Search

Immune molecules have been discovered recently to play critical roles in the development, function, and plasticity of the cerebral cortex. MHC class I (MHCI) molecules are expressed in the central nervous system and regulate activity-dependent refinement of visual projections during late postnatal development. They have also been implicated in neurodevelopmental diseases such as schizophrenia and autism. Despite the excitement generated by these unique roles for immune proteins in the brain, little is known about how these molecules regulate cortical connections. The first step toward elucidating the mechanism is to identify the spatial and temporal distribution of MHCI proteins throughout development. Using a pan-specific antibody that recognizes many MHCI variants for biochemistry and immunohistochemistry, we found that MHCI proteins are expressed in the rat visual cortex at all ages examined-during the peak of synaptogenesis, the critical period of synaptic refinement, and adulthood. Their abundance in the cortex peaked during early postnatal development, declining during periods of plasticity and adulthood. In contrast to current assumptions, pre- and postembedding immunogold electron microscopy (EM) revealed that MHCI proteins were present both pre- and postsynaptically at all ages examined. They were often found in the postsynaptic density and were closely associated with synaptic vesicles in the presynaptic terminal. These results suggest a previously undescribed model in which MHCI molecules function on both sides of the synapse to regulate connectivity in the mammalian visual cortex before, during, and after the establishment of connections.

New neurons are continuously generated in restricted regions of the adult mammalian brain. Although these adult-born neurons have been shown to receive synaptic inputs, little is known about their synaptic outputs. Using retrovirus-mediated birth-dating and labeling in combination with serial section electron microscopic reconstruction, we report that mossy fiber en passant boutons of adult-born dentate granule cells form initial synaptic contacts with CA3 pyramidal cells within 2 weeks after their birth and reach morphologic maturity within 8 weeks in the adult hippocampus. Knockdown of Disrupted-in-Schizophrenia-1 (DISC1) in newborn granule cells leads to defects in axonal targeting and development of synaptic outputs in the adult brain. Together with previous reports of synaptic inputs, these results demonstrate that adult-born neurons are fully integrated into the existing neuronal circuitry. Our results also indicate a role for DISC1 in presynaptic development and may have implications for the etiology of schizophrenia and related mental disorders.

Differential composition of GABA(A) receptor (GABA(A)R) subunits underlies the variability of fast inhibitory synaptic transmission; alteration of specific GABA(A)R subunits in localized brain regions may contribute to abnormal brain states such as absence epilepsy. We combined immunocytochemistry and high-resolution ImmunoGold electron microscopy to study cellular and subcellular localization of GABA(A)R alpha1, alpha3, and beta2/beta3 subunits in ventral posterior nucleus (VP) and reticular nucleus (RTN) of control rats and WAG/Rij rats, a genetic model of absence epilepsy. In control rats, alpha1 subunits were prominent at inhibitory synapses in VP and much less prominent in RTN; in contrast, the alpha3 subunit was highly evident at inhibitory synapses in RTN. beta2/beta3 subunits were evenly distributed at inhibitory synapses in both VP and RTN. ImmunoGold particles representing all subunits were concentrated at postsynaptic densities with no extrasynaptic localization. Calculated mean number of particles for alpha1 subunit per postsynaptic density in nonepileptic VP was 6.1 +/- 3.7, for alpha3 subunit in RTN it was 6.6 +/- 3.4, and for beta2/beta3 subunits in VP and RTN the mean numbers were 3.7 +/- 1.3 and 3.5 +/- 1.2, respectively. In WAG/Rij rats, there was a specific loss of alpha3 subunit immunoreactivity at inhibitory synapses in RTN, without reduction in alpha3 subunit mRNA or significant change in immunostaining for other markers of RTN cell identity such as GABA or parvalbumin. alpha3 immunostaining in cortex was unchanged. Subtle, localized changes in GABA(A)R expression acting at highly specific points in the interconnected thalamocortical network lie at the heart of idiopathic generalized epilepsy.

A key step in glutamatergic synapse maturation is the replacement of developmentally expressed N-methyl-D-aspartate receptors (NMDARs) with mature forms that differ in subunit composition, electrophysiological properties and propensity to elicit synaptic plasticity. However, the mechanisms underlying the removal and replacement of synaptic NMDARs are poorly understood. Here we demonstrate that NMDARs containing the developmentally regulated NR3A subunit undergo rapid endocytosis from the dendritic plasma membrane in cultured rat hippocampal neurons. This endocytic removal is regulated by PACSIN1/syndapin1, which directly and selectively binds the carboxy-terminal domain of NR3A through its NPF motifs and assembles a complex of proteins including dynamin and clathrin. Endocytosis of NR3A by PACSIN1 is activity dependent, and disruption of PACSIN1 function causes NR3A accumulation at synaptic sites. Our results reveal a new activity-dependent mechanism involved in the regulation of NMDAR expression at synapses during development, and identify a brain-specific endocytic adaptor that confers spatiotemporal and subunit specificity to NMDAR endocytosis.

Plexin signaling is required for stereotyped pruning of long axon collaterals in the vertebrate CNS; however, a cellular basis for plexins on stereotyped pruning has not been determined. Using quantitative electron microscopy and immunocytochemistry, we found that infrapyramidal mossy fiber axon collaterals form transient synaptic complexes with basal dendrites of CA3 pyramidal cells in the early postnatal mouse hippocampus. At later postnatal ages, these synaptic complexes stop maturing and are removed before stereotyped pruning by a mechanism that does not involve axon degeneration and glial cell engulfment. In knock-out mice that lack plexin-A3 signaling, the synaptic complexes continue to mature, and, as a result, the collaterals are not pruned. Thus, our results suggest that intact plexin-A3 signaling contributes to synaptic complex elimination, which is associated with stereotyped axon pruning.

Switching of the NMDA receptor 2A (NR2A) and NR2B subunits at NMDA receptors is thought to underlie the functional changes that occur in NMDA receptor properties during the developmental epoch when neural plasticity is most pronounced. The cellular expression of NR2A and NR2B and the NR2 synaptic binding protein postsynaptic density-95 (PSD-95) was examined in the mouse somatosensory cortex and thalamus from postnatal day 2 (P2) to P15 using reverse transcription-PCR, in situ hybridization histochemistry, and immunocytochemistry. The localization of NR2A and NR2B subunits and PSD-95 was then studied at synapses in layer IV of somatosensory cortex and in the ventral posterior nucleus of the thalamus using high-resolution immunoelectron microscopy. At both cortical and thalamic synapses, a quantitative switch in the dominant synaptic subunit from NR2B to NR2A was accompanied by a similar change in the cellular expression of NR2A but not of NR2B. Synaptic PSD-95 developed independently, although both NR2A and NR2B colocalized with PSD-95. Displacement of NR2B subunits from synapses was not accompanied by an increase in an extrasynaptic pool of this subunit. Thus, the switch in synaptic NR2 subunit predominance does not occur by changes in expression or displacement from synapses and may reflect the formation of new synapses from which NR2B is lacking.

The trafficking of glutamate receptors in neurons is of the utmost importance for synapse formation and synaptic plasticity. Recently, we demonstrated that both NMDA and AMPA receptors reside in mobile transport packets that are recruited rapidly and independently to nascent synapses. Here, we show that a large proportion of the glutamate receptor clusters in young cortical neurons are present on the surface of dendrites before synapses are formed and these surface-exposed transport packets are mobile. Exocytosis of glutamate receptors to the dendritic surface occurs via a SNARE [soluble n-ethylmaleimide-sensitive factor attachment protein (SNAP) receptor]-dependent SNAP-23-mediated mechanism. Endocytosis occurs rapidly after surface exposure; >50% of surface-labeled NMDA receptors (NMDARs) are endocytosed within 5 min. NMDARs are transported along microtubules on large tubulovesicular organelles, as indicated by immunoelectron microscopy, and are associated with EEA1 (early endosomal antigen 1) and SAP102 (synapse-associated protein 102), as indicated by immunocytochemistry. Most surprisingly, a large proportion of these transport packets cycle through the dendritic plasma membrane before synapse formation. These results suggest a novel model in which NMDARs cycle with the plasma membrane during pauses of movement along microtubules while trafficking.

Low-frequency thalamocortical oscillations that underlie drowsiness and slow-wave sleep depend on rhythmic inhibition of relay cells by neurons in the reticular nucleus (RTN) under the influence of corticothalamic fibers that branch to innervate RTN neurons and relay neurons. To generate oscillations, input to RTN predictably should be stronger so disynaptic inhibition of relay cells overcomes direct corticothalamic excitation. Amplitudes of excitatory postsynaptic conductances (EPSCs) evoked in RTN neurons by minimal stimulation of corticothalamic fibers were 2.4 times larger than in relay neurons, and quantal size of RTN EPSCs was 2.6 times greater. GluR4-receptor subunits labeled at corticothalamic synapses on RTN neurons outnumbered those on relay cells by 3.7 times, providing a basis for differences in synaptic strength.

Quantitative electron microscopy was used to examine the relative contributions of different types of synapses to the circuitry of the thalamic reticular nucleus (RTN) in the rat. Single RTN cells were injected with Lucifer Yellow (LY) in fixed brain slices and examined after photoconversion; corticothalamic axons and terminals were labeled by anterograde transport of Phaseolus vulgaris-leucoagglutinin (PHA-L); and gamma-aminobutyric acid (GABA)ergic terminals were labeled by postembedding immunocytochemistry. Three types of synapses, made by morphologically distinguishable small terminals (ST), large terminals (LT), and GABAergic terminals, were distributed on all portions of the dendritic trees of injected RTN cells. ST and LT terminals formed asymmetrical, presumed excitatory, synaptic contacts. On proximal dendrites, approximately 50% of the synapses were ST, 30-40% were LT, and 10-25% were GABAergic. On distal dendrites, 60-65% were ST, 20% were LT, and 15% were GABAergic. PHA-L labeling showed that labeled corticothalamic terminals and ST terminals have identical morphological features and the same distribution patterns on RTN dendrites, indicating that the majority of excitatory afferents to RTN neurons are derived from the cerebral cortex. The LT terminals found in smaller numbers are probably derived from collateral axons of thalamocortical relay cells. GABAergic terminals formed by LY-labeled, intra-RTN axon collaterals were relatively few in number, and no dendrodendritic synapses were observed., (Copyright 1999 Wiley-Liss, Inc.)

In mouse thalamocortical slices in vitro, the potassium channel blocker 4-AP and GABAA receptor antagonist bicuculline together induced spontaneous prolonged depolarizations in layer VI neurons from postnatal day 2 (P2), in ventroposterior nucleus neurons (VP) from P7, and in reticular nucleus neurons (RTN) from P8. Dual whole-cell recordings revealed that prolonged bursts were synchronized in layer VI, VP, and RTN. Bursts were present in cortex isolated from thalamus, but not in thalamus isolated from cortex, indicating that bursts originated in cortex and propagated to thalamus. Prolonged bursts were synchronized in layer VI when vertical cuts extended from pia mater through layers IV or V, but were no longer synchronized when cuts extended through layer VI and white matter. In voltage-clamp recordings before P10, burst conductance of all three neuronal populations was dominated by the NMDA receptor-mediated conductance, and therefore synapses were "silent". In cortex and RTN, after P10, bursts were associated with strong AMPA/kainate receptor-mediated conductances, and synapses had become "functional"; silent synapses persisted in a large proportion of VP cells after P10. Before P9, the NMDA receptor antagonist APV or the non-NMDA receptor antagonist CNQX blocked the prolonged bursts. After P9, CNQX continued to block the prolonged bursts, but APV merely shortened their duration. Thus, NMDA receptor-based silent synapses are essential for paroxysmal corticothalamic activity during early postnatal development, and connections between layer VI neurons are sufficient for horizontal cortical synchronization.

The development of receptor function at corticothalamic synapses during the first 20 days of postnatal development is described. Whole cell excitatory postsynaptic currents (EPSCs) were evoked in relay neurons of the ventral posterior nucleus (VP) by stimulation of corticothalamic fibers in in vitro slices of mouse brain from postnatal day 1 (P1). During P1-P12, excitatory postsynaptic conductances showed strong voltage dependence at peak current and at 100 ms after the stimulus and were almost completely antagonized by -2-amino-5-phosphonopentoic acid (APV), indicating that N-methyl--aspartate (NMDA) receptor-mediated currents dominate corticothalamic EPSCs at this time. After P12, in 42% of cells, excitatory postsynaptic conductances showed no voltage-dependence at peak current but still showed voltage-dependence 100-ms poststimulus. This voltage-dependent conductance was antagonized by APV. The nonvoltage-dependent component was APV resistant, showed fast decay, and was antagonized by the nonNMDA antagonist, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX). In the remaining 58% of cells after P12, excitatory postsynaptic conductances showed moderate voltage dependence at peak conductance and strong voltage dependence 100 ms after the stimulus. Analysis of EPSCs before and after APV showed a significant increase in the relative contribution of the non-NMDA conductance after the second postnatal week. From P1 to P16, there was a significant decrease in the time constant of decay of the NMDA EPSC but no change in the voltage dependence of the NMDA response. After P8, slow EPSPs, 1.5-30 s in duration and mediated by metabotropic glutamate receptors (mGluRs), could be evoked by high-frequency stimulation of corticothalamic fibers in the presence of APV and CNQX. Similar slow depolarizations could be evoked by local application of the mGluR agonist (+/-)-1-aminocyclopentane-trans-1,3-dicarboxylic acid (t-ACPD) but from P0. Both conductances were blocked by the mGluR antagonist, (RS)-alpha-methyl-4-carboxyphenylglycine. Hence functional mGluR receptors are present on VP cells from birth, but their synaptic activation at corticothalamic synapses can only be detected after P8. In voltage clamp, the extrapolated reversal potential of the t-ACPD current, with potassium gluconate-based internal solution, was +12 +/- 10 (SE) mV, and the measured reversal potential with cesium gluconate-based internal solution was 1.5 +/- 9.9 mV, suggesting that the mGluR-mediated depolarization was mediated by a nonselective cation current. Replacement of NaCl in the external solution caused the reversal potential of the current to shift to -18 +/- 2 mV, indicating that Na+ is a charge carrier in the current. The current amplitude was not reduced by application of Cs+, Ba2+, and Cd2+, indicating that the t-ACPD current was distinct from the hyperpolarization-activated cation current (IH) and distinct from certain other previously characterized mGluR-activated, nonselective cation conductances.

It is widely believed that inhibitory synapses are not present or present in only small numbers in the rodent cerebral cortex during the early postnatal period when the cortex is being innervated by thalamocortical fibers. Quantitative electron microscopy was carried out on the posteromedial barrel subfield of mouse somatosensory cortex from postnatal day 4 (P4) when thalamocortical innervation of the barrels is becoming established, through to sexual maturity (>P32), and in adulthood. Both asymmetrical (putatively excitatory) and symmetrical (putatively inhibitory) synapses were present in all layers from P4. The symmetrical synapses were immunoreactive for GABA at all ages. There was a progressive increase in both asymmetrical and symmetrical synapses up to P32, density in all layers increasing 16-fold, with the production of asymmetrical synapses leading and greatly outstripping that of symmetrical. From P32 to P120, the oldest age studied, synaptic numbers declined by 18% to 13 times the P4 level, but this affected predominantly layers II/III, IV and V, and mainly involved asymmetrical synapses. The relative percentage of asymmetrical to symmetrical synapses from P4 to P8 was 57%/43% but at P32 it was 89.5%/10.5% and in adulthood 85.4%/14.6%. These data indicate that inhibitory synaptogenesis in the rodent cortex begins earlier than previously thought, a basis for inhibition being present from the earliest period. Pruning of all synapses occurs well after thalamocortical innervation is established and inhibitory synapses are less affected by the pruning process.

Pre-embedding immunoperoxidase staining for CAM II kinase-alpha and post-embedding immunogold staining for glutamate and GABA, were used to reveal the subcellular distribution of CAM II kinase-alpha at transmitter-characterized synapses in the CA1 region of rat hippocampus. Immunoelectron microscopy showed that the majority of CAM II kinase-alpha-immunostained neuronal profiles were dendritic spines presumably derived from pyramidal cells. CAM II kinase-alpha immunoreactivity was mainly localized in postsynatic densities associated with glutamatergic axon terminals. No CAM II kinase-alpha immunoreactivity was detected in GABA-immunoreactive profiles or at GABAergic synapses. This study provides morphological evidence that CAM II kinase-alpha is involved only in excitatory neuronal transmission in the CA1 region. The enzyme is unlikely to be involved in plasticity at GABA synapses.

The alpha subunit of type II calcium/calmodulin-dependent protein kinase (CAM II kinase-alpha) plays an important role in longterm synaptic plasticity. We applied preembedding immunocytochemistry (for CAM II kinase-alpha) and postembedding immunogold labeling [for glutamate or gamma-aminobutyric acid (GABA)] to explore the subcellular relationships between transmitter-defined axon terminals and the kinase at excitatory and inhibitory synapses in thalamus and cerebral cortex. Many (but not all) axon terminals ending in asymmetric synapses contained presynaptic CAM II kinase-alpha immunoreactivity; GABAergic terminals ending in symmetric synapses did not. Postsynaptically, CAM II kinase-alpha immunoreactivity was associated with postsynaptic densities of many (but not all) glutamatergic axon terminals ending on excitatory neurons. CAM II kinase-alpha immunoreactivity was absent at postsynaptic densities of all GABAergic synapses. The findings show that CAM II kinase-alpha is selectively expressed in subpopulations of excitatory neurons and, to our knowledge, demonstrate for the first time that it is only associated with glutamatergic terminals pre- and postsynaptically. CAM II kinase-alpha is unlikely to play a role in plasticity at GABAergic synapses.

This study was aimed at determining the synaptic circuitry that contributes to the alterations in thalamic function that accompany changes in behavioral states. The somatosensory sector of the thalamic reticular nucleus (RTN) was identified by microelectrode recording in cats and injected with Phaseolus vulgaris-leucoagglutinin (PHA-L). The axons of labeled RTN cells gave rise to collaterals within the RTN and continued into the dorsal thalamus where they terminated predominately in the ventral posterior lateral nucleus (VPL). After small injections in the upper limb representation of RTN, most labeled terminations in VPL were confined to its medial part, suggesting the presence of a topographic organization in the projection. Terminations were concentrated in localized, focal aggregations of boutons. Combined electron microscopic immunocytochemistry, using immunogold labeling for gamma-aminobutyric acid (GABA), showed that the PHA-L labeled boutons were GABA-positive terminals that ended in symmetrical synapses. Eighty-two percent of these synapses were on dendrites of relay neurons, 8.5% on dendrites of interneurons, and 9.3% on somata. The terminals of RTN axons form the majority of axon terminals ending in symmetrical synapses in VPL. Their concentration on relay neurons probably underlies the capacity of the RTN projection to reduce background activity of VPL relay neurons in the awake state and to maintain oscillatory behavior of these neurons in drowsiness and early phases of sleep.

This study was aimed at providing quantitative data on the thalamic circuitry that underlies the central processing of somatosensory information. Four physiologically identified thalamocortical relay neurons in the ventral posterior lateral nucleus (VPL) of the cat thalamus were injected with horseradish peroxidase and subjected to quantitative electron microscopy after pre- or postembedding immunostaining for gamma-aminobutyric acid to reveal synaptic terminals of thalamic inhibitory neurons. The four cells all had rapidly adapting responses to light mechanical stimuli applied to their receptive fields, which were situated on hairy or glabrous skin or related to a joint. Their dendritic architecture was typical of cells previously described as type I relay cells in VPL, and they lacked dendritic appendages. Terminals ending in synapses on the injected cells were categorized as RL (ascending afferent), F (inhibitory), PSD (presynaptic dendrite), and RS (mainly corticothalamic) types and were quantified in reconstructions of serial thin sections. RL and F terminals formed the majority of the synapses on proximal dendrites (approximately 50% each). The number of synapses formed by RL terminals declined on intermediate dendrites, but those formed by F terminals remained relatively high, declining to moderate levels (20-30%) on distal dendrites. RS terminals formed moderate numbers of the synapses on intermediate dendrites and the majority (> 60%) of the synapses on distal dendrites. Synapses formed by PSDs were concentrated on intermediate dendrites and were few in number (approximately 6%). They formed synaptic triads with F terminals and rarely with RL terminals. On somata, only a few synapses were found, all made by F terminals. The total number of synapses per cell was calculated to be 5,584-8,797, with a density of 0.6-0.9 per micrometer of dendritic length. Of the total, RL terminals constituted approximately 15%, F terminals approximately 35%, PSD terminals approximately 5%, and RS terminals approximately 50%. These results provide the first quantitative assessment of the synaptic architecture of thalamic somatic sensory relay neurons and show the basic organizational pattern exhibited by representatives of the physiological type of relay neurons most commonly encountered in the VPL nucleus.

Correlative light and electron microscopic immunocytochemical methods were used to analyze the 5-HT innervation of the primary auditory area (AI) of the cat cerebral cortex and to examine the synaptic relationships of 5-HT basket terminations on target neurons in that area. Three morphological types of 5-HT-immunoreactive fibers are present: type I, which is very thin and very finely beaded; type II, which is thin and coarsely beaded; and type III, which has a relatively thick main shaft and very few beads. Type I is the most abundant, type II is relatively less common, and type III is the least abundant type. The 3 types of fibers are present through the thickness of AI and in the subjacent white matter, but the densest plexus is found in layers I-III. One of the most characteristic features of type II fibers is that they commonly form small, dense clusters that resemble baskets apposed to the somata and primary dendrites of unstained neurons. The basket formations are more frequently found in layers I and II, and they vary in complexity. Simultaneous immunostaining for GABA and 5-HT reveals that many 5-HT baskets surround the somata and dendrites of GABA neurons. In 2-microns-thick plastic sections, each basket formation can be seen surrounding 1 or a group of 2 or 3 cells. In the latter case, one cell is much larger and at the electron microscope level is identified as a neuron, while the other cells are neuroglial cells. Reconstructions were made from serial electron micrographs of 135 5-HT-immunoreactive boutons. Of these boutons, 110 belonged to basket formations, 14 to type I axons located in the neuropil, and the remaining 11 to type II fibers located in the white matter. Only 4 of the 135 boutons made conventional synaptic contacts. These were of the asymmetrical type. Most of the boutons made very small, indistinct membrane specializations or none at all. The present results therefore suggest a strong interaction between 5-HT axon terminals and specific GABA neurons, which may be mediated by release sites that are not associated with morphologically distinct synaptic contacts.

Neurons in the monkey and rat cerebral cortex immunoreactive for somatostatin tetradecapeptide (SRIF) and for neuropeptide Y (NPY) were examined in the light and electron microscope. Neurons immunoreactive for either peptide are found in all areas of monkey cortex examined as well as throughout the rat cerebral cortex and in the subcortical white matter of both species. In monkey and rat cortex, SRIF-positive neurons are morphologically very similar to NPY-positive neurons. Of the total population of SRIF-positive and NPY-positive neurons in sensory-motor and parietal cortex of monkeys, a minimum of 24% was immunoreactive for both peptides. Most cell bodies are small (8 to 10 micron in diameter) and are present through the depth of the cortex but are densest in layers II-III, in layer VI, and in the subjacent white matter. From the cell bodies several processes commonly emerge, branch two or three times, become beaded, and extend for long distances through the cortex. The fields formed by these processes vary from cell to cell; therefore, the usual morphological terms bipolar, multipolar, and so on do not adequately characterize the full population of neurons. Virtually every cell, however, has at least one long vertically oriented process, and most processes of white matter cells ascent into the cortex. No processes could be positively identified with the light microscope as axons. The processes of the peptide-positive neurons form dense plexuses in the cortex. In each area of monkey cortex, SRIF-positive and NPY-positive processes form a superficial plexus in layers I and II and a deep plexus in layer VI. These plexuses vary in density from area to area. All appear to arise from cortical or white matter cells rather than from extrinsic afferents. In some areas such as SI and areas 5 and 7, the superficial plexus extends deeply into layers III and IV; and in area 17, two very prominent middle plexuses occur in layers IIIB through IVB and in the upper one-third of layer V; these are separated by layer IVC, a major zone of thalamic terminations, which contains very few SRIF- or NPY-positive processes. The density of the plexuses is greater for NPY-positive processes than for SRIF-positive processes in all areas. In the rat, the plexuses do not display a strict laminar organization but generally are densest in the supragranular layers (I to III) and decline steadily in the deeper layers.(ABSTRACT TRUNCATED AT 400 WORDS)

In the monkey neocortex, immunoreactivity for the 28-kDa vitamin D-dependent calcium binding protein (Calbindin) is contained in a set of GABAergic intrinsic neurons whose small size and laminar locations render them very distinct from a second set of GABAergic intrinsic neurons that show immunoreactivity for another calcium-binding protein, parvalbumin. A conspicuous feature of many calbindin-immunoreactive cells is their possession of long, vertically oriented bundles of immunoreactive processes that descend or ascend vertically through several cortical layers. These are components of the radial fasciculi of the cortex and are here shown by correlative electron microscopic immunocytochemistry to consist of both immunoreactive dendrites and unmyelinated axons. The morphology of the bundles and the parent cells indicates that the cells are classical double bouquet cells. The calbindin-positive axons in the radial fasciculi in the present study formed symmetric synapses on unlabeled dendritic shafts (62%) and spines (38%). Despite the close-packed nature of the immunoreactive axons, relatively few terminals of the same axon converged on a single postsynaptic profile. The postsynaptic profiles were identified in certain cases as side branches of pyramidal cell apical and basal dendrites. Mainstem apical dendrites generally did not receive synapses derived from the calbindin-positive axons. These results indicate that double bouquet cells can be distinguished both by their GABAergic character and by their possession of calbindin immunoreactivity. They are probably major contributors to the vertical flow of inhibitory influences across laminae of the cerebral cortex.

Neurons in the monkey somatic sensory and motor cortex were labelled immunocytochemically for the GABA synthesizing enzyme, glutamic acid decarboxylase (GAD), and examined with the electron microscope. The somata and dendrites of many large GAD-positive neurons of layers III-VI receive numerous asymmetric synapses from unlabelled terminals and symmetric synapses from GAD-positive terminals. Comparisons with light and electron microscopic studies of Golgi-impregnated neurons suggest that the large labelled neurons are basket cells. Small GAD-positive neurons generally receive few synapses on their somata and dendrites, and probably conform to several morphological types. GAD-positive axons from symmetric synapses on many neuronal elements including the somata, dendrites and initial segments of pyramidal cells, and the somata and dendrites of non-pyramidal cells. Synapses between GAD-positive terminals and GAD-positive cell bodies and dendrites are common in all layers. Many GAD-positive terminals in layers III-VI arise from myelinated axons. Some of the axons form pericellular terminal nests on pyramidal cell somata and are interpreted as originating from basket cells while other GAD-positive myelinated axons synapse with the somata and dendrites of non-pyramidal cells. These findings suggest either that the sites of basket cell terminations are more heterogeneous than previously believed or that there are other classes of GAD-positive neurons with myelinated axons. Unmyelinated GAD-positive axons synapse with the initial segments of pyramidal cell axons or form en passant synapses with dendritic spines and small dendritic shafts and are interpreted as arising from the population of small GAD-positive neurons which appears to include several morphological types.

Immunocytochemical and electron microscopic methods were used to examine neurons in regio superior of rat hippocampus displaying cholecystokinin octapeptide-like immunoreactivity. Cholecystokinin-immunoreactive synaptic terminals and somata are found in all layers of regio superior but are most numerous in stratum pyramidale. The vast majority of terminals form symmetric synaptic contacts onto the somata and proximal dendrites of hippocampal pyramidal cells and onto smaller dendrites which may also arise from pyramidal cells. A very small number of cholecystokinin-immunoreactive terminals form synapses that appear asymmetric and contact dendritic shafts or spines. The somata of some pyramidal cells receive symmetric synapses from cholecystokinin-immunoreactive terminals that are joined by cytoplasmic bridges to form parts of pericellular baskets. These and adjacent pyramidal cell somata are also contacted by terminals that are not immunoreactive for cholecystokinin. No cholecystokinin-positive terminals contacted the initial segments of pyramidal cell axons. Cholecystokinin-immunoreactive cells are found in all layers of regio superior. Their somata receive a few symmetric synapses, most of which are formed by terminals not immunoreactive for cholecystokinin. Their dendrites receive a greater number of both symmetric and asymmetric contacts, some of which are immunoreactive for cholecystokinin. We conclude the following: The localization of cholecystokinin immunoreactivity in synaptic terminals contacting the somata and dendrites of hippocampal pyramidal cells is consistent with the suggestion that cholecystokinin acts as a neurotransmitter at these sites and at sites in other parts of the cerebral cortex. Results from the present and previous studies suggest that cholecystokinin-like immunoreactivity may co-exist with gamma-aminobutyrate in some non-pyramidal neurons of regio superior. Cholecystokinin-immunoreactive terminals arise mainly from non-pyramidal cells intrinsic to the hippocampus, one class of which appears to be a type of basket cell.