Projections from ventral hippocampus to medial prefrontal cortex but not nucleus accumbens remain functional after fornix lesions in rats.

Sensorimotor gating, as measured by prepulse inhibition (PPI) of startle, is deficient in human beings with schizophrenia and is greatly reduced in rats after bilateral infusion of N-methyl-D-aspartate (NMDA) into the ventral hippocampus (VH). The disruption of PPI by bilateral VH NMDA infusion is blocked by bilateral medial prefrontal cortex (mPFC) lesions, but not by bilateral lesions of the fornix, which is the principal output pathway of the hippocampal formation of the VH. Tract-tracing studies have shown the presence of additional nonfornical pathways by which the VH and neighboring structures of the amygdala may reach forebrain regions that regulate PPI, including the mPFC. To determine whether these nonfornical pathways might mediate forebrain activation after VH NMDA infusion, we examined the effects of bilateral VH NMDA infusion on c-Fos protein expression in the mPFC and nucleus accumbens (NAC) after sham vs. bilateral fornix lesions. Significant increases of c-Fos expression were observed in both the mPFC and NAC after bilateral VH NMDA infusions. Fornix lesions blocked enhanced c-Fos expression in the NAC but not the mPFC after VH NMDA infusion. The results suggest that an intact fornix may be necessary for VH activation of the NAC, but that the VH uses additional nonfornical projections to activate PPI-regulatory circuits within the mPFC.

Granule cells in the cochlear nucleus sensitive to sound activation detected by Fos protein expression.

Granule cells are the smallest neuronal type in the cochlear nucleus (CN). Due to their small size, it is extremely difficult to record their sound-evoked activity with microelectrodes. Compared with large, non-granule cells, much less is known about their response properties to sound stimulation. Here, we use Fos, the nuclear regulatory protein, as a neuronal activity marker to determine the responsiveness of granule cells to sound in comparison to the larger neurons. The present study determined the threshold sensitivity and activation pattern of neurons in the three subdivisions of the CN with free-field sound stimulation in monaural, awake rats. Immunocytochemical localization of Fos was used as our metric for "sound activation." Neuronal types upregulating Fos expression in response to sound stimulation were further identified with Nissl counterstaining. Our results show that most CN cell types can upregulate Fos expression when sound activated and the number of Fos-expressing neurons is directly related to sound intensity. The threshold for Fos activation in granule cells is lower than that for non-granule cells. The number of Fos activated granule cells saturates at high sound intensity, while the number of Fos activated non-granule cells is a monotonic function. By comparing the patterns of sound-induced Fos expression in different CN cell types, it may be possible to predict features of sound-evoked activity in granule cells.

Laminar inputs from dorsal cochlear nucleus and ventral cochlear nucleus to the central nucleus of the inferior colliculus: two patterns of convergence.

The central nucleus of the inferior colliculus is a laminated structure composed of oriented dendrites and similarly oriented afferent fibers that provide a substrate for tonotopic organization. Although inputs from many sources converge in the inferior colliculus, how axons from these sources contribute to the laminar pattern has remained unclear. Here, we investigated the axons from the cochlear nuclei that terminate in the central nucleus of the cat and rat. After characterization of the best frequency of the neurons at the injection sites in the cochlear nucleus, the neurons were labeled with dextran in order to visualize their axons and synaptic boutons in the central nucleus. Quantitative methods were used to determine the size and distribution of the boutons within the laminar organization. Two components in the laminae were identified: (1) a narrow axonal lamina that included the largest fibers and largest boutons; (2) a wide axonal lamina, surrounding the narrow lamina, composed of thin fibers and only small boutons. The wide lamina was approximately 30-40% wider than the narrow lamina, and it often extended more than 100 microm beyond the larger boutons on each side. The presence of both thick and thin fibers within the acoustic striae following these injections suggests that large and small fibers/boutons within these bands may originate from different neuronal types in the dorsal and ventral cochlear nucleus. We conclude that the narrow laminae that contain large fibers and boutons originate from larger cell types in the cochlear nucleus. In contrast, the wide lamina composed exclusively of small boutons may represent an input from other, perhaps smaller neurons in the cochlear nucleus. Thus, two types of inferior colliculus laminar structures may originate from the cochlear nucleus, and the small boutons in the wide laminae may contribute a functionally distinct input to the neurons of the inferior colliculus.

Prefrontal D1 and ventral hippocampal N-methyl-D-aspartate regulation of startle gating in rats.

BACKGROUND: Sensorimotor gating, as measured by prepulse inhibition of the startle reflex, is deficient in schizophrenia patients, and in rats after specific manipulations of limbic cortico-striato-pallido-thalamic circuitry. For example, prepulse inhibition in rats is disrupted after D1 blockade in the medial prefrontal cortex, and after N-methyl-D-aspartate infusion into the ventral hippocampus. In the present study, we examined whether these two substrates form part of an integrated circuit regulating sensorimotor gating, which might contribute to the loss of prepulse inhibition in patient populations. METHODS: Prepulse inhibition was assessed in male Sprague-Dawley rats after systemic or intra-medial prefrontal cortex administration of the D1 antagonist, SCH 23390. Separate rats received intra-medial prefrontal cortex infusion of the retrograde transported label Fluoro-Gold. In rats with sham or electrolytic lesions of the medial prefrontal cortex, prepulse inhibition was tested after infusion of N-methyl-D-aspartate or vehicle into ventral hippocampus regions that were determined to send projections to the medial prefrontal cortex. RESULTS: Prepulse inhibition was disrupted after systemic SCH 23390 treatment and after infusion of SCH 23390 into medial prefrontal cortex sites within the prelimbic and cingulate cortices. Fluoro-Gold infusion into these medial prefrontal cortex sites labeled cells in the ventral hippocampus complex, including regions CA1 and entorhinal cortex. N-methyl-D-aspartate infusions into these ventral hippocampus regions disrupted prepulse inhibition in rats after sham but not electrolytic lesions of the medial prefrontal cortex. CONCLUSIONS: Prepulse inhibition appears to be regulated by interacting substrates within the ventral hippocampus and MPFC. Specifically, NMDA activation of the ventral hippocampus appears to disrupt prepulse inhibition in a manner that is dependent on the integrity of infralimbic or cingulate cortical regions that also support a D1-mediated regulation of prepulse inhibition. Conceivably, dysfunction within these hippocampal-frontal circuits may contribute to sensorimotor gating deficits in schizophrenia.

Input-output relationships of the dorsal nucleus of the lateral lemniscus: possible substrate for the processing of dynamic spatial cues.

One organizing principle of the auditory system is the progressive representation of best tuning frequency. Superimposed on this tonotopy are nucleotopic organizations, some of which are related to the processing of different spatial cues. In the present study, we correlated asymmetries in the outputs of the dorsal nucleus of the lateral lemniscus (DNLL) to the two inferior colliculi (ICs), with asymmetries in the inputs to DNLL from the two lateral superior olives (LSOs). The positions of DNLL neurons with crossed and uncrossed projections were plotted from cases with unilateral injections of retrograde tracers in the IC. We found an orderly dorsal-to-ventral progression to the output that recapitulated the tonotopy of DNLL. In addition, we found a nucleotopic organization in the ventral (high-frequency) part of DNLL. Neurons with projections to the ventromedial (high-frequency) part of the contralateral IC were preferentially located ventrolaterally in DNLL; those with projections to the ventromedial part of the ipsilateral IC were preferentially located ventromedially in DNLL. This partial segregation of outputs corresponded with a partial segregation of inputs from the two LSOs in cases which received closely matched bilateral injections of anterograde tracers in LSO. The ventral part of DNLL received a heavy projection medially from the opposite LSO and a heavy projection laterally from the ipsilateral LSO. The findings suggest a direct relationship in the ventral part of the DNLL between inputs from the two LSOs and outputs to the two ICs. Possible roles for this segregation of pathways in DNLL are discussed in relation to the processing of static and dynamic spatial cues.

Spatial representation of frequency in the rat dorsal nucleus of the lateral lemniscus as revealed by acoustically induced c-fos mRNA expression.

The conventional view, based largely on studies in cats, holds that the dorsal nucleus of the lateral lemniscus (DNLL) is tonotopically organized with a dorsal (low-frequency) to ventral (high-frequency) representation. Based on the topography of projections between the DNLL and inferior colliculus, it has been proposed that the rat DNLL has a concentric, inside-to-outside, tonotopic organization with high frequencies represented along the rind and low frequencies represented in the core. We used acoustic stimulation and c-fos mRNA expression to examine this issue. Results suggest that the rat DNLL does have a crude tonotopic organization and that this tonotopy has a concentric component. Following high-frequency stimulation, labeled neurons were found most frequently along the margins of DNLL, although they also tended to be more concentrated ventrally. Many fewer neurons labeled following middle-frequency stimulation, and these tended to be more uniformly distributed throughout the nucleus. Still fewer neurons labeled after low-frequency stimulation and these tended to be scattered mostly in the dorsal half of the nucleus. We conclude that: (i) many more neurons in the rat DNLL are responsive to high-frequency than to low-frequency acoustic stimulation; and (ii) that the frequency representation of the rat DNLL has both concentric and dorsal-to-ventral components.

Effects of stimulus frequency and intensity on c-fos mRNA expression in the adult rat auditory brainstem.

Induction of the cellular fos gene (c-fos) is one of the earliest transcriptional changes observed following neuronal excitation. Although not an activity marker in the strict electrophysiological sense, many neurons in the central nervous system increase their c-fos expression after periods of sustained stimulation at physiological levels of intensity. In the present study, induction of c-fos mRNA expression was examined in the auditory brainstem after 1 hour of continuous free-field acoustic stimulation. Sprague-Dawley rats were exposed to pure tones of 2, 8, 16, or 32 kHz or half-octave noise bands centered on 2, 8, or 32 kHz at 80-120 dB SPL. Stimulation-induced c-fos mRNA expression was evident at all levels of the auditory brainstem, and this expression was intensity dependent. In some brain areas, induced expression manifested a clear tonotopic organization, i.e., in dorsal, posteroventral, and anteroventral cochlear nuclei, and in the medial nucleus of the trapezoid body. The inferior colliculus exhibited multiple tonotopic representations. The dorsal nucleus of the lateral lemniscus had a crude tonotopy. Although expression was present, tonotopy was not evident in periolivary nuclei or in the ventral or intermediate nuclei of the lateral lemniscus. Free-field diotic stimulation did not induce c-fos mRNA expression in the medial or lateral superior olivary nuclei. Expression was induced in the lateral superior olive by dichotic stimulation (after a unilateral cochlear ablation), and that expression was tonotopically organized. The results suggest that stimulation-induced c-fos mRNA expression can be an effective way of mapping neuronal activity in the central auditory system under both normal and pathological conditions.

Cochlear ablation alters acoustically induced c-fos mRNA expression in the adult rat auditory brainstem.

Expression of c-fos mRNA was studied in the adult rat brain following cochlear ablations by using in situ hybridization. In normal animals, expression was produced by acoustic stimulation and was found to be tonotopically distributed in many auditory nuclei. Following unilateral cochlear ablation, acoustically driven expression was eliminated or decreased in areas normally activated by the ablated ear, e.g., the ipsilateral dorsal and ventral cochlear nuclei, dorsal periolivary nuclei, and lateral nucleus of the trapezoid body and the contralateral medial and ventral nuclei of the trapezoid body, lateral lemniscal nuclei, and inferior colliculus. These deficits did not recover, even after long survivals up to 6 months. Results also indicated that neurons in the dorsal cochlear nucleus could be activated by contralateral stimulation in the absence of ipsilateral cochlear input and that the influence of the contralateral ear was tonotopically organized. Results also indicated that c-fos expression rose rapidly and persisted for up to 6 months in neurons in the rostral part of the contralateral medial nucleus of the trapezoid body following a cochlear ablation, even in the absence of acoustic stimulation. This response may reflect a release of constitutive excitatory inputs normally suppressed by missing afferent input or changes in homeostatic gene expression related to sensory deprivation. Instances of transient, surgery-dependent increases in c-fos mRNA expression in the absence of acoustic stimulation were observed in the superficial dorsal cochlear nucleus and the cochlear nerve root on the ablated side.

Substrate for rapid feedforward inhibition of the auditory forebrain.

A sizeable, feedforward, GABAergic projection exists from the inferior colliculus to the medial geniculate body in cats. We compared the dimensions of GABA-immunoreactive and non-immunoreactive axons in the brachium of the cat inferior colliculus, and demonstrate that GABA-immunoreactive axons are among the largest of brachial axons. We propose that, based on their large size, GABAergic tectothalamic axons provide a substrate for rapid feedforward inhibition of geniculate neurons.

GABA- and glycine-immunoreactive projections from the superior olivary complex to the cochlear nucleus in guinea pig.

Retrograde transport of horseradish peroxidase was combined with immunocytochemistry to identify the origins of potential gamma-aminobutyric acid (GABA) -ergic and glycinergic inputs to different subdivisions of the cochlear nucleus. Projection neurons in the inferior colliculus, superior olivary complex, and contralateral cochlear nucleus were examined, but only those from the superior olivary complex contained significant numbers of GABA- or glycine-immunoreactive neurons. The majority of these were in periolivary nuclei ipsilaterally, with a sizeable contribution from the contralateral ventral nucleus of the trapezoid body. Overall, 80% of olivary neurons projecting to the cochlear nucleus were immunoreactive for GABA, glycine, or both. Most glycine-immunoreactive projection neurons were located ipsilaterally, in the lateral and ventral nuclei of the trapezoid body and the dorsal periolivary nucleus. This suggests that glycine is the predominant neurotransmitter used by ipsilateral olivary projections. Most GABA-immunoreactive cells were located bilaterally in the ventral nuclei of the trapezoid body. The contralateral olivary projection was primarily GABA-immunoreactive and provided almost half the GABA-immunoreactive projections to the cochlear nucleus. This suggests that GABA is the predominant neurotransmitter used by contralateral olivary projections. The present results suggest that the superior olivary complex is the most important extrinsic source of inhibitory inputs to the cochlear nucleus. Individual periolivary nuclei differ in the strength and the transmitter content of their projections to the cochlear nucleus and may perform different roles in acoustic processing in the cochlear nucleus.

Patterns of gamma-aminobutyric acid and glycine immunoreactivities reflect structural and functional differences of the cat lateral lemniscal nuclei.

The three nuclei of the cat lateral lemniscus (dorsal, intermediate, and ventral) were distinguished by their immunoreactivities for the putative inhibitory transmitters, gamma-aminobutyric acid (GABA) and glycine. Each nucleus had a distinct pattern of somatic and perisomatic labeling. The dorsal nucleus contained mostly GABA-immunoreactive neurons (85%), with moderate numbers of GABA- and glycine-immunoreactive puncta along their somata. The remaining neurons were nonimmunoreactive (15%). The intermediate nucleus contained mostly nonimmunoreactive neurons (82%), and these had numerous glycine-immunoreactive and few GABA-immunoreactive perisomatic puncta. The remaining neurons were immunoreactive for GABA only (10%), glycine only (2%), or both (6%). The ventral nucleus contained mostly glycine-immunoreactive neurons (81%), and about half of these were also GABA-immunoreactive. The remaining neurons were either nonimmunoreactive (8%) or GABA-immunoreactive only (11%). Neurons in the ventral nucleus had fewer immunoreactive perisomatic puncta than neurons in either the dorsal or the intermediate nuclei. These differences in neuronal immunoreactivity and in the relative abundance of GABA-and glycine-immunoreactive perisomatic puncta among the three nuclei of the lateral lemniscus support connectional and electrophysiological evidence that each nucleus has a different functional role in auditory processing. In particular, this study demonstrates that the intermediate nucleus of the cat is cytochemically distinct from the dorsal and ventral nuclei in terms of the somatic and perisomatic immunoreactivity of its neurons for these two important inhibitory transmitters and may provide novel inputs to the inferior colliculus.

Glutamatergic connections of the auditory midbrain: selective uptake and axonal transport of D-[3H]aspartate.

D-[3H]aspartate was used to identify potential glutamatergic connections of the chinchilla inferior colliculus (IC). High-affinity uptake of D-[3H]aspartate is considered a selective marker for glutamatergic synapses, and neurons retrogradely labeled from such injections are believed to use glutamate, or a closely related compound, as a transmitter. Injections of D-[3H]aspartate suggest that glutamatergic endings in the IC arise primarily from intrinsic connections, the opposite IC, layer 5 of temporal cortex, nucleus sagulum, and lateral lemniscal nuclei. Neurons giving rise to the principal sensory (lemniscal) projections to the IC, i.e., those from the cochlear nuclei, superior olive, and the majority of projections from the lateral lemniscal nuclei, did not label in these experiments, indicating that their synapses do not recognize D-[3H]aspartate as a suitable substrate and may use inhibitory or other excitatory transmitters. After IC injections, fiber and diffuse labeling was found ipsilaterally in the medial geniculate body, superior colliculus, and dorsolateral pontine nuclei, contralaterally in the IC, and bilaterally in the superior olive and cochlear nuclei. Such labeling was attributed to anterograde transport of D-[3H]aspartate within the efferent collaterals of labeled IC neurons, suggesting that many of the IC's efferent projections may also be glutamatergic. This interpretation was confirmed in separate experiments in which D-[3H]aspartate, injected in the medial geniculate body, retrogradely labeled neurons in the IC as well as in layer 6 of temporal cortex. Finally, the mesencephalic trigeminal nucleus and tract labeled in some cases and may have local glutamatergic connections.

GABAergic feedforward projections from the inferior colliculus to the medial geniculate body.

A novel and robust projection from gamma-aminobutyric acid-containing (GABAergic) inferior colliculus neurons to the media] geniculate body (MGB) was discovered in the cat using axoplasmic transport methods combined with immunocytochemistry. This input travels with the classical inferior colliculus projection to the MGB, and it is a direct ascending GABAergic pathway to the sensory thalamus that may be inhibitory. This bilateral projection constitutes 10-30% of the neurons in the auditory tectothalamic system. Studies by others have shown that comparable input to the corresponding thalamic visual or somesthetic nuclei is absent. This suggests that monosynaptic inhibition or disinhibition is a prominent feature in the MGB and that differences in neural circuitry distinguish it from its thalamic visual and somesthetic counterparts.

Morphology of GABAergic neurons in the inferior colliculus of the cat.

The goal of the present study was to provide a comprehensive and quantitative description of neurons immunoreactive for gamma-aminobutyric acid (GABA) in the inferior colliculus (IC) of the cat. Neurons were investigated with two different antisera and two different incubation methods. Free-floating frozen or vibratome-cut sections were incubated either with an antiserum to glutamic acid decarboxylase (GAD) or to GABA conjugated to protein with glutaraldehyde. Additional 1.5-microns-thick sections were incubated with the GABA antiserum after embedding and removal of the plastic. Quantitative data were obtained from much of this material. Despite the use of these different antisera and reaction methods, the results obtained were remarkably similar. The results show that GAD- or GABA-positive neurons represent a significant population of cells in the central nucleus of the IC, up to 20% of the neurons. Most of these neurons have large or medium-sized perikarya. In contrast, immunonegative neurons are medium-sized or small. Many GABA-positive neurons had proximal dendrites or somata oriented in parallel to the fibrodendritic laminae of the central nucleus and are presumed to be disc-shaped neurons. Other have an orthogonal orientation and are presumed to be stellate cells. Large GABA-positive neurons form two groups, those with many axosomatic endings and those with few. Collectively, these observations suggest that there are several types of GABAergic neuron in the central nucleus and, by extension, that these may participate in many types of inhibitory circuits.

Glycine immunoreactive projections from the dorsal to the anteroventral cochlear nucleus.

The aim of the present study was to investigate whether projections from the dorsal cochlear nucleus (DCN) to the anteroventral cochlear nucleus (AVCN) use either of two inhibitory transmitters, glycine or GABA. Retrograde HRP labeling of DCN-to-AVCN projection neurons was combined with postembedding immunocytochemistry in the DCN of guinea pigs. Following injections of HRP in the anterior or posterior divisions of AVCN, large numbers of neurons were labeled in the DCN. All of these were located in the deep layer, except for a few granule cells. Nearly all (96%) of the projection neurons were immunoreactive for glycine and most had dendritic and somatic morphologies corresponding to those of elongate neurons (so-called 'corn' cells); only a few resembled small stellate neurons. Few (3%) retrogradely labeled neurons were immunoreactive for GABA. The results suggest that projections from the deep DCN to the AVCN are formed primarily by glycinergic elongate neurons. These projections could have a substantial inhibitory influence on the output of neurons in the AVCN.

Neurotransmitter-specific uptake and retrograde transport of [3H]glycine from the inferior colliculus by ipsilateral projections of the superior olivary complex and nuclei of the lateral lemniscus.

Neurotransmitter-specific uptake and retrograde axonal transport of [3H]glycine were used to identify glycinergic projections to the inferior colliculus in chinchillas and guinea pigs. Six h after injection of [3H]glycine in the inferior colliculus, autoradiographically labeled cells were found ipsilaterally in the ventral nucleus of the lateral lemniscus, the lateral superior olive and the dorsomedial periolivary nucleus. These 3 regions accounted for 95% of the labeled projection neurons, with the remainder scattered elsewhere in the ipsilateral superior olivary complex. No labeled cells were found contralaterally even after survival times as long as 24 h. Retrograde transport of HRP from the inferior colliculus in these same cases confirmed the presence of additional projections that did not accumulate [3H]glycine. These included ipsilateral projections from the medial superior olive and cochlear nucleus and contralateral projections from the inferior colliculus, dorsal nucleus of the lateral lemniscus, lateral superior olive, periolivary nuclei and cochlear nucleus. The results implicate uncrossed projections from the ventral nucleus of the lateral lemniscus, lateral superior olive, and dorsomedial periolivary nucleus as the principal sources of inhibitory glycinergic inputs to the inferior colliculus.

The form and distribution of GABAergic synapses on the principal cell types of the ventral cochlear nucleus of the cat.

The distribution of GABAergic endings was examined histochemically in the ventral cochlear nucleus (VCN) of the cat using an antibody to glutamate decarboxylase (GAD), the synthetic enzyme for GABA. Immunoreactive (GAD+) endings appeared in all subdivisions of the cat VCN. Each of the principal cell types had a characteristic labeling pattern, based on the size, concentration, and distribution of GAD+ endings on its soma. Spherical bushy cell somata were typically contacted by many small (less than 1.5 microns in diameter) and medium-sized (1.5-2 microns in diameter) endings, many of which aggregated into tight clusters. Globular bushy cells had a similar pattern, but the clusters of GAD+ endings were less tightly packed. Reactive endings on stellate cells were more evenly distributed. GAD+ endings on octopus cells were larger (up to 2.5 microns in diameter) than those on the bushy cells and tended to aggregate into small clusters or rows on the somata and dendrites. Reactive endings contained small pleomorphic vesicles and formed symmetrical synaptic contacts on each of the cell types examined. The patterns formed by GAD+ endings on each type of neuron resemble those of certain types of non-cochlear axons previously described with the Golgi methods as projecting from the dorsal cochlear nucleus and the trapezoid body.

Glycine-immunoreactive projection of the cat lateral superior olive: possible role in midbrain ear dominance.

Neurons in the lateral superior olive are optimally excited by stimulation of the ipsilateral ear, as are those in the inferior colliculus by stimulation of the contralateral ear. This reversal of ear dominance may result, in part, from distinct crossed excitatory and uncrossed inhibitory pathways ascending from the lateral superior olive. To explore this possibility, immunoreactivity for two putative inhibitory neurotransmitters, glycine and GABA, was examined in projection neurons that retrogradely transported horseradish peroxidase from the cat inferior colliculus. The results suggest that the projection from the lateral superior olive can be segregated, immunocytochemically, into three components: 1) a crossed, glycine-negative (-) projection; 2) an uncrossed, glycine-positive (+) projection; and 3) an uncrossed, glycine(-) projection. Additional evidence suggests that the terminal fields of the two uncrossed projections may distribute differently within the inferior colliculus. Glycine(+) or glycine(-) projection neurons, crossed or uncrossed, do not differ in the size, shape, or location of their somata. However, most glycine(-) neurons are heavily encrusted with glycine(+) endings; glycine(+) neurons have 40-60% fewer of these endings. Glycine(-) neurons located in the lateral limb have fewer glycine (+) perisomatic endings than those in the medial limb. Few projection neurons are GABA(+), and GABA(+) perisomatic endings are rare in the lateral superior olive. Thus, there is a heavy uncrossed projection from the cat lateral superior olive to the inferior colliculus that may be glycinergic and inhibitory. Furthermore, there is a bilateral projection that is not glycinergic or GABAergic, which may be excitatory. The potential contribution of these pathways to contralateral ear dominance in the inferior colliculus is discussed.

The morphology and synaptic connections of spiny stellate neurons in monkey visual cortex (area 17): a Golgi-electron microscopic study.

Based on a gold-toning, Golgi-electron microscope examination of 12 small and medium-sized spiny stellate neurons in laminae 4A, 4B, and 4C of the monkey visual cortex (area 17), the ultrastructure of the cell somata, dendrites, and axons of these neurons is described. Particular attention is paid to the synapses involving the surface of different parts of these neurons. Only symmetric synapses occur on the somata of spiny stellate neurons, and these occur with a frequency of 11.0-15.9 synapses/100 microns2 perikaryal surface. Symmetric synapses also occur on dendritic shafts and, occasionally, on dendritic spines. Asymmetric synapses are occasionally present along the dendritic shafts of spiny stellate neurons, but the majority of asymmetric synapses (75-95%) occur on their dendritic spines. The initial axon segments of the smallest spiny stellate neurons possess no axo-axonal synapses, but several symmetric synapses are present along the initial segment of a medium-sized, spiny stellate neuron in layer 4B. Fifty-three synapses made by boutons of the axons of these spiny stellate neurons have been identified, and all are asymmetric. Sixty per cent of the synapses are formed by boutons en passant and the remainder by the terminal swellings of spine-like axonal appendages, boutons terminaux. Of the synapses formed by the axons of spiny stellate cells, axo-spinous synapses outnumber axo-dendritic synapses two to one, and axo-dendritic synapses involve both spinous and aspinous dendrites. Evidence is presented which suggests that many of the axon terminals forming asymmetric synapses with the dendritic shafts and spines of spiny stellate neurons are derived from other spiny stellate neurons.

Interneuronal and glial-neuronal gap junctions in the lamina ganglionaris of the compound eye of the housefly, Musca domestica.

The cell-body layer of the lamina ganglionaris of the housefly, Musca domestica, contains the perikarya of five types of monopolar interneuron (L1-L5) along with their enveloping neuroglia (Strausfeld 1971). We confirm previous reports (Trujillo-Cenóz 1965; Boschek 1971) that monopolar cell bodies in the lamina form three structural classes: Class I, Class II, and midget monopolar cells. Class-I cells (L1 and L2) have large (8-15 microns) often crescent-shaped cell bodies, much perinuclear cytoplasm and deep glial invaginations. Class-II cells (L3 and L4) have smaller perikarya (4-8 microns) with little perinuclear cytoplasm and no glial invaginations. The 'midget' monopolar cell (L5) resides at the base of the cell-body layer and has a cub-shaped cell body. Though embedded within a reticulum of satellite glia, the L1-L4 monopolar perikarya and their immediately proximal neurites frequently oppose each other directly. Typical arthropod (beta-type) gap junctions are routinely observed at these interfaces. These junctions can span up to 0.8 micron with an intercellular space of 2-4 nm. The surrounding nonspecialized interspace is 12-20 nm. Freeze-fracture replicas of monopolar appositions confirm the presence of beta-type gap junctions, i.e., circular plaques (0.15-0.7 micron diam.) of large (10-15 nm) E-face particles. Gap junctions are present between Class I somata and their proximal neurites, between Class I and Class II somata and proximal neurites, and between Class II somata. Intercartridge coupling may exist between such monopolar somata. The cell body and proximal neurite of L5 were not examined. We also find that Class I and Class II somata are extensively linked to their satellite glia via gap junctions. The gap width and nonjunctional interspace between neuron and glia are the same as those found between neurons. The particular arrangement and morphology of lamina monopolar neurons suggest that coupling or low resistance pathways between functionally distinct neurons and between neuron and glia are probably related to the metabolic requirements of the "nuclear" layer and may play a role in wide field signal averaging and light adaptation.