Activity-dependent localization in spines of the F-actin capping protein CapZ screened in a rat model of dementia Takuma Kitanishi1,a, Jun Sakai2, Shinichi Kojima2, Yoshito Saitoh3, Kaoru Inokuchi3, Masahiro Fukaya4, Masahiko Watanabe4, Norio Matsuki1 and Maki K. Yamada1,5,b* 1

Laboratory of Chemical Pharmacology, Graduate School of Pharmaceutical Sciences, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan 2 Dainippon Sumitomo Pharma Co., Ltd, 3-1-98 Kasugade Naka, Konohana-ku, Osaka 554-0022, Japan 3 Mitsubishi Kagaku Institute of Life Sciences, MITILS, 11 Minamiooya, Machida, Tokyo 194-8511, Japan 4 Department of Anatomy and Embryology, Hokkaido University Graduate School of Medicine, Sapporo 060-8638, Japan 5 PRESTO, JST, 4-1-8 Honcho Kawaguchi, Saitama 332-0012, Japan

Actin reorganization in dendritic spines is hypothesized to underlie neuronal plasticity. Actinrelated proteins, therefore, might serve as useful markers of plastic changes in dendritic spines. Here, we utilized memory deficits induced by fimbria-fornix transection (FFT) in rats as a dementia model to screen candidate memory-associated molecules by using a two-dimensional gel method. Comparison of protein profiles between the transected and control sides of hippocampi after unilateral FFT revealed a reduction in the F-actin capping protein (CapZ) signal on the FFT side. Subsequent immunostaining of brain sections and cultured hippocampal neurons revealed that CapZ localized in dendritic spines and the signal intensity in each spine varied widely. The CapZ content decreased after suppression of neuronal firing by tetrodotoxin treatment in cultured neurons, indicating rapid and activity-dependent regulation of CapZ accumulation in spines. To test input specificity of CapZ accumulation in vivo, we delivered high-frequency stimuli to the medial perforant path unilaterally in awake rats. This path selectively inputs to the middle molecular layer of the dentate gyrus, where CapZ immunoreactivity increased. We conclude that activity-dependent, synapse-specific regulation of CapZ redistribution might be important in both maintenance and remodeling of synaptic connections in neurons receiving specific spatial and temporal patterns of inputs.

Introduction The vast majority of excitatory synapses in the central nervous system occur on dendritic spines (Harris & Stevens 1989). Recent findings have revealed that the shape of spines (e.g., head size, length and neck width) is closely associated with their functional properties such as size of excitatory currents (Matsuzaki et al. 2001) and biochemical and electrical properties Communicated by : Noriko Osumi *Correspondence: [email protected] a Present address : Kavli Institute for Systems Neuroscience and Centre for the Biology of Memory, Norwegian University of Science and Technology, 7489 Trondheim, Norway. b Present address: PRESTO, Department of Cellular Neurobiology, Graduate School of Medicine and Faculty of Medicine, The University of Tokyo, 7-3-1 Hongo Bunkyo-ku, Tokyo 113-0033, Japan.

(Bloodgood & Sabatini 2005; Noguchi et al. 2005; Araya et al. 2006). Furthermore, spines undergo activity-dependent morphological plasticity. The induction of long-term potentiation (LTP) is associated with long-lasting enlargement of existing spines and de novo emergence of spines (Engert & Bonhoeffer 1999; Matsuzaki et al. 2004), while long-term depression leads to shrinkage and retraction of them (Nagerl et al. 2004; Zhou et al. 2004). Filamentous actin (F-actin) is a crucial component of the spine cytoskeleton, and its dynamic behavior regulates the shape of spines (Matus et al. 2000). Thus, quantitative monitoring of actin-related proteins in spines may provide a useful readout of the ongoing plastic changes in neural networks that are responsible for memory formation and maintenance. Fimbria-fornix transection (FFT) in animals has long been used as a model of memory deficit (Gaffan

DOI: 10.1111/j.1365-2443.2010.01411.x  2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Genes to Cells (2010) 15, 737–747

737

T Kitanishi et al.

1972; Cassel et al. 1997). It involves the axotomy of septohippocampal projection neurons, leading to a loss of theta-wave activity (Oddie et al. 2002) and inhibition of c-fos expression in the hippocampus ipsilateral to unilateral FFT (Vann et al. 2000), both of which are implicated in memory formation. Within 7–10 days after FFT surgery, deficits in hippocampus-dependent spatial learning and memory function become apparent without detectable histological damage to the hippocampal trisynaptic glutamatergic circuits (Cassel et al. 1997). Bilateral FFT has been shown to induce memory deficit similar to that observed in Alzheimer’s disease within 10 days (Gaffan 1972). In addition, our previous study of the bilateral FFT protocol and its subsequent effects on animal behavior confirmed the initiation of learning deficits by 10 days after the surgery (Nakao et al. 2001). The animals with bilateral FFT at 10 days show marked learning deficits, which would accompany changes in protein markers of plasticity independent of compensatory mechanisms. In this study, we designed an experiment with animals that had undergone unilateral FFT instead of bilateral transection to obtain control protein samples within the same animals. We believe that comparison of protein profiles between the ipsilateral and contralateral hippocampi after unilateral FFT in the same rats can reduce the variability associated with inter-individual genetic, physiological and environmental differences. Using a proteomic approach (fluorescent two-dimensional differential gel electrophoresis, 2D-DIGE), we identified the F-actin capping protein CapZ (Schafer & Cooper 1995) among the candidate proteins showing changes after FFT. CapZ is known to regulate actin dynamics in non-neuronal cells (Mejillano et al. 2004) and regulate growth cone morphology and neurite outgrowth in cultured hippocampal neurons (Davis et al. 2009). However, its localization and activity-dependent regulation in mature neurons are not yet clear. To evaluate its ability to mark a specific subset of synapses undergoing activity-dependent changes, we further examined the localization and dynamics of CapZ in hippocampal neurons, both in vitro and in vivo.

Results Changes in 2D-DIGE profile of CapZ after unilateral FFT

By using 2D-DIGE, we analyzed the difference in protein profiles in an intact hippocampus and a tran738

Genes to Cells (2010) 15, 737–747

sected side at 10 days after unilateral FFT in rats. Of more than 400 spots detected by the 2D-DIGE system (Fig. 1A), 27 spots exhibited statistically significant differences in signal intensity between the transected and intact hemispheres in the same animal (P < 0.01 by Student’s t-test, N = 5 rats). While the intensity in the lesioned side had increased in most of proteins (e.g., GFAP, vimentin, HSP90 and 14-3-3), it decreased in seven proteins: drebrin-like SH3P7r3, CRMP4, profilin Iia, the guanine nucleotide-binding protein transducin, chaperonin-containing Tcp1, the

(A)

(B)

(C)

Figure 1 2D-DIGE analysis showing a decrease in the amount of CapZ beta 2 in the fimbria-fornix transection (FFT) hippocampus. (A) A representative image of the 2DDIGE analysis showing protein spots at 10 days after unilateral FFT. The CapZ beta spot is indicated by a red box. M.W.: molecular weight. (B) Three-dimensional views of the signal intensity around the CapZ beta spot indicated in A show that it was well isolated from neighboring spots. The spot from the transected hippocampus had lower intensity than that from the control side. (C) The signal intensity of the spots in B was significantly decreased in the transected sides of all rats (N = 5 rats, **P < 0.01 by Student’s t-test). The mean signal intensities for each condition are connected by a line. Cont: control side; Trans: transected side.

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Localization of CapZ in dendritic spines

transcriptional activator protein PUR-alpha and F-actin capping protein (CapZ). Among these candidate proteins, CapZ is directly involved in the reorganization of the F-actin network and is likely to play an important role in stabilizing changes in neuronal morphology, especially at synaptic connections. Therefore, we subsequently examined the localization and dynamic redistribution of CapZ during activitydependent alterations in the neural network. CapZ is a heterodimer consisting of alpha and beta subunits. Here, we analyzed the CapZ splicing variant beta 2 subunit, which is more abundantly expressed in the brain than the other variant, beta 1 (Hart & Cooper 1999; Huang et al. 1999). The intensity of the spot corresponding to the beta subunit was significantly decreased in the transected side of all five rats (Fig. 1B,C). However, 1D SDS-PAGE immunoblot analysis of the same samples as those used in the 2D-DIGE analysis of CapZ beta 2 showed no statistically significant differences in the signal intensity of this protein (104 ± 1% of that in the control, N = 5 rats, P = 0.12 in Student’s t-test, Supplementary Fig. S1B). The specificity of the monoclonal antibody for CapZ beta 2, 3F2.3, was confirmed (see Supplementary Fig. S1A), wherein a single band of approximately 30 kDa (deduced molecular weight is 31.3 kDa) was detected. Three smaller spots with identical molecular weight but with different isoelectric values in 2D immunoblot of CapZ suggest protein modifications (Supplementary Fig. S1C). Thus, the decrease in the signal observed in the 2D-DIGE

(A)

analysis was not because of the change in the net amount of CapZ protein but because of a change in the protein state (e.g., tight association with other molecules) that hampers the isoelectric focusing of the protein spot in one-dimensional separation. Enrichment of CapZ in a subset of dendritic spines and its dependency on neuronal activity

To examine the expression of CapZ in neurons and its subcellular localization, we immunocytochemically analyzed dissociated and cultured hippocampal neurons with anti-CapZ beta 2 antibody. A punctate pattern of CapZ immunoreactivity (i.r.) was observed along the dendrites and in dendritic spines, where all protrusions whose length exceeded 0.5 lm were defined as spines (Fig. 2A. rhodamine-phalloidin was used to detect F-actin, which is abundant in spines.). CapZ i.r. was also observed in cell body with moderate reactivity and in thin processes corresponding to isolated and bundled axons with low reactivity. Glial cells showed less i.r. than neurons (data not shown). To exclude the possibility that the immunoreactive puncta corresponded to focal adhesion contacts, we examined the distance in the z-direction between CapZ-immunopositive puncta and the culture surface. The presence of a space between them indicated that CapZ was accumulated in structures distinct from focal adhesion contacts. Interestingly, the intensity of the i.r. signal in each spine (average pixel intensity in a spine) varied

(B)

Figure 2 Heterogeneous localization of CapZ among spines and its regulation by neuronal activity. (A) Dissociated hippocampal neuronal cultures immunostained with anti-CapZ beta 2 antibody and counterstained with rhodamine-phalloidin (purple). Punctate CapZ signal (green) was observed in dendrites and heterogeneously in dendritic spines. Some spines were labeled strongly (arrowheads), while others were moderately labeled (yellow arrows). The right panels represent the CapZ immunoreactivity (i.r.) in each spine against its head size and length. N = 265 spines from 16 cells in three independent experiments. Scale bar, 2 lm. (B) CapZ i.r in the spines of hippocampal neurons decreased after treatment with the sodium channel blocker tetrodotoxin (TTX; 1 lM) for 60 min. Each symbol represents an i.r. value from a spine (N = 265 spines from 16 cells for the control group and 219 spines from 11 cells for the TTX-treated group in 3 independent experiments) and **P < 0.01 by Mann–Whitney’s U-test. The line in the middle of each box-and-whisker plot indicates the median, that at the top of the box indicates the 75th quartile, that at the bottom of the box indicates the 25th quartile, and the whiskers indicate the extent of the 10th and 90th percentiles, respectively. The right panel shows the frequency distribution (P) of the i.r. values in percentage from the control and TTX-treated spines.  2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Genes to Cells (2010) 15, 737–747

739

T Kitanishi et al.

widely, ranging from a value below the level of detection to a value 10-fold greater than the median intensity level (Fig. 2A, right graphs). When the intensity was plotted against the spine head size and spine length, a positive (N = 265 spines from 16 cells, r2 = 0.14, P < 0.01 by Pearson’s correlation test) and a negative (r2 = 0.13, P < 0.01) correlations were found, respectively. Considering that short spines tend to have a large head (N = 265 spines, r2 = 0.046, P < 0.01 by Pearson’s correlation test), these data indicate that short and large-head spines tend to have relatively high levels of CapZ. However, considerable variation in CapZ intensity was observed among spines having similar head size and length (Fig. 2A, white arrow heads and yellow arrows), suggesting that spine morphology is not the sole determinant of the CapZ content. Conversely, the observation of spines with widely varying morphology but similar CapZ reactivity indicates that the CapZ content is less likely to determine spine morphology. To rule out the possibility that the observed localization might be an aberration observed because of antibody accessibility, we co-expressed CapZ tagged with EGFP together with monomeric red fluorescent protein and found similar expression pattern (Supplementary Fig. S2). As we found that spine morphology is not the sole determinant of CapZ accumulation, we hypothesized that previous synaptic transmission may regulate the CapZ content in spines. To test this hypothesis, cultured neurons were treated with 1 lM of the sodium channel blocker tetrodotoxin (TTX) for 60 min. The level of CapZ beta 2 i.r. in each individual spine was found to be significantly lower in the TTX-treated group than in nontreated controls (Fig. 2B), indicating rapid and activity-dependent regulation of CapZ localization. Enrichment of CapZ in a subset of dendritic spines in vivo

On the basis of the heterogeneity of the postsynaptic CapZ concentration observed in the culture system, we next investigated the extent of heterogeneity of CapZ beta 2 accumulation in hippocampal spines in vivo. The observations in the culture system revealed the presence of a mechanism underlying the accumulation of CapZ in a subset of spines that tend to be shorter and have larger heads. To examine whether a similar mechanism exists in the intact hippocampus of adult animals, we used Thy1-mGFP transgenic mice, which sparsely express membrane-targeted GFP (mGFP) in neurons. A punctate and discrete CapZ 740

Genes to Cells (2010) 15, 737–747

signal was found in the spines of CA1 pyramidal and dentate granule cells (Fig. 3A, N = 4 cells each from four mice from three independent experiments for CA1 apical, CA1 basal and dentate apical dendrites). Distinct CapZ signal inside spines was observed only in 25% (peri-spine, 21%; no detectable signal, 54%) of all spines (N = 120 spines of apical dendrites from 4 CA1 pyramidal cells). In the mushroom-shaped spines, which corresponded 30% of all spines, CapZ signals were frequently found (53%, N = 36 spines). Thus, CapZ localized heterogeneously among spines and a specific subset of spines also showed preferential accumulation of CapZ in vivo. CapZ i.r. was observed throughout the hippocampus in all mice (N = 6) and rats (N = 7) examined (Fig. 3B). Immunoelectron microscopy confirmed that CapZ signals were localized in spines (Fig. 3C). Input layer-selective increase in CapZ after LTP-inducing stimuli

Detailed in vitro and in vivo analysis of CapZ localization revealed its accumulation in a specific subset of spines. Together with the experiment with TTX treatment (Fig. 2B), these results raised the possibility that CapZ accumulation was closely regulated by the activity-dependent signaling system. As neighboring spines along the dendritic shafts showed wide heterogeneity in the content of CapZ, synapse-specific signaling mechanisms might lead to heterogeneous nature of CapZ expression in selective spines. To determine whether selective stimulation of particular axons can induce input-specific CapZ accumulation in individual spines, we unilaterally delivered a highfrequency stimulus (HFS) to the medial perforant path (MPP) axons in awake animals (Fig. 4). The axons in the MPP selectively make synapses on dendritic segments of the middle molecular layer (MML) of ipsilateral dentate gyrus granule cells. The HFS used here has been shown to induce robust LTP and actinrelated structural reorganization at MPP-MML synapses (Fukazawa et al. 2003). At 45 min after HFS delivery, the CapZ beta 2 i.r. increased significantly at the ipsilateral MML (Fig. 4A–D). The i.r. at the ipsilateral MML increased at multiple test time points (30 and 60 min, N = 2 rats each) after HFS delivery. The CapZ signal did not increase in the neighboring inner and outer molecular layers or in the contralateral MML. The staining intensity of rhodamine-phalloidin, with which F-actin is visualized, was also increased only in the ipsilateral MML (Fig. 4E–H), and this finding is

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Localization of CapZ in dendritic spines (A)

(B)

(C)

Figure 3 Distribution and localization of CapZ in the hippocampus in vivo. (A) The CapZ i.r. (purple) in spines was examined in the brain sections of Thy1-mGFP (membrane-bound GFP; green) mice. Subsets of spine heads were labeled with anti-CapZ antibody (white arrowheads), while some spines had no detectable labeling (yellow arrowhead). The right images are examples of labeled spines that had varied labeling intensities from strong to weak. The CapZ i.r. pattern is also shown in reconstructed xz plane images. The graphs on the right of each image represent the fluorescent intensity profiles of CapZ (purple) and mGFP (green) along the white line in the image. The numbers in the right panels correspond to the numbers of spines in the left figures, and the other two images are from different cells. Scale bar, 2 lm. (B and C) Immunofluorescence (B) and postembedding immunoelectron microscopic analyses (C) were performed with anti-CapZ beta 2 antibody. CapZ was ubiquitously observed in the mouse hippocampus. DG: dentate gyrus, Sub: subiculum. CapZ beta 2 was often detected at the postsynaptic density (arrows) and was also found in the extrasynaptic plasma membrane and cytosol (arrowheads) of dendritic spines. Ra: stratum radiatum, Sp: dendritic spine, NT: presynaptic nerve terminal. Scale bars, 500 lm in (B) and 0.1 lm in (C).

consistent with previous reports (Fukazawa et al. 2003). Thus, within a single dentate gyrus granule cell, CapZ selectively accumulated in the MML that received synaptic stimulation (Fig. 4D). Schematic illustration of dentate granule cells and the stimulated axons within MML are shown in Fig. 4I. The staining was seen as puncta that matched the sizes of spines, and the intensity of each puncta increased in the MML (Fig. 4J). In Fig. 4K, we have shown partial colocalization of CapZ and phosphorylated Cofilin (pCofilin), that is induced by LTP-inducing stimulation and is mostly restricted to dendritic spines (Chen et al. 2007b; Fedulov et al. 2007). According to Fedulov et al., pCofilin was sparsely found as 1.4 puncta ⁄ 100 lm3 in a normal condition, while total spines are 300 ⁄ 100 lm3. The result shows that CapZ and pCofilin had similar density and co-localized only in a limited portion in the ipsilateral MML, suggesting that CapZ accumulation in spines is limited and heterogeneous even in the MML that received HFS. The difference in the distribution between two activity-dependent molecules may be attributable to the difference in the time course, because pCofilin is

rapidly accumulated (within 2–7 min) and disappeared to half at 15–30 min after the stimulation.

Discussion CapZ is a well-known F-actin capping protein named after the Z-band of muscle fibers, where it stabilizes, holds and aligns F-actin (Casella et al. 1987). Few studies have focused on the role of CapZ in the CNS. In the present study, we found a decrease in the intensity of the CapZ spot in 2D-DIGE analysis of the hippocampus immediately after the induction of dementia, and we revealed that CapZ accumulates in a subset of dendritic spines and is regulated by neuronal activity. Previous studies have revealed that CapZ is found in the postsynaptic density (Jordan et al. 2004; Li et al. 2004; Peng et al. 2004; Yoshimura et al. 2004), and its content is decreased in the brains of individuals with fetal Down’s syndrome (Gulesserian et al. 2002). Interestingly, CapZ was recently reported to bind to disabled-1 (Sato et al. 2007), which regulates amyloid beta production (Hoe et al. 2006). These reports together with our present findings suggest that CapZ regulation

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Genes to Cells (2010) 15, 737–747

741

T Kitanishi et al. (A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

(I)

(J)

(K)

Figure 4 CapZ accumulation in response to high-frequency stimulation. Histochemical analysis of CapZ beta 2 (A–D) and F-actin (E–H) at 45 min after HFS application to the MPP, which induces LTP at the ipsilateral MML synapses of the dentate gyrus, in freely moving rats. (I) Schematic illustration of the MPP and granule cells in the dentate gyrus. (A and E) Signals of antiCapZ antibody (A) and rhodamine-phalloidin to visualize F-actin (E) in the dentate gyrus ipsilateral to the HFS. (B, C, F and G) Images of the ipsi- and contralateral sides of the dentate gyrus. Fluorescence intensity profiles for each image are shown on the right. O: outer molecular layer, M: MML, I: inner molecular layer, GCL: granule cell layer. In panels D and H, the fluorescent intensity was quantified across all molecular layers. The intensity outside the dentate gyrus (stratum lacunosum-moleculare of the CA1 area) in each image was used for normalization (1.0). Open and filled symbols indicate the ipsi- and contralateral sides, respectively. Data are represented as mean ± SEM of the values from four rats; *P < 0.05 vs. contralateral side by Student’s t-test after two-way ANOVA. (J) High-power image shows the difference in intensity of CapZ puncta between MML and IML of the ipsilateral DG. (K) Higher power images of MML of ipsilateral DG co-stained with pCofilin (red). Arrowheads indicate colocalized puncta of CapZ and pCofilin. Rightmost panel shows 3D confirmation of the colocalization for the puncta marked with an open arrowhead. Scale bar, 4 lm in (J) and 2 lm in (K).

might be associated with the neural network remodeling underlying higher cognitive processes such as learning and memory. Possible CapZ function in spine shape regulation

CapZ binds to the fast-growing barbed end of actin filaments (Caldwell et al. 1989; Schafer et al. 1993). Gelsolin, the other well-known capping protein, exhibits severing activities (Yin et al. 1981) and has been shown to destabilize F-actin in spines (Hayashi et al. 1996; Star et al. 2002). In contrast, CapZ has been found to exhibit only stabilizing activities, e.g., on F-actin in muscle cells. Thus, CapZ might preferentially function to stabilize F-actin in the CNS as well. Considering that the shape of dendritic spines is primarily regulated by F-actin (Ackermann & Matus 742

Genes to Cells (2010) 15, 737–747

2003), it is logical to suggest that CapZ may be involved in the regulation of spine shape. In a melanocytoma cell line, depletion of CapZ by short hairpin RNA caused loss of lamellipodia and an explosive increase in the number of filopodia (Mejillano et al. 2004). We therefore suggest that CapZ suppresses the formation of thin actin-fiber bundles and negatively regulates filopodia formation in neurons. In the present study, large amounts of CapZ were rarely found in thin filopodia (Figs 2,3). In some cases, CapZ was found at the tip of thin filopodia with small globular expansions (Fig. 2A, arrowhead at the bottom left), where it may inhibit further elongation and thereby promote spherical expansion at the tip to generate new spine heads. The CapZ content varied greatly among large spines, even those with similar head size and length

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Localization of CapZ in dendritic spines

(Figs 2,3), suggesting that this parameter is not the sole or critical determinant of spine shape. Therefore, the role(s) played by CapZ in some large dendritic spines is an interesting subject for future investigation. Typically, the F-actin in dendritic spines is in dynamic equilibrium with globular actin (G-actin). That is, the actin components are treadmilling from one end of the fiber to another, continually turning over. Synaptic inputs shift the F-actin ⁄ G-actin equilibrium toward F-actin predominance (Okamoto et al. 2004) and slow down the turnover rate of the fiber in an NMDA receptor-dependent manner (Star et al. 2002). Thereafter, stable F-actin appears in spines in association with enlargement of the spine head (Honkura et al. 2008). CapZ may contribute to the activity-regulated stabilization of actin architecture. The results obtained here are consistent with the above-mentioned hypothetical role of CapZ. We found that LTP-inducing electrical stimulation led to CapZ accumulation in the stimulated layer (MML) within 45 min and conversely that the CapZ content in spines decreased after a 60-min blockade of neuronal activity. This activity-dependent CapZ accumulation supports the possible role of CapZ in the regulation of actin dynamics in response to synaptic inputs. CapZ as a possible molecular marker of spine plasticity

The characteristic distribution of CapZ shown in this study indicates the heterogeneity of CapZ content in spines even with similar morphological characteristics. Interestingly, this variability might reflect the history of input given to the individual synapses. Our in vivo experiments involving an LTP-inducing stimulus suggest that the input-specific marking of synapses by CapZ accumulation occurs in a time scale of 30 min to 1 h. In turn, the clearance of CapZ accumulation by the overall suppression of neuronal firing observed in vitro also occurs in a similar time scale. The kinetics of CapZ accumulation and clearance is comparable to the previously reported time course of spine structural remodeling (Honkura et al. 2008), suggesting the role of CapZ in marking spines with previous synaptic inputs. Previously, we reported that, in only a limited number of spines in Arc-expressing cells (Kitanishi et al. 2009), exploring activity of an animal was accompanied by morphological alterations in its dendritic spines in the hippocampal neurons. In addition, we showed that learning enhances brain-derived neurotrophic factor expression (Chen et al. 2007a) in a limited number

of neurons. Together, these data imply that natural sensory and learning stimuli cause changes only in a limited number of synapses. To capture such a limited behavior-dependent synaptic plasticity in vivo, a highly specific molecular marker is necessary. Although 2 actin-related proteins, namely, profilin (Ackermann & Matus 2003; Lamprecht et al. 2006) and cofilin (Chen et al. 2007b; Fedulov et al. 2007), have been proposed to show activity-dependent accumulation or phosphorylation in spines, the input specificity of accumulation in vivo has not been proved so far. The findings of the present study show that CapZ is heterogeneously localized and shows input layerspecific accumulation as well as a response to increased or decreased neuronal activity. We suggest that CapZ serves as a useful marker for activitydependent spine plasticity and possibly for memoryrelated neuronal network reorganization.

Experimental procedures Unilateral FFT All experiments were performed according to the Japan Neuroscience Society guide for the care and use of laboratory animals. Wistar ⁄ ST male rats (aged 8–9 weeks; SLC, Shizuoka, Japan) were deeply anesthetized with pentobarbital (50 mg ⁄ kg i.p.; Dainippon Sumitomo Pharma, Osaka, Japan) and placed in a stereotaxic apparatus. Unilateral FFT was performed by vertical insertion of a coronally positioned razor blade (5.0mm wide, an end at the midline) through a lined skull burr hole into the right hemisphere (6.0-mm deep at a position 1.1 mm posterior to bregma) (Nakao et al. 2002). The rats after surgery were not tested in any behavioral paradigm. We have purposely left rats in novel environment (laboratory room) at least for 1 h before sampling.

2D-DIGE and LC-MS ⁄ MS (liquid chromatography with tandem mass spectrometry) analysis Ten days after unilateral FFT, each side of the hippocampus was rapidly dissected, separately frozen in liquid N2, powdered and homogenized in cold lysis buffer (7 M urea, 2 M thiourea, 4% CHAPS, 5 mM magnesium acetate, 2 mM sodium vanadate, 2 mM petabloc SC, 5 mM Tris ⁄ Cl (pH 8)). After centrifugation (20 min, 12 000 g, 4C), each supernatant from 10 samples was labeled with the indodicarbocyanine dye Cy5. As a standard, a mixed pool containing equal amounts of some samples (6 of the 10 samples; from three rats) labeled with the indocarbocyanine Cy3 was used to prevent labeling bias. Each Cy5-labeled sample and the Cy3-labeled standard (50 lg each) were mixed and applied to an 18-cm immobilized pH gradient strip (pH 4–7). Isoelectric focusing was carried out using Multiphor II according to the manufacturer’s protocol (GE

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Genes to Cells (2010) 15, 737–747

743

T Kitanishi et al. Healthcare Bioscience, Tokyo, Japan). Electrophoresis in the second dimension (molecular weight) was carried out with 12% SDS-PAGE gels. Gel images were collected by scanning with a 2920-2D Master Imager (GE Healthcare Bioscience). Quantitative statistical analysis was performed using DeCyder software (GE Healthcare Bioscience). Spots of differentially expressed proteins were excised from the 2D gels by using an Ettan Spot Picker (GE Healthcare Bioscience) and digested with trypsin (Promega, Mannheim, Germany). After digestion, the products were recovered from the gels by sequential extraction with 25 mM ammonium bicarbonate and 100% acetonitrile. The extracts were dried in a SpeedVac evaporator (Thermo Savant, Waltham, MA, USA) and resuspended in 10 lL of 0.1% TFA. The digested samples were injected onto a capillary high-performance liquid chromatography system equipped with a HP1100 solvent delivery pump (Agilent Technologies, Waldbronn, Germany), an Acurate flow splitter (LC Packings, Amsterdam, The Netherlands), a peptide trap column (0.5 · 5 mm; Michrom BioResources, Auburn, CA, USA) and a PepMap analytical column (0.075 · 150 mm; LC Packings). The extracted peptides were analyzed using an LCQ ion-trap mass spectrometer (ThermoFisher Scientific, San Jose, CA, USA) in the data-dependent MS ⁄ MS mode. The MASCOT program (Matrix Science, London, UK) was used for protein identification, and mammalian proteins in the NCBI nr database were used as references.

Primary culture and immunocytochemistry Cultured hippocampal neurons (DIV14) were prepared as described previously (Ohba et al. 2005) with minor modifications. Briefly, whole brains were isolated from embryonic day 18 Wistar rats (SLC); the hippocampi were removed and treated with 0.25% trypsin (Difco Laboratories, Detroit, MI, USA) and 0.01% deoxyribonuclease I (Sigma, St. Louis, MO, USA) at 37C for 30 min. The cells were suspended in Neurobasal medium (Invitrogen, Eugene, OR, USA) containing 10% fetal bovine serum (Sanko-junyaku, Tokyo, Japan) and were plated at a density of 4.5–5.0 · 104 cells ⁄ cm2 on polyethyleneimine (Sigma)-coated glass coverslips (Matsunami Glass, Osaka, Japan), each equipped with flexiPERM (Sartorius, Go¨ttingen, Germany), which created eight wells of dimensions 0.8 cm2 · 1.0 cm2. At 24 h after plating, the medium was changed to serum-free Neurobasal medium supplemented with 2% B27 (Invitrogen). Half of the medium was changed every 4 days. For immunocytochemical analysis, the cells were treated with 4% PFA at 4C for 30 min. Fluorescent images were acquired using an Eclipse TE 300 fluorescent microscope (Nikon, Tokyo, Japan) with a 100· magnification, 1.3 NA, oil immersion objective and an Orca II cooled CCD camera (Hamamatsu Photonics, Shizuoka, Japan) or an MRC-1000 confocal microscope (Carl Zeiss, Oberkochen, Germany) system equipped with a 60· magnification, 1.2 NA, water immersion objective and 488-nm argon and 543-nm helium ⁄ neon lasers. Images (16-bit) were acquired with a fixed expo-

744

Genes to Cells (2010) 15, 737–747

sure time that was determined to avoid saturation. At first, rhodamine-phalloidin (33 lM) images were processed with a convolution filter (Laplacian) to detect the edges of protrusions: those whose length exceeded 0.5 lm were objectively defined as dendritic spines. After the background was automatically subtracted from the CapZ beta 2 immunocytochemical images, the average intensity in each spine was measured.

Immunohistochemical analysis Male Thy1-mGFP mice [5 months old, line 21 (DePaola et al. 2003)], which express membrane-targeted green fluorescent protein (mGFP) in a subset of neurons, were used to visualize fine dendritic structures including spines. Anesthetized mice or rats were transcardially perfused with 4% PFA in 0.1 M PB. Dissected brains were soaked in fixative for 2 h. For antigen retrieval, Microslicer (Dosaka, Japan) sections of thickness 100 (Fig. 3A) or 50 (Fig. 3B) lm were treated with LAB solution (Polysciences, Warrington, PA, USA) for 5 min, 0.1% Triton X-100 for 30 min and 1 mg ⁄ mL of pepsin (DAKO, Carpinteria, CA, USA) in 0.2 N HCl at 37C for 5 min (Watanabe et al. 1998; Fukaya & Watanabe 2000). The sections were then sequentially incubated with 2% or 10% normal goat or donkey serum for 60 min, anti-CapZ beta 2 antibody (0.3 lg ⁄ mL; Developmental Studies Hybridoma Bank, University of Iowa, IA, USA) and rat monoclonal anti-GFP antibody [1:400; Nacalai Tesque, Japan, (Fig. 3A)] or anti-pCofilin (1:100, Abcam, Cambridge, UK) overnight at 4C, and finally with Alexa-594 anti-mouse IgG and Alexa-488 anti-rat IgG antibodies (1 : 400; Invitrogen) or Cy3-labeled donkey anti-mouse IgG (1 : 200; Jackson ImmunoResearch, West Grove, PA, USA). The sections were then mounted in Permafluor (Thermo Shandon, Pittsburgh, PA, USA). Fluorescent images were acquired using a confocal laser scanning microscope (LSM510 Meta, Carl Zeiss) equipped with a 63· magnification, 1.4 NA, oil immersion objective (FV1000; Olympus, Tokyo, Japan). The voxel size was set to 0.05 · 0.05 · 0.25 lm. For immunoelectron microscopic examination, the fixative was 4% PFA and 0.1% glutaraldehyde in 0.1 M PB. For postembedding immunogold processing, the slices were cryoprotected by immersion in 30% sucrose in 0.1 M PB and frozen rapidly in liquid propane in a Leica EM CPC unit. The frozen sections were immersed in 0.5% uranyl acetate in methanol at )90C in a Leica AFS freeze substitution unit, embedded at )45C in Lowicryl HM-20 resin (Lowi, Waldkraiburg, Germany) and polymerized with UV light. After etching with saturated sodium ethanolate solution for 3 s, ultra-thin sections on nickel grids were successively treated with 1% human serum albumin (Wako, Osaka, Japan) in 0.1% Tween 20 in Tris-buffered saline (pH 7.5) for 1 h, anti-CapZ antibody (15 lg ⁄ mL) overnight and colloidal gold (10 nm)conjugated anti-mouse IgG (1 : 100; British Biocell International, Cardiff, UK) for 2 h. Finally, the grids were stained with uranyl acetate for 15 min. Electron micrographs were taken randomly with an H7100 electron microscope (Hitachi, Tokyo, Japan).

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Localization of CapZ in dendritic spines

Dentate gyrus LTP and histochemical analysis LTP experiments on freely moving rats were carried out as described previously (Fukazawa et al. 2003). Briefly, a bipolar tungsten electrode (8.7 mm posterior, 5.3 mm lateral and 5.3 mm ventral to bregma) and a recording electrode (4.0 mm posterior, 2.5 mm lateral and 3.8 mm ventral to bregma) were unilaterally implanted into the rats under pentobarbital anesthesia (i.p., 50 mg ⁄ kg body weight). The rats were individually housed in cages and allowed to recover for at least 2 weeks. After the stability of the basal fEPSP was monitored, an HFS was applied; this stimulus consists of 500 pulses (10 trains of 5 bursts of ten 400-Hz pulses with a 1-s interburst interval and a 1-min intertrain interval) and has been shown to induce LTP. The rats were sacrificed under anesthesia at 30, 45, or 60 min after stimulation. The brain was immediately removed, frozen in dry ice and cut coronally with a cryostat microtome into 10- to 30-lm sections. These were fixed in 4% PFA in 0.1 M PB for 30 min. The subsequent histochemical procedures were as described for Fig. 3A. When F-actin was to be detected, the step involving LAB solution was omitted. The secondary antibody solution contained NeuroTrace 435 ⁄ 455 blue fluorescent Nissl stain (1:50; Invitrogen) to enable visualization of cell layers. Images acquired using an LSM 510 Meta microscope (Carl Zeiss) were analyzed using the MetaMorph software (Universal Imaging, West Chester, PA, USA).

Acknowledgements We thank Dr. Kazuto Nakao (National Institute of Health, USA) for providing technical advice, Dr. Yugo Fukazawa (National Institute for Physiological Sciences, Japan) for his useful discussions on LTP experiments, Dr. Shigeo Okabe for providing constructive comments on the manuscript, Drs. V. de Paola and P. Caroni for Thy1-mGFP mice and the Developmental Studies Hybridoma Bank (University of Iowa) for anti-CapZ beta 2 antibody. This work was supported by grants from PRESTO, Kato Memorial Bioscience Foundation and Grants-in-Aid KAKENHI (19590060 and 20019016 to MKY) from MEXT, Japan.

References Ackermann, M. & Matus, A. (2003) Activity-induced targeting of profilin and stabilization of dendritic spine morphology. Nat. Neurosci. 6, 1194–1200. Araya, R., Jiang, J., Eisenthal, K.B. & Yuste, R. (2006) The spine neck filters membrane potentials. Proc. Natl Acad. Sci. USA 103, 17961–17966. Bloodgood, B.L. & Sabatini, B.L. (2005) Neuronal activity regulates diffusion across the neck of dendritic spines. Science 310, 866–869. Caldwell, J.E., Heiss, S.G., Mermall, V. & Cooper, J.A. (1989) Effects of CapZ, an actin capping protein of muscle, on the polymerization of actin. Biochemistry 28, 8506–8514.

Casella, J.F., Craig, S.W., Maack, D.J. & Brown, A.E. (1987) Cap Z(36 ⁄ 32), a barbed end actin-capping protein, is a component of the Z-line of skeletal muscle. J. Cell Biol. 105, 371–379. Cassel, J.C., Duconseille, E., Jeltsch, H. & Will, B. (1997) The fimbria-fornix ⁄ cingular bundle pathways: a review of neurochemical and behavioural approaches using lesions and transplantation techniques. Prog. Neurobiol. 51, 663– 716. Chen, J., Kitanishi, T., Ikeda, T., Matsuki, N. & Yamada, M.K. (2007a) Contextual learning induces an increase in the number of hippocampal CA1 neurons expressing high levels of BDNF. Neurobiol. Learn. Mem. 88, 409–415. Chen, L.Y., Rex, C.S., Casale, M.S., Gall, C.M. & Lynch, G. (2007b) Changes in synaptic morphology accompany actin signaling during LTP. J. Neurosci. 27, 5363–5372. Davis, D.A., Wilson, M.H., Giraud, J., Xie, Z., Tseng, H.C., England, C., Herscovitz, H., Tsai, L.H. & Delalle, I. (2009) Capzb2 interacts with beta-tubulin to regulate growth cone morphology and neurite outgrowth. PLoS Biol. 7, e1000208. DePaola, V., Arber, S. & Caroni, P. (2003) AMPA receptors regulate dynamic equilibrium of presynaptic terminals in mature hippocampal networks. Nat. Neurosci. 6, 491– 500. Engert, F. & Bonhoeffer, T. (1999) Dendritic spine changes associated with hippocampal long-term synaptic plasticity. Nature 399, 66–70. Fedulov, V., Rex, C.S., Simmons, D.A., Palmer, L., Gall, C.M. & Lynch, G. (2007) Evidence that long-term potentiation occurs within individual hippocampal synapses during learning. J. Neurosci. 27, 8031–8039. Fukaya, M. & Watanabe, M. (2000) Improved immunohistochemical detection of postsynaptically located PSD95 ⁄ SAP90 protein family by protease section pretreatment: a study in the adult mouse brain. J. Comp. Neurol. 426, 572– 586. Fukazawa, Y., Saitoh, Y., Ozawa, F., Ohta, Y., Mizuno, K. & Inokuchi, K. (2003) Hippocampal LTP is accompanied by enhanced F-actin content within the dendritic spine that is essential for late LTP maintenance in vivo. Neuron 38, 447– 460. Gaffan, D. (1972) Loss of Recognition Memory in Rats with Lesions of Fornix. Neuropsychologia 10, 327–341. Gulesserian, T., Kim, S.H., Fountoulakis, M. & Lubec, G. (2002) Aberrant expression of centractin and capping proteins, integral constituents of the dynactin complex, in fetal down syndrome brain. Biochem. Biophys. Res. Commun. 291, 62–67. Harris, K.M. & Stevens, J.K. (1989) Dendritic spines of Ca-1 pyramidal cells in the rat hippocampus – serial electronmicroscopy with reference to their biophysical characteristics. J. Neurosci. 9, 2982–2997. Hart, M.C. & Cooper, J.A. (1999) Vertebrate isoforms of actin capping protein beta have distinct functions in vivo. J. Cell Biol. 147, 1287–1298.

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Genes to Cells (2010) 15, 737–747

745

T Kitanishi et al. Hayashi, K., Ishikawa, R., Ye, L.H., He, X.L., Takata, K., Kohama, K. & Shirao, T. (1996) Modulatory role of drebrin on the cytoskeleton within dendritic spines in the rat cerebral cortex. J. Neurosci. 16, 7161–7170. Hoe, H.S., Tran, T.S., Matsuoka, Y., Howell, B.W. & Rebeck, G.W. (2006) DAB1 and Reelin effects on amyloid precursor protein and ApoE receptor 2 trafficking and processing. J. Biol. Chem. 281, 35176–35185. Honkura, N., Matsuzaki, M., Noguchi, J., Ellis-Davies, G.C.R. & Kasai, H. (2008) The subspine organization of actin fibers regulates the structure and plasticity of dendritic spines. Neuron 57, 719–729. Huang, M., Yang, C., Schafer, D.A., Cooper, J.A., Higgs, H.N. & Zigmond, S.H. (1999) Cdc42-induced actin filaments are protected from capping protein. Curr. Biol. 9, 979–982. Jordan, B.A., Fernholz, B.D., Boussac, M., Xu, C., Grigorean, G., Ziff, E.B. & Neubert, T.A. (2004) Identification and verification of novel rodent postsynaptic density proteins. Mol. Cell Proteomics 3, 857–871. Kitanishi, T., Ikegaya, Y., Matsuki, N. & Yamada, M.K. (2009) Experience-dependent, rapid structural changes in hippocampal pyramidal cell spines. Cereb. Cortex 19, 2572– 2578. Lamprecht, R., Farb, C.R., Rodrigues, S.M. & LeDoux, J.E. (2006) Fear conditioning drives profilin into amygdala dendritic spines. Nat. Neurosci. 9, 481–483. Li, K.W., Hornshaw, M.P., Van Der Schors, R.C., Watson, R., Tate, S., Casetta, B., Jimenez, C.R., Gouwenberg, Y., Gundelfinger, E.D., Smalla, K.H. & Smit, A.B. (2004) Proteomics analysis of rat brain postsynaptic density. Implications of the diverse protein functional groups for the integration of synaptic physiology. J. Biol. Chem. 279, 987– 1002. Matsuzaki, M., Ellis-Davies, G.C., Nemoto, T., Miyashita, Y., Iino, M. & Kasai, H. (2001) Dendritic spine geometry is critical for AMPA receptor expression in hippocampal CA1 pyramidal neurons. Nat. Neurosci. 4, 1086–1092. Matsuzaki, M., Honkura, N., Ellis-Davies, G.C. & Kasai, H. (2004) Structural basis of long-term potentiation in single dendritic spines. Nature 429, 761–766. Matus, A., Brinkhaus, H. & Wagner, U. (2000) Actin dynamics in dendritic spines: a form of regulated plasticity at excitatory synapses. Hippocampus 10, 555–560. Mejillano, M.R., Kojima, S., Applewhite, D.A., Gertler, F.B., Svitkina, T.M. & Borisy, G.G. (2004) Lamellipodial versus filopodial mode of the actin nanomachinery: pivotal role of the filament barbed end. Cell 118, 363–373. Nagerl, U.V., Eberhorn, N., Cambridge, S.B. & Bonhoeffer, T. (2004) Bidirectional activity-dependent morphological plasticity in hippocampal neurons. Neuron 44, 759–767. Nakao, K., Ikegaya, Y., Yamada, M.K., Nishiyama, N. & Matsuki, N. (2001) Spatial performance correlates with long-term potentiation of the dentate gyrus but not of the CA1 region in rats with fimbria-fornix lesions. Neurosci. Lett. 307, 159–162.

746

Genes to Cells (2010) 15, 737–747

Nakao, K., Ikegaya, Y., Yamada, M.K., Nishiyama, N. & Matsuki, N. (2002) Hippocampal long-term depression as an index of spatial working memory. Eur. J. Neurosci. 16, 970–974. Noguchi, J., Matsuzaki, M., Ellis-Davies, G.C. & Kasai, H. (2005) Spine-neck geometry determines NMDA receptordependent Ca2+ signaling in dendrites. Neuron 46, 609– 622. Oddie, S.D., Kirk, I.J., Gorny, B.P., Whishaw, I.Q. & Bland, B.H. (2002) Impaired dodging in food-conflict following fimbriafornix transection in rats: A novel hippocampal formation deficit. Brain Res. Bull. 57, 565–573. Ohba, S., Ikeda, T., Ikegaya, Y., Nishiyama, N., Matsuki, N. & Yamada, M.K. (2005) BDNF locally potentiates GABAergic presynaptic machineries: Target-selective circuit inhibition. Cereb. Cortex 15, 291–298. Okamoto, K., Nagai, T., Miyawaki, A. & Hayashi, Y. (2004) Rapid and persistent modulation of actin dynamics regulates postsynaptic reorganization underlying bidirectional plasticity. Nat. Neurosci. 7, 1104–1112. Peng, J., Kim, M.J., Cheng, D., Duong, D.M., Gygi, S.P. & Sheng, M. (2004) Semiquantitative proteomic analysis of rat forebrain postsynaptic density fractions by mass spectrometry. J. Biol. Chem. 279, 21003–21011. Sato, Y., Taoka, M., Sugiyama, N., Kubo, K., Fuchigami, T., Asada, A., Saito, T., Nakajima, K., Isobe, T. & Hisanaga, S. (2007) Regulation of the interaction of Disabled-1 with CIN85 by phosphorylation with Cyclin-dependent kinase 5. Genes Cells 12, 1315–1327. Schafer, D.A. & Cooper, J.A. (1995) Control of actin assembly at filament ends. Annu. Rev. Cell Dev. Biol. 11, 497–518. Schafer, D.A., Waddle, J.A. & Cooper, J.A. (1993) Localization of CapZ during myofibrillogenesis in cultured chicken muscle. Cell Motil. Cytoskeleton 25, 317–335. Star, E.N., Kwiatkowski, D.J. & Murthy, V.N. (2002) Rapid turnover of actin in dendritic spines and its regulation by activity. Nat. Neurosci. 5, 239–246. Vann, S.D., Brown, M.W., Erichsen, J.T. & Aggleton, J.P. (2000) Using Fos imaging in the rat to reveal the anatomical extent of the disruptive effects of fornix lesions. J. Neurosci. 20, 8144–8152. Watanabe, M., Fukaya, M., Sakimura, K., Manabe, T., Mishina, M. & Inoue, Y. (1998) Selective scarcity of NMDA receptor channel subunits in the stratum lucidum (mossy fibre-recipient layer) of the mouse hippocampal CA3 subfield. Eur. J. Neurosci. 10, 478–487. Yin, H.L., Hartwig, J.H., Maruyama, K. & Stossel, T.P. (1981) Ca2+ control of actin filament length. Effects of macrophage gelsolin on actin polymerization. J. Biol. Chem. 256, 9693–9697. Yoshimura, Y., Yamauchi, Y., Shinkawa, T., Taoka, M., Donai, H., Takahashi, N., Isobe, T. & Yamauchi, T. (2004) Molecular constituents of the postsynaptic density fraction revealed by proteomic analysis using multidimensional liquid chromatography-tandem mass spectrometry. J. Neurochem. 88, 759–768.

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Localization of CapZ in dendritic spines Zhou, Q., Homma, K.J. & Poo, M.M. (2004) Shrinkage of dendritic spines associated with long-term depression of hippocampal synapses. Neuron 44, 749–757. Received: 20 December 2009 Accepted: 28 March 2010

Supporting Information/Supplementary material The following Supporting Information can be found in the online version of the article:

Figure S1 Immunoblot for CapZ in 1D (A, B) and 2D (C) SDS-PAGE. Figure S2 Localization and dynamics of EGFP-tagged CapZ in cultured hippocampal neurons. Additional Supporting Information may be found in the online version of this article. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article

 2010 The Authors Journal compilation  2010 by the Molecular Biology Society of Japan/Blackwell Publishing Ltd.

Genes to Cells (2010) 15, 737–747

747