Synaptic plasticity after chemical deafening and electrical stimulation of the auditory nerve in cats.

The effects of deafness on brain structure and function have been studied using animal models of congenital deafness that include surgical ablation of the organ of Corti, acoustic trauma, ototoxic drugs, and hereditary deafness. This report describes the morphologic plasticity of auditory nerve synapses in response to ototoxic deafening and chronic electrical stimulation of the auditory nerve. Normal kittens were deafened by neonatal administration of neomycin that eliminated auditory receptor cells. Some of these cats were raised deaf, whereas others were chronically implanted with cochlear electrodes at 2 months of age and electrically stimulated for up to 12 months. The large endings of the auditory nerve, endbulbs of Held, were studied because they hold a key position in the timing pathway for sound localization, are readily identifiable, and exhibit deafness-associated abnormalities. Compared with those of normal hearing cats, synapses of ototoxically deafened cats displayed expanded postsynaptic densities, a 35.4% decrease in synaptic vesicle (SV) density, and a reduction in the somatic size of spherical bushy cells (SBCs). In comparison with normal hearing cats, ototoxically deafened cats that received cochlear stimulation had endbulbs that expressed postsynaptic densities (PSDs) that were statistically identical in size, showed a 48.1% reduction in SV density, and whose target SBCs had a 25.5% reduction in soma area. These results demonstrate that electrical stimulation via a cochlear implant in chemically deafened cats preserves PSD size but not other aspects of synapse morphology. This determination further suggests that the effects of ototoxic deafness are not identical to those of hereditary deafness.

Revealing the molecular layer of the primate dorsal cochlear nucleus.

In nonprimate mammals, the dorsal cochlear nucleus (DCN) is thought to play a role in the orientation of the head toward sounds of interest by integrating acoustic and somatosensory information. Humans and higher primates might not use this system because of reported phylogenetic changes in DCN cytoarchitecture [Moskowitz N (1969) Comparative aspects of some features of the central auditory system of primates. Ann N Y Acad Sci 167:357-369; Moore JK, Osen KK (1979) The cochlear nuclei in man. Am J Anat 154:393-418; Moore JK (1980) The primate cochlear nuclei: loss of lamination as a phylogenetic process. J Comp Neurol 193:609-629]. In this study, we re-evaluated this question from a comparative perspective and examined the rhesus monkey (cercopithecoid primate) using more sensitive probes and higher resolution imaging methods. We used electron microscopy to identify parallel fibers and their synapses, and molecular markers to determine that primates exhibit the main components of excitatory neurotransmission as other mammals. We observed that characteristics of the monkey molecular layer resembled what has been reported for nonprimates: (1) immunohistochemistry revealed many unmyelinated, thin axons and en passant glutamatergic synapses on dendritic spines; (2) immunohistochemistry for phosphodiesterase (PDE10A) showed the nuclei of granule cells distributed in the external molecular layer and the deep layers in the DCN; (3) antibodies for the inositol trisphosphate receptor (IP3r) and calbindin immunostained cartwheel cells; (4) postembedding immunogold labeling revealed synaptic expression of AMPA and delta glutamate receptor subunits on spines in parallel fiber endings; and (5) parallel fibers use vesicular glutamate transporter 1 (VGLUT1) to package glutamate into the synaptic vesicles and to mediate glutamate transport. These observations are consistent with the argument that the rhesus monkey DCN has neuronal features similar to those of other nonprimate mammals.

Projections of low spontaneous rate, high threshold auditory nerve fibers to the small cell cap of the cochlear nucleus in cats.

The marginal shell of the anteroventral cochlear nucleus houses small cells that are distinct from the overlying microneurons of the granule cell domain and the underlying projection neurons of the magnocellular core. This thin shell of small cells and associated neuropil receives auditory nerve input from only the low (<18 spikes/s) spontaneous rate (SR), high threshold auditory nerve fibers; high SR, low threshold fibers do not project there. It should be noted, that most of these auditory nerve terminations reside in the neuropil and intermix with dendrites that originate outside the shell. Consequently, electron microscopy is necessary to determine the synaptic targets. For this report, the terminations of intracellularly labeled low SR auditory nerve fibers in the small cell of cats cap were mapped through serial sections using a light microscope. The terminals were then examined with an electron microscope and found to form synapses with the somata and dendrites of small cells. Moreover, the small cell dendrites were identifiable by an abundance of microtubules and the presence of polyribosomes that were free or associated with membranous cisterns. These data contribute to the concept of a high threshold feedback circuit to the inner ear, and reveal translational machinery for local control of activity-dependent synaptic modification.

Hearing molecules: contributions from genetic deafness.

Considerable progress has been made over the past decade identifying many genes associated with deafness. With the identification of these hereditary deafness genes and the proteins they encode, molecular elements of basic hearing mechanisms emerge. As functional studies of these molecular elements become available, we can put together the pieces of the puzzle and begin to reach an understanding of the molecular mechanisms of hearing. The goal of this review is to discuss studies over the past decade that address the function of the proteins implicated in genetic deafness and to place them in the context of basic molecular mechanisms in hearing. The first part of this review highlights structural and functional features of the cochlea and auditory nerve. This background will provide a context for the second part, which addresses the molecular mechanisms underlying cochlear function as elucidated by genetic causes of deafness.

Postnatal development of a large auditory nerve terminal: the endbulb of Held in cats.

The endbulbs of Held are formed by the ascending branches of myelinated auditory nerve fibers and represent one of the largest synaptic endings in the brain. Most of the developmental changes in structure occur during the first 30 postnatal days of age. The neonatal endbulb begins as a flattened expansion with many filopodia, resembling a growth cone and characterized by numerous puncta adherentia and synapses associated with small postsynaptic densities; the most impressive feature of the ending at this age is its highly irregular plasma membrane that interdigitates with that of the postsynaptic spherical bushy cell. During these first 30 days, the number of puncta adherentia diminishes, postsynaptic densities nearly double in size, intermembraneous cisternae emerge, and plasma membranes flatten. These features endow the endbulb with an adult-like appearance. On the other hand, synaptic vesicle density increases progressively from approximately 50/microm2 at birth to 100/microm2 at adulthood. Mitochondria size remains constant over this developmental period but mitochondrial volume fraction increases until 60 days postnatal. Although many features of endbulb morphology stabilize by 30 days, other features suggest that endbulb development continues into the third month of age. Many of these observations correlate with the maturation of physiological response properties and suggest issues for further study.

Restoration of auditory nerve synapses in cats by cochlear implants.

Congenital deafness results in abnormal synaptic structure in endings of the auditory nerve. If these abnormalities persist after restoration of auditory nerve activity by a cochlear implant, the processing of time-varying signals such as speech would likely be impaired. We stimulated congenitally deaf cats for 3 months with a six-channel cochlear implant. The device used human speech-processing programs, and cats responded to environmental sounds. Auditory nerve fibers exhibited a recovery of normal synaptic structure in these cats. This rescue of synapses is attributed to a return of spike activity in the auditory nerve and may help explain cochlear implant benefits in childhood deafness.

Projections of the pontine nuclei to the cochlear nucleus in rats.

In the cochlear nucleus, there is a magnocellular core of neurons whose axons form the ascending auditory pathways. Surrounding this core is a thin shell of microneurons called the granule cell domain (GCD). The GCD receives auditory and nonauditory inputs and projects in turn to the dorsal cochlear nucleus, thus appearing to serve as a central locus for integrating polysensory information and descending feedback. Nevertheless, the source of many of these inputs and the nature of the synaptic connections are relatively unknown. We used the retrograde tracer Fast Blue to demonstrate that a major projection arises from the contralateral pontine nuclei (PN) to the GCD. The projecting cells are more densely located in the ventral and rostral parts of the PN. They also are clustered into a lateral and a medial group. Injections of anterograde tracers into the PN labeled mossy fibers in the contralateral GCD. The terminals are confined to those parts of the GCD immediately surrounding the ventral cochlear nucleus. There is no PN projection to the dorsal cochlear nucleus. These endings have the form of bouton and mossy fiber endings as revealed by light and electron microscopy. The PN represent a key station between the cerebral and cerebellar cortices, so the pontocochlear nucleus projection emerges as a significant source of highly processed information that is introduced into the early stages of the auditory pathway. The cerebropontocerebellar pathway may impart coordination and timing cues to the motor system. In an analogous way, perhaps the cerebropontocochlear nucleus projection endows the auditory system with a timing mechanism for extracting temporal information.

Progressive cerebellar, auditory, and esophageal dysfunction caused by targeted disruption of the frizzled-4 gene.

Wnt signaling has been implicated in the control of cell proliferation and in synapse formation during neural development, and these actions are presumed to be mediated by frizzled receptors. In this paper we report the phenotype of mice carrying a targeted deletion of the frizzled-4 (fz4) gene. fz4(-/-) mice exhibit three distinct defects: (1) progressive cerebellar degeneration associated with severe ataxia, (2) absence of a skeletal muscle sheath around the lower esophagus associated with progressive esophageal distension and dysfunction, and (3) progressive deafness caused by a defect in the peripheral auditory system unaccompanied by loss of hair cells or other auditory neurons. As assayed using a lacZ knock-in reporter, fz4 is widely expressed within the CNS. In particular, fz4 is expressed in cerebellar Purkinje cells, esophageal skeletal muscle, and cochlear inner hair cells, and the absence of Fz4 in these cells is presumed to account for the fz4(-/-) phenotype. In contrast to the early cell proliferation and patterning effects classically ascribed to Wnts, the auditory and cerebellar phenotypes of fz4(-/-) mice implicate Frizzled signaling in maintaining the viability and integrity of the nervous system in later life.

The effects of congenital deafness on auditory nerve synapses and globular bushy cells in cats.

It is well known that auditory deprivation affects the structure and function of the central nervous system. Congenital deafness represents one form of deprivation, and in the adult white cat, it has been shown to have a clear effect upon the synaptic interface between endbulbs of Held and spherical bushy cells. It is not known, however, whether all primary synapses are affected and/or whether they are affected in the same way and to the same extent. Thus, we studied a second neuronal circuit in the deaf white cat involving modified (small) endbulbs and globular bushy cells. Compared to normal hearing cats, modified endbulbs of congenitally deaf cats were 52.2% smaller but unchanged in structural complexity. There was also a striking loss of extracellular space between ending and cell body. The somata of postsynaptic globular bushy cells were 13.4% smaller and had enlarged postsynaptic densities. These data reveal that axosomatic synapses demonstrate abnormal structure as a consequence of deafness and that the extent of the abnormalities can vary with respect to the circuits involved. The implication of these observations is that synaptic anomalies would introduce differential delays within separate circuits, thereby desynchronizing neural activity from sound stimuli. This loss of synchronization could in turn disrupt temporal processing and compromise a host of related functions, including language comprehension.

Development of primary axosomatic endings in the anteroventral cochlear nucleus of mice.

The endbulb of Held is a large synaptic ending that arises from the myelinated auditory nerve fibers. Endbulbs exhibit an elaborate pattern of terminal branching and produce extensive contact with the postsynaptic cell body. These structural features appear to underlie the tight coupling between presynaptic activity and postsynaptic spike discharges. As a first step toward understanding the relationship between environmental sounds and the development of these neural elements, we examined the age-related changes in the morphology of endbulbs of Held in CBA/J mice, a strain known to retain good hearing throughout life. Neurobiotin was injected into the modiolus of the cochlea in CBA/J mice ranging in age from postnatal day 1 to 7 months. Light microscopic analyses suggest that endbulbs of the CBA/J mice develop from small bouton endings at birth into large, highly branched structures in adults. This increase in structural complexity occurs mostly during the second through eighth postnatal weeks, and general stages of development can be defined. In addition, we compared endbulb structure between adult CBA/J mice and adult shaker-2 mice (Myo15sh2/sh2) and heterozygous littermates (Myo15+/sh2). The shaker-2 mouse carries a mutated myosin 15 gene that results in congenital deafness, presumably due to abnormally short stereocilia in hair cell receptors. Neurobiotin was injected into the modiolus of adult CBA/J, Myo15sh2/sh2, and Myo15+/sh2 mice. Endbulbs of deaf adult Myo15sh2/sh2 mice exhibited a striking reduction in terminal branching compared with those of CBA/J and Myo15+/sh2 mice. Notably, the abnormal endbulbs of Myo15sh2/sh2 mice do not resemble immature endbulbs of normal-hearing mice, suggesting that deafness does not simply arrest development.

Inhibitory synaptic interactions between cochlear nuclei: evidence from an in vitro whole brain study.

