V-1, a protein expressed transiently during murine cerebellar development, regulates actin polymerization via interaction with capping protein.

V-1 is a 12-kDa protein consisting of three consecutive ANK repeats, which are believed to serve as the surface for protein-protein interactions. It is thought to have a role in neural development for its temporal profile of expression during murine cerebellar development, but its precise role remains unknown. Here we applied the proteomic approach to search for protein targets that interact with V-1. The V-1 cDNA attached with a tandem affinity purification tag was expressed in the cultured 293T cells, and the protein complex formed within the cells were captured and characterized by mass spectrometry. We detected two polypeptides specifically associated with V-1, which were identified as the alpha and beta subunits of the capping protein (CP, alternatively called CapZ or beta-actinin). CP regulates actin polymerization by capping the barbed end of the actin filament. The V-1.CP complex was detected not only in cultured cells transfected with the V-1 cDNA but also endogenously in cells as well as in murine cerebellar extracts. An analysis of the V-1/CP interaction by surface plasmon resonance spectroscopy showed that V-1 formed a stable complex with the CP heterodimer with a dissociation constant of 1.2 x 10(-7) m and a molecular stoichiometry of approximately 1:1. In addition, V-1 inhibited the CP-regulated actin polymerization in vitro in a dose-dependent manner. Thus, our results suggest that V-1 is a novel component that regulates the dynamics of actin polymerization by interacting with CP and thereby participates in a variety of cellular processes such as actin-driven cell movements and motility during neuronal development.

The V-1 protein was originally identified in the murine cerebellum as one of the proteins expressed significantly at the initial stage of postnatal development (1), particularly in the regions where synaptic formation and neuronal migration occur actively during neurogenesis (2,3). V-1 consists of 117 amino acids containing three contiguous repeats of the ANK motif, alternatively called the cdc10/SWI6 motif (1,2), which is crucial for a large number of protein-protein interactions (4). Previous studies suggested the potential role of V-1 in the signal transduction pathways leading to catecholamine synthesis or to cardiac hypertrophy. For example, Yamakuni et al. (5)(6)(7) demonstrate that the overexpression of V-1 caused a significant increase in the catecholamine level in PC12 cells, presumably through the transcriptional activation of the genes for catecholamine synthesis. In other reports (8 -11), V-1 was designated as "myotrophin" and was shown to participate in the cell signaling pathways for the NFkB-mediated activation of protein synthesis in the myocytes. Thus, both of these studies (5,9) suggested the roles of V-1 in the biological events taking place in the nucleus. However, no biological function has been attributed to V-1 in the cytoplasm in which this molecule predominantly resides within the cells and tissues (5,9).
In this study, we screened for V-1-binding proteins by a novel proteomic approach that combined the tandem affinity purification (TAP) 1 procedure (12) and mass spectrometry (MS). Following this strategy, we identified capping protein (CP) as a V-1-binding protein. We confirmed the existence of the V-1⅐CP complex not only in cultured cells transfected with the TAPtagged V-1 but also endogenously in cells and rat cerebellar extracts. Furthermore, we found that V-1 inhibited the CPregulated actin polymerization. On the basis of these results, the possible role of V-1 in neuronal development is discussed.

EXPERIMENTAL PROCEDURES
Materials-Actin monomer was prepared from the acetone-dried powder of rabbit skeletal muscle according to the procedure of Spudich and Watt (13). Pyrene-labeled actin was purchased from Cytoskeleton Inc. (Denver, CO). The polyclonal antibody against the V-1 protein was raised in rabbits injected with the recombinant V-1 protein. The antibody was purified by ammonium sulfate fractionation (0 -50% saturation) followed by affinity chromatography on Sepharose beads coupled with the recombinant V-1 protein. The anti-CP monoclonal antibody (14) was obtained from the Developmental Studies Hybridoma Bank, developed under the auspices of the National Institute of Child Health and Human Development, and maintained by The University of Iowa, Department of Biological Sciences. Oligonucleotides were purchased from Sigma.
Plasmid Constructions-To construct the cDNA for TAP-tagged V-1, the TAP cDNA (pBS1479) (12) was first digested with BamHI and HindIII and was inserted into the BamHI and EcoRI sites of the mammalian expression vector pcDNA3 (Invitrogen) after blunting the HindIII and EcoRI sites (termed pcDNA3-TAP). The V-1 cDNA was then generated by PCR using the oligonucleotides 5Ј-GCGAAGCTTA-TGTGCGACAAGGAGTTCAT-3Ј and 5Ј-CTGGATCCCTGGAGAAGAG-CTTTGATTG-3Ј and the rat V-1 cDNA (2). The PCR fragment was digested with HindIII and BamHI and was inserted into the cloning site of pcDNA3-TAP. The cDNA for GST-tagged V-1 was constructed by PCR amplification of the V-1 cDNA using the oligonucleotides 5Ј-TAGGATCCTGTGCGA-CAAGGAGTTCATG-3Ј and 5Ј-GCGAATTCTCACTGGAGAAGAGCTT-TGA-3Ј and the rat V-1 cDNA. The PCR fragment was digested with BamHI and EcoRI and was inserted into the cloning sites of the bacterial expression vector pGEX-3X (Amersham Biosciences). The V-1 cDNA was also amplified by PCR using the oligonucleotides 5Ј-TACC-ATGGGATGTGATAAAGAGTTCATG-3Ј and 5Ј-GTTCATGGATCCAT-CACTGGAGAAGAGC-3Ј and the rat V-1 cDNA. The resulting fragment was digested with NdeI and BamHI and was inserted into the cloning sites of the bacterial expression vector, pET 15b (Novagen, Madison, WI).
Cell Culture and Immunoprecipitation-Human embryonic kidney 293 T cells (293T) were maintained in Dulbecco's modified Eagle's medium containing 10% fetal calf serum (Invitrogen) as described previously (15). The cells (2 ϫ 10 6 cells) were cultured in a 100-mm dish and were transfected the following day with 8 g of cDNA using the Polyfect transfection reagent (Qiagen, Hilden, Germany) according to the manufacturer's instructions. After 48 h, the cells were washed twice with phosphate-buffered saline and were immediately scraped into 500 l of lysis buffer containing 50 mM Tris-HCl (pH 8.0), 10% glycerol, 1% Triton X-100, 150 mM NaCl, 100 mM NaF, 5 M ZnCl 2 , 1 mM Na 3 VO 4 , 10 mM EGTA, 2 g/ml leupeptin, and 4 mM phenylmethylsulfonyl fluoride. The cell lysate was centrifuged at 100,000 ϫ g for 20 min at 4°C, and the supernatant (ϳ5-mg protein) was incubated with 20 l of IgG-Sepharose beads (Amersham Biosciences) for 2 h at 4°C. The IgG beads were washed five times with 500 l of a buffer containing 20 mM Tris-HCl (pH 7.5), 5% glycerol, 0.1% Triton X-100, and 150 mM NaCl and twice with 50 mM Tris-HCl (pH 8.0). The beads were then mixed with tobacco etch virus protease (Invitrogen) at room temperature for 1 h in 50 mM Tris-HCl (pH8.0) to release the protein complex into solution. The complex was subsequently incubated with 20 l of calmodulin-agarose beads (Stratagene, La Jolla, CA) in a buffer containing 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, and 1 mM CaCl 2 . After washing the beads twice with the same buffer, the bound proteins were eluted from the beads with this buffer containing 4 mM EDTA instead of 1 mM CaCl 2 .
In-gel Digestion and Tandem Mass Spectrometry-Protein bands (0.2 ϫ 0.8 cm) were excised from the SDS gel, dehydrated with acetonitrile, and vacuum-dried. The gel pieces were swollen in 10 l of 100 mM Tris-HCl (pH 8.8), and the proteins were in-gel digested overnight with 250 ng of trypsin (Promega, Madison, WI) at 37°C. The peptide fragments were extracted from the gel with 100 l of 50% acetonitrile in 5% formic acid and were concentrated by adsorption onto 0.1 l of reversed-phase beads (POROS R2, Applied Biosystems, Foster, CA). After the beads were washed twice with 400 l of 0.1% trifluoroacetic acid, the bound peptides were recovered with 2 l of 50% methanol in 5% acetic acid. The peptide mixture was then analyzed by tandem mass spectrometry on a electrospray ionization quadrupole time-of-flight mass spectrometer (Q-TOF2, Micromass, Manchester, United Kingdom) equipped with a nanospray tip. All of the MS/MS spectra were searched against the non-redundant protein sequence data base maintained at the National Center for Biotechnology Information to identify proteins by the Mascot program (Matrixscience, London, United Kingdom). The MS/MS signal assignments were confirmed manually.
Expression and Purification of Recombinant Proteins-Recombinant V-1 protein was expressed in Escherichia coli (BL21) and was purified by ammonium sulfate fractionation (50 -85% saturation) followed by anion exchange chromatography on a DEAE-5PW column (0.75 ϫ 15 cm, TOSOH, Tokyo, Japan). The GST-tagged V-1 protein (GST-V-1) was purified with glutathione-Sepharose beads (Amersham Biosciences) as described previously (16). The CP heterodimer consisting of ␣ and ␤ subunits was produced in E. coli (BL21) using the simultaneous expression system as described by Soeno et al. (17). For the experiment shown in Fig. 3, A and B, the ␣ and ␤ subunits were each expressed separately in E. coli (BL21). Because the expressed ␣ and ␤ subunits were each recovered in the insoluble fraction, the precipitates were collected and dissolved in buffer A (2 mM dithiothreitol, 50 mM Tris-HCl (pH 7.5)) containing 8 M urea. The concentration of urea was then decreased to 0.06 M by a two-step dilution. First, the solution was diluted with buffer A containing 6 M urea to a protein concentration of 200 nmol protein/ml, and the insoluble material was removed by centrifugation. The soluble fraction then was diluted 100-fold with the surface plasmon resonance (SPR) buffer (150 mM NaCl, 3 mM EDTA, 0.005% Polysorbate 20, 10 mM HEPES (pH 7.4)). For the experiment shown in Fig. 3B, the ␣ and ␤ subunits were mixed and reconstituted to the heterodimer at 4°C for 10 min.
Surface Plasmon Resonance Analysis-All of the studies were performed on a Biacore 2000 instrument (Biacore, Uppsala, Sweden). The GST or GST-fused V-1 protein was immobilized on the CM5 sensor chip via an anti-GST antibody (Biacore). The concentration of the immobilized protein was adjusted to yield ϳ180 resonance units. The V-1/CP interaction analysis was carried out at 25°C by introducing CP␣, CP␤, or the reconstituted CP heterodimer (1 M) at a flow rate of 20 l/min into the Biacore instrument, and the changes in the SPR signals were recorded. For the kinetic analysis, various concentrations of the CP heterodimer were loaded on the sensor surface to equilibrate the V-1/CP interaction, and the concentration of the complex (Req) was measured as the response units. The correlation among Req, the concentration of ligand (C) passed over the sensor surface, and the total binding capacity (R max ) of the immobilized protein was: Req/C ϭ K a R max Ϫ K a Req (18). Thus, the association constant (K a ) was determined from the plot of Req/C versus Req estimated at different CP concentrations by a Scatchard plot analysis.
Assay of CP-mediated Actin Nucleation and Depolymerization-The CP-mediated actin nucleation and depolymerization were analyzed essentially according to the procedures described by Higashi and Oosawa (19) and Schafer et al. (14), respectively. For the actin nucleation assay The polymerized mixture was centrifuged for 30 min at 100,000 ϫ g, and the precipitates were analyzed by SDS-PAGE (CBB staining). SDS-PAGE and Western blotting were carried out as described previously (20). Silver staining was performed without glutaraldehyde (21).

Identification of CP as a V-1-binding Protein-To search for
the V-1-binding protein, we carried out an affinity purification experiment using the TAP method (12). The TAP method utilizes two affinity modules (a calmodulin-binding peptide and a protein A epitope) separated by a cleavage site for the tobacco etch virus protease. Therefore, the purification produces much less background than the conventional single epitope tags such as Myc or hemagglutinin. To adapt this method for our purpose, we constructed a mammalian expression vector encoding V-1 with a carboxyl-terminal TAP tag, transiently transformed 293T cells with the vector, and purified the tagged V-1 with its binding proteins from the cell lysate by the TAP method. Fig.  1A, lane 2, shows the SDS-PAGE photographs of the purified protein complex. In addition to the tagged V-1 used as bait, the 36-and 33-kDa polypeptide bands were reproducibly detected in the cells transfected with a TAP-tagged V-1 but not in the control cells (Fig. 1A, lane 1), suggesting that these polypeptides were associated with the expressed V-1 protein. These bands (assigned as bands 1 and 2 in Fig. 1A) were excised from the gel and subjected to in-gel tryptic digestion, and the peptide fragments thus generated were analyzed by nanoelectrospray tandem mass spectrometry, respectively.
Two doubly charged peptide ions with m/z 599.40 and 786.00, were observed from band 1. The data base analysis of the MS/MS spectrum of the peptide ion with m/z 599.40 showed that it corresponded to the sequence LLLNNDNLLR of CP␣ (GenBank TM accession number P52907) at residues 38 -47. The manual assignment of the fragment ions also yielded the same sequence (Fig. 1B). Likewise, the MS/MS spectrum of the other peptide ion with m/z 786.00 was assigned to the sequence FTITPPTAQVVGVLK of CP␣ at residues 179 -193 (data not shown), confirming that band 1 contained CP␣. From band 2, two doubly charged peptide ions with m/z 586.20 and 677.30 were observed. The analysis of their collision-induced dissociation fragments indicated that one (m/z 586.20) corresponded to the sequence STLNEIYFGK of CP␤ (GenBank TM accession number P47756) at residues 226 -235 (Fig. 1C) and the other (m/z 677.30) corresponded to the sequence SGSGTMN-LGGSLTR of the same polypeptide at residues 182-195 (data not shown). The molecular weights of bands 1 and 2 estimated by SDS-PAGE (Fig. 1A) were also agreed with those of the respective CP subunits (32, 902 and 31,331). Because CP is a stable heterodimer of ␣ and ␤ subunits (22,23) and because these polypeptides appear in almost equimolar amounts in the captured V-1 complex (Fig. 1A), we anticipated that V-1 bound the CP heterodimer itself.
Detection of the Endogenous V-1⅐CP Complex-The overex-pression of a particular protein sometimes induces artificial interactions among proteins, which do not occur under physiological conditions. To exclude the possibility that the observed interaction between V-1 and CP might be artificial, we sought to detect the endogenous V-1⅐CP complex in cultured 293T cells as well as in rat cerebella. Soluble extracts were prepared from 293T cells or from rat cerebella at a postnatal day 12, and the V-1 protein was immunoprecipitated from the extracts with the anti-V-1 IgG immobilized on Sepharose beads. The precipitate was then analyzed by SDS-PAGE followed by immunoblotting with the antibody against the CP␤ subunit. As shown in Fig. 2, the CP␤ was clearly detected in the precipitates derived both  from the 293T cells and the rat cerebella, suggesting that the V-1⅐CP complex was present under physiological conditions in vivo.
Biochemical Properties of the Interaction between V-1 and CP-To study the biochemical characteristics of the V-1⅐CP complex in further detail, we prepared a number of materials including V-1, GST-tagged V-1, and the CP␣ and ␤ subunits. These proteins were expressed in the E. coli cells and were purified to near homogeneity as observed by SDS-PAGE (Fig.  3A). First, we studied the complex formation of V-1 and CP by SPR. For the direct binding assay on the SPR biosensor, the GST-V-1 protein was attached to the sensor surface via the GST antibody, and the recombinant CP was passed over the sensor chip. As shown in Fig. 3B, positive binding signals were detected when the CP heterodimer, reconstituted from the recombinant CP␣ and ␤ subunits, was introduced to the sensor chip. Interestingly, however, the purified CP␣ or CP␤ alone gave no SPR signals (Fig. 3B) even with repeated binding experiments using different subunit preparations, suggesting that V-1 bound specifically to the heterodimeric CP molecule. To obtain kinetic data of the complex formation, the steadystate resonance was measured using various concentrations of CP (Fig. 3C). The results showed that the immobilized V-1 was saturable with respect to the CP binding and the association constant calculated from the slope of the straight line of the corresponding Scatchard plot (inset) was 8.4 ϫ 10 6 M. The K a value was not influenced by the amount of immobilized V-1 protein (data not shown). The equilibrium dissociation constant (K d ) calculated from 1/K a was 1.2 ϫ 10 Ϫ7 M, indicating that the V-1⅐CP complex is relatively stable.
To clarify the formation of the V-1⅐CP complex and to determine the molecular stoichiometry, the "native-PAGE assay" (24) was performed (Fig. 3D). When the recombinant V-1 was mixed with CP under physiological conditions and was analyzed by gel electrophoresis without chaotropic reagents, a new protein band appeared that migrated between the V-1 and CP heterodimer (Fig. 3D, left panel). The subsequent analysis of this protein band by SDS-PAGE revealed that it was a V-1⅐CP complex, because it contained V-1 and the CP␣ and ␤ subunits (Fig. 3D, right panel). The semi-quantitative analysis of the CBB-stained bands indicated that the density of each polypeptide band was 1.0:0.8:1.0 for CP␣, ␤, and V-1, respectively. This finding suggests that the binding stoichiometry of the V-1⅐CPheterodimer complex is ϳ1:1, i.e. one V-1 molecule associates with one CP heterodimer. V-1 exists in the mouse cerebellum at a concentration of 3.7 ϫ 10 Ϫ7 M as estimated by quantitative immunoblotting, 2 and this concentration is almost the same level as that of CP (2 Х 20 ϫ 10 Ϫ7 M) reported in previous studies (14,32,33). Thus, the V-1/CP interaction could occur under physiological condition in terms of cellular levels of V-1 and CP and the binding kinetics of these proteins.
Suppression of CP Activity by V-1-CP nucleates actin polymerization and caps the barbed ends of actin filaments (14,23,25,26). To examine the effects of V-1 protein on the activities of CP, we assayed the initial rates of polymerization with the CP-nucleated actin and depolymerization of the CP-capped F-actin in the presence and absence of the V-1 protein. For the CP-nucleated actin polymerization, we monitored the changes in optical absorption to measure the initial rate of actin polymerization. As shown in Fig. 4Aa, CP markedly increased the initial rate of actin polymerization and the V-1 protein canceled this CP function. The difference in the initial rate of polymerization between CP-nucleated actin in the presence and absence of the V-1 protein was statistically significant (Fig. 4Ab). The V-1 protein reduced the nucleation activity of CP in a dose-dependent manner, whereas V-1 alone had no effect on the actin polymerization (data not shown). Besides this activity, we also measured the effects of V-1 on the dilution-induced depolymerization of the CP-capped actin filaments by monitoring the changes in fluorescence (Fig. 4B). After the preincubation with V-1, CP reduced its capping activity to depolymerize the actin filaments. Thus, the capping activity of CP was dependent on the concentration of V-1, whereas V-1 alone had no effect on the actin depolymerization. These results demonstrate that the V-1 protein suppressed the CP activities of both actin nucleation and F-actin capping.
Interaction between the V-1⅐CP Complex and F-actin-We studied whether the V-1⅐CP complex bound F-actin by using a co-sedimentation assay. The actin monomer was polymerized and co-sedimented with CP in the presence or absence of the V-1 protein, and aliquots of the precipitates were analyzed by SDS gel electrophoresis. As shown in Fig. 5, CP co-sedimented with F-actin only in the absence of the V-1 protein. Thus, V-1 appeared to form a stable complex with CP and to prevent CP binding to F-actin. DISCUSSION CP is one of the F-actin-binding proteins that caps the barbed end of actin filaments and nucleates the actin polymerization in a Ca 2ϩ -independent (22, 23, 27, 28) and a phosphatidylinositol 4,5-bisphosphate-dependent manner (29 -32). This activity is thought to be functionally significant, because the actin-based movement of Dictyostelium is proportional to the expression level of CP (33) and because CP is essential for the in vitro reconstitution of the cell movement (34,35). In this study, we have shown that the V-1 protein forms a stable stoichiometric complex with CP in vitro as well as in vivo (Figs. 2 and 3) and inhibits the CP-mediated nucleation of actin polymerization. Therefore, we assume that V-1 participates in the regulation of actin dynamics in the cells via the interaction with CP.
Our strategy to identify the V-1-interacting molecules was based on the tandem affinity purification of the V-1 complex with a TAP tag followed by protein identification by mass spectrometry. The TAP method was originally developed to analyze interactions among yeast proteins (12). We constructed a mammalian expression vector for the TAP method and applied it to the mammalian 293T cell line. Even though the V-1 protein is a rather minor cellular component and a transient expression system was used for the assay, the method enabled us to isolate a sufficient amount of the V-1⅐CP complex for characterization by nanospray tandem mass spectrometry (Fig.  1). This affinity-tag technique coupled with mass spectrometry is useful to detect novel protein interactions not only in yeast but also in mammalian cells.
The V-1 protein consists of three consecutive ANK repeats with an additional short stretch of sequence (1,2,37). The ANK repeat is a structural motif found in many proteins (36) and mediates specific interactions with a diverse array of protein targets (4). In the tertiary structure of V-1 determined by NMR spectroscopy (37), the ANK repeats comprise the hairpin-helixloop-helix modules where the ␣ helices lie along one side providing a structural framework and the hairpins protrude on the other side of the molecule. The hairpins and the surface of the ␣ helices form a groove-like structure, which is believed to be responsible for the contact with the target molecule. This study identified CP as a potential target of V-1. Interestingly, V-1 bound to the functional CP heterodimer consisting of the ␣ and ␤ subunits but did not bind each of the two subunits. This finding is comparable with previous observations that each of the CP subunits was unstable and did not bind actin in vitro and in vivo (26,38,39). Thus, it seems likely that V-1 recognizes the structural interface of the CP heterodimer by its groove-like ANK repeats and covers the F-actin binding surface located in the carboxyl-terminal region of the CP␤ subunit (26). However, whether this is the molecular mechanism by which V-1 inhibits the interaction of CP and F-actin awaits further structural investigation.
Two distinct proteins, carmil (40) and twinfilin (24), are known to bind CP. Carmil is a scaffold protein containing the CP-binding site, the myosin I-binding site, the short sequence commonly found in several actin monomer-binding proteins, and the acidic stretch that can activate Arp2/3-dependent actin nucleation (40). The NH 2 -terminal region of carmil binds CP, but its activity with CP is unknown. Twinfilin is a ubiquitous actin monomer-binding protein composed of two ADF/cofilinlike domains connected by a short linker region and has a role in the regulation of actin turnover (41). Twinfilin also forms a stable complex with CP but does not affect its activity. V-1 is neither a scaffold protein nor an actin-binding protein, and it lacks structural homology to either carmil or twinfilin. Thus, V-1 belongs to a novel CP-binding protein category. Recently, Ena/VASP was reported as an anti-capping molecule, which promotes actin filament elongation by associating with the barbed ends of actin and shielding them from CP (42). The actin cytoskeleton in the Ena/VASP-deficient cell contained shorter, more highly branched filaments than those in the control cells. V-1 resembles Ena/VASP in the activity of CPregulated actin polymerization, but whether a similar phenotype can be attributed to the V-1-deficient cell is currently unknown.
Previous studies suggested the potential roles of V-1 in nuclear events such as the transcriptional activation of a set of enzymes involved in catecholamine synthesis (5)(6)(7) or the regulation of de novo protein synthesis (8 -10, 43). This study suggests for the first time that V-1 functions in the molecular events taking place in the cytoplasm in which the major portion of V-1 resides within the cells as revealed by previous immunohistochemical studies (5,9). V-1 also appeared to be a typical cytoplasmic protein in terms of amino acid sequence with no apparent nuclear localization signal. Our previous studies revealed the characteristic temporal profile of V-1 expression in the developing murine cerebellum at postnatal days 7-12 (2). Namely, V-1 expression is particularly significant during the migration of progenitor granule cells from the external to internal granular layer to make synaptic contacts with the target Purkinje cells. Likewise, the dynamics of actin polymerization play a pivotal role in the maturation of granule cells, because the modulation of the barbed ends of actin filaments with cytochalasins changed the behavior of the growth cone (44) and the migration of granule cells (45). These observations coupled with the results reported here suggest that V-1 may have a role in the CP-mediated actin-driven cell movements and motility such as granular cell migration and synapse formation.