Most somatostatin-immunoreactive neurons in the rat fascia dentata do not contain the calcium-binding protein parvalbumin

Gingko biloba extract (EGb 761) attenuates ischemic brain injury-induced reduction in Ca(2+) sensor protein hippocalcin

Gingko biloba extract 761 (EGb 761) protects neuronal cells from ischemic brain injury via a number of neuroprotective mechanisms. Hippocalcin is a calcium sensor protein that regulates intracellular calcium concentrations and apoptotic cell death. We investigated whether EGb 761 regulates hippocalcin expression in cerebral ischemia. Male Sprague-Dawley rats were treated with vehicle or EGb 761 (100 mg/kg) prior to middle cerebral artery occlusion (MCAO), and cerebral cortex tissues were collected 24 h after MCAO. A proteomic approach demonstrated reduction in hippocalcin expression in vehicle-treated animals during MCAO, whereas EGb 761 treatment prevented injury-induced decreases in hippocalcin expression. RT-PCR and Western blot analyses indicated that EGb 761 attenuates injury-induced decrease in hippocalcin. These results suggest that the maintenance of hippocalcin during cerebral ischemia contributes to the neuroprotective role of EGb 761.

Altered intrinsic properties of neuronal subtypes in malformed epileptogenic cortex

Neuronal intrinsic properties control action potential firing rates and serve to define particular neuronal subtypes. Changes in intrinsic properties have previously been shown to contribute to hyperexcitability in a number of epilepsy models. Here we examined whether a developmental insult producing the cortical malformation of microgyria altered the identity or firing properties of layer V pyramidal neurons and two interneuron subtypes. Trains of action potentials were elicited with a series of current injection steps during whole cell patch clamp recordings. Cells in malformed cortex identified as having an apical dendrite had firing patterns similar to control pyramidal neurons. The duration of the second action potential in the train was increased in paramicrogyral (PMG) pyramidal cells, suggesting that these cells may be in an immature state, as was previously found for layer II/III pyramidal neurons. Based on stereotypical firing patterns and other intrinsic properties, fast-spiking (FS) and low threshold-spiking (LTS) interneuron subpopulations were clearly identified in both control and malformed cortex. Most intrinsic properties measured in malformed cortex were unchanged, suggesting that subtype identity is maintained. However, LTS interneurons in lesioned cortex had increased maximum firing frequency, decreased initial afterhyperpolarization duration, and increased total adaptation ratio compared to control LTS cells. FS interneurons demonstrated decreased maximum firing frequencies in malformed cortex compared to control FS cells. These changes may increase the efficacy of LTS while decreasing the effectiveness of FS interneurons. These data indicate that differential alterations of individual neuronal subpopulations may endow them with specific characteristics that promote epileptogenesis.

Parvalbumin-, calbindin-, and calretinin-immunoreactive hippocampal interneuron density in autism

BACKGROUND

There has been a long-standing interest in the possible role of the hippocampus in autism and both postmortem brain and neuroimaging studies have documented varying abnormalities in this limbic system structure.

AIMS

This study investigates the density of subsets of hippocampal interneurons, immunostained with the calcium binding proteins, calbindin (CB), calretinin (CR) and parvalbumin (PV) to determine whether specific subpopulations of interneurons are impacted in autism.

MATERIALS AND METHODS

Unbiased stereological techniques were used to quantify the neuronal density of these immunoreactive subpopulations of gamma-aminobutyric acid-ergic (GABAergic) interneurons analyzed in the CA and subicular fields in postmortem brain material obtained from five autistic and five age-, gender- and postmortem interval-matched control cases.

RESULTS

Results indicate a selective increase in the density of CB-immunoreactive interneurons in the dentate gyrus, an increase in CR-immunoreactive interneurons in area CA1, and an increase in PV-immunoreactive interneurons in areas CA1 and CA3 in the hippocampus of individuals with autism when compared with controls.

DISCUSSION/CONCLUSIONS

Although our sample size is small, these findings suggest that GABAergic interneurons may represent a vulnerable target in the brains of individuals with autism, potentially impacting upon their key role in learning and information processing. These preliminary findings further suggest the need for future more expanded studies in a larger number of postmortem brain samples from cases of autism and controls.

Region- and age-specific deficits in γ-aminobutyric acidergic neuron development in the telencephalon of theuPAR-/- mouse

We have previously shown that in adult mice with a null mutation in the urokinase-type plasminogen activator receptor (uPAR) gene, maintained on a C57BL/6J/129Sv background, there is a selective loss of GABAergic interneurons in anterior cingulate and parietal cortex, with the parvalbumin-expressing subpopulation preferentially affected. Here, we performed a more detailed anatomical analysis of uPAR(-/-) mutation on the congenic C57BL/6J background. With glutamic acid decarboxylase-67 and gamma-aminobutyric acid (GABA) immunostaining, there is a similar region-selective loss of cortical interneurons in the congenic uPAR(-/-) mice from the earliest age examined (P21). In contrast, the loss of parvalbumin-immunoreactive cells is observed only in adult cortex, and the extent of this loss is less than in the mixed background. Moreover, earlier in development, although there are normal numbers of parvalbumin cells in the uPAR(-/-) cortex, fewer cells coexpress GABA, suggesting that the parvalbumin subpopulation migrates appropriately to the cortex, but does not differentiate normally. Among the other forebrain regions examined, only the adult hippocampus shows a loss of GABAergic interneurons, although the somatostatin, rather than the parvalbumin, subpopulation contributes to this loss. The data suggest that uPAR function is necessary for the normal development of a subpopulation of GABAergic neurons in the telencephalon. It is likely that the late-onset parvalbumin phenotype is due to the effects of an altered local environment on selectively vulnerable neurons and that the extent of this loss is strain dependent. Thus, an interplay between complex genetic factors and the environment may influence the phenotypic impact of the uPAR mutation both pre- and postnatally.

Long-term behavioral and morphological consequences of nonconvulsive status epilepticus in rats

The aims of the present study were to ascertain whether nonconvulsive status epilepticus (NCSE) could give rise to long-term behavioral deficits and permanent brain damage. Two months after NCSE was elicited with pilocarpine (15 mg/kg i.p.) in LiCl-pretreated adult male rats, animals were assigned to either behavioral (spontaneous behavior, social interaction, elevated plus-maze, rotorod, and bar-holding tests) or EEG studies. Another group of animals was sacrificed and their brains were processed for Nissl and Timm staining as well as for parvalbumin and calbindin immunohistochemistry. Behavioral analysis revealed motor deficits (shorter latencies to fall from rotorod as well as from bar) and disturbances in the social behavior of experimental animals (decreased interest in juvenile conspecific). EEGs showed no apparent abnormalities. Quantification of immunohistochemically stained sections revealed decreased amounts of parvalbumin- and calbindin-immunoreactive neurons in the motor cortex and of parvalbumin-positive neurons in the dentate gyrus. Despite relatively inconspicuous manifestations, NCSE may represent a risk for long-term deficits.

Recurrent excitation of granule cells with basal dendrites and low interneuron density and inhibitory postsynaptic current frequency in the dentate gyrus of macaque monkeys

Temporal lobe epilepsy is often associated with pathological changes in the dentate gyrus, and such changes may be more common in humans than in some nonprimate species. To examine species-specific characteristics that might predispose the dentate gyrus to epileptogenic damage, we evaluated recurrent excitation of granule cells with and without basal dendrites in macaque monkeys, measured miniature inhibitory postsynaptic currents (mIPSCs) of granule cells in macaque monkeys and compared them to rats, and estimated the granule cell-to-interneuron ratio in macaque monkeys and rats. In hippocampal slices from monkeys, whole-cell patch recording revealed antidromically evoked excitatory PSCs that were four times larger and inhibitory PSCs that were over two times larger in granule cells with basal dendrites than without. These findings suggest that granule cells with basal dendrites receive more recurrent excitation and, to a lesser degree, more recurrent inhibition. Miniature IPSC amplitude was slightly larger in monkey granule cells with basal dendrites than in those without, but mIPSC frequency was similar and only 26% of that reported for rats. In situ hybridization for glutamic acid decarboxylase and immunocytochemistry for somatostatin, parvalbumin, and neuronal nuclei revealed interneuron proportions and distributions in monkeys that were similar to those reported for rats. However, the interneuron-to-granule cell ratio was lower in monkeys (1:28) than in rats (1:11). These findings suggest that in the primate dentate gyrus, recurrent excitation is enhanced and inhibition is reduced compared with rodents. These primate characteristics may contribute to the susceptibility of the human dentate gyrus to epileptogenic injuries.

Time-dependent distribution and neuronal localization of c-fos protein in the rat hippocampus following 4-aminopyridine seizures

The immunohistochemical localization of c-fos protein in the CNS neurons was studied in a model of generalized epilepsy induced by the intraperitoneal injection of 4-aminopyridine to adult Wistar rats. This specific blocker of the voltage-dependent potassium channels proved to be suitable for use in the investigation of epileptogenesis. Following the treatment of adult rats with 5 mg kg of 4-aminopyridine, the animals experienced generalized seizures. At the end of the experiment, the rats were briefly anesthetized and perfused with fixative. Frozen coronal plane sections were cut and processed for immunohistochemistry, using polyclonal c-fos antibody. The number and distribution of immunostained cell nuclei in the hippocampus were analyzed in detail with the help of a digital microscope camera and a morphometry program. The highest level of immunostaining was detected in most of the structures at 3 h, but the level had decreased to the control level by 5 h following 4-aminopyridine injection. In the dentate fascia, immunostaining was highest at 1 h and then decreased slowly until 5 h post-injection. The activated neuronal assemblies were analyzed with the aid of parvalbumin c-fos double immunostaining. These countings revealed the highest inhibitory interneuronal activation in every part of the hippocampus (including the dentate fascia) at 3 h post-injection. The results indicate that systemic 4-aminopyridine induces limbic seizures, which are probably initiated in the entorhinal cortex.

Parvalbumin-immunoreactive neurons in the hippocampal formation of Alzheimer's diseased brain

The number and topographic distribution of immunocytochemically stained parvalbumin interneurons was determined in the hippocampal formation of control and Alzheimer's diseased brain. In control hippocampus, parvalbumin interneurons were aspiny and pleomorphic, with extensive dendritic arbors. In dentate gyrus, parvalbumin cells, as well as a dense plexus of fibers and puncta, were associated with the granule cell layer. A few cells also occupied the molecular layer. In strata oriens and pyramidale of CA1-CA3 subfields, parvalbumin neurons gave rise to dendrites that extended into adjacent strata. Densely stained puncta and beaded fibers occupied stratum pyramidale, with less dense staining in adjacent strata oriens and radiatum. Virtually no parvalbumin profiles were observed in stratum lacunosum-moleculare or the alveus. Numerous polymorphic parvalbumin neurons and a dense plexus of fibers and puncta characterized the deep layer of the subiculum and the lamina principalis externa of the presubiculum. In Alzheimer's diseased hippocampus, there was an approximate 60% decrease in the number of parvalbumin interneurons in the dentate gyrus/CA4 subfield (P<0.01) and subfields CA1-CA2 (P<0.01). In contrast, parvalbumin neurons did not statistically decline in subfields CA3, subiculum or presubiculum in Alzheimer's diseased brains relative to controls. Concurrent staining with Thioflavin-S histochemistry did not reveal degenerative changes within parvalbumin-stained profiles. These findings reveal that parvalbumin interneurons within specific hippocampal subfields are selectively vulnerable in Alzheimer's disease. This vulnerability may be related to their differential connectivity, e.g., those regions connectionally related to the cerebral cortex (dentate gyrus and CA1) are more vulnerable than those regions connectionally related to subcortical loci (subiculum and presubiculum).

A Golgi study of the short-axon interneurons of the cell layer and inner plexiform layer of the medial cortex of the lizard Podarcis hispanica

The medial cortex of lizards is a three-layered brain region displaying cyto- and chemoarchitectonical, connectional, and ontogenetic characteristics that relate it to the hippocampal fascia dentata of mammals. Three interneuron types located in the cell layer and ten others in the inner plexiform layer (six in the juxtasomatic zone and four in the deep zone) are described in this study. The granuloid neurons, web-axon neurons, and deep-fusiform neurons lay within the cell layer. These neurons were scarce; they were probably gamma-aminobutyric acid (GABA)-, and parvalbumin-immunoreactive and presumably participated in feed forward as well as in feed back inhibition of the principal projection cells of the lizard medial cortex. In the juxtasomatic inner plexiform layer, the smooth vertical neurons, smooth horizontal neurons, small radial neurons, large radial neurons, pyramidal-like radial neurons, and spheroidal neurons were found. They were all probably GABA-, and parvalbumin-immunoreactive and were involved in feed forward inhibition of principal medial cortex cells. In the deep inner plexiform layer lay the giant-multipolar neurons, long-spined polymorphic neurons, periventricular neurons, and alveus-horizontal neurons. These neurons were probably GABA-immunoreactive and either neuropeptide- (somatostatin-neuropeptide Y) or parvalbumin-immunoreactive. They seemed to be involved in feed back or even occasionally in feed forward inhibition phenomena.

Neurons co-localizing calretinin immunoreactivity and reduced nicotinamide adenine dinucleotide phosphate diaphorase (NADPH-d) activity in the hippocampus and dentate gyrus of the rat

Co-localization of calretinin immunoreactivity and nicotinamide adenine dinucleotide phosphate diaphorase (NADPH-d) activity was studied in the rat hippocampus and dentate gyrus. Neurons co-expressing both markers (CR/NADPH-d) were observed throughout the hippocampus and dentate gyrus. However, they were more abundant in the stratum pyramidale and radiatum of CA3, stratum pyramidale of CA1, and in the juxtagranular zone of the hilus. The NADPH-d activity appeared in 37% of the calretinin immunoreactive neurons in CA3, 42% in CA1, and 36% in the dentate gyrus, whereas calretinin immunoreactivity occurred in 41% of the NADPH-d positive neurons in the hippocampus, and 16% in the dentate gyrus. The morphology and location of the double marked cells could not be used as a characteristic of the co-localizing neurons. The heavily stained NADPH-d neurons occurring mainly in CA1 do not show calretinin immunoreactivity. NADPH-d fiber swellings could be observed in close apposition to calretinin immunoreactive neurons and dendrites, suggesting synaptic contacts. It has been reported that calretinin immunoreactivity and NADPH-d activity co-localize infrequently in other areas such as the neocortex, striatum, hypothalamus and tegmental nucleus. The relatively high proportion of double marked cells found in the hippocampus and dentate gyrus could be indicative of the importance of the CR/NADPH-d interneurons in the circuitries of these areas.

Rapid and transient increases in cellular immediate early gene and neuropeptide mRNAs in cortical and limbic areas after amygdaloid kindling seizures in the rat

Changes in transcription factor and neuropeptide gene expression are likely to be involved in the cascade of genetic and molecular events leading to permanent changes in neuronal activity associated with kindling and epilepsy. Both acute-transient and delayed-sustained changes in transcription factor or immediate early gene (IEG) activity have previously been reported in response to different stimuli. In the present study in situ hybridization was used to investigate the possible time course (30 min-8 week) of IEG and neuropeptide mRNA induction in forebrain in a kindling model of epilepsy. Kindling was produced by daily unilateral stimulation of the amygdala. IEG mRNAs were detected using [35S]-labelled oligonucleotide probes specific for c-fos, c-jun, NGFI-A (PC1) and PC3 transcripts. Possible changes in the level of mRNAs encoding the neuropeptides somatostatin (SOM) and neuropeptide Y (NPY) were also studied. Stimulation-induced seizures produced dramatic bilateral increases in all IEG mRNAs in the dentate gyrus after 30 min to 1 h. Ipsilateral or bilateral increases in c-fos and PC3 mRNA were observed in the piriform cortex of individual animals at 30 min post-stimulation. While the distribution and apparent basal expression of the different IEGs varied (NGFI-A and c-jun > c-fos and PC3), the degree of induction in the dentate gyrus was similar for all IEGs studied (i.e. 200-300%). No long-term changes in IEG mRNA expression were detected beyond 2 h and up to 8 week after the last seizure. Increased levels of preproSOM and preproNPY mRNAs were consistently observed in hilar interneurons, but not in pyramidal or granule cells of the hippocampus, after 1-2 h. These increases were not maintained at later times. The short-term effects on IEG and neuropeptide mRNAs observed suggest that these changes are consequence of seizure activity with the development of kindling. In contrast, no evidence was found of any substantial, long-lasting effects on these parameters associated with the established kindled state.

Prolonged induction of c-fos in neuropeptide Y- and somatostatin-immunoreactive neurons of the rat dentate gyrus after electroconvulsive stimulation

Induction of c-fos mRNA and Fos was studied in the hilus and granular layer of the dentate gyrus at various times up to 24 h after single electroconvulsive stimulation (ECS) using in situ hybridization and immunocytochemistry. In both areas of the dentate gyrus, a prominent induction of c-fos mRNA and Fos was observed. Compared to the granular layer, however, c-fos mRNA and Fos in hilar cells reached maximum later and remained elevated considerably longer. Several neurochemically distinct populations of hilar neurons have been described, some of which contain neuropeptide Y (NPY) and/or somatostatin (SS). Using double-labelling immunocytochemistry, we examined to what extent Fos was induced in these hilar neurons after ECS. Although a minor population of non-NPY non-SS cells displayed Fos induction early after ECS, prolonged induction of Fos almost exclusively occurred in NPY or SS neurons. The Fos-immunoreactive NPY or SS neurons only amounted to about 50% of the total hilar population of NPY or SS neurons. The present observations suggest that a subpopulation of hilar NPY and SS neurons may be central to the actions of electroconvulsive seizures in the dentate gyrus.

Distribution of substance P-immunoreactive neurons and fibers in the monkey hippocampal formation

Substance P containing neurons was visualized by immunocytochemistry in the monkey hippocampus, subicular complex, and entorhinal cortex. Immunoreactive neurons were found solely in the hilar region of the dentate gyrus, and in strata oriens and pyramidale of Ammon's horn. In the subicular complex, immunoreactive neurons were located in those layers which were close to the alveus, whereas in the entorhinal cortex most of the substance P-positive neurons appeared in the second and third layers above the lamina dissecans. The majority of substance P-containing neurons were large multipolar cells, but small bipolar and multipolar cells also occurred in Ammon's horn, subiculum and entorhinal cortex. Dendrites of immunoreactive cells were smooth and displayed a few small, faintly stained spines which were hard to identify in the light microscopic preparations, but were visible with electron microscopy. Substance P-positive dendrites were exclusively found in the hilar region and never observed in the upper two-thirds of the molecular layer of the dentate gyrus. Moreover, immunoreactive dendrites rarely penetrated the stratum lacunosum-moleculare of Ammon's horn. In the electron microscopic preparations, somal and dendritic features of substance P-positive neurons were similar to those observed for GABAergic local circuit neurons. Axons of the substance P-immunoreactive local circuit neurons were thin and richly arborized in the upper two-thirds of the molecular layer of the dentate gyrus, in the stratum lacunosum-moleculare of Ammon's horn as well as in the subpial layers of the subicular complex and entorhinal cortex. Their terminals formed exclusively symmetric synapses with dendrites and spines. However, substance P-immunoreactive boutons were not found to make symmetric, axosomatic synapses on the granule cells of the dentate gyrus and very few were present on the pyramidal neurons of Ammon's horn, subicular complex, and entorhinal cortex. Hippocampal neurons, which were immunoreactive for substance P, also contained the neuropeptide somatostatin. However, not all of the somatostatin-containing neurons were substance P-immunoreactive. Thus, substance P-positive neurons are a subpopulation of somatostatin immunoreactive, GABAergic neurons. In conclusion, substance P-immunoreactive neurons are ideally suited for feed-back dendritic inhibition which may control the effectiveness of the main excitatory cortical input to the granule cells of the dentate gyrus and pyramidal neurons of the Ammon's horn.

AMPA receptors in the rat and primate hippocampus: a possible absence of GluR2/3 subunits in most interneurons

Amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptors are assembled from the four subunits GluR1, 2, 3, 4 (or GluRA, B, C, D). AMPA channels that do not contain the GluR2 subunit are permeable to calcium. Recent studies indicate that excitotoxic as well as epileptic and ischemic cell damage may be mediated not only by N-methyl-Daspartate receptors, but also by AMPA receptors. The majority of interneurons in the hippocampus are resistant, but subsets of interneurons are consistently damaged in different disease states. Single immunolabeling using antibodies against AMPA receptor subunits, together with double immunolabeling for calcium-binding proteins (parvalbumin, calbindin and calretinin) and the neuropeptide somatostatin, were performed to study GluR1-4 immunoreactivity in interneuronal populations and principal cells. The ultrastructure of GluR1-4 labeled neurons was also examined using electron microscopy. With the exception of calbindin-positive interneurons, GluR2/3 was absent from hippocampal interneurons in both rat and monkey. In the rat, interneurons were more strongly immunoreactive against GluR1 than principal cells. In the monkey, immunoreactivity for GluR4 in interneurons was stronger than for GluR1. All GluR subunits were confined to spines, dendritic membrane and cytoplasm surrounding the nucleus but absent from axons and presynaptic terminals. Our findings suggest that hippocampal principal cells and interneurons express different complements of AMPA receptor subunits. Furthermore, the absence of GluR2 and/or GluR3 in both vulnerable and resistant interneurons subtypes indicates that knowledge of receptor subunit composition is not sufficient to predict neuronal vulnerability.

Pyramidal cell dendrites are the primary targets of calbindin D28k-immunoreactive interneurons in the hippocampus

The axonal arborization and postsynaptic targets of calbindin D28k (CB)-immunoreactive nonprincipal neurons have been studied in the rat dorsal hippocampus. Two types of neurons were distinguished on the basis of soma location, the characteristics of the dendritic free, and the axon arborisation pattern. Type I cells were located in stratum radiatum of the CA1 and CA3 regions and occasionally in strata pyramidale and oriens. These cells had multipolar or bitufted dendritic trees primarily located in stratum radiatum. Their axons could be followed for a considerable distance, arborised within stratum radiatum, and were covered with regularly spaced small boutons. As demonstrated with postembedding immunogold staining, their axon terminals were gamma-aminobutyric acid (GABA) immunoreactive, and formed symmetrical synapses predominantly on proximal and distal dendrites of pyramidal cells (28% and 58%, respectively), and occasionally on spines (9%) or on GABA-positive dendrites (5%). Type II cells were found exclusively in stratum oriens of the CA1 and CA3 regions and possessed large, fusiform cell bodies and long, horizontally oriented dendrites. Their axon initial segments turned towards the alveus and disappeared in a myelin sheet, which was often possible to follow into the white matter. We conclude that type I CB-immunoreactive cells are likely to represent a major source of inhibitory synapses in the dendritic region of pyramidal cells, which are responsible for the control of dendritic electrogenesis. The distribution of local collaterals of type II cells-if they have any-remains unknown, but their main axon is likely to project to the medial septum.

Co-localization of somatostatin mRNA and parvalbumin in the dorsal rat hippocampus after cerebral ischemia

Following transient global ischemia most of the neurons containing somatostatin in the fascia dentata of the dorsal hippocampal formation die, while somatostatinergic neurons in the CA1 region survive. The neurons react to ischemia with a transiently reduced expression of somatostatin mRNA and peptide. We have tested the hypothesis that this selective vulnerability is solely related to those somatostatinergic neurons which do not express the calcium-binding protein parvalbumin. Postischemic changes were studied in rat dorsal hippocampus at 2 and 16 days after 10 min of global cerebral ischemia using a four-vessel occlusion model. We performed a double-staining visualizing the mRNA coding for somatostatin by non-radioactive in situ hybridization and parvalbumin protein by immunocytochemistry. Only 5% of the somatostatinergic cells in the fascia dentata contained parvalbumin. The number of somatostatinergic cells was permanently reduced following ischemia. Among surviving neurons we found cells with and without parvalbumin expression. Thus, expression of parvalbumin is not predictive for survival of somatostatinergic cells in the fascia dentata. In contrast, in CA1, 37% of the somatostatinergic cells contained parvalbumin. These cells were unaffected by the transient ischemic period. The somatostatinergic cells lacking parvalbumin showed transiently reduced mRNA levels at day 2, but recovered to control values at the 16th postischemic day. Thus, expression of the calcium-buffering protein parvalbumin coincides with resistance of somatostatinergic neurons in CA1 to transient effects of ischemia. We conclude that the calcium-buffering capacity of parvalbumin may partially contribute to the protection of somatostatinergic neurons from ischemia in the dorsal hippocampus. However, the survival of somatostatinergic cells without parvalbumin indicates the importance of other factors as well.

Connection matrix of the hippocampal formation: I. The dentate gyrus

The hippocampal formation presents a special opportunity for realistic neural modeling since its structure, connectivity, and physiology are better understood than that of other cortical components. A review of the quantitative neuroanatomy of the rodent dentate gyrus (DG) is presented in the context of the development of a computational model of its connectivity. The DG is a three-layered folded sheet of neural tissue. This sheet is represented as a rectangle, having a surface area of 37 mm2 and a septotemporal length of 12 mm. Points, representing cell somata, are distributed in the model rectangle in a roughly uniform fashion. Synaptic connectivity is generated by assigning each presynaptic cell a spatial zone representing its axonal arbor. For each postsynaptic cell, a list of potential presynaptic cells is compiled, based on which arbor zones the given postsynaptic cell falls within. An appropriate number of presynaptic inputs are then selected at random. The principal cells of the DG, the granule cells, are represented in the model, as are non-principal cells, including basket cells, chandelier cells, mossy cells, and GABAergic peptidergic polymorphic (GPP) cells. The neurons of layer II of the entorhinal cortex are included also. The DG receives its main extrinsic input from these cells via the perforant path. The basket cells, chandelier cells, and GPP cells receive perforant path and granule cell input and exert both feedforward and feedback inhibition onto the granule cells. Mossy cells receive converging input from granule cells and send their output back primarily to distant septotemporal levels, where they contact both granule cells and non-principal cells. To permit numerical simulations, the model must be scaled down while preserving its anatomical structure. A variety of methods for doing this exist. Hippocampal allometry provides valuable clues in this regard.

Alterations of inhibitory synaptic responses in the dentate gyrus of temporal lobe epileptic patients

The number of orthodromically evoked population spikes was used to classify brain slice tissue from the dentate gyrus of temporal lobe epileptic patients as "more excitable" (multiple population spikes) or "less excitable" (a single population spike). During orthodromic stimulation, "more excitable" tissue exhibited less paired-pulse depression in comparison to "less excitable" tissue. During antidromic stimulation, both multiple population spikes and paired-pulse depression were observed in "more excitable" tissue. "Less excitable" tissue exhibited a single antidromic spike and often no antidromically evoked paired-pulse depression. The strength of antidromic paired-pulse depression was correlated positively with the number of antidromic spikes and was correlated negatively with orthodromic paired-pulse depression. Although orthodromic and antidromic paired-pulse depression were correlated to the number of orthodromically evoked population spikes, this correlation was not as strong as that between orthodromic paired-pulse depression, antidromic paired-pulse depression, and number of antidromically evoked population spikes. The antidromic paired-pulse depression observed in tissue exhibiting antidromically evoked multiple population spikes was enhanced rather than blocked by bicuculline. In addition, the blockade of the antidromic paired-pulse depression by CNQX indicated that this inhibition is mediated by an AMPA-type glutamatergic synapse. We suggest that alterations in circuitry occur in the dentate gyrus of some temporal lobe epileptic patients and were manifested by both a loss of inhibitory input as well as an increase of inhibition, which was dependent on the pathway of stimulation. The results of pairing antidromic and orthodromic stimuli were consistent with these conclusions.

Substance P-containing hypothalamic afferents to the monkey hippocampus: an immunocytochemical, tracing, and coexistence study

In order to identify the synaptic connections of substance P-containing afferents within the hypothalamo-hippocampal projection of the monkey, we performed a combined light and electron microscopic, immunocytochemical study, made lesions of the fimbriafornix, and employed retrograde tracing using WGA-HRP. Furthermore, coexistence studies for substance P and GAD were performed to identify the putative transmitters of these hypothalamic projection neurons. A plexus of large substance P-immunoreactive terminals was identified in both the innermost portion of the molecular layer and in CA2. Axon terminals in both plexuses established exclusively asymmetric synapses with spines and dendritic shafts. Substance P-immunoreactive boutons were degenerating 5 days after lesioning, and had disappeared 10 days after ipsilateral fimbria-fornix transection. Thus, these terminals were of extrinsic origin. In contrast, immunoreactive fibers in the outer third of the dentate molecular layer remained unaffected by the lesion. Retrograde tracing combined with immunostaining for substance P revealed the parent cell bodies of the extrinsic substance P-containing afferents in the supramammillary nucleus. Colocalization studies employing a consecutive semi-thin sections technique indicate that these large substance P-containing projection neurons lack GABA as an inhibitory transmitter. These results suggest that hypothalamic afferents of the monkey hippocampus contain substance P. Because these afferents lack GABA as an inhibitory transmitter and establish exclusively asymmetric synapses, this projection may excite hippocampal target neurons.

Colocalization of nitric oxide synthase and somatostatin immunoreactivity in rat dentate hilar neurons

Distribution of nitric oxide synthase (NOS), somatostatin (SSN), and parvalbumin (PV) was studied in the rat hippocampus by immunohistochemical methods. The aim was to explore the interrelationship between SSN-immunoreactive (SSN-IR) neurons in the dentate hilus, which have been shown to be vulnerable to a number of pathophysiological insults, and the presence or absence of NOS and/or PV in the same subset of dentate hilar neurons. Small NOS-IR neurons were scattered in the pyramidal, oriens, and radiatum layers of the CA1-CA3 areas and in the subiculum, where larger NOS-IR neurons were occasionally noted. In the area dentata, NOS-IR neurons, which were composed of small and large polymorphic cells, appeared as a single file at the hilar border with the granule cell layer and clustered in the hilus in fairly high density. Double-labeling techniques showed that most NOS-IR neurons in the hilus were SSN-IR, whereas coexistence of NOS and PV immunoreactivity or SSN and PV immunoreactivity was low in dentate hilar neurons. In other areas of the hippocampus, colocalization of NOS and SSN in the same neurons was much less frequent. Thus, SSN-IR neurons in the dentate hilus constitute a population of neurons that contain the enzyme NOS as well. The presence of NOS coupled to the lack or low level of PV in this group of neurons may provide a neurochemical basis for their high susceptibility to certain pathophysiological insults.

Parvalbumin-containing neurons in the cerebral cortex of the lizard Podarcis hispanica: morphology, ultrastructure, and coexistence with GABA, somatostatin, and neuropeptide Y

The morphology, fine structure, and degree of colocalization with GABA, somatostatin, and neuropeptide Y of parvalbumin-containing cells were studied with immunocytochemistry in the cerebral cortex of the lizard Podarcis hispanica. Parvalbumin-containing cells make up a morphologically heterogeneous population of spine-free neurons, displaying the morphological features of nonprincipal cells previously described in Golgi studies. Electron microscopically, parvalbumin-immunoreactive cell bodies are similar in all cortical areas and layers. The perisomatic input is moderate in number, and boutons with either round clear vesicles or flattened vesicles were observed making asymmetric or symmetric synaptic contacts, respectively. Parvalbumin-immunoreactive dendrites are smooth and almost completely covered with synaptic boutons of different types, most of which establish asymmetric contacts. Parvalbumin-immunoreactive boutons are concentrated around cell bodies of principal cells. They are large, containing abundant mitochondria and small pleomorphic vesicles, and establishing symmetric synaptic contacts with somata, proximal dendritic shafts, and axon initial segments of principal cells. Colocalization studies revealed that all the parvalbumin-containing cells are GABA-immunoreactive, representing only a fraction of the GABA-immunopositive cell population, and that parvalbumin- and peptide- (somatostatin and neuropeptide Y) containing cells show a negligible overlap. These results demonstrate that in the cerebral cortex of the lizard Podarcis hispanica, parvalbumin-containing cells represent a subset of nonprincipal GABAergic neurons largely involved in perisomatic inhibition, which are different from the peptide-containing cells, and suggest that they may include both axosomatic and axoaxonic cells.

Subpopulations of GABA neurons containing somatostatin, neuropeptide Y, and parvalbumin in the dorsomedial cortex of the lizard Psammodromus algirus

Different subpopulations of GABA neurons containing the neuropeptides somatostatin and neuropeptide Y, and the calcium binding protein parvalbumin were studied by immunocytochemistry using light and electron microscopy in the dorsomedial cortex of the lizard Psammodromus algirus to investigate the connectivity of different subsets of GABA neurons in the lizard dorsomedial cortical circuitry and to compare cortical regions of reptiles and mammals. GABA neurons were classified into different subsets by using the peroxidase anti-peroxidase immunohistochemical method on adjacent Araldite-embedded semithin sections. GABA neurons in the dorsomedial cortex fall into three major subsets: 1) neurons with somatostatin (and neuropeptide Y), which accounted for about 44% of the GABA population; 2) neurons with parvalbumin, which accounted for about 13% of the GABA neurons; and 3) neurons without parvalbumin or neuropeptides, which represented 40% of all GABA cells. This division of GABA neurons in non-overlapping subpopulations of neuropeptide- and parvalbumin-containing cells is similar to that found in the mammalian hippocampal formation. On the basis of the nerve terminal fields, somatostatin- and parvalbumin-immunoreactive neuronal populations appear to be functionally different, acting on different portions of the projection neurons. Parvalbumin-immunoreactive neurons inhibit the pyramidal neurons at the cell body level, whereas somatostatin-immunoreactive neurons inhibit them on distal dendrites. The results of the present study add more similarities between the lizard dorsomedial cortex and parts of the mammalian hippocampus.

Calretinin immunoreactivity in the monkey hippocampal formation--II. Intrinsic GABAergic and hypothalamic non-GABAergic systems: an experimental tracing and co-existence study

Our light and electron microscopic studies (Seress L., Nitsch R. and Leranth C. (1993) Neuroscience 55, 775-796.) indicated that in the hippocampus of the African Green monkey, calretinin is exclusively present in non-pyramidal cells. Calretinin-positive axons formed a prominent band at the border of the dentate molecular and granule cell layers and in the pyramidal layer of CA2, and established asymmetric synapses with different postsynaptic targets. The goal of this study is to determine the cells of origin of this presumably extrinsic innervation, and subsequently, the characterization of their neurochemical features. We were able to demonstrate that calretinin-immunoreactive axon terminals in the inner molecular layer of the dentate gyrus and in the pyramidal layer of CA2 disappear 10 days after fimbria-fornix transection. Retrograde tracing revealed their cells of origin to be in the supramammillary nucleus. Co-localization studies employing the cryostat consecutive, semithin section technique provided evidence that these large projecting neurons contained both calretinin and substance-P but lack GABA as an inhibitory transmitter. In contrast, co-localization studies revealed that almost all of the intrinsic calretinin-positive neurons in different areas of the primate hippocampus contained GAD or GABA. These results suggest that there are two separate calretinin-containing systems in the primate hippocampus, i.e. an intrinsic inhibitory and an extrinsic excitatory one, the latter deriving from the supramammillary nucleus of the hypothalamus.

Calretinin immunoreactivity in the monkey hippocampal formation--I. Light and electron microscopic characteristics and co-localization with other calcium-binding proteins

Calretinin-containing neurons were visualized by immunocytochemistry in the monkey hippocampal formation, subicular complex, and entorhinal cortex. Calretinin-immunoreactivity was present exclusively in non-granule cells of the dentate gyrus and in non-pyramidal cells of Ammon's horn, subiculum and entorhinal cortex. Most frequently, calretinin-positive neurons were found at the hilar border of the dentate granule cell layer and in the stratum radiatum of CA1-3 areas. In the subicular complex, immunoreactive neurons were evenly distributed in all layers, whereas in the entorhinal cortex, they were accumulated in external layers above the lamina dissecans. Distinct bands of calretinin-positive fibers occupied the supragranular zone of the molecular layer in dentate gyrus, the pyramidal cell layer of the CA2 area in Ammon's horn and the upper two layers of presubiculum. The majority of calretinin-immunoreactive neurons were small, bipolar or fusiform neurons with a dendritic tree oriented parallel to the dendrites of principal cells (granule cells in dentate gyrus and pyramidal neurons elsewhere). Dendrites were smooth or sparsely spiny, displaying small spines of conventional type. Co-existence studies showed that these neurons were completely devoid of other calcium-binding proteins, parvalbumin and calbindin. Electron microscopic analysis revealed somata of immunoreactive neurons which contained a large nucleus and a small cytoplasmic rim, which contained only few organelles. The nucleus displayed deep infoldings and intranuclear rods. Input synapses of immunoreactive neurons were rare both on somata and dendrites and large surface areas were frequently apposed by glial processes. This was very prominent in the dentate gyrus and Ammon's horn. Axons of calretinin-positive neurons were thin, arborized in all layers and had small varicosities. Their terminals formed symmetric synaptic contacts mainly with dendrites and less frequently with somata of principal cells. Axon terminals of calretinin-immunoreactive fiber bundles in the supragranular layer, as well as in the pyramidal layer of the CA2 area, formed asymmetric synaptic contacts with dendritic shafts. In addition, they established asymmetric axospinous and axosomatic synaptic contacts with granule cells of the dentate gyrus. In the presubiculum, the calretinin-positive axon bundle included a large number of immunoreactive myelinated axons, as well as axon terminals. The characteristic location and features of synapses suggests that these fibers derive from extra-hippocampal afferents (Nitsch, R. and Leranth C. (1993) Neuroscience 55, 797-812) and not from the calretinin-immunoreactive neurons of the hippocampal formation.

Changes in parvalbumin-immunoreactive neurons in the rat hippocampus following a kainic acid lesion

Changes in a sub-population of hippocampal non-pyramidal neurons following a unilateral lesion with kainic acid were examined using an antibody raised against the Ca-binding protein parvalbumin. A loss of 71-97% of the parvalbumin-immunoreactive neurons occurred at the three post-lesion times studied (1, 2 and 4 weeks) in all areas of the ipsilateral hippocampus, but no such loss was observed in the dentate gyrus. Resistant parvalbumin-immunoreactive neurons occurred principally in stratum pyramidale and displayed altered morphology from the normal with swollen dendrites and dendritic varicosities. The contralateral hippocampus exhibited losses of parvalbumin-immunoreactive cells, but this was restricted to stratum oriens of CA1. This data demonstrates the loss of a specific and important population of non-pyramidal neurons which might be responsible for the chronic loss of functional inhibition seen in this animal model of temporal lobe epilepsy.

Distribution of parvalbumin-immunoreactive cells and fibers in the monkey temporal lobe: the hippocampal formation

The distribution of parvalbumin-immunoreactive cells and fibers in the various fields of the hippocampal formation was studied in the macaque monkey. Parvalbumin-immunoreactive neurons had aspiny or sparsely spiny dendrites that often had a beaded appearance; most resembled classically identified interneurons. Parvalbumin-immunoreactive fibers and terminals were confined to certain laminae in each field and generally had a pericellular distribution. In the dentate gyrus, there was a dense pericellular plexus of immunoreactive terminals in the granule cell layer. Except for a narrow supragranular zone, there was a marked paucity of terminals in the molecular and polymorphic cell layers. Immunoreactive neurons were mainly located immediately subjacent to the granule cell layer and comprised a variety of morphological cell types. The three fields of the hippocampus proper (CA3, CA2, and CA1) demonstrated differences in their parvalbumin staining characteristics. In CA3, there was a prominent pericellular terminal plexus in the pyramidal cell layer that was densest distally (closer to CA2). Immunoreactive cells were located either in the pyramidal cell layer, where many had a pyramidal shape and prominent apical and basal dendrites, or in stratum oriens. CA2 had a staining pattern similar to that in CA3, though both the number of labeled cells and the density of the pericellular terminal plexus were greater in CA2. In CA1, there was a markedly lower number of parvalbumin-labeled cells than in CA3 and CA2 and the cells tended to be located in the deep part of the pyramidal cell layer or in stratum oriens. The pyramidal cell layer of CA1 contained a pericellular terminal plexus that was substantially less dense than in CA3 and CA2. At the border between CA1 and the subiculum there was a marked increase in the number of parvalbumin-immunoreactive neurons. The positive cells were scattered throughout the pyramidal cell layer of the subiculum and comprised a variety of sizes and shapes. Terminal labeling was higher in the pyramidal cell layer of the subiculum than in CA1. Layer II of the presubiculum had one of the highest densities of fiber and terminal labeling in the hippocampal formation. The density of staining was lower in the superficial portion of the layer where linear cartridges of presumed axo-axonic synapses were common. A large number of parvalbumin-immunoreactive cells were scattered throughout layer II of the presubiculum; small, spherical, multipolar cells were commonly observed in layer I. The parasubiculum had a somewhat lower density of positive cells and fibers than the presubiculum.(ABSTRACT TRUNCATED AT 400 WORDS)

GABAergic neurons of the rat dorsal hippocampus express muscarinic acetylcholine receptors

The expression of muscarinic acetylcholine receptors (mAChRs) in glutamic acid decarboxylase (GAD)-positive cells in the different strata of CA1, CA3, and the dentate gyrus (DG) of the dorsal hippocampus is examined by way of quantitative immunofluorescent double labeling employing M35, the monoclonal antibody raised against purified mAChR protein. Of all GAD-positive neurons, 97.5% express mAChRs. Conversely, 92.9% of the muscarinic cholinoceptive nonpyramidal neurons express GAD. These results indicate that the vast majority of the gamma-aminobutyric acid (GABA)ergic neurons express mAChRs. In addition to GAD, parvalbumin (PARV) and somatostatin (SOM) are two neurochemical substances notably expressed in GABAergic neurons. In order to examine whether the entire muscarinic cholinoceptive nonpyramidal cell group can be characterized by these three GABAergic markers, a cocktail of GAD, PARV, and SOM was used in a fluorescent double-labeling experiment with M35. These results show that 97.2% of all muscarinic cholinoceptive nonpyramidal neurons can be neurochemically characterized by the content of GAD, PARV, and SOM. In conclusion, nearly all GABAergic cells express mAChRs and, conversely, virtually the entire muscarinic cholinoceptive nonpyramidal cell group belongs to the GABAergic cell population. This study, therefore, provides anatomical evidence for an extensive neuronal connectivity of the hippocampal muscarinic cholinoceptive nonpyramidal system and the inhibitory GABAergic circuitry.

Distribution of GABAergic interneurons immunoreactive for calretinin, calbindin D28K, and parvalbumin in the cerebral cortex of the lizard Podarcis hispanica

The types and distribution of cells containing three calcium-binding proteins, calretinin, calbindin D28K, and parvalbumin, have been studied by immunocytochemistry in different areas of the cerebral cortex of lizards. Cross-reactivity of the antisera has been excluded by demonstrating the existence of several cell groups immunoreactive for one but not the other two calcium-binding proteins. In the dorsal and dorsomedial cortices all three proteins coexist in a single subpopulation of gamma-aminobutyric acid (GABA)ergic neurons, the terminals of which form pericellular baskets around cell bodies of bipyramidal neurons. The somata of these neurons are largely restricted to the cellular and inner plexiform layers, but the dendrites usually penetrate all layers, allowing the neurons to sample input from all possible sources. A small number of parvalbumin-containing neurons in the outer plexiform layer do not contain the other two proteins. The medial cortex, which is likely to be homologous to the mammalian dentate gyrus, only contains parvalbumin-immunoreactive neurons. The dendritic trees of these cells appear to avoid the Timm-positive fields receiving input from zinc-rich fiber collaterals, originating from principal cells. The lateral cortex contains calbindin D28K-immunoreactive GABAergic neurons, which lack the other two calcium-binding proteins. These neurons have horizontally running dendrites in the outer plexiform layer, but their axon terminals could not be visualized. The present study uncovered important similarities and differences between the lizard and the mammalian archicortex in the types of neurons containing calcium-binding proteins. As in mammals, different cell types evolved in the lizard to inhibit the perisomatic versus the distal dendritic region of principal cells, the calcium-binding protein-containing neurons being responsible for the former, and neuropeptide-containing neurons for the latter. The results also suggest that further neurochemical diversion of GABAergic interneurons coupled to a functional specialization took place during phylogenetic development from reptiles to mammals.

Calcium-binding proteins in the nervous system

Among the many calcium-binding proteins in the nervous system, parvalbumin, calbindin-D28K and calretinin are particularly striking in their abundance and in the specificity of their distribution. They can be found in different subsets of neurons in many brain regions. Although it is not yet known whether they play a 'triggering' role like calmodulin, or merely act as buffers to modulate cytosolic calcium transients, they are valuable markers of neuronal subpopulations for anatomical and developmental studies.

Immunohistochemical markers in rat cortex: co-localization of calretinin and calbindin-D28k with neuropeptides and GABA

Calretinin and calbindin-D28k are two calcium-binding proteins which are present in separate populations of interneurons in cerebral cortex and hippocampus. To identify these cells with the populations expressing different transmitters, two-colour immunofluorescence was done with antibodies against the calcium-binding proteins plus antibodies against vasoactive intestinal peptide (VIP), somatostatin (SRIF), or gamma-aminobutyric acid (GABA). In neocortex, calretinin is partially co-localized with VIP (especially in the deeper layers) and is not co-localized with SRIF. Calbindin is largely co-localized with SRIF, and not with VIP. Both calretinin and calbindin are partially co-localized with GABA. In piriform and entorhinal cortex, the patterns resemble those in neocortex. In hippocampus, preliminary data indicate greater heterogeneity, especially in the ventral part; at least a few double-positive cells are present for every combination of calcium-binding protein and neuropeptide. These results expand the known diversity of local-circuit neurons in cortical regions.

The synaptology of parvalbumin-immunoreactive neurons in the primate prefrontal cortex

Electron microscopy and immunocytochemistry with a monoclonal antibody against parvalbumin (PV) were combined to analyze the distribution and morphology of PV-immunoreactive (PV-IR) neurons and the synaptology of PV-IR processes in the principal sulcus of the macaque prefrontal cortex. Parvalbumin-IR neurons are present in layers II-VI of the macaque principal sulcus (Walker's area 46) and are concentrated in a band centered around layer IV. PV-IR cells are exclusively non-pyramidal in shape and are morphologically heterogeneous with soma sizes ranging from less than 10 microns to greater than 20 microns. Well-labeled neurons that could be classified on the basis of soma size and dendritic configuration resembled large basket and chandelier cells. A novel finding is that supragranular PV-IR neurons exhibit dendritic patterns with predominantly vertical orientations, whereas infragranular cells exhibit mostly horizontal or oblique dendritic orientations. PV-IR cells within layer IV exhibit a mixture of dendritic arrangements. Vertical rows of PV-IR puncta, 15-30 microns in length, resembling the "cartridges" of chandelier cell axons were most dense in layers II, superficial III, and the granular layer IV but were not observed in the infragranular layers. Cartridges were often present beneath unlabeled, presumed pyramidal cells. PV-IR puncta also formed pericellular nests around pyramidal cell somata and proximal dendrites, suggestive of basket cell innervation. PV-IR axons were occasionally observed in the white matter underlying the principal sulcus. Electron microscopic analysis revealed that PV-IR somata and dendrites are densely innervated by nonimmunoreactive terminals forming asymmetric (Gray type I) synapses as well as by fewer terminals forming symmetric (Gray type II) synapses. The majority of terminals forming symmetric synapses with PV-IR post-synaptic structures were not immunolabeled; however, some of these boutons did contain PV-immunoreactivity. PV-IR boutons exclusively form symmetric synapses and heavily innervate layer II/III pyramidal cells. PV-IR axon cartridges formed numerous axo-axonic synapses with the axon initial segments of pyramidal cells 15-20 microns beneath the axon hillock and also terminated on large axonal spines of the initial segment. Furthermore, we failed to observe a mixture of PV-immunoreactive and non-immunoreactive boutons composing a single axon cartridge. Pyramidal cell somata and proximal dendrites were also heavily innervated by PV-IR boutons forming symmetric synapses, again, consistent with basket cell innervation. In addition, PV-IR axon terminals frequently formed symmetric synapses with dendritic shafts and spines of unidentified neurons.(ABSTRACT TRUNCATED AT 400 WORDS)

Induction of immediate-early gene proteins in dentate granule cells and somatostatin interneurons after hippocampal seizures

The expression of the protein products of the immediate-early genes c-fos, Fos B, Fos-related proteins (FRAs), c-jun, jun B, jun D and krox-24 was investigated in the rat hippocampus at various times after electrically-induced hippocampal seizures. Hippocampal seizures induced all the immediate-early gene proteins in dentate granule cells with differing time-courses. In addition, Krox-24, Fos and Jun D were also induced in somatostatin-containing interneurons throughout the hippocampus and also in a small percentage of parvalbumin-containing interneurons. Thus, hippocampal seizures induce waves of immediate-early gene protein expression in dentate granule cells and a selective expression of krox-24, Fos and Jun D in hippocampal somatostatin interneurons. These results suggest that biochemical and/or morphological changes occurring in dentate granule cells and somatostatin interneurons after seizures may be regulated by immediate-early gene expression, and that these immediate-early gene proteins may be involved in seizure development in the nervous system.

Parvalbumin-positive neurons in rat dorsal hippocampus contain muscarinic acetylcholine receptors

The present study describes the colocalization of muscarinic acetylcholine receptors (mAChRs) and the calcium-binding protein parvalbumin (PARV) in nonpyramidal neurons of the rat dorsal hippocampus by means of dual-label immunocytochemistry. Fifty-two percent of all muscarinic cholinoceptive nonpyramidal neurons contain parvalbumin. Conversely, the vast majority (88.4%) of all PARVergic neurons possess mAChRs. The PARVergic neurons embedded within the CA1 pyramidal cell layer are nearly all (98.9%) immunopositive for mAChRs. These results indicate that the cholinergic septohippocampal projection represents a major afferent system upon the inhibitory PARVergic neurons.

Neuropeptide Y (NPY)-immunoreactive neurons in the primate fascia dentata; occasional coexistence with calcium-binding proteins: a light and electron microscopic study

Neuropeptide Y (NPY)-containing neurons are known to be highly vulnerable following sustained electrical stimulation in rats and in humans suffering from temporal lobe epilepsy. This has been related to a strong excitatory input. In contrast, there is evidence that neurons containing calcium-binding proteins exhibit a high resistance under experimental seizure and hypoxia conditions. The aim of this study was to determine the coexistence of NPY and calcium-binding proteins in inhibitory neurons of the primate fascia dentata and their synaptic connections. Vibratome sections of hippocampi of African green monkeys (Cercopithecus aethiops) were immunostained with antibodies against NPY, PARV, and CB. A quantitative coexistence study was performed for NPY and PARV on consecutive semithin sections. In contrast to the rodent hippocampus, NPY-immunoreactive neurons were found exclusively in the hilus of fascia dentata with horizontally oriented dendrites which did not extend into the granular and molecular layer. Conversely, PARV-immunoreactive neurons were also present in the granular and inner molecular layer and extended their dendrites far out in the molecular layer and the hilus. Axon terminals immunoreactive for NPY were mostly concentrated in the middle and outer molecular layer and the hilar region and were rare in the granular layer. PARV-immunoreactive boutons were basically restricted to the granular layer where they formed typical baskets. The antibody against calbindin stained almost exclusively granule cells. Coexistence of NPY- and PARV-immunoreactivity was found only in hilar neurons and was rare (9 out of 152 cells analyzed). These results suggest that most NPY-immunoreactive neurons do not contain calcium-binding proteins. NPY-containing neurons exhibited ultrastructural characteristics as described for inhibitory neurons. Their dendrites were only sparsely contacted by mostly asymmetric synaptic terminals, including a very small number of mossy fiber axon terminals. In turn, numerous NPY-immunoreactive axon terminals formed symmetric synapses with spines and dendritic shafts of unlabeled neurons in the middle and outer molecular layer, whereas no contact with granule cell bodies was evident. Thus, we conclude that the vulnerability of NPY-containing inhibitory neurons may be due more to the lack of calcium-binding proteins than to a strong excitatory innervation. As their axons may contribute to the inhibitory control of the major excitatory input from the entorhinal cortex, their loss following overstimulation may play a role in perpetuating hippocampal seizure activity.

Vigabatrin pre-treatment prevents hilar somatostatin cell loss and the development of interictal spiking activity following sustained simulation of the perforant path

Somatostatin-containing neurons in the hilus of the dentate gyrus are known to be exceptionally vulnerable in experimental models of epilepsy, as well as in human temporal lobe epilepsy. The position of these cells in the circuitry of the dentate gyrus is ideal for gating the activation evoked by afferents from the entorhinal cortex. In the present study we have shown that the loss of hilar somatostatin-containing neurons, and the development of interictal spiking activity induced by sustained perforant pathway stimulation can be prevented by high doses (500 mg/kg), but not by low doses (100 mg/kg) of vigabatrin, an irreversible inhibitor of GABA-transaminase.

Non-pyramidal cells in the CA3 region of the rat hippocampus: relationships of fine structure, synaptic input and chemical characteristics

The non-pyramidal cells of the hippocampus are heterogeneous with respect to their morphology, peptide content, physiological properties, and postsynaptic targets. Here we demonstrate that the content of peptides (cholecystokinin, somatostatin) and calcium-binding proteins (parvalbumin and calbindin) of non-pyramidal cells is not related to a characteristic fine structure or synaptic input. Varying numbers of GABA-negative and GABA-positive input synapses of non-pyramidal cells indicate that these neurons are differently integrated in inhibitory and disinhibitory circuits.

Area-specific morphological and neurochemical maturation of non-pyramidal neurons in the rat hippocampus as revealed by parvalbumin immunocytochemistry

The time course of the morphological differentiation of non-pyramidal neurons in the rat hippocampus shows an area specificity. Thus, non-pyramidal neurons in CA3 appear more mature than in CA1 at early postnatal stages. Physiological data provide evidence for an earlier maturation of GABA-mediated inhibition in CA3 in comparison to CA1. As the calcium-binding protein parvalbumin (PARV) is thought to be a marker for highly active inhibitory neurons, we analyzed the area-specific appearance of PARV in GABAergic neurons during development. Employing combined light and electron microscopic immunocytochemistry, we revealed an area specificity in the time course of the neurochemical and morphological maturation of this functionally important subpopulation of non-pyramidal cells. The first appearance of PARV-immunoreactivity was observed at P7 and was exclusively located in cell bodies in CA3. At P8, neurons in CA3 exhibited PARV-immunoreactivity in cell bodies and dendrites, but very rarely in axon terminals. These neurons displayed the typical light and electron microscopic characteristics of GABAergic non-pyramidal cells. At P10, axon terminals formed typical baskets surrounding the pyramidal cells. The appearance of PARV-immunoreactivity in cell bodies, dendrites and axon terminals in CA1 was noticed about 1 to 2 days later. In the fascia dentata, non-granule cells displayed immunoreactivity not before P10. These data indicate a sequential neurochemical and morphological maturation of non-pyramidal neurons that may be related to differences in the maturation of inhibition during hippocampal development.

Late appearance of parvalbumin-immunoreactivity in the development of GABAergic neurons in the rat hippocampus

The calcium-binding protein parvalbumin (PARV) is supposed to have a protective function under conditions of experimental seizure and hypoxia in a subgroup of GABAergic inhibitory neurons in the adult rat hippocampus. Here we studied the appearance of PARV immunoreactivity in rat hippocampal non-pyramidal cells during postnatal development in comparison to glutamate decarboxylase (GAD) immunoreactivity. PARV-immunoreactive neurons were not observed before postnatal day 7 whereas GAD-positive neurons and terminal-like puncta were present at postnatal day 2 (P2) and were frequent around P5. From other studies it is known that all GABAergic neurons are formed prenatally. Our data thus indicate that in the early postnatal period GABAergic non-pyramidal cells are poorly protected by calcium-binding proteins against a pathological calcium influx.