: 355–372, 1972. OSEN, K. K., OTTERSEN, O. P., AND STORM-MATHISEN, J. Colocalization of glycine-like and GABA-like immunoreactivities: a semiquantitative study of individual neurons in the dorsal cochlear nucleus of cat. In: Glycine Neurotransmission, edited by O. P. Ottersen and J. Storm-Mathisen. New York: Wiley, 1990, p. 417–451. PALOMBI, P. S. AND CASPARY, D. M. GABAa receptor antagonist bicuculline alters response properties of posteroventral cochlear nucleus neurons. J. Neurophysiol. 67: 738–746, 1992. RAMAN, I. M., ZHANG, S., AND TRUSSELL, L. O. Pathway-specific variants of AMPA receptors and their contribution to neuronal signaling. J. Neurosci. 14: 4998–5010, 1994. RHODE, W. S. AND GREENBERG, S. Encoding of amplitude modulation in the cochlear nucleus of the cat. J. Neurophysiol. 71: 1797–1825, 1994. RHODE, W. S. AND SMITH, P. H. Encoding timing and intensity in the ventral cochlear nucleus of the cat. J. Neurophysiol. 56: 261–286, 1986. ROTH, G. L., AITKIN, L. M., ANDERSEN, R. A., AND MERZENICH, M. M. Some features of the spatial organization of the central nucleus of the inferior colliculus of the cat. J. Comp. Neurol. 182: 661–680, 1978. RYUGO, D. K., WILLARD, F. H., AND FEKETE, D. M. Differential afferent projections to the inferior colliculus from the cochlear nucleus in the albino mouse. Brain Res. 210: 342–349, 1981. SAINT MARIE, R. L., BENSON, C. G., OSTAPOFF, E.-M., AND MOREST, D. K. Glycine immunoreactive projections from the dorsal to the anteroventral cochlear nucleus. Hear. Res. 51: 11–28, 1991. SAINT MARIE, R. L., OSTAPOFF, E. M., MOREST, D. K., AND WENTHOLD, R. J. Glycine-immunoreactive projection of the cat lateral superior olive: possible role in midbrain ear dominance. J. Comp. Neurol. 279: 382– 396, 1989. SCHOFIELD, B. R. AND CANT, N. B. Origins and targets of commissural connections between the cochlear nuclei in guinea pigs. J. Comp. Neurol. 375: 128–146, 1996. SHOFNER, W. P. AND YOUNG, E. D. Excitatory/Inhibitory response types in the cochlear nucleus: relationships to discharge patterns and responses to electrical stimulation of the auditory nerve. J. Neurophysiol. 54: 918– 939, 1985. 63