J Biol Chem, Vol. 274, Issue 43, 30914-30918, October 22, 1999

nArgBP2, a Novel Neural Member of Ponsin/ArgBP2/Vinexin Family That Interacts with Synapse-associated Protein 90/Postsynaptic Density-95-associated Protein (SAPAP)^*

Hiroshi Kawabe, Yutaka Hata^§¶, Masakazu Takeuchi^§, Nobuyuki Ide^§, Akira Mizoguchi, and Yoshimi Takai^§**

From the  Department of Molecular Biology and Biochemistry, Osaka University Graduate School of Medicine/Faculty of Medicine, Suita 565-0871, Japan, the ^§ Takai Biotimer Project, ERATO, Japan Science and Technology Corporation, c/o JCR Pharmaceuticals Co. Ltd., Kobe 651-2241, Japan, the ^¶ Department of Medical Biochemistry, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, Japan, and the  Department of Anatomy and Neurobiology, Graduate School, Kyoto University, Kyoto 606-8315, Japan

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Postsynaptic density (PSD)-95/synapse-associated protein (SAP) 90 and synaptic scaffolding molecule (S-SCAM) are synaptic membrane-associated guanylate kinases. Both the proteins interact with SAP90/PSD-95-associated protein (SAPAP) (also called guanylate kinase-associated protein/Dlg-associated protein). SAPAP is a protein highly enriched in the PSD fraction and may link PSD-95/SAP90 and S-SCAM to Triton X-100-insoluble structures. We found here a novel SAPAP-interacting protein, which was specifically expressed in neural tissue and was present in the postsynaptic density fraction in brain. This protein had a sorbin homology domain in the N terminus, a zinc finger motif in the middle region, and three src homology (SH) 3 domains in the C terminus and was homologous to the ponsin/ArgBP2/vinexin family proteins. We named this protein nArgBP2 because it was the most homologous to ArgBP2. nArgBP2 is a neural member of a growing family of SH3-containing proteins. nArgBP2 bound to the proline-rich region of SAPAP via its third SH3 domain and was coimmunoprecipitated with SAPAP from the extract of rat brain. Furthermore, nArgBP2 was colocalized with SAPAP at synapses in cerebellum. nArgBP2 bound to not only SAPAP but also vinculin and l-afadin, known to bind to ponsin and vinexin. nArgBP2 may be implicated in the protein network around SAPAP in the PSD.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Postsynaptic density (PSD)¹-95/synapse-associated protein (SAP) 90 is a major component of the PSD and has a chimerical molecular structure of three PSD-95/Dlg-A/ZO-1 (PDZ), one src homology (SH) 3, and one guanylate kinase (GK) domain (1, 2). SAP90/PSD-95-associated protein (SAPAP) (also called guanylate kinase-associated protein (GKAP)/Dlg-associated protein (DAP)) was initially identified as a protein interacting with the GK domain of PSD-95/SAP90 (3-5). Later, we have found that SAPAP also interacts with the GK domain of synaptic scaffolding molecule (S-SCAM) (6). S-SCAM has a molecular structure similar to that of PSD-95/SAP90 and is composed of six PDZ, two WW, and one GK domains. Both PSD-95/SAP90 and S-SCAM interact with N-methyl-D-aspartate (NMDA) receptors, neuroligin, and membrane-associated guanylate kinase-interacting protein (MAGUIN)-1 and may be implicated in the assembly of synaptic components (6, 7). SAPAP has at least four isoforms (SAPAP1-4), and these isoforms are about 50% homologous to each other. SAPAP has five repeats of 14 amino acids in the middle region, and the interaction with PSD-95/SAP90 is mediated by these repeats (3). SAPAP has a proline-rich region immediately after the 14 amino acid repeats. The sequence of this region is consistent with the consensus motif for the binding of the SH3 domain but does not bind to that of PSD-95/SAP90 (4, 8). The C-terminal sequences of all the isoforms of SAPAP are well conserved. We have recently found an SH3- and PDZ-containing protein interacting with this region and have named it synamon (9). The sequence of SAPAP predicts a hydrophilic protein, but SAPAP is highly resistant to Triton X-100 extraction in the PSD fraction as well as in transfected cells (10), and the detergent-insolubility is conferred by the N-terminal region. PSD-95/SAP90 and S-SCAM are Triton X-100-soluble, when expressed in transfected cells, but become Triton X-100-insoluble, when coexpressed with SAPAP (10). A brain-enriched guanylate kinase-interacting protein (BEGAIN) is a protein interacting with the GK domain of PSD-95/SAP90 and is Triton X-100-soluble (10). BEGAIN is recruited to the Triton X-100-insoluble fraction, when coexpressed with both of SAPAP and PSD-95/SAP90. Whether the Triton X-100-insolubility of SAPAP reflects the association of SAPAP with the cytoskeleton, the detergent-resistant membrane, or other structures, has not been clarified.

To further identify SAPAP-interacting molecules, we have performed a yeast two-hybrid screening using SAPAP3 as a bait. We have obtained a novel brain-specific protein. This protein has three SH3 domains in the C-terminal region. For the last 2 years, several proteins with three SH3 domains in the C-terminal region have been identified. They are ArgBP2 (11), vinexin (12), and ponsin/c-Cbl-associated protein (CAP) (13-15). ArgBP2, having two splicing variants, ArgBP2A and -B, is a protein interacting with tyrosine kinases, Arg and c-Abl, and shows the highest homology to the protein which we have identified here (11). Therefore, we have named this protein nArgBP2 which stands for neural ArgBP2. Vinexin, having two splicing variants, vinexin  and -, has been identified as a vinculin-binding protein (12). Vinculin is an actin-binding protein which is associated with a Ca²⁺-dependent homophilic cell-cell adhesion molecule, cadherins, through catenins and with a cell-matrix adhesion molecule, integrins, through talin, another actin-binding protein (for a review, see Ref.16). Ponsin, having at least 13 splicing variants, ponsin-1 to -12, has been discovered as a protein interacting with l-afadin and vinculin (13). l-Afadin is an actin-binding protein which is associated with a Ca²⁺-independent homophilic cell-cell adhesion molecule, nectin (17). CAP had been reported as SH3P12 and has been renamed after its interaction with c-Cbl proto-oncogene (14, 15). Ponsin is the same as CAP, and in this paper, we use ponsin to describe this protein just for simplicity. Thus, nArgBP2 is a new member of a growing family of proteins with three SH3 domains interacting with cytoskeletons and signaling molecules.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

DNA Constructions-- Various expression constructs were prepared by conventional molecular biology techniques and PCR method using pBTM116, pCMV Myc, pClneo Myc, pCMV HA, pGex4T-1, and pGex5X-3. pBTM116 SAPAP1-7, -2-1, -3-1, and -4-1 contain the full-length coding region of SAPAP1, -2, -3, and -4, respectively. The following constructs contain the following amino acids of SAPAP1: pCMV Myc SAPAP1-1, 1-992; pCMV Myc SAPAP1-2, 1-477; pCMV Myc SAPAP1-3, 1-556; pCMV Myc SAPAP1-6, 478-992; pCMV Myc SAPAP1-8, 568-992; and pMal SAPAP1-3, 551-992. pCMV HA SAPAP4 contains the full-length coding region of SAPAP4. The following constructs contain the following amino acids of nArgBP2: pGex4T-1 nArgBP2-1, 1-342; pGex5X-3 nArgBP2-11, 957-1124; pGex5X-3 nArgBP2-12, 1032-1196; and pGex5X-3 nArgBP2-13, 1123-1196. pCMV Myc l-afadin and pCMV Myc vinculin-C2 were described previously (13).

Yeast Two-hybrid Screening and cDNA Cloning-- Yeast two-hybrid screening was performed using a rat brain cDNA library as described previously (4). To obtain a full-length cDNA of nArgBP2, a rat brain cDNA library (Stratagene) was screened using [-³²P]dCTP-labeled insert from the pPrey clone as a probe.

Antibodies-- A mouse monoclonal antibody was raised against the product of pGex 4T-1 nArgBP2-1. Rabbit polyclonal antibodies were raised against the products of pGex 5X-3 nArgBP2-11 and pMal SAPAP1-3. Mouse monoclonal anti-Myc and anti-HA antibodies were obtained from American Type Culture Collection and Roche Molecular Biochemicals, respectively.

In Vitro Binding Assay-- COS cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal bovine serum and 100 units/ml penicillin and streptomycin under 5% CO₂ at 37 °C and then transfected with various Myc-tagged constructs of SAPAP1 or pCMV HA SAPAP4 using the DEAE-dextran method. COS cells of three 10-cm dishes were homogenized in 0.1 ml of Buffer A (50 mM Tris/HCl, pH 7.5, containing 1 mM EDTA, 1 mM DTT, 1% (w/v) Triton X-100, 6 M urea, 10 µM -amidinophenylmethanesulfonyl fluoride hydrochloride (APMSF), 10 µg/ml aprotinin, and 10 µg/ml leupeptin). The homogenate was centrifuged at 100,000 × g for 10 min, and the supernatant was dialyzed against 2 liters of Buffer B (20 mM Tris/HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, and 1 mM DTT) containing 10 µM APMSF overnight and centrifuged again at 100,000 × g for 15 min. 1.5 nmol of one of various glutathione S-transferase (GST) fusion constructs of nArgBP2 or GST was applied to 50 µl of glutathione-Sepharose 4B columns (Amersham Pharmacia Biotech) and washed with Buffer C (20 mM Tris/HCl, pH 7.5, containing 1 mM EDTA, 1 mM DTT, 1% (w/v) Triton X-100, 150 mM NaCl, 10 µM APMSF, 10 µg/ml aprotinin, and 10 µg/ml leupeptin) to prepare GST-fusion protein columns. 0.1 ml of the supernatant was applied to each GST-fusion protein column and washed with Buffer D (20 mM Tris/HCl, pH 7.5, containing 1 mM EDTA, 1 mM DTT, 0.3% (w/v) Triton X-100, 150 mM NaCl, 10 µM APMSF, 10 µg/ml aprotinin, and 10 µg/ml leupeptin). The proteins attached to the columns were detected with the anti-Myc tag antibody.

Coimmunoprecipitation-- The synaptic plasma membrane (SPM) fraction was prepared from adult rat brains as described previously (4). The SPM fraction (5 mg of protein) was homogenized in 16 ml of Buffer A. The homogenate was centrifuged at 100,000 × g for 30 min, and the supernatant was dialyzed against 2 liters of Buffer B containing 10 µM APMSF overnight and centrifuged again at 100,000 × g for 30 min. Either 2 µl of the mouse monoclonal anti-Myc antibody, 2 µl of the mouse monoclonal anti-nArgBP2 antibody, 23 µl of the rabbit polyclonal anti-SAPAP antibody, or 23 µl of the rabbit preimmune serum was incubated with 50 µl of protein G-Sepharose 4 Fast Flow beads (Amersham Pharmacia Biotech) and washed with Buffer C. 1 ml of the supernatant was incubated with each antibody fixed on the beads. After the beads were washed with Buffer D, the proteins attached to the beads were detected by immunoblottings.

Immunohistochemistry-- Immunofluorescence microscopy of frozen sections of mouse brain was done as described (13). Briefly, samples were frozen using liquid nitrogen, and the frozen sections were cut in a cryostat. The samples were mounted on glass slides, air dried, and fixed with 95% (v/v) ethanol at 4 °C for 30 min and with acetone at room temperature for 1 min. The samples were washed with phosphate-buffered saline containing 10 mg/ml bovine serum albumin, incubated with either the rabbit polyclonal anti-nArgBP2 or the anti-SAPAP antibody at 1:500 dilution, washed again with phosphate-buffered saline containing 10 mg/ml bovine serum albumin and 0.01% (w/v) saponin, and incubated with Texas Red-conjugated secondary antibodies (Amersham Pharmacia Biotech). The samples were washed with phosphate-buffered saline, embedded, and viewed with a confocal imaging system, MRC-1024 (Bio-Rad Laboratories).

Miscellaneous Procedures-- The subcellular fractionation of rat brain, SDS-polyacrylamide gel electrophoresis, and protein determination were performed as described (4). Western blotting was performed using ECL reagents (Amersham Pharmacia Biotech).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Isolation of nArgBP2-- We performed a yeast two-hybrid screening using pBTM116 SAPAP3-1 as a bait from a rat brain cDNA library. 11 independent positive clones were obtained from 2 × 10⁶ transformants. Among them, four clones were novel and we focused on one clone, pPrey 4015, containing SH3 domains. To obtain a full-length clone of pPrey 4015, we screened a rat brain cDNA library by plaque hybridization using an insert of pPrey 4015 as a probe. No single clone contained a full-length coding region, but two overlapping clones contained a putative initiation methionine and a putative stop codon. The sequences around the putative initiation and stop codons were confirmed by sequencing more than one clone. We also performed PCR using rat brain cDNAs to obtain a product containing a full-length coding region. The full-length clone of p4015 encoded a protein composed of 1,196 amino acids (Fig. 1A). Analysis of the sequence of nArgBP2 revealed several structural characters. The N-terminal region containing the amino acids 190-325 showed similarity to a peptide hormone called sorbin (Fig. 1A, shaded). nArgBP2 had one zinc finger motif in the middle region (Fig. 1A, solid underline) and three SH3 domains in the C-terminal region (Fig. 1A, boxed). The data base search revealed several mammalian proteins which had three SH3 domains in the C-terminal region. ArgBP2, vinexin, and ponsin belong to this family of proteins (11-15). ArgBP2, vinexin, and ponsin have a sorbin homology domain in the N terminus and three SH3 domains in the C terminus, which are highly conserved, but they do not have a zinc finger motif (Fig. 1B). nArgBP2 is the most homologous to ArgBP2 but has two additional regions which are not found in ArgBP2A. There are an additional 68 amino acids (amino acid 1-68) at the N-terminal region of nArgBP2 and an insert of 606 amino acids (amino acid 322-927) at the middle region containing a zinc finger motif, in comparison with the amino acid sequence of ArgBP2A. Except for these regions, nArgBP2 is about 90% identical to ArgBP2A, 50% to ponsin-1, and 40% to vinexin . Considering that nArgBP2 cDNA is cloned from a rat brain cDNA library, and the ArgBP2A cDNA from the human heart library (11), nArgBP2 could be a neural splice variant of the rat counterpart of ArgBP2A.

View larger version (76K): [in this window] [in a new window]   Fig. 1.   Structure of nArgBP2. A, sequence of nArgBP2. Residues are shown in single letter codes. The residues of a sorbin homology domain, a zinc finger motif, and SH3 domains are shaded, solid underlined, and boxed, respectively. The region contained in the clone from the yeast two-hybrid screening is dashed underlined. B, schematic description of nArgBP2, ArgBP2A, ponsin-1, and vinexin .

Interaction of nArgBP2 with SAPAP-- nArgBP2 and SAPAP were coimmunoprecipitated from the rat SPM fraction, suggesting that these proteins interact with each other in vivo (Fig. 2A). nArgBP2 was obtained from the yeast two-hybrid screening with SAPAP3 as a bait, but the interactions of SAPAP1, -2, and -4 with nArgBP2 were confirmed in the -galactosidase assays in yeast (data not shown). Because the original pPrey clone contained only the C-terminal region (Fig. 1, dashed underline), the SH3 domains were expected to be sufficient for the interaction of nArgBP2 with SAPAP. To confirm this assumption, the interactions of HA-tagged SAPAP4 with various GST-fusion proteins containing the SH3 domains of nArgBP2 were tested. The GST-fusion protein with the second and third SH3 domains interacted with SAPAP4, whereas the GST-fusion protein with the first and second SH3 domains did not (Fig. 2B). The GST-fusion protein containing the third SH3 domain alone also bound SAPAP4. The SAPAP-binding activity of the third SH3 domain alone was less than that of the second and third SH3 domains. The reason for this difference is not clear, but the second SH3 domain and the region between the second and third SH3 domains may be also involved in the interaction. Next, to map the nArgBP2-interacting region of SAPAP, various Myc-tagged constructs of SAPAP1 were prepared (Fig. 3A), and the interactions with the GST-fusion proteins containing the second and third SH3 domains of nArgBP2 were tested. Among the five constructs tested, pClneo Myc SAPAP1-1 and SAPAP1-6, both of which contained the proline-rich region after the five 14-amino acid repeats, bound to nArgBP2 (Fig. 3B). These results indicate that nArgBP2 interacts with the proline-rich region in the middle region of SAPAP1.

View larger version (37K): [in this window] [in a new window]   Fig. 2.   Interaction of nArgBP2 with SAPAP. A, coimmunoprecipitation of nArgBP2 and SAPAP from rat brain. The urea/detergent extract of the rat SPM fraction was incubated with either the mouse monoclonal anti-Myc antibody, the mouse monoclonal anti-nArgBP2 antibody, the rabbit preimmune serum, or the rabbit polyclonal anti-SAPAP antibody fixed on protein G-Sepharose beads, and the proteins attached to the beads were immunoblotted with the anti-SAPAP or anti-nArgBP2 antibody. A-a, the immunoblot with the anti-SAPAP antibody; lane 1, the original extract before the incubation; lane 2, the precipitate with the anti-Myc antibody; and lane 3, the precipitate with the anti-nArgBP2 antibody. A-b, the immunoblot with the anti-nArgBP2 antibody; lane 1, the original extract before the incubation; lane 2, the precipitate with the rabbit preimmune serum; and lane 3, the precipitate with the anti-SAPAP antibody. B, SAPAP-interacting domain of nArgBP2. The extract of COS cells expressing HA-tagged SAPAP4 was incubated with either GST-nArgBP2-11 (the first and second SH3 domains), -12 (the second and third SH3 domains), or -13 (the third SH3 domain) fixed on the glutathione beads, and the proteins attached to the beads were detected by the anti-HA antibody. Lane 1, the original sample before the incubation; lane 2, the precipitate with GST-nArgBP2-11; lane 3, the precipitate with GST-nArgBP2-12; and lane 4, the precipitate with GST-nArgBP2-13.

View larger version (49K): [in this window] [in a new window]   Fig. 3.   nArgBP2-interacting domain of SAPAP1. A, schematic description of various Myc-tagged constructs of SAPAP1. The region containing the five repeats of 14 amino acids (amino acid 482 to 556) is shown as an open box. The proline-rich regions are shown as black boxes. Panel B, the extract of COS cells expressing various Myc-tagged constructs of SAPAP1 were incubated with either GST or GST-nArgBP2-12 fixed on the glutathione beads, and the proteins attached to the beads were detected with the anti-Myc antibody. Lanes 1-3, pCMV Myc SAPAP1-1; lanes 4-6, pCMV Myc SAPAP1-2; lanes 7-9, pCMV Myc SAPAP1-3; lanes 10-12, pCMV Myc SAPAP1-6; lanes 13-15, pCMV Myc SAPAP1-8; lanes 1, 4, 7, 10, and 13, the original extract before the incubation; lanes 2, 5, 8, 11, and 14, the precipitate with GST; lanes 3, 6, 9, 12, and 15, the precipitate with GST-nArgBP2-12.

Tissue Distribution and Subcellular Localization of nArgBP2-- We next examined tissue distribution of nArgBP2. Northern blot analysis was performed using the probe corresponding to the amino acid residues 453-958 of nArgBP2. A 7.4-kilobase transcript was detected only in brain, and no message was detected in heart, spleen, lung, liver, skeletal muscles, or kidney (Fig. 4A). Western blot analysis using this antibody recognized a protein with a molecular mass of 160 kDa only in brain (Fig. 4B). In rat brain subcellular fractions, nArgBP2 was detected in the synaptosomal cytosol but was the most enriched in the PSD fraction (Fig. 5A). SAPAP was also the most enriched in the PSD fraction but was not detected in the synaptosomal cytosol fraction (Fig. 5B). In rat cerebellum, nArgBP2 was detected in the soma and the dendrites of Purkinje cells and in the glomerulus of the granular layer (Fig. 6A). The pattern of nArgBP2 immunoreactivity was punctate, and the distribution of SAPAP in the same region was similar to that of nArgBP2 (Fig. 6B). These results suggest that nArgBP2 is a neural protein localized at cell bodies, neurites, and synapses.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Tissue distribution of nArgBP2. A, Northern blot analysis of nArgBP2. A blot with 2 µg of mRNA from each rat tissue was hybridized with a uniformly labeled probe corresponding to the amino acids 453-958 of nArgBP2 and exposed to Fuji image plate overnight. B, Western blot analysis of nArgBP2. Homogenates of various rat tissues (50 µg of protein) were immunoblotted with the mouse monoclonal anti-nArgBP2 antibody. Lane 1, heart; lane 2, brain; lane 3, spleen; lane 4, lung; lane 5, liver; lane 6, skeletal muscle; and lane 7, kidney.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Subcellular localization of nArgBP2 in rat brain. Equal aliquots of the subcellular fractions of rat brain (25 µg of the protein each) were immunoblotted with the mouse monoclonal anti-nArgBP2 or the rabbit polyclonal anti-SAPAP antibody. A, the immunoblot with the mouse monoclonal anti-nArgBP2 antibody. B, the immunoblot with the rabbit polyclonal anti-SAPAP antibody. Lane 1, the homogenate fraction; lane 2, the crude synaptosomal fraction; lane 3, the nuclear pellet fraction; lane 4, the synaptosomal cytosol fraction; lane 5, the crude synaptosomal pellet fraction; lane 6, the lysed synaptosomal membrane fraction; lane 7, the crude synaptic vesicle fraction; lane 8, the SPM fraction; lane 9, the 0.5% (w/v) Triton X-100-soluble fraction of the SPM; lane 10, the 0.5% (w/v) Triton X-100-insoluble fraction of the SPM; lane 11, the 1% (w/v) Triton X-100-soluble fraction of the SPM; and lane 12, the 1% (w/v) Triton X-100-insoluble fraction of the SPM.

View larger version (87K): [in this window] [in a new window]   Fig. 6.   Localization of nArgBP2 and SAPAP in vivo. Frozen sections of rat cerebellum were incubated with the rabbit polyclonal anti-nArgBP2 or the anti-SAPAP antibody and visualized with the Texas Red-conjugated secondary antibody. A, localization of nArgBP2 in rat cerebellum. B, localization of SAPAP in rat cerebellum. Open arrowhead, Purkinje cell; open arrow, glomerulus. The scale bar indicates 30 µm.

Interaction of nArgBP2 with l-Afadin and Vinculin-- The sequences of the SH3 domains are highly conserved among the members of ponsin/ArgBP2/vinexin family (11-15). The interactions of ponsin with l-afadin and vinculin are mediated via the C-terminal SH3 domains (13). We finally examined whether nArgBP2 also interacted with l-afadin and vinculin. Myc-tagged l-afadin or vinculin was expressed in COS cells, and the extracts of COS cells were incubated with the GST-fusion constructs of nArgBP2. The GST-fusion protein containing the second and third SH3 domains of nArgBP2 bound l-afadin and vinculin (Fig. 7, A and B).

View larger version (44K): [in this window] [in a new window]   Fig. 7.   Interaction of nArgBP2 with l-afadin and vinculin. The extract of COS cells expressing Myc-tagged l-afadin or vinculin was incubated with either GST or GST-nArgBP2-12 (the second and third SH3 domains) fixed on the glutathione beads. A, interaction of nArgBP2 with l-afadin. B, interaction of nArgBP2 with vinculin. Lane 1, the original extract before incubation; lane 2, the precipitate with GST; and lane 3, the precipitate with GST-nArgBP2-12.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this paper, we have identified a novel neural protein, nArgBP2, which interacts with SAPAP. The data base search using BLASTp has revealed that 11 mammalian proteins in GenBank^TM have three SH3 domains in the C-terminal region. These 11 proteins include 4 independent proteins, namely ArgBP2, vinexin, ponsin/CAP, and a protein named SH3-containing adaptor protein (11-15). The multiple alignment using CLUSTRAW shows that nArgBP2 and ArgBP2 form a subfamily different from that of vinexin and ponsin (Fig. 8).

View larger version (8K): [in this window] [in a new window]   Fig. 8.   Dendrogram analysis of the ponsin/ArgBP2/vinexin family using the CLUSTRAW program. The diagram depicts the degree of sequence difference between the indicated proteins. The length of each horizontal line corresponds to the sequence distance between the different proteins.

The experiments using the GST-fusion proteins of the SH3 domains of nArgBP2 and Myc-tagged constructs of SAPAP1 suggest that the interaction between nArgBP2 and SAPAP is mediated via the third SH3 domain of nArgBP2 and the proline-rich region of SAPAP, although we cannot exclude the possibility that the regions of nArgBP2 besides the SH3 domains also bind to some region of SAPAP. The third SH3 domain of nArgBP2 interacts with l-afadin and vinculin in vitro, but it remains to be determined whether nArgBP2 interacts simultaneously with vinculin or l-afadin and SAPAP. Because these molecules interact with the same domain of nArgBP2, they may compete in the binding to nArgBP2, and there may be some regulatory mechanism, such as phosphorylation of nArgBP2, to determine which molecule binds to nArgBP2. The precise role of nArgBP2 in organization of synaptic components in the PSD is not known, but it may be intriguing to speculate that by interacting with SAPAP, nArgBP2 may link PSD-95/SAP90, S-SCAM, synamon, and their interacting molecules to cell adhesion systems, such as the cadherin-catenin-vinculin (16) and nectin-afadin (13, 17) systems, and to signaling molecules, such as c-Cbl, Arg, and c-Abl (for reviews, see Refs. 18 and 19).

SAPAP is highly enriched in the PSD fraction and has the ability to recruit PSD-95/SAP90 and S-SCAM to the Triton X-100-insoluble structures in transfected cells, suggesting that SAPAP may link PSD-95/SAP90 and S-SCAM to the PSD in vivo (10). SAPAP is composed of three regions with distinct characters. The N-terminal region is highly resistant to Triton X-100 extraction. The middle region contains five repeats of 14 amino acids followed by a proline-rich region. The C-terminal region also has proline-rich regions and is Triton X-100-soluble. PSD-95/SAP90 and S-SCAM bind to the five repeats of 14 amino acids. Synamon binds to the C terminus of SAPAP (9). The nArgBP2-interacting region is different from either the PSD-95/SAP90- and S-SCAM-interacting region or the synamon-interacting region. nArgBP2 may form a complex with PSD-95/SAP90, S-SCAM, and synamon via SAPAP and may link the actin cytoskeleton to the components of the PSD.

	ACKNOWLEDGEMENTS

We thank Dr. Kenji Mandai (ERATO) for the constructs of pCMV Myc l-afadin and pCMV Myc vinculin-C2. We also thank Dr. Hiroto Kawashima (Department of Bioregulation, Biomedical Research Center, Osaka University Medical School, Suita, Japan) for kind advice on making monoclonal antibodies.

	FOOTNOTES

^* The work performed at Osaka University Graduate School of Medicine/Faculty of Medicine was supported by grants-in-aid for Scientific Research and for Cancer Research from the Ministry of Education, Science, Sports, and Culture, Japan (1998) and by grants from the Human Frontier Science Program (1998).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in this paper has been submitted to the GenBank^TM/EMBL Data Bank with accession number(s) AF026505.

^** To whom correspondence should be addressed. Tel.: 81-6-6879-3410; Fax: 81-6-6879-3419; E-mail: ytakai@molbio.med.osak-u.ac.jp.

	ABBREVIATIONS

The abbreviations used are: PSD, postsynaptic density; SAP, synapse-associated protein; PDZ, PSD-95/Dlg-A/ZO-1; SH, src homology; GK, guanylate kinase; SAPAP, SAP90/PSD-95-associated protein; GKAP, guanylate kinase-associated protein; DAP, hDLG-associated protein; S-SCAM, synaptic scaffolding molecule; NMDA, N-methyl-D-aspartate; MAGUIN, membrane-associated guanylate kinase-interacting protein; BEGAIN, brain-enriched guanylate kinase-interacting protein; CAP, c-Cbl-associated protein; APMSF, -amidinophenylmethanesulfonyl fluoride hydrochloride; GST, glutathione S-transferase; SPM, synaptic plasma membrane; DTT, dithiothreitol.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES