An electron microscopic study of lesion-induced synaptogenesis in the dentate gyrus of the adult rat. I. Magnitude and time course of degeneration

Fine structure and synaptic connections of identified neurons in the rat fascia dentata

A survey is given of the synaptic connections of identified neurons in the rat fascia dentata based on our own Golgi/electron microscopic and light and electron microscopic immunocytochemical findings as well as on results obtained from the literature. The report largely deals with the dominating cell type in the region, the dentate granule cell. Of the various types of hilar cells, the GABAergic neurons, particularly the inhibitory basket cells, are taken into account. Differences in fine structure between granule cells and basket cells as well as mutual synaptic connections between these two types of dentate neurons are elaborated. This survey may provide a basis for further neurophysiological and pharmacological studies on these cells.

Synaptic reorganization in the medial amygdaloid nucleus after lesion of the accessory olfactory bulb of adult rat. II. New synapse formation in the medial amygdaloid nucleus by fibers from the bed nucleus of the stria terminalis

The rearrangement of terminations from the bed nucleus of stria terminalis (BST) was examined in the medial amygdaloid nucleus (MAN) at 2 months following the lesion of the accessory olfactory bulb (AOB) using an electron microscopy and degeneration study. At 2 days following a BST lesion, the number of degenerating synapses was 0.7 +/- 0.1 (mean +/- S.E.M.) per unit area (2500 microns2) in the molecular layer, and 3.0 +/- 0.3 in the cellular part. At 2 months after an AOB lesion, the degenerating synapses from the AOB had completely disappeared from the MAN. The BST was then lesioned at 2 months after the AOB lesion and, at 2 days following this BST lesion, the degenerating synapses were counted in MAN. The numbers observed were 3.3 +/- 0.6 per unit area in the molecular layer and 4.5 +/- 0.4 in the cellular part. Therefore, the number of these degenerating synapses increased significantly within the molecular layer, though, in the cellular part the number of synapses was not significantly elevated over control. No differences in postsynaptic profiles (ratio of synapses on dendritic spine to dendritic shaft) were observed after the AOB lesion. These results indicate that the BST fibers formed new synapses in the molecular layer following the denervation of AOB fibers. The possibility of new synapse formation by other afferent fibers in addition to the AOB fibers is discussed as is the relationship between lesion induced synaptic reorganization and functional recovery after injury.

Synaptic reorganization in the medial amygdaloid nucleus after lesion of the accessory olfactory bulb of adult rat. I. Quantitative and electron microscopic study of the recovery of synaptic density

The time course of the appearance of degenerating synapses and the loss and reappearance of intact synapses was investigated in the medial amygdaloid nucleus (MAN) following lesions of the accessory olfactory bulb (AOB). Degenerating synapses were concentrated in the molecular layer of MAN. A quantitative electron microscopic analysis of the molecular layer was performed at 2, 4, 8, 16, 32 and 64 days after the placement of kainic acid injections within AOB. At each survival time, counts of synapses were made from photomicrographic montages. The degenerating synapses were observed at 2 days after lesion. Their density reached a peak at 2 or 4 days after the lesion and decreased thereafter. No degenerating synapses were found at 64 days following the lesion. Shrinkage of the thickness of molecular layer was observed following the AOB lesion. This was most pronounced at 4 or 8 days after the lesion at which time the thickness was 70-75% of the contralateral molecular layer. The shrinkage receded moderately (85% of contralateral layer at 64 days after lesion). The compensated synaptic density of intact synapses was reduced to less than half of the control density at 4 or 8 days after the lesion. Thereafter synaptic density increased gradually and reached more than 80% of control density at 64 days after lesion. The recovery in number of intact synapses following the lesion suggests the possibility of synaptic reorganization by remaining afferent fibers.

An immunocytochemical and biochemical study of the microtubule-associated protein Tau during post-lesion afferent reorganization in the hippocampus of adult rats

A monoclonal antibody against the microtubule-associated protein (MAP) Tau was used to examine the fate of this molecule during post-lesion afferent reorganization in the hippocampus of adult rats. An immunocytochemical analysis was carried out in the dentate gyrus after unilateral destruction of the entorhinal cortex (EC). In the non-denervated hippocampus, Tau immunoreactivity was detected in parallel axons and mossy fibers; no staining was present in neuronal cell bodies and dendrites. A significant decrease in Tau immunoreactivity was detected in the outer 2/3 of the ipsilateral dentate gyrus molecular layer (ML) 2 days after an EC lesion, whereas staining in the inner 1/3 of the same ML increased considerably. This was followed by a very rapid recovery of Tau immunoreactivity in the outer 2/3 of the denervated ML, which by 10 days post-lesion was almost identical to that of the contralateral non-denervated ML. A similar phenomenon was observed in other regions of the hippocampus denervated by the EC lesion. The modifications in Tau immunoreactivity in the denervated hippocampus were also accompanied by changes in the polypeptide composition of this heterogeneous group of MAPs, as revealed by immunoblot analysis of hippocampal extracts obtained at different post-lesion intervals; these changes involved a rapid and significant increase in low molecular weight migrating Tau-immunoreactive polypeptides. The present observations indicate that important modifications in Tau proteins occur in the deafferented hippocampus, a phenomenon that may well be related with the regulation of microtubule polymerization during post-lesion axonal growth.

The response of the associational afferents to the dentate gyrus to simultaneous or sequential elimination of the commissural and entorhinal afferents

Previous studies have shown that following the removal of the commissural afferents to the dentate gyrus, the ipsilateral association afferents, which are normally distributed within the same region of the molecular layer, sprout new collateral branches and in time occupy essentially all the vacated synaptic sites. It is also known that when the entorhinal afferents to the dentate gyrus are interrupted the associational and commissural fibers can both undergo a similar phase of reactive synaptogenesis and give rise to new collaterals which extend for some distance into the denervated zone. Since the associational fibers can sprout after the removal of either the commissural or entorhinal afferents experiments were designed to determine their capacity for sprouting in newborn and young adult rats when both groups of afferents were eliminated either simultaneously or sequentially (i.e., after an interval of 8 weeks). The resulting changes in the terminal field of the associational afferents were assessed, two months after the last deafferentation, by measuring in autoradiographs the width of the zone occupied by the associational afferents labeled with [3H]proline, and by estimating the volume of this region in Timm-stained sections. The results indicate that under these conditions the associational afferents are capable of expanding their terminal field not only to occupy essentially all of the synaptic sites made available by the elimination of commissural fibers, but also to occupy a significant proportion of the space vacated by the removal of the entorhinal afferents. This suggests that the capacity of the associational afferents for reactive synaptogenesis is greater than that expressed after either commissural of entorhinal lesions alone.

Nerve connections between mouse and rat hippocampal brain tissue: ultrastructural observations after intracerebral xenografting

Fascia dentata tissue from embryonic mice was grafted to the hippocampal region of newborn Kyoto rats. After 1-7 months the recipients received lesions of the entorhinal cortex on the side of transplantation. Three days later their brains were processed for electron microscopy. The xenografts were identified by their content of mouse dentate granule cells which have smaller cell nuclei with more nucleoli than the corresponding host rat granule cells. Electron dense, degenerating host rat entorhinal fibers terminated in the outer parts of the mouse dentate molecular layer corresponding to the normal perforant path zones. They formed the normal type of synaptic contacts with dendritic spines. The findings demonstrate that precise synaptic contacts can be made across a species barrier.

Brain spectrin, calpain and long-term changes in synaptic efficacy

This chapter discusses the possibility that proteolytic digestion of cytoskeletal proteins, in particular spectrin, is part of the mechanisms through which physiological activity elicits structural and chemical changes in brain synapses. Recent work from several laboratories has produced a description of the initial events that trigger the long-term potentiation (LTP) of synaptic responses that appears in hippocampus after brief episodes of high frequency electrical stimulation. A likely sequence is as follows: suppression of IPSPs, prolongation of EPSPs, activation of N-methyl-D-aspartate (NMDA) receptors, influx of calcium into target cells. After briefly describing the evidence for this triggering sequence, the review takes up the question of what types of calcium sensitive chemistries are available to synaptic region that could produce functional changes lasting for weeks (i.e., for LTP). It is argued that the partial degradation of spectrin by a calcium-activated protease (calpain) provides a mechanism of this type. Spectrin is a substrate for calpain and both it and a breakdown product comparable to that produced by calpain are found in postsynaptic densities. Moreover, there is substantial evidence that spectrin regulates the surface chemistry and morphology of cells and thus its partial degradation would be expected to produce pronounced and persistent modifications in synapses. To reinforce this point, the review discusses recent findings suggesting that calpain mediated proteolysis of spectrin and other cytoskeletal proteins produces substantial changes in the shape of blood-borne cells and the distribution of their surface receptors.(ABSTRACT TRUNCATED AT 250 WORDS)

Time course of reactive synaptogenesis in the subcortical somatosensory system

These experiments were designed to determine when synaptogenesis begins in the adult rat ventral posterolateral nucleus of the thalamus following lesions of the dorsal column nuclei. Given the relatively uncomplicated structure of the neuropil in the ventral posterolateral nucleus of the rat, the specificity of reactive synaptogenesis of the lemniscal input and the effect of the loss of lemniscal terminals on terminals from other sources could be determined. By use of morphometric analysis of electron micrographs, the numerical density of the 3 terminal types in the neuropil was determined at a series of postlesion survival times ranging from 12 hours to 50 days. Synaptogenesis began about 30 days after the lesions of the dorsal column nuclei and was complete by 50 days. The slow onset of synaptogenesis was in response to a loss of the lemniscal terminals, which account for only 3% of the total number of synapses in the ventral posterolateral nucleus. The low level of synaptogenesis early in the recovery process differs from the recovery seen in other central nervous system sites, which show an early rapid increase in synapses in response to much greater denervation. The loss of lemniscal terminals has relatively little effect on the numerical density or distribution of the terminals of other types. The new terminals that are formed come both from axons that originate from the undamaged portion of the dorsal column nuclei and from axons originating in the spinal cord.

Lesion-induced transneuronal plasticity in the adult rat hippocampus

The process of reactive synaptogenesis has been demonstrated in several areas of the central nervous system, including the hippocampal dentate gyrus. After a complete unilateral entorhinal lesion, approximately 85% of the input to the outer two-thirds of the ipsilateral dentate molecular layer is lost. Bilateral fluctuations in synaptic density within non-denervated zones of the dentate molecular layer predict further alterations in neural circuitry at sites located transneuronally to the denervated dentate granule cells. Using quantitative electron microscopy, our study demonstrates a complete cycle of synapse loss and reacquisition within the ipsilateral but not contralateral CA4/hilus region of the hippocampal formation. This area is one of the terminal fields for the dentate granule cell mossy fiber axons. In addition the granule cell mossy fiber axons sprout during the postlesion time course and form a significantly increased number of new mossy fiber terminals within the ipsilateral and contralateral CA4/hilus area. Our results indicate that responses to brain injury may no longer be confined to a local denervated site, but probably include polyneuronal circuitry loops, which may encompass one or more areas of the central nervous system. Previous difficulties in providing a close behavioral or functional correlation to localized structural events may be explained by a more global brain response to an injury.

Thalamic control of dopaminergic functions in the caudate-putamen of the rat--III. The effects of lesions in the parafascicular-intralaminar nuclei on D2 dopamine receptors and high affinity dopamine uptake

Dopamine receptor binding in the caudate-putamen was studied following bilateral lesions of the thalamostriatal pathway. Receptor binding was assayed using [3H]spiperone and defined with both (+)-butaclamol and S(-)-sulpiride. Radiofrequency lesions resulted in an increase in the Bmax of [3H]spiperone binding defined with both (+)-butaclamol and S(-)-sulpiride between 7 and 14 days following surgery. At longer survival times a fluctuating response was seen in which a decrease in receptor binding was observed at 28 days following lesion and a further rise again at 70 days. At no time point was significant change in Kd recorded. Further experiments were carried out to control for the possible effects of damage to fibres of passage and for inadvertent damage to habenula, as well as to define the receptor subtype involved. Ibotenic acid lesions resulted in similar effects to those reported with the radiofrequency method. Thus, 7 days following lesion, Bmax for (+)-butaclamol-defined [3H]spiperone binding increased by approximately 14-20% over that recorded in sham-lesioned animals. Using S(-)-sulpiride to define binding, Bmax was found to increase 13-17% in the same membrane preparations. Similar results were obtained in experiments at 14 days following ibotenic acid induced lesions. Again, no change in Kd was recorded. When radiofrequency lesions were made, which were largely restricted to habenula and associated fibres of passage, only small [(+)-butaclamol defined] or insignificant [S(-)-sulpiride defined] changes in Bmax were recorded. Combined radiofrequency lesions of habenula and ibotenic acid lesions of the thalamus resulted in a larger increase in Bmax for (+)-butaclamol defined binding than with S(-)-sulpiride defined binding. Our interpretation of these findings, in the light of the histology of the lesions, is that the predominant effect of removing thalamic input to the caudate-putamen is an increase in the number of D2 receptors, but without any change of affinity. A small component of the change in Bmax defined with (+)-butaclamol found with radiofrequency lesions may be due to a response at non-dopamine sites (possibly a 5-hydroxytryptamine receptor subtype) following damage to other caudate-putamen afferents which pass near the habenula or fasciculus retroflexus. Following unilateral ibotenic acid lesions of the thalamus, the number of high affinity uptake sites for dopamine was increased at long survival times.(ABSTRACT TRUNCATED AT 400 WORDS)

Reactive synaptogenesis in hippocampal area CA1 of aged and young adult rats

Selective lesions that result in a partial loss of neuronal input appear to signal residual, undamaged inputs to sprout and replace synaptic connections that have been lost. Previous investigations have compared this process of reactive synaptogenesis in young and old animals in the hippocampal dentate gyrus and have demonstrated that the aged brain has a diminished capacity for reinnervation following massive denervation of a target area. This investigation has focused on the lesion-induced plasticity of an adjacent area of the hippocampal formation, area CA1 of regio superior, in young adult and aged rats. Young adult aged Fischer 344 rates were subjected to a unilateral, intraventricular injection of kainic acid that selectively destroyed the CA3-CA4 hippocampal pyramidal neurons. Following a 2-day interoperative interval, the rats sustained an ipsilateral transection of the fimbria-fornix. Animals were killed at 4, 10, 30, and 60 days following the second transection and processed for electron microscopic analysis. Photographic montages were constructed of area CA1 extending from the alveus to the hippocampal fissure. The density of synapses, both intact and degenerating, was determined and analyzed as a function of age, days postlesion, and zone of analysis. Synaptic density decreased 30-40% contralaterally and 60-70% ipsilaterally in both aged and young adult rats. While both age groups restored synaptic density to preoperative levels, aged subjects required significantly more time. Aged rats appeared to be retarded in the initial phases of synaptic replacement. The delay in the aged animals' reactive response was not due to any differences in degeneration clearance between the age groups.

Altered regulation of lesion-induced synaptogenesis by adrenalectomy and corticosterone in young adult rats

Quantitative electron microscopy was used to examine the effect of circulating glucocorticoids on the removal of degenerating synapses and the replacement of lost synaptic contacts in young adult rats that follow partial denervation of the hippocampal dentate gyrus. Subjects were adrenalectomized prior to subcutaneous implantation of pellets containing a specified concentration of corticosterone and subsequent unilateral ablation of the entorhinal cortex. Animals maintained at high circulating concentrations of glucocorticoids were significantly retarded in the early phase of degenerating synapse removal and in the rate of synaptic replacement. Subjects maintained at extremely low concentrations of glucocorticoids were also significantly retarded in the early stages of synapse removal but showed an early replacement of lost synaptic contacts followed by a dramatic decrease in the rate of replacement. By 60 days after the lesion both groups of animals showed synapse replacement equivalent to young adult controls while significant amounts of degenerating synapses still remained in the denervated neuropil. The results demonstrate that circulating glucocorticoids can exert a marked influence on lesion-induced synaptic replacement in the hippocampal dentate gyrus.

Induced homotypic sprouting of serotonergic fibers in hippocampus. II. An immunocytochemistry study

The dorsal hippocampus of the rat normally receives its 5-HT innervation from two homologous groups of cells in the median raphe nucleus via the cingulum bundle-induseum griseum (CB-IG) and the fornix-fimbria (FF) (J. Comp. Neurol., 179 (1978) 641-667 and Brain Res. Bull., 10 (1983) 445-451). 5-HT immunoreactive (IR) fibers are distributed in a laminar pattern in the hippocampus. These fibers have large varicosities and are densely distributed in the infragranular layer of dentate gyrus, in the stratum lacunosum-moleculare of the cornu Ammonis and in the area fasciola cinerea (FC). The present study provides evidence that the density of the 5-HT-IR fibers in the dorsal hippocampus is greatly decreased but maintained a similar laminar pattern 3 days after lesioning by microinjection of 4 micrograms of 5,7-dihydroxytryptamine (5,7-DHT) into the CB-IG. An apparently normal density and distribution pattern of the 5-HT-IR fiber is seen by 42 days postlesion. The FC in the hippocampus is among the first regions reinnervated by 5-HT-IR fibers with very dense and large varicosities. The restitution of 5-HT-IR fibers in the dorsal hippocampus after the 5,7-DHT lesion in the CB-IG is accompanied by a marked increase in the number and intensity of 5-HT-IR fibers in the FF. No evidence of a regeneration of 5-HT-IR fibers is seen distal to the injection site in the CB-IG. These observations provide direct evidence for homotypic collateral sprouting in the CNS induced by removal of a single fiber type.

The cholinergic innervation of the rat fascia dentata: identification of target structures on granule cells by combining choline acetyltransferase immunocytochemistry and Golgi impregnation

A monoclonal antibody against choline acetyltransferase (ChAT), the acetylcholine-synthesizing enzyme, was used to study cholinergic synapses on identified (Golgi stained) granule cells in the rat fascia dentata. Choline acetyltransferase immunocytochemistry was applied to 40-microns Vibratome sections cut perpendicular to the longitudinal axis of the hippocampus. Light microscopy revealed fine varicose ChAT-immunoreactive axons in all layers of the fascia dentata, i.e., in the stratum moleculare, the stratum granulosum, and the subgranular polymorph zone. Most fibers were observed in the vicinity of granule cell bodies where they ran mainly parallel to the granular layer. Next, the immunostained Vibratome sections were sandwiched between small pieces of Parafilm and piled to form a block that was covered with agar and Golgi stained. After that, the sections were separated by cutting away the agar and removing the Parafilm. Sections containing well-impregnated granule cells were gold-toned (Fairén et al., '77), embedded in Araldite, and subjected to ultrathin sectioning for electron microscopy. A total of 14 gold-toned granule cells were examined in the electron microscope for synaptic contacts with cholinergic afferents. Choline acetyltransferase-immunoreactive axon terminals were observed that established symmetric synaptic contacts with the cell bodies and dendritic shafts of the gold-toned identified granule cells. Two types of contact were observed on spines arising from gold-toned granule cell dendrites. Immunoreactive terminals established asymmetric synaptic contacts with the head of small spines and symmetric contacts with the stalk of large, complex spines. The boutons forming asymmetric synaptic contacts with the cup-shaped spine head of the complex spines were not found to be immunoreactive. Our results demonstrate that cholinergic fibers to the rat fascia dentata establish characteristic types of synaptic contact with different postsynaptic elements of granule cells, suggesting a complex function of this afferent system.

Compensation in the number of presynaptic dense projections and synaptic vesicles in remaining parallel fibres following cerebellar lesions

Our previous investigations demonstrated an increase in the size of remaining synaptic sites as an intermediate or possible alternative to sprouting plasticity. The total amount of postsynaptic contact area remained relatively constant for each target neuron even though there was a marked decrease in the number of sites on these neurons. In addition, enlarged boutons containing numerous synaptic vesicles were positioned adjacent to enlarged postsynaptic sites. The question posed by this study was to determine whether dense projections, parts of the presynaptic grids of the remaining parallel fibres, spread to cover the enlarged postsynaptic sites, or if the number of these densities increased on each site to maintain the structural organization of the presynaptic grid. In addition, the number of synaptic vesicles per bouton was quantitated to determine whether they compensated by increasing their number in relationship to the increased area of the presynaptic grid. The number of parallel fibre synapses on Purkinje cells was reduced by transection of a narrow bundle of parallel fibres accompanied by a small lesion undercutting the molecular layer to destroy granule cells contributing to this bundle. The number of presynaptic dense projections was quantitated in control and lesioned preparations (using ethanolic acid staining) in order to determine their correlation to the area of each site. In addition, the average number of synaptic vesicles in boutons was compared to the average size of boutons and the average contact area of the synaptic sites. At 3 to 7 days following partial deafferentation of Purkinje cells in adult rats, the density of dense projections of parallel fibre synapses on Purkinje cell spines remained uniform. This occurred throughout a range of reduction in the number of synapses in conjunction with a reciprocal increase in the size of sites. The finding of a uniform density of these projections and an increase in the size implies that each granule cell axon must gain dense projections. In addition, the remaining presynaptic boutons had a uniform density of synaptic vesicles even though the volume of the boutons and the area of the synaptic contact doubled. Thus, the number of synaptic vesicles gained in proportion to the total enlargement of the contact site and the bouton size. These results strongly suggest that deficits or losses in synaptic connections of parallel fibre on Purkinje cell spines produces a compensation in the total number of synaptic vesicles and presynaptic dense projections of the remaining boutons.(ABSTRACT TRUNCATED AT 400 WORDS)

Mossy fiber sprouting in the fascia dentata after unilateral entorhinal lesions: quantitative analysis using computer-assisted image processing

Axon sprouting typically occurs in a brain region that has been partially denervated. The present study demonstrates, quantitatively, evidence for sprouting outside the region of deafferentation. A modification of the Timm sulfide silver histochemical method was used to monitor an increase in the mossy fiber terminal field in the fascia dentata of adult rats following severe deafferentation of the outer three-fourths of stratum moleculare by unilateral entorhinal lesions. Computer-assisted image processing techniques were used to quantify mossy fiber sprouting. In stratum granulosum, and to a lesser extent in the deep (supragranular) portion of stratum moleculare (areas separated from the zone of deafferentation), there was a three-fold increase in the area of mossy fiber staining on the side of the lesion compared to the non-operated side (and unoperated animals). Much of the increased staining was located near the tip of the infrapyramidal (ventral) blade of the fascia dentata. Since mossy fiber sprouting apparently occurs in the absence of degeneration-produced synaptic dilution in that region, it may represent an example of post-lesion growth initiated by conditions fundamentally different from those normally believed to induce sprouting.

Cell biology of synaptic plasticity

The nervous system of mammals retains throughout the animals' life-span the ability to modify the number, nature, and level of activity of its synapses. Synaptic plasticity is most evident after injury to the nervous system, and the cellular and molecular mechanisms that make it possible are beginning to be understood. Transplantation of brain tissue provides a powerful approach for studying mechanisms of synaptic plasticity. In turn, understanding the response of the central nervous system to injury can be used to optimize transplant survival and integration with the host brain.

Cellular and connective organization of slice cultures of the rat hippocampus and fascia dentata

This study examined the cellular and connective organization of hippocampal tissue taken from 6-8-day-old rats and cultured by the roller tube technique for 3-6 weeks. In the cultures containing the fascia dentata and the hippocampus proper (CA1, CA3, CA4) the main cell and neuropil layers were organotypically organized when observed in ordinary cell stains. The normal distribution of smaller cell populations of AChE-positive neurons and somatostatin-reactive neurons was demonstrated by histochemical and immunohistochemical methods. Both cell types were mainly confined to str. oriens of CA3 and CA1 and the dentate hilus (CA4). Individual dentate granule cells and hippocampal pyramidal cells were injected with lucifer yellow and HRP, revealing great stability of the dendritic patterns of these cells in the culture condition. The same was found for the axonal branching and termination of HRP-filled mossy fibers arising from an HRP-injected granule cell. The preservation of organotypic afferent patterns in the cultures was also shown by Timm staining of the terminal distribution of the mossy fiber system. Mossy fiber terminals, with characteristic ultrastructural features verified in the electron microscope, were thus found in the hilus (CA4) and along the CA3 pyramidal cell layer onto the CA3-CA1 transition. Depending on the amount of dentate tissue relative to CA3 the terminals could stop before reaching CA1 (small fascia dentata) or take up additional intra and infrapyramidal locations along CA3 (small CA3). In cultures with a gap in the CA3 pyramidal cell layer some mossy fiber terminals were found in contact with the CA3 pyramidal cells beyond the gap. In all cultures there was an aberrant projection of supragranular mossy fibers. This projection is analogous to the one known from lesion and transplant studies to form in the absence of the entorhinal perforant path input to the dentate molecular layer. Also, in accordance with these studies the Timm staining pattern of the outer parts of the dentate molecular layer and the entire molecular layer of the hippocampus was altered corresponding to the spread of afferents normally confined to the inner zone of the dentate and str. radiatum of CA3 and CA1. Possibly as a consequence of the lack of normal targets for projections from CA1, this subfield contained an unusually dense Timm staining suggestive of autoinnervation.(ABSTRACT TRUNCATED AT 400 WORDS)

Regulation of axonal ingrowth into area dentata as studied by sequential, double intraocular brain tissue transplantation

The anterior eye chamber was used as a model environment to study, in isolation, the interaction of embryonic area dentata transplants with transplants of one of three important sources of in situ innervation: entorhinal cortex, locus coeruleus or septal nuclei. None of these brain regions significantly affected the morphogenesis or in oculo growth of area dentata transplants. All three brain regions innervated the area dentata transplant. Entorhinal cortical transplants sent nerve fibers into a limited, and apparently specific, region of area dentata that was adjacent to the entorhinal transplant. This light innervation contrasts to the predominant innervation of area dentata by entorhinal cortex in situ. The fluorescent, noradrenergic neurons of locus coeruleus provided the area dentata transplant with an abundance of fine varicose nerve fibers. Given about 100 noradrenergic neurons in the locus coeruleus transplant and 4 to 6 months joint survival, the area dentata transplant was noradrenergically hyperinnervated. The cholinergic neurons of the septal nuclei transplant had a prolific ingrowth of acetylcholinesterase (AChE)-positive nerve fibers to the area dentata transplant. There appeared to be a mutual exclusion between the extrinsic AChE-positive fibers and the intrinsic Timm's-positive granule cell mossy fibers in the area dentata transplant. We conclude that isolated replicas of the coeruleo-, septo-, and entorhinal cortico-dentate pathways can be made through sequential intraocular double grafting. The nature of the in oculo connectivity between these replicates offers clues as to the mechanisms that might account for the regulation of nerve growth.

Reciprocal relationship between size of postsynaptic densities and their number: constancy in contact area

Plasticity in the size of postsynaptic membrane specializations (postsynaptic densities) was analyzed by quantitation following lesioning of parallel fiber afferent axons to Purkinje cells in the cerebellum. Double sectioning of parallel fibers in the same folium as well as longitudinal undercutting of the molecular layer to destroy granule cells and their parallel fibers were used to produce various levels of afferent reduction to Purkinje cells. The length of profiles of membrane densities was measured utilizing semi-automated, computer-electron microscopy and the number of synapses was determined from their volume density and changes in volume estimated from cortical thickness. Correlation between the number of synapses on Purkinje cells and their average contacts area revealed a reciprocal relationship throughout a range of 0-67% reductions in parallel fiber synapses. Larger reduction levels had a progressive decrease in average size of contacted postsynaptic densities and an accompanying increase in the number of vacant postsynaptic specializations. Total absence of parallel fibers resulted in nearly all vacant sites on spines (except for a few connections with boutons having irregular shaped vesicles). These vacant sites were, on the average, only half the size of controls but their number was approximately double the control amount. This study confirms that location and size of synapses are not permanent and that plasticity in size of contacts allows reorganization in circuitry to compensate alterations in the number of inputs following a number of perturbations. The finding of a reciprocal relationship indicates that total contact area on Purkinje cells remained relatively constant throughout the entire range of reductions in the number of afferents. A 'constancy principle for total postsynaptic contact area' is envisioned to stabilize functional aspects of circuitry during developmental organization and to direct compensation following reduction in pre- or postsynaptic neurons from environmental effects or attrition in aging. Constancy in target area provides a realm under which the size of individual synapses can be modified as functional adaptations in circuitry even without changes in the number of connections.

Synaptic density and axonal sprouting in rat hippocampus: stability in adulthood and decline in late adulthood

Synaptic density and axonal sprouting remain stable in the rat dentate gyrus up to 12 months of age. At 18-24 months, an 8-10% loss in synaptic density occurs, and synaptic reinnervation could not be clearly detected at 15 days postlesion, compared to a 40% synaptic recovery in rats lesioned at 6 or 12 months postnatal. Since the amount of synaptic degeneration remaining in the denervated zone was the same at 15 days postlesion in all rats examined, synaptic reinnervation is not primarily dependent on the rate at which degeneration is removed. The results indicate that the decline in the growth response in the rat hippocampus is not a progressive trend from early adulthood into old age, but appears after the midpoint of life.

Delayed neuronal death in the rat hippocampus following transient forebrain ischemia

An unusual, slowly progressing neuronal damage has been reported to occur in the gerbil hippocampus following ischemia (Kirino 1982). Delayed neuronal death following ischemia has also been noticed in the rat four-vessel occlusion model (Pulsinelli et al. 1982). By light microscopy this slow neuronal injury in the rat was not different from the previously known neuronal ischemic cell change. This report lead us to the question as to whether neurons in the rat hippocampus are damaged rapidly following an initial latent period or deteriorate slowly and progressively until they display overt changes. To clarify this point, observation was done on the hippocampal CA1 sector of the rat following ischemia. Rats were subjected to four-vessel occlusion, and those which developed ischemic symptoms were perfusion-fixed. Although the change appeared very slowly and lacked microvacuolation of the cytoplasm, neuronal alteration was practically not different from classical ischemic cell change. By electron microscopy, however, the change was detectable when the neurons still appeared intact by light microscopy. An increase in the membranous organelles and deposition of dark substances were the initial manifestations. It seemed that the CA1 neurons deteriorated very slowly and progressively, and that they retained partial viability in the initial phase of the change. In spite of the difference in light-microscopic findings, the mechanisms underlying delayed neuronal death in the rat and gerbil hippocampus seemed to be identical.

Uptake of D-aspartate and L-glutamate in excitatory axon terminals in hippocampus: autoradiographic and biochemical comparison with gamma-aminobutyrate and other amino acids in normal rats and in rats with lesions

High affinity uptake sites for 3H-labelled amino acids were studied in synaptosome-containing homogenates processed biochemically or in surface autoradiograms of incubated slices of hippocampus. D-aspartate and L-glutamate had apparently identical distributions. In normal rat hippocampus the highest uptake was in the terminal fields of axons from the pyramidal cells of regio inferior and hilus fasciae dentatae, while there was a moderate uptake in the terminal fields of the medial and lateral perforant paths, slight uptake in the mossy fibre layer and no uptake in the terminal fields of the basket cells. Uptake sites for gamma-aminobutyrate were concentrated in the latter fields, and in the most superficial cortical layers. The present method shows no uptake in cell bodies. The uptake activities were strongly inhibited by recognized blockers of (neuronal) high affinity uptake of glutamate or gamma-aminobutyrate. Autoradiographically, several other amino acids showed negligible uptake. The uptake of D-aspartate was reduced by 80% in stratum oriens and stratum radiatum of regio superior 4-14 days (70% at 3 days) after transection of the afferent pyramidal cell axons from the ipsi-and contralateral regio inferior. The reduction was in the number of uptake sites, not in their affinity. Uptake of gamma-aminobutyrate was not reduced. Lesions affecting regio superior caused a loss of D-aspartate uptake in subiculum at a site known to receive hippocampal afferents. Autoradiographically, the uptake of D-aspartate was strongly reduced in the inner zone (i.e. the target zone), but increased in the middle zone of the dentate molecular layer after lesions of the hilus fasciae dentatae. At 4 days and longer after transection of the entorhinal afferents, there was a conspicuous reduction of D-aspartate and L-glutamate uptake in the target zones of both the medial and lateral contingent of these fibres. In the same animals, the terminal zone of afferents from hilus fasciae dentatae had an increased radioactivity and was slightly wider than normally. Concomitantly, the gamma-aminobutyrate uptake was increased in the target zones of the degenerating perforant path fibres. The results demonstrate that uptake sites for D-aspartate and L-glutamate are highly localized in axon terminals of regio inferior pyramidal cells and in perforant path afferents. The latter category of terminals has a lower density of acidic amino acid uptake sites than the former. Uptake sites for gamma-aminobutyrate are localized in terminals of intrinsic neurones, including the axosomatic terminals of basket cells.(ABSTRACT TRUNCATED AT 400 WORDS)

Glial contribution to amino acid content and metabolism of the deafferented dentate gyrus

The time course of tissue content and evoked release of endogenous amino acids was analyzed in the partially deafferented dentate gyrus of the rat hippocampus 2-24 days following unilateral lesion of the perforant path. Amino acids in tissue extracts and perfusates were determined after precolumn derivatization and hplc separation. The astrocytic glial cell reaction was monitored with immunohistochemistry of S-100. The tissue content of glutamate decreased significantly on the lesioned side, whereas only a moderate reduction in taurine, aspartate, and alanine occurred. Glutamine was significantly elevated at 7 days. The evoked efflux of glutamate was reduced at 2 and 7 days, whereas no change was seen at longer survival periods. The evoked release of GABA and aspartate increased on the denervated side after 12 and 24 days. The rate of carbon utilization into amino acid pools was followed with 14C-glucose and 14C-acetate. The incorporation of acetate showed a peak 2-9 days following lesion, which paralleled in time the hypertrophic glial cells. The incorporation of glucose decreased during this period. The metabolic events are discussed in relation to the morphological changes in synapses and glial cells.

Kindled seizure-induced reduction of muscarinic cholinergic receptors in rat hippocampal formation: evidence for localization to dentate granule cells

The binding of [3H] quinuclidinyl benzilate ( [3H] QNB) to muscarinic cholinergic receptors in dentate gyrus of rat hippocampal formation was analyzed by membrane binding assay and in vitro autoradiography. The destruction of dentate granule cells, either by neonatal irradiation or colchicine injection, resulted in nearly complete elimination of [3H] QNB binding sites in the molecular and granule cell layers. By contrast, neither perforant path transection nor destruction of the septal-hippocampal cholinergic afferents caused a decline of [3H] QNB binding sites. Amygdala kindled seizures resulted in a 30% reduction of [3H] QNB binding sites which was distributed uniformly across the entire molecular and granule cell layers. Thus, most, if not all, of the muscarinic cholinergic receptors present in dentate gyrus appear to reside on the somata and dendritic trees of the dentate granule cells. We propose that this kindled seizure-induced decline of muscarinic receptors represents an endogenous compensatory mechanism designed to stabilize granule cell excitability.

Evidence for differential localization of two binding sites for L-[3H]glutamate in rat fascia dentata

Two binding sites for L-[3H]glutamate were tentatively localized in rat fascia dentata by determining the effects of selective lesions on specific binding. Both destruction of dentate granule cells with colchicine and ablation of the ipsilateral entorhinal cortex markedly reduced radioligand binding to a quisqualate-sensitive site (GLU A), but only the entorhinal lesion significantly reduced binding to a site that is less sensitive to quisqualate (GLU B). These results suggest that GLU A binding sites are localized mainly on the dentate granule cells, whereas GLU B binding sites are localized, in part, on the perforant path fibers, but not on granule cells.

Entorhinal lesions result in shrinkage of the outer molecular layer of rat dentate gyrus leading subsequently to an apparent increase of glutamate decarboxylase and cytochrome oxidase activities

In intact dentate gyrus, glutamate decarboxylase immunoreactivity (GAD) and cytochrome oxidase activity (CyO) showed different distributions patterns. Entorhinal lesions caused increases of GAD and CyO in the outer molecular layer (OML) of the ipsilateral side. Submicroscopical localization of these enzymes did not change, except for CyO labeling more astrocytic mitochondria. The increase in numerical density of GAD puncta correlated quantitatively with shrinkage of OML, whereas in the whole molecular layer the number of GAD puncta remained unchanged. Hence, the localized increase of enzyme activities and lysosomes is apparently related to shrinkage of OML, but does not indicate plasticity of GABAergic neurons.

Enhanced acetylcholinesterase staining in the hippocampal perforant pathway zone after combined lesions of the septum and entorhinal cortex

A lesion of the septum or a transection of the fimbria-fornix diminishes most, but not all, acetylcholinesterase (AChE) staining in the hippocampal formation. The residual AChE is located in the outer part of the molecular layer of the hippocampal CA1 zone and adjacent subicular field (zone 31). We report that following combined lesions of the septum and entorhinal cortex, the residual hippocampal AChE staining pattern expands and occupies the zone innervated normally by perforant pathway terminals from the entorhinal cortex.

Rate of synaptic replacement in denervated rat hippocampus declines precipitously from the juvenile period to adulthood

Synaptic contacts per unit area in the rat dentate gyrus reach adult numbers by the end of the first month after birth and remain constant thereafter. This experiment demonstrated that the rate at which synapses were replaced by sprouting after a lesion declined dramatically from 35 to 90 days of age. Thus, the juvenile period of the rat's life is marked by a considerable change in neuronal plasticity. This may be related to age-dependent effects in recovery from brain damage.

Relative slowness of heterotypic synaptogenesis in the septal nuclei

The dorsolateral quadrant of the lateral septal nucleus receives a bilateral projection from the fimbria. When the fimbria of one side is cut, the axons of the remaining fimbria take over its synaptic sites preferentially, but when both fimbrias are cut the sites are reinnervated by non-fimbrial axons. To explore the basis of this preference, the present study plots the time courses of the appearance and disappearance of degenerating synapses, and the loss and recovery of non-degenerating synapses after ipsi-, contra- and bi-lateral fimbrial lesions. A preliminary investigation showed that at any time after these three lesions there was no change in the numerical density per unit area of 'control' structures such as shaft synapses (which do not degenerate) and neuronal perikarya (which neither shrink nor degenerate). This indicates that the changes in the numerical density of fimbrial (spine) synapses can be used as a measure of the processes of deafferentation and reinnervation without the danger of the numerical data being distorted by shrinkage. In the sampled area, the ipsilateral fimbrial axons account for about 45% of the synapses and the contralateral fimbrial axons for 25%. The number of degenerating synapses appearing at any one time underestimates the loss of non-degenerating synapses by about one-third, and a photographic simulation of degeneration suggests that a major factor in this discrepancy is the difficulty in recognizing degenerating synapses. Our main finding is that there is a major delay in the rate of removal of degeneration, and in the rate of reinnervation, after bilateral as opposed to unilateral lesions. This delay cannot be accounted for in any simple way by the greater amounts of degeneration. Thus after unilateral lesions, which cause the turnover of 25% (contralateral) or 45% (ipsilateral) of the synapses, 50% of the degeneration is removed in 1-2 days after the peak, whereas after bilateral lesions, which affect 70% of the synapses, it takes 20 days for 50% of the degeneration to be removed. That the synaptic changes after bilateral lesions involve a qualitatively different mechanism is also suggested by the observations of a much greater proportional increase in the multiple synapse index, and a decreased astroglial response.

Increases in protein-precursor incorporation in the denervated neuropil of the dentate gyrus during reinnervation

Cellular metabolic events accompanying postlesion synaptogenesis in the hippocampus were studied by analyzing incorporation of protein precursor ([3H]leucine) in the dentate gyrus. Adult rats were injected intravenously with [3H]leucine at periods from 2 to 60 days following unilateral destruction of the entorhinal cortex, and were killed 30 min later. Precursor incorporation was quantified autoradiographically by counting silver grains over the cell bodies and dendrites of dentate granule cells ipsi- and contralateral to the lesion. The relative grain density was increased over the denervated portion of the neuropil at 6-12 days postlesion, corresponding to the early phase of terminal proliferation and reactive synaptogenesis. Whereas incorporation was increased over the denervated neuropil, the availability of [3H]leucine was decreased relative to the contralateral side in autoradiographic preparations designed to reveal the concentration of the unincorporated 3H-labeled precursor and its diffusible degradation products. Silver grains were not selectively associated with glial cells bodies or vascular elements, but rather were distributed diffusely throughout the neuropil. Increases in grain density over the denervated zone were observed when animals were killed 8 min after the leucine injection, suggesting that the increases were not due solely to rapid transport from granule cell bodies to dendrites. We propose that an increased incorporation of protein precursor occurs primarily within the denervated dendrites of granule cells during the early phase of reinnervation, and that protein synthetic activity in these cells might be involved in the process of reinnervation.

Reinnervation of the partially deafferented hippocampus by compensatory collateral sprouting from spared cholinergic and noradrenergic afferents

The cholinergic and adrenergic afferents innervating the hippocampal formation in the rat reach the target region via three distinctly separate routes, two dorsal and one ventral one. Partial deafferentation of the hippocampus obtained by destruction of the dorsal routes (through the fimbria-fornix and the supracallosal striae) resulted in removal of 90% and 60% of the cholinergic and adrenergic innervations, respectively, within one month. By 6-10 months after lesion, the remaining cholinergic and adrenergic inputs, reaching the target via the ventral route, had expanded more than two-fold, resulting in a significant recovery in the original cholinergic and adrenergic innervation patterns. Because of its slow and protracted time-course and its ability to re-establish innervation also in initially denervated areas, this compensatory collateral sprouting phenomenon may be of particular interest for the understanding of the long-term, protracted functional recovery that is seen both after experimental brain lesions as well as in patients with severe brain injuries.

Lesion-induced mossy fibers to the molecular layer of the rat fascia dentata: identification of postsynaptic granule cells by the Golgi-EM technique

The axons of the dentate granule cells, the hippocampal mossy fibers, sprout "backward" into the dentate molecular layer when this is heavily denervated. Using the combined Golgi-electron microscopy (EM) technique we now demonstrate that these aberrant supragranular mossy fibers at least in part terminate on granule cell dendrites. Sprouting of mossy fibers into the dentate molecular layer was induced in adult rats by simultaneous surgical removal of the commissural and entorhinal afferents to the fascia dentata. After at least 7 weeks survival, the presence of mossy fiber terminals in the inner part of the dentate molecular layer was demonstrated by light microscopy. In the electron microscope the mossy fiber terminals were identified by their unique structural characteristics, namely, the unusually large size of the terminals, the dense packing of clear synaptic vesicles with a few dense core vesicles intermingled, the presence of asymmetric synaptic contacts with spines and desmosome-like contacts with dendritic shafts, and the continuity with a thin unmyelinated preterminal axon. Golgi-stained granule cells were first identified in the light microscope, and then, after deimpregnation, the same cells were examined in the electron microscope. In ultrathin, serial sections lesion-induced mossy fiber terminals were found in synaptic contact with spines on proximal dendritic segments of such identified Golgi-impregnated granule cells. From this we conclude that the aberrant, supragranular mossy fibers can innervate dendrites of the parent cell group, the dentate granule cells. The results, moreover, provide an example of reactive synaptogenesis where both the sprouted afferents and its postsynaptic element have been identified.

Lysosomal enzyme changes in young and aged control and entorhinal-lesioned rats

In the hippocampus of young and aged rats, lysosomal acid phosphatase and beta-glucuronidase were localized by enzyme histochemistry to the pyramidal and granule cells and to the mossy fiber zone (acid phosphatase only). Following a unilateral entorhinal lesion, enhancement in enzyme staining was observed within GFA (glial fibrillary acidic) protein-positive astrocytes and in cells resembling microglia and oligodendroglia in the ipsilateral dentate outer molecular layer and the hippocampal stratum lacunosum-moleculare. In young animals, these changes, first detected on day 2 post-lesion, were maximum on days 4 and 10 post-lesion and declined subsequently. The aged animals exhibited a delayed time-course of enzyme changes with minimal increases on days 2 and 4 post-lesion and greater increases on days 10 and 30 and gradual diminution thereafter. The delayed lysosomal response in the aged animal paralleled a reduction in the induction of glial cells following the lesion. Our findings suggest that the previously reported impairment in the clearing of degeneration debris in the aged animal is mediated by reduced glial induction and delayed glial lysosomal activation.

Neuronal response of the hippocampal formation to injury: blood flow, glucose metabolism, and protein synthesis

The reaction of the hippocampal formation to entorhinal lesions was studied from the viewpoints of cerebral blood flow ([123I]isopropyl-iodoamphetamine[IMP])-glucose utilization ([14C]2-deoxyglucose), and protein synthesis ([14C]leucine), using single- and double-label autoradiography. Our study showed (i) decreased glucose utilization in the inner part, and increased glucose utilization in the outer part of the molecular layer of the dentate gyrus, starting 3 days after the lesion; (ii) increased uptake of [123I]IMP around the lesion from 1 to 3 days postlesion; and (iii) starting 3 days after the lesion, marked decrease in [14C]leucine incorporation into proteins and cell loss in the dorsal CA1 and dorsal subiculum in about one-half of the rats. These changes were present only in animals with lesions which invaded the ventral hippocampal formation in which axons of CA1 cells travel. By contrast, transsection of the 3rd and 4th cranial nerves resulted, 3 to 9 days after injury, in a striking increase in protein synthesis in the oculomotor and trochlear nuclei. These results raise the possibility that in some neurons the failure of central regeneration may result from the cell's inability to increase its rate of protein synthesis in response to axonal injury.

Turnover of brain postsynaptic densities after selective deafferentation: detection by means of an antibody to antigen PSD-95

Antibodies to antigen PSD-95, a neuronal protein present only in postsynaptic densities (PSDs), have been used to follow immunohistochemically the loss and recovery of PSDs after selective deafferentation of both the hippocampal formation and the lateral septal nucleus of the rat. Three days after unilateral entorhinal ablation, densitometry of brain tissue sections stained by the peroxidase-antiperoxidase (PAP) method showed a decrease of about 25% in the outer 2/3 of the ipsilateral dentate gyrus molecular layer (ML), whereas staining in the inner 1/3 of the same ML layer actually increased about 16%. The intensely staining inner 1/3 of the deafferented ML expanded over time so that by 30 days postlesion the expansion had reached 60-70% of the ML. In the lateral septal nucleus, unilateral fimbria transection did not change either the pattern of anti-PSD PAP staining or that of [125I]protein A binding as revealed by autoradiography or by microdissection of the lateral septal nuclei to determine bound radioactivity by gamma-counting. Bilateral intraventricular injection of kainic acid to destroy area CA3 of the ipsilateral hippocampus caused very little loss of anti-PSD stain in the dentate gyrus ML, but decreased by 45% and 34% the intensity of stain in stratum radiatum and stratum oriens of the hippocampus, respectively. However, bilateral injections of doses of kainate high enough to destroy areas CA3, CA4 and part of CA1, caused 53% loss of stain in the inner 1/3 of the dentate gyrus ML. Recovery from the PSD loss caused by kainate in area CA1 was slow, only 67-81% of control by 90 days post-lesion. The immunohistochemical results in dentate gyrus and septum corresponded closely with quantitative data obtained by electron microscopy. Therefore, the response of PSDs to the loss of their presynaptic counterpart appears to depend on the overall extent of deafferentation of the neuron, the zone of the dendritic tree where deafferentation occurs and, perhaps, other specific features of the target cells.

Synaptic remodelling and astrocytic hypertrophy in rat cerebral cortex from early to late adulthood

Ultrastructural observation of the molecular layer of the parietal cortex of rats, aged 3, 6, 10 and 17 months, revealed various atypical synaptic profiles besides typical synapses. The atypical synapses were frequently in the vicinity of hypertrophied astrocytic profiles, and were sometimes completely surrounded by astrocytic processes. The presynaptic terminal contained either no vesicles or a few small distorted vesicles. Vacant postsynaptic terminals were occasionally seen. The total surface area of astrocytic profiles and the numbers of atypical synapses increased significantly between 3 and 10 months. The astrocytic acquisition of degenerating terminals was repeatedly observed over this period. Since there was no decrease in total synaptic number at this age, the astrocytic phenomenon may represent a stage in a continuous cycle of synaptic loss and replacement in the normal brain. By 17 months, when total synapse numbers decrease, synaptic replacement may be less than optimal.

Perforated postsynaptic densities: probable intermediates in synapse turnover

The molecular layer of the dentate gyrus of normal rats shows a large incidence of perforated postsynaptic densities (PSDs). The perforations or discontinuities occur almost exclusively in PSDs located in spines showing a U- or W-shaped junctional profile (complex PSDs). Perforated PSDs account for 16-25% of the total complex PSD profiles in young adult rats and 12-29% of those in aged animals. The frequency of perforations in the inner molecular layer of the dentate gyrus undergoes significant changes during a cycle of nondegenerative synapse turnover induced by ipsilateral ablation of the entorhinal cortex. During the first 2 days postlesion nonperforated PSDs (simple PSDs) decrease sharply, whereas perforated PSDs change little. However, at later times (4-10 days) there is a significant increase in the number of perforated PSDs that balances the number of simple PSDs lost. Beyond 10 days postlesion the proportion of both types of PSD is restored slowly to normal--i.e., nonperforated PSDs increase in number and perforated PSDs decrease, returning to the values in unoperated animals by 120 days postlesion. This inverse relationship between small nonperforated PSDs and large perforated PSDs suggests a precursor-product relationship between them. We propose that perforated PSDs are intermediates in an ongoing cycle of synapse turnover that is a part of the normal maintenance and adaptation of the nervous system.