Show Summary Details

Page of

PRINTED FROM OXFORD HANDBOOKS ONLINE ( © Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Handbooks Online for personal use (for details see Privacy Policy and Legal Notice).

date: 26 May 2019

In Vitro Studies of Neuromodulation and Plasticity in the Dorsal Cochlear Nucleus

Abstract and Keywords

The dorsal cochlear nucleus (DCN), a division of the cochlear nuclear complex, has been the subject of intense interest for its role in auditory processing and hearing disorders. The tonotopic layout of DCN principal cells and the refinement of processing of auditory signals by interneurons are together thought to permit encoding of sound source elevation. However, the many cell types and complex connectivity of the DCN suggest more diverse functions than localization. A prominent non-auditory input to the DCN has been proposed to assist in such functions as orienting to sounds of interest, detecting moving sounds, or cancelling self-generated sounds. Synaptic plasticity in the DCN may be essential for dynamic tuning of non-auditory input. Indeed, long-term changes in synaptic or membrane properties could underlie tinnitus, which is associated with hyperactivity in the DCN in some animal models. Finally, the DCN is invested with wide-ranging neuromodulatory mechanisms, suggesting that changes in the behavioral state of animals associated with such neuromodulatory systems might alter sensory processing at the earliest stages of the auditory pathway. This review will focus on studies that have utilized the in vitro brain slice approach to identify basic mechanisms of synaptic plasticity and neuromodulation in the DCN.

Keywords: cochlear nucleus, sound localization, synaptic plasticity, synapses, inhibition, modulation, tinnitus

The dorsal cochlear nucleus (DCN) is a mammalian-specific structure that has long fascinated auditory physiologists due to its rich cellular diversity, the striking nonlinearity of its neural responses, and its highly complex circuitry. Detailed reviews of the structure of the DCN and its function in hearing are available (Trussell & Oertel, 2018; Young & Davis, 2002; Young & Oertel, 2010), so only a brief overview will be presented here to provide context for a discussion of plasticity and modulation. Mammals feature higher-frequency hearing than birds, reptiles, and turtles, and meet the attendant challenges to encoding high-frequency sounds with unique circuitry. Brain regions like the lateral superior olive and medial nucleus of the trapezoid body use high-frequency binaural cues like interaural level and amplitude modulation phase differences to localize azimuthal sound sources. By contrast, the DCN is a monaural processing region that appears to take advantage of the pinna’s remarkable capacity to create sound source–dependent filtering of signals before they reach the eardrum, particularly for higher-frequency sounds. A prominent feature of this filter, known as the “head related transfer function,” is the presence of spectral notches whose position in the sound spectrum is highly dependent on the location of the sound source, particularly in elevation (Rice, May, Spirou, & Young, 1992). Accordingly, principal cells of the DCN are quite sensitive to these notches, and thus may encode elevation cues (Nelken & Young, 1994).

The cellular basis of this sensitivity centers on the connectivity of principal cells and a set of inhibitory interneurons (Trussell & Oertel, 2018; Young & Davis, 2002). The DCN is a layered structure, with a superficial molecular layer that lies against the ependymal layer of the brainstem, followed by a cell body layer and finally a deep layer (Brawer, Morest, & Kane, 1974; Osen, 1969; Wickesberg & Oertel, 1988). Figure 1 illustrates the cell types of the DCN and their synaptic innervation. The primary principal cell of the DCN is the fusiform cell (also called pyramidal cell), whose soma is situated in the cell body layer, and whose dendrites extend basally to the deep layer, and apically to the molecular layer. Auditory nerve fibers tuned to different sound frequencies innervate the array of fusiform cells on their basal dendrites, and thereby form a frequency map (Wickesberg & Oertel, 1988). Acoustic responses of fusiform cell from auditory nerve inputs may be reinforced or modified by additional tonotopic inputs from excitatory T-stellate neurons of the ventral cochlear nucleus (Oertel, Wright, Cao, Ferragamo, & Bal, 2011). Fusiform cells in turn project to the central nucleus of the inferior colliculus, contributing to the frequency map at that midbrain station. Auditory-evoked fusiform cell activity is tempered by two types of glycinergic interneuron. Vertical cells (also called tuberculoventral cells) of the deep layer contact the soma and basal dendrites of fusiform cells (Rubio & Juiz, 2004). These cells are narrowly tuned, possibly to frequencies just below the best frequency of the fusiform cell they innervate (Young & Voigt, 1982). Further inhibition of fusiform cells is made by the broadly tuned D-stellate of the ventral cochlear nucleus (Nelken & Young, 1994). Together, the threshold and tuning characteristics of these inhibitory cells are believed to contribute to the complex response profiles observed in fusiform cells (Young & Davis, 2002).

In Vitro Studies of Neuromodulation and Plasticity in the Dorsal Cochlear NucleusClick to view larger

Figure 1. Schematic of microcircuitry of the DCN. Left side of figure shows the granule cell domain, including the multimodal input and its preprocessing circuitry. Right side shows multisensory integration domain. Labeled gray boxes indicate cell types, open boxes indicate extrinsic inputs. Arrowheads are excitatory synapses, filled circles are inhibitory synapses, and grey line is an electrical (gap junction mediated) synapse. Abbreviations: Aud, auditory nerve; CWC, cartwheel cell; D-St, D-stellate cell; FC, fusiform cell; GoC, Golgi cell; GrC, granule cell; MMI, multimodal inputs; SSC, superficial stellate cell; T-St, T-stellate cell; UBC, unipolar brush cell; VC, vertical cell.

In addition to the auditory input to DCN, there also exists multimodal input that terminates on auditory granule cells, excitatory neurons whose axons form the parallel fibers of the DCN molecular layer and terminate on the apical dendrites of fusiform cells (Trussell & Oertel, 2018). Parallel fibers also activate two types of interneurons, the cartwheel cell and the superficial stellate cell, whose dendrites ramify in the molecular layer, and whose axon terminals inhibit the fusiform cell by co-releasing GABA and glycine. Given these two categories of input, auditory and multimodal, one view of the DCN is that it acts as a “multisensory” integrator, with apical and basal dendrites responsible for receiving multimodal and auditory input, respectively.

However, this simple view must be weighted by an appreciation of the cellular components of the auditory granule cell domain and molecular layer of the DCN, which closely resemble the circuitry of the cerebellar cortex. Auditory granule cells are distributed in up to seven regions in and around the entire cochlear nuclear complex, all feeding their axons into the DCN molecular layer through different routes (Mugnaini, Warr, & Osen, 1980). The inputs to the granule cells originate from across the central nervous system and comprise cortical, pontine, collicular, vestibular, proprioceptive fibers, as well as the fine, type II auditory nerve fibers (Shore & Zhou, 2006). These fibers form bouton or mossy endings onto granule cell dendrites. In addition, multimodal inputs also terminate on excitatory unipolar brush cells (UBCs), which in turn activate granule cells via an intrinsic mossy fiber ending (Mugnaini, Diño, & Jaarsma, 1997). Inhibitory Golgi cells receive both auditory and parallel fiber input and terminate on granule cells and UBCs (Yaeger & Trussell, 2015). Accordingly, the signals transmitted in the molecular layer are both highly diverse and highly preprocessed at this stage.