Using guinea-pig isolated whole brain preparation in vitro, synaptic responses to electrical stimulation of auditory nerves were examined in intracellularly recorded and stained neurons of posteroventral and dorsal divisions of the cochlear nucleus. Stimulation of the contralateral auditory nerve evoked exclusively IPSPs in 70% of neurons, with amplitude of 2.3+/-1.2mV. Neurons of all major cell types were inhibited from the contralateral side. In the majority of responding cells (78%) IPSPs were induced at latencies of 3-9 ms suggesting di- and trisynaptic connections from contralateral auditory afferents or, respectively, mono- and disynaptic connections from the contralateral cochlear nucleus. Few cells responded with long-latency IPSPs (13.5-23ms), indicating involvement of polysynaptic pathways. These data demonstrate the existence of functional, direct and indirect inhibitory connections between the cochlear nuclei.

Glycine immunoreactivity of multipolar neurons in the ventral cochlear nucleus which project to the dorsal cochlear nucleus.

Certain distinct populations of neurons in the dorsal cochlear nucleus are inhibited by a neural source that is responsive to a wide range of acoustic frequencies. In this study, we examined the glycine immunoreactivity of two types of ventral cochlear nucleus neurons (planar and radiate) in the rat which project to the dorsal cochlear nucleus (DCN) and thus, might be responsible for this inhibition. Previously, we proposed that planar neurons provided a tonotopic and narrowly tuned input to the DCN, whereas radiate neurons provided a broadly tuned input and thus, were strong candidates as the source of broadband inhibition (Doucet and Ryugo [1997] J. Comp. Neurol. 385:245-264). We tested this idea by combining retrograde labeling and glycine immunohistochemical protocols. Planar and radiate neurons were first retrogradely labeled by injecting biotinylated dextran amine into a restricted region of the dorsal cochlear nucleus. The labeled cells were visualized using streptavidin conjugated to indocarbocyanine (Cy3), a fluorescent marker. Sections that contained planar or radiate neurons were then processed for glycine immunocytochemistry using diaminobenzidine as the chromogen. Immunostaining of planar neurons was light, comparable to that of excitatory neurons (pyramidal neurons in the DCN), whereas immunostaining of radiate neurons was dark, comparable to that of glycinergic neurons (cartwheel cells in the dorsal cochlear nucleus and principal cells in the medial nucleus of the trapezoid body). These results are consistent with the hypothesis that radiate neurons in the ventral cochlear nucleus subserve the wideband inhibition observed in the dorsal cochlear nucleus.

Differential cellular expression of isoforms of inositol 1,4,5-triphosphate receptors in neurons and glia in brain.

Inositol 1,4,5-trisphosphate receptors (IP3R) are mediators of second messenger-induced intracellular calcium release. Three isoforms are known to be expressed in brain, but their regional distributions and cellular localizations are little known. In order to better understand the roles of IP3 receptor isoforms in brain function, a first step is to define their distributions. We have used affinity-purified antibodies directed against peptides unique to each isoform to determine their sites of expression in rat brain. Type 1 IP3R (IP3R1) is dramatically enriched in Purkinje neurons in cerebellum and neurons in other regions, consistent with previous studies. By contrast, IP3R2 is only detected in glia, whereas IP3R3 is predominantly neuronal, with little detected in glia. IP3R3 is enriched in neuropil, especially in neuronal terminals (which often contain large dense core vesicles) in limbic and basal forebrain regions including olfactory tubercle, central nucleus of the amygdala, and bed nucleus of the stria terminalis. In addition, IP3R1 and IP3R3 have clearly distinct time courses of expression in developing brains. These data suggest separate roles for inositol 1,4,5-trisphosphate receptor isoforms in development, and for glial and neuronal function. The IP3R3 may be involved in regulation of neurotransmitter or neuropeptide release in terminals within specific nuclei of the basal forebrain and limbic system.

Single unit recordings in the auditory nerve of congenitally deaf white cats: morphological correlates in the cochlea and cochlear nucleus.

It is well known that experimentally induced cochlear damage produces structural, physiological, and biochemical alterations in neurons of the cochlear nucleus. In contrast, much less is known with respect to the naturally occurring cochlear pathology presented by congenital deafness. The present study attempts to relate organ of Corti structure and auditory nerve activity to the morphology of primary synaptic endings in the cochlear nucleus of congenitally deaf white cats. Our observations reveal that the amount of sound-evoked spike activity in auditory nerve fibers influences terminal morphology and synaptic structure in the anteroventral cochlear nucleus. Some white cats had no hearing. They exhibited severely reduced spontaneous activity and no sound-evoked activity in auditory nerve fibers. They had no recognizable organ of Corti, presented >90% loss of spiral ganglion cells, and displayed marked structural abnormalities of endbulbs of Held and their synapses. Other white cats had partial hearing and possessed auditory nerve fibers with a wide range of spontaneous activity but elevated sound-evoked thresholds (60-70 dB SPL). They also exhibited obvious abnormalities in the tectorial membrane, supporting cells, and Reissner's membrane throughout the cochlear duct and had complete inner and outer hair cell loss in the base. The spatial distribution of spiral ganglion cell loss correlated with the pattern of hair cell loss. Primary neurons of hearing-impaired cats displayed structural abnormalities of their endbulbs and synapses in the cochlear nucleus which were intermediate in form compared to normal and totally deaf cats. Changes in endbulb structure appear to correspond to relative levels of deafness. These data suggest that endbulb structure is significantly influenced by sound-evoked auditory nerve activity.

Ultrastructural analysis of primary endings in deaf white cats: morphologic alterations in endbulbs of Held.

Changes in structure and function of the auditory system can be produced by experimentally manipulating the sensory environment, and especially dramatic effects result from deprivation procedures. An alternative deprivation strategy utilizes naturally occurring lesions. The congenitally deaf white cat represents an animal model of sensory deprivation because it mimics a form of human deafness called the Scheibe deformity and permits studies of how central neurons react to early-onset cochlear degeneration. We studied the synaptic characteristics of the endbulb of Held, a prominent auditory nerve terminal in the cochlear nucleus. Endbulbs arise from the ascending branch of the auditory nerve fiber and contact the cell body of spherical bushy cells. After 6 months, endbulbs of deaf white cats exhibit alterations in structure that are clearly distinguishable from those of normal hearing cats, including a diminution in terminal branching, a reduction in synaptic vesicle density, structural abnormalities in mitochondria, thickening of the pre- and postsynaptic densities, and enlargement of synapse size. The hypertrophied membrane densities are suggestive of a compensatory response to diminished transmitter release. These data reveal that early-onset, long-term deafness produces unambiguous alterations in synaptic structure and may be relevant to rehabilitation strategies that promote aural/oral communication.

Projections from the ventral cochlear nucleus to the dorsal cochlear nucleus in rats.

Local circuit interactions between the dorsal and ventral divisions of the cochlear nucleus are known to influence the evoked responses of the resident neurons to sound. In the present study, we examined the projections of neurons in the ventral cochlear nucleus to the dorsal cochlear nucleus by using retrograde transport of biotinylated dextran amine injected into restricted but different regions of the dorsal cochlear nucleus. In all cases, we found retrogradely labeled granule, unipolar brush, and chestnut cells in the granule cell domain, and retrogradely labeled multipolar cells in the magnocellular core of the ventral cochlear nucleus. A small number of the labeled multipolar cells were found along the margins of the ventral cochlear nucleus, usually near the boundaries of the granule cell domain. Spherical bushy, globular bushy, and octopus cells were not labeled. Retrogradely-labeled auditory nerve fibers and the majority of labeled multipolar neurons formed a narrow sheet extending across the medial-to-lateral extent of the ventral cochlear nucleus whose dorsoventral position was topographically related to the injection site. Labeled multipolar cells within the core of the ventral cochlear nucleus could be divided into at least two distinct groups. Planar neurons were most numerous, their somata found within the associated band of labeled fibers, and their dendrites oriented within this band. This arrangement mimics the organization of isofrequency contours and implies that planar neurons respond best to a narrow range of frequencies. In contrast, radiate neurons were infrequent, found scattered throughout the ventral cochlear nucleus, and had long dendrites oriented perpendicular to the isofrequency contours. This dendritic orientation suggests that radiate neurons are sensitive to a broad range of frequencies. These structural differences between planar and radiate neurons suggest that they subserve separate functions in acoustic processing.

Ultrastructural changes in primary endings of deaf white cats.

Changes in brain structure occur as a consequence of altered experience. During maturation of the auditory nervous system, sensory deprivation is known to cause cell loss, abnormal axonal projections, and synaptic alterations. These animal data may be relevant to clinical observations that cochlear implants provide superior benefit for individuals who become deaf postlingually compared with those who become deaf prelingually. That is, implantation appears most efficacious if it occurs after functional connections are established but before deprivation-induced changes in the central auditory system. After this period, synaptic reorganization may underlie the diminished effectiveness of cochlear implants.

Peripheral course of genioglossal motor axons within the hypoglossal nerve of the rat.

The spatial compartmentalization of motor axons in cranial nerves has not been previously demonstrated. In the present study, motor axons in the medial lingual ramus of the hypoglossal nerve of rats were labeled using horseradish peroxidase and diaminobenzidine staining methods. Neuronal axons were segregated into fascicles within the main nerve trunk (average total length: 10.4 mm). Light microscopic examination of stained nerve sections revealed reaction product within individual axons and showed that the grouping of labeled nerve fibers in the medial ramus was maintained for at least the peripheral half of the nerve before the stain faded. The authors propose that the fascicular anatomy of the rat hypoglossal cranial nerve resembles that of peripheral spinal nerves in humans. This functional and structural compartmentalization may have a clinical impact on the repair and treatment of cranial nerve pathologies.

Morphological changes in the cochlear nucleus of congenitally deaf white cats.

Investigations in animal models and humans have indicated that congenital deafness produces degenerative changes in the central auditory pathway. The cochlear nucleus is the first central structure that receives cochlear input, and may be considered the origin of ascending auditory pathways. In this context, we studied congenitally deaf white cats, who express early onset cochlear receptor loss, in order to assess the nature of structural changes in cells of the cochlear nucleus. It is conceivable that pathologic alterations in higher auditory structures are transneuronally distributed through this nucleus. The cochlear nuclei of nonwhite cats with normal hearing were compared to those of deaf white cats exhibiting hearing loss in excess of 70 dB SPL. The cochlear nuclei of the deaf white cats were smaller in volume by roughly 50%, with the ventral and dorsal divisions being equally affected. Cell body silhouette area was determined for spherical bushy cells of the anteroventral cochlear nucleus (AVCN), pyramidal cells of the dorsal cochlear nucleus (DCN), sensory neurons from the principal trigeminal nucleus, and motoneurons of the facial nucleus. We found no statistical difference in neuronal cell body size between nonauditory neurons of these two groups of cats, whereas auditory neurons of deaf white cats were 30.8-39.4% smaller than those of normal cats. These data imply that neuronal changes in congenitally deaf cats are specific to the auditory pathway. Although cochlear nucleus volume loss was uniform for both divisions, there was a differential effect on cell density: AVCN cell density increased by 40%, whereas DCN cell density was relatively unaffected (10% increase). Astrocyte density was also greater in the AVCN (52%) compared to that in the DCN (5%). These observations reveal a differential impact on cells in the cochlear nucleus to congenital deafness, suggesting selective processing impairment at this level. If similar patterns of degeneration occur in humans, such pathologies may underlie reduced processing of input from cochlear implants in congenitally deaf adults.

Immunocytochemical localization of glycine in a subset of cartwheel cells of the dorsal cochlear nucleus in rats.

Glycine is an inhibitory neurotransmitter and a glutamate cofactor for N-methyl-D-aspartate (NMDA) receptors in the central nervous system. The distribution of glycine in the auditory system will therefore provide clues as to synaptic mechanisms underlying auditory signal processing. Previous studies have reported the immunocytochemical presence of glycine in the dorsal cochlear nucleus of a variety of mammals, but the specificity with respect to particular cell types has proven elusive at the light microscopic level. We sought to identify cell types in the superficial regions of the dorsal cochlear nucleus that were immunoreactive to glycine using light and electron microscopy in the rat. At the light microscopic level, glycine immunoreactivity was present in some but not all medium-sized cells in layers I and II. The somata of pyramidal and granule cells were not stained. At the electron microscopic level, using previously published ultrastructural criteria, we examined the glycine-labeled cells and determined that many but not all cartwheel cells were labeled. We also observed unlabeled unipolar brush cells, Golgi cells, and stellate cells. As some of the labeled cells could not be identified, we could not determine whether unipolar brush cells, Golgi cells or stellate cells had both labeled and unlabeled subpopulations. Our observations indicate that within the population of cartwheel cells, only a subset are glycine-immunoreactive.