Involvement of the c-Src-Crk-C3G-Rap1 signaling in the nectin-induced activation of Cdc42 and formation of adherens junctions.

Nectins, Ca(2+)-independent immunoglobulin-like cell-cell adhesion molecules, induce the activation of Cdc42 and Rac small G proteins, enhancing the formation of cadherin-based adherens junctions (AJs) and claudin-based tight junctions. Nectins recruit and activate c-Src at the nectin-based cell-cell contact sites. c-Src then activates Cdc42 through FRG, a Cdc42-GDP/GTP exchange factor. We showed here that Rap1 small G protein was involved in the nectin-induced activation of Cdc42 and formation of AJs. Rap1 was recruited to the nectin-based cell-cell contact sites and locally activated through the c-Src-Crk-C3G signaling there. The activation of either c-Src or Rap1 alone was insufficient for and the activation of both molecules was essential for the activation of FRG. The activation of Rap1 was not necessary for the c-Src-mediated phosphorylation or recruitment of FRG. The inhibition of the Crk, C3G, or Rap1 signaling reduced the formation of AJs. These results indicate that Rap1 is activated by nectins through the c-Src-Crk-C3G signaling and involved in the nectin-induced, c-Src- and FRG-mediated activation of Cdc42 and formation of AJs.

activated in this process. We have recently found that nectins directly bind Par-3, raising the possibility that Cdc42 activated by the action of nectins is involved in the activation of this cell polarity protein complex (22). Thus, Cdc42 and Rac activated by the action of nectins are likely to play important roles in the formation of cell-cell junctions and cell polarity.
Extending these earlier observations, we have recently found that c-Src is recruited and activated by nectins at the nectinbased cell-cell contact sites and that c-Src activated in this way then tyrosine phosphorylates and activates FRG, a Cdc42-GEF, resulting in the activation of Cdc42 (23). Along this line, we examined here whether Rap1 is involved in the nectininduced, c-Src-and FRG-mediated activation of Cdc42 and formation of AJs.

EXPERIMENTAL PROCEDURES
Plasmid Constructions and Protein Purification-V12Rap1B, in which a glycine residue at amino acid 12 of bovine Rap1B was replaced by a valine (V), was a constitutively active mutant of Rap1 and prepared using the QuikChange site-directed mutagenesis kit (Stratagene). The cDNA of Rap1 GTPase-activating protein (GAP) was a gift from Dr. P. Casey (Duke University, Durham, NC). Expression vectors for GFPtagged Rap1 (pEGFP-Rap1), GFP-tagged V12Rap1 (pEGFP-V12Rap1), and GFP-tagged Rap1GAP (pEGFP-Rap1GAP) were constructed by inserting the cDNA fragments encoding a wild-type Rap1, V12Rap1, and Rap1GAP into pEGFP, respectively. An expression vector for Myctagged Rap1GAP (pEFBOS-Myc-Rap1GAP) was constructed by inserting the cDNA encoding Rap1GAP into pEFBOS-Myc. Expression vec-tors for FLAG-tagged FRG (pEFBOS-Flag-FRG), Raichu-Rap1 (pRaichu-Rap1), FLAG-tagged dominant negative mutant of CrkI (pIRM21-Flag-CrkI-W169L), a constitutively active mutant of C3G (pCAGGS-C3G-CA), and a dominant negative mutant of C3G (pCAGGS-C3G-dCD) were prepared as described previously (23)(24)(25). Expression vectors for GFP-tagged constitutively active mutants of Cdc42 (pEGFP-V12Cdc42) and Rac1 (pEGFP-V12Rac1) were kindly supplied by Dr. S. Narumiya (Kyoto University, Kyoto, Japan). Expression vectors for a wild-type c-Src (pcDNA-c-Src) and a constitutively active mutant of c-Src (pUSE-c-Src-CA) were constructed by inserting the full-length cDNA encoding c-Src into pcDNA plasmid and purchased from Upstate Biotechnology, respectively. pEGFP-CAAX was prepared as described previously (26). The CAAX motif consists of the polybasic sequence of amino acids (C is Cys, A is an aliphatic amino acid, and X is a variety of amino acids) and preferentially targets proteins to the plasma membrane (27). Expression vectors for glutathione S-transferase (GST)-RalGDS-Ras-binding domain (pGEX-RalGDS-RBD) and GST-PAK-Cdc42/Rac interactive binding region (pGEX-PAK-CRIB) were kind gifts from Dr. A. Kikuchi (Hiroshima University, Hiroshima, Japan) and Dr. T. Takenawa (The University of Tokyo, Tokyo, Japan), respectively. All the constructs used here were confirmed by sequencing. The GST-RalGDS-RBD fusion protein and GST-PAK-RBD fusion protein were overexpressed and purified from Escherichia coli according to the manufacturer's protocol (Amersham Biosciences). Protein concentrations were determined by using the BCA protein assay kit (Pierce) with bovine serum albumin as a reference protein.
Antibodies-Rabbit anti-nectin-1 and C3G polyclonal antibodies (pAbs) were prepared as described previously (24,28). A rabbit anti-Rap1/Krev-1 pAb (Santa Cruz Biotechnology), a rabbit anti-phospho- Nectin-1-L cells transiently overexpressing GFP-Rap1GAP or GFP were cultured on the Nef-3-or IgG-coated coverslips for 1 h. The cells were fixed and stained for actin filaments (F-actin) with rhodaminephalloidin. Scale bars, 10 m. Bars in the graph represent percentage of cells with filopodia (F) or lamellipodia (L) of the total 100 cells counted and are expressed as means Ϯ S.E. of three independent experiments. B and C, pull-down assay. Nectin-1-L cells transiently overexpressing Myc-Rap1GAP or not were cultured with clustered Nef-3 or IgG for 30 min and then subjected to the pull-down assay using GST-PAK-CRIB followed by Western blotting using the anti-Cdc42 and anti-Rac1 mAbs. B, Cdc42; C, Rac1. Bars in the graphs of B and C represent the relative intensity of GTP-Cdc42 or GTP-Rac1 normalized for the total amount of Cdc42 or Rac1, respectively, as compared with a value of clustered IgG stimulation, which is expressed as 1. The results shown are representative of three independent experiments.
Cell Culture and Fluorescent Resonance Energy Transfer (FRET) Imaging-L or MDCK cells stably expressing exogenous nectin-1 (nectin-1-L or nectin-1-MDCK cells, respectively) were prepared as described previously (28). Transfection and immunofluorescence microscopy of cultured cells with a confocal imaging system (Radian 2000, Bio-Rad) were performed as described previously (28). FRET imaging was also performed as described previously (23,30). In brief, nectin-1-L cells were transfected with pRaichu-Rap1. After a 24-h transfection, the cells were replated on the dishes coated with 50 g/ml Nef-3 or 50 g/ml IgG. The cells were then imaged with an Olympus IX71 inverted microscope equipped with a cooled charge-coupled device camera, Cool-SNAP HQ (Roper Scientific), controlled by MetaMorph software (Universal Imaging). For dual emission ratio imaging, we used a 440AF21 excitation filter, a 455DRLP dichroic mirror, and two emission filters, 480AF30 for CFP and 535AF25 for YFP (Omega Optical Inc.). The cells were illuminated with a 75-watt xenon lamp through a 6% neutral density filter (Omega Optical Inc.) and a 60ϫ oil immersion objective lens. Exposure times for 3 ϫ 3 binning were 200 ms to obtain images of CFP and YFP and 50 ms to obtain images of a differential interference contrast. After background subtraction, the ratio image of YFP/CFP was created with the MetaMorph software and used to represent FRET efficiency.
Assay for Cell Spreading and Bead-Cell Contact-The cell spreading assay was performed as described previously (23). For inhibition of the activities of Src family kinases, the cells were treated with 20 M PP2 (Calbiochem-Novabiochem) dissolved in 0.2% Me 2 SO, 20 M PP3 (Calbiochem-Novabiochem) dissolved in 0.2% Me 2 SO, 10 M SU6656 (Sigma) dissolved in 0.2% Me 2 SO, or 0.2% Me 2 SO as a control in Dulbecco's modified Eagle's medium (DMEM) for 1 h before the trypsin-EDTA treatment. The cells were also treated with 20 M PP2 or 20 M PP3 during the trypsin-EDTA treatment and cultured on the coverslips coated with Nef-3 or IgG. The bead-cell contact was assayed as described previously (23). Briefly nectin-1-L cells were seeded on each well of a 24-well plate and cultured in DMEM containing 10% fetal calf serum for 12 h. Latex-sulfate microbeads coated with the recombinant extracellular fragment of nectin-3 fused to the human IgG (Nef-3) or IgG were added to each well. After a 30-min incubation, the cells were fixed and immunostained.
Pull-down Assay for Small G Proteins-The pull-down assay was performed as described previously (23). In brief, 26 g of Nef-3 or human IgG (Fc-specific) (Sigma) was clustered using 9 g of the antihuman IgG pAb (Sigma) in 50 l of phosphate-buffered saline at room temperature for 1 h. Nectin-1-L cells (5 ϫ 10 5 cells/60-mm dish) were transiently transfected with the plasmid of Rap1GAP and cultured for 8 h. After a 16-h serum starvation, the medium was replaced by 1 ml of DMEM containing clustered Nef-3 or IgG, and the cells were incubated for the indicated periods of time. The cells were then washed with 1 ml of ice-cold phosphate-buffered saline containing 1 mM sodium vanadate, lysed in Buffer A (50 mM Tris/HCl, pH 7.4, 150 mM NaCl, 5 mM MgCl 2 , 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium vanadate) containing 10 g GST fusion proteins, GST-RalGDS-RBD for Rap1, or GST-PAK-CRIB for Cdc42 and Rac, and incubated at 2°C for 30 min. The cell extract was obtained by centrifugation at 20,000 ϫ g at 0°C for 5 min and incubated with 50 l of glutathione-agarose beads (Amersham Biosciences) at 2°C for 1 h. After the beads were washed with Buffer A, proteins bound to the beads were eluted with Laemmli buffer and subjected to SDS-PAGE followed by Western blotting.
Assay for Immunoprecipitation and GEF Activity-The immunoprecipitation assay was performed as described previously (23). In brief, nectin-1-L cells (1.5 ϫ 10 6 cells/100-mm dish) were transfected with the combination of pEFBOS-Flag-FRG, pcDNA-c-Src, pEGFP, pEGFP-V12Rap1, or pEGFP-Rap1GAP. After an 8-h culture, the cells were serum-starved for 16 h. The medium was then replaced by 3 ml of DMEM containing clustered Nef-3 or IgG, and the cells were incubated for the indicated periods of time. The cells were washed twice with 3 ml of ice-cold phosphate-buffered saline containing 1 mM sodium vanadate, lysed in 0.5 ml of Buffer B (20 mM Tris/HCl, pH 7.4, 150 mM NaCl, 2 mM EDTA, 1% TritonX-100, 1% sodium deoxycholate, 0.1% SDS, 10 g/ml leupeptin, 10 g/ml aprotinin, 10 M (p-amidinophenyl)methanesulfonyl fluoride hydrochloride, 20 mM ␤-glycerophosphate/sodium, 10 mM sodium fluoride, 1 mM sodium vanadate, and phosphatase inhibitor mixture 1), and incubated on ice for 15 min. The cell extract was obtained by centrifugation at 20,000 ϫ g at 4°C for 15 min, incubated with 5 l of the anti-FLAG mAb at 4°C for 1 h, and further incubated with 50 l of protein G-Sepharose 4 Fast Flow beads (Amersham Biosciences) for 1 h. After the incubation, the beads were washed with Buffer B. Proteins bound to the beads were eluted with Laemmli buffer and subjected to SDS-PAGE followed by Western blotting. The GEF activity was assayed as described previously (31). The binding of [ 35 S]GTP␥S to Cdc42 (5 pmol) was assayed by measuring the radioactivity of [ 35 S]GTP␥S bound to Cdc42 after incubation for the indicated periods of time in the presence of each immunoprecipitant from the cells treated with clustered Nef-3 or IgG for 30 min.

Necessity of Rap1 for the Nectin-induced Activation of Cdc42
and Rac-Rap1GAP has been shown to hydrolyze GTP bound to Rap1 and to inactivate it (32). We first examined by using Rap1GAP whether Rap1 is involved in the nectin-induced activation of Cdc42 and Rac. We have previously shown by using Nef-3 that nectins induce the formation of filopodia and lamellipodia through the respective activation of Cdc42 and Rac in nectin-1-L and nectin-1-MDCK cells (23,33). Nectin-1-L cells were transfected with the plasmid of GFP-Rap1GAP or GFP as a control. After a 24-h culture, the cells were cultured on the Nef-3-coated coverslips. Neither filopodia nor lamellipodia were formed in nectin-1-L cells transiently overexpressing GFP-Rap1GAP, whereas these protrusions were markedly formed in nectin-1-L cells transiently overexpressing GFP (Fig.  1A). When nectin-1-L cells transiently overexpressing GFP-Rap1GAP or GFP were cultured on the IgG-coated coverslips as a control, neither filopodia nor lamellipodia were formed. The essentially same results were obtained for nectin-1-MDCK cells FIG. 3. No restoration by a constitutively active mutant of Rap1 or C3G of Src family kinase inhibition on the Nef-3-induced formation of filopodia and lamellipodia and activation of Cdc42 and Rac. A, cell spreading assay. Nectin-1-L cells transiently overexpressing GFP-V12Rap1, C3G-CA, or GFP were cultured on the Nef-3-coated coverslips in the presence of PP2, PP3, or SU6656 for 1 h. The cells were fixed and stained for F-actin and C3G-CA with rhodamine-phalloidin and the anti-C3G pAb, respectively. Scale bars, 10 m. Bars in the graph represent percentage of cells with filopodia (F) and lamellipodia (L) of the total 100 cells counted and are expressed as means Ϯ S.E. of three independent experiments. B, pull-down assay. Nectin-1-L cells overexpressing GFP-V12Rap1, Myc-Rap1GAP, c-Src-CA, or not were cultured with clustered Nef-3 or IgG in the presence or absence of PP2, PP3, SU6656, or Me 2 SO (DMSO) for 30 min. After the cells were washed, they were lysed and subjected to the pull-down assay using GST-PAK-CRIB followed by Western blotting using the anti-Cdc42 or anti-Rac1 mAbs. that were transiently overexpressing GFP-Rap1GAP or GFP as a control (data not shown). We then confirmed by the pull-down assay using GST-PAK-CRIB that Rap1GAP inhibited the Nef-3-induced activation of Cdc42 and Rac in nectin-1-L cells (Fig.  1, B and C). These results indicate that Rap1 is necessary for the nectin-induced activation of Cdc42 and Rac and the subsequent formation of filopodia and lamellipodia, respectively.
Rap1 as an Upstream Regulator of Cdc42 and Rac and a Downstream Regulator of c-Src-If Rap1 functions upstream of Cdc42 and Rac, overexpression of V12Cdc42 and V12Rac1 in nectin-1-L cells expressing Rap1GAP would suppress the inhibitory effect of Rap1GAP on the Nef-3-induced formation of filopodia and lamellipodia. To test this possibility, nectin-1-L cells transiently co-expressing GFP-V12Cdc42, GFP-V12Rac1, or GFP as a control with Myc-Rap1GAP were cultured on the Nef-3-coated coverslips. Neither filopodia nor lamellipodia were formed in the cells co-expressing GFP with Myc-Rap1GAP, whereas both protrusions were formed in the cells expressing GFP alone as described above (Fig. 2 and see also Fig. 1A). Both filopodia and lamellipodia were formed to large and moderate extents, respectively, in the cells co-expressing GFP-V12Cdc42 with Myc-Rap1GAP, whereas lamellipodia were mainly formed to a large extent in the cells co-expressing GFP-V12Rac1 with Myc-Rap1GAP. We have previously shown that the activation of Cdc42 is necessary, but not sufficient, for the activation of Rac in nectin-1-L cells and that the activation of Rac is not necessary for the activation of Cdc42 (23,33). Taken together, these results indicate that Rap1 functions upstream of Cdc42 and Rac in the nectin-induced formation of filopodia and lamellipodia.

