Intracellular Ca2+-dependent cellular responses are often mediated by the ubiquitous protein CaM (calmodulin), which, upon binding Ca2+, can interact with and alter the function of numerous proteins. In the present study, using a newly developed functional proteomic screen of rat brain extracts, we identified PRG-1 (plasticity-related gene-1) as a novel CaM target. A CaM-overlay and an immunoprecipitation assay revealed that PRG-1 is capable of binding the Ca2+/CaM complex in vitro and in transfected cells. Surface plasmon resonance and zero-length cross-linking showed that the C-terminal putative cytoplasmic domain (residues 466–766) of PRG-1 binds equimolar amounts of CaM in a Ca2+-dependent manner, with a relatively high affinity (a Kd value for Ca2+/CaM of 8 nM). Various PRG-1 mutants indicated that the Ca2+/CaM-binding region of PRG-1 is located between residues Ser554 and Gln588, and that Trp559 and Ile578 potentially anchor PRG-1 to CaM. This is supported by pronounced changes in the fluorescence emission spectrum of Trp559 in the PRG-1 peptide (residues 554–588) upon binding to Ca2+/CaM, showing the stoichiometrical binding of the PRG-1 peptide with Ca2+/CaM. Immunoblot analyses revealed that the PRG-1 protein is abundant in brain, but is weakly expressed in the testes. Immunohistochemical analysis revealed that PRG-1 is highly expressed in forebrain structures and in the cerebellar cortex. Furthermore, PRG-1 localizes at the postsynaptic compartment of excitatory synapses and dendritic shafts of hippocampal neurons, but is not present in presynaptic nerve terminals. The combined observations suggest that PRG-1 may be involved in postsynaptic functions regulated by intracellular Ca2+-signalling.
- Ca2+ signalling
- calmodulin (CaM)
- plasticity-related gene-1 (PRG-1)
CaM (calmodulin) is a multifunctional and ubiquitously expressed Ca2+-signal transducer that is composed of four EF-hand Ca2+-binding motifs . An increase in intracellular Ca2+ induced by various extracellular stimuli leads to a Ca2+-loaded form of CaM, followed by CaM interaction with its cellular targets . The Ca2+/CaM complex interacts with numerous cellular targets to regulate their function, resulting in the modulation of a large number of Ca2+-dependent signal transduction processes such as secretion, contraction, cell cycle and gene expression [3,4]. Accordingly, CaM has been recognized as playing an important role in physiological responses in many tissues and cells and as mediating a number of actions of Ca2+ as an intracellular second messenger.
Numerous studies have shown that Ca2+/CaM-mediated signal transduction is important for the plasticity of the central nervous system, including for the processes of learning and memory [5,6]. Brain tissue contains a high concentration of CaM (~0.5 mg/g of tissue) , as well as a variety of CaM-binding proteins including Ca2+/CaM-dependent protein kinases [8,9], protein phosphatase (calcineurin) , nitric oxide synthase [11,12], adenylate cyclase , phosphodiesterase , ion transporters , membrane receptors [16,17] and cytoskeletal proteins . Therefore it is quite possible that brain contains as yet unidentified CaM-regulated proteins that are involved in Ca2+-dependent modulation of the central nervous system.
Previously, both Berggård et al.  and our group  have searched for novel CaM-mediated intracellular pathways by independently performing comprehensive analyses of CaM targets, which were identified using a proteomics approach. Using LC-MS/MS (liquid chromatography-tandem MS), we identified 36 known Ca2+/CaM-binding proteins, as well as a novel CaM-binding protein, wolframin, in a rat brain extract that had been partially purified using CaM-affinity chromatography .
In the present study, in a continuation of our functional proteomics approach, we identified PRG-1 (plasticity-related gene-1) as a novel CaM-binding protein. PRG-1 was originally identified by differential screening of a cDNA library derived from lesioned hippocampus . PRG-1 is a neuron-specific membrane protein, which has 766 (rat and mouse) or 763 (human) amino acid residues and an apparent molecular mass of ~80 kDa. It has been reported that PRG-1 possesses lipid phosphatase activity and is involved in axonal outgrowth as well as in neuron sprouting .
We further biochemically characterized the binding ability of PRG-1 and Ca2+/CaM, and immunohistochemically analysed PRG-1 tissue and subcellular distribution.
MATERIALS AND METHODS
Recombinant rat CaM was expressed in Escherichia coli BL21 (DE3) cells (Stratagene) using the pET-CaM vector (kindly provided by Dr Nobuhiro Hayashi, Fujita Health University, Toyoake, Japan) and was purified using phenyl–Sepharose column chromatography . Rat CaM fused with GST (glutathione transferase) by a Gly×6 spacer (CaM–GST) was constructed by PCR amplification using PrimeSTAR HS DNA polymerase (Takara) as described below. A cDNA-encoded rat CaM with Gly×3 at its C-terminus was PCR-amplified using pET-CaM as a template and the primers 5′-GGCCCACCATGGCTGACCAACTGACTGA-3′ (sense) and 5′-ACCACCACCCTTCGCTGTCATCATTTGTAC-3′ (antisense). A cDNA-encoded GST with Gly×3 at its N-terminus was amplified using pGEX-2T (GE Healthcare UK) as a template and the primers 5′-GGTGGTGGTATGTCCCCTATACTAGGTTAT-3′ (sense) and 5′-GGCTCGAGTCAGGATCCACGCGGAACCAG-3′ (antisense). The cDNA was digested with NcoI and XhoI (restriction sites in the primers are underlined) respectively, and was ligated into the NcoI/XhoI site of the vector pET16b (Novagen Merck). The recombinant CaM–GST was expressed in E. coli BL21-CodonPlus(DE3)-RIL (Stratagene), and was purified using glutathione–Sepharose chromatography followed by phenyl–Sepharose chromatography. Mouse PRG-1 cDNA (pTA-mPRG-1, GenBank® accession number AF541279) was kindly provided by Dr Olaf Ninnemann and Dr Robert Nitsch (Charité-Universitätsmedizin, Berlin, Germany) . Human PRG-1 cDNA (KIAA0455, GenBank® accession number AB007924) was obtained from the Kazusa DNA Research Institute (Chiba, Japan). The anti-HA (haemagglutinin) (12CA5) and anti-FLAG (clone M2) antibodies were obtained from Roche Applied Sciences and Sigma–Aldrich respectively. The anti-His tag antibody (27E8) and anti-HSP70 (heat-shock protein 70) antibody (C92F3A-5) were purchased from Cell Signaling Technology and Assay Designs respectively. Biotinylated CaM was purchased from Calbiochem. PRG-1 peptide (S554ARAKWLKAAEKTVACNRGNNQPRIMQVIAMSKQQ588) was synthesized and obtained with 97% purity, estimated by HPLC and MS (SCRUM). All other chemicals were obtained from standard commercial sources.
Whole rat brains (5.3 g) were homogenized in 20 ml of Buffer A [150 mM NaCl, 50 mM Tris/HCl (pH 7.5), 1 mM DTT (dithiothreitol), 0.2 mM PMSF and 10 μg/ml leupeptin] containing 1 mM EDTA, 1 mM EGTA and 1% Nonidet P40, followed by centrifugation at 35000 rev./min for 60 min at 4 °C. Purified CaM–GST (19 mg) was added and mixed with the supernatant, followed by the addition of CaCl2 (5 mM final concentration). The mixture of rat brain extract with CaM–GST was applied to glutathione–Sepharose columns (2 ml bed volume, GE Healthcare) and the columns were then washed with 20 ml of Buffer B (Buffer A containing 0.2 mM CaCl2). After washing with 20 ml of Buffer B containing 1 M NaCl, the column was washed with 20 ml of Buffer B. Elution of CaM-binding proteins from the column was carried out using Buffer A containing 2 mM EGTA. Fractions (2 ml) were collected, SDS/PAGE sample buffer (36 μl) and 1 M DTT (4 μl) were added to the eluate, and each sample was then stored at −80 °C until analysis by MS or CaM overlay .
A 24 μl sample of the eluate from the CaM–GST-coupled glutathione–Sepharose column described above was concentrated, separated by SDS/PAGE (10% gels) and lightly stained with Coomassie Brilliant Blue. Next, 20 gel slices were excised from each sample lane in the 30–200 kDa range, followed by in-gel digestion with 10 μg/ml trypsin (Promega) overnight at 37 °C . The digested peptides were eluted with 0.1% formic acid and were subjected to LC-MS/MS analysis, which was performed on a Q-Tof 2 quadrupole/time-of-flight hybrid mass spectrometer (Micromass) interfaced with a CapLC capillary reverse-phase LC system (Micromass). A 90 min linear gradient from 5 to 45% acetonitrile in 0.1% formic acid was produced and split at a 1:20 ratio. The gradient solution was then injected into a NanoLC column (PepMap C18, 75 μm×150 mm; LC Packings) at 100 nl/min. The eluted peptides were sprayed directly into the mass spectrometer. MS/MS data were acquired by MassLynx software (Micromass) and were converted into a single text file (containing the observed precursor peptide m/z, the fragment ion m/z and intensity values) by ProteinLynx software (Micromass). The file was analysed using the Matrix Science Mascot MS/MS Ion Search (http://www.matrixscience.com) to search and assign the obtained peptides to the NCBI non-redundant database. The search parameters were set as follows: database, NCBInr; taxonomy, all; enzyme, trypsin; fixed modifications, carbamidomethyl (C); variable modifications, oxidation (M); peptide tol., ±0.2 Da; and MS/MS tol., ±0.2 Da.
Cloning and construction of PRG-1 cDNAs
Rat PRG-1 cDNA (GenBank® accession number AY266268) encoding residues 2–766 was obtained by RT (reverse-transcriptase)–PCR using PrimeSTAR HS DNA polymerase (Takara), rat brain cDNA (QUICK-Clone, Clontech Laboratories) as a template and the primers 5′-CAGCGCGCTGGCTCCAGCGGTGCCCGC-3′ (sense) and 5′-CCTCTAGATCAATCCTTATAAGGCCTCGTG-3′ (antisense, Xba1 restriction site underlined). The resulting PCR fragment was subcloned into the EcoRV/XbaI site of the pME18s-HA vector. Full-length rat PRG-1 was created by PCR using a sense primer containing the first ATG of PRG-1 followed by subcloning into pcDNA3.1 (Clontech Laboratories). Expression plasmids for HA-tagged mouse (residues 2–766) and human (residues 2–763) PRG-1 were generated by PCR using the appropriate cDNA as a template followed by subcloning into the EcoRV/XbaI site of the pME18s-HA vector. For expression of GST-rat PRG-1 C-terminal fragments, PCR fragments encoding residues 364–766, 386–766, 442–766, 531–766, 554–766, 560–766 or 564–766 were introduced into the BamHI/SmaI site of the vector pGEX-KG-PreS-His6, resulting in the addition of GHHHHHH to the C-terminal end (after Asp766) of GST–rat PRG-1. C-terminal truncation and point mutants of HA–rat PRG-1 were created by PCR using PrimeSTAR HS DNA polymerase or site-directed mutagenesis (GeneEditor™ system; Promega) and pME18s-HA-rat PRG-1 wild-type (residues 2–766) as a template. The rat PRG-1 fragment (residues 466–766) with GHHHHHH at the C-terminal end (after Asp766) was created by PCR using pGEX-PRG-1 364-766-His6 as a template and was introduced into the NcoI/BamHI site of pET16b (Novagen). The nucleotide sequences of all constructs used in the present study were confirmed by sequencing using an ABI PRISM 310 automated sequencer (Applied Biosystems).
Purification of recombinant PRG-1
GST–rat PRG-1 C-terminal fragment cDNAs (pGEX-PRG-1-His6) or the rat PRG-1 466–766-His6 cDNA were introduced into E. coli BL21-CodonPlus (DE3)-RIL (Stratagene), and expression of the recombinant proteins was induced by the addition of IPTG (isopropyl β-D-thiogalactopyranoside). An E. coli pellet containing GST-fusion protein was lysed with PBS and then purified using glutathione–Sepharose column chromatography as described in the manufacturer's protocol. Further purification of the GST-fused proteins was then carried out using Ni-NTA (Ni2+-nitrilotriacetate)-agarose column chromatography (Qiagen). The cytoplasmic domain of rat PRG-1 (residues 466–766, wild-type) with added GHHHHHH was purified using CaM–Sepharose column chromatography (GE Healthcare) followed by Ni-NTA-agarose column chromatography (Qiagen) according to the manufacturer's protocol. Two mutants (W559A and I578A) were purified by Ni-NTA column chromatography followed by Ultrogel AcA 54 gel-filtration column chromatography (Pall Corporation) (2.5 cm×88 cm).
Expression of PRG-1 in COS-7 cells
COS-7 cells were maintained in DMEM (Dulbecco's modified Eagle's medium) containing 10% fetal bovine serum. Transfection of pME-HA-PRG-1 wild-type and mutant constructs, or of pcDNA3 rat PRG-1 (10 μg), into COS-7 cells (10 cm dishes) was carried out using the Lipofectamine™ reagent (Invitrogen). After 40 h of incubation, the cells were washed with PBS and then collected in Buffer C [20 mM Tris/HCl (pH 7.5), 150 mM NaCl, 2 mM EGTA, 2 mM EDTA, 0.2 mM PMSF, 10 mg/l leupeptin and 10 mg/l trypsin inhibitor] followed by sonication. Cell extracts were centrifuged at 21880 g for 15 min at 4 °C, and the precipitated membrane fraction was then dissolved in 50 μl of SDS/PAGE buffer. An aliquot of this sample (5 μl) was used for CaM-overlay or Western blot analyses.
Samples were first separated on SDS/PAGE followed by transfer of the proteins on to a nitrocellulose membrane (Hybond C, GE Healthcare). The membrane was then incubated with a buffer containing 150 mM NaCl and 20 mM Tris/HCl (pH 7.5) supplemented with 5% BSA in the presence of either 1 mM CaCl2 or 2 mM EGTA for 1 h and was then incubated with 0.5 μg/ml biotinylated CaM in the same buffer without BSA for 1 h. After extensive washing, the membrane was incubated with avidin–horseradish peroxidase in either CaCl2- or EGTA-containing buffer for 1 h. After washing the membrane again, a chemiluminescence reagent (PerkinElmer Life Sciences) was used for detection of the CaM-binding signal.
SPR (surface plasmon resonance) analysis
PRG-1–CaM binding interactions were measured by SPR using a Biacore 2000 system (Biacore AB). The recombinant, His6-tagged, cytoplasmic domains of rat PRG-1 (residues 466–766) including wild-type, W559A and I578A mutants were covalently immobilized on CM5 research grade chips (Biacore) to a level represented by 1170 (wild-type), 1014 (W559A) and 1057 (I578A) response units using NHS (N-hydroxysuccinamide) and EDC [1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide] (Biacore amine coupling kit). CaM (0.15–2.32 μM) was injected over the sensor surface at a flow rate of 30 μl/min in binding buffer [HBS-P: 10 mM Hepes (pH 7.4), 150 mM NaCl and 0.005% Surfactant P20] containing either 1 mM CaCl2 or 2 mM EGTA. CaM was allowed to interact with the surface of the sensor chip for 2 min, after which CaM-free binding buffer was injected over the sensor surface to monitor CaM dissociation. The sensor surface was regenerated using a calcium-free flow buffer (HBS-P containing 5 mM EGTA) between injections. BIAEvaluation Software (Version 4.1; Biacore) was used to process and analyse the raw binding data to obtain a Kd value for CaM.
Two-step zero-length cross-linking was carried out according to a method described by Grabarek and Gergely . After CaM (39 μg) was treated with 50 mM NHS and 30 mM EDC at 25 °C for 15 min in a solution (40 μl) containing 150 mM NaCl, 25 mM Hepes (pH 7.5) and 0.2 mM CaCl2, activation was terminated by the addition of 2-mercaptoethanol (10 μl). Activated CaM (16 μg) was then incubated with the recombinant His6-tagged cytoplasmic domain of PRG-1 (residues 466–766, 15 μg) in the presence of 1 mM CaCl2 at 25 °C for 1 h. After termination of the reaction by the addition of SDS/PAGE sample buffer, samples were analysed by SDS/PAGE (12.5% gels).
Tryptophan fluorescence emission spectra were recorded using an Hitachi 650-10 fluorescence spectrophotometer with excitation at 290 nm (bandwidth 4 nm) and emission scanned from 300 to 400 nm (bandwidth 4 nm). Spectra were recorded at 20 °C in quartz cuvettes in a solution containing 150 mM NaCl, 20 mM Tris/HCl (pH 7.5), 1 mM CaCl2 and 1.4 μM PRG-1 peptide with various concentrations (0–2.81 μM) of CaM.
CD spectra were recorded on a Jasco J-725 spectropolarimeter at 20 °C in 150 mM NaCl, 20 mM Tris/HCl (pH 7.5). Far-UV CD spectra (200–260 nm) were measured using 1 mm cuvettes with 0.5 mg/ml rat PRG-1 fragments (residues 466–766) including wild-type, W559A and I578A mutants.
Production of an anti-PRG-1 antibody
An anti-PRG-1 guinea pig polyclonal antibody was produced by immunization with a purified, His6-tagged, rat PRG-1 C-terminal fragment (residues 466–766) and was purified using antigen-coupled affinity-column chromatography.
Preparation of rat tissue lysates
All experimental procedures involving animals were performed according to the guidelines for the care and use of animals as established by the Kagawa University Animal Ethics Committee and the Animal Experimentation and Ethics Committee of the Kitasato University School of Medicine.
Rat tissues were homogenized in 7 volumes of lysis buffer [50 mM Tris/HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, 1% sodium deoxycholate, 1 mM PMSF and a protease inhibitor cocktail (Sigma; 10.4 mM AEBSF, 8 μM aprotinin, 0.2 mM leupeptin, 0.4 mM bestatin, 0.15 mM pepstatin A and 0.14 mM E-64) and were then centrifuged at 21880 g for 60 min. The supernatants were collected and analysed by Western blotting.
Male Wistar rats, at postnatal week 7, were fixed by transcardial perfusion with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) under deep anaesthesia. After overnight incubation with the same fixative, the brains were cryoprotected with 30% (w/v) sucrose in phosphate buffer and were cut in coronal or sagittal planes to a thickness of 25 μm using a cryostat. The sections were solubilized in 0.3% Triton X-100 for 30 min, blocked with 5% normal goat serum, and were incubated overnight with guinea pig polyclonal antibody against PRG-1 at a final concentration of 1 μg/ml. After overnight incubation with the primary antibody, the sections were incubated with biotinylated anti-guinea pig IgG (Vector Laboratories), followed by incubation with horseradish peroxidase-conjugated streptavidin (Dako). Immunoreactions were visualized using DAB (3,3′-diaminobenzidine) tetrahydrochloride chromogenic substrate (Dako). For pre-embedding immunoelectron microscopic analysis, the sections were subjected to the same procedure described above, except that Triton X-100 was omitted. After visualization with DAB, the sections were post-fixed with 1% osmium tetroxide, then incubated with 2% uranyl acetate, dehydrated and embedded in epoxy resin. Ultrathin sections were examined using a JEM-1230 electron microscope (JEOL Limited). For immunofluorescent staining of brain sections, combinations of the following primary antibodies were used: guinea pig polyclonal anti-PRG-1 IgG, mouse monoclonal anti-synaptophysin IgG (clone SVP-38, Sigma–Aldrich), rabbit polyclonal anti-PSD-95 (postsynaptic density 95) IgG  or rabbit polyclonal anti-gephyrin IgG. Immunoreactions were visualized using Alexa Fluor® 488-conjugated anti-guinea pig IgG or Alexa Fluor® 594-conjugated anti-mouse or anti-rabbit IgG (Molecular Probes). For immunofluorescent staining of cultured neurons, the plates were fixed 17 days after plating using 4% paraformaldehyde for 10 min and were then incubated with guinea pig polyclonal anti-PRG-1 IgG, mouse monoclonal anti-NF200 (neurofilament 200) IgG (clone NE14, Sigma–Aldrich) and rabbit polyclonal anti-MAP2 (microtubule-associated protein 2) IgG (Chemicon) or with guinea pig polyclonal anti-PRG-1 IgG and mouse monoclonal anti-synaptophysin IgG (Sigma–Aldrich). To visualize the immunoreaction, the slides were subsequently incubated with Alexa Fluor® 488-conjugated anti-guinea pig IgG, Alexa Fluor® 594-conjugated anti-mouse IgG and Alexa Fluor® 647-conjugated anti-rabbit IgG (Molecular Probes). Nuclei were counterstained with DAPI (4′,6-diamidino-2-phenylindole). Immunofluorescent images (512×512 pixels) were obtained using a confocal laser-scanning microscope (LSM 710, Zeiss) and ×10, ×20 and ×63 plan-apochromat objective lenses. The brightness and contrast of the final images were adjusted using Photoshop CS4 software (Adobe Systems).
Hippocampal culture and immunostaining
Hippocampal neurons were prepared from the hippocampi of Wistar rats on embryonic day 18 according to a procedure that has been previously described in detail . The neurons were plated on to poly-D-lysine and laminin-coated plastic dishes at a density of 0.75×106 cells per 35-mm dish and were maintained at 37 °C in a growth medium consisting of neurobasal medium (Life Technologies) supplemented with 2×B27 supplement and 1 mM Glutamax-I (Gibco) in a humidified incubator with 5% CO2.
Western blot analysis was performed with the primary antibodies indicated and with horseradish peroxidase-conjugated anti-mouse IgG (GE Healthcare UK) or anti-guinea pig IgG (Jackson Immunoresearch) as the secondary antibody. A chemiluminescence reagent (PerkinElmer Life Sciences) was used for signal detection. The protein concentration was estimated by staining the samples with Coomassie Brilliant Blue (Bio-Rad Laboratories) using BSA as a standard.
Identification of the product of PRG-1 as a novel CaM target
We recently performed comprehensive analyses of CaM-binding proteins that were identified using a newly developed functional proteomic screen to search for novel CaM-mediated signalling pathways . Affinity purification of CaM targets using CaM-fused GST as an affinity ligand, followed by MS analysis, was used for identification of CaM-interacting proteins. In addition to the targets of CaM from rat brain extract that we previously identified using this functional proteomic approach , we obtained five peptides (Figure 1A; T-1, T-2, T-3, T-4 and T-5) that completely matched regions of the amino acid sequence of the product of PRG-1 (Figure 1A). PRG-1 has not previously been shown to interact with CaM.
In order to examine the Ca2+-dependency of the interaction of PRG-1 with CaM, we analysed membrane fractions from PRG-1-transfected COS-7 cells by a CaM-overlay assay. As shown in Figure 1(B), HA-tagged PRG-1 of rat, mouse or human origin expressed in COS-7 cells gave a CaM-binding signal at a position of ~80 kDa in the presence of 1 mM CaCl2 (Figure 1B, left-hand and centre panels), but not if Ca2+ was chelated with EGTA (Figure 1B, right-hand panel). We next tested whether PRG-1 is capable of binding to Ca2+/CaM in intact cells by immunoprecipitation analysis (Figure 1C). HA-tagged rat PRG-1 was expressed with or without FLAG-tagged CaM in COS-7 cells and then FLAG–CaM was immunoprecipitated from the transfected cell lysate using an anti-FLAG antibody in the presence of 2 mM CaCl2 or 2 mM EGTA. Western blot analysis with an anti-HA antibody showed that HA-tagged rat PRG-1 specifically co-immunoprecipitated with FLAG–CaM in the presence, but not in the absence, of Ca2+ (Figure 1C, bottom panel). These results indicate that PRG-1 is capable of binding the Ca2+/CaM complex in cells, as well as in vitro.
Identification of the CaM-binding region in rat PRG-1
In order to precisely map the Ca2+/CaM-binding site in PRG-1, we constructed and expressed HA-tagged rat PRG-1 encoding various C-terminal truncation mutants in COS-7 cells followed by CaM-overlay analysis in the presence of 1 mM CaCl2. The CaM-binding ability of PRG-1 truncated at residue 588 (2–588) was similar to that of wild-type PRG-1 (2–766) or of PRG-1 (2–592). However, truncation of a further three amino acid residues (2–585) completely abolished PRG-1–CaM binding (Figure 2A). This result indicates that the C-terminal putative cytoplasmic domain of PRG-1 contains the CaM-binding site. To confirm this possibility, we constructed and purified GST-fused PRG-1 C-terminal deletion mutants and analysed their CaM-binding by CaM-overlay analysis in the presence of 1 mM CaCl2 (Figure 2B). A GST-fused C-terminal fragment of PRG-1 (554–766) bound Ca2+/CaM in a similar manner to longer PRG-1 fragments (531–766, 442–766, 386–766, 364–766), but a PRG-1 fragment with deletion of a further six amino acids (560–766) failed to bind Ca2+/CaM (Figure 2B). These results indicate that the minimum Ca2+/CaM-binding region of rat PRG-1 is located between Ser554 and Gln588, which is conserved in mouse and human PRG-1 (results not shown). Next we analysed the stoichiometry of CaM binding to PRG-1 by a two-step zero-length cross-linking method using EDC in the presence of NHS . Since PRG-1 is a transmembrane protein , it is difficult to prepare soluble recombinant full-length PRG-1. Therefore the C-terminal domain of PRG-1 (residues 466–766) fused with a His-tag and purified using CaM–Sepharose followed by Ni-NTA agarose chromatography, was used for this analysis. Although the apparent molecular mass of the C-terminal domain of rat PRG-1 on SDS/PAGE is calculated to be ~35 kDa, the molecular mass of the cross-linked product in the presence of Ca2+ and CaM (~17 kDa) was increased to 58 kDa (Figure 2C). These observations clearly suggest that PRG-1 bound to CaM in equimolar amounts.
Alanine-scanning mutagenesis of CaM-binding in PRG-1
Based on the mapping of the CaM-binding region in PRG-1 (Figures 2A and 2B), we were able to synthesize the PRG-1 peptide (Ser554–Gln588, Figure 3B) to directly examine its Ca2+/CaM-binding by analysis of the fluorescence emission spectrum of the single tryptophan residue (Trp559) in the peptide (Figure 3A). Upon binding to equimolar amounts of Ca2+/CaM, the fluorescence emission maximum of the Trp559 in the PRG-1 peptide shifts from 359 to 334 nm, and the emission intensity increases, as shown in Figure 3A (insert). This signal intensification and blue-shift in the emission maximum are consistent with the Trp559 being in a more hydrophobic environment when the PRG-1 peptide is bound to CaM. Changes in the fluorescence emission intensity at 325 nm as a function of added CaM are shown in Figure 3(A). The major changes in Trp559 fluorescence occurs as the [CaM]/[PRG-1 peptide] ratio changes from 0.8 to 1, indicating 1:1 stoichiometrical binding of CaM to PRG-1 peptide, consistent with a result of cross-linking analysis shown in Figure 2(C). Structural studies of the Ca2+/CaM complex bound to CaM-binding peptides, which were derived from CaM-dependent protein kinases, including CaMKII (Ca2+/CaM-dependent protein kinase II) (1-10 motif) , MLCKs (myosin light-chain kinases) (1-14 motif) [28,29] and CaMKKs (CaM kinase kinases) (1-16 motif) [30,31] have been reported. These studies indicate that CaM-binding peptides are composed of 20–26 amino acid residue amphipathic α-helices, which contain two key hydrophobic residues that insert into hydrophobic pockets of CaM . A very recent report indicated a novel CaM-binding motif in which Ca2+/CaM is wrapped around helical CaM-binding peptide in the PMCA pump (plasma-membrane Ca2+-ATPase pump) using two anchoring residues from the peptide at relative positions 18 and 1 (18-1 motif) . Compared with these well-characterized CaM-binding peptides, the CaM-binding region in PRG-1 identified in the present study (residues 554–588) is apparently longer than conventional CaM-binding sequences (Figure 3B). We therefore attempted to identify key hydrophobic residues of PRG-1 that would function in anchoring PRG-1 to CaM using alanine-scanning mutagenesis of HA–rat PRG-1 (Figure 3C). Although the expression level of all of the mutants was comparable with that of wild-type PRG-1 (Figure 3C, upper panel), the W559A and I578A mutants completely failed to bind CaM (Figure 3C, lower panel). Similar results were obtained when we analysed the C-terminal fragments (residues 466–766) of PRG-1 including wild-type, W559A and I578A mutants by CaM overlay (Figure 4B). Far-UV CD measurements were also used to examine the structure of PRG-1 fragments. A random coil sequence yields a very typical CD spectrum, with a sharp negative minimum at approx. 200 nm. The recorded CD spectra of purified C-terminal fragments (residues 466–766) of PRG-1 wild-type, W559A and I578A mutants (Figure 4A) are very similar and show this characteristic profile (Figure 4C). The CD results therefore clearly indicated that the PRG-1 C-terminal fragments (residues 466–766) are unstructured in solution and mutations of Trp559 and Ile578 to alanine did not significantly affect the structure of the C-terminal domain of PRG-1. Taken together, these results suggest that Trp559 and Ile578 are the essential hydrophobic residues that anchor PRG-1 to the hydrophobic pockets of CaM.
Characterization of CaM-binding of the rat PRG-1 cytoplasmic domain
In order to examine the kinetics of PRG-1–CaM binding, we analysed CaM binding of the rat PRG-1 C-terminal fragment (residues 466–766) including wild-type, W559A and I578A mutants by SPR (Figure 5). Purified PRG-1 C-terminal fragments as shown in Figure 4(A) were immobilized on an SPR sensor chip (CM5) by amine coupling. Figure 5 shows the binding responses of the C-terminal domains (residues 466–766) of rat PRG-1 to injection of a series of CaM concentrations with 1 mM CaCl2 or with 2 mM EGTA (results not shown). Non-specific-binding responses to the sensor chip surface were subtracted (see the Materials and methods section). CaM bound with rapid kinetics to the C-terminal domain (residues 466–766) of rat PRG-1 wild-type in a concentration-dependent manner (Figure 5, left-hand panel). Binding of CaM to the C-terminal domain of PRG-1 wild-type was completely calcium-dependent (results not shown), with a calculated Kd for CaM of 8 nM. In addition, we could observe Ca2+/CaM-binding of W559A and I578A mutants with calculated Kd values for CaM of 662 nM and 80 nM respectively (Figure 5, middle and right-hand panels). These Kd values of the mutants were approximately one to two orders of magnitude higher than that of the wild-type PRG-1 C-terminal domain, which was due to higher dissociation rate constants of the mutants (W559A mutant: kd=4.5×10−1 s−1; I578A mutant: kd=8×10−2 s−1) as compared with that of wild-type (kd=6×10−3 s−1). The SPR data therefore indicates that PRG-1 bound to CaM with a relatively high affinity and strengthen the conclusion from CaM-overlay analysis (Figures 3C and 4B) that Trp559 and Ile578 are crucial residues for the CaM binding.
Tissue distribution and immunohistochemical localization of PRG-1 in the adult rat brain
In addition to PRG-1 biochemical characterization, we further performed immunohistochemical analyses of rat PRG-1 using a newly generated polyclonal antibody against the C-terminal fragment (residues 466–766) of rat PRG-1. The specificity of this antibody was first confirmed by Western blotting analysis in which a single immunoreactive band at ~80 kDa was detected in both lysates of rat cerebrum and of rat PRG-1-transfected COS-7 cells, but not in lysates of non-transfected COS-7 cells (Figure 6A). Using this antibody, the tissue distribution of PRG-1 was then determined by Western blot analysis of various tissue lysates of the adult male rat. The anti-PRG-1 antibody strongly cross-reacted with a single immunoreactive band in cerebrum and cerebellum, and weakly cross-reacted with a band in testes, but did not cross-react with any band in other peripheral tissues (Figure 6B, upper panel). This protein expression pattern is consistent with the Northern blot analysis of PRG-1 mRNA . Next, to examine the localization of PRG-1 in the brain, parasagittal sections of the adult rat brain were subjected to immunoperoxidase staining with an anti-rat PRG-1 antibody. Prominent immunostaining of PRG-1 was observed in cerebellar cortex and forebrain structures such as the olfactory bulb, the hippocampus, the cerebral cortex and the caudate putamen (Figure 7A). In control experiments, the immunostaining was completely abolished when the PRG-1 antibody was omitted (results not shown).
Localization of PRG-1 in CA1 hippocampal neurons
The tissue localization of PRG-1 was further examined by immunofluorescence analysis in the hippocampus, where PRG-1 was expressed at the highest level (Figures 7B–7I). PRG-1 was densely distributed in the strata radiatum and oriens of the CA1–3 (CA is cornu ammonis) regions and in the dentate molecular layer, but was distributed at a lower density in the stratum lacunosum-moleculare (Figure 7B). Immunoreactivity for PRG-1 was weak in the stratum pyramidale and the dentate granule cell layer. At high magnification of the dendritic field, PRG-1 appeared as tiny puncta, which delineated immunonegative dendritic shafts (Figure 7C). In double immunofluorescent staining for PRG-1 and PSD-95, an excitatory postsynaptic marker, PRG-1-immunoreactive puncta largely co-labelled with PSD-95-immunoreactive puncta (Figures 7G–7I) and localized close to synaptophysin-immunoreactive nerve terminals (Figures 7D–7F). In contrast, PRG-1 distribution did not overlap with gephyrin, an inhibitory postsynaptic marker (results not shown). To further confirm the dendritic localization of PRG-1, cultured hippocampal neurons at DIV17 (DIV is days in vitro) were immunohistochemically co-stained for PRG-1 and a marker of axons, NF200 or of dendrites, MAP2. As shown in Figures 8(A)–8(D), the intense PRG-1 immunoreactive signal strongly overlapped with MAP2-immunoreactive dendrites but with few, if any, NF200-immunoreactive axons. At high magnification of immunoreactive dendrites, PRG-1 immunoreactivity occurred as numerous spots in dendritic shafts and spines (Figure 8H). Double immunofluorescent staining showed that PRG-1-immunoreactive puncta partially overlapped with, and were closely apposed to, synaptophysin-immunoreactive nerve terminals in hippocampal neurons (Figures 8E–8J). All of these findings suggest that PRG-1 is localized at excitatory synapses of hippocampal neurons.
To further examine the precise PRG-1 synaptic localization, we performed pre-embedding immunoelectron microscopy in the stratum radiatum of the hippocampal CA1 region (Figure 9). Immunoreactive deposits for PRG-1 were densely accumulated on the postsynaptic membranes of asymmetrical synapses that were formed at the tips of dendritic spines as well as in dendritic shafts, whereas PRG-1 staining was rarely seen in presynaptic nerve terminals.
Numerous studies have demonstrated that the four EF-handed Ca2+-binding protein CaM is conserved in diverse species and functions as a transducer of biological Ca2+ signals. Based on increasing database information concerning CaM and CaM targets, it is clear that CaM regulates a large number of pathways that involve Ca2+ signal transduction and that CaM regulation of these pathways is complicated. In particular in the central nervous system, which has the highest tissue content of CaM , there are a variety of CaM targets including signal transduction enzymes such as protein kinases and a phosphatase, as well as metabolic enzymes and receptors. It has been shown that CaM-mediated Ca2+ signalling plays important roles in various neuronal functions including synaptogenesis , dendritic arborization  and cortical axon elongation , as well as in synaptic plasticity . The previous comprehensive search for novel CaM targets using a functional proteomic approach opened up the possibility of discovering novel Ca2+ signalling pathways [19,20]. In the present study we characterized the CaM-binding properties of PRG-1, which was identified as a potential CaM-binding protein in a functional proteomic screen.
We confirmed that PRG-1 shows Ca2+-dependent CaM binding by affinity chromatography, CaM-overlay, fluorescence measurement and SPR analysis. We further showed that the C-terminal putative cytoplasmic domain (residues 466–766) of PRG-1 binds equimolar amounts of CaM with a relatively high affinity (Kd=8 nM), which is comparable with that for CaM-regulated enzymes such as CaM kinases. We also confirmed that PRG-1 and CaM could be co-immunoprecipitated from transfected COS-7 cells, suggesting their in vivo interaction. However, we cannot rule out the possibility that CaM and PRG-1 might have interacted following homogenization of the cells. Therefore more direct proof of the in vivo interaction of CaM and PRG-1 will require further analyses. CaM-overlay and SPR analysis showed that substitution of Trp559 and Ile578 of rat PRG-1 with alanine completely impaired CaM binding and significantly reduced the binding affinity of PRG-1 for CaM respectively, indicating that both residues are apparently key hydrophobic residues for anchoring PRG-1 to the hydrophobic pocket of the CaM molecule. Whereas we could observe weak Ca2+/CaM-binding of W559A and I578A mutants by SPR analysis, the results obtained from CaM overlay showed that both mutations apparently completely abolished the CaM binding. This is probably due to using a low concentration of biotinylated CaM (~30 nM) for the CaM-overlay assay. This 18-residue separation between two key hydrophobic groups of a CaM-binding protein is unique and is the longest among previously determined CaM-binding proteins, indicating a new and distinct class of Ca2+/CaM target recognition motifs (1-20 motif) . Accumulated evidence has indicated novel CaM-binding motifs, except for conventional α-helical CaM-binding sequences . For example, a single CaM molecule interacts simultaneously with two target peptides derived from petunia glutamate decarboxylase, adopting an α-helical structure  and the CaM-binding peptide from MARCKS (myristoylated alanine-rich C-kinase substrate) adopted a unique non-α-helical structure . Thus a novel target sequence of PRG-1 (1-20 motif) found in the present study may indicate that it did not adopt an α-helical structure.
PRG-1 was initially discovered as a neuron-specific membrane-associated LPP (lipid phosphate phosphatase) that attenuates phospholipid-induced axon collapse, resulting in axon outgrowth , we therefore tested the effect of Ca2+/CaM on the LPP activity of PRG-1 expressed in COS-7 cells using 32P-labelled oleoyl-LPA (lysophosphatidic acid) as a substrate. However, we could not detect any [32P]LPA hydrolysing activity of PRG-1 either in the presence or absence of Ca2+/CaM even though [32P]LPA hydrolysing activity of the control PAP-2a (phosphatidic acid phosphatase type 2a isozyme)  was detected (results not shown). These results suggest that an as yet unidentified activity of PRG-1 may be modulated by Ca2+/CaM-binding.
Regarding the subcellular localization of PRG-1, our present immunohistochemical analysis revealed that rat PRG-1 is highly expressed in forebrain structures and in the cerebellar cortex and that it localizes exclusively at dendrites, particularly excitatory postsynapses of dendritic spines, in hippocampal neurons, although PRG-1 has not been found in a proteomics study of postsynaptic density purified from adult rat forebrain and cerebellum . During the course of the present study, Trimbuch et al.  reported a similar immunohistochemical localization of PRG-1 in mouse brain. They also showed that deletion of prg-1 in mice leads to epileptic seizures and augmentation of EPSCs (excitatory postsynaptic currents), indicating that PRG-1 plays an important role in modulation of hippocampal excitability. The postsynaptic compartment contains numerous CaM-binding proteins including Ca2+ channel and NMDA (N-methyl-D-aspartate)-type glutamate receptors, regulatory enzymes such as CaMKII, Ca2+/CaM-dependent protein phosphatase (calcineurin), Ca2+/CaM-dependent cAMP phosphodiesterase, adenylate cyclase, nitric oxide synthase and RasGRF1, all of which are generally accepted to regulate neuronal transmission and plasticity initiated by an increase in intracellular Ca2+ concentration . Combined with our present results, these data suggest that PRG-1 regulates excitatory synaptic transmission and that this regulation may be modulated via PRG-1–CaM binding and controlled by changes in the concentration of intracellular Ca2+.
In summary, our findings may shed light on a novel property of PRG-1 as a CaM-binding protein at the postsynaptic membrane of excitatory synapses. However, it is difficult to speculate a specific role for CaM binding in PRG-1 function since PRG-1 has no well-defined biochemical function at present. Further studies specifically designed to explore the molecular function and regulation of PRG-1 by the Ca2+/CaM complex will be necessary to understand the mechanism and function of Ca2+/CaM-PRG-1 pathways in the central nervous system.
Hiroshi Tokumitsu designed the study and performed the PRG-1–CaM binding assays together with Tomohito Fujimoto. Naoya Hatano and Saki Yurimoto contributed to the functional proteomic analysis using LC-MS/MS. Mitsumasa Tsuchiya performed SPR analyses. Hiroyuki Sakagami and Naoki Ohara performed immunohistochemical analyses. Ryoji Kobayashi provided conceptual input and suggestions for the completion of the manuscript. Hiroshi Tokumitsu and Hiroyuki Sakagami wrote the manuscript.
This work was supported, in part, by Grants-in-aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan [grant numbers 21570143 (to H.T.), 19300119 (to H.S.), 21659049 (to H.S.)]; and Kagawa University Characteristic Prior Research Fund 2009 (to H.T.). This work was also supported by Comprehensive Brain Science Network (to H.T. and H.S.).
We thank Yukako Yurie and Momoko Nitta (Kagawa University, Kagawa, Japan) for excellent technical assistance and Dr Mitsu Ikura (Toronto University, Toronto, ON, Canada) for valuable discussions. We also thank Dr Masahiro Kai and Dr Fumio Sakane (Sapporo Medical University, Sapporo, Japan) for assaying the LPP activity of PRG-1. We also thank Dr Takao Ojima, Dr Hiroyuki Tanaka, Dr Kunihiko Konno and Dr Chun Hong Yuan (Hokkaido University, Hokkaido, Japan) for fluorescence and CD measurements.
Abbreviations: CA, cornu ammonis; CaM, calmodulin; CaMKII, Ca2+/CaM-dependent protein kinase II; DAB, 3,3′-diaminobenzidine; DAPI, 4′,6-diamidino-2-phenylindole; DTT, dithiothreitol; EDC, 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide; GST, glutathione transferase; HA, haemagglutinin; HSP70, heat-shock protein 70; LC-MS/MS, liquid chromatography-tandem MS; LPA, lysophosphatidic acid; LPP, lipid phosphate phosphatase; MAP2, microtubule-associated protein 2; MLCK, myosin light-chain kinase; NF200, neurofilament 200; NHS, N-hydroxysuccinamide; Ni-NTA, Ni2+-nitrilotriacetate; PMCA, plasma-membrane Ca2+-ATPase; PRG-1, plasticity-related gene-1; PSD-95, postsynaptic density 95; SPR, surface plasmon resonance
- © The Authors Journal compilation © 2010 Biochemical Society