This brief overview of DCN circuitry will be revisited in later sections, particularly with respect to modulation and plasticity. However, it should be emphasized that significant features of the circuitry remain to be clarified, and this gap in knowledge limits how well we can understand the impact of plasticity in the DCN. For example, it is not known whether specific modalities are conveyed to parallel fibers as a “labeled line” or whether granule cells integrate multiple modalities, as in the cerebellar cortex (Arenz, Bracey, & Margrie, 2009; Huang et al., 2013). Moreover, it remains unclear whether molecular layer interneurons serve as feedforward inhibitors of fusiform cell activity, or whether they receive different signals than those received by fusiform cells, setting the stage for lateral inhibition (Roberts & Trussell, 2010). Also unclear is whether molecular layer interneurons act purely within isofrequency laminae of the fusiform cell array, or act across best-frequency borders. Circuit-level questions like these are helpful to keep in mind when reflecting on the functional impact of modulation and plasticity.

Short-Term Plasticity

Short-term plasticity refers to use-dependent changes in synaptic strength that last for tens of milliseconds to seconds (Zucker & Regehr, 2002). Examples to be explored here include synaptic facilitation, synaptic depression, post-tetanic potentiation, and, as described in a later section, depolarization-induced suppression of release (Lovinger, 2008; Zucker & Regehr, 2002). With some exceptions, these forms of plasticity are interpreted as changes in the amount of neurotransmitter released. The topic is of considerable interest because of the potential for short-term plasticity to contribute to the response profile of neurons when activated by different patterns of acoustic stimuli in vivo. For example, a transient response to sound might reflect depression, while a “build-up” response could, in principle, reflect facilitation, and these purely synaptic phenomena would then be integrated with the intrinsic properties of the neuron to determine the in vivo response, as characterized, for example, in a post stimulus time histogram (Zhou, Li, Yuan, Tao, & Zhang, 2015). Indeed, an important future area of work will be to find ways to untangle the synaptic and intrinsic contributions to the classical response types of auditory neurons.

Synaptic Depression

Depression has been analyzed primarily using in vitro brain slice preparations, and these studies universally describe profound synaptic depression at auditory nerve terminals in the VCN (Wang & Manis, 2008; Xie & Manis, 2017; Yang & Xu-Friedman, 2009). In DCN, EPSCs recorded in fusiform or vertical cells exhibited depression when fibers in the deep layer were stimulated (Kuo, Lu, & Trussell, 2012; Sedlacek & Brenowitz, 2014; Tang & Trussell, 2017). There is some uncertainty regarding whether this kind of stimulation activates primary afferents or collaterals of T-stellate cells, and likely both contribute to the responses. Regardless of the types of fibers contributing to the EPSC in fusiform and vertical cells, their depression has consequences for synaptic inhibition of fusiform cells. Inhibitory synapses between vertical and fusiform cells exhibit little short-term plasticity (Kuo et al., 2012; Sedlacek & Brenowitz, 2014). However, when vertical cells are driven by depressing excitatory inputs, the resulting EPSPs gradually fall subthreshold, and resulting in an apparent depression of the feed forward inhibition to the fusiform cell (Sedlacek & Brenowitz, 2014). This result illustrates how short-term plasticity may have consequences to downstream responses in a circuit.

Brain slices and the solutions in which they are bathed clearly do not mimic normal tissue and its normal environment. Does depression occur in vivo or only in the in vitro brain slice preparation? In fact, enormous differences in the extent of depression in vivo versus in vitro were documented at the auditory nerve—VCN bushy cell synapse. Kuenzel et al. (Kuenzel, Borst, & van der Heijden, 2011), using juxtacellular recordings in guinea pig VCN, demonstrated a complete absence of short term EPSP depression in vivo. This study carefully discussed the possible reasons for this discrepancy between preparations—including differences in extracellular Ca2+, presence of neuromodulators, and the effects of spontaneous activity of auditory nerve fibers—concluding that all three factors could be important. For example, depression is reduced when presynaptic release probability is lowered by lowering extracellular Ca2+ from the ~2 mM typical of many brain slice studies to the ~1–1.5 mM level found in brain extracellular fluid (Borst, 2010). Potential additional effects of anesthesia during recordings in vivo may also contribute to these differences. However, the studies of Kuenzel et al. (2011) were made with juxtacellular recording of axosomatic end-bulb synapses, and it remains unclear what is the level of synaptic depression exhibited by synapses made by afferents to dendrites of neurons in DCN. Presumably, if there is depression, the added excitatory influence of T-stellate cells recruited during a stimulus might mitigate a decline in afferent synaptic strength.

Prominent depression has also been described for inhibitory postsynaptic currents (IPSCs) generated by cartwheel cell synapses on fusiform cells or on other cartwheel cells (Mancilla & Manis, 2009; Roberts, Bender, & Trussell, 2008). Notably, depression was observed even when bath Ca2+ was reduced to in vivo levels. Depression was effectively modeled by assuming that synapses have two pools of vesicles which differ in their release probability (Pr) (Lu & Trussell, 2016). High-Pr vesicles are readily depleted and therefore account for depression of IPSCs, while low-Pr vesicles serve to maintain sustained inhibition to target cells. Interestingly, since cartwheel cells are spontaneously active in vitro and in vivo, it is likely that its synapses operate under a continual state of depression. As described in the section Neuromodulation, changes in spontaneous firing provides a mechanism for controlling synaptic strength and the level of inhibition provided to postsynaptic neurons. A smaller degree of depression was found for synapses made by inhibitory Golgi cells of the DCN and granule cell regions, when simulated with high-frequency trains (Yaeger & Trussell, 2015).