FIG. 4. Nef-3-induced activation and recruitment of Rap1.
A, pull-down assay. Nectin-1-L cells were washed and subjected to the pull-down assay using GST-RalGDS-RBD followed by Western blotting using the anti-Rap1 pAb. a, nectin-1-L cells cultured with clustered Nef-3 or IgG for the indicated periods of time; b, nectin-1-L cells cultured with clustered Nef-3 or IgG for 30 min in the presence of PP2, PP3, or SU6656; c, nectin-1-L cells overexpressing c-Src-CA or an empty vector cultured with clustered Nef-3 or IgG for 30 min. The results shown are representative of three independent experiments. B, FRET imaging. Nectin-1-L cells transiently overexpressing Raichu-Rap1 were cultured on the Nef-3-coated dishes for 15 min and imaged for YFP, CFP, and differential interference contrast (DIC) for the indicated periods of time after plating. YFP/CFP ratio images are shown in the pseudocolor mode with the corresponding differential interference contrast images. In the intensity modulated display (IMD) mode, eight colors from red to blue were used to represent the YFP/ CFP ratio with the intensity of each color indicating the mean intensity of YFP and CFP. High YFP/CFP ratio shown in red color indicates high FRET efficiency of the probe, reflecting high GTP/GDP ratio of Rap1. The upper and lower limits of ratio range are shown. Scale bars, 10 m. The results shown are representative of 10 independent experiments. C, bead-cell assay. Nectin-1-L cells transiently overexpressing GFP-Rap1 were incubated with the Nef-3-or IgG-coated beads for 30 min. The cells were fixed and immunostained for nectin-1 with the anti-nectin-1 pAb. Scale bars, 10 m. Positions of the beads are marked with asterisks. Bars in the graphs represent percentage of the beadcell contact sites with the signal for nectin-1 of the total bead-cell contact sites counted (left) or percentage of the beadcell contact sites with the signal for GFP of the bead-cell contact sites with the signal for nectin-1 counted (right) and are expressed as means Ϯ S.E. of three independent experiments in which a total of 50 bead-cell contact sites were counted.
We have previously shown that c-Src functions downstream of nectins and upstream of Cdc42 and Rac (23). To clarify that Rap1 functions downstream of c-Src, we examined whether the inhibitory effect of Rap1GAP on the Nef-3-induced formation of filopodia and lamellipodia in nectin-1-L cells was not suppressed by co-expression of c-Src-CA. When nectin-1-L cells co-expressing Myc-Rap1GAP with c-Src-CA were cultured on the Nef-3-coated coverslips, neither filopodia nor lamellipodia were formed (Fig. 2). In addition, c-Src-CA enhanced the Nef-3-induced activation of Rap1 in nectin-1-L cells as estimated by the pull-down assay for Rap1 using the GST-RalGDS-RBD (see Fig. 4A, c). These results indicate that Rap1 functions downstream of c-Src.
Insufficiency of the Activation of Rap1 Alone for the Nectininduced Activation of Cdc42 and Rac-We have previously shown that PP2, an inhibitor of Src family kinases, inhibits the Nef-3-induced formation of filopodia and lamellipodia (23). We examined whether the activation of Rap1 alone is sufficient for the Nef-3-induced formation of filopodia and lamellipodia. Nectin-1-L cells expressing GFP-V12Rap1 or GFP as a control were cultured on the Nef-3-coated coverslips in the presence of PP2 or PP3, an inactive analogue of PP2. In nectin-1-L cells expressing GFP, filopodia and lamellipodia were markedly formed in the presence of PP3, whereas both protrusions were negligibly formed in the presence of PP2 (Fig. 3A). This inhibitory effect of PP2 on the Nef-3-induced formation of filopodia and lamellipodia was confirmed by SU6656, another inhibitor of Src family kinases (34). This inhibitory effect of PP2 was not rescued by GFP-V12Rap1 or C3G-CA, a constitutively active mutant of Rap1-GEF (24). We obtained the essentially same results for nectin-1-MDCK cells (data not shown). These results indicate that the activation of Rap1 is necessary, but not sufficient, for the nectin-induced formation of filopodia and lamellipodia.
We further confirmed that the activation of Rap1 is necessary, but not sufficient, for the Nef-3-induced activation of Cdc42 and Rac. The Nef-3-induced activation of Cdc42 and Rac was inhibited by PP2 and SU6656, but not by PP3, as estimated by the pull-down assay using GST-PAK-CRIB in nectin-1-L cells (Fig. 3B). The inhibition by PP2 of the Nef-3-induced activation of Cdc42 and Rac was not restored by V12Rap1. On the other hand, co-expression of c-Src-CA and Rap1GAP did not induce the activation of Cdc42 or Rac. Taken together, these results indicate that the activation of either c-Src or Rap1 alone is insufficient for and the activation of both is required for the nectin-induced activation of Cdc42 and Rac and formation of filopodia or lamellipodia, respectively.
Nectin-induced, c-Src-mediated Activation of Rap1-We then examined whether Nef-3 indeed activates Rap1. Nectin-1-L cells were incubated with clustered Nef-3 or IgG as a control and then subjected to the pull-down assay using GST-RalGDS-RBD. The incubation of clustered Nef-3 increased the GTP-bound form of Rap1 at 30 min, whereas the incubation of clustered IgG did not increase the GTP-bound form of Rap1 (Fig. 4A, a). To confirm again that c-Src is involved in the nectin-induced activation of C3G-dCD, GFP-V12Rap1, or GFP were cultured on the Nef-3-coated coverslips for 1 h. The cells were fixed and stained for F-actin, Flag-CrkI-W169L, and C3G-dCD with rhodamine-phalloidin, the anti-FLAG mAb, and the anti-C3G pAb, respectively. Scale bars, 10 m. Bars in the graph represent percentage of cells with filopodia (F) and lamellipodia (L) of the total 100 cells counted and are expressed as means Ϯ S.E. of three independent experiments. B, pull-down assay. Nectin-1-L cells overexpressing Flag-CrkI-W169L, C3G-dCD, or an empty vector were cultured with clustered Nef-3 or IgG for 30 min and then subjected to the pull-down assay using GST-RalGDS-RBD for Rap1 and GST-PAK-CRIB for Cdc42 and Rac followed by Western blotting using the anti-Rap1 pAb, anti-Cdc42 mAb, and anti-Rac1 mAb. The results shown are representative of three independent experiments. Rap1, nectin-1-L cells treated with PP3, PP2, or SU6656 were cultured with clustered Nef-3 and then subjected to the pulldown assay. The Nef-3-induced activation of Rap1 was inhibited by PP2 and SU6656, but not by PP3, in nectin-1-L cells (Fig. 4A,  b). Thus, the Nef-3-induced activation of Rap1 requires c-Src. We next monitored the activation of Rap1 in living nectin-1-L cells cultured on the Nef-3-coated dishes by FRET imaging using the FRET probe Raichu-Rap1 (30). The Nef-3-induced activation of Rap1 was observed at the periphery of the cells where filopodia were formed (Fig. 4B). In addition, Rap1 was also activated at the internal perinuclear region of the cells as described previously (30).
Because Rap1 was activated by Nef-3 near the formation of filopodia, we speculated that Rap1 would be recruited to the nectin-based cell-cell contact sites. To examine this hypothesis, nectin-1-L cells transiently overexpressing GFP-Rap1 or GFP as a control were cultured with microbeads coated with Nef-3 or IgG as a control. GFP-Rap1 accumulated at the contact sites between the Nef-3-coated beads and nectin-1-L cells but not between the IgG-coated beads and nectin-1-L cells (Fig. 4C). GFP did not accumulate at the contact sites between the Nef-3-coated beads and nectin-1-L cells (data not shown). Taken together, these results indicate that Rap1 is recruited to the nectin-based cell-cell contact sites and locally activated there.
Activation of Rap1 by C3G through Crk-C3G is a Rap1-GEF, which is activated through Crk adaptor protein (35,36). Many types of stimuli are reported to induce the binding of the Crk-C3G complex to phosphotyrosine-containing proteins, including c-Src (37). Thus, we reasoned that C3G would be a Rap1-GEF that might function downstream of nectins and c-Src. We examined whether CrkI-W169L or C3G-dCD (24) inhibits the Nef-3-induced, Cdc42-or Rac-mediated formation of filopodia or lamellipodia, respectively. Nectin-1-L cells transiently co-expressing Flag-CrkI-W169L or C3G-dCD with GFP-V12Rap1 or GFP as a control were cultured on the Nef-3-coated coverslips. Filopodia and lamellipodia were markedly formed in the cells expressing GFP-V12Rap1 or GFP alone, whereas neither protrusions were formed in nectin-1-L cells expressing Flag-CrkI-W169L or C3G-dCD with GFP (Fig. 5A). The inhibitory effects of Flag-CrkI-W169L and C3G-dCD were suppressed by co-expression of GFP-V12Rap1. The essentially same results were obtained for nectin-1-MDCK cells (data not shown). Taken together, these results indicate that Rap1 is activated by nectins through the c-Src-Crk-C3G pathway. This conclusion was further confirmed by the pull-down assay for Rap1, Cdc42, and Rac. The activation of Rap1, Cdc42, and Rac was measured by the pull-down assay using GST-RalGDS-RBD or GST-PAK-CRIB in nectin-1-L cells overexpressing Flag-CrkI-W169L, C3G-dCD, or an empty vector incubated with clustered Nef-3 or IgG. The Nef-3-induced activation of Rap1, Cdc42, and Rac was all reduced by Flag-CrkI-W169L or C3G-dCD (Fig. 5B).
Requirement of Rap1 for the Nectin-induced, c-Src-mediated Activation of FRG-We have previously shown that nectins stimulate the GEF activity of FRG, a GEF for Cdc42 (23). To confirm that Rap1 is indeed involved in the nectin-induced activation of FRG for Cdc42, we measured the GEF activity of FRG for Cdc42 in nectin-1-L cells incubated with clustered Nef-3 or IgG as a control. Nectin-1-L cells overexpressing Flag-FRG and c-Src with GFP-V12Rap1 or GFP as a control were incubated with clustered Nef-3 or IgG in the presence or absence of PP2 and subjected to the immunoprecipitation using the anti-FLAG mAb. The GEF activity of FRG for Cdc42 in the Nef-3-treated cells expressing GFP was higher than that from the IgG-treated cells, indicating that Nef-3 activates FRG (Fig.  6, A and B). This Nef-3-induced activation of FRG was inhibited by PP2, and this inhibitory effect of PP2 was not suppressed by GFP -V12Rap1 (Fig. 6A). GFP-V12Rap1 alone did not affect the GEF activity of FRG. Furthermore the Nef-3induced activation of FRG was inhibited by GFP-Rap1GAP, and this inhibitory effect of Rap1GAP was not suppressed by c- Src-CA (Fig. 6B). These results indicate that the activation of either Rap1 or c-Src alone is insufficient for and the activation of both is necessary for the nectin-induced activation of FRG.
No Requirement of Rap1 for the Nectin-induced, c-Src-mediated Recruitment or Phosphorylation of FRG-We examined whether Rap1 is necessary for the recruitment of FRG to the nectin-based cell-cell adhesion and for the c-Src-mediated tyrosine phosphorylation of FRG. Nectin-1-L cells overexpressing Flag-FRG with GFP-Rap1GAP or GFP-CAAX as a control were cultured with the Nef-3-or IgG-coated beads. The immunofluorescence signals for nectin-1 and Flag-FRG were observed at the contact sites between the Nef-3-coated beads and nectin-1-L cells overexpressing GