Transient appearance of Ca2+ -permeable AMPA receptors is crucial for the production of repetitive LTP-induced synaptic enhancement (RISE) in cultured hippocampal slices.

We have previously shown that repetitive induction of long-term potentiation (LTP) by glutamate (100 µM, 3 min, three times at 24-hr intervals) provoked long-lasting synaptic enhancement accompanied by synaptogenesis in rat hippocampal slice cultures, a phenomenon termed RISE (repetitive LTP-induced synaptic enhancement). Here, we examined the role of Ca2+ -permeable (CP) AMPA receptors (AMPARs) in the establishment of RISE. We first found a component sensitive to the Joro-spider toxin (JSTX), a blocker of CP-AMPARs, in a field EPSP recorded from CA3-CA1 synapses at 2-3 days after stimulation, but this component was not found for 9-10 days. We also observed that rectification of AMPAR-mediated current appeared only 2-3 days after stimulation, using a whole-cell patch clamp recording from CA1 pyramidal neurons. These findings indicate that CP-AMPAR is transiently expressed in the developing phase of RISE. The blockade of CP-AMPARs by JSTX for 24 hr at this developing phase inhibited RISE establishment, accompanied by the loss of small synapses at the ultrastructural level. These results suggest that transiently induced CP-AMPARs play a critical role in synaptogenesis in the developing phase of long-lasting hippocampal synaptic plasticity, RISE.

Reduction in NPY-positive neurons and dysregulation of excitability in young senescence-accelerated mouse prone 8 (SAMP8) hippocampus precede the onset of cognitive impairment.

The senescence-accelerated mouse prone 8 (SAMP8) strain is considered a neurodegeneration model showing age-related cognitive deficits with little physical impairment. Young SAMP8 mice, however, exhibit signs of disturbances in development such as marked hyperactivity and reduced anxiety well before the onset of cognitive impairment. As the key enzyme in local regulation of thyroid hormone (TH) signaling, type 2 deiodinase, was significantly reduced in the SAMP8 hippocampus relative to that of the normally aging SAM-resistant 1 (SAMR1), we used these two strains to compare the development of the hippocampal GABAergic system, which is known to be strongly affected by hypothyroidism. Among GABAergic components, neuronal K+ /Cl- co-transporter 2 was down-regulated in SAMP8 transiently at 2 weeks. Although distribution of total GABAergic neurons was similar in both strains, 22-30% reduction was observed in the neuropeptide Y (NPY)-positive subpopulation of GABAergic neurons in SAMP8. Electrophysiological studies on hippocampal slices obtained at 4 weeks revealed that epileptiform activity, induced by high-frequency stimulation, lasted four times longer in SAMP8 compared with SAMR1, indicating a dysregulation of excitability that may be linked to the behavioral abnormalities of young SAMP8 and to neurodegeneration later on in life. Local attenuation of TH signaling may thus impact the normal development of the GABAergic system.

A simplified method to generate serotonergic neurons from mouse embryonic stem and induced pluripotent stem cells.

We have developed a new simple method to induce serotonergic neurons from embryonic stem (ES) and induced pluripotent stem cells. When ES or induced pluripotent stem cells were cultured on a thick gel layer of Matrigel, most colonies extended TuJ1-positive neurites. We found that noggin, a known antagonist of bone morphogenic protein, induces ES cells to express genes involved in serotonergic differentiation, such as Nkx2.2, Pet-1, Sonic hedgehog, tryptophan hydroxylase 2, and serotonin transporter, as well as increases high potassium-induced release of serotonin. To concentrate serotonergic neurons, ES cells carrying Pet-1-enhancer-driven enhanced green fluorescent protein were differentiated and sorted into about 80% pure cultures of serotonergic neurons. Whole cell voltage-clamp recordings showed a voltage-dependent current in dissociated neurons. This simplified method provides an alternative option for serotonergic differentiation of pluripotent stem cells and will likely contribute a deeper understanding regarding the nature of serotonergic neurons and open new therapeutic perspectives for the treatment of psychiatric disorders.

Local establishment of repetitive long-term potentiation-induced synaptic enhancement in cultured hippocampal slices with divided input pathways.

Long-term potentiation (LTP) in the rodent hippocampus is a popular model for synaptic plasticity, which is considered the cellular basis for brain memory. Because most LTP analysis involves acutely prepared brain slices, however, the longevity of single LTP has not been well documented. Using stable hippocampal slice cultures for long-term examination, we previously found that single LTP disappeared within 1 day. In contrast, repeated induction of LTP led to the development of a distinct type of plasticity that lasted for more than 3 weeks and was accompanied by the formation of new synapses. Naming this novel plastic phenomenon repetitive LTP-induced synaptic enhancement (RISE), we proposed it as a model for the cellular processes involved in long-term memory formation. However, because in those experiments LTP was induced pharmacologically in the whole slice, it is not known whether RISE has input-pathway specificity, an essential property for memory. In this study, we divided the input pathway of CA1 pyramidal neurons by a knife cut and induced LTP three times, the third by tetanic stimulation in one of the divided pathways to express RISE specifically. Voltage-sensitive dye imaging and Golgi-staining performed 2 weeks after the three LTP inductions revealed both enhanced synaptic strength and increased dendritic spine density confined to the tetanized region. These results demonstrate that RISE is a feasible cellular model for long-term memory.

Involvement of the p75(NTR) signaling pathway in persistent synaptic suppression coupled with synapse elimination following repeated long-term depression induction.

Synaptic plasticity, especially structural plasticity, is thought to be a basis for long-lasting memory. We previously reported that, in rat hippocampus slice cultures, repeated induction of long-term depression (LTD) by application of a metabotropic glutamate receptor (mGluR) agonist led to slowly developing, long-lasting synaptic suppression coupled with synapse elimination. We referred to this phenomenon as LOSS (LTD-repetition-operated synaptic suppression) to discriminate it from conventional single LTD and proposed it as a model for analyzing structural plasticity. Recently, proneurotrophin-activated p75(NTR) signaling has been gaining attention as a possible pathway for the regulation of both neuronal apoptosis and synaptic plasticity. In this study, we examined whether this signaling has a role in the establishment of LOSS. The application of anisomycin indicated that, for LOSS to occur, novel protein synthesis is needed within 6 hr after the induction of mGluR-dependent LTD, which demonstrates that LOSS is an active process and therefore is not due to withering in response to a shortage of trophic factors. Furthermore, we found that pro-BDNF (a species of proneurotrophins) is newly synthesized within 6 hr after the induction of LTD. We therefore exogenously applied a cleavage-resistant form of pro-BDNF, finding synaptic suppression similar to LOSS. LOSS could be abolished by the application of an antibody that binds to and neutralizes p75(NTR) following repeated LTD induction. These results suggest involvement of the p75(NTR) signaling pathway in the long-lasting decremental form of synaptic plasticity.

Analysis of gene expression changes associated with long-lasting synaptic enhancement in hippocampal slice cultures after repetitive exposures to glutamate.

We have previously shown that repetitive exposures to glutamate (100 muM, 3 min, three times at 24-hr intervals) induced a long-lasting synaptic enhancement accompanied by synaptogenesis in rat hippocampal slice cultures, a phenomenon termed RISE (for repetitive LTP-induced synaptic enhancement). To investigate the molecular mechanisms underlying RISE, we first analyzed the time course of gene expression changes between 4 hr and 12 days after repetitive stimulation using an original oligonucleotide microarray: "synaptoarray." The results demonstrated that changes in the expression of synapse-related genes were induced in two time phases, an early phase of 24-96 hr and a late phase of 6-12 days after the third stimulation. Comprehensive screening at 48 hr after the third stimulation using commercially available high-density microarrays provided candidate genes responsible for RISE. From real-time PCR analysis of these and related genes, two categories of genes were identified, 1) genes previously reported to be induced by physiological as well as epileptic activity (bdnf, grm5, rgs2, syt4, ania4/carp/dclk) and 2) genes involved in cofilin-based regulation of actin filament dynamics (ywhaz, ssh1l, pak4, limk1, cfl). In the first category, synaptotagmin 4 showed a third stimulation-specific up-regulation also at the protein level. Five genes in the second category were coordinately up-regulated by the second stimulation, resulting in a decrease in cofilin phosphorylation and an enhancement of actin filament dynamics. In contrast, after the third stimulation, they were differentially regulated to increase cofilin phosphorylation and enhance actin polymerization, which may be a key step leading to the establishment of RISE.

Repetitive induction of late-phase LTP produces long-lasting synaptic enhancement accompanied by synaptogenesis in cultured hippocampal slices.

Long-term plasticity of synaptic transmission is assumed to underlie the formation of long-term memory. Although the cellular mechanisms underlying short-term plasticity have been analyzed in detail, the mechanisms underlying the transformation from short-term to long-term plasticity remain largely unrevealed. We propose the novel long-lasting phenomenon as a model system for the analysis of long-term plasticity. We previously reported that the repetitive activation of cAMP-dependent protein kinase (PKA) by forskolin application led to an enhancement in synaptic strength coupled with synaptogenesis that lasted more than 3 weeks in cultured rat hippocampal slices. To elucidate whether this long-lasting synaptic enhancement depended on the induction of long-term potentiation (LTP) or on the pharmacological effect of forskolin, we applied glutamate (Glu) and correlated its dose with the production of the long-lasting synaptic enhancement. When the dose of Glu was low (10, 30 muM), only transient excitation or early-phase LTP (E-LTP) was induced by a single application and no long-lasting synaptic enhancement was produced by three applications. When the dose was raised to 100 or 300 muM, late-phase LTP (L-LTP) was induced by a single application and long-lasting synaptic enhancement was produced by three applications. The Glu-produced enhancement was accompanied by an increase in the frequency (but not the amplitude) of miniature EPSC and the number of synaptic structures. The enhancement depended on the interval of repetition and protein synthesis immediately after the Glu applications. These results indicate that the repetitive induction of L-LTP, but not E-LTP or transient excitation, triggers cellular processes leading to the long-lasting synaptic enhancement and the formation of new synapses.

Ultrastructural features of hippocampal CA1 synapses with respect to synaptic enhancement following repeated PKA activation.

We reported previously that repeated activations, but not a single activation, of cyclic AMP-dependent protein kinase (PKA), led to a slowly developing (requiring approximately 1 week to develop) long-lasting (lasting > or = 3 weeks) enhancement of synaptic transmission efficiency in the organotypic slice culture of the rat hippocampus. It was accompanied by an increase in the number of synapses identified immunohistochemically. To answer the question of whether the "perforated synapse", which is known to occur transiently after the induction of long-term potentiation (LTP) in combination with the enlargement of postsynaptic density (PSD), is involved also in this slow/persistent synaptic enhancement, we examined the ultrastructural changes after the repeated activations of PKA. The answer was partially yes (occurrence of perforated synapses was increased) but partially no (the increase in the number of perforated synapses was not transient but persistent; mean apparent size of PSD did not increase). These results suggest that the mechanism of the slow/persistent synaptogenesis shares limited features with the mechanism of the quick/transient morphogenesis after LTP.

Possible involvement of BDNF release in long-lasting synapse formation induced by repetitive PKA activation.

For the analysis of the cellular mechanism underlying long-term synaptic plasticity, a model system that allows long-lasting pursuit is required. Previously we reported that, in hippocampal neurons under dissociated cell culture conditions, repeated (but not a single) transient activation of protein kinase A (PKA) led to an increase in the number of synapses that lasted >3 weeks, and hence we proposed that this phenomenon should serve as an appropriate model system. Here we report that repeated pulsatile application of brain-derived neurotrophic factor (BDNF) leads to persistent synapse formation equivalent to that after the repeated transient activation of PKA. A BDNF-scavenging substance applied concomitantly with PKA activation abolished the synapse formation. The release of BDNF upon PKA activation was confirmed by phosphorylation of TrkB. These results indicate that the release of BDNF is involved in the putative signaling cascade connecting PKA activation and synapse formation.

An in vitro reproduction of stress-induced memory defects: Effects of corticoids on dendritic spine dynamics.

Previously, in organotypic slice culture of rodent hippocampus we found that three repeated inductions of LTP, but not a single induction, led to a slow-developing long-lasting enhancement of synaptic strength coupled with synapse formation. Naming this structural plasticity RISE (repetitive LTP-induced synaptic enhancement) and assuming it to be a potential in vitro reproduction of repetition-dependent memory consolidation, we are analyzing its cellular mechanisms. Here, we applied a glucocorticoid to the culture to mimic acute excess stress and demonstrated its blockade of RISE. Since excess stress interferes with behavioral memory consolidation, the parallelism between RISE in vitro and memory consolidation in vivo is supported. We recently reported that RISE developed after stochastic processes. Here we found that the glucocorticoid interfered with RISE by suppressing the increment of dendritic spine fluctuation that precedes a net increase in spine density. The present study provides clues for understanding the mechanism of stress-induced memory defects.

Long-lasting synapse formation in cultured rat hippocampal neurons after repeated PKA activation.

Recently, we reported that the repeated activation of cyclic-AMP-dependent protein kinase (PKA) in the rat hippocampus under tissue culture conditions induced the enhancement of excitatory postsynaptic potential (EPSP), which lasted more than 2 weeks and was accompanied by the formation of morphologically identifiable synapses. Here we examined whether an equivalent synapse formation is induced in dissociated cell cultures of rat hippocampal neurons. Brief (15-min) application of Sp-cAMPS (a membrane-permeable analog of cyclic AMP) induced an increase in the number of synaptic sites (identified by the apposition of immunocytochemically labeled pre- and postsynaptic structures). There were two types of increase: a short-lasting one that lasted less than 24 h after a single application of Sp-cAMPS, and a long-lasting one that lasted more than 2 weeks after repeated applications. The long-lasting increase in synaptic sites was dependent on the time and interval of application and was suppressed by Rp-cAMPS (a PKA inhibitor). The synapses were judged to be active based on the endocytosis of FM1-43, a fluorescent dye. Electron microscopy confirmed the increase in the number of synaptic ultrastructures. The present results show that the synaptogenesis induced by repeated PKA activation is reproducible in a neuronal network that is reconstituted under dissociated cell culture conditions. This experimental system, together with the synaptogenesis in the slice culture system described previously, serves as a good in vitro model for the analysis of the process of conversion from short-lasting plasticity (lasting for hours) into a long-lasting one (lasting for days-weeks).

Repetition of mGluR-dependent LTD causes slowly developing persistent reduction in synaptic strength accompanied by synapse elimination.

Synaptic plasticity, the cellular basis of memory, operates in a bidirectional manner. LTP (long-term potentiation) is followed by structural changes that may lead to the formation of new synapses. However, little is known whether LTD (long-term depression) is followed by morphological changes. Here we show that the repetitive induction of metabotropic glutamate receptor (mGluR)-dependent LTD in stable cultures of rat hippocampal slices led to a slowly developing persistent (ranging over weeks) reduction in synaptic strength that was accompanied by the loss of synaptic structures. LTD was induced pharmacologically 1-3 times at 24-h intervals by applying aseptically ACPD (1-aminocyclopentane-1,3-dicarboxylic acid), an agonist of group I/II mGluR, and APV (2-amino-5-phosphonovalerate), an antagonist of the NMDA (N-methyl-D-aspartate) receptor. One ACPD/APV application induced LTD that lasted less than 24 h. After three LTD inductions, however, a gradual attenuation of the fEPSP (field excitatory postsynaptic potential) amplitude and a decrease in the number of pre- and postsynaptic structures were observed. The blockade of LTD by an mGluR antagonist or a protein phosphatase 2B inhibitor abolished the development of the synaptic attenuation. In contrast to our previous finding that the repetitive LTP induction triggered a slowly developing persistent synaptic enhancement, the incremental and decremental forms of synaptic plasticity appeared to occur symmetrically not only on the minutes-hours time order but also on the days-weeks time order.

Dendritic spine dynamics leading to spine elimination after repeated inductions of LTD.

Memory is fixed solidly by repetition. However, the cellular mechanism underlying this repetition-dependent memory consolidation/reconsolidation remains unclear. In our previous study using stable slice cultures of the rodent hippocampus, we found long-lasting synaptic enhancement/suppression coupled with synapse formation/elimination after repeated inductions of chemical LTP/LTD, respectively. We proposed these phenomena as useful model systems for analyzing repetition-dependent memory consolidation. Recently, we analyzed the dynamics of dendritic spines during development of the enhancement, and found that the spines increased in number following characteristic stochastic processes. The current study investigates spine dynamics during the development of the suppression. We found that the rate of spine retraction increased immediately leaving that of spine generation unaltered. Spine elimination occurred independent of the pre-existing spine density on the dendritic segment. In terms of elimination, mushroom-type spines were not necessarily more stable than stubby-type and thin-type spines.

The dendritic layer-specific persistent enhancement of synaptic transmission induced by repetitive activation of protein kinase A.

Synaptic plasticity, the cellular basis of brain memory, is established through at least two phases: short-term and long-term plasticity. It is assumed that the short-term plasticity instantaneously provoked in pre-existing synapses, as represented by a long-term potentiation (LTP) in the mammalian hippocampus, is converted to the long-term plasticity that develops slowly accompanying the formation of new synapses. However, this conversion has scarcely been analyzed primarily because of the lack of the model system. Recently, we found that a repeated activation of protein kinase A (PKA), but not a single activation of PKA, led to a slowly-developing long-lasting enhancement of synaptic strength coupled with synaptogenesis in cultured rat hippocampus and proposed that this phenomenon would serve as the required model system. In the present study, we investigated the geographical aspect of this phenomenon using a high-speed voltage-sensitive dye (VSD) imaging methodology. Before doing this, we had to overcome the difficulties in applying this methodology to the quantitative analysis on the cultured hippocampal slices. Those difficulties are multiple types of signal decay and a large variance in the number of cells among specimens. After resolving these problems we found that the enhancement of synaptic efficacy in the CA1 stratum radiatum occurred predominantly in the proximal dendritic layer.

Non-synaptic exocytosis enhanced in rat cerebellar granule neurons cultured under survival-promoting conditions.

Cultured rat cerebellar granule neurons (CGNs) are often used to analyze activity-dependent neuronal selection occurring during brain development. The CGNs survive long only when the culture medium contains a depolarizing agent. However, it is argued whether the depolarization critical for survival is of presynaptic or postsynaptic compartment. Since CGNs form no synapses among them, it is generally assumed that the latter would be the case. But it is possible that the depolarization would induce exocytosis of survival-promoting substances whether or not CGNs form synapses. Here we directly examined the exocytotic activities of CGNs under survival-promoting and survival-limiting conditions by electron microscopy to support this possibility. CGNs possessed clusters of synaptic vesicle-like vesicles (SVVs) in neuritic varicosities. CGNs cultured in high-KCl medium had significantly smaller SVV clusters than those cultured in low-KCl medium. The number of SVVs increased when the high KCl-cultured CGNs were transferred to low-KCl medium, indicating a sustained high rate of exocytosis in high-KCl medium. The majority of the varicosities containing SVVs were not apposed to definite postsynaptic structures, indicating that exocytosis occurs from a non-synaptic surface. Fluorescence Ca(2+) imaging revealed that the high KCl-cultured CGNs had spots of high Ca(2+) along their neurites, corresponding to the varicosities.

Repetitive activation of protein kinase A induces slow and persistent potentiation associated with synaptogenesis in cultured hippocampus.

Mammalian brain memory is hypothesized to be established through two phases; short-term plasticity, as exemplified by long-term potentiation (LTP) where pre-existing synapses change transmission efficiency, and long-lasting plasticity where new synapses are formed. This hypothesis, however, has not been verified experimentally. Using cultured hippocampal slices, we show that the repeated induction of late-phase LTP by brief applications of forskolin (FK) led to a slowly-developing long-lasting synaptogenesis, as judged from electrophysiological, cytological and ultrastructural indices. These indices include (1) field postsynaptic potential standardized by field action potential, which should represent the number of synapses per neuron; (2) the amounts of synaptic marker proteins; (3) the number of synaptophysin-immunopositive puncta; (4) the number of dendritic spines per length; (5) the density of synaptic ultrastructures; (6) ultrastructures similar to synapse perforation. Increment in these indices occurred approximately 10 days after FK-application and outlasted the following weeks. The increment depended on the times and intervals of FK-application. A biologically inert FK analogue failed to produce the similar effect. An inhibitor for cyclic AMP-dependent protein kinase (PKA) blocked the synaptogenesis. The cultured brain slice repeatedly exposed to FK should serve as a good model system for the analysis of persistent synaptogenesis possibly related to long-term memory in mammalian CNS.

Survival promotion of rat cerebellar granule neurons by co-culture with pontine explant.

Cultured cerebellar granule neuron (CGN) of the rat is the most frequently used model system for analysis of activity-dependent neuronal survival. CGNs do not survive longer than 2 weeks in a standard culture medium unless KCl (or other excitants such as glutamate) is added. It is assumed that KCl represents synaptic activity, but no tests have been made on whether the survival of CGNs really depends on the synaptic input. Here we co-cultured CGNs with an explant of the pons including the basilar pontine nucleus (BPN), which is one of the input sources of CGNs in vivo, to confirm if synaptic input is really a determinant for the survival of these cells. In this co-culture system, the viability of CGNs was significantly increased without the addition of KCl. The survival promotion was confined to the population of CGNs having contact with neurites of BPN and was cancelled by an application of tetrodotoxin or antagonists of glutamate receptors, indicating that the survival depended on synaptic activity. Explants of other glutamatergic tissues including the hippocampus failed to promote the survival, although neurites grew out from these explants as vigorously as from the BPN explants. Calcium and FM1-43 imaging examinations revealed that the CGNs had formed functional synapses with the BPN explant but not with the hippocampal explant. These results, confirming the assumption that synaptic activity determines neuronal survival, provide evidence for presynaptic contribution to the survival.

Ultrafilter co-culture, a new method for estimating the molecular mass of bioactive substances, indicates a small molecule neurotrophic substance is released from cultured cerebellar granule neurons of the BALB/c mouse.

Cultured rat cerebellar granule neurons (CGNs), which require a depolarizing agent in the medium for long-term survival, are widely used for the analysis of mechanisms underlying the activity-dependent survival of neurons. It was recently found that this is not the case for BALB/c mouse CGNs, which survive without a depolarizing agent. Co-culture experiments indicated that the mouse cells release a neurotrophic substance. However, the substance is apparently short-living in the medium, making its molecular identification difficult. Here a novel co-culture method was devised for estimating the relative molecular masses of biologically active substances, using a commercially available dialysis membrane filter unit to separate substance-donor from substance-recipient cells. By this simultaneous fractionation/bioassay, the molecular mass of the assumed neurotrophic substance was estimated to be <3 kDa. Neurotrophic substances previously reported to be effective in rat CGNs, including neurotrophins, pituitary adenylate cyclase-activating polypeptide, parathyroid hormone-related polypeptide, glutamic acid, gamma-aminobutyric acid, and D-serine, were excluded as candidate molecules. Estrogen, however, remained a candidate. It should be stressed that the requirements for the activity-dependent survival of CGNs are species-dependent. Care should be taken in the analysis of activity-dependent neuronal survival using transgenic animals.

Involvement of TrkB- and p75(NTR)-signaling pathways in two contrasting forms of long-lasting synaptic plasticity.

The repetition of experience is often necessary to establish long-lasting memory. However, the cellular mechanisms underlying this repetition-dependent consolidation of memory remain unclear. We previously observed in organotypic slice cultures of the rodent hippocampus that repeated inductions of long-term potentiation (LTP) led to a slowly developing long-lasting synaptic enhancement coupled with synaptogenesis. We also reported that repeated inductions of long-term depression (LTD) produced a long-lasting synaptic suppression coupled with synapse elimination. We proposed these phenomena as useful in vitro models for analyzing repetition-dependent consolidation. Here, we hypothesized that the enhancement and suppression are mediated by the brain-derived neurotrophic factor (BDNF)-TrkB signaling pathway and the proBDNF-p75(NTR) pathway, respectively. When we masked the respective pathways, reversals of the enhancement and suppression resulted. These results suggest the alternative activation of the p75(NTR) pathway by BDNF under TrkB-masking conditions and of the TrkB pathway by proBDNF under p75(NTR)-masking conditions, thus supporting the aforementioned hypothesis.

Dendritic spine dynamics in synaptogenesis after repeated LTP inductions: dependence on pre-existing spine density.

Not only from our daily experience but from learning experiments in animals, we know that the establishment of long-lasting memory requires repeated practice. However, cellular backgrounds underlying this repetition-dependent consolidation of memory remain largely unclear. We reported previously using organotypic slice cultures of rodent hippocampus that the repeated inductions of LTP (long-term potentiation) lead to a slowly developing long-lasting synaptic enhancement accompanied by synaptogenesis distinct from LTP itself, and proposed this phenomenon as a model system suitable for the analysis of the repetition-dependent consolidation of memory. Here we examined the dynamics of individual dendritic spines after repeated LTP-inductions and found the existence of two phases in the spines' stochastic behavior that eventually lead to the increase in spine density. This spine dynamics occurred preferentially in the dendritic segments having low pre-existing spine density. Our results may provide clues for understanding the cellular bases underlying the repetition-dependent consolidation of memory.