Desensitization is a form of synaptic depression in which postsynaptic transmitter-gated ion channels close despite the continued presence of the transmitter. In the DCN, profound desensitization occurs at synapses between mossy fibers and UBCs (Lu, Balmer, Romero, & Trussell, 2017). Two subtypes of UBC are found in the DCN: ON UBCs are excited by glutamate released from mossy fiber stimulation while OFF UBCs are inhibited. Transmission at the mossy fiber to the ON UBC synapse employs an odd form of depression that results in sustained transmission rather than loss of transmission. Repeated stimuli of presynaptic mossy fibers results in complete depression of the fast EPSC. Upon cessation of synaptic stimulation, a slow EPSC gradually rises over 100–200 ms and then decays over a 1-second period. This slow EPSC is sufficient to drive the majority of the excitatory response of the UBC. Slow EPSCs result from an unusual response of the AMPA receptors of UBCs, in which synaptic glutamate initially desensitizes the receptors (accounting for depression), but then allows reactivation of receptors as glutamate is slowly cleared from the synaptic cleft. This feature of the AMPA receptors requires co-expression of the γ2 transmembrane AMPA receptor regulatory protein (TARP) (Lu et al., 2017).

Synaptic Facilitation

Facilitation of EPSCs is characteristic of parallel fiber synapses made by auditory granule cells firing at rates between 10–100 Hz, and has been described for synapses onto Golgi, cartwheel, fusiform, and superficial stellate cells in vitro (Apostolides & Trussell, 2014; Manis, 1989; Roberts & Trussell, 2010; Yaeger & Trussell, 2015). The mechanism of facilitation was explored by making paired recordings between single granule and Golgi cells (Yaeger & Trussell, 2015). Spikes triggered in the granule cell led to EPSCs in Golgi cell whose mean amplitude facilitated by 250% over 5 stimuli at 90 Hz. Because the number of synapses and their release probability were small, it was possible to detect failures of transmission and ascribe them to failure to release transmitter (rather than to a stimulus failure). The fraction of failures could then be used to determine changes in release probability. This analysis showed clearly that facilitation was associated with a matching increase in release probability at the parallel fiber synapse. While the mechanism of this enhancement is not clear, it may reflect buildup of residual Ca2+ in terminals, saturation of presynaptic Ca2+ buffers, or binding of Ca2+ to a vesicle release protein that can augment exocytosis. A recent study of facilitation at cerebellar parallel fiber synapses showed that like many brain synapses, facilitation partially depends on the presynaptic protein synaptotagmin 7, but also requires as yet unknown factors (Turecek & Regehr, 2018). The physiological function of facilitation remains unexplored in the DCN at this time. However, given the multimodal flavor of the mossy fiber/parallel fiber pathway in DCN, facilitation may represent a way to provide a short-term strengthening or “reward” to parallel fibers that convey a particular modality. Alternatively, facilitation may be a mechanism for tuning parallel fiber synapses to particular patterns of activity rather than merely duration of activity: since the facilitated state activates over a few spikes, and lasts typically for less than 100 ms, repeated, brief bursts of spikes might represent an ideal mode of parallel fiber transmission. Such bursts are observed in granule cells of the cerebellar cortex in response to sensory stimuli in vivo (Chadderton, Margrie, & Häusser, 2004).

Long-Term Plasticity

Long-term plasticity in the DCN was proposed based on the close parallels between DCN circuitry and that of other cerebellum like structures. For example, plasticity is thought to underlie cerebellum-dependent motor learning and electrosensory lobe (ELL)-dependent cancellation of signals associated with self-generated electrical fields (Bell, Han, & Sawtell, 2008).

The starting point for exploring the possibility of such plasticity in the DCN is found in a study by Fujino & Oertel (2003). Using mouse brain slices, the authors recorded either from fusiform cells or cartwheel cells with patch pipettes and stimulated either auditory afferents or parallel fibers while monitoring EPSC amplitudes. Two induction protocols based on studies in hippocampus were explored: pairing of high-frequency synaptic stimuli (100 Hz) with a lengthy postsynaptic depolarization to –30 mV versus low-frequency stimuli (1 Hz) paired with briefer postsynaptic depolarization also to –30 mV. Several important conclusions came from this analysis. Both cell types responded to the high-frequency stimulation of parallel fibers with a long-term (> 1 hr) potentiation (LTP) of EPSC amplitude, and both responded to the low-frequency stimuli of parallel fibers with a long-term depression (LTD) of EPSC amplitude. Moreover, in both cells, plasticity was reversible, such that by changing stimulus protocols, LTP and LTD could be quickly expunged. Most interestingly, when the same protocol was applied to fusiform cells with stimulation of auditory afferents instead of parallel fibers, no plasticity was observed. Thus, both inhibitory and excitatory neurons could exhibit bidirectional plasticity, and in the case of the fusiform cell, the plasticity was input specific. Since cartwheel cells do not receive direct auditory afferents, the question of input specificity for those cells could not be addressed. In any case, the results clearly demonstrated that the multimodal signaling to the DCN was subject to prominent use-dependent plasticity, whereas auditory inputs were not.

The fact that auditory input was not plastic might suggest a presynaptic mechanism of LTP/LTD induction or expression, that is, a change in the parallel fiber terminals. However, further experiments by Fujino and Oertel (2003) indicated that the plasticity they documented could be inhibited by alterations in postsynaptic signaling pathways. In both cell types, LTP and LTD were reduced by inhibition of NMDA and metabotropic glutamate receptors (mGluRs). Both of these receptors appeared to contribute to EPSPs (see section on neuromodulation below). Moreover, depleting internal Ca2+ stores by including thapsigargin and caffeine in the patch pipette completely prevented plasticity. These results suggest that the activation of synaptic receptors and voltage dependent Ca2+ channels during the induction protocol led to activation of LTP and LTD pathways in a Ca2+-dependent manner. Presumably, the two protocols led to different levels of intracellular Ca2+ and thus different synaptic outcomes (Lisman, 1989). However, these manipulations do not clarify what specific pathways are required. For example, Ca2+ spikes occurring during the release of Ca2+ induced by caffeine prior to induction might itself have interfered with plasticity mechanisms. Moreover, it is not clear from this work whether subtle or even naturally occurring patterns of signals might activate the cells in a manner that could lead to plasticity.

Other potential mechanisms of DCN plasticity were suggested by Tagoe, Deeping, & Hamann (2017). Field potential recordings from rat DCN slices before and after tetanic stimulation of parallel fibers revealed that LTP was associated with an apparent decrease in release probability, as estimate by paired-pulse ratios. This aspect of LTP was reverse upon application of an NMDA receptor antagonist. Moreover, LTP could not be induced if the slices were bathed in a high-Ca2+ solution, suggesting that release probability at parallel fiber synapses must be low to permit expression of LTP. Patch clamp recordings on identified cell types suggested that this effect was seen in fusiform cells; it remains possible that cartwheel cells also exhibit this dependence on release probability.

Spike Timing Dependent Plasticity

Previous work on rodent cerebellum and electrosensory lobe (ELL) of mormyrid electric fish had suggested the presence of spike timing dependent plasticity (STDP; Han, Grant, & Bell, 2000). This form of plasticity differs from that described earlier in that it is dependent not merely on postsynaptic depolarization but on postsynaptic spikes, and specifically the precise timing of those spikes in relation to the synaptic stimulus. In different preparations, the timing dependence could lead to LTP or LTD, and have timing requirements in the range of milliseconds to hundreds of milliseconds. In ELL, medium ganglion cells exhibit LTD of parallel fiber evoked EPSPs following a pairing protocol in which parallel fiber stimuli arise 5–30 ms before evoking a broad, Ca2+-dependent spike in the ganglion cell. Other timing relationships, either positive or negative, resulted only in a weak non-associative LTP. Cartwheel cells in the DCN circuit are similar to medium ganglion cells of the ELL, as both are inhibitory, feature Ca2+-dependent spikes, and have as their synaptic target a principal cell that receives afferent input as well as parallel fiber input. Given that the Fujino and Oertel study had found similar forms of plasticity in both excitatory and inhibitory neurons, it was important to determine whether differences in the synaptic responses in the two cell types might be revealed using the more subtle STDP analyses. Tzounopulos and colleagues (Tzounopoulos, Kim, Oertel, & Trussell, 2004) explored whether cartwheel cells could exhibit STDP by pairing parallel fiber stimuli with postsynaptic complex spikes. LTD was only observed when the spike was triggered 5 ms after the EPSP for modest (10 Hz) stimulus rates. Higher-frequency (40 Hz) pairings resulted in LTP with little timing dependence. Fusiform cells also exhibited STDP but with a very different timing profile: the same protocol that led to LTD in the cartwheel cell caused LTP in the fusiform cell, and reversing the timing relationship caused fusiform cell LTD. One interpretation then is that when spikes follow parallel fiber EPSPs, plasticity enhances the excitation of the fusiform cells and reduces its likelihood of inhibition.

How does this kind of LTP/LTD associated with STDP compare to that described by Fujino and Oertel (2003) in terms of site of expression and signaling mechanisms? These questions are difficult to address without side-by-side experimental analysis, but is probable that they differ considerably. Tzounopoulos, Rubio, Keen, & Trussell (2007) explored the sites of expression and signaling pathways for fusiform and cartwheel cell STDP. LTP in fusiform cells was expressed as an enhancement of postsynaptic AMPA receptors, and required activation of NMDA receptors, intracellular Ca2+, and activity of calcium-calmodulin dependent protein kinase II (CAMKII). STDP in cartwheel cells was much more complex in its signaling requirements. LTD, induced by the 5-ms delayed EPSP-complex spike sequence, was expressed presynaptically as reduced glutamate release, and required postsynaptic NMDA receptor activation and an intracellular Ca2+ rise. Unlike plasticity with the strong pairing protocols used in Fujino and Oertel’s work, mGluRs were not required for STDP.

These results in cartwheel cells indicated the presence of a retrograde messenger. Indeed, LTD in cartwheel cells triggered with the STPD protocol was converted to LTP upon blocking endocannabinoid receptors (CB1R). This “revealed” LTP was expressed postsynaptically and had a mechanism similar to that of fusiform cell LTP. Apparently, the shape of the timing plot for STPD in cartwheel cells represented the summation of pre- and postsynaptic plasticity mechanisms with slightly different timing requirements. Immuno-electron microscopic analysis of CB1R distribution showed that CB1R are found on parallel fiber terminals apposing cartwheel cell spines but less so on terminals apposed to fusiform cell spines. Thus, synapse specific expression of CB1R is part of a mechanism to endow the inhibitory and excitatory neurons with distinct forms of STDP.


The DCN is invested with catecholaminergic, serotonergic and cholinergic fibers (Klepper & Herbert, 1991; Mellott, Motts, & Schofield, 2011; Thompson & Thompson, 2001). In vitro studies have shown that the transmitters associated with these fibers modify the activity of the DCN in remarkably diverse ways that go beyond simple alterations of excitability; indeed, these modulators differentially affect multiple aspects of neuronal circuitry, and thus could control the output of the DCN. This discussion will focus on effects of neuromodulators on in vitro slice preparations, primarily because their cellular mechanisms are most readily worked out in vitro. Our definition of neuromodulator is intentionally broad: compounds that are released at synapses, besides the conventional fast-acting transmitters, that alter the function of circuits. By this definition, substances such as Zn2+ or endocannabinoids, released by neurons whose cell bodies are within the DCN, also serve as neuromodulators. As will be seen, modulation impacts all levels of DCN function, and thus cannot be considered separately from either baseline function of the DCN circuit, or from the DCN while undergoing long-term plasticity. However, a significant caveat for in vitro studies of modulation, particularly for studies of modulators released by extrinsic inputs to DCN, is that the effects are generally studied by bath application of compounds to the brain slice bathing solution. Synaptic release of endogenous neuromodulators, the “real thing,” could potentially result in different effects than those studied in vitro, depending on the location of sites of release, the modulators’ concentration, and their receptors’ distribution.


In recordings of IPSCs in fusiform cells, bath application noradrenaline was observed to have paradoxical effects (Kuo & Trussell, 2011). On the one hand, spontaneous IPSCs originating from spontaneously firing cartwheel cells were nearly eliminated in the presence of noradrenaline, suggesting presynaptic inhibition by the modulator. However, when cartwheel cells were electrical stimulated, the resulting evoked IPSCs were significantly enhanced by noradrenaline. These seemingly opposing effects were mediated by alpha-2 noradrenergic receptors. This paradox was resolved by the discovery that, rather than suppressing the synapses of cartwheel cells, noradrenaline suppresses their spontaneous action potentials. As noted previously, cartwheel cells are prone to synaptic depression. Thus, by tuning down spontaneous firing, cartwheel terminals recovered from depression and thus release more transmitter. Moreover, because spontaneous IPSCs in the fusiform cells were reduced in frequency, the “signal to noise” of inhibition was increased several fold. An additional outcome of this modulation was also described. Multiple cartwheel cells converge onto each fusiform cell, and effective postsynaptic inhibition requires a critical number of synchronously active inhibitory neurons. By using a combination of el++ectrophysiological and optogenetic approaches, Lu & Trussell (2016) showed that under control conditions, that critical number was four cartwheel cells, whereas in the presence of noradrenaline a single cartwheel provided enough inhibition to block spiking in the fusiform cell. Thus, noradrenaline alters the effective size of the inhibitory microcircuit. If these different cartwheel cells were activated by different subsets of parallel fibers, then noradrenaline’s actions could be seen as enhancing the salience of individual modalities, as reflected in their ability to control the fusiform cell’s output.


Dopamine’s action on cartwheel cells is completely different from that of noradrenaline (Bender, Ford, & Trussell, 2010; Bender, Uebele, Renger, & Trussell, 2012). Spontaneous action potentials in cartwheel cells appear as either simple Na+-based spikes or as complex spikes, a mixture of Na+ spikelets riding on a slower T-type Ca2+ channel dependent depolarization (Golding & Oertel, 1997; Kim & Trussell, 2007; Manis, Spirou, Wright, Paydar, & Ryugo, 1994). Application of dopamine activated the D3 receptor and transformed spontaneous spiking from mixtures of simple and complex spikes to all simple spikes (Bender et al., 2012). Moreover, dopamine resulted in elevation of threshold for evoking a complex spike. By reducing complex spike activation, inhibitory synaptic events at cartwheel cell terminals appeared more in isolation rather than as a cluster, reducing the potency of inhibitory signals in the fusiform cell (Bender et al., 2010). Again, a paradox was presented, as dopamine’s effects were similar to that produced by pharmacologically blocking T-type Ca2+ channels, even though the somatically recorded T-type current was not inhibited by dopamine. In this case, it was discovered that dopamine selectively inhibited the T-type channels of the axon initial segment, and it was specifically this restricted population of channels that was responsible for originating the complex spike. Importantly, similar effects were obtained by driving release of dopamine from dopaminergic fibers. Thus, two catecholamines, dopamine and noradrenaline, both target cartwheel cell and influence their spike activity, but in totally different ways and with very different outcomes for synaptic inhibition.


Serotonin’s actions have been explored extensively in the auditory system (see chapter by Hurley, this volume), yet until recently the effects of this compound had not been investigated in the DCN. Tang & Trussell (2015) showed that serotonin enhanced spontaneous firing in fusiform cells and increased fusiform cell sensitivity to electrical stimulation. This effect was mediated by the action of 5HT2A and 5HT7 receptors, which resulted in enhancement of the so-called IH current mediated by HCN (hyperpolarization and cyclic-nucleotide gated) channels. The interaction between receptor and channel required cAMP and Src kinase signaling pathways. In an in vivo context, this mechanism would be expected to enhance background signaling from the DCN and reduce the threshold for acoustic responses. However, a second study (Tang & Trussell, 2017) suggested a much more complex effect of serotonin in DCN. In this work, the authors systematically examined serotonin effects on granule, vertical, cartwheel and fusiform cells, and their respective synaptic contacts. They discovered very selective targets of serotonin. Granule and cartwheel cells and their synapses were unaffected. Vertical cells, which mediate feedforward inhibition of fusiform cells and are excited by auditory afferent fibers, had a response to serotonin identical to that of fusiform cells. Finally, auditory afferent input to fusiform cells was inhibited. The outcome of these multiple effects was striking: responses in fusiform cells to stimulation of auditory fibers were inhibited by serotonin, a consequence of suppression of auditory synapses on fusiform cells, and of recruitment of more inhibitory vertical cells. By contrast, parallel fiber EPSPs were more effective in bringing the fusiform cells to spike threshold due to the effects on HCN just described. Thus, serotonin appears to shift the balance of input between the two sensory streams, favoring the multimodal signals carried through the granule cell system.


Endocannabinoids are neuromodulators that are produced intrinsically within the DCN rather than from extrinsic sources, unlike as serotonin or catecholamines. Besides their role in STDP described previously, endocannabinoids also play a role in a short-term plasticity called depolarization induced suppression of excitation (DSE). Here, a period of strong postsynaptic depolarization results in a depression of EPSP amplitude lasting about 10 seconds. In DCN, prominent DSE is seen for parallel fiber to cartwheel cell synapses, but less so for parallel fiber-fusiform synapses, and it is absent at auditory nerve-fusiform synapses (Zhao, Rubio, & Tzounopoulos, 2009). DSE is blocked by the CB1 receptor antagonist AM-251 and is mediated by 2-arachidonoyl-glycerol (2-AG), an endocannabinoid synthesized by diacylglycerol lipase alpha and beta (DGLα and DGLβ). Accordingly, these enzymes are localized specifically to cartwheel cell spines.

The depolarizing stimuli used in revealing DSE are severe, and therefore one may ask whether normal patterns of postsynaptic spiking could drive sufficient endocannabinoid release to mediate DSE. This was examined by Sedlacek, Tipton, & Brenowitz (2011), who monitored postsynaptic spikes, EPSPs in cartwheel cells, and found that ongoing spike activity could generate a “tonic” endocannabinoid dependent DSE proportional to spike rate. Interestingly, these investigators also monitored intracellular Ca2+, and discovered that the Ca2+ requirements for this more physiological DSE were lower than that of DSE induced by strong stimuli. Together these results suggest that endocannabinoid signaling serves not only to guide long term plasticity but also to set synaptic strength of parallel fibers on an ongoing basis.


Cholinergic fibers in the DCN likely originate as either collaterals of cochlear efferent fibers originating in the superior olive, or projections from the pedunculopontine nuclei (Mellott et al., 2011). Two notable studies have examined cholinergic modulation and its interaction with plasticity mechanisms in the DCN. In recordings of cartwheel cells, He and colleagues simultaneously monitored both glutamatergic EPSPs and synaptic spine Ca2+ changes following from parallel fiber activation (He, Wang, Petralia, & Brenowitz, 2014). The muscarinic acetylcholine receptor agonist oxotremorine-m increased both EPSPs and spine Ca2+. The mechanism of this effect was quite complex. Spine Ca2+ increases were due to the activation of NMDA receptors during the EPSP depolarization, which caused relief of Mg2+ block of the receptors. However, the EPSP also activated L-type Ca2+ channels, and Ca2+ flux through these channels activated BK-type K+ channels, which then opposed the EPSP, repolarizing the membrane. Blockade of BK channels or L-type Ca2+ channels enhanced EPSPs and spine Ca2+. He et al. (2014) found that muscarinic receptors inhibited the L-type channels through a PKA dependent mechanism, thereby reducing BK activation and allowing a greater depolarization and NMDA receptor activation.

Very different functions of muscarinic receptors in DCN were described in Zhao & Tzounopoulos (2011). The inhibitory analog of DSE, depolarization induced suppression of inhibition (DSI), is normally absent for inhibitory inputs to fusiform cells. However, application of oxotremorine-M revealed a latent endocannabinoid-dependent DSI at these inputs. Moreover, the agonist also enhanced DSE onto fusiform cells. This regulation of short-term plasticity by cholinergic receptors illustrates how extrinsic and intrinsic neuromodulator systems may interact to shape circuit activity.

As already described, fusiform cells exhibit Hebbian synaptic plasticity, with STDP protocols producing anti-Hebbian plasticity (i.e., plasticity induced by a spike-EPSP timing relation opposite to that of Hebbian plasticity) in cartwheel cells and not fusiform cells. Zhao & Tzounopoulos (2011) found that when the agonist oxotremorine-m was applied during the STDP stimulus protocol, the fusiform cell now exhibited anti-Hebbian plasticity. This effect was NMDA receptor and postsynaptic Ca2+ dependent. The authors inferred that the muscarinic receptor agonist led to phospholipase C activation and consequent release of endocannabinoids. Endocannabinoids would then act on the presynaptic parallel fiber terminals to inhibit glutamate release in a manner analogous to the mechanism described for cartwheel cells, despite the reduced expression of CB1 receptors on terminals ending on fusiform cells as compared to cartwheel cells (Tzounopoulos et al., 2007). This confound notwithstanding, the study of Zhou et al. (2015) is significant as it raises the possibility that “learning rules” that guide plasticity in the auditory system are themselves plastic, and can be modified by neuromodulators such as acetylcholine that are associated with changes in the behavioral state of animals.


Zn2+ ions are densely accumulated at parallel fiber terminals of the cochlear nucleus (Frederickson, Howell, Haigh, & Danscher, 1988; Rubio & Juiz, 1998). Indeed, their concentration at those synapses is one of the clearest factors that differentiates auditory from cerebellar granule cells, as the latter do not accumulate Zn2+. Tzounopoulos and colleagues have published a series of studies that explored the remarkably complex effects of Zn2+ in the DCN, taking advantage of new reagents for selective buffering or detection of Zn2+, as well as mouse lines with genetic deletion of the requisite Zn2+ receptors and transporters.

A common approach in the study of synaptic Zn2+ is to compare the physiological responses of synapses before and after chelation of Zn2+. Brain slice recordings in which fusiform cells are patch clamped and parallel fibers are stimulated revealed that chelation of Zn2+ by the novel, high-affinity chelator ZX1 enhanced the amplitude of both AMPA and NMDA receptor mediated synaptic responses (Anderson et al., 2015; Kalappa, Anderson, Goldberg, Lippard, & Tzounopoulos, 2015). These and other experiments indicated that endogenous Zn2+, released during synaptic transmission, directly inhibits glutamate receptor function. These effects were absent in mice lacking the presynaptic zinc transporter ZnT3, supporting the notion that ZX1 was indeed inhibiting a response to endogenous Zn2+. However, additional inhibition of NMDA receptor occurred from ambient Zn2+, whose source was not determined.

Consistent with the idea that endogenous Zn2+ is contained in vesicles, fluorescent reporters of Zn2+ showed bright labeling in the DCN molecular layer, and this fluorescent signal was absent in the ZnT3 KO mice (Kalappa et al., 2015). Stimulation of parallel fibers led to an increase in extracellular zinc-dependent fluorescence when detected with a cell-impermeable probe. Surprisingly, both of these Zn2+ signals were reduced in mice that had been noise exposed, indicating that damage to the auditory system can impact multimodal integration in the auditory granule cell system.

Other, more complex effects of Zn2+ were observed in the DCN. Perez-Rosello et al. (2013) showed that bath application of exogenous Zn2+ decreased EPSC amplitude, an effect that required expression of a postsynaptic metabotropic Zn2+ receptor, mZnR. This effect had a surprising mechanism: the Zn2+ binding to mZnR promoted release of the endocannabinoid 2-arachidonoylglycerol and presynaptic inhibition of glutamate release. This effect is distinct from the direct effect of Zn2+ on postsynaptic receptors described earlier. Pre- and postsynaptic effects of Zn2+ were further revealed by Kalappa & Tzounopoulos (2017). These authors found that high-frequency stimulation of parallel fibers in elevated bath Ca2+ (2.4 mM) led to facilitation, and this facilitation was reduced after chelation of Zn2+ by ZX1. Facilitation in physiological Ca2+ was less sensitive to ZX1. The authors proposed that Zn2+ release and subsequent suppression of glutamate release by through retrograde transmission of endocannabinoids normally allows parallel fiber terminals to facilitate to a greater extent without vesicle depletion. However, other interpretations could be considered: endocannabinoids might directly modify the facilitation mechanism, or potentiate a distinct subpopulation of synaptic vesicles that enables transmission during high-frequency stimuli (Lu & Trussell, 2016).


The microcircuits of the DCN are distinguished by their sensitivity to use-dependent plasticity and to neuromodulators. These effects are remarkably interactive: modulators can affect both short and long-term plasticity, while plasticity can impact the release of intrinsic neuromodulators, and perhaps also extrinsic neuromodulators. Defining this playing field for fine control of DCN function invites deeper investigation. For example, it is highly unlikely that the host of neuromodulators described here act independently. Not only could multiple compounds be released at once, but these may involve overlapping sets of intracellular transduction pathways. Any proposed effect of a neuromodulator must eventually be confirmed by demonstrating effects of endogenously released neuromodulators, perhaps using optogenetic methods. Synaptic plasticity mechanisms must be demonstrated using activity patterns in cell populations like those in vivo, if they are to be relevant to auditory function. Indeed, such studies have been performed, although using paradigms conceptually quite different from the in vitro models. Lastly, it will be important to further examine synaptic plasticity and modulation following noise induced hearing loss or in tinnitus animal models, to assess how pathological changes in auditory function are due to changes at the level of the synapse.


I would like to thank Tavita Garrett for helpful comments on the manuscript. My work was supported by NIH grants DC004450 and NS028901.


Anderson, C. T., Radford, R. J., Zastrow, M. L., Zhang, D. Y., Apfel, U.-P., Lippard, S. J., & Tzounopoulos, T. (2015). Modulation of extrasynaptic NMDA receptors by synaptic and tonic zinc. Proceedings of the National Academy of Sciences of the United States of America, 112(20), E2705–2714. this resource:

Apostolides, P. F., & Trussell, L. O. (2014). Chemical synaptic transmission onto superficial stellate cells of the mouse dorsal cochlear nucleus. Journal of Neurophysiology, 111(9), 1812–1822. this resource:

Arenz, A., Bracey, E. F., & Margrie, T. W. (2009). Sensory representations in cerebellar granule cells. Current Opinion in Neurobiology, 19(4), 445–451. this resource:

Bell, C. C., Han, V., & Sawtell, N. B. (2008). Cerebellum-like structures and their implications for cerebellar function. Annual Review of Neuroscience, 31, 1–24. this resource:

Bender, K. J., Ford, C. P., & Trussell, L. O. (2010). Dopaminergic modulation of axon initial segment calcium channels regulates action potential initiation. Neuron, 68(3), 500–511. this resource:

Bender, K. J., Uebele, V. N., Renger, J. J., & Trussell, L. O. (2012). Control of firing patterns through modulation of axon initial segment T-type calcium channels. The Journal of Physiology, 590(1), 109–118. this resource:

Borst, J. G. G. (2010). The low synaptic release probability in vivo. Trends in Neurosciences, 33(6), 259–266. this resource:

Brawer, J. R., Morest, D. K., & Kane, E. C. (1974). The neuronal architecture of the cochlear nucleus of the cat. The Journal of Comparative Neurology, 155(3), 251–300. this resource:

Chadderton, P., Margrie, T. W., & Häusser, M. (2004). Integration of quanta in cerebellar granule cells during sensory processing. Nature, 428(6985), 856–860. this resource:

Frederickson, C. J., Howell, G. A., Haigh, M. D., & Danscher, G. (1988). Zinc-containing fiber systems in the cochlear nuclei of the rat and mouse. Hearing Research, 36(2–3), 203–211.Find this resource:

Fujino, K., & Oertel, D. (2003). Bidirectional synaptic plasticity in the cerebellum-like mammalian dorsal cochlear nucleus. Proceedings of the National Academy of Sciences of the United States of America, 100(1), 265–270. this resource:

Golding, N. L., & Oertel, D. (1997). Physiological identification of the targets of cartwheel cells in the dorsal cochlear nucleus. Journal of Neurophysiology, 78(1), 248–260.Find this resource:

Han, V. Z., Grant, K., & Bell, C. C. (2000). Reversible associative depression and nonassociative potentiation at a parallel fiber synapse. Neuron, 27(3), 611–622.Find this resource:

He, S., Wang, Y.-X., Petralia, R. S., & Brenowitz, S. D. (2014). Cholinergic modulation of large-conductance calcium-activated potassium channels regulates synaptic strength and spine calcium in cartwheel cells of the dorsal cochlear nucleus. Journal of Neuroscience, 34(15), 5261–5272. this resource:

Huang, C.-C., Sugino, K., Shima, Y., Guo, C., Bai, S., Mensh, B. D., … Hantman, A. W. (2013). Convergence of pontine and proprioceptive streams onto multimodal cerebellar granule cells. ELife, 2, e00400. this resource:

Kalappa, B. I., Anderson, C. T., Goldberg, J. M., Lippard, S. J., & Tzounopoulos, T. (2015). AMPA receptor inhibition by synaptically released zinc. Proceedings of the National Academy of Sciences of the United States of America, 112(51), 15749–15754. this resource:

Kalappa, B. I., & Tzounopoulos, T. (2017). Context-dependent modulation of excitatory synaptic strength by synaptically released zinc. ENeuro, 4(1). this resource:

Kim, Y., & Trussell, L. O. (2007). Ion channels generating complex spikes in cartwheel cells of the dorsal cochlear nucleus. Journal of Neurophysiology, 97(2), 1705–1725. this resource:

Klepper, A., & Herbert, H. (1991). Distribution and origin of noradrenergic and serotonergic fibers in the cochlear nucleus and inferior colliculus of the rat. Brain Research, 557(1–2), 190–201.Find this resource:

Kuenzel, T., Borst, J. G. G., & van der Heijden, M. (2011). Factors controlling the input-output relationship of spherical bushy cells in the gerbil cochlear nucleus. Journal of Neuroscience, 31(11), 4260–4273. this resource:

Kuo, S. P., Lu, H.-W., & Trussell, L. O. (2012). Intrinsic and synaptic properties of vertical cells of the mouse dorsal cochlear nucleus. Journal of Neurophysiology, 108(4), 1186–1198. this resource:

Kuo, S. P., & Trussell, L. O. (2011). Spontaneous spiking and synaptic depression underlie noradrenergic control of feed-forward inhibition. Neuron, 71(2), 306–318. this resource:

Lisman, J. (1989). A mechanism for the Hebb and the anti-Hebb processes underlying learning and memory. Proceedings of the National Academy of Sciences of the United States of America, 86(23), 9574–9578.Find this resource:

Lovinger, D. M. (2008). Presynaptic modulation by endocannabinoids. Handbook of Experimental Pharmacology, 184, 435–477. this resource:

Lu, H.-W., Balmer, T. S., Romero, G. E., & Trussell, L. O. (2017). Slow AMPAR synaptic transmission is determined by stargazin and glutamate transporters. Neuron, 96(1), 73–80.e4. this resource:

Lu, H.-W., & Trussell, L. O. (2016). Spontaneous activity defines effective convergence ratios in an inhibitory circuit. Journal of Neuroscience, 36(11), 3268–3280. this resource:

Mancilla, J. G., & Manis, P. B. (2009). Two distinct types of inhibition mediated by cartwheel cells in the dorsal cochlear nucleus. Journal of Neurophysiology, 102(2), 1287–1295.Find this resource:

Manis, P. B. (1989). Responses to parallel fiber stimulation in the guinea pig dorsal cochlear nucleus in vitro. Journal of Neurophysiology, 61(1), 149–161.Find this resource:

Manis, P. B., Spirou, G. A., Wright, D. D., Paydar, S., & Ryugo, D. K. (1994). Physiology and morphology of complex spiking neurons in the guinea pig dorsal cochlear nucleus. The Journal of Comparative Neurology, 348(2), 261–276. this resource:

Mellott, J. G., Motts, S. D., & Schofield, B. R. (2011). Multiple origins of cholinergic innervation of the cochlear nucleus. Neuroscience, 180, 138–147. this resource:

Mugnaini, E., Diño, M. R., & Jaarsma, D. (1997). The unipolar brush cells of the mammalian cerebellum and cochlear nucleus: Cytology and microcircuitry. Progress in Brain Research, 114, 131–150.Find this resource:

Mugnaini, E., Warr, W. B., & Osen, K. K. (1980). Distribution and light microscopic features of granule cells in the cochlear nuclei of cat, rat, and mouse. Journal of Comparative Neurology, 191(4), 581–606.Find this resource:

Nelken, I., & Young, E. D. (1994). Two separate inhibitory mechanisms shape the responses of dorsal cochlear nucleus type IV units to narrowband and wideband stimuli. Journal of Neurophysiology, 71(6), 2446–2462.Find this resource:

Oertel, D., Wright, S., Cao, X.-J., Ferragamo, M., & Bal, R. (2011). The multiple functions of T stellate/multipolar/chopper cells in the ventral cochlear nucleus. Hearing Research, 276(1–2), 61–69. this resource:

Osen, K. K. (1969). Cytoarchitecture of the cochlear nuclei in the cat. The Journal of Comparative Neurology, 136(4), 453–484. this resource:

Perez-Rosello, T., Anderson, C. T., Schopfer, F. J., Zhao, Y., Gilad, D., Salvatore, S. R., … Tzounopoulos, T. (2013). Synaptic Zn2+ inhibits neurotransmitter release by promoting endocannabinoid synthesis. Journal of Neuroscience, 33(22), 9259–9272. this resource:

Rice, J. J., May, B. J., Spirou, G. A., & Young, E. D. (1992). Pinna-based spectral cues for sound localization in cat. Hearing Research, 58(2), 132–152.Find this resource:

Roberts, M. T., Bender, K. J., & Trussell, L. O. (2008). Fidelity of complex spike-mediated synaptic transmission between inhibitory interneurons. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 28(38), 9440–9450. this resource:

Roberts, M. T., & Trussell, L. O. (2010). Molecular layer inhibitory interneurons provide feedforward and lateral inhibition in the dorsal cochlear nucleus. Journal of Neurophysiology, 104(5), 2462–2473. this resource:

Rubio, M. E., & Juiz, J. M. (1998). Chemical anatomy of excitatory endings in the dorsal cochlear nucleus of the rat: Differential synaptic distribution of aspartate aminotransferase, glutamate, and vesicular zinc. Journal of Comparative Neurology, 399(3), 341–358.Find this resource:

Rubio, M. E., & Juiz, J. M. (2004). Differential distribution of synaptic endingscontaining glutamate, glycine, and GABA in the rat dorsal cochlear nucleus. Journal of Comparative Neurology, 477(3), 253–272. this resource:

Sedlacek, M., & Brenowitz, S. D. (2014). Cell-type specific short-term plasticity at auditory nerve synapses controls feed-forward inhibition in the dorsal cochlear nucleus. Frontiers in Neural Circuits, 8, 78. this resource:

Sedlacek, M., Tipton, P. W., & Brenowitz, S. D. (2011). Sustained firing of cartwheel cells in the dorsal cochlear nucleus evokes endocannabinoid release and retrograde suppression of parallel fiber synapses. Journal of Neuroscience, 31(44), 15807–15817. this resource:

Shore, S. E., & Zhou, J. (2006). Somatosensory influence on the cochlear nucleus and beyond. Hearing Research, 216–217, 90–99. this resource:

Tagoe, T., Deeping, D., & Hamann, M. (2017). Saturation of long-term potentiation in the dorsal cochlear nucleus and its pharmacological reversal in an experimental model of tinnitus. Experimental Neurology, 292, 1–10. this resource:

Tang, Z.-Q., & Trussell, L. O. (2015). Serotonergic regulation of excitability of principal cells of the dorsal cochlear nucleus. Journal of Neuroscience, 35(11), 4540–4551. this resource:

Tang, Z.-Q., & Trussell, L. O. (2017). Serotonergic modulation of sensory representation in a central multisensory circuit is pathway specific. Cell Reports, 20(8), 1844–1854. this resource:

Thompson, A. M., & Thompson, G. C. (2001). Serotonin projection patterns to the cochlear nucleus. Brain Research, 907(1–2), 195–207.Find this resource:

Trussell, L. O., & Oertel, D. (2018). Mirocircuits of the dorsal cochlear nucleus. In D. L. Oliver, N. B. Cant, R. R. Fay, & A. N. Popper (Eds.), The Mammalian Auditory Pathways: Synaptic Organization and Microcircuits (Vol. 65, pp. 73–99). New York, NY: Springer-Verlag.Find this resource:

Turecek, J., & Regehr, W. G. (2018). Synaptotagmin 7 Mediates Both Facilitation and Asynchronous Release at Granule Cell Synapses. Journal of Neuroscience, 38(13), 3240–3251. this resource:

Tzounopoulos, T., Kim, Y., Oertel, D., & Trussell, L. O. (2004). Cell-specific, spike timing-dependent plasticities in the dorsal cochlear nucleus. Nature Neuroscience, 7(7), 719–725. this resource:

Tzounopoulos, T., Rubio, M. E., Keen, J. E., & Trussell, L. O. (2007). Coactivation of pre- and postsynaptic signaling mechanisms determines cell-specific spike-timing-dependent plasticity. Neuron, 54(2), 291–301. this resource:

Wang, Y., & Manis, P. B. (2008). Short-term synaptic depression and recovery at the mature mammalian endbulb of Held synapse in mice. Journal of Neurophysiology, 100(3), 1255–1264. this resource:

Wickesberg, R. E., & Oertel, D. (1988). Tonotopic projection from the dorsal to the anteroventral cochlear nucleus of mice. The Journal of Comparative Neurology, 268(3), 389–399. this resource:

Xie, R., & Manis, P. B. (2017). Radiate and planar multipolar neurons of the mouse anteroventral cochlear nucleus: Intrinsic excitability and characterization of their auditory nerve input. Frontiers in Neural Circuits, 11, 77. this resource:

Yaeger, D. B., & Trussell, L. O. (2015). Single granule cells excite Golgi cells and evoke feedback inhibition in the cochlear nucleus. Journal of Neuroscience, 35(11), 4741–4750. this resource:

Yang, H., & Xu-Friedman, M. A. (2009). Impact of synaptic depression on spike timing at the endbulb of Held. Journal of Neurophysiology, 102(3), 1699–1710. this resource:

Young, E. D., & Davis, K. A. (2002). Circuitry and function of the dorsal cochlear nucleus. In D. Oertel, R. R. Fay, & A. N. Popper (Eds.), Integrative Functions in the Mammalian Auditory Pathway (Vol. 15, pp. 160–206). New York, NY: Springer.Find this resource:

Young, E. D., & Oertel, D. (2010). Cochlear nucleus. In G. M. Shepherd & S. Grillner (Eds.), Handbook of Brain Microcircuits (pp. 215–223). New York, NY: Oxford University Press.Find this resource:

Young, E. D., & Voigt, H. F. (1982). Response properties of type II and type III units in dorsal cochlear nucleus. Hearing Research, 6(2), 153–169.Find this resource:

Zhao, Y., Rubio, M. E., & Tzounopoulos, T. (2009). Distinct functional and anatomical architecture of the endocannabinoid system in the auditory brainstem. Journal of Neurophysiology, 101(5), 2434–2446. this resource:

Zhao, Y., & Tzounopoulos, T. (2011). Physiological activation of cholinergic inputs controls associative synaptic plasticity via modulation of endocannabinoid signaling. Journal of Neuroscience, 31(9), 3158–3168. this resource:

Zhou, M., Li, Y.-T., Yuan, W., Tao, H. W., & Zhang, L. I. (2015). Synaptic mechanisms for generating temporal diversity of auditory representation in the dorsal cochlear nucleus. Journal of Neurophysiology, 113(5), 1358–1368. this resource:

Zucker, R. S., & Regehr, W. G. (2002). Short-term synaptic plasticity. Annual Review of Physiology, 64, 355–405. this resource: