Selective retrograde transport of D-aspartate in spinal interneurons and cortical neurons of rats

Kynurenines and Glutamate: Multiple Links and Therapeutic Implications

Glutamate is firmly established as the major excitatory neurotransmitter in the mammalian brain and is actively involved in most aspects of neurophysiology. Moreover, glutamatergic impairments are associated with a wide variety of dysfunctional states, and both hypo- and hyperfunction of glutamate have been plausibly linked to the pathophysiology of neurological and psychiatric diseases. Metabolites of the kynurenine pathway (KP), the major catabolic route of the essential amino acid tryptophan, influence glutamatergic activity in several distinct ways. This includes direct effects of these "kynurenines" on ionotropic and metabotropic glutamate receptors or vesicular glutamate transport, and indirect effects, which are initiated by actions at various other recognition sites. In addition, some KP metabolites affect glutamatergic functions by generating or scavenging highly reactive free radicals. This review summarizes these phenomena and discusses implications for brain physiology and pathology.

Neurotensin modulation of acetylcholine, GABA, and aspartate release from rat prefrontal cortex studied in vivo with microdialysis

The effects of the peptide transmitter neurotensin (NT) on the release of acetylcholine (ACh), gamma-aminobutyric acid (GABA), glutamate (Glu), aspartate (Asp), and taurine from the prefrontal cortex (PFC) of freely moving rats were studied by transversal microdialysis. Neurotensin (0.2 and 1 microM) administered locally in the PFC produced a concentration-dependent increase in the extracellular levels of ACh, GABA, and Asp, but not of Glu or taurine. The increase produced by 1 microM NT reached a maximum of about 240% for ACh, 370% for GABA, and 380% for Asp. Lower doses of NT (0.05 microM) did not cause a significant change in ACh, GABA, or Asp output in the PFC. Higher concentrations of NT (2 microM) did not induce further increases in the level of neurotransmitters. A high-affinity selective neurotensin receptor (NTR1) antagonist SR 48692 (0.5 microM) perfused locally blocked neurotensin (1 microM)-evoked ACh, GABA, and Asp release. Local infusion of the sodium channel blocker tetrodotoxin (TTX) (1 microM) decreased the release of ACh, had no significant effect on GABA or Asp release, and prevented the 1 microM neurotensin-induced increase in ACh, GABA, and Asp output. Removal of calcium from the Ringer's solution prevented the peptide from having any effects on the neurotransmitters. Thus, in vivo NT plays a modulatory role in the PFC by interacting with cortical neurons releasing GABA and Asp and with ACh-containing neurons projecting to the PFC. The NT effects are of neural origin, as they are TTX-sensitive, and mediated by the NTR1 receptor, as they are antagonized by SR 48692.

Primary somatosensory cortex modulation of tactile responses in nucleus gracilis cells of rats

Corticofugal influences from the primary somatosensory cortex to the gracilis nuclei were studied with single unit recordings performed in urethane-anaesthetized rats. Two types of neurons were identified: low firing rate (LF) neurons, which could be activated antidromically by medial lemniscus stimulation; and high firing rate (HF) neurons. The effects of electrically stimulating the contralateral primary somatosensory cortex were studied in two situations: when the stimulated cortical area and specific gracilis cells had overlapping receptive fields and when the receptive fields of the cells and primary somatosensory cortex did not overlap. Cortical stimulation facilitated cortical and tactile responses in most gracilis neurons (68% and 58% for LF and HF neurons, respectively) with overlapping receptive fields. When receptive fields were different, cortical stimulation inhibited tactile response in most LF neurons (58%) and some HF neurons (20%). Trains of cortical shocks during sensory stimulation demonstrated that the facilitatory and inhibitory effects outlasted the stimulation period by 5 min. The facilitatory effect was decreased by iontophoretic application of the N-methyl-D-aspartate (NMDA) receptor antagonist APV (50 mm). However, APV did not modify the intensity of the tactile response inhibition in cells with nonoverlapping receptive fields, although, its duration was decreased (<5 min). Iontophoretic application of the gamma-aminobutyric acid (GABA)(A) antagonist bicuculline (20 mm) blocked the cortically evoked inhibition in cells with nonoverlapping receptive fields. The results indicate that the somatosensory cortex precisely controls somatosensory transmission throughout the gracilis nucleus by means of NMDA and GABA(A) receptor activation.

The caudal ventrolateral medulla as an important inhibitory modulator of pain transmission in the spinal cord

The caudal ventrolateral medulla (VLM) has emerged during the last decade as one of the main components of the endogenous pain control system. Profound and long-lasting analgesia is produced by mild stimulation of the VLM. The VLMlat, the reticular formation located between the spinal trigeminal nucleus and the lateral reticular nucleus (LRt), appears to play a major role in that antinociceptive action. The projections to spinal cord laminae involved in nociceptive transmission originate exclusively in the VLMlat. The VLMlat participates in a disynaptic pathway involving spinally projecting pontine A5 noradrenergic neurons, which appears to convey alpha(2)-adrenoreceptor-mediated analgesia produced from the VLM. Neurons in the VLMlat and in lamina I are reciprocally connected by a closed loop that is likely to mediate feedback control of supraspinal nociceptive transmission. On the other hand, the LRt, which is targeted by ventral (lamina VII) and deep dorsal (laminae IV to V) horn inputs, projects to the premotor lamina VII. Nociceptive input ascending from the cord and increases in blood pressure are discussed as possible physiologic triggers of the analgesia produced by the VLM. The overall role of the VLM as a center for integration of nociceptive, cardiovascular, and motor functions is discussed. The putative therapeutic benefits of manipulating the VLM for the control of chronic pain are envisaged.

Possible glutamatergic/aspartatergic projections to the supramammillary nucleus and their origins in the rat studied by selective [(3)H]D-aspartate labelling and immunocytochemistry

The supramammillary neurons projecting directly to the hippocampus or indirectly via the septum participate in the regulation of hippocampal theta activity. Inputs to the supramammillary nucleus are only partly specified neurochemically. Glutamate appears to be an excitatory transmitter in this cell group, however, the origin of the glutamatergic afferents is unknown. The present investigations were devoted to study this question. The transmitter-selective [(3)H]D-aspartate retrograde transport method was used injecting the tracer into the lateral subregion of the nucleus. The radioactive tracer was visualized by autoradiography. Non-selective retrograde tracing experiments were also performed for reference injecting wheat germ agglutinin-conjugated colloidal gold into the same supramammillary region. Retrogradely radiolabelled neurons in various numbers were detected in several brain regions including medial septum-diagonal band complex, lateral septum, rostral part of medial and lateral preoptic areas, lateral habenula, ventral premammillary nucleus, apical subregion of interpeduncular nucleus, laterodorsal tegmental nucleus, and dorsal and median raphe nuclei. Radiolabelled neurons in the mentioned raphe nuclei were serotonin-immunonegative. In the non-selective retrograde tracing experiments combined with immunocytochemistry, about 50% of the retrogradely labelled neurons in the raphe nuclei was serotonin-immunonegative, showing that not only serotonergic raphe neurons project to the supramammillary nucleus. The findings indicate that a significant part of the afferents from telencephalic, diencephalic and brainstem regions to the supramammillary nucleus may contain glutamate/aspartate as neurotransmitter. The most important functional implications of these observations concern the role of the supramammillary nucleus in controlling the electrical activity of the hippocampus, and in particular the generation and maintenance of the theta rhythm.

Excitatory Amino Acid Pathway from the Hippocampus to the Prefrontal Cortex. Contribution of AMPA Receptors in Hippocampo-prefrontal Cortex Transmission

Previous experiments in the rat have demonstrated that field CA1 and the subiculum project to the prefrontal cortex and that this direct unilateral pathway is excitatory. In the present study, anatomical and electrophysiological approaches were used to determine the transmitter mediating the excitatory responses in prefrontal cortex neurons to low-frequency stimulation of the hippocampus. The method of selective retrograde d-[3H]aspartate labelling was used to identify putative glutamatergic and/or aspartatergic hippocampal afferent fibres to the prefrontal cortex. Unilateral microinjection of d-[3H]aspartate into the prelimbic area of the prefrontal cortex resulted in the retrograde labelling of a fraction of hippocampal neurons. Some labelled cell bodies were distributed in field CA1 and the subiculum but larger numbers of neurons were detected in the ventral and intermediary subiculum. In a second series of experiments, the excitatory transmission from the hippocampus to the prefrontal cortex was pharmacologically analysed to provide further evidence for the involvement of glutamate and/or aspartate in the pathway. All prefrontal cortex neurons responding to the stimulation of the hippocampus were activated by selective agonists of the glutamate receptor subtypes alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) and N-methyl-d-aspartate (NMDA), and these effects were selectively antagonized by 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and 2-amino-5-phosphonopentanoic acid (APV) respectively. Most of the excitatory responses of prefrontal cortex neurons to single and paired-pulse stimulation of the hippocampus were antagonized by CNQX. APV only affected the excitatory response in a few cells. These results suggest that the hippocampal input to the prefrontal cortex utilizes glutamate and/or aspartate as a transmitter. Even though prefrontal cortex neurons responding to the stimulation of the hippocampus appear to have both AMPA and NMDA receptors, low-frequency stimulation of the hippocampo-prefrontal cortex pathway activates cortical neurons mostly through AMPA receptors.

Properties and plasticity of synaptic inputs to rat dorsal column neurones recorded in vitro

1. The mechanisms regulating the flow of sensory signals and their modification by synaptic interactions in the dorsal column nuclei are incompletely understood. Therefore, we examined the interactions between EPSPs evoked by stimulation of dorsal column and corticofugal fibres in the dorsal column nuclei cells using an in vitro slice technique. 2. Dorsal column EPSPs had briefer durations at depolarised membrane potentials than corticofugal EPSPs. Superfusion of the NMDA receptor antagonist 2D(-)-2-amino-5-phosphonovaleric acid (AP5) did not modify dorsal column EPSPs but reduced corticofugal EPSPs. Application of the AMPA/kainate receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) abolished both dorsal column and corticofugal EPSPs in cells held at the resting potential. Therefore, dorsal column EPSPs were mediated by non-NMDA receptors but corticofugal EPSPs revealed both non-NMDA- and NMDA-dependent components. 3. Paired-pulse stimulation of dorsal column fibres elicited a depression of the second EPSP at pulse intervals of < 50 ms; however, paired-pulse stimulation of corticofugal fibres evoked facilitation of the second EPSP at pulse intervals of < 30 ms. When stimulation of the corticofugal fibres preceded stimulation of the dorsal column fibres, facilitation of the dorsal column EPSP was observed at pulse intervals of < 100 ms. This facilitation was blocked at hyperpolarised membrane potentials or in the presence of AP5, suggesting activation of NMDA receptors. There was a depression of corticofugal EPSPs by previous dorsal column stimulation. 4. Dorsal column EPSPs were gradually depressed during stimulation with barrages at frequencies of > 10 Hz, while corticofugal EPSPs were facilitated and summated at frequencies > 30 Hz. Hyperpolarisation and application of AP5 prevented the facilitation of corticofugal EPSPs. High frequency stimulation of the corticofugal input elicited a short-lasting AP5-sensitive facilitation of both corticofugal and dorsal column EPSPs. Depolarising current facilitated dorsal column EPSPs but not corticofugal EPSPs. 5. These results indicate that synaptic interactions include different forms of activity-dependent synaptic plasticity, with the participation of NMDA receptors and probably Ca(2+) inflow through voltage-gated channels. These complex synaptic interactions may represent the cellular substrate of the integrative function of the dorsal column nuclei observed in vivo.

Localization of glutamatergic/aspartatergic neurons projecting to the hypothalamic paraventricular nucleus studied by retrograde transport of [3H]D-aspartate autoradiography

Morphological and functional data indicate that glutamatergic innervation of the hypothalamic paraventricular nucleus plays an important role in the control of this prominent cell group. Sources of this neural input are unknown. The present investigations were aimed at studying this question. The retrograde tracer [3H]D-aspartate, which is selectively taken up by the terminals of neurons that use glutamate or aspartate as a neurotransmitter, and is retrogradely transported to their perikarya, was injected into the paraventricular nucleus. The brain was examined for labelled neurons visualized by autoradiography. Labelled neurons were detected in the paraventricular nucleus itself, in several hypothalamic areas including medial and lateral preoptic area, suprachiasmatic nucleus, anterior hypothalamic area, ventromedial nucleus, dorsomedial nucleus, lateral hypothalamic area, posterior part of arcuate nucleus, ventral premammillary nucleus and supramammillary nucleus. Outside the hypothalamus labelled neurons were found in the thalamic paraventricular nucleus and in certain telencephalic regions including lateral septum, bed nucleus of the stria terminalis and amygdala. All of them are known to project to the hypothalamic paraventricular nucleus. We failed to detect labelled neurons in the lower brainstem. From these findings we conclude that firstly, there are glutamatergic/aspartatergic interneurons in the paraventricular nucleus; secondly, all intrahypothalamic and telencephalic, but not lower brainstem afferents to this nucleus contain glutamatergic/aspartatergic fibres; and thirdly, the glutamatergic/aspartatergic innervation of this heterogeneous cell group is extremely complex.

Immunoreactivity for the group III metabotropic glutamate receptor subtype mGluR4a in the superficial laminae of the rat spinal dorsal horn

Studies indicate that metabotropic glutamate receptors (mGluRs) may play a role in spinal sensory transmission. We examined the cellular and subcellular distribution of the mGluR subtype 4a in spinal tissue by means of a specific antiserum and immunocytochemical techniques for light and electron microscopy. A dense plexus of mGluR4a-immunoreactive elements was seen in the dorsal horn, with an apparent accumulation in lamina II. The immunostaining was composed of sparse immunoreactive fibres and punctate elements. No perikaryal staining was seen. Immunostaining for mGluR4a was detected in small to medium-sized cells but not in large cells in dorsal root ganglia. At the electron microscopic level, superficial dorsal horn laminae demonstrated numerous immunoreactive vesicle-containing profiles. Labelling was present in the cytoplasmic matrix, but accretion of immunoreaction product to presynaptic specialisations was commonly observed. Axolemmal labelling was confirmed by using a preembedding immunogold technique, which revealed distinctive deposits of gold immunoparticles along presynaptic thickenings with an average centre-to-centre distance of 41 nm (41.145 +/- 13.59). Immunoreactive terminals often formed synaptic contacts with dendritic profiles immunonegative for mGluR4a. Immunonegative dendritic profiles were observed in apposition to both mGluR4a-immunoreactive and immunonegative terminals. Diffuse immunoperoxidase reaction product was also detected in dendritic profiles, some of which were contacted by mGluR4a-immunoreactive endings, but only occasionally were they observed to accumulate immunoreaction product along the postsynaptic density. Terminals immunoreactive for mGluR4a also formed axosomatic contacts. The present results reveal that mGluR4a subserves a complex spinal circuitry to which the primary afferent system seems to be a major contributor.

The supramammillo-hippocampal and supramammillo-septal glutamatergic/aspartatergic projections in the rat: a combined [3H]D-aspartate autoradiographic and immunohistochemical study

It is well established that the supramammillary nucleus plays a critical role in hippocampal theta rhythm generation/regulation by its direct and indirect (via the septal complex) connections to the hippocampus. Previous morphological and electrophysiological studies indicate that both the supramammillo-hippocampal and supramammillo-septal efferents contain excitatory transmitter. To test the validity of this assumption, transmitter specific retrograde tracer experiments were performed. [3H]D-aspartate was injected into different locations of the hippocampus (granular and supragranular layers of the dentate gyrus and CA2 and CA3a areas of the Ammon's horn) and septal complex (medial septum and the area between the medial and lateral septum) that are known targets of the supramammillary projection. Consecutive vibratome sections prepared from the entire length of the posterior hypothalamus, including the supramammillary area, were immunostained for calretinin, tyrosine hydroxylase, or calbindin, and further processed for autoradiography. Radiolabeled, radiolabeled plus calretinin-containing, and calretinin-immunoreactive neurons were plotted at six different oro-caudal levels of the supramammillary area. The results demonstrated that following both hippocampal and septal injection of the tracer, the majority of the retrogradely radiolabeled (glutamatergic/aspartatergic) cells are immunoreactive for calretinin. However, non-radiolabeled calretinin-containing neurons and radiolabeled calretinin-immunonegative cells were also seen, albeit at a much lower density. These observations clearly indicate the presence of glutamatergic/aspartatergic projections to both the hippocampus and septal complex. It may be assumed that this transmitter could play a role in hippocampal theta rhythm generation/regulation.

Corticofugal action on somatosensory response properties of rat nucleus gracilis cells

Single unit recordings were performed in the nucleus gracilis (Gr) of anesthetized rats to study the influences of the sensorimotor corticofugal projections on sensory responses of those cells. The effects of electrical stimulation of contralateral primary sensory cortex were studied in two conditions: when the receptive fields of the stimulated cortical area and the gracilis cells overlapped (matched) or when they were completely different (unmatched). Cortical stimulation at low intensities (<50 microA) evoked spike firing only in gracilis neurons with matched receptive fields. When the receptive fields were unmatched, the intensity of the stimulation had to be increased above 50 microA to elicit spike firing. To study the corticofugal actions on the responses of Gr neurons, the onset of peripheral stimulation was likened to a single cortical shock in the sensorimotor cortex. When receptive fields matched, cortical stimulation facilitated the cellular responses to the natural sensory stimulation of their RF in most of the Gr neurons (86%). In the unmatched receptive fields, cortical stimulation could either inhibit (66.7%), facilitate (20.8%) or did not modify (12.5%) the sensory response at all. Trains of cortical shocks during sensory stimulation demonstrated that the facilitatory and inhibitory effects on Gr neurons outlasted the period of stimulation by 30-60 s. Results indicate that the sensorimotor cortex exercises a very precise control of sensory transmission throughout the Gr nucleus and suggest that the corticofugal projection may play an important role in the plasticity of the sensorimotor system.

GluR1 and GluR2/3 subunits of the AMPA-type glutamate receptor are associated with particular types of neurone in laminae I-III of the spinal dorsal horn of the rat

GluR1 and GluR2 subunits of the alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptor are expressed at high levels by neurones in laminae I-III of rat spinal dorsal horn, an area which contains numerous, densely packed small neurones. In order to determine whether these subunits are expressed by inhibitory or excitatory neurones, we combined pre-embedding immunocytochemistry with antibodies that recognize either GluR1, or an epitope common to GluR2 and 3, with postembedding detection of gamma-aminobutyric acid (GABA) and glycine. Most (78%) of the neurones with GluR1-immunoreactivity were GABA-immunoreactive, and some of these were also glycine-immunoreactive, whereas nearly all (97%) of the GluR2/3-immunoreactive neurones were not GABA- or glycine-immunoreactive. We carried out double-immunofluorescence and confocal microscopy to provide further information on the neurochemistry of cells that express these subunits. As expected, all neurotensin- and virtually all somatostatin-immunoreactive cells (which are thought to be excitatory interneurones) were GluR2/3- but not GluR1-immunoreactive, whereas parvalbumin-containing cells (most of which are GABAergic) possessed GluR1-, but usually not GluR2/3-immunoreactivity. Neurones that contained nitric oxide synthase (most of which are GABAergic) were more variable, with 57% GluR1-immunoreactive and 41% GluR2/3-immunoreactive. Cholinergic neurones in lamina III (which are also GABAergic) invariably showed each type of GluR-immunoreactivity. These results suggest that neuronal populations in laminae I-III have characteristic patterns of GluR expression: GluR1 is particularly associated with inhibitory neurones, and GluR2 with excitatory neurones. This makes it likely that some of the AMPA receptors present on the inhibitory interneurones lack the GluR2 subunit, and may therefore have significant Ca2+-permeability.

A population of supramammillary area calretinin neurons terminating on medial septal area cholinergic and lateral septal area calbindin-containing cells are aspartate/glutamatergic

The excitatory amino acid, aspartate/glutamate content of septal complex calretinin (CR)-, choline acetyltransferase plus substance P-, and Leu-enkephalin (Leu-enk)-containing extrinsic afferents was examined. Experiments were carried out using the transmitter-specific [3H]-D-aspartate retrograde tracer technique in combination with immunostaining for CR, choline acetyltransferase, and Leu-enk. The extrinsic and intrinsic CR innervation of the same brain areas were elucidated on control rats and on animals in which the septum was surgically separated from its ventral afferents. Correlated light and electron microscopic double-immunostaining experiments were used to determine the synaptic connections between CR axon terminals and lateral septal area calbindin (CB)- and medial septal area choline acetyltransferase-immunoreactive neurons. Furthermore, to determine the synaptic power of supramammilloseptal aspartate/glutamatergic neurons on the septal complex, semiquantitative analyses were performed in the supramammillary area on retrogradely (1) [3H]-D-aspartate-radiolabeled and (2) HRP-labeled material. The results demonstrated that a population of the extrinsic CR axons originating in the supramammillary area are aspartate/glutamatergic. These fibers forming asymmetric synaptic contacts terminate on both CB and cholinergic neurons. Intraseptal CR neurons, which establish symmetric synapses, innervate only lateral septal area neurons, including the CB-containing cells. These observations, together with other published data, raise the possibility of a hippocampus-lateral septal (GABAergic CB-containing neurons)-supramammillary area (aspartate/glutamatergic cells)-medial septal (cholinergic neurons)-hippocampus signal loop, which might be involved in the generation and regulation of hippocampal theta rhythm activity.

AMPA receptor subunits underlying terminals of fine-caliber primary afferent fibers

Postembedding immunogold electron microscopy was used to determine the relation of primary afferent terminals in superficial laminae of the spinal dorsal horn with AMPA receptor subunits. Immunogold particles coding for GluR1 and GluR2/3 were concentrated at synaptic sites, between 30 nm outside and 40 nm inside the postsynaptic membrane. Immunopositive synapses displayed round vesicles and asymmetric specializations, characteristic of terminals releasing excitatory neurotransmitters; symmetric synapses, characteristic of terminals releasing inhibitory amino acids, were immunonegative. In superficial laminae, large terminals of two main types at the center of a synaptic glomerulus originate from primary afferents: C1 terminals are mainly endings of unmyelinated afferent fibers; C2 terminals are mainly endings of thinly myelinated afferent fibers. Terminals of both types were presynaptic to AMPA subunits, but in different proportions: C1 terminals were related more to GluR1 than to GluR2/3, whereas the reverse was true for C2 terminals. These results suggest that functional properties of peripheral afferents to the spinal cord may be specified by the density and combination of receptor subunits in the postsynaptic membrane, and raise the possibility that calcium-permeable AMPA channels may play a special role in the mediation of sensory input by unmyelinated fibers.

Evidence for release of glutamic acid, aspartic acid and substance P but not gamma-aminobutyric acid from primary afferent fibres in rat spinal cord

In vitro superfusion release experiments and autoradiography were carried out on spinal cords of neonatally capsaicin-treated rats. Electrical and chemical stimulations significantly increased the release of aspartate, glutamate and gamma-aminobutyric acid (GABA) from hemisected dorsal horn slices of vehicle-treated animals. In capsaicin-treated rats, the evoked release of aspartate, glutamate and substance P but not GABA, were significantly lower. Capsaicin (1 microM) stimulated the release of aspartate and glutamate, as reported for substance P, in control slices but this effect was not as apparent in tissues from capsaicin-treated rats. Evoked GABA release was not affected in either case. alpha-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), kainate, dizocilpine and GABAB binding sites were highly localised in the substantia gelatinosa. Capsaicin treatment did not affect the affinity of the binding sites in all four cases but significantly reduced the density of kainate, dizocilpine and GABAB binding sites. The data suggest that capsaicin-sensitive primary afferent fibres release aspartate, glutamate and Substance P following high-intensity stimulations and that this release might be modulated by presynaptic glutamate and GABAB receptors present on these terminals.

The types of neuron in spinal dorsal horn which possess neurokinin-1 receptors

In order to provide further information about the types of spinal neuron which possess neurokinin-1 receptors, we have carried out pre-embedding immunocytochemistry on sections of rat lumbar spinal cord with an antiserum raised against a synthetic peptide corresponding to part of the sequence of the receptor, and combined this with post-embedding immunocytochemistry to detect GABA and glycine. Numerous neuronal cell bodies showing neurokinin-1 receptor-immunoreactivity were seen in lamina I, laminae III-VI, the lateral spinal nucleus and the area around the central canal. Most of the cells observed in lamina III were small and had relatively restricted dendritic trees which could often not be followed into lamina II, however some larger cells in laminae III and IV had dendrites which extended through lamina II and into lamina I. Cells of the latter type are likely to represent a major target of substance P released from small-diameter primary afferents in the superficial dorsal horn. The great majority (255 out of 283) of spinal neurons which possessed neurokinin-1 receptor-immunoreactivity, including all of those in lamina I, were not GABA- or glycine-immunoreactive, however a few cells in the deep part of the dorsal horn and the lateral spinal nucleus and several cells near the central canal were GABA-immunoreactive, and some of these were also glycine-immunoreactive. These results suggest that substance P acts through neurokinin-1 receptors mainly on excitatory neurons within the spinal cord.

Some inhibitory neurons in the spinal cord develop c-fos-immunoreactivity after noxious stimulation

In order to determine which types of spinal neuron produce c-fos in response to noxious stimulation, we have combined pre-embedding detection of c-fos-like immunoreactivity with post-embedding immunocytochemistry using antibodies against GABA and glycine, 2 h after subcutaneous injection of formalin into a hindpaw of anaesthetized rats. Throughout the spinal cord, the majority of c-fos-immunoreactive neurons (72-81%) did not possess GABA- or glycine-like immunoreactivity, while the remaining cells contained one or both types of immunoreactivity. In the superficial dorsal horn (laminae I and II) and dorsal white matter, between 14 and 20% of c-fos-immunoreactive neurons were GABA-immunoreactive, and some of these were also glycine-immunoreactive. A single neuron in lamina I in one animal was glycine- but not GABA-immunoreactive. In the remainder of the spinal cord, between 21 and 35% of the c-fos-immunoreactive cells were GABA- or glycine-immunoreactive, and the majority of these neurons contained both types of immunoreactivity. These results suggest that some inhibitory neurons in both the superficial and deep parts of the dorsal horn are activated by noxious stimuli. It is known that some of the cells which produce c-fos in response to noxious stimulation are projection neurons, with axons ascending to the brainstem or thalamus, however, because of the large number of c-fos-immunoreactive cells in the dorsal horn, it is likely that many are interneurons, and some of these are probably excitatory cells which use glutamate as a transmitter. It therefore appears that after noxious stimulation c-fos is produced in several types of spinal neuron, including projection cells and both excitatory and inhibitory interneurons.

Alternatively spliced isoforms of the N-methyl-D-aspartate receptor subunit 1 are differentially distributed within the rat spinal cord

N-Methyl-D-aspartate-activated ionotropic glutamate receptors play a crucial role in synaptic transmission in the spinal cord. Molecular cloning has identified two polymorphic subunits--N-methyl-D-aspartate receptor subunits 1 and 2--the products of alternative splicing (subunit 1a-4b) or of different genes (subunit 2 A-D). While the distribution of N-methyl-D-aspartate receptor subunit 1 splice variants is unknown in the spinal cord, that of subunit 2 appears controversial. We examined, by means of in situ hybridization, the distribution of messenger RNAs encoded by these genes in rat cervical spinal cord. Most neurons throughout all the laminae express predominantly type b variants of subunit 1 (dorsal horn: 3b; ventral horn: 4b) and the 2A subunit, although some neurons in laminae 2 and 9 also express subunit 2B. Our findings demonstrate that subunit 1 splice variants are differentially distributed in the rat cervical cord and, since they fall into two physiologically and pharmacologically distinct groups, may reveal the distribution of antagonist- and agonist-preferring N-methyl-D-aspartate receptor subclasses. They also indicate the co-distribution of receptor subunits 1 and 2, suggesting the existence of heteromeric N-methyl-D-aspartate receptor complexes. Thus, in the spinal cord, different combinations of subunit 1 isoforms as well as subunit 2 may form N-methyl-D-aspartate receptors with different physiological and pharmacological properties. If this structural diversity of presumptive N-methyl-D-aspartate receptors exists in human spinal cord, it might identify potential targets for drug therapy.

Neurotransmitters in subcortical somatosensory pathways

Investigations during recent years indicate that many different neuroactive substances are involved in the transmission and modulation of somesthetic information in the central nervous system. This review surveys recent developments within the field of somatosensory neurotransmission, emphasizing immunocytochemical findings. Increasing evidence indicates a widespread role for glutamate as a fast-acting excitatory neurotransmitter at different levels in somatosensory pathways. Several studies have substantiated a role for glutamate as a neurotransmitter in primary afferent neurons and in corticofugal projections, and also indicate a neurotransmitter role for glutamate in ascending somatosensory pathways. Other substances likely to be involved in somatosensory neurotransmission include the neuropeptides. Many different peptides have been detected in primary afferent neurons with unmyelinated or thinly myelinated axons, and are thus likely to be directly involved in primary afferent neurotransmission. Some neurons giving rise to ascending somatosensory pathways, primarily those with cell bodies in the dorsal horn, are also immunoreactive for peptides. Recent investigations have shown that the expression of neuropeptides, both in primary afferent and ascending tract neurons, may change as a result of various kinds of peripheral manipulation. The occurrence of neurotransmitters in intrinsic neurons and neurons providing modulating inputs to somatosensory relay nuclei (the dorsal horn, the lateral cervical nucleus, the dorsal column nuclei and the ventrobasal thalamus) is also reviewed. Neurotransmitters and modulators in such neurons include acetylcholine, monoamines, GABA, glycine, glutamate, and various neuropeptides.

Compartmentation of glutamate and glutamine in the lateral cervical nucleus: further evidence for glutamate as a spinocervical tract neurotransmitter

Previous observations indicate that spinocervical tract terminals contain relatively high levels of glutamate. To examine whether these high glutamate levels are likely to represent a neurotransmitter pool or an elevated metabolic pool, the distributions of glutamate- and glutamine-like immunoreactivities were examined in adjacent immunogold-labeled sections of the lateral cervical nucleus. Spinocervical tract terminals were identified by anterograde transport of horseradish peroxidase and wheat germ agglutinin-horseradish peroxidase conjugate from the spinal cord. Spinocervical tract terminals were found to contain significantly higher levels of glutamate-like immunoreactivity than other examined tissue compartments (large neuronal cell bodies, terminals with pleomorphic vesicles, astrocytes, and average tissue level). In contrast, the highest levels of glutamine-like immunoreactivity were detected in astrocytes. The different analyzed tissue elements formed three groups with respect to glutamate:glutamine ratios: one high ratio group including spinocervical tract terminals, a second group with intermediate ratios consisting of neuronal cell bodies and terminals containing pleomorphic synaptic vesicles, and a third low ratio group including astrocytes. Our findings indicate the presence of a compartmentation of glutamate and glutamine in the lateral cervical nucleus, similar to that postulated in biochemical studies of the central nervous system. The results also show that spinocervical tract terminals have high glutamate: glutamine ratios, similar to those previously observed in putative glutamatergic terminals in the cerebellar cortex. Thus, spinocervical tract terminals display biochemical characteristics that would be expected of glutamatergic terminals and the present findings therefore provide further evidence for glutamate as a spinocervical tract neurotransmitter.

Evidence for excitatory amino acid neurotransmitters in forward and feedback corticocortical pathways within rat visual cortex

It is a commonly accepted notion that cells which make projections between the multiple cortical areas found in the mammalian visual system are excitatory, but there is little direct evidence that this is the case. Here we demonstrate using retrograde tracing with D-[3H]aspartate that connections in the rat which project from lower to higher visual areas (i.e. forward) and those which project from higher to lower areas (i.e. feedback) may use excitatory amino acid neurotransmitters. Following injection into the primary visual cortex, clusters of retrogradely labelled cells were found in several extrastriate areas within the cytoarchitectonic subdivisions 18a ('areas' LM, AL, PX, FLX, RL, AX) and 18b ('area' MX), and in the retrosplenial cortex. In all of these areas D-[3H]aspartate-labelled cells were surrounded by diffuse label which may represent anterograde labelling of axon terminals. This suggests that both legs of reciprocal intracortical circuits have similar chemospecificity. To directly demonstrate excitatory amino acid localization in forward projections, D-[3H]aspartate was injected into extrastriate area LM. As expected, the results revealed retrogradely labelled neurons within area 17. Outside area 17, LM injections labelled neurons in AL, PX, FLX, RL, AX and MX. Taken in the context of the hierarchy of areas in rat cerebral cortex (Coogan and Burkhalter, J. Neurosci., 13, 3749-3772, 1993), these results show that D-[3H]aspartate labels: (1) forward connections from area 17 to LM, AL, PX, RL, AX and MX, (2) feedback connections from LM, AL, FLX, PX, RL, AX and MX to area 17, (3) feedback connections from AL, PX, RL, AX and MX to LM, and (4) lateral connections between FLX and LM. These findings strongly indicate that both forward and feedback connections as well as lateral connections at several different levels of the cortical hierarchy use excitatory amino acid neurotransmitters.

Glutaminase-like immunoreactivity in rat spinomesencephalic tract cells

Retrograde transport of the fluorescent tracer Fluorogold was used in combination with immunohistochemical staining for the enzyme glutaminase to identify putative glutamatergic neurons belonging to the rat spinomesencephalic tract. Glutaminase-like staining in spinal projection neurons suggests that the relay of nociceptive information from the spinal cord to midbrain may involve the excitatory amino acid glutamate.

Inhibitory and excitatory amino acid neurotransmitter binding sites in cynomolgus monkey (Macaca fascicularis) cervical spinal cord

Autoradiography of inhibitory and excitatory amino acid neurotransmitter binding sites in the cervical spinal cord of M. fascicularis spinal cord revealed inhomogeneous distribution of all binding sites in spinal gray matter. Quisqualate-sensitive [3H]glutamate binding, [3H]MK-801 binding, benzodiazepine binding, kainate binding, and GABAB binding had highest levels in the superficial layers of the dorsal horn (laminae 1 and 2) and substantially lower levels in other laminae. [3H]Strychnine binding was more uniformly distributed throughout all laminae with highest levels in the superficial layers of the dorsal horn. These results are similar to those found in other mammals.

Direct observations of synapses between L-glutamate-immunoreactive boutons and identified spinocervical tract neurones in the spinal cord of the cat

Four spinocervical tract cells in lumbosacral spinal cords of adult cats were physiologically characterized and intracellularly labelled with horseradish peroxidase. The neurones were examined with a light microscope and reconstructed. Selected regions were chosen for ultrastructural analysis. Thin sections were treated to reveal the presence of L-glutamate by using the postembedding immunogold method. Two antisera, which specifically recognise the presence of fixed glutamate in tissue, were used in the study. Somata, proximal, and distal dendrites of all four neurones received synaptic contacts from boutons which displayed an obvious immunogold reaction. These boutons formed between 35% and 48% of all synaptic contacts onto spinocervical tract cells. Glutamate-enriched boutons were associated with gold particle densities which were 2-3 times greater than the average densities associated with the surrounding neuropil. Their profiles had a mean diameter of 1.68 microns, contained round agranular synaptic vesicles, and formed asymmetrical synaptic junctions. However, not all boutons displaying these characteristics were enriched with glutamate. Immunogold studies of alternate thin sections, which were incubated with glutamate or GABA antiserum, demonstrated that synaptic boutons on spinocervical tract cells were either enriched with GABA or with glutamate and formed two separate populations which had distinct morphological characteristics. GABA-containing boutons contained irregularly shaped agranular vesicles and formed symmetrical synaptic junctions, whereas glutamate-enriched boutons corresponded to those described above. A further population of boutons, containing highly flattened vesicles, was not immunoreactive for GABA or glutamate. The evidence supports the idea that much of the excitatory transmission into the SCT is mediated by L-glutamate.

Excitatory amino acids as neurotransmitters of cortical and cerebellar projections to the red nucleus: an immunocytochemical study in the guinea pig

We combined a retrograde labeling technique with peroxidase immunocytochemistry to verify whether cortical and cerebellar neurons projecting to the red nucleus (RN) contain high concentrations of glutamate and aspartate as possible neurotransmitters. Injections of a tracer, colloidal gold-labeled enzymatically inactive horseradish peroxidase conjugated to wheatgerm agglutinin, into the RN of adult guinea pigs produced retrograde labeling of layer V cortical neurons, with a large predominance in the ipsilateral hemisphere. Corticorubral neurons were located in the granular parietal cortex (Gr), agranular frontal cortex (Ag), agranular cingulate cortex (Cg), and retrobulbar cortex (Rb). Large numbers of retrogradely labeled neurons were concentrated in contralateral interpositus and dentate cerebellar nuclei. We found the majority of corticorubral neurons to be immunostained by antibodies raised in rabbits against glutamate or aspartate conjugated to invertebrate hemocyanin by glutaraldehyde, supporting the hypothesis that excitatory amino acids are neurotransmitters of corticorubral projections. With either antiserum, immunostaining was found in 58-72% of corticorubral neurons in Ag and Gr; higher percentages were observed in Rb (80-85%) and Cg (up to 96%). Cross-sectional area measurements indicated that the perikarya of corticorubral neurons were larger in Ag and Gr than in Rb and Cg; in each area, soma size values of immunopositive corticorubral neurons tended to be larger than those of immunonegative ones. In the cerebellar nuclei, virtually all retrogradely labeled neurons were immunostained by glutamate and aspartate antisera, suggesting that excitatory amino acids might also be considered as possible neurotransmitters for cerebellorubral projections.

Post cardiac arrest hyperoxic resuscitation enhances neuronal vulnerability of the respiratory rhythm generator and some brainstem and spinal cord neuronal pools in the dog

Selective neuronal vulnerability of the motor cortex, basal ganglia, brainstem, medulla, cerebellum, C6 and L6 segments of the spinal cord were studied after 15 min of cardiac arrest followed by 1 h of normoxic or hyperoxic resuscitation using the suppressive Nauta method in dogs. Hyperoxic resuscitation causes characteristic somatodendritic argyrophilia of the interneuronal pool in the spinal cord and lower medulla. Cuneate, lateral reticular, supraspinal, and caudal trigeminal nuclei as well as the dorsal and ventral respiratory neuronal groups were heavily involved. Similarly, the Purkinje cells, neurons in the middle and deep portions of the mesencephalic tectum, perirubral, pretectal, posterior commissure, middle-sized striatal and giant pyramidal (Betz's) neurons in the motor cortex became argyrophilic. Hyperoxic resuscitation versus normoxic resuscitation causes statistically significant somatodendritic argyrophilia of the dorsal respiratory group, cuneate, dorsal lateral geniculate and thalamic reticular nuclei.

Organization of the serotonergic innervation of spinal neurons in rats--I. Neuropeptide coexistence in varicosities innervating some spinothalamic tract neurons but not in those innervating postsynaptic dorsal column neurons

Previous studies have suggested that peptides such as substance P and thyrotropin-releasing hormone coexist with serotonin in the same varicosities in the ventral horn and intermediate gray of the spinal cord in rat. However, coexistence of these peptides with serotonin is rare in fibers in the superficial dorsal horn. Since it has been proposed that serotonergic fibers in the superficial dorsal horn act to modulate nociception, it was hypothesized that the serotonergic neurons that contain neither substance P nor thyrotropin-releasing hormone might constitute a specifically antinociceptive subset of serotonergic neurons. This being the case, it would be expected that different types of serotonergic neurons innervate nociceptive and non-nociceptive spinal neurons. In order to test this hypothesis, a group of cells that include nociceptive neurons (spinothalamic tract neurons) and a group of predominantly non-nociceptive neurons (postsynaptic dorsal column neurons) in the spinal cord of rat were retrogradely labeled. Sections of the spinal cord containing retrogradely labeled spinothalamic tract or postsynaptic dorsal column neurons were stained for serotonin and either substance P or thyrotropin-releasing hormone using two-color immunohistochemistry. A retrogradely labeled cell was classified as "apposed" if there was no discernible distance between an immunohistochemically labeled varicosity and the cell. Eighty per cent of spinothalamic tract and 83% of postsynaptic dorsal column profiles were apposed by serotonin-immunoreactive varicosities in the spinal cord. Thirty-one per cent of the spinothalamic tract profiles that were apposed by serotonergic varicosities were apposed by serotonergic varicosities that were also stained for thyrotropin-releasing hormone. The distribution of the latter spinothalamic neurons was similar to that reported for spinothalamic tract neurons responsive to joint movement. In addition, at least 63% of the spinothalamic tract profiles which were apposed by serotonergic varicosities were apposed by "serotonin-only" varicosities, including most spinothalamic tract neurons in the marginal zone, suggesting that at least some "serotonin-only" neurons are antinociceptive. However, contrary to the hypothesis, at least 94% of the postsynaptic dorsal column profiles apposed by serotonergic varicosities were apposed by "serotonin-only" varicosities. These findings suggest that there may be a relationship between the sensory modality to which a spinal neuron responds and the type of serotonergic innervation it receives. However, it appears that "serotonin-only" neurons may not constitute a specifically antinociceptive category of serotonergic neurons.

Ultrastructural and immunocytochemical study of lamina II islet cells in rat spinal dorsal horn

In order to compare the ultrastructure of GABA-immunoreactive and nonimmunoreactive islet cells in lamina II of the rat dorsal horn, a combined ultrastructural and immunocytochemical study of nine Golgi-stained neurones was performed. Cell bodies of these neurones were tested with antiserum to GABA, and in most cases with antiserum to glycine, while parts of the cell body and dendritic tree were examined with the electron microscope. Four of the neurones had cell bodies that were immunoreactive with GABA antiserum, and 2 of these were also glycine-immunoreactive, while 2 were not. Cell bodies of the remaining five neurones were not immunoreactive with GABA antiserum, nor, in the 3 cases tested, with glycine antiserum. Three of the GABA-immunoreactive cells possessed vesicle-containing dendrites and were presynaptic at dendrodendritic synapses, whereas no vesicles were observed in the dendrites of any of the neurones that were not GABA-immunoreactive. The axon of one of the nonimmunoreactive cells was found with the electron microscope. It gave rise to boutons that contained round agranular vesicles and a few dense-cored vesicles. Three synapses formed by this axon were identified and all were asymmetric. No obvious differences were detected in the types of profile that were presynaptic to GABA-immunoreactive and nonimmunoreactive cells. These results suggest that GABAergic islet cells are a source of presynaptic dendrites in lamina II of the rat and that some presynaptic dendrites contain GABA and glycine, while others contain GABA without glycine. The nonimmunoreactive islet cells presumably represent a distinct functional class of neurones and some of these may release an excitatory amino acid transmitter, possibly in addition to one or more neuropeptides.

Neuroanatomical evidence that vagal afferent nerves do not possess a high affinity uptake system for glutamate

The ability of vagal and glossopharyngeal afferent neurons to retrogradely transport 3H-D-aspartate from the nucleus tractus solitarius to the nodose and petrosal ganglia was examined. Injections of 3H-D-aspartate centered in the medial NTS at the rostral-caudal level of the area postrema failed to consistently label cells in the nodose and petrosal ganglia. In 5 of the 10 rats studied no retrogradely labeled neurons were observed in these ganglia ipsilateral to the injection site, while in the other 5 rats a small number of cells (less than 3%) were labeled. Injections of 3H-D-aspartate into the NTS consistently produced retrograde labeling of neurons in the ipsilateral paratrigeminal area. In addition, many heavily labeled neurons were observed in the injected as well as the contralateral NTS. Injections of 3H-D-asparate into the spinal trigeminal nucleus consistently labeled neurons in the trigeminal ganglion. Since the uptake and retrograde transport of 3H-D-aspartate appears to be characteristic of neurons that use glutamate or aspartate as a neurotransmitter, these results suggest that vagal and glossopharyngeal afferents are not glutamatergic or aspartatergic.

Acoustic chiasm V: inhibition and excitation in the ipsilateral and contralateral projections of LSO

When this series of experiments was begun in 1984, the activity of each lateral superior olive (LSO) in the mammalian hindbrain was known to encode the hemifield of acoustic space containing a sound source. However, the almost random bilaterality of its ascending projections seemed to jumble that identification before reaching the midbrain. At the same time, electrophysiological studies of LSO and its efferent target in the inferior colliculus, along with the strictly contralateral deficits in sound localization resulting from unilateral lesions above the level of the superior olives, indicated that hemifield allegiance was largely maintained (though reversed) at the midbrain. Here we present seven lines of biochemical evidence, some combined with prior ablations, supporting the notion that the anatomical segregation of the ipsilateral and contralateral fibers ascending from the LSO is accompanied by a corresponding segregation of their neurotransmitters: most of the ascending ipsilateral projection is probably glycinergic and, hence, inhibitory in effect, while most of the contralateral projection is probably glutamatergic/aspartergic and, hence, excitatory in effect. Taken together, the inhibitory ipsilateral projections and the excitatory contralateral projections serve to amplify functional contralaterality at the higher levels of the auditory system.

Different populations of parvalbumin- and calbindin-D28k-immunoreactive neurons contain GABA and accumulate 3H-D-aspartate in the dorsal horn of the rat spinal cord

The colocalization of parvalbumin (PV), calbindin-D28k (CaBP), GABA immunoreactivities, and the ability to accumulate 3H-D-aspartate selectively were investigated in neurons of laminae I-IV of the dorsal horn of the rat spinal cord. Following injection of 3H-D-aspartate into the basal dorsal horn (laminae IV-VI), perikarya selectively accumulating 3H-D-aspartate were detected in araldite embedded semithin sections by autoradiography, and consecutive semithin sections were treated to reveal PV, CaBP and GABA by postembedding immunocytochemistry. Perikarya accumulating 3H-D-aspartate were found exclusively in laminae I-III, and no labelled somata were found in deeper layers or in the intermediolateral column although the labelled amino acid clearly spread to these regions. More than half of the labelled cells were localized in lamina II. In this layer, 16.4% of 3H-D-aspartate-labelled perikarya were also stained for CaBP. In contrast to CaBP, PV or GABA was never detected in neurons accumulating 3H-D-aspartate. A high proportion of PV-immunoreactive perikarya were also stained for GABA in laminae II and III (70.0% and 61.2% respectively). However, the majority of CaBP-immunoreactive perikarya were GABA-negative. GABA-immunoreactivity was found in less than 2% of the total population of cells stained for CaBP in laminae I-IV. A significant proportion of the GABA-negative but PV-immunoreactive neurons also showed CaBP-immunoreactivity in laminae II and IV. These results show that out of the two calcium-binding proteins, CaBP is a characteristic protein of a small subpopulation of neurons using excitatory amino acids and PV is a characteristic protein of a subpopulation of neurons utilizing GABA as a transmitter. However, both proteins are present in additional subgroups of neurons, and neuronal populations using inhibitory or excitatory amino acid transmitters are heterogeneous with regard to their content of calcium-binding proteins in the dorsal horn of the rat spinal cord.

Sources of presumptive glutamatergic/aspartatergic afferents to the magnocellular basal forebrain in the rat

The distribution of presumptive glutamatergic and/or aspartatergic neurons retrogradely labeled following injections of [3H]-D-aspartate into the magnocellular basal forebrain of the rat was compared with the distribution of neurons labeled by comparable injections of the nonspecific retrograde axonal tracer wheat germ agglutinin conjugated to horseradish peroxidase. Cells retrogradely labeled by wheat germ agglutinin-horseradish peroxidase were found in a wide range of limbic and limbic-related structures in the forebrain and brainstem. In the telencephalon, labeled neurons were seen in the orbital, medial prefrontal, and agranular insular cortical areas, the amygdaloid complex, and the hippocampal formation. Labeled cells were also seen in the olfactory cortex, the lateral septum, the ventral striatopallidal region, and the magnocellular basal forebrain itself. In the diencephalon, neurons were labeled in the midline nuclear complex of the thalamus, the lateral habenular nucleus, and the hypothalamus. In the brainstem, labeled cells were found bilaterally in the ventral midbrain, the central gray, the reticular formation, the parabrachial nuclei, the raphe nuclei, the laterodorsal tegmental nucleus, and the locus coeruleus. A significant fraction of the afferents to the magnocellular basal forebrain appear to be glutamatergic and/or aspartatergic. Only a few of the regions labeled with wheat germ agglutinin-horseradish peroxidase were not also labeled with [3H]-D-aspartate in the comparable experiments. Most prominent among the non-glutamatergic/aspartatergic projections were those from fields CA1 and CA3 of the hippocampus, the hilus of the dentate gyrus, the dorsal subiculum, the tuberomammillary nucleus, and the ventral pallidum. In addition, most of the lateral hypothalamic and brainstem projections to the magnocellular basal forebrain were not significantly labeled with [3H]-D-aspartate. In addition to these inputs, a commissural projection from the region of the contralateral nucleus of the horizontal limb of the diagonal band was confirmed with both wheat germ agglutinin-horseradish peroxidase and the anterograde axonal tracer Phaseolus vulgaris leucoagglutinin. This projection did not label with [3H]-D-aspartate or [3H]-GABA, suggesting that it is not glutamatergic/aspartatergic or GABAergic. Furthermore, double labeling experiments with the fluorescent retrograde tracer True Blue and antibodies against choline acetyltransferase indicate that the projection is not cholinergic.

Excitatory amino acid transmitters and their receptors in neural circuits of the cerebral neocortex

In 1954, L-glutamate (Glu) and L-aspartate (Asp) were first suggested as being excitatory synaptic transmitters in the cerebral cortex. Since then, evidence has mounted steadily in favor of the view that Glu and Asp are major excitatory transmitters in the neocortex. Many of the experimental studies which reported how Glu/Asp came to satisfy the criteria for transmitters in the neocortex are reviewed here, according to the methods employed. Since the question of which particular synaptic sites in cortical neural circuits Glu/Asp operate as excitatory transmitters has not previously been reviewed, particular attention is given to efferent, afferent and intrinsic neural circuits of the visual and somatosensory cortices, where circuitry is relatively clearly delineated. Recent studies using chemical assays of released amino acids, high-affinity uptake mechanisms of Glu/Asp from nerve terminals, the direct micro-iontophoretic administration of Glu/Asp antagonists, and immunocytochemical techniques have demonstrated that almost all corticofugal efferent projections employ Glu/Asp as excitatory synaptic transmitters. Evidence indicating that thalamocortical afferent projections, including geniculocortical projections and some intrinsic connections are glutamatergic, is also reviewed. Thus, the results highlighted here indicate that the main framework of neocortical circuitry is operated by Glu/Asp. Pharmacological studies indicate that synaptic receptors for Glu/Asp can be classified into a few subtypes, including N-methyl-D-aspartate (NMDA) and quisqualate/kainate (non-NMDA) types. Some evidence indicating the sites of operation of NMDA and non-NMDA receptors in neocortical circuitry is reviewed, and the distinct, functional significance of these two types of Glu/Asp receptors in information processing in the neocortex is proposed.

Distribution of phosphate-activated glutaminase in the human cerebral cortex

Phosphate-activated glutaminase (PAG), which catalyses conversion of glutamine to glutamate, is a potential marker for glutamatergic, and possibly GABA, neurons in the central nervous system. A polyclonal antibody, raised in rabbits against rat brain PAG, was applied to postmortem human brain tissue to reveal the distribution of PAG in the cerebral cortex. PAG immunoreactivity was observed in pyramidal and non-pyramidal neurons but not in glial cells. In the neocortex, large to medium-sized pyramidal neurons in layers III and V were stained most intensely, while the majority of smaller pyramidal cells were labeled either lightly or moderately. Such modified pyramids as the giant Betz cells, the large pyramidal cells of Meynert, and the solitary cells of Ramón y Cajal were also stained intensely. Fusiform cells in layer VI showed moderate to intense labeling. A number of cortical non-pyramidal neurons of various sizes stained moderately to intensely. These included large basket cells which were identified by their characteristic morphology and size in primary cortical areas. Pyramidal cells in the hippocampal formation as well as basket cells of the stratum oriens stained moderately to intensely. Since pyramidal cells are believed to be glutamatergic and large basket cells GABAergic, these results suggest that PAG plays a role in generating not only transmitter glutamate, but also GABA precursor glutamate.

Ascending afferents to the lateral cervical nucleus are enriched in glutamate-like immunoreactivity: a combined anterograde transport-immunogold study in the cat

To investigate whether glutamate (Glu) may be a transmitter in terminals of ascending afferents to the lateral cervical nucleus (LCN), these terminals were identified by anterograde transport of wheatgerm agglutinin-horseradish peroxidase from the spinal cord, and their content of Glu-like immunoreactivity (Glu-LI) was assessed at the ultrastructural level by the immunogold technique. The gold particle density over the peroxidase-positive terminals of the spinocervical tract (SCT) was significantly higher (by a factor of 2.44) than over a reference terminal population containing flattened or pleomorphic vesicles. Further, LCN neurons were densely labeled by the Glu antiserum, although the gold particle density over neuronal cell bodies was not as high as in the SCT terminals. Previous investigations have shown enrichment of Glu-LI in putative glutamatergic terminals in other parts of the CNS. Hence, the present observations indicate that Glu may be a transmitter in the synapses between SCT terminals and LCN neurons. The cell body labeling in the LCN is more difficult to interpret because of possible interference of metabolic pools of glutamate.

Light microscope study of the coexistence of GABA-like and glycine-like immunoreactivities in the spinal cord of the rat

The distributions of GABA-like and glycine-like immunoreactivities in the rat spinal cord were compared by using postembedding immunohistochemistry on semithin sections. In laminae I, II, and III, the proportions of GABA immunoreactive cells were 28%, 31%, and 46%, respectively, whereas for glycine immunoreactive cells the proportions were 9%, 14%, and 30%. Nearly all of the glycine immunoreactive cells in this area were also immunoreactive with the anti-GABA antiserum. In lamina II, some Golgi-stained islet cells were glycine immunoreactive, whereas others were not. Immunoreactive cell bodies were also present in the remainder of the grey matter. Some of these reacted with anti-GABA or antiglycine antiserum; others showed immunoreactivity with both antisera. Immunoreactive axons were found in the ventral and lateral funiculi of the white matter. Many large axons reacted with antiglycine antiserum, whereas GABA-immunoreactive axons were mostly of small diameter. Some large and small axons showed both types of immunoreactivity. These results suggest that the inhibitory neurotransmitters GABA and glycine coexist within cell bodies and axons in the rat spinal cord.

Excitatory amino acids and synaptic transmission: the evidence for a physiological function

For 30 years physiological techniques have been used to investigate excitatory amino acids as neurotransmitters. In the last ten years progress on the definition of receptor subtypes and the availability of more selective agonists and antagonists has fuelled physiological, neurochemical and histochemical approaches to elucidating the involvement of excitatory amino acids at synaptic sites throughout the vertebrate CNS. Here Max Headley and Sten Grillner assess the advances made in defining the roles of excitatory amino acids as functional transmitters, taking examples mainly from studies on the spinal cord, and comment on the limitations of the types of approach that are used in such studies.

Afferents to the median raphe nucleus of the rat: retrograde cholera toxin and wheat germ conjugated horseradish peroxidase tracing, and selective D-[3H]aspartate labelling of possible excitatory amino acid inputs

Afferents to the median-paramedian raphe nuclear complex, which contains the B8 serotonergic cell group, were investigated in the rat with neuroanatomical and transmitter-selective retrograde labelling techniques. Injection of sensitive retrograde tracers, cholera toxin genoid or wheat germ agglutinin conjugated horseradish peroxidase into the median raphe resulted in labelling of neurons in a large number of brain regions. Projections from 26 of these regions are supported by available orthograde tracing data; the cingulate cortex, bed nucleus of stria terminalis, medial septum and diagonal band of Broca, ventral pallidum, medial and lateral preoptic areas, lateral hypothalamus, dorsomedial nucleus of hypothalamus, lateral habenula, interpeduncular nucleus, substantia nigra, central (periaqueductal) gray, and laterodorsal tegmental nucleus seem to represent major sources of afferents to the median-paramedian raphe complex. Retrogradely labelled cells were also observed in a number of regions for which anterograde tracing data are not available, including the perifornical hypothalamic nucleus, ventral premammillary nucleus, supramammillary and submammillothalamic nuclei and the B9 area. Possible excitatory amino acid afferents were identified with retrograde D-[3H]aspartate labelling. Microinjection of D-[3H]aspartate at a low concentration, 10(-4) M in 50 nl, resulted in retrograde labelling of a limited number of median raphe afferents. The most prominent labelling was observed in the lateral habenula and the interpeduncular nucleus, but retrogradely labelled cells were also noted in the medial and lateral preoptic areas, lateral and dorsal hypothalamus, ventral tegmental area, laterodorsal tegmental nucleus, medial parabrachial nucleus, and the pontine tegmentum. After injections of 10(-3) M D-[3H]aspartate selective labelling also appeared in more distant afferent regions, including cells in cingulate cortex, and in some regions located at shorter distances, such as the supramammillary nucleus. Injections of D-[3H]aspartate at high concentration, 10(-2) M, resulted in the appearance of weakly to moderately labelled cells in most afferent areas which were devoid of labelled cells after injections of lower concentrations, suggesting that this labelling may be non-specific. It was concluded that the median-paramedian raphe receives afferents from a large number of forebrain and hypothalamic regions, while relatively few brain stem regions project to this nuclear complex. The selectivity of retrograde labelling with D-[3H]aspartate was found to be concentration dependent, and it is suggested that the connections showing high affinity for D-[3H]aspartate may use excitatory amino acids as transmitters. Excitatory amino acid inputs from lateral habenula and interpeduncular nucleus may play predominant roles in the control of ascending serotonergic and non-serotonergic projections originating in the median and paramedian raphe nuclei.

Excitatory amino acid projections to the periaqueductal gray in the rat: a retrograde transport study utilizing D[3H]aspartate and [3H]GABA

The afferents to the periaqueductal gray utilizing excitatory amino acid transmitters have been described in rat brain by autoradiography following microinfusion and retrograde transport of D[3H]aspartate. Parallel experiments employing injections of [3H]GABA established that the retrograde labelling found with D[3H]aspartate was transmitter-selective. Following infusion of D[3H]aspartate, perikaryal labelling was found in nine subcortical areas, particularly infralimbic and cingulate cortices, with a predominance of ipsilateral labelled perikarya. Heaviest cortical labelling was localized in perirhinal cortex, in an extensive band of cells adjoining the rhinal sulcus. The hypothalamus contained the heaviest perikaryal labelling within brain: D[3H]aspartate labelled cells in 11 hypothalamic and mammillary nuclei. Intense bilateral labelling was obtained in ventromedial hypothalamus, although the number of perikarya was lower contralaterally. D[3H]Aspartate also produced heavy ipsilateral labelling of perikarya in posterior hypothalamus. Labelling patterns in cortex and hypothalamus were precise and topographic, and [3H]GABA never labelled cells in these regions. Other telencephalic and diencephalic areas containing prominent, retrogradely labelled cells were the lateral septum, amygdala, zona incerta and lateral habenula. The relative density of labelled cells in mesencephalic areas was much lower than that found in cortex and hypothalamus, although D[3H]aspartate labelled a moderate number of perikarya in the inferior colliculus and cuneiform nucleus. A smaller number of heavily labelled cells was found in the parabrachial nuclei, Kolliker-Fuse nucleus and laterodorsal tegmental nucleus. Only occasional labelled perikarya were observed in the myencephalon. Low densities of labelled cells were found after the injection of [3H]GABA into the periaqueductal gray, and the only regions in which a small number of perikarya were labelled by both [3H]GABA and D[3H]aspartate were the dorsal raphe and parabrachial nuclei. Overall, the retrograde transport of D[3H]aspartate revealed a complex topographic and convergent network of afferent pathways to the periaqueductal gray likely to utilize an excitatory amino acid transmitter. Our findings confirm the selectivity of this neurochemical mapping technique and provide evidence that hypothalamic, habenular, subthalamic and cuneiform afferents to the periaqueductal gray utilize an acidic amino acid as their transmitter. They also confirm that corticofugal afferents to periaqueductal gray utilize an excitatory amino acid.

Excitatory transmitter amino acid-containing neurons in the rat visual cortex: a light and electron microscopic immunocytochemical study

The distribution and morphology of neurons labelled with antisera to glutamate or aspartate were examined, at the light and electron microscope levels, in the rat visual cortex. Using widely accepted light microscopic features as well as well-established nuclear, cytoplasmic, and synaptic criteria, we noted that glutamate-immunoreactive neurons were pyramidal cells distributed in layers II-VI, with an increased concentration in layers II and III. Aspartate immunoreactivity was localized chiefly to pyramidal neurons in layers II-VI. However, approximately 10% of immunolabeled cells were nonpyramidal neurons scattered throughout the cortex. Cell-body measurements revealed that, for both groups of neurons, layer V contained the largest labelled neurons, whereas layers IV and VI contained the smallest. Furthermore, in every layer, aspartate-stained neurons were larger than glutamate-positive cells. Finally, glutamate- and aspartate-labelled axon terminals formed asymmetrical synapses, which are presumably excitatory in nature, primarily with dendritic spines. These findings, together with recent detailed studies of the projections of glutamate- and aspartate-labelled cortical neurons, may provide essential background information for studies aimed to elucidate the function(s) of excitatory amino acids in the cortex and their role in pathological conditions.

Autoradiographic mapping of a selective cyclic adenosine monophosphate phosphodiesterase in rat brain with the antidepressant [3H]rolipram

Rolipram is a clinically effective antidepressant with selective cAMP phosphodiesterase (PDE) inhibiting properties. (+/-)-[3H]Rolipram binds with high affinity (Kd = 2.52 +/- 0.47 nM) to sections of rat brain (Hill number = 0.90 +/- 0.05). Binding is stereospecific. Association of (+/-) [3H]rolipram to sections is rapid (47% of specific binding in the first minute, kobs = 0.52 min-1). Dissociation of (+/-)-[3H]rolipram exhibits non first order kinetics (3 component model; t1/2 = 2.5 min, 50 min and 6 h, respectively). A number of PDE inhibitors reduce (+/-)-[3H]rolipram binding to the level of nonspecific binding ((-)-rolipram, IC50 = 0.9 nM; (+/-)-rolipram, IC50 = 1.5 nM; Ro 20-1724, IC50 = 11 nM; ICI 63.197, IC50 = 35 nM; medazepam, IC50 = 240 nM; diazepam, IC50 = 1200 nM; IBMX, IC50 = 3800 nM). In vitro autoradiography reveals high binding site densities in the cerebellum, olfactory bulb, lateral septal nucleus, frontal cortex, subiculum and CA1 of hippocampus. Most of the labeled structures are part of the limbic system. In vivo autoradiography of (+/-)-[3H]rolipram binding shows much more nonspecific binding than in vitro, nevertheless the distribution pattern of (+/-)-[3H]rolipram binding sites is similar. A comparison of the distribution pattern of (+/-)-[3H]rolipram binding sites with that of an antidepressant (monoamine oxidase inhibitor, monoamine uptake inhibitor) reveals no overlap. Limited, though significant correlations exist with the distribution of beta 1-adrenergic, adenosine1 and glutamate/quisqualate receptors as well as protein kinase C, but not with beta 2-adrenergic receptors and forskolin binding sites.

Glutamate-positive neurons and axon terminals in cat sensory cortex: a correlative light and electron microscopic study

Immunocytochemical methods were used to perform a correlative light and electron microscopic study of neurons and axon terminals immunoreactive to the antiglutamate (Glu) serum of Hepler et al. ('88) in the visual and somatic sensory areas of cats. At the light microscopic level, numerous Glu-positive neurons were found in all layers except layer I of both cortical areas. On the basis of the dendritic staining of Glu-positive cells, two major morphological categories were found: pyramidal cells, which were the most frequent type of immunostained neuron, and multipolar neurons, which were more numerous in layer IV of area 17 than in any other layer. A large number of Glu-positive neurons, however, did not display dendritic labelling and were considered unidentified neurons. Counts of labelled neurons were performed in the striate cortex; approximately 40% were Glu-positive. Numerous lightly stained punctate structures were observed in all cortical layers: the majority of these Glu-positive puncta were in the neuropil. After resectioning the plastic sections for electron microscopy it was observed that: 1) the majority of neurons unidentifiable at light microscopic level were indeed pyramidal neurons except in layer IV of area 17, where many stained cells were probably spiny stellate neurons. Some Glu-positive neurons, however, exhibited clear ultrastructural features of nonspiny nonpyramidal cells; 2) all synaptic contacts made by Glu-positive axon terminals were of the asymmetric type, but not all asymmetric synaptic contacts were labelled. The vast majority of postsynaptic targets of Glu-positive axons were unlabelled dendritic spines and shafts. The present results provide further evidence that Glu (or a closely related compound) is probably the neurotransmitter of numerous excitatory neurons in the neocortex.

Cortical glutaminase, beta-glucuronidase and glucose utilization in Alzheimer's disease

Large pyramidal neurons of rat and human neocortex stain immunohistochemically for phosphate-activated glutaminase (PAG). In a limited number of postmortem brains, we find large reductions in cortical PAG activity in Alzheimer's disease (AD). This finding is consistent with histological evidence that pyramidal neurons are affected in AD. The reductions are greater than those found in the same samples in choline acetyltransferase (ChAT) but the possible deleterious effects of coma and similar premortem factors on human PAG activity have yet to be assessed. The activity of beta-glucuronidase, a lysosomal enzyme which occurs in reactive astrocytes, is elevated in the same samples. Positron emission tomography (PET) studies, using 18F-fluorodeoxyglucose (FDG), have demonstrated significant deficiencies in glucose metabolism in the cortex in AD, with the parietal, temporal and some frontal areas being particularly affected. We found in serial scans of 13 AD cases, including one relatively young (44-46 year old) familial case, an exacerbation of the defect over time in most cases. We have found a negative correlation between the regional metabolic rates for glucose (LCMR(s] measured premortem and the beta-glucuronidase activities measured postmortem on a few AD cases that have come to autopsy. The correlations between LCMR(s) and PAG and ChAT activities tend to be positive. The results are consistent with previous suggestions that decreased LCMR(s) in AD reflect local neuronal loss and gliosis.

Glutamate and aspartate immunoreactivity in corticospinal neurons of rats

A combination of retrograde tracers and immunostaining was employed to test whether corticospinal tract (CST) neurons in rats may use amino acid excitatory neurotransmitters. CST neurons were identified following injections of either Diamidino Yellow (DY) or colloidal gold-labeled enzymatically inactive horseradish peroxidase conjugated to wheat germ agglutinin (WGAapoHRP-Au) in the spinal cord. As retrograde tracers, the two substances seemed to be equally effective, but WGAapoHRP-Au was better suited than DY as a tracer to use in combination with immunocytochemistry. Sections through the primary sensorimotor cortex, which contained the bulk of identified CST neurons, and the secondary somatosensory cortex were processed with antisera raised in rabbits against glutamate (Glu) or aspartate (Asp) conjugated by glutaraldehyde to hemocyanin. In rats with DY injections, about 60-75% of the CST neurons were Glu-immunopositive, with higher ratios in SI and MI than in SII. Similar results were obtained in all areas examined from the rats with injections of WGAapoHRP-Au. Only sections from rats with injections of WGAapoHRP-Au were processed for Asp immunostaining. In this material, between 65 and 75% of the CST neurons were Asp-immunopositive, with a slightly higher ratio in SI and MI than in SII. The possibility that these results might reflect limited penetration of the antiserum and/or staining of the same population of CST neurons by either antiserum was addressed in sections processed with both the Glu and Asp antisera. In sections incubated in a mixture of the two antisera, the percentage of immunostained CST neurons was higher, about 90%, than in sections processed for only one of the two antisera. Furthermore, in rats in which Glu and Asp antibodies were visualized by two distinguishable immunostainings, four populations of CST neurons were identifiable: 1) neurons only immunopositive for Glu, 2) neurons only immunopositive for Asp, 3) neurons likely to be stained by both, and 4) neurons immunonegative for both antisera. Twenty-five to 30% of CST neurons were positive for only one antiserum, and about 50% were positive for both. No preferential distribution was evident for any one of these populations of neurons. However, perikaryal cross-sectional areas were larger for the double-stained than for the single-stained CST neurons. Glutamergic and aspartergic transmission in CST neurons has been proposed in several publications in which methods other than immunocytochemistry were employed.(ABSTRACT TRUNCATED AT 400 WORDS)