Aging, spatial learning, and total synapse number in the rat CA1 stratum radiatum.

The aim of this study was to determine whether spatial learning deficits in aged rats are associated with a loss of hippocampal synapses. The Morris water maze task was used to assess the spatial learning capacity of young and aged rats and to attribute aged animals to learning-impaired and learning-unimpaired groups. The number of axospinous synapses in the entire volume of the CA1 stratum radiatum was estimated with unbiased stereological techniques. The results show that the total number of all axospinous synapses and of their perforated and nonperforated subtypes remains constant in the CA1 stratum radiatum of aged learning-impaired rats as compared to aged learning-unimpaired rats and to young adults. Thus, neither age-related deficits in spatial learning nor advanced chronological age are associated with a loss of axospinous synapses from the rat CA1 stratum radiatum.

Synapses with a segmented, completely partitioned postsynaptic density express more AMPA receptors than other axospinous synaptic junctions.

Axospinous perforated synapses of one morphological subtype exhibit multiple transmission zones, each one being formed by an axon terminal protrusion apposing a postsynaptic density (PSD) segment and separated from others by complete spine partitions. Such segmented, completely partitioned (SCP) synapses have been implicated in synaptic plasticity and postulated to be exceptionally efficacious. The present study explored the validity of this supposition. Postembedding immunogold electron microscopy was used for quantifying the postsynaptic AMPA receptor (AMPAR) expression, which is widely regarded as a major determinant of synaptic efficacy. Various subtypes of axospinous synapses were examined in the rat CA1 stratum radiatum. The results showed that the number of immunogold particles for AMPARs in SCP synapses markedly and significantly exceeded that in other perforated subtypes (by 101% on the average) and in nonperforated immunopositive synapses (by 1086%). Moreover, the particle number per single PSD segment, each of which also contained NMDA receptors, was significantly higher than that per nonperforated PSD (by 485%). SCP synapses also exhibited a higher particle density per unit PSD area, as well as a larger overall PSD area as compared with other synaptic subtypes. Analysis of covariance revealed that the high AMPAR expression in SCP synapses was related to the segmented PSD configuration, not only to the PSD size. Moreover, the subpopulations of SCP and other perforated synapses with either overlapping or equal PSD sizes differed in AMPAR content and concentration, with both measures being significantly higher in SCP synapses. Thus, the elevated AMPAR expression in SCP synapses is associated with the presence of separate PSD segments, not only with their large PSD area. These findings are consistent with the idea that SCP synapses have a relatively greater efficacy and may support maximal levels of synaptic enhancement characteristic of certain forms of synaptic plasticity such as the early LTP phase.

Associative learning elicits the formation of multiple-synapse boutons.

The formation of new synapses has been suggested to underlie learning and memory. However, previous work from this laboratory has demonstrated that hippocampus-dependent associative learning does not induce a net gain in the total number of hippocampal synapses and, hence, a net synaptogenesis. The aim of the present work was to determine whether associative learning involves a specific synaptogenesis confined to the formation of multiple-synapse boutons (MSBs) that synapse with more than one dendritic spine. We used the behavioral paradigm of trace eyeblink conditioning, which is a hippocampus-dependent form of associative learning. Conditioned rabbits were given daily 80-trial sessions to a criterion of 80% conditioned responses in a session. During each trial, the conditioned stimulus (tone) and the unconditioned stimulus (corneal airpuff) were presented with an intervening trace interval of 500 msec. Brain tissue was taken for morphological analyses 24 hr after the last session. Unbiased stereological methods were used for obtaining estimates of the total number of MSBs in the stratum radiatum of hippocampal subfield CA1. The results showed that the total number of MSBs was significantly increased in conditioned rabbits as compared with pseudoconditioned or unstimulated controls. This conditioning-induced change, which occurs without a net synaptogenesis, reflects a specific synaptogenesis resulting in MSB formation. Models of the latter process are proposed. The models postulate that it requires spine motility and may involve the relocation of existing spines from nonactivated boutons or the outgrowth of newly formed spines for specific synaptogenesis with single-synapse boutons activated by the conditioning stimulation.

Structural synaptic modifications associated with hippocampal LTP and behavioral learning.

An important problem in the neurobiology of memory is whether cellular mechanisms of learning and memory include the formation of new synapses or the remodeling of existing ones. To elucidate this problem, numerous studies have examined alterations in the number and structure of synapses following behavioral learning and hippocampal long-term potentiation (LTP), which is viewed as a synaptic model of memory. The data reported in the literature and obtained in this laboratory are analyzed here to evaluate what kind of structural modification is likely to account for synaptic plasticity associated with learning and memory. It has been demonstrated that LTP induction elicits the formation of additional synapses between activated axon terminals and newly emerging dendritic spines. Similarly, some forms of learning have been shown to increase the number of synapses. Although many ultrastructural studies examining the effect of LTP or learning failed to find a change in total synapse number, this population measure might not detect an increase in a small proportion of synapses established by activated terminals. LTP and learning have also been shown to induce a remodeling of synapses. This process is proposed to involve the transformation of certain synaptic subtypes into more efficacious ones, including the conversion of 'silent' synapses into functional synapses. It appears, therefore, that cellular mechanisms of learning and memory are likely to include both synaptogenesis and synapse remodeling.

Remodeling of hippocampal synapses after hippocampus-dependent associative learning.

The aim of this study was to determine whether hippocampus-dependent associative learning involves changes in the number and/or structure of hippocampal synapses. A behavioral paradigm of trace eyeblink conditioning was used. Young adult rabbits were given daily 80 trial sessions to a criterion of 80% conditioned responses in a session. During each trial, the conditioned (tone) and unconditioned (corneal airpuff) stimuli were presented with a stimulus-free or trace interval of 500 msec. Control rabbits were pseudoconditioned by equal numbers of random presentations of the same stimuli. Brain tissue was taken for morphological analyses 24 hours after the last session. Synapses were examined in the stratum radiatum of hippocampal subfield CA1. Unbiased stereological methods were used to obtain estimates of the total number of synapses in this layer as well as the area of the postsynaptic density. The data showed that the total numbers of all synaptic contacts and various morphological subtypes of synapses did not change in conditioned animals. The area of the postsynaptic density, however, was significantly increased after conditioning in axospinous nonperforated synapses. This structural alteration may reflect an addition of signal transduction proteins (such as receptors and ion channels) and the transformation of postsynaptically silent synapses into functional ones. The findings of the present study indicate that cellular mechanisms of hippocampus-dependent associative learning include the remodeling of existing hippocampal synapses. Further studies examining various time points along the learning curve are necessary to clarify the issue of whether these mechanisms also involve the formation of additional synaptic contacts.

Unbiased stereological estimation of the total number of synapses in a brain region.

Modern stereological methods have been used to make unbiased estimates of the total number of synapses in the striatum radiatum of the hippocampal CA1 region of five rabbits. The approach used involved a two stage analysis and is generally applicable to all parts of the nervous system. During the first stage of the analysis, the reference volume was estimated by point counting, at the light microscope level, according to the Cavalieri principle. During the second stage, the numerical density of synapses was estimated with dissectors at the electron microscopic level. The total number of synapses was calculated as the product of the numerical density and the volume of the region. The sampling with points and dissectors was carried out in all three dimensions of the entire CA1 region in a manner that ensured that all parts of the region and all synapses within it had equal probabilities of being sampled. An analysis of the precision of the estimate of total synapse number has been performed in terms of the variances of volume and synaptic numerical density at different levels of sampling, i.e. at the level of points, sections, individual animals and group of animals. Detailed descriptions of the procedures used to estimate the total number of synapses, evaluate the precision of the estimates, and optimize the sampling scheme are provided.

Synapse restructuring associated with the maintenance phase of hippocampal long-term potentiation.

Synapses in the middle molecular layer of the rat dentate gyrus were analyzed by electron microscopy during the maintenance phase of long-term potentiation (LTP). LTP was induced by high-frequency stimulation of the medial perforant path carried out on each of 4 consecutive days. The dentate gyrus was examined electron microscopically 13 days following the fourth stimulation. At this time point, synaptic responses were still significantly enhanced relative to baseline, although the extent of their potentiation was lower than 1 hour after the last high-frequency stimulation. Stimulated, but not potentiated, rats served as controls. Using the stereological double disector method, estimates of the number of different morphological types of synapses per postsynaptic neuron were obtained. The number of asymmetrical axodendritic synapses increased (by 28%) during LTP maintenance, whereas the number of other synaptic types was not significantly altered. Our previous work demonstrated that the induction of LTP is followed by a selective increase in the number of axospinous perforated synapses with multiple, completely partitioned, transmission zones. Thus, the induction and maintenance phases of LTP are characterized by different structural synaptic alterations. These alterations may be related to each other as indicated by another finding of the present study regarding the existence of perforated synapses that appear to be transitional between axospinous and axodendritic junctions. This suggests a model of structural synaptic plasticity associated with LTP in which some axospinous perforated synapses increase in numbers shortly after the induction of LTP and are then converted into axodendritic ones during LTP maintenance.

Perforated axospinous synapses with multiple, completely partitioned transmission zones: probable structural intermediates in synaptic plasticity.

Analysis of axospinous synapses in the rat dentate gyrus, using three-dimensional reconstructions from electron micrographs of serial sections, revealed a novel synaptic subtype. Synapses of this subtype exhibit partitions that emanate from the postsynaptic spine head and invaginate the presynaptic axon terminal, dividing its portion contracted by the spine into distinct protrusions. Such complete spine partitions provide barriers between two to four discrete transmission zones, each one being formed by a separate axon terminal protrusion and delineated by a separate segment of the postsynaptic density (PSD). Spine partitions that differ from the complete ones were found in two other synaptic subtypes. One of these is characterized by a sectional partition the base of which is placed between the arms of a horseshoe-shaped PSD. Synapses of another subtype exhibit a focal partition the base of which is restricted to a perforation in a fenestrated PSD. Although both sectional and focal partitions invaginate a presynaptic axon terminal, they do not divide into separate protrusions and do not split a single transmission zone into disjointed entities. All three subtypes of partitioned synapses have nonpartitioned counterparts exhibiting segmented, horseshoe-shaped, or fenestrated PSDs. These observations suggest a model of structural modifications underlying synaptic plasticity. According to this model, synapses with multiple, completely partitioned transmission zones that appear to be designed as elements of an unusually high strength, represent pivotal structural intermediates in synaptic plasticity. The formation of such synapses from those that belong to other subtypes is postulated to result in a sustained increase in the efficacy of synaptic transmission. Conversely, a disassembly of complete partitions with the transformation of multiple transmission zones into a single one is proposed to lead to a persistent depression of synaptic responses.

Structural synaptic correlate of long-term potentiation: formation of axospinous synapses with multiple, completely partitioned transmission zones.

Synapses were analyzed in the middle molecular layer (MML) and inner molecular layer (IML) of the rat dentate gyrus following the induction of long-term potentiation (LTP) by high-frequency stimulation of the medial perforant path carried out on each of 4 consecutive days. Potentiated animals were sacrificed 1 hour after the fourth high frequency stimulation. Stimulated but not potentiated and implanted but not stimulated animals served as controls. Using the stereological disector technique, unbiased estimates of the number of synapses per postsynaptic neuron were differentially obtained for various subtypes of axospinous junctions: For atypical (giant) nonperforated synapses with a continuous postsynaptic density (PSD), and for perforated ones distinguished by (1) a fenestrated PSD and focal spine partition, (2) a horseshoe-shaped PSD and sectional spine partition, (3) a segmented PSD and complete spine partition(s), and (4) a fenestrated, (5) horseshoe-shaped, or (6) segmented PSD without a spine partition. The major finding of this study is that the induction of LTP in the rat dentate gyrus is followed by a significant and marked increase in the number of only those perforated axospinous synapses that have multiple, completely partitioned transmission zones. No other synaptic subtype exhibits such a change as a result of LTP induction. Moreover, this structural alteration is limited to the terminal synaptic field of activated axons (MML) and does not involve an immediately adjacent one (IML) that was not directly activated by potentiating stimulation. The observed highly selective modification of synaptic connectivity involving only one particular synaptic subtype in the potentiated synaptic field may represent a structural substrate of the long-lasting enhancement of synaptic responses that characterizes LTP.

Age-related loss of axospinous synapses formed by two afferent systems in the rat dentate gyrus as revealed by the unbiased stereological dissector technique.

Previous attempts to elucidate whether a loss of hippocampal synapses occurs during aging provided conflicting results, possibly due to the unavailability, at the time, of unbiased methods for synapse quantitation. This study was designed to reexamine the issue by means of modern technical procedures that provide unbiased estimates of synaptic numbers. Groups of 14 young adult (5 months old) and 14 aged (28 months old) male Fischer-344 rats were compared. Synapses were examined in the middle (MML) and inner (IML) molecular layer of the hippocampal dentate gyrus, where synaptic contacts are predominantly formed by different systems of afferents, the entorhinal and commissural-associational fibers, respectively. The number of synapses per neuron was estimated with the aid of the stereological dissector technique. The results showed that the total number of synaptic contacts per neuron was significantly diminished in the MML (by 23.6%) and IML (by 22.7%) of aged rats relative to young adults. This age-related synaptic loss involved axospinous, but not axodendritic, junctions of the MML (-24.4%) and IML (-24.0%). Both perforated and nonperforated axospinous synapses (distinguished by a discontinuous or continuous postsynaptic density, respectively) exhibited an age-dependent decrease in numbers, though this decrease did not reach statistical significance in the case of perforated junctions of the IML. The observed age-related loss of axospinous synapses may underlie the reduction in the amplitude of excitatory postsynaptic potentials and the decline in functional synaptic plasticity detected in the dentate gyrus of senescent rats.

Structural synaptic plasticity associated with the induction of long-term potentiation is preserved in the dentate gyrus of aged rats.

Changes in synaptic numbers were examined in the hippocampal dentate gyrus of aged (28 months old) rats following the induction of long-term potentiation (LTP) by high-frequency stimulation of the medial perforant path carried out on each of 4 consecutive days. Potentiated animals were sacrificed 1 hour after the fourth stimulation. Stimulated but not potentiated and implanted but not stimulated rats of the same chronological age served as controls. Synapses were analyzed in the middle (MML) and inner (IML) molecular layer of the dentate gyrus. Using the stereological dissector technique, unbiased estimates of the number per neuron were obtained for the following morphological varieties of synapses: axodendritic synaptic junctions involving dendritic shafts, nonperforated axospinous synapses having a continuous postsynaptic density (PSD), and perforated ones distinguished by a fenestrated, horseshoe-shaped, or segmented PSD. The induction of LTP resulted in a selective increase in the number of synapses with segmented PSDs. This change was detected only in the potentiated synaptic field (MML), but not in an immediately adjacent one (IML), which was not directly stimulated during the induction of LTP. Comparison of these data with the results of our previous LTP study in young adult rats (Geinisman, Y. et al., 1991, Brain Res. 566:77-88) showed that the only significant difference in the absolute number of synaptic contacts per neuron between potentiated animals of the two chronological ages was an age-related reduction in segmented synapses of the MML. Relative increases in the number of segmented synapses per neuron were, however, virtually of the same magnitude in potentiated rats of both ages as compared with their respective controls. This finding may explain why senescent rats can be potentiated to the same extent as young ones.

Increase in the number of axospinous synapses with segmented postsynaptic densities following hippocampal kindling.

Kindling results from intermittent electrical stimulation of a local brain region and leads to a virtually permanent augmentation of synaptic responsiveness in the stimulated circuit. It has been hypothesized that an increase in the number of synapses may represent a structural basis for the enduring expression of synaptic plasticity following kindling, but such an alteration has not been demonstrated unequivocally. The present report provides evidence that hippocampal kindling is indeed accompanied by an increase in synaptic numbers. Young adult rats were kindled via medial perforant path stimulation and sacrificed 4 weeks after reaching a criterion of 5 generalized seizures. Stimulated but not kindled and implanted but not stimulated rats served as controls. Synapses were analyzed in the middle (MML) and inner (IML) molecular layer of the hippocampal dentate gyrus. Using the stereological disector technique, unbiased estimates of the number of synapses per neuron were differentially obtained for 3 morphological subtypes of perforated axospinous synapses characterized by a fenestrated, horseshoe-shaped or segmented postsynaptic density (PSD). A significant increase in synaptic numbers was found to selectively involve only those perforated synapses which are distinguished by a segmented PSD consisting of 2-5 discrete plates. This structural modification was restricted to the terminal synaptic field of stimulated axons (MML), but was not observed in an immediately adjacent synaptic field (IML) which was not directly stimulated during kindling. Since synapses distinguished by a segmented PSD may represent specialized synaptic contacts of an unusually high efficacy, a selective increase in their numbers is likely to provide a structural substrate of the augmented synaptic gain associated with kindling.

Induction of long-term potentiation is associated with an increase in the number of axospinous synapses with segmented postsynaptic densities.

Long-term potentiation (LTP) is characterized by a long-lasting enhancement of synaptic efficacy which may be due to an increase in synaptic numbers. The present study was designed to verify the validity of this suggestion using recently developed unbiased methods for synapse quantitation. LTP was elicited in young adult rats by high-frequency stimulation of the medial perforant path carried out on each of 4 consecutive days. Potentiated animals were sacrificed 1 h after the fourth stimulation. Stimulated but not potentiated and implanted but not stimulated rats served as controls. Synapses were examined in the middle (MML) and inner (IML) molecular layer of the hippocampal dentate gyrus. Using the stereological disector technique, unbiased estimates of the number of synapses per neuron were differentially obtained for the following morphological synaptic types: axodendritic synapses involving dendritic shafts, non-perforated axospinous synapses exhibiting a continuous postsynaptic density (PSD) and perforated axospinous synapses distinguished by a fenestrated, horseshoe-shaped or segmented PSD. A major finding of this study is that the induction of LTP is accompanied by a selective increase in the number of synapses with segmented PSDs. This change was detected only in the potentiated synaptic field (MML), but not in an immediately adjacent one (IML) which was not directly stimulated during the induction of LTP. It is strongly suggested by the latter finding that the increase in the number of axospinous synapses exhibiting segmented PSDs is associated with LTP. Such a highly selective modification of connectivity, which involves only one particular subtype of synapses in the potentiated synaptic field, is likely to represent a structural substrate of the enduring augmentation of synaptic efficacy typical of LTP.

The brain's record of experience: kindling-induced enlargement of the active zone in hippocampal perforated synapses.

Kindling is a consequence of intermittent electrical stimulation of a local forebrain area leading to a durable augmentation of synaptic responsiveness in the stimulated circuit. The basis for this functional change is unknown, but there is evidence suggesting that it entails a structural modification of synapses. The present report demonstrates that hippocampal kindling induces a selective enlargement of active zones in perforated axospinous synapses formed by stimulated axons. Since the active zone is the site of intracellular transmission, its enlargement involving only a certain subpopulation of synapses provides a likely structural substrate of synaptic plasticity associated with kindling.

Increase in the relative proportion of perforated axospinous synapses following hippocampal kindling is specific for the synaptic field of stimulated axons.

A comparative analysis of axospinous synapses was performed in the middle (MML) and inner (IML) molecular layer of the hippocampal dentate gyrus of rats kindled via medial perforant path stimulation and sacrificed 4 weeks after reaching a criterion of 5 generalized seizures. The MML was a directly stimulated structure, while the IML was not. Both are immediately adjacent synaptic fields likely to be equally susceptible to any generalized effects of convulsions and hypoxia. In these two subdivisions of the molecular layer, the so-called perforated and non-perforated synapses, distinguished respectively by a discontinuous or continuous postsynaptic density, were differentially quantified. In the MML, the ratio of perforated to non-perforated synapses was found to be markedly increased in kindled rats relative to controls. In the IML, however, no change in this ratio was detected following kindling. Thus, the shift in the relative preponderance of perforated synapses over non-perforated ones is not a consequence of generalized phenomena accompanying the kindling process.

Perforated synapses on double-headed dendritic spines: a possible structural substrate of synaptic plasticity.

Examination of axospinous synapses in serial sections obtained from the middle molecular layer of the rat dentate gyrus has revealed that some of them involve double-headed dendritic spines. Each spine head is apposed by a separate axon terminal with which it always forms a perforated synaptic contact distinguished by a discontinuous postsynaptic density. The number of perforated synapses on double-headed spines was estimated as a synapse-to-neuron ratio with the aid of the disector technique and found to be significantly increased in rats kindled via medial perforant path stimulation. These results support the notion that perforated synapses involving double-headed dendritic spines represent a structural modification related to enhanced synaptic efficacy.

Age-dependent alterations in hippocampal synaptic plasticity: relation to memory disorders.

In this paper, we review the evidence indicating that the common disturbance in recent memory associated with aging is a consequence of functional and structural impairment in the hippocampal formation. In the Fischer 344 rat, an experimental model of the human age-related memory disorder was developed. The majority of aged rats of this strain show impaired performance in the 8-arm radial maze in a manner typical of young rats with bilateral hippocampal lesions. Aged animals also exhibit rapid decay of LTP and slower kindling of the perforant path-dentate synapse. Furthermore, quantitative morphometric analysis of the hippocampal synaptic architecture revealed that aged, memory-impaired rats had a specific loss of perforated axospinous synapses in the middle third of the dentate gyrus molecular layer; the extent of loss was directly related to the degree of memory dysfunction. Most important was the fact that the electrophysiological and morphological abnormalities did not appear in equally old animals with good memory.