HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 276, Issue 35, 33093-33100, August 31, 2001

This Article Abstract Full Text (PDF) All Versions of this Article: 276/35/33093    most recent M105792200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jannatipour, M. Articles by Rouleau, G. A. Search for Related Content PubMed PubMed Citation Articles by Jannatipour, M. Articles by Rouleau, G. A. Social Bookmarking                      What's this?

Schwannomin Isoform-1 Interacts with Syntenin via PDZ Domains^*

Mehrdad Jannatipour^§, Patrick Dion, Saad Khan, Hitesh Jindal^¶, Xueping Fan, Janet Laganière, Athar H. Chishti^¶, and Guy A. Rouleau

From the  Center for Research in Neuroscience, McGill University and the Montreal General Hospital, 1650 Cedar Avenue, Montreal, Quebec, Canada H3G 1A4 and the ^¶ Section of Hematology-Oncology Research, St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston, Massachusetts 02135

Received for publication, June 21, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The neurofibromatosis type 2 gene (NF2) is involved in the pathogenesis of benign tumors of the human nervous system. The NF2 protein, called schwannomin or merlin, is inactivated in virtually all schwannomas and meningiomas. The molecular mechanisms by which schwannomin functions as a tumor suppressor is unknown but believed to involve plasma membrane-cytoskeletal interactions. Two major alternatively spliced isoforms of schwannomin differing in their C termini have been reported. Using the yeast two-hybrid system, we have identified syntenin as a binding partner for schwannomin isoform-1 (sch-1). Syntenin is an adapter protein that couples transmembrane proteoglycans to cytoskeletal components and is involved in intracellular vesicle transport. The C terminus 25 amino acids of sch-1 and the two PDZ domains of syntenin mediate their binding, and mutations introduced within the VAFFEEL region of sch-1 defined a sequence crucial for syntenin recognition. We have showed that the two proteins interacted in vitro and in vivo and localized underneath the plasma membrane. Fibroblast cells expressing heterologous antisense syntenin display alterations in the subcellular distribution of sch-1. Together, these results provide the first functional clue to the existence of schwannomin isoforms and could unravel novel pathways for the transport and subcellular localization of schwannomin in vivo.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Neurofibromatosis type 2 (NF2)¹ is an autosomal dominant disorder that leads to the development of schwannomas, meningiomas, ependymomas, and other tumors of the central nervous system (1, 2). Inactivating mutations of the NF2 gene resulting in loss of function of its protein product causes the disease. The NF2 gene product is referred to as NF2 protein or schwannomin or merlin. We will use the nomenclature schwannomin-1 (sch-1) throughout the text. Tumors from patients with inherited disease (3, 4), as well as sporadic schwannomas and meningiomas, show loss of heterozygosity of the NF2 gene region on chromosome 22 and inactivating mutations on the second NF2 allele, strongly suggesting that schwannomin functions as a tumor suppressor protein (5, 6). Schwannomin bears a striking similarity to ERM proteins (ezrin, radixin, and moesin) that are members of the protein 4.1 superfamily. ERM proteins function as molecular adaptors linking integral membrane proteins to the cytoskeleton (7). Members from this family modulate cell shape and membrane specializations by regulating the distribution of adhesion molecules on the cell surface (8, 9). The amino-terminal end of ERM proteins, termed as the FERM domain, has been shown to interact with the plasma membrane proteins such as CD43, CD44, intercellular adhesion molecule-1, and intercellular adhesion molecule-2. ERM proteins also contain actin-binding sites mapped to the carboxyl-terminal domains (8, 10-12). This actin-binding site is highly conserved, but the corresponding site in schwannomin is partially divergent, indicating a slightly different function. Schwannomin is localized to actin-rich plasma membrane processes (3, 4, 13) and also binds microtubules (14). Schwannomin has also been shown to interact with numerous membrane-associated proteins such as CD44 (15), NHERF (EBP50) (16), II-spectrin (17), and 1-integrin (18). These observations suggest that schwannomin exerts its tumor suppressor effects presumably by modulating some aspects of the receptor-cytoskeletal linkage critical for cell growth and adhesion pathways.

Several alternatively spliced transcripts of schwannomin have been identified. The two most prevalent isoforms differ by the presence (isoform-1 or sch-1) or the absence (isoform-2 or sch-2) of exon 16 (19, 20). Human sch-1 codes for 595 amino acids whereas shorter sch-2 consists of 590 amino acids. Both the primary structure and the tissue distribution of each isoform are conserved among many species. In addition, these two transcripts show distinct tissue specificity and temporal expression during rat development (19). These features underscore the importance of distinct schwannomin C termini in cellular function. Moreover in mouse NIH-3T3 or in rat schwannoma cell lines, the heterologous expression of sch-1 but not of sch-2 decreases the growth rate (21-23), stressing the importance of the C termini of sch-1 in the protein tumor suppressor activity (6).

Based of these observations, we set out to search for proteins that specifically interact with the carboxyl-terminal end of sch-1. Here we report the identification of syntenin, an adapter protein with two tandem PDZ domains, as a novel binding partner that specifically interacts with the C terminus of sch-1. Syntenin is known to bind to several proteins including the cytoplasmic domains of cell surface syndecans (24), EphA7 of ephrin receptor tyrosine kinases (25), and the C termini B class ephrin ligands (26). In addition, syntenin plays a role in the targeting and maturation processes of proTGF- from early compartments of the secretory pathway to the cell surface (27). More recently, syntenin has been localized to the apical recycling compartment of MDCK epithelial cells suggesting a role in the intracellular trafficking and targeting pathways (28). Direct and specific association between syntenin and schwannomin isoform-1 begins to suggest a functional role for the differentially spliced isoforms of schwannomin and may have implications in the assembly of cytoskeletal complexes, trafficking, and subcellular targeting events necessary for the regulation of cell growth and adhesion processes.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Yeast Two-hybrid Screening and Cloning of Syntenin-- Two constructs of human sch-1 were engineered; one consists of the carboxyl-terminal 30 amino acids (C30-1), and the other consists of the carboxyl-terminal 125 amino acids (C125-1). These respective cDNA clones were amplified by polymerase chain reaction (PCR) and cloned into the EcoRI-XhoI sites of the pEG202LexA (Origene Technologies, Inc.) fusion plasmid vector. pEG202LexA-C30-1 insert was generated with primers 5'-AAGTTCTCGAGCTCCCACGGAAACCCCCAGCCA-3' and 5'-ACAATGAATTCTCCGACAGGGGTGGCAGCAGCAAGCAC-3'. pEG202LexA-C125-1 insert was generated with primers 5'-AAGTTCTCGAGCTCCCACGGAAACCCCCAGCCA-3' and 5'-AGCGAGAATTCAAGCAA AGCTCCTGGAGATTGCCACC-3'. Using the same set of primers as used for pEG202LexA-C30-1 construct, the cDNA for human schwannomin isoform-2 (clone 767295; GenBank^TM accession number AA418421) was amplified. This construct, designated as pEG202LexA-C25-2, encodes the carboxyl-terminal 25 amino acids of sch-2 (C25-2). The human fetal kidney cDNA library cloned into vector pJG4-5 was screened (2 × 10⁷ primary transformants) in yeast strain EGY48. The insert from one of the selected clones, syntenin (120), was completely sequenced. A BLAST search identified I.M.A.G.E. clones 33203 and 566968 as coding for the full-length cDNA (2,193 base pairs) of syntenin, which were purchased from Research Genetics, Inc.

Recombinant cDNA Constructs-- Different LexA-syntenin fusion proteins containing defined regions of syntenin were made by PCR amplification and cloned into EcoRI-XhoI restriction sites of pJG4-5 vector. For confocal microscopy studies, peGFP-C1 or peGFP-C2 vectors (CLONTECH) were used to fuse different regions of schwannomin or syntenin coding sequences to the carboxyl-terminal end of enhanced green fluorescent protein (eGFP). The carboxyl-terminal end of sch-1 (peG202LexA-C30-1 or peG202LexA-C125-1) and the coding region of syntenin consisting of two PDZ domains (120) were subcloned in frame as an EcoRI-XhoI fragment into peGFP-C2 vector. Full-length syntenin fused to the eGFP was constructed by subcloning a BamHI-XhoI fragment from pGEX-5X-1-syntenin (see below) into BglII-SalI sites of pEGFP-C2. Full-length human sch-1 fused to eGFP was constructed by PCR amplification and introduction of restriction sites BamHI-EcoRI and cloning into BglII-EcoRI of pEGFP-C1 expression vector. The antisense syntenin cDNA was cloned into pTRE expression vector for use in the tetracycline-regulated expression system (CLONTECH). Full-length syntenin cDNA (I.M.A.G.E. clone 566968 constructed in pBS-SK) and pTRE vector were digested with the restriction enzymes XhoI and SacII, respectively. After blunt ending with T4 DNA polymerase, both plasmids were digested with EcoRI. In the case of syntenin, because of an internal restriction site, partial EcoRI digestion was carried out to isolate full-length syntenin cDNA for cloning into pTRE (pTRE-syntenin (AS)).

Northern Blotting-- Multiple tissue northern blot was purchased from CLONTECH Inc. A single-stranded full-length human syntenin cDNA probe was prepared, and hybridization was performed according the manufacturer's recommendations.

Expression and Purification of GST-Syntenin Fusion Proteins-- Full-length syntenin was amplified by PCR (5'-TCGGATCCCCATGTCTCTCTATCCATCTC TCGAAG-3' and 5'-CCGCTCGAGTTAAACCTCAGGAATGGTGTGG-3') and cloned into BamHI-XhoI restriction sites of plasmid pGEX-5X-1 (Amersham Pharmacia Biotech). The syntenin coding region containing both PDZ domains, syntenin (120), was subcloned as an EcoRI-XhoI fragment into pGEX-5X-1 expression vector. The Escherichia coli BL21 strain and bulk GST purification module (Amersham Pharmacia Biotech) were used for expression and purification of GST fusion proteins. Purified fusion proteins GST-syntenin (FL) and GST-syntenin (120) were confirmed by Western blot using immunoaffinity-purified anti-syntenin-N and anti-syntenin-C antibodies.

In Vitro Binding Assays-- For pull-down experiments, the interaction between syntenin and schwannomin was examined by incubation of purified GST-syntenin fusion protein (1 µg) and biotinylated sch-1 peptide (0.1 µg) in 1 ml of phosphate-buffered saline containing 75 mM NaCl on ice for 1 h. Immunopure immobilized streptavidin beads (50 µl of 50% agarose slurry from Pierce) was added, and incubation was continued for another hour on ice. Beads were washed four times with phosphate-buffered saline containing 75 mM NaCl, boiled for 5 min in 50 µl of SDS sample buffer, and resolved by 10% SDS-PAGE, followed by Western blot analysis using an anti-syntenin-C antibody. For blot overlay assays, purified GST protein and GST-syntenin (FL) fusion proteins (1 µg each) were resolved by SDS-PAGE (10%) and transferred onto a nitrocellulose membrane. The transferred protein were visualized using Ponceau S, and the membranes were blocked in buffer containing 6% casein, 1% polyvinyl pyrrolidone-40, 10 mM EDTA in phosphate-buffered saline, pH 7.4. Blots were incubated overnight at 4 °C with biotinylated peptide corresponding to the 16 amino acids carboxyl-terminal of sch-1 (biotin-LTLQSAKSRVAFFEEL) at a concentration of 2 µg/ml in phosphate-buffered saline containing 75 mM NaCl. The blots were washed three times for 10 min with PBS containing 75 mM NaCl, before incubation for 1 h at room temperature with streptavidin-horseradish peroxidase (Pierce; 1:1500). After rinsing with PBS containing 75 mM NaCl, the presence of the biotinylated peptide was tested using ECL Western blotting detection reagent (PerkinElmer Life Sciences).

Binding of Native Schwannomin with the GST-Syntenin Fusion Protein-- Glutathione-Sepharose beads coupled to GST alone, or GST-syntenin (FL) was blocked with 0.5% bovine serum albumin in binding buffer (50 mM Tris-HCl, pH 7.5, 100 mM KCl, 1 mM EGTA, 1 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride). This step was followed by incubation either at 4 °C overnight or at 23 °C for 90 min with native schwannomin protein (4.8 µg) purified from human erythrocyte membranes (29). After incubation, the beads were washed three times with the binding buffer containing 1% Nonidet P-40 detergent and once with the binding buffer without Nonidet P-40 detergent. Proteins were resolved by 10% SDS-PAGE and transferred onto nitrocellulose membrane. Blots were blocked with blocking buffer (6% casein, 1% polyvinyl pyrrolidone-40, 10 mM EDTA in PBS, pH 7.4) and incubated with anti-schwannomin-1 antibody (C-18; SC-332; Santa Cruz Biotechnology) diluted 1:1000 in blocking buffer for 2 h at room temperature. After incubation, the blots were washed with buffer containing 10 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 0.1% Tween 20. Finally blots were incubated for 1 h with goat anti-rabbit IgG-horseradish peroxidase and visualized using ECL detection reagent.

Production of Antibodies Against the Carboxyl- and Amino-terminal Segments of Syntenin-- A peptide corresponding to 30 amino acids from the amino-terminal end of syntenin, NH₂-MSLYPSLEDLK VDKVIQAQTAFSANPANPA-COOH, was used to produce anti-syntenin-N. A second peptide corresponding to 30 amino acids from the carboxyl-terminal end of syntenin, NH₂-IMPAFIFEHIIKR MAPSIMKSLMDHTIPEV-COOH, was used to produce anti-syntenin-C. Both antibodies were produced in rabbits, and in both cases the serum was immunopurified using their respective immobilized immunogen before their uses in experiments.

Coimmunoprecipitation of Syntenin and Schwannomin-- HeLa cells (1 × 10⁶) were lysed in 1 ml of immunoprecipitation buffer (IB) (50 mM Hepes, pH 7.4, 150 mM NaCl, 5 mM EDTA), 1% Nonidet P-40, and protease inhibitors and centrifuged for 1 h at 4 °C. Proteins from the supernatant (500 µg) were incubated with immunopurified anti-syntenin-N (2 µg) and protein A/G-Sepharose beads (50 µl of 50% slurry solution; Amersham Pharmacia Biotech) overnight at 4 °C. The immunoprecipitated material was washed four times with IB buffer containing 0.1% Nonidet P-40 before the bound proteins were eluted by boiling in SDS sample buffer. Finally the samples were resolved on a 7.5% SDS-PAGE, transferred onto a nitrocellulose filter (Schleicher & Schuell), and immunoblotted at 4 °C overnight with polyclonal anti-sch-1 (C-18; SC-332; Santa Cruz Biotechnology Inc, 1:5000). Detection for 1 h at room temperature with horseradish peroxydase-conjugated anti-rabbit antibody (1:5000) followed and ECL detection reagent (PerkinElmer Life Sciences) was used to detect bound antibodies on Western blots.

Cell Culture, Transfection, and Fluorescence Microscopy-- HeLa cells were cultured in complete Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Stable tet-off cell lines (3T3-L1) were established according to the manufacturer's recommendations (CLONTECH Inc.). Stable tet-off cells (3T3-L1) were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum with 2 µg/ml of doxycycline and 200 µg/ml of hygromycin for 24 h. For the inductions the cells were washed with Hanks' balanced salt solution and then transferred in medium without doxycycline for 5 days. During induction tet system-approved fetal bovine serum (CLONTECH Inc.) was used. For immunocytochemistry, the cells were fixed (3% formaldehyde, 1× PBS) and permeabilized (0.2% Triton X-100 in PBS) for 15 min at room temperature. This was followed by a blocking step with 10% goat serum for 1 h at room temperature and then incubation overnight at 4 °C with polyclonal anti-sch-1 (C-18; 0.4 µg/ml). After this the cells were rinsed and incubated first with biotinylated anti-rabbit antibody (1:750 dilution) and then with streptavidin-DTAF (1:500 dilution) for 1 h at room temperature (Jackson Immunoresearch Laboratories Inc.). Slides were finally mounted and viewed with a polyvar fluorescent microscope. For confocal microscopy studies, glass coverslips were seeded with 2 × 10⁵ HeLa cells per well. After incubation at 37 °C for 18 h, cells were transfected and allowed to recover for 24 h. Transfected HeLa cells were fixed (3% formaldehyde in PBS) at room temperature for 15 min and assessed for localization of green fluorescent protein using a confocal laser-scanning microscope (Leica).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Syntenin Interacts with the C Terminus of Schwannomin Isoform-1-- We performed a yeast two-hybrid screen using the carboxyl-terminal 30 amino acids of sch-1 as bait (NH₂-SDRGGSSKHNTI KKLTLQSAKSRVAFFEEL-COOH, pEG202LexA-sch-C30-1) to screen an oligo(dT)-primed human kidney cDNA library (30). It is noteworthy that a second screen was also attempted using a longer segment of sch-1 as bait (125 amino acids; pEGF202LexA-sch-C125-1) but was abandoned because of the invariable self-activation of this longer bait. Nonetheless the screen using the carboxyl-terminal 30 amino acids resulted in the isolation of eight clones specifically interacting with sch-C30-1 but not with LexA or an unrelated protein. Subsequent sequencing of those clones showed that they all encompassed the coding sequence of a protein known as syntenin or mda-9 (24, 27). Full-length cDNA for syntenin (2,193 base pairs) was then obtained from Integrated Molecular Analysis of the human genome and its expression (I.M.A.G.E.). Syntenin cDNA encodes a 298-amino acid protein containing two PDZ domains in tandem. These two domains span amino acids 113 to 193 (PDZ1) and 198 to 273 (PDZ2), respectively (Fig. 1A). All eight clones isolated from the screen lacked the first 119 amino terminus residues of syntenin, but they all contained the two complete PDZ domains. To make sure that the observed syntenin-schwannomin interaction was indeed specific to isoform-1, we fused LexA to the last 25 amino acids from sch-2 (SDRGGSSKHNTIKKPQAQGRRPICI-COOH; pEG202LexA-sch-C25-2) and observed no interaction with the PDZ domains present in a syntenin (120) (see below for details). A search in the BLAST data base for expressed sequence tags yielded syntenin-matching sequence tags in many fetal and adult human tissue cDNA libraries suggesting widespread expression of syntenin at the mRNA level. To determine the tissue distribution of syntenin, a multiple tissue RNA blot (CLONTECH) was probed with a complete syntenin cDNA probe. The results indicate that syntenin is present as a single species of 2.4 kilobases with strongest expression observed in pancreas, skeletal muscle, placenta, and heart with limited expression seen in brain, lung, liver, and kidney (data not shown).

View larger version (29K): [in this window] [in a new window]   Fig. 1.   A, amino acid sequence of human syntenin isolated during the screen for sch-1 interacting partners. The human fetal kidney library constructed in vector pJG4-5, which was used for the yeast two-hybrid screen, generated clones with sequences coding for residues located between position 120 and 298. The complete sequence was obtained from the cDNA clone 566968 at I.M.A.G.E. The gray boxes indicate the two distinct PDZ domains within the protein; PDZ domain 1 is underlined once, and PDZ domain 2 is underlined twice. B, details of the syntenin constructs used to analyze the interaction between syntenin and sch-1. In this section and in C, the extent of the interaction is expressed as -galactosidase (-gal) activity and growth on Leu plates. +, interaction detected; ±, interaction substantially lower than + detected; , interaction not detected. C, association between syntenin and the products of sch-1 constructs where specific amino acids located within the VAFFEEL region alanine were independently substituted for alanine. The sequence shows the amino acid residues forming the carboxyl-terminal portion of schwannomin isoform-1. The amino acid residues specific to sch-1 are underlined.

Both PDZ Domains of Syntenin Are Required for Interaction with Schwannomin Isoform-1-- Sequences derived from all previously obtained clones showed that the region upstream to the first PDZ domains of syntenin is not involved in the association with sch-1. We used the vector pJG4-5 to subclone regions of syntenin downstream from the beginning of the first PDZ domain and narrow down the region necessary for maintaining the association between syntenin and the carboxyl-terminal end of sch-1 (Fig. 1B). First we tested a construct, syntenin (120), where both the upstream and downstream sequences to the PDZ domains were missing. The loss of the downstream sequences flanking the PDZ domains of syntenin dramatically weakened its interaction with sch-1. To test whether the presence of both PDZ domains was essential for the interaction with sch-1 to occur, two separate constructs were prepared. One construct, syntenin 113-193, contained only the first PDZ domain whereas the syntenin 198-273 construct contained only the second PDZ domain. The results from the interactions using these single PDZ domain constructs indicated that they do not sustain the association with schwannomin. Additional constructs were engineered, syntenin (1) and syntenin (172), so that each one of the PDZ domains was complemented by partial flanking sequences from the second PDZ domain. These partially complemented PDZ domains constructs of syntenin were used to test whether they could rescued the interaction of with sch-1, and both showed that no interactions between their product and sch-1 occurred (Fig. 1B). Finally, we examined the requirement of the last 25 amino acids of syntenin for binding to sch-1. A new construct of syntenin was made, syntenin (1), lacking only the last 42 amino acids, also failed to interact with sch-1. In fact, absolutely no interaction between the product from this construct and schwannomin was observed. By comparison, an interaction between the product from the syntenin (120) construct and schwannomin was observed, although it was weaker when compared with the interactions observed between schwannomin and either syntenin (FL) or syntenin (120) products (Fig. 1B). Therefore, the 21 amino acids residues located between positions 257 and 273 are necessary for the interaction between syntenin and sch-1 to occur as their removal completely prevented the interaction between the two proteins.

Schwannomin-Syntenin Interaction Is Mediated by the VAFFEEL Segment of Schwannomin Isoform-1-- By generating a series of alanine substitutions within the last 30 amino acids sch-1, we further analyzed the interactions between syntenin and sch-1. These interactions were again measured using the yeast two-hybrid system (Fig. 1C). These modified amino acids within the VAFFEEL domain showed that substitutions of both phenylalanine residues at positions 3 and 4 for alanine can completely abrogate the interaction with syntenin. Not all substitution within the VAFFEEL completely abrogated the interaction. Alanine substitution of either valine or glutamic acid residues at respective positions 6 and 1 of sch-1 simply weakened it as seen by the completed absence of blue from the -galactosidase reporter gene in combination with weaker growth over Leu plates. The last alanine substitutions within the VAFFEEL of both the leucine and glutamic acid residues at positions 0 and 2, respectively, had no effect on the interaction between syntenin sch-1. The interactions between syntenin and any of the products from sch-1 constructs where either one of the remaining 23 amino acids (located outside of the VAFFEEL region) were substituted for alanine were all tested and found to be positives (data not shown). All together, these results indicate that there is a specific signature sequence at the C terminus of sch-1, which is crucial for its recognition by the PDZ domains of syntenin.

In Vitro Interaction between Syntenin and Schwannomin Isoform-1 by Pull-down and Blot Overlay Assays-- Both to confirm the true nature of the interaction between syntenin and sch-1 and to further investigate the role of the VAFFEEL domain, we tested it in vitro. To do this we used a biotinylated sch-1 peptide containing the VAFFEEL domain (biotin-LTLQSAKS RVAFFEEL) and both full-length syntenin or syntenin (120) GST fusion proteins. These GST fusion proteins were independently incubated in the presence of the sch-1 biotinylated peptide (Fig. 2A, lanes 1-9) so that the peptide-protein complexes could be pulled down using streptavidin beads. Before loading the proteins were boiled for 5 min in 50 µl of SDS sample buffer so that the complexes would be disrupted. Before any pull-down, a single syntenin band corresponding to either GST-syntenin (FL) or GST-syntenin (120) was detected when using both the anti-GST antibody (Fig. 2A, lanes 2 and 3) and the anti-syntenin-C antibody (Fig. 2A, lanes 5 and 6). When the streptavidin bead pull-down was performed (Fig. 2A, lanes 7-9), and the proteins were detected using the same anti-syntenin-C antibody (Fig. 2A, lanes 8 and 9), the detected bands matched exactly with the bands detected before the pull-down. This was true for both syntenin (FL) (Fig. 2A, lanes 3, 6, and 9) and syntenin (120) (Fig. 2A, lanes 2, 5, and 8). The biotinylated sch-1 peptide did not interact with the GST alone (Fig. 2A, lane 7).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   In Vitro interaction between GST-syntenin fusion proteins and a biotinylated sch-1 peptide. A, purified GST-syntenin (FL) and GST-syntenin (120) fusion proteins were independently incubated in the presence of biotinylated sch-1 peptides (lanes 1-9). On the third gel (lanes 7-9) a streptavidin bead pull-down was performed on the combined proteins. The material was loaded on SDS-PAGE gels, transferred, and detected using specific antibodies. When no pull-down was performed (lanes 1-6), the fusion proteins were detected with anti-GST (lanes 1-3) and anti-syntenin-C antibodies (lanes 4-6). In the case where the combined proteins were pulled down they were detected using an anti-syntenin-C antibody (lanes 7-9). B, blot overlay interaction assay between immobilized GST-syntenin (FL) and a biotinylated sch-1 peptide. Purified GST protein alone and GST-syntenin (FL) fusion proteins were resolved on a SDS-PAGE gel and transferred to a nitrocellulose membrane. Before incubation of the membrane with the biotinylated peptide, the membrane was treated with Ponceau S to reveal all proteins (lane 1 and 2). The same membrane was incubated, overnight at 4 °C, in the presence of a biotinylated sch-1 peptide, and the interaction with the fusion protein was detected with streptavidin-horseradish peroxidase and ECL Western blotting detection reagent (lane 3 and 4).

The in vitro interaction between syntenin and schwannomin was further tested using a blot overlay assays. This method allows for the detection of interactions between a biotinylated sch-1 peptide and GST-syntenin (FL) fusion proteins immobilized on nitrocellulose. Binding of biotinylated sch-1 peptide is detected using the streptavidin-horseradish peroxidase and the ECL Western blotting detection reagent (PerkinElmer Life Sciences). The location and quality of the GST fusion proteins was first established by the Ponceau S staining (Fig. 2B, lanes 1 and 2). Specific binding resulting from the interaction between two proteins was observed in the lane containing GST-syntenin (Fig. 2B, lane 4). Purified GST control did not show any detectable binding with the biotinylated peptide (Fig. 2B, lane 3). This result was further investigated using another blot overlay assay where the nitrocellulose membrane containing full-length GST-syntenin was incubated with native schwannomin isolated from human erythrocyte membranes (Fig. 3A). Binding was examined under two different incubation conditions, 4 °C overnight or 23 °C for 90 min. Native schwannomin specifically bound to the GST-syntenin but did not interact with GST alone (Fig. 3A, lane 2 and 4). The schwannomin-syntenin interaction was significantly stronger at 23 (Fig. 3A, lane 3 and 5) than at 4 °C. Together, these results indicate that schwannomin specifically interacts with syntenin in vitro.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   In Vitro interaction between purified GST-syntenin (FL) and native schwannomin protein derived from human erythrocyte. A, glutathione-Sepharose beads coupled to GST alone (lanes 2 and 4) or GST-syntenin (FL) (lanes 3 and 5) were blocked with 0.5% bovine serum albumin in binding buffer and then incubated with native schwannomin protein purified from human erythrocyte membranes. The incubation was done either overnight at 4 °C (lanes 2 and 3) or for 90 min at 23 °C (lanes 4 and 5). After these incubations, a GST pull-down using glutathione-Sepharose beads was performed before the products were resolved on SDS-PAGE gels, transferred, and detected using anti-sch-1. The schwannomin from the erythrocyte extract migrates in parallel to the schwannomin obtained from the GST pull-down material (lane 1). B, an anti-syntenin-N polyclonal antibody was used to coimmunoprecipitate schwannomin from 3T3-L1 tet-off cells extracts. The immunoprecipitated protein (lane 2) was resolved on an SDS-PAGE gel, transferred, and detected using a commercial anti-sch-1 (C-18) antibody. The positive and negative controls were, respectively, total cell extract (lane 1) and protein A/G-seahorse alone (lane 3).

Syntenin Associates with Schwannomin In Vivo-- To determine whether the interaction observed in vitro using both the yeast two-hybrid system and the pull-down assays occurs in vivo, we performed a coimmunoprecipitation experiment. A purified anti-syntenin-N antibody directed against the amino-terminal segment of syntenin was used to immunoprecipitate the potential complex to be formed by endogenous schwannomin and syntenin in 3T3L1 cells. The coimmunoprecipitated material was boiled in SDS sample buffer and loaded on SDS-PAGE gels detected with anti-sch-1 C-18. The pull-down of the syntenin and its bound partners by anti-syntenin-N did reveal the presence of schwannomin (Fig. 3B, lane 2). Although the coimmunoprecipitated sch-1 band was relatively weak it migrated to the exact same position as the sch-1 band from total 3T3L1 cell lysate (Fig. 3B, lane 1). In view of our previous in vitro demonstration of the interaction between syntenin and sch-1, this in vivo result suggest that syntenin may form a complex with schwannomin in the living cell.

The Carboxyl-terminal End of Schwannomin Isoform-1 Is Sufficient for Its Localization to the Plasma Membrane-- Wild type schwannomin is mainly concentrated at the cytoplasmic face of the plasma membrane, in close association with the actin cytoskeleton (14). Schwannomin is also found around the membrane-ruffling region and in as yet uncharacterized granules/vesicles at the perinuclear region (4). To verify whether syntenin is localized to similar structures, we transfected HeLa cells with various constructs of both schwannomin and syntenin eGFP fusion proteins (Fig. 4). We initially tested and confirmed that HeLa cells do express endogenous syntenin (data not shown). HeLa cells transfected with the pEGFP-C1 vector alone showed green fluorescence uniformly in the cytoplasm but not specifically underneath the plasma membrane (Fig. 4A). When HeLa cells were transfected with the sch-C30-1-GFP construct that expresses the last 30 amino acids of sch-1 (Fig. 4B), fluorescence was observed at the plasma membrane, as well as in some cytoplasmic punctate structures/vesicles of unknown identity. When a longer segment of sch-1 (sch-C125-1-GFP) encoding the carboxyl-terminal 125 amino acids was expressed, the majority of the signal was observed at the plasma membrane with some fluorescence visible in the punctate granular vesicle-like structures (Fig. 4C). Finally, when a full-length construct of sch-1 tagged to GFP was expressed in the same cells, the fluorescence signal was largely granular and located at the cytoplasmic face of the plasma membrane and throughout the cytoplasm (Fig. 4D). Overall these results demonstrate that the carboxyl-terminal segment of sch-1 is sufficient for mediating its localization underneath the plasma membrane and to punctate/vesicular structures. It is noteworthy that although a relatively weaker signal was observed under the plasma membrane with the shorter sch-C30-GFP construct, in comparison with the signal from the sch-C125-GFP construct, it was nonetheless sufficient to direct a portion of this sch-1 peptide to this location. HeLa cells were also transfected with syntenin (FL)-GFP and with a syntenin (120)-GFP. Both syntenin fusion proteins were also partially localized to the plasma membrane and in the case of syntenin (FL)-GFP to what appears to be punctate/intracellular vesicular structures (Fig. 4, E and F). The subcellular localization of sch-1 and syntenin (FL) was observed under the plasma membrane and in punctate structures, which suggests that these two compartments could be the ones providing a suitable environment for those protein to functionally associate with each other in vivo.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Subcellular localization of syntenin and schwannomin-1 in HeLa cells. The cells were separately transfected with specific syntenin and schwannomin-eGFP fusion protein, and 18 h later the fluorescence signal was directly observed using a confocal laser-scanning microscope (Leica). Pictures taken from HeLa cells expressing, in order, the following: A, the vector pEGFP-C1 alone; B, schwannomin-C30-1-GFP; C, schwannomin-C125-1-GFP; D, full-length schwannomin-GFP; E, syntenin (120)-GFP; and F, syntenin (FL)-GFP.

Down-regulation of Syntenin Causes Mislocalization of Schwannomin Isoform-1-- We wanted to assess the biological significance of the syntenin-schwannomin interaction in living cells and hypothesized that syntenin might be involved in the trafficking of sch-1 to the cytoplasmic membrane. Therefore, we anticipated a disruption of sch-1 localization if down-regulation of syntenin was to occur in cells. To test this hypothesis, we established a tet-inducible system to control the expression of an antisense cDNA of human syntenin in a mouse fibroblast 3T3-L1 tet-off cell line (CLONTECH). Before any localization studies were performed, Western blot detections using anti-syntenin-C were prepared from different 3T3-L1 tet-off transfectant cell lines to verify the efficiency of the syntenin antisense expression by looking at the level of syntenin protein detectable (Fig. 5). A 3T3-L1 tet-off cell line transfected with the pTRE vector alone (line V1C11) shows a normal level of syntenin expression (Fig. 5, lane 1). In contrast, two 3T3-L1 tet-off-derived cell lines (A1C1 and A1C3) transfected with the pTRE vector encoding for the conditional expression of antisense syntenin upon doxycycline removal showed significant suppression of syntenin protein (Fig. 5, lanes 2 and 3).

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Down-regulation of syntenin and the effect on schwannomin isoform-1 localization. Stable clones of 3T3-L1 tet-off cells were obtained from transfections with either a pTRE vector with the cDNA for antisense syntenin or with the pTRE vector alone. Following doxycycline removal to activate the antisense expression, total proteins were extracted from the cell lines. The proteins were resolved on an SDS-PAGE gel transferred and detected using an anti-syntenin-C antibody. The endogenous level of syntenin expression in the 3T3-L1-tet-off cell line V1C11 that expresses the control pTRE vector is shown in lane 1. After doxycycline removal, the endogenous syntenin level from two of 3T3-L1 tet-off cell lines that express the syntenin antisense was also verified, A1C1 (lane 2) and A1C3 (lane 3).

A full week after the removal of doxycycline from the culture medium of cell lines V1C11, A1C1, and A1C3, the cells were examined for any change in sch-1 localization. This was observed through fluorescent microscopy with the anti-sch-1 antibody described earlier. In both the V1C11 cell line (where the syntenin antisense expression is not possible) (Fig. 6A) and in the A1C3 cell lines where doxycycline was not removed (Fig. 6B), sch-1 distribution was observed both in the cytoplasm and underneath the plasma membrane with some signal visible in the filopodia extension. In sharp contrast to this, cell lines A1C3 (Fig. 6, C, D, and E) and A1C1 (Fig. 6F) expressing antisense syntenin displayed a more flattened morphology, and more importantly, sch-1 signal was largely granular, localizing predominantly within the cytoplasm with much less signal detectable at the cytoplasmic face of plasma membrane. Together, these results support the hypothesis that the syntenin-schwannomin interaction plays a critical role in the subcellular trafficking and targeting of the schwannomin isoform-1 to the plasma membrane.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Schwannomin subcellular distribution in 3T3-L1 tet-off cell transfectants 1 week after doxycycline removal. The localization of sch-1 was observed by immunocytochemistry using an anti-sch-1 antibody (C-18). Pictures were taken from A, V1C11 pTRE vector control cell line; B, A1C3 cell line before doxycycline was removed from the media; C-F, cell lines A1C3 (C, D, and E) and A1C1 (F) after doxycycline removal to induce the syntenin antisense expression.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Schwannomin (also called merlin/NF2 protein) is a member of the ERM family of proteins that are believed to play important roles in the reorganization of cortical cytoskeleton (14, 23). Over the last two decades, extensive biochemical studies on erythrocyte protein 4.1, the prototypical member of the ERM superfamily, laid the foundation for the function of this family of proteins as linkers of membrane-cytoskeleton (31). Schwannomin contains an amino-terminal FERM domain preceding an -helical stretch interrupted by a proline-rich region and ends with a charged C terminus. The two most abundant isoforms of schwannomin lack residues corresponding to either exon 16 (sch-1) or exon 17 (sch-2), and therefore the last 16 amino acids of sch-1 differentiate these two isoforms (3, 6). More importantly, only sch-1 has been shown to suppress cell growth suggesting that the last 16 amino acids of sch-1 are essential for its tumor suppressor function in vivo (21). Like other ERM proteins, sch-1 can form intramolecular interactions between its amino and carboxyl termini thus creating a "closed" conformation that must be disrupted to produce an "open" conformation of sch-1 (32). In contrast, sch-2 does not undergo intramolecular interactions and therefore remains locked in open conformation, permitting constitutive and perhaps unregulated interactions with actin and other cytoplasmic proteins (33). For example, sch-2 associates tightly with the II-spectrin through its C terminus whereas the association between sch-1 and II-spectrin is significantly weaker (17). It is believed that the weaker binding of sch-1 could be a consequence of the masking of the ligand-binding site because of its closed conformation. The open conformation of sch-1 is also a prerequisite for its interactions with the hepatocyte growth factor-regulated tyrosine kinase substrate and EBP50, respectively (34). Similarly, the carboxyl-terminal residues of sch-1 are known to participate in the homotypic interaction between sch-1 molecules and also in the heterotypic interaction with either ezrin and sch-2 (35, 36). Together, these observations indicate that the C terminus of sch-1 participates in the formation of both homo- and heterotypic associations thus permitting the assembly of large protein complexes in vivo.

To gain further insights into the functional significance of schwannomin isoform diversity, we used the carboxyl-terminal 30 amino acids of sch-1 as bait in a yeast two-hybrid assay to search for interacting proteins in the human fetal kidney cDNA library. The yeast two-hybrid screen resulted in the identification of syntenin as a new binding partner for sch-1 (Fig. 1). Syntenin is a 30-kDa cytosolic adaptor protein that contains a tandem repeat of PDZ domains, which bind directly the cytoplasmic domains of syndecans receptors (37). Syndecans are a family of transmembrane cell-surface heparin sulfate proteoglycans that participate in multiple cellular functions (38). Interestingly, the cytoplasmic domain of syndecan-2 could also bind to the PDZ domain of human CASK/LIN-2 that in turn associates with the FERM domain of protein 4.1 thus providing a coupling mechanism between the extracellular matrix and the actin cytoskeleton (39). More recent evidence indicates that syntenin could serve as a marker protein for the apical-basolateral polarity in the MDCK epithelial cells thus suggesting a functional role of syntenin in the intracellular trafficking and targeting of receptors to the plasma membrane (28).

The two PDZ domains of syntenin, encompassing two thirds of the protein, are crucial for its intracellular scaffolding function linking transmembrane receptors to signaling components or cytoskeleton (40, 41). In general, the PDZ domains bind to the C termini of numerous proteins/or internal peptides of defined sequence, dimerize with other PDZ domains, and interact with other protein domains such as SH3, GUK (guanylate kinase-like), FERM, and leucine zipper motifs (42-44). The PDZ domain-seeded multimeric protein complexes are involved in the organization, positioning, and clustering of receptors (Wingless, Notch, proTGF-) and in the recruitment of cytosolic proteins (GTPase-activating protein and guanine nucleotide exchange factor) to the plasma membrane (45). Our yeast two-hybrid and the GST pull-down results indicate that the carboxyl-terminal peptide of sch-1 binds to the region overlapping the two PDZ domains of syntenin (Figs. 2 and 3). This observation is consistent with the earlier demonstration that both PDZ domains of syntenin are required for the binding of syntenin to syndecans and proTGF- (24, 27). In contrast, the neurofascin adhesion receptor binds exclusively to the PDZ2 domain of syntenin indicating the complexity of PDZ domain-mediated interactions within a single molecule (46). The last 25 amino acids of syntenin, presumably located outside its PDZ2 domain, are also important in strengthening its interaction with sch-1, or they might be required for proper folding of the two PDZ domains. In contrast, sch-2 failed to bind to syntenin indicating that the carboxyl-terminal sequence of sch-1 plays a critical role in this interaction. The in vitro association between sch-1 and syntenin was confirmed in vivo by coimmunoprecipitation and colocalization studies (Figs. 3 and 4). The relatively weak in vivo interaction between syntenin and sch-1 could be a consequence of factors such as intra- and intermolecular associations of sch-1, undefined cell culture conditions such as cell density, serum growth factor, adhesion substrate, and cell shape (47), and the phosphorylation status of sch-1. Future studies investigating the effects of these factors on syntenin and sch-1 interaction may provide answers to these questions.

The PDZ domains are classified into two subclasses based on the sequence of peptides with which they bind. Class I PDZ domains tend to prefer the carboxyl-terminal sequence (S/T)XV. In contrast, class II PDZ domains preferentially bind (F/Y)X(F/V/A) (48). Based on previously published studies, the PDZ domains of syntenin have been classified as class II PDZ domains with preference for bulky hydrophobic residues at positions 0 and 2 of the binding peptide (48, 49). A close inspection of the last 16 amino acids of sch-1 reveals an intriguing internal sequence, SRV, that may fold into a -hairpin "finger," allowing its interaction with the class I PDZ domains (50). Whether this internal SRV motif of sch-1 binds to the class II PDZ domains of syntenin remains to be established. The very C terminus of sch-1 is likely to provide the binding interface for the peptide binding pockets of class II PDZ domains of syntenin. Interestingly, mutagenesis of the Glu-593 residue at the 2 position in sch-1 does not affect its binding to syntenin whereas the replacement of Phe-592 located at the 3 position completely abolished the binding (Fig. 1C). This result is consistent with published studies demonstrating that a variety of amino acids (except Pro and Lys) could replace the Thr residue at the 2 position without disrupting the interaction of proTGF- with syntenin (27). These observations suggest that one or both PDZ domains of syntenin may display a unique and relaxed specificity at the 2 position consistent with the observed specificity of other class II PDZ domains (27). Alternatively, the direct interaction between syntenin and sch-1 as reported here may unravel a new mode of sequence motif of carboxyl-terminal peptides recognized by the PDZ domains of syntenin. Future structural studies of peptide-bound PDZ domains may provide clues to some of these questions.

Although the existence of the two splice variants of schwannomin has been known for some time, the functional basis of their distinguishing feature at the C termini remains poorly understood. The present study provides the first evidence demonstrating direct interaction between the C terminus of sch-1 and the two PDZ domains of syntenin. The lack of binding between the C terminus of sch-2 and syntenin may begin to shed light on the intriguing observation that only sch-1 functions as a growth suppressor protein (22). Syntenin was originally localized to both plasma membrane and early secretory structures of MDCK cells (24, 27), and our results also show punctate vesicular staining of syntenin within the cytoplasm of HeLa cells. The proposed role of syntenin in the intracellular trafficking pathways may be relevant to the function of sch-1 in vivo. sch-1 is generally localized to the plasma membrane, but the mechanism of its delivery to the cell periphery remains unknown. Our results demonstrate that the exogenous sch-1 is predominantly localized to the plasma membrane consistent with the partial localization of syntenin to the plasma membrane. Indeed, the down-regulation of syntenin by antisense cDNA expression leads to punctate/vesicular accumulation of sch-1 in the cytoplasm (Fig. 6) implying that the syntenin and sch-1 interaction may be critical for some of the subcellular targeting functions in vivo. Considering the vast number of protein-protein interactions afforded by the PDZ domains of syntenin, and the presence of open/close conformational switch in sch-1, our findings may provide a framework for the trafficking and subcellular targeting of large protein complexes necessary for the efficient mode of signal transduction in vivo.

Our results indicate that the adhesion properties of 3T3L1 cells are also affected, as evident by their flattened morphology, when the expression of syntenin is suppressed and the membrane localization of sch-1 is altered (Fig. 6, C and D). This observation is consistent with the previous reports showing that some mutant, and rare isoforms of schwannomin lead to weakened cell surface adhesion properties (23, 52, 53), and some isoforms of schwannomin are colocalized with 1-integrin (18). Moreover, one of the binding partners of syntenin, syndecan-4, signals in a Rho-dependent manner with 1-integrin thus modulating the assembly of focal adhesions and microfilaments (51). Together, the PDZ domain-mediated interaction between syntenin and sch-1 could rationalize the existence of a large protein complex with syndecan-4, proTGF-, ephrin ligands, and neurofascin, and the regulation of this multimeric complex is likely to underlie the molecular basis of growth suppression by sch-1 in brain tumors.

	FOOTNOTES

^* This work was supported in part by funds from the National Cancer Institute. G. A. R. was supported by the Canadian Institutes of Health Research. M. J. was the recipient of a Neuroscience Foundation fellowship. National Institutes of Health Grants CA66263 and HL60755 provided support for A. H. C.

^§ Present address: Dept. of Neurobiology and Behavior, Laboratories of Molecular Neuropatho-genesis and Molecular and Cellular Neurobiology, 1109 Gillespie Neuroscience Research Facility, Irvine, CA 92697-4545.

To whom correspondence should be addressed. Tel.: 514-937-6011 (ext. 4235); E-mail: mi32@musica.mcgill.ca.

Published, JBC Papers in Press, June 29, 2001, DOI 10.1074/jbc.M105792200

	ABBREVIATIONS

The abbreviations used are: NF2, neurofibromatosis type 2; sch, schwannomin; PCR, polymerase chain reaction; eGFP, enhanced green fluorescent protein; GST, glutathione S-transferase; PAGE, polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline; ECL, enhanced chemiluminescence; FL, full-length.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				D. Sarkar, H. Boukerche, Z.-z. Su, and P. B. Fisher mda-9/Syntenin: More than Just a Simple Adapter Protein When It Comes to Cancer Metastasis Cancer Res., May 1, 2008; 68(9): 3087 - 3093. [Abstract] [Full Text] [PDF]

				J. M. Beekman and P. J. Coffer The ins and outs of syntenin, a multifunctional intracellular adaptor protein J. Cell Sci., May 1, 2008; 121(9): 1349 - 1355. [Abstract] [Full Text] [PDF]

				N. Chatterjee, J. Stegmuller, P. Schatzle, K. Karram, M. Koroll, H. B. Werner, K.-A. Nave, and J. Trotter Interaction of Syntenin-1 and the NG2 Proteoglycan in Migratory Oligodendrocyte Precursor Cells J. Biol. Chem., March 28, 2008; 283(13): 8310 - 8317. [Abstract] [Full Text] [PDF]

				N. Latysheva, G. Muratov, S. Rajesh, M. Padgett, N. A. Hotchin, M. Overduin, and F. Berditchevski Syntenin-1 Is a New Component of Tetraspanin-Enriched Microdomains: Mechanisms and Consequences of the Interaction of Syntenin-1 with CD63 Mol. Cell. Biol., October 15, 2006; 26(20): 7707 - 7718. [Abstract] [Full Text] [PDF]

I. Gimferrer, A. Ibanez, M. Farnos, M.-R. Sarrias, R. Fenutria, S. Rosello, P. Zimmermann, G. David, J. Vives, C. Serra-Pages, et al. The Lymphocyte Receptor CD6 Interacts with Syntenin-1, a Scaffolding Protein Containing PDZ Domains J. Immunol.,