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

J. Biol. Chem., Vol. 275, Issue 27, 20578-20587, July 7, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/27/20578    most recent M001887200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lim, R. W. L. Articles by Halpain, S. Search for Related Content PubMed PubMed Citation Articles by Lim, R. W. L. Articles by Halpain, S. Social Bookmarking                      What's this?

Regulated Association of Microtubule-associated Protein 2 (MAP2) with Src and Grb2: Evidence for MAP2 as a Scaffolding Protein^*

Rita W. L. Lim and Shelley Halpain

From the Department of Cell Biology, The Scripps Research Institute, La Jolla, California 92037

Received for publication, March 7, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Microtubule-associated protein 2 (MAP2) and tau, which is involved in Alzheimer's disease, are major cytoskeletal proteins in neurons. These proteins are involved in microtubule assembly and stability. To further characterize MAP2, we took a strategy of identifying potential MAP2 binding partners. The low molecular weight MAP2c protein has 11 PXXP motifs that are conserved across species, and these PXXP motifs could be potential ligands for Src homology 3 (SH3) domains. We tested for MAP2 interaction with SH3 domain-containing proteins. All neuronal MAP2 isoforms bound specifically to the SH3 domains of c-Src and Grb2 in an in vitro glutathione S-transferase-SH3 pull-down assay. Interactions between endogenous proteins were confirmed by co-immunoprecipitation using brain lysate. All three proteins were also found co-expressed in neuronal cell bodies and dendrites. Surprisingly, the SH3 domain-binding site was mapped to the microtubule-binding domain that contains no PXXP motif. Src bound primarily the soluble, non-microtubule-associated MAP2c in vitro. This specific MAP2/SH3 domain interaction was inhibited by phosphorylation of MAP2c by the mitogen-activated protein kinase extracellular signal-regulated kinase 2 but not by protein kinase A. This phosphorylation-regulated association of MAP2 with proteins of intracellular signal transduction pathways suggests a possible link between cellular signaling and neuronal cytoskeleton, with MAP2 perhaps acting as a molecular scaffold upon which cytoskeleton-modifying proteins assemble and dissociate in response to neuronal activity.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In the mammalian nervous system, the neuron-specific microtubule-associated protein MAP2¹ regulates microtubule dynamics (1) and is attributed with an important role in neurite outgrowth and dendrite development (2-6). MAP2 belongs to a family of heat stable microtubule-associated proteins found in eukaryotic cells. Other family members include MAP4, a ubiquitous protein, and tau, a neuronal protein involved in Alzheimer's disease (4, 5). Members of this family share a common microtubule-binding domain (MTBD) consisting of approximately 100-130 amino acid residues near the carboxyl terminus of the molecule. The MTBD contains three or four tandem imperfect repeats of 18 amino acid residues separated by a variable inter-repeat region of approximately 12 amino acid residues (7).

Recently, tau has been shown to bind the Src family of non-receptor tyrosine kinases (Src, Lyn, and Fyn (8)) and protein phosphatases PP1 and PP2A (9-11). The interactions among PP1, PP2A, tau, and microtubules were postulated to regulate the phosphorylation state of tau, and alterations to these interactions could lead to the development of tauopathies seen in neurological diseases such as Alzheimer's (12-14). This increasing body of evidence suggests that microtubule-based structural proteins like MAP2 and tau could play other physiological roles in addition to the regulation of microtubule assembly.

Like tau, MAP2 is a phosphoprotein, and its cellular activities are regulated by phosphorylation. The phosphorylation state of MAP2 is modulated by neural activity in vivo (15-20). In vitro, several serine/threonine protein kinases and phosphatases are capable of using MAP2 as a substrate (21-28). Tyrosine kinases such as insulin receptor kinase, epidermal growth factor receptor kinase, and Src have been reported to phosphorylate MAP2 in vitro (29-31). Besides microtubules, several additional proteins have been reported to interact with MAP2 in vitro, the best characterized of which is the type II regulatory subunit of PKA (32-34). Other proteins include F-actin (35-38), calmodulin (39), and the mitochondrial outer membrane protein porin (40). The precise binding sites on MAP2 for these proteins have yet to be identified. Except for the MTBD at the carboxyl terminus and the type II regulatory subunit of PKA binding region (30 amino acids) in the amino terminus of MAP2, the rest of the molecule (65%) has no assigned function. It is possible that MAP2 possesses some as yet uncovered role in the neuron, in addition to its contribution to cellular structural integrity.

This paper presents evidence of a novel interaction of MAP2 with the non-receptor protein tyrosine kinase c-Src and the adaptor protein Grb2. The data support the emerging idea that members of signal transduction pathways interact with microtubule-associated proteins and could thereby regulate the organization and dynamics of the cytoskeleton.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- DNA restriction enzymes, PKA, MAPK (extracellular signal-regulated kinase 2, p42), and phosphotyrosine monoclonal antibody P-Tyr-100 were from New England Biolabs. The Pfu DNA polymerase and Escherichia coli strains XL1 Blue, XL2 Blue, DH5, JM109, and BL21(DE3) were from Stratagene. Protein expression vectors pET-3a, pET-11a, and pET-32a were from Novagen. The pGEX-2T vector, glutathione-Sepharose, SP ion exchange resin, HRP-conjugated goat anti-mouse and anti-rabbit antibodies, and thrombin were from Amersham Pharmacia Biotech. Monoclonal anti-polyhistidine antibody, HIS-1, anti-MAP2 antibody, HM2, PMSF, reduced glutathione, protein A-Sepharose, ampicillin, carbenicillin, leupeptin, aprotinin, pepstatin A, benzamide, benzamidine, -caproic acid, GTP, ATP, L-glutamine, pyruvate, GSK-3, and lysozyme were from Sigma. Superfect reagent was from Qiagen. Paclitaxel was from Calbiochem. Sypro Orange was from Bio-Rad. BCA protein assay reagent was from Pierce. Nitrocellulose membranes (0.45 µm, BA-85) were from Schleicher & Schuell. Alkaline phosphatase-conjugated goat anti-rabbit and anti-mouse antibodies were from Promega. Alexa-568 conjugated goat anti-mouse IgG and Alexa-488 conjugated goat anti-rabbit IgG were from Molecular Probes. Rabbit anti-Src sc-18 and rabbit anti-Grb2 sc-255 antibodies were from Santa Cruz Biotechnology. Monoclonal anti-phosphotyrosine antibody 4G10 was from Upstate Biotechnology. HRP-conjugated recombinant anti-phosphotyrosine PY20 (RC20-HRP) and monoclonal anti-Grb2 antibodies were from Transduction Laboratories. Bovine tubulin was from Cytoskeleton Inc. Talon spin columns (0.5 ml) were from CLONTECH Laboratories. [-³²P]ATP was from NEN Life Science Products.

Construction of Expression Vectors-- All molecular methods were performed according to Ref. 41. Construction of the expression plasmid pET3aMAP2c, which encodes the full-length rat MAP2c protein, has been described (1). Mutant forms of MAP2c were made by PCR cloning using pET3aMAP2c or one of the subsequently constructed expression plasmids described below as PCR templates. The open reading frames in all constructs were verified by automated DNA sequencing conducted at Scripps Nucleic Acid Core facility with PE Biosystems instrument 373 A using the Taq DNA polymerase, fluorescence dye detection, and dideoxy terminator chemistry (42). Custom designed primers with unique restriction enzyme recognition sites were used to amplify specific regions of MAP2c cDNA and to facilitate ligation of the amplified PCR fragments into expression vectors. DNA amplifications were conducted in a model 9600 DNA thermal cycler (PE Biosystems). The amplification protocol consisted of 30 cycles of the following conditions: denaturation at 95 °C for 1 min, primer annealing at 57 °C for 1 min, and chain extension at 72 °C for 1 min (5 min for the last cycle) with recombinant Pfu DNA polymerase according to the manufacturer's procedure. The PCR fragments encoding the amino acids 1-299, 1-310, 154-467, and 296-467 of MAP2c were digested with NdeI and BamHI and ligated into a previously NdeI/BamHI-digested pET-11a plasmid. The PCR fragment encoding amino acids 300-400, the MTBD, was digested with NcoI and BamHI and ligated into a previously NcoI/BamHI-digested pET-32a plasmid. The new constructs were transformed into cloning host E. coli strain DH5, XL-1 Blue, XL-2 Blue, or JM109 and were also transformed into expression host E. coli BL21(DE3) by the CaCl₂ transformation method. Stable transformants were selected on LB plates supplemented with 50 µg/ml carbenicillin.

To make internal deletion MAP2c constructs, separate DNA fragments of MAP2c were amplified by PCR and then ligated together. The expression plasmid 134-141 encoding MAP2c, lacking the amino acids 134-141, was made by amplifying the DNA sequence encoding the amino acids 1-133 and 142-467. Both PCR products were gel-purified and digested with BssSI. The 1-133 fragment was further digested with NdeI, whereas the 142-467 fragment was digested with BamHI. The digested DNA fragments were gel-purified again and ligated into the NdeI/BamHI-digested expression vector pET-3a. The ligation of these two fragments introduced an additional valine at the junction between the amino acids 133 and 142. The 134-141 construct was transformed into JM109 and BL21(DE3). The expression plasmids 219-290, 290-297, and 290-400, which encode MAP2c lacking regions 219-290, 290-297, and 290-400, respectively, were constructed using a similar strategy, except that the PCR products were initially digested with BsiWI prior to digestion with NdeI or BamHI. In making the double deletion construct 134-141, 219-290, a PCR fragment corresponding to the region 142-467, 219-290, was amplified from the 219-290 single deletion plasmid described above. This PCR fragment was digested with BssSI and BamHI, gel-purified, and then ligated to the NdeI/BssSI-digested fragment 1-133. For the double deletion construct 134-141, 290-297, a PCR fragment corresponding to the region 1-289, 134-141, was amplified from the 134-141 single deletion plasmid described above. This PCR fragment was sequentially digested with BsiWI followed by NdeI, gel-purified, and then ligated to the BsiWI/BamHI-digested fragment 298-467. To construct the 219-297 plasmid, the 219-290 plasmid was digested with NdeI and BsiWI, and the larger fragment, which encodes the region 1-218, was gel-purified. The 290-297 plasmid was digested with BsiWI and BamHI, and the smaller fragment, which encodes the region 298-467, was gel-purified. The 1-218 NdeI/BsiWI fragment was ligated to the 298-467 BsiWI/BamHI fragment and then inserted into an NdeI/BamHI-digested pET-3a plasmid. To construct the triple deletion 134-141, 220-297 plasmid, the 134-141, 219-290 plasmid was digested with NdeI and BsiWI, and the larger fragment, which encodes the region 1-218, 134-141, was gel-purified. This fragment was ligated to the 298-467 BsiWI/BamHI fragment and then inserted into an NdeI/BamHI-digested pET-3a plasmid. These constructs were transformed into cloning bacteria E. coli JM109 and expression host bacteria BL21(DE3).

Expression and Purification of Recombinant MAP2c Proteins-- Induction of protein expression in bacteria and purification of full-length and mutant non-fusion MAP2c proteins were performed as described (1), with modifications. The concentration of NaCl was reduced to 0.1 M. The boiled supernatant collected was further purified by SP ion exchange chromatography. The supernatant was filtered through a 0.45 µm filter and loaded onto a 0.6 × 1.0-cm SP column pre-equilibrated with 5 ml of 50 mM sodium acetate, pH 5.5, 1 mM EDTA, 1 mM EGTA, and 5 mM -mercaptoethanol (Buffer A) at room temperature. The rMAP2c bound to the SP column at pH 5.5, whereas most of the bacterial proteins were found in the run-through. The column was washed with 20 ml of Buffer A containing 0.15 M NaCl, and rMAP2c protein was step eluted with Buffer A containing 0.3 M NaCl. Fractions of 300 µl were collected and analyzed by SDS-PAGE. Fractions containing rMAP2c were pooled and dialyzed against 500 ml of 50 mM Tris-HCl, pH 8, 1 mM EGTA, 1 mM EDTA, 0.02% NaN₃, 1 mM dithiothreitol, and 150 mM NaCl at 4 °C overnight, divided into small aliquots, and stored at -80 °C. The rMAP2c proteins obtained were of 90% purity, and the average yield was 0.7 mg/50 ml of bacterial culture.

The MAP2c MTBD fragment cloned into the expression plasmid pET-32a was expressed as a histidine-tagged thioredoxin fusion protein (HisTrxMTBD). The induction of HisTrxMTBD and the histidine-tagged thioredoxin (HisTrx) protein expression from pET-32a was initiated at A₆₀₀ = 0.6-0.9 with 1 mM isopropyl-1-thio--D-galactopyranoside. Protein expression was allowed to proceed at 30 °C for 3 h. Bacteria harvested from a 50-ml culture was lysed in 10 ml of 20 mM Tris-HCl, pH 8.0, 100 mM NaCl (Buffer B) supplemented with 1 mM PMSF, 10 µg/ml aprotinin, 10 µg/ml pepstatin A, 10 µg/ml leupeptin. Cell lysis using lysozyme and sonication were performed as described (1). The lysate was centrifuged at 15,000 × g for 30 min at 4 °C. The supernatant was collected, filtered through a 0.45 µm filter, and loaded onto a 0.5 × 1.5-cm Talon-NX column that had been pre-equilibrated with Buffer B. The column was washed with 5 ml of Buffer B followed by 5 ml of Buffer B containing 10 mM imidazole, pH 8.0, and the histidine-tagged protein was step eluted with Buffer B containing 100 mM imidazole, pH 8.0. Fractions of 500 µl were collected and analyzed by SDS-PAGE. Fractions containing the histidine-tagged protein were pooled, dialyzed against 250 ml of Buffer B with 5 mM EDTA and 0.02% NaN₃ at 4 °C overnight, divided into small aliquots, and stored at -80 °C. The proteins were of 95% purity, and the average yield was 0.5 mg/50 ml of bacterial culture.

GST-SH3 Fusion Proteins and Thrombin Cleavage-- Expression and purification of the GST and GST-SH3 fusion proteins using glutathione-Sepharose beads were carried out according to the manufacturer's procedure. A 5-µl aliquot of a ~50% (v/v) slurry of protein-bound glutathione-Sepharose beads contained about 15-20 µg of GST-SH3 fusion protein as determined by SDS-PAGE. For the microtubule binding competition experiments, the GST-SH3 fusion proteins were eluted from the beads with 10 mM reduced glutathione in 50 mM Tris-HCl, pH 8.0. The eluted proteins were dialyzed against 1× PBS at 4 °C overnight and stored at -80 °C.

To obtain the thrombin-cleaved Grb2-N-SH3 fragment, the GST-Grb2-N-SH3 fusion protein was affinity-purified from 5 liters of bacterial culture and dialyzed against 4 liters of 0.5× PBS at 4 °C, with two changes over a period of 12 h, to remove the reduced glutathione. The protein concentration of the dialyzed fusion protein was then diluted to 0.4 mg/ml with 10 mM Tris-HCl, pH 8.0. Thrombin was added to a final concentration of 10 units/mg protein, and the digestion was allowed to proceed at room temperature for 36 h. The digested mixture was filtered through a 0.22 µm filter and applied onto a 1.0 × 5.0-cm glutathione-Sepharose column to remove the larger GST fragment. The run-through solution containing the Grb2-N-SH3 fragment was concentrated using a 3000 MWCO membrane and further purified by gel filtration on Sephadex 30 (2 × 50 cm) equilibrated with 25 mM Tris-HCl, pH 8.0, 0.1 mM EDTA, pH 8.0, and 1 mM NaN₃ at a flow rate of 90 ml/h. Fractions of 3 ml were collected and analyzed by SDS-PAGE. The Grb2-N-SH3 fragment eluted at about 125 ml. Fractions containing Grb2-N-SH3 were pooled, concentrated, and stored at -80 °C.

In Vitro GST-SH3 Binding Assay-- Approximately 5 µg of total rMAP2c bacterial lysate, 1-2 µg of partially purified rMAP2c (boiled, high salt bacterial supernatant), 1-2 µg of SP purified rMAP2c, and 1 µg of phosphorylated or control-phosphorylated rMAP2c were incubated with 5 µl of a 50% slurry of protein bound glutathione-Sepharose beads in a 200-µl volume containing 50 mM Tris-HCl, pH 7.5, 75 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 1 mM dithiothreitol, 7.5 mM MgCl₂, 5% glycerol, 1 mM Na₃VO₄, 0.5% bovine serum albumin, and 1% Nonidet P-40 (Buffer C) at 4 °C for 2 h with gentle mixing. In competition assays using HisTrxMTBD or HisTrx proteins, 0.4 µg (80 nM) of rMAP2c was mixed with increasing amounts of HisTrxMTBD or HisTrx protein (0.2-6 µM) and incubated with 0.8 µl of the 50% slurry of protein bound glutathione-Sepharose beads in a final volume of 100 µl. In the binding assays in which rat brain or transfected COS-7 cell lysates were used, 0.3-1.0 mg of total lysate protein were incubated with 40 µl of a 50% slurry of protein bound glutathione-Sepharose beads in a final volume of 500 µl. The beads were collected by centrifugation, washed twice with 1.3 ml of Buffer C containing 150 mM NaCl, and washed a third time with Buffer C at 150 mM NaCl but minus Nonidet P-40. The presence of rMAP2c bound on the beads was analyzed by SDS-PAGE, followed by immunoblotting using monoclonal anti-MAP2 antibody HM2 (1:1000). The presence of histidine-tagged protein was detected by monoclonal anti-polyhistidine antibody HIS-1 (1:3000).

Cell Cultures, Transfection, and Immunofluorescence Microscopy-- COS-7 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 2 mM L-glutamine, and 2 mM pyruvate at 37 °C in a humidified 5% CO₂ atmosphere. When the cultures reached approximately 80% confluency, they were transfected with 8.5 µg of pEGFP-N1-MAP2c plasmid in the presence of 42 µl of Superfect reagent for 3 h at 37 °C according to the manufacturer's protocol. Approximately 60-70 h posttransfection, cells were rinsed once in ice-cold 1× PBS and harvested in ice-cold lysis buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.5, 0.5% sodium deoxycholate, 1% Nonidet P-40, 1 mM Na₃VO₄, 1 mM PMSF, 1 mM NaF, 10 µg/ml aprotinin, 10 µg/ml pepstatin A, and 10 µg/ml leupeptin). A pulse of sonication (Fisher Sonifer, setting 6, 5 s) was given to ensure complete cell lysis, and the lysate was centrifuged at 13,500 × g for 30 min at 4 °C. Expression of the rMAP2c-GFP fusion protein in the transfected COS-7 cells was verified by immunoblot analysis using monoclonal anti-MAP2 antibody HM2. The clarified lysate was used in the GST-SH3 binding assay.

Preparation of dissociated rat hippocampal cultures from E18 rat embryos, and growth and maintenance of these cultures were performed as described (43). At 15 days in vitro, the cultures were double-stained for MAP2 (HM2, 1:1000) and Src (sc-18, 1:1000), or MAP2 and Grb2 (sc-255, 1:1000) as described (43) using Alexa-568 conjugated goat anti-rabbit IgG (1:1000) and Alexa-488 conjugated goat anti-mouse IgG (1:1000) as secondary antibodies. Digital images were obtained using a 60× oil immersion objective on an Olympus IX-70 microscope coupled with a Photometrix PXL cooled CCD camera (43).

Preparation of Rat Brain Lysates-- Adult rat forebrain or hippocampus and neonatal (postnatal day 1 (P1)) brain tissue were lysed in 50 mM Tris-HCl, pH 7.5, 50 mM NaCl, 50 mM NaF, 5 mM Na₃VO₄, 5 mM EGTA, 5 mM EDTA, 1 mM dithiothreitol, 1% Nonidet P-40, 1 mM PMSF, 10 µg/ml aprotinin, 10 µg/ml pepstatin A, and 10 µg/ml leupeptin at 4 °C. The homogenates were sonicated to ensure complete lysis and centrifuged as described above. The clarified brain lysates were used in the GST-SH3 binding and co-immunoprecipitation assays.

Co-immunoprecipitation of MAP2 with Src or Grb2 from Rat Brain Lysate-- Clarified brain lysates (1 mg of total protein) were incubated with 25 µg of rabbit anti-Src (sc-18), rabbit anti-Grb2 (sc-255), rabbit anti-MAP2 (no. 4170),² irrelevant IgG, rabbit preimmune serum (no. 4170pis), or lysis buffer alone in a final volume of 400 µl at 4 °C overnight with gentle mixing. A 100-µl aliquot of a 50% slurry of protein A-Sepharose beads in 1× PBS and 1% bovine serum albumin was added, and the mixture was incubated at 4 °C for 1 h with gentle agitation. The beads were collected by centrifugation at 4000 × g for 5 min, washed six times briefly with 1 ml of 1× PBS, boiled in Laemmli SDS-sample buffer for 2 min (44), and analyzed by SDS-PAGE and immunoblotting.

In Vitro Phosphorylation of rMAP2c-- Phosphorylation reactions were carried out in a 50-µl volume containing 50 mM HEPES, pH 7.4, 10 mM MgCl₂, 1 mM EDTA, 1 mM EGTA, 2 mM dithiothreitol, 0.1 mM Na₃VO₄, 1 mM PMSF, 1 mM benzamidine, 1 mM benzamide, 5 mM -caproic acid, 15 µg of rMAP2c, and 45 units of either PKA or MAPK (p42, extracellular signal-regulated kinase 2) or 12 units of GSK-3, and 0.8 µM ATP with or without 0.5 µCi of [-³²P]ATP (6000 Ci/mmol, NEN Life Science Products) at 30 °C. At the indicated times, 13 µl were removed and boiled at 95 °C for 5 min to terminate the reaction. Samples were incubated on ice for 10 min and centrifuged at 13,500 × g for 15 min, and the supernatant collected was used in the GST-SH3 binding assay described above. Control reactions were carried out in the absence of added ATP.

Microtubule Binding Assay-- Microtubules were polymerized from tubulin at a concentration of 5 mg/ml in assembly buffer (20 mM NaPO₄, 100 mM L-glutamate, 1 mM EGTA, 1 mM MgCl₂, and 1 mM GTP) and stabilized by 20 µM Taxol as described (45). Binding assay was performed according to Ref. 46 using 1 µM full-length rMAP2c and 5 µM tubulin in an assay volume of 50 µl. The SDS-PAGE gels were stained with Sypro Orange, and the fluorescence was quantified on a FluorImager 595 (Molecular Dynamics, Sunnyvale, CA). In assays in which GST, GST-Src-SH3, or Grb2-N-SH3 was added to the assay, a 10 µM concentration of these proteins were used. Data are expressed as a percentage of rMAP2c found in the microtubule pellet fraction: (rMAP2c in pellet)/(rMAP2c in pellet + supernatant) × 100. One-way analysis of variance was used to assess statistical significance.

Protein Analytical Methods-- Protein concentrations were determined by the BCA method (47) using bovine serum albumin as the protein standard. Protein samples were analyzed by SDS-PAGE according to Ref. 44 and visualized by Coomassie Brilliant Blue or Sypro Orange staining. Electrotransfer and immunoblot analyses were performed as described (48) using the respective primary antibodies and alkaline phosphatase-conjugated or HRP-conjugated secondary antibodies. When alkaline phosphatase-conjugated antibodies were used, the 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium colorimetric detection method was used. When HRP-conjugated antibodies were used, the chemiluminescence detection method was used.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Recombinant MAP2c Interacts Specifically with SH3 Domains of Src and Grb2-- We hypothesized that other cellular proteins interact with MAP2 in addition to those that have been reported thus far. Analysis of the amino acid sequence of rat MAP2c, the smallest known isoform of MAP2 (also known as juvenile MAP2 or embryonic MAP2, due to its perinatal expression pattern), revealed 13 PXXP motifs distributed in several clusters in the amino-terminal two-thirds of the molecule. These polyproline clusters lie within regions of unknown function. Eleven of these PXXP motifs are conserved across species and across alternatively spliced isoforms of MAP2. Because PXXP is a known consensus motif of SH3 domain ligands (49), we investigated the possibility that MAP2c would interact with a variety of SH3 domain-containing proteins.

Using a GST-SH3 binding assay, we tested the ability of rMAP2c protein in crude bacterial lysate to co-precipitate with SH3 domains of several candidate proteins of varying cellular functions (Fig. 1). In this assay, the GST protein alone did not precipitate any detectable amount of rMAP2c. However, rMAP2c reliably co-precipitated together with GST-SH3 fusion proteins derived from the protein tyrosine kinase Src and the adaptor protein Grb2 (Fig. 1A). In contrast, the SH3 domains of the protein tyrosine kinase Abl, the adaptor protein Nck, and the GTPase activating protein GAP exhibited little or no detectable binding to rMAP2c. Although the fusion protein containing the second of the three SH3 domains in Nck, GST-Nck-SH3(2), reproducibly precipitated trace amounts of rMAP2c, neither the fusion protein containing the first SH3 domain Nck-SH3(1) nor that containing both the second and third SH3 domains of Nck, Nck-SH3(2 + 3), produced detectable binding with rMAP2c. This suggests that perhaps the second SH3 domain of Nck was conformationally inaccessible in the larger Nck-SH3(2 + 3) molecule. In addition to the intact Grb2, the SH3 domains located at the amino and carboxyl termini of Grb2 were analyzed independently. Grb2-N-SH3 exhibited stronger binding than Grb2-C-SH3 and may well contribute to the majority of the SH3 domain interaction observed for the intact Grb2.

View larger version (39K): [in this window] [in a new window]   Fig. 1.   Bacterially expressed recombinant MAP2c interacts specifically with the SH3 domains of Src and Grb2. In vitro GST-SH3 binding assay in which non-heat-denatured rMAP2c bacterial lysate was incubated with GST or various GST-SH3 fusion proteins bound to glutathione-Sepharose beads and analyzed by SDS-PAGE as described under "Experimental Procedures." A, nitrocellulose immunoblot probed with anti-MAP2 antibody HM2 and detected by colorimetric reaction. B, Coomassie Brilliant Blue-stained 12% SDS-PAGE gel. Positions of rMAP2c, GST, and assorted GST-SH3 fusion proteins are indicated at the right. Positions of molecular size markers (in kDa) are indicated at the left.

Both the recombinant and native forms of MAP2 are heat-stable and retain their microtubule binding properties after being subjected to high temperature (90-105 °C; 10 min) in the presence of high salt, a step often used in the purification of MAP2 and other heat-stable microtubule-associated proteins (50). As with microtubule binding, heat treatment did not abolish the ability of rMAP2c to interact with SH3 domains. The same SH3 domain selectivity profile was observed for rMAP2c partially purified from bacterial lysates via a high salt boiling step and also for the subsequently SP-purified rMAP2c (data not shown), suggesting that SH3 binding by MAP2c was not dependent on heat-labile secondary structure.

MAP2 Isoforms Expressed in Mammalian Cells Interact with SH3 Domains of Src and Grb2-- Among the SH3 domains tested, we observed that the Src and Grb2 SH3 domains gave specific and strongest binding signals for rMAP2c in our GST-SH3 binding assay. We therefore proceeded to analyze whether native isoforms of MAP2 expressed in eukaryotic cells behaved likewise. Recombinant MAP2c-GFP expressed in transfected COS-7 cells as well as both the high and low molecular weight isoforms of MAP2 present in rat brain lysate specifically bound the SH3 domains of Src and Grb2 (Fig. 2). In transfected COS-7 cells, the rMAP2c-GFP fusion protein was highly expressed and was detected as a doublet or a triplet of ~90-95 kDa (Fig. 2A), indicative of post-translational modification of the GFP fusion protein in COS-7 cells. These posttranslationally modified rMAP2c-GFPs from COS-7 cell lysate bound specifically to GST-Src-SH3 and GST-Grb2-N-SH3 (Fig. 2A). Adult rat brains express predominantly the high molecular weight isoforms MAP2a and MAP2b, and both of these isoforms were detected bound to GST-Src-SH3 and GST-Grb2-N-SH3 (Fig. 2B). Insignificant trace amounts of MAP2a and MAP2b were detected in GST alone control conditions. Similarly, the high molecular weight isoform MAP2b and the low molecular weight isoform MAP2c from neonatal rat brain lysate bound to GST-Src-SH3, GST-Grb2, GST-Grb2-N-SH3, and GST-Grb2-C-SH3 but not to GST (Fig. 2C).

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Recombinant MAP2c-GFP fusion protein expressed from transiently transfected COS-7 cells and native MAP2 isoforms from rat brain interact specifically with the SH3 domains of Src and Grb2. In vitro GST-SH3 binding assays were performed with transfected COS-7 cell lysates (A), adult rat hippocampal lysate (B), and neonatal rat brain lysate (C), as described under "Experimental Procedures." The immunoblots were stained with anti-MAP2 antibody HM2 and detected by colorimetric reaction. Positions of rMAP2c-GFP, various rat brain isoforms of MAP2 (MAP2a, MAP2b, and MAP2c) are indicated at the left. Positions of molecular size markers (in kDa) are indicated at the right.

Co-immunoprecipitation of Endogenous MAP2 with Endogenous Src or Grb2-- Results obtained from the in vitro GST-SH3 binding assay above strongly suggest that the SH3 domains of Src and Grb2 could mediate direct association with MAP2. We next asked whether endogenous Src or Grb2 also bind to native MAP2 by performing co-immunoprecipitation experiments (Fig. 3). Under non-denaturing conditions, antibodies to Src or Grb2 precipitated MAP2 isoforms in the immune complexes. A fraction of MAP2 in the adult and neonatal rat brains associated with Src or Grb2 and was found in the immunoprecipitates (Fig. 3A, lanes 3, 4, 7, and 8). No MAP2 was detected in a control immunoprecipitation experiment using an irrelevant IgG (Fig. 3A, lane 5). Similarly, no detectable MAP2 was bound on the protein A-Sepharose beads alone (Fig. 3A, lane 6). In the reciprocal co-immunoprecipitation experiments using rabbit anti-MAP2 antibody (no. 4170), Grb2 was found co-precipitating with MAP2 in the immune complex (Fig. 3B, lane 2). Preimmune serum IgG precipitated no MAP2 and a negligible amount of Grb2 (Fig. 3B, lane 3). A similar experiment to detect co-precipitation of Src using anti-MAP2 antibodies was not interpretable due to interference of cross-reacting IgG bands in the vicinity of Src on the blot.

View larger version (60K): [in this window] [in a new window]   Fig. 3.   Endogenous Src and Grb2 interact with native MAP2 isoforms in rat brain. The interactions of endogenous Src and Grb2 with native MAP2 proteins in rat brains were assessed by co-immunoprecipitation as described under "Experimental Procedures." A, the immunoblot was stained for MAP2 using anti-MAP2 antibody HM2 and detected by colorimetric reaction (lanes 1-6) or by the chemiluminescence method (lanes 7 and 8). Lane 1, adult rat hippocampal lysate (20 µg); lane 2, neonatal P1 brain lysate (20 µg); lane 3, IP of hippocampal lysate with anti-Src antibody sc-18; lane 4, IP of P1 brain lysate with anti-Src antibody sc-18; lane 5, IP of hippocampal lysate with an irrelevant IgG; lane 6, IP of hippocampal lysate without added antibody; lane 7, IP of hippocampal lysate with rabbit anti-Grb2 sc-255; lane 8, IP of P1 brain lysate with anti-Grb2 antibody sc-255. B, reciprocal co-immunoprecipitation assay of rat forebrain lysate using rabbit anti-MAP2 (no. 4170) antibody. Immune complexes were separated on a 4-12% gradient gel, electroblotted onto nitrocellulose membrane, and immunoblotted. The blot was divided and probed separately for MAP2 or Grb2. The top blot was stained for MAP2 using anti-MAP2 antibody HM2, and the bottom blot was stained for Grb2 using mouse anti-Grb2 antibody. Both blots were detected by the chemiluminescence method. Lane 1, adult forebrain lysate (30 µg); lane 2, IP of adult forebrain lysate with rabbit anti-MAP2 antibody (no. 4170); lane 3, IP of adult forebrain lysate with rabbit preimmune serum (no. 4170pis). Positions of MAP2 isoforms and Grb2 are indicated at the left. Positions of molecular size markers (in kDa) are indicated at the right.

Native MAP2 from Rat Brains Are Not Phosphorylated on Tyrosine Residues-- Because MAP2 from rat brain bound Src, we asked whether endogenous MAP2 was a natural substrate of Src kinase activity. In vitro, MAP2 is a substrate of the tyrosine kinases insulin receptor kinase, epidermal growth factor receptor kinase, and Src kinase (29-31). Native MAP2 and rMAP2c-GFP were immunoprecipitated from rat brain and transfected COS-7 cells lysates, respectively, in the presence of the tyrosine phosphatase inhibitor sodium vanadate. The immune complexes were separated on SDS-PAGE gels, transferred to nitrocellulose membranes, and immunoblotted with three different anti-phosphotyrosine antibodies: P-Tyr-100, 4G10, and PY20-derived RC20-HRP. No detectable signal above background was observed for any of these three antibodies (data not shown), suggesting that under our experimental conditions, native MAP2 is not a substrate of Src or other tyrosine kinases in vivo. These results were in agreement with experiments in which phosphoamino acid analyses failed to detect phosphotyrosine in MAP2 isolated from ³²P-labeled rat hippocampal slices (16).

MAP2, Src, and Grb2 Are Co-expressed in the Dendrites of Cultured Neurons-- The results of the in vitro GST-SH3 binding assays and the co-immunoprecipitation experiments strongly suggest that MAP2 could indeed interact with Src and Grb2 in vivo. For these interactions to occur in neurons, these proteins must be in close proximity with each other. We proceeded to establish the distribution and localization of these proteins in cultured rat hippocampal neurons. Neurons were double immunostained for MAP2 and Src or MAP2 and Grb2. As described previously (51), MAP2 was found exclusively in the somatodendritic compartment and was absent from axons (Fig. 4). Src immunoreactivity exhibited a punctate distribution in the cell body, axons, and dendrites (Fig. 4A). Grb2 immunoreactivity was largely concentrated in the cell body with modest levels throughout the dendrites; axons were stained only weakly (Fig. 4B). Thus, MAP2, Src, and Grb2 were all present in neuronal cell bodies and dendrites, where they could potentially engage in functional interactions.

View larger version (90K): [in this window] [in a new window]   Fig. 4.   MAP2, Src, and Grb2 proteins are co-expressed in the dendrites of cultured neurons. Rat hippocampal neurons were cultured at low density for 15 days in vitro and double-stained for MAP2 and Src or for MAP2 and Grb2, as described under "Experimental Procedures." A, neuron double-stained for Src with anti-Src antibody sc-18 (left) and for MAP2 with anti-MAP2 antibody HM2 (right). Src immunoreactivity is present throughout the cell body, axons, and dendrites. Axons are labeled only weakly. Arrowheads indicate processes that are positive for Src immunoreactivity but negative for MAP2, identifying them as axons. B, neuron double-stained for Grb2 with anti-Grb2 antibody sc-255 (left) and for MAP2 with anti-MAP2 antibody HM2 (right). Grb2 immunoreactivity is strongest in the cell body and moderate in the dendrites. The scale bar represents 20 µm.

The MTBD of rMAP2c Binds Src-SH3 and Grb2-N-SH3 Domains-- There are 13 potential SH3 domain binding PXXP motifs arranged in clusters in the amino-terminal two-thirds of the rat MAP2c, and 11 of these PXXP motifs are conserved across species (Fig. 5A). To identify the Src-SH3 and Grb2-SH3 binding region in MAP2c, we created a series of recombinant, non-fusion MAP2c proteins that have amino- or carboxyl-terminal truncations or have internal deletions at the polyproline-rich clusters (Fig. 5B). These mutants were assessed for SH3 domain binding activity by the GST-SH3 binding assay. All mutant MAP2c proteins were expressed at high levels in expression host bacteria (10-20% of total protein), remained primarily in the soluble fraction, and were purified from the soluble fraction by the high salt boiling method (1) followed by SP ion exchange chromatography. The rMAP2c in clarified bacterial lysates and SP-purified rMAP2c were used separately in the GST-SH3 binding assays, and both samples of rMAP2c proteins gave essentially similar results. Surprisingly, the carboxyl-terminally truncated mutants that contained all 13 potential SH3 domain binding PXXP motifs, 1-299 and 1-310, exhibited no detectable co-precipitation with GST-Src-SH3 or GST-Grb2-N-SH3 in the GST binding assay (Fig. 5B). In contrast, the amino-terminally truncated mutants, 154-476 and 296-467, efficiently co-precipitated with GST-Src-SH3 and GST-Grb2-N-SH3. The 296-467 MAP2c contains no PXXP motif. These data suggest that binding of Src-SH3 and Grb2-N-SH3 domains by MAP2c did not involve PXXP motifs and that the binding region resides elsewhere in the MAP2c molecule, possibly in the carboxyl terminus. This hypothesis was verified using mutant MAP2c proteins lacking specific polyproline-rich regions, amino acids 134-141, 219-290, and 290-297. All the polyproline-rich deletion mutants were still able to bind to GST-Src-SH3 and GST-Grb2-N-SH3, and the observed binding of the mutant rMAP2c proteins was comparable to that of the full-length rMAP2c (Fig. 5B). A candidate region in the carboxyl terminus of MAP2c that could contribute to SH3 domain binding is the MTBD. A mutant 290-400 rMAP2c lacking both the polyproline-rich cluster amino acids 290-297 and the MTBD was constructed and tested for SH3 domain binding. The 290-400 rMAP2c exhibited no detectable co-precipitation with GST-Src-SH3 and GST-Grb2-N-SH3 in the GST-SH3 binding assay (Fig. 5B).

View larger version (34K): [in this window] [in a new window]   Fig. 5.   Mapping the region on MAP2c that binds specifically to the SH3 domain of Src and Grb2. A, schematic diagram of MAP2c showing the polyproline-rich clusters in the amino-terminal two-thirds of the molecule. The amino acid sequences of the 11 conserved SH3 domain-binding PXXP motifs are shown. Other functional domains of MAP2c shown are the binding region for the type II regulatory subunit (RII) of the cAMP-dependent protein kinase (black box) near the amino terminus (33, 34), and MTBD containing three imperfect repeats (gray boxes) and inter-repeats near the carboxyl terminus (7). B, summary of the interactions of the full-length and the various mutant rMAP2c proteins with the SH3 domains of Src and Grb2. The amino- or carboxyl-terminally truncated rMAP2c are denoted by the amino acid residue positions corresponding to the specific region on the full-length MAP2c that are present in the mutant proteins (left). Mutants with internal deletion are denoted by the region of amino acid residues that are removed. A positive sign (+) indicates that there was detectable rMAP2 co-precipitation in the GST binding assay and that the amounts co-precipitated were comparable to those observed for the full-length rMAP2c. A negative sign (-) indicates that there was no detectable co-precipitation with GST-SH3 fusion proteins. ND, not determined. Results were compiled from at least three independent GST-SH3 binding assays using different batches of SP-purified rMAP2c proteins.

To further test that only the MTBD in rMAP2c contributes to Src-SH3 and Grb2-SH3 domain binding, a recombinant MTBD, amino acids 300-400, expressed as a histidine-tagged thioredoxin fusion protein (HisTrxMTBD), was made. However, the leader fusion protein HisTrx consistently gave a background of non-specific binding to the GST-SH3 fusion proteins used in the GST-SH3 binding assay. To circumvent this problem, the purified HisTrxMTBD protein was used in competition with the full-length rMAP2c in the GST-SH3 binding assay. If the MTBD in rMAP2c specifically interacts with the SH3 domains of Src and Grb2, we would expect a concentration-dependent inhibition of full-length rMAP2c binding to the above-mentioned SH3 domains in the presence of excess HisTrxMTBD but not in the presence of excess HisTrx. Increasing concentrations of HisTrxMTBD significantly reduced the amount of full-length rMAP2c co-precipitating with GST-Src-SH3 (Fig. 6A) or GST-Grb2-N-SH3 (Fig. 6B). At the same time, a corresponding increase of HisTrxMTBD in the pellet with GST-Src-SH3 was detected (Fig. 6A). Increasing amounts of HisTrx had no effect on the interaction of the full-length rMAP2c with the SH3 domains of Src (data not shown) or Grb2 (Fig. 6C) as the amount of full-length rMAP2c co-precipitated remained constant. Therefore, we conclude that the Src-SH3 and Grb2-SH3 domain-binding region is localized to the MTBD, amino acids 300-400 of MAP2c.

View larger version (39K): [in this window] [in a new window]   Fig. 6.   MAP2 microtubule-binding domain binds the SH3 domain of Src and Grb2. Recombinant MAP2c was incubated in the absence or presence of an increasing amount HisTrxMTBD or HisTrx proteins in the GST-SH3 binding assay. Proteins on the immunoblots were detected by the chemiluminescence method. A, immunoblot of proteins co-precipitated with GST-Src-SH3. The blot was divided and probed separately for MAP2 using anti-MAP2 antibody HM2 (top) or stained for HisTrxMTBD using anti-polyhistidine HIS-1 (bottom). B and C, immunoblots of proteins co-precipitated with GST-Grb2-N-SH3. Blots were stained for MAP2 using anti-MAP2 antibody HM2.

Src and Grb2 Bind Soluble Non-microtubule-bound MAP2-- Because microtubules, Src-SH3 domain, and Grb2-SH3 domain bind within the same region in MAP2c, we were interested to know whether these associations were mutually or non-mutually exclusive. MAP2c-microtubule binding assays were performed in the absence and presence of 10-fold molar excess of GST or GST-Src-SH3. Because GST is known to form dimers (52, 53), the dimeric GST-Src-SH3 might be sterically excluded from the MTBD in the presence of microtubules. We attempted to obtained the Src-SH3 and Grb2-N-SH3 fragments from the GST fusion proteins by thrombin cleavage for use in this assay. Only the GST-Grb2-N-SH3 protein was digestible by thrombin to liberate a larger, 27-kDa GST leader and a smaller, 6-kDa Grb2-N-SH3 protein. The thrombin cleavage site between the GST leader and Src-SH3 domain in GST-Src-SH3 fusion was presumed to be eliminated during the construction of the expression vector. Under the microtubule-binding experimental conditions, rMAP2c bound to microtubules with an approximate stoichiometry of 1 mol of rMAP2c to 4 mol of tubulin dimer. No significant difference in rMAP2c binding to microtubules was observed in the presence of excess GST, GST-Src-SH3, or Grb2-N-SH3 (p  0.05; n = 5; one-way analysis of variance) (Fig. 7). We further determined whether microtubule-bound rMAP2c could mediate GST-Src-SH3 co-precipitation in the presence of microtubules. GST-Src-SH3 alone did not bind microtubules, and it did not co-precipitate along with MAP2 in the microtubule pellet (data not shown). These data indicate that microtubule binding and SH3 domain binding are mutually exclusive and that rMAP2c shows preference to microtubule binding even in the presence of excess Src-SH3 or Grb2-N-SH3 domains. Thus, Src or Grb2 would interact with soluble, non-microtubule-associated MAP2.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Recombinant MAP2c binds microtubules in the presence of GST-Src-SH3 and Grb2-N-SH3. The microtubule binding activity of full-length rMAP2c was determined in the absence and presence of GST, GST-Src-SH3, and Grb2-N-SH3 proteins as described under "Experimental Procedures." The supernatant and pellet samples were analyzed by SDS-PAGE. The separated proteins were stained with Sypro Orange and quantified by fluorometry. The data are presented as a percentage of microtubule-bound rMAP2c over total rMAP2c. There was no significant difference in rMAP2c binding to microtubules in the presence of excess GST, GST-Src-SH3, or Grb2-N-SH3 (p  0.05; n = 5; one-way analysis of variance).

Phosphorylation of rMAP2c by MAPK Reduces Binding to Src and Grb2-- MAP2 proteins isolated from brain tissues are highly and differentially phosphorylated (16, 54, 55). Phosphorylation regulates microtubule binding (56), and activity-dependent phosphorylation and dephosphorylation of MAP2 have been observed (15-20). Therefore, we tested the effects of phosphorylation on the SH3 domain binding activity of MAP2. The cAMP-dependent protein kinase (PKA) rapidly phosphorylated rMAP2c within 1 h (Fig. 8A). In contrast, the mitogen-activated, proline-directed protein kinase MAPK (extracellular signal-regulated kinase 2, p42) achieved comparable phosphorylation of rMAP2c after 8 h under similar kinase reaction conditions in vitro (Fig. 8A). Under the conditions used, both protein kinases maximally phosphorylate rMAP2c to a stoichiometry of approximately 3 mol of phosphate/mol.³ The PKA-phosphorylated rMAP2c bound SH3 domains of Src and Grb2. All rMAP2c incubated in the absence of ATP (as control-phosphorylated rMAP2c) also bound SH3 domains of Src and Grb2 (Fig. 8, B and C). However, rMAP2c phosphorylated by MAPK for 8 h exhibited significantly reduced interaction with the SH3 domains of both Src and Grb2 (Fig. 8C).

View larger version (33K): [in this window] [in a new window]   Fig. 8.   Phosphorylation by MAPK (extracellular signal-regulated kinase 2, p42) regulates rMAP2c association with SH3 domains of Src and Grb2. Recombinant MAP2c was incubated in the absence (control) or presence of ATP and with either PKA or MAPK for various times, as indicated, prior to use in the GST-SH3 binding. Immunoblots were stained for MAP2c with anti-MAP2 antibody HM2 and detected by colorimetric reaction. A, autoradiograph showing time-dependent phosphorylation of rMAP2c by PKA or MAPK. B, immunoblots of phosphorylated and mock-phosphorylated (control) rMAP2c co-precipitated with GST-Src-SH3. C, immunoblots of phosphorylated and control rMAP2c co-precipitated with GST-Grb2-N-SH3. Results in B and C are representative of three independent experiments.

Because MAPK is a proline-directed protein kinase, phosphorylation by MAPK would theoretically occur at serine and threonine residues that are adjacent to proline residues. To investigate whether phosphorylation near proline residues is involved in regulating MAP2/SH3 domain interactions, proline-rich deletion mutant rMAP2c proteins were phosphorylated by MAPK for 8 h and then tested for SH3 domain binding. The SH3 domain binding activities of all proline-rich deletion mutant rMAP2c proteins were inhibited by MAPK phosphorylation (data not shown). In addition, full-length rMAP2c was maximally phosphorylated using another proline-directed kinase GSK-3 for 24 h and analyzed for Src-SH3 and Grb2-N-SH3 domain binding activity. The GSK-3-phosphorylated rMAP2c demonstrated binding to the selected SH3 domains that was comparable to that of the full-length, non-phosphorylated rMAP2c (data not shown). Together, these data suggest that the addition of phosphate groups to regions other than the proline-rich areas in MAP2c regulate the interaction with SH3 domains. Key phosphorylation sites may lie within the MTBD/SH3 domain-binding region itself.

In MAP2c, the MTBD/SH3 domain-binding region contains three KXGS motifs that are targets of multiple kinases. Phosphorylation of the serine residues in the KXGS motifs abolishes microtubule binding in vitro (22). Furthermore, full-length rMAP2c containing serine to glutamate mutations in the KXGS motifs exhibit greatly reduced microtubule binding in vivo.² We tested the SH3 domain binding activity of these mutants and found that they retained SH3 domain binding activity in vitro that was indistinguishable from that of the full-length, wild type rMAP2c (data not shown). This observation indicates that phosphorylation at the KXGS motifs in the MTBD is unlikely to modulate the binding of Src-SH3 and Grb2-SH3 domains and that other candidate residues must be considered.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

This study presents evidence of novel and direct interactions of the intracellular signaling proteins Src and Grb2 with soluble, non-microtubule-bound MAP2 in neurons. These interactions were mediated through the SH3 domains of Src and Grb2 and the MTBD of MAP2 and were regulated by MAPK-dependent phosphorylation. Previous studies showed that MAP2 is a substrate of MAPK in vivo (18, 19).

SH3 domains are well documented protein-protein interaction modules found in multiple signaling pathways in cells, and they mediate formation of large multiprotein complexes (57). MAP2c contains eleven PXXP motifs that are conserved across species. The PXXP motif has been demonstrated to be the minimum consensus ligand motif of SH3 domains (58). Surprisingly, none of the PXXP motifs in MAP2c were involved in binding the SH3 domains of Src or Grb2 under the described experimental conditions. Instead, the SH3 domain-binding region in MAP2c was localized to the same region responsible for microtubule binding, the MTBD, amino acids 300-400. However, it seems likely that the binding of microtubules versus SH3 domains to this region of MAP2 involve distinct conformations, because serine to glutamate mutations in the KXGS motifs that severely disrupt microtubule binding had no effect on SH3 domain binding.

The related neuronal microtubule-associated protein tau was also reported to bind the Src family of protein tyrosine kinases, Fyn, Lck, and Src, through SH3 domains of the kinases (8). The Fyn-SH3 domain-binding site in tau was mapped to a proline-rich region of amino acids 170-178, ¹⁷⁰VVRTPPKSP¹⁷⁸, in the 352-amino acid isoform of human tau. This region accounted for ~90% of the Fyn-SH3 domain binding activity in tau (8). A similar sequence, ²⁸⁶IIRTPPKSP²⁹⁴, is also found in MAP2c (Fig. 9). However, this sequence does not appear to participate in binding the SH3 domains of Src and Grb2, because the 290-297 rMAP2c mutant bound to both Src-SH3 and Grb2-N-SH3, and the binding activities were indistinguishable from those of the full-length rMAP2c protein (Fig. 5B). The amino acids in the vicinity of ²⁸⁶IIRTPPKSP²⁹⁴ in MAP2c differ from those of tau (Fig. 9). It is possible that neighboring amino acid composition influences the binding specificity of Src in MAP2c versus tau.

View larger version (46K): [in this window] [in a new window]   Fig. 9.   Amino acid sequence similarity between human tau and rat MAP2c. Sequence alignment of the carboxyl half of the human tau protein (GenBankTM accession number J03778) (73) and rat neonatal MAP2c (GenBankTM accession number X17682) (74). The boxed regions denote the imperfect repeats of the MTBD, and the underlined region denotes the tyrosine kinase Fyn-SH3 domain-binding region in the human tau protein (352-amino acid isoform) (8). Note that this region is similar in both proteins, although significant differences occur in the flanking regions.

Whereas PXXP motifs are ideal binding ligands for SH3 domains, non-proline-rich ligands have also been documented. For example, the Sab protein preferentially associates with Bruton's tyrosine kinase through a non-proline-rich sequence (59). Similarly, the c-Src-SH3 domain intramolecularly binds a stretch of 14 amino acids (KPQTQGLAKDAWEI) in the SH2-kinase linker region in the molecule (60, 61). In an ideal PXXP-SH3 interaction, the PXXP ligand assumes a polyproline II left-handed helix such that the alternating proline and arginine residues make contact with tyrosine, tryptophan, and aspartate residues in the hydrophobic pocket of the SH3 domain (62, 63). In the inactive c-Src, the SH2-kinase linker region adopts a partial pseudo-polyproline II left-handed helix conformation and binds the SH3 domain to form an intramolecular interaction (60, 61). It is possible that certain regions in the MTBD of MAP2c assume a similar pseudo-polyproline II left-handed helix conformation that is capable of interacting with the hydrophobic pockets of Src-SH3 and Grb2-SH3. Structural studies of the MTBD/SH3 domain complexes will be needed to reveal the true nature of these interactions.

The fact that microtubules and the SH3 domains of Src and Grb2 bind to the MTBD of MAP2c has two major implications. First, the 11 conserved PXXP motifs on MAP2c remain available to interact with other cellular proteins. With MAP2 acting as a scaffold protein for Src, these proteins could become potential substrates for the MAP2c-bound Src tyrosine kinase activity. Second, because microtubule binding and SH3 domain binding are mutually exclusive, the non-microtubule-bound pool of MAP2 could assume the additional physiological role of a scaffold protein for Src and Grb2. Both microtubule binding and SH3 domain binding are regulated by phosphorylation. Differential phosphorylation could presumably regulate the dual roles of MAP2 in vivo (scaffold versus cytoskeletal). Our data suggest that MAPK may regulate MAP2/Src/Grb2 interaction in vivo. Phosphorylation within the three KXGS motifs in the MTBD releases MAP2 from microtubules (22) and thus is predicted to increase the pool of soluble MAP2 available for binding and organizing signaling molecules. In addition, a subpopulation of non-microtubule-bound MAP2c is observed to co-localize with microfilament-based structures in transfected cells² and neurons (64). This suggests that MAP2 could also function to shuttle signaling molecules to the microfilament system in the dendrites of neurons.

Although MAP2 did not appear to be a target of Src in vivo, other potential cellular substrates of Src in the dendrites include the microtubules, microfilaments, and proteins directly or indirectly associated with MAP2. Phosphorylation could release Src from MAP2 to augment an increase in membrane-associated Src at the growth cone. In this way, MAP2-bound Src in developing dendrites might be directed to regions of active microtubule dynamics, such as nerve growth cones, for participation in neuronal morphogenesis. - and -tubulins, the building blocks of microtubules, have been reported to be phosphorylated by pp60^c-Src at nerve growth cone membranes (65). Because soluble MAP2 associates with microfilaments, actin and actin-binding proteins could also be targeted by Src. For example, F-actin and the actin-membrane linker protein ezrin are substrates of tyrosine kinases in situ (66-68). Growth factor-induced increases in protein tyrosine phosphorylation and a consequent reorganization of the actin cytoskeleton are well documented (69). Other proteins that bind MAP2, including the type II regulatory subunit of PKA, calmodulin, and the mitochondrial outer membrane protein porin, might also be substrates of Src. In addition, a Src-dependent signaling cascade might affect other yet-to-be-identified proteins associated with MAP2 and the microtubule cytoskeleton. The adult isoforms of MAP2 are large molecules (~200 kDa), and the function of more than 90% of the molecule is still unknown. The binding of MAP2 with other proteins could occur via PXXP motifs or other regions in the molecule. These proteins could then become substrates of Src. The discovery of new MAP2 binding partners would have implications for understanding the relationship between cell signaling and the neuronal cytoskeleton, particularly in areas of neurite growth and synaptic plasticity.

Interactions of proteins with the adaptor protein Grb2 on MAP2 could potentially generate multimeric signaling complexes on soluble MAP2. Grb2 has two SH3 domains and a SH2 domain and is known to bridge receptor and cytoplasmic tyrosine kinases with the Ras signaling pathway and MAPK activation (70, 71). The amino-terminal SH3 domain exhibited stronger binding to rMAP2c than the carboxyl-terminal SH3 domain. The amino-terminal SH3 domain could be MAP2-bound, leaving the carboxyl-terminal SH3 domain free for binding other proteins. In addition, the SH2 domain could bind tyrosine-phosphorylated proteins. Thus, in vivo, Grb2 might mediate the indirect interaction between MAP2 and cellular proteins, such as activated tyrosine kinases, tyrosine phosphatases, and other adaptor molecules (57, 70-72). Phosphorylation of MAP2 by MAPK would then release the multimeric signaling complexes into the cytoplasmic compartment, and the complexes could then join up with the appropriate effector molecules downstream of the signaling cascades.

In conclusion, the demonstrated direct interactions between MAP2/Src and MAP2/Grb2, and the proposed phosphorylation-dependent scaffolding role of MAP2 provide a plausible mechanism for modulation of the neuronal cytoskeleton by extracellular signals. The interaction also provides a mechanism for the local organization of signaling complexes in the neuron. Signal-dependent modulation of the neuronal cytoskeleton is necessary during the growth of dendrites in order to form intricate neural circuitry. Future experiments aimed at uncovering new proteins associated with MAP2 and the neuronal cytoskeleton will no doubt provide further insight into the physiological significance of these novel interactions in the developing nervous system.

	ACKNOWLEDGEMENTS

We thank Dr. Gary M. Bokoch for gifts of GST-SH3 expression E. coli; Dr. Klaus Hahn for the COS-7 cells; Rachel Ozer for the pEGFP-N1-MAP2c and the various serine mutated rMAP2c plasmids; Arlene Hipolito for help with cell culture, staining, and fluorescence microscopy; Amy Batinica for the construction of the 219-290 rMAP2c expression plasmid; Dr. Brian Ceresa for a gift RC20-HRP; Dr. Jawdat Al-bassam for bovine tubulin; Dr. Jennifer Kowalski for assistance in purification of Grb2-N-SH3-cleaved fragments; Camilo Orozco for expert computer assistance; and Melissa Paine for help with the purification of HisTrxMTBD protein.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant MH 50861 (to S. H.).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.

To whom correspondence should be addressed: Department of Cell Biology, The Scripps Research Institute, 10550 N. Torrey Pines Rd., La Jolla, CA 92037. Tel.: 858-784-2420; Fax: 858-784-2513; E-mail: shelley@scripps.edu.

Published, JBC Papers in Press, April 25, 2000, DOI 10.1074/jbc.M001887200

² R. Ozer and S. Halpain, submitted for publication.

³ A. Batinica and S. Halpain, unpublished data.

	ABBREVIATIONS

The abbreviations used are: MAP2, microtubule-associated protein 2; MTBD, microtubule-binding domain; PKA, protein kinase A; MAPK, mitogen-activated protein kinase; rMAP2c, recombinant bacterially expressed MAP2c; SP, sulfopropyl; HRP, horseradish peroxidase; PMSF, phenylmethylsulfonyl fluoride; GSK-3, glycogen synthase kinase-3; PBS, phosphate-buffered saline; PAGE, polyacrylamide gel electrophoresis; GST, glutathione S-transferase; HisTrx, histidine-tagged thioredoxin fusion protein; HisTrxMTBD, histidine-tagged thioredoxin-MTBD fusion protein; IP, immunoprecipitate; GFP, green fluorescent protein; SH, Src homology.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				B. Wu, B. Decourt, M. A. Zabidi, L. T. Wuethrich, W. H. Kim, Z. Zhou, K. MacIsaac, and D. M. Suter Microtubule-mediated Src Tyrosine Kinase Trafficking in Neuronal Growth Cones Mol. Biol. Cell, November 1, 2008; 19(11): 4611 - 4627. [Abstract] [Full Text] [PDF]

				M. H. Soltani, R. Pichardo, Z. Song, N. Sangha, F. Camacho, K. Satyamoorthy, O. P. Sangueza, and V. Setaluri Microtubule-Associated Protein 2, a Marker of Neuronal Differentiation, Induces Mitotic Defects, Inhibits Growth of Melanoma Cells, and Predicts Metastatic Potential of Cutaneous Melanoma Am. J. Pathol., June 1, 2005; 166(6): 1841 - 1850. [Abstract] [Full Text] [PDF]

				S. P. Zamora-Leon, A. Bresnick, J. M. Backer, and B. Shafit-Zagardo Fyn Phosphorylates Human MAP-2c on Tyrosine 67 J. Biol. Chem., January 21, 2005; 280(3): 1962 - 1970. [Abstract] [Full Text] [PDF]

				T. Phung-Koskas, A. Pilon, C. Pous, C. Betzina, M. Sturm, M.-L. Bourguet-Kondracki, G. Durand, and A. Drechou STAT5B-mediated Growth Hormone Signaling Is Organized by Highly Dynamic Microtubules in Hepatic Cells J. Biol. Chem., January 14, 2005; 280(2): 1123 - 1131. [Abstract] [Full Text] [PDF]

				L. M. Salvador, M. P. Flynn, J. Avila, S. Reierstad, E. T. Maizels, H. Alam, Y. Park, J. D. Scott, D. W. Carr, and M. Hunzicker-Dunn Neuronal Microtubule-associated Protein 2D Is a Dual A-kinase Anchoring Protein Expressed in Rat Ovarian Granulosa Cells J. Biol. Chem., June 25, 2004; 279(26): 27621 - 27632. [Abstract] [Full Text] [PDF]

				E. Laurine, D. Lafitte, C. Gregoire, E. Seree, E. Loret, S. Douillard, B. Michel, C. Briand, and J.-M. Verdier Specific Binding of Dehydroepiandrosterone to the N Terminus of the Microtubule-associated Protein MAP2 J. Biol. Chem., August 8, 2003; 278(32): 29979 - 29986. [Abstract] [Full Text] [PDF]

				J. Al-Bassam, R. S. Ozer, D. Safer, S. Halpain, and R. A. Milligan MAP2 and tau bind longitudinally along the outer ridges of microtubule protofilaments J. Cell Biol., June 24, 2002; 157(7): 1187 - 1196. [Abstract] [Full Text] [PDF]

				D. Belanger, C. A. Farah, M. D. Nguyen, M. Lauzon, S. Cornibert, and N. Leclerc The projection domain of MAP2b regulates microtubule protrusion and process formation in Sf9 cells J. Cell Sci., January 4, 2002; 115(7): 1523 - 1539. [Abstract] [Full Text] [PDF]

				S. P. Zamora-Leon, G. Lee, P. Davies, and B. Shafit-Zagardo Binding of Fyn to MAP-2c through an SH3 Binding Domain. REGULATION OF THE INTERACTION BY ERK2 J. Biol. Chem., October 19, 2001; 276(43): 39950 - 39958. [Abstract] [Full Text] [PDF]

				D. Fang, J. Hallman, N. Sangha, T. E. Kute, J. A. Hammarback, W. L. White, and V. Setaluri Expression of Microtubule-Associated Protein 2 in Benign and Malignant Melanocytes : Implications for Differentiation and Progression of Cutaneous Melanoma Am. J. Pathol., June 1, 2001; 158(6): 2107 - 2115. [Abstract] [Full Text] [PDF]

				R. S. Ozer and S. Halpain Phosphorylation-dependent Localization of Microtubule-associated Protein MAP2c to the Actin Cytoskeleton Mol. Biol. Cell, October 1, 2000; 11(10): 3573 - 3587. [Abstract] [Full Text]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/27/20578    most recent M001887200v1 Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted 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 Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Lim, R. W. L. Articles by Halpain, S. Search for Related Content PubMed PubMed Citation Articles by Lim, R. W. L. Articles by Halpain, S. Social Bookmarking                      What's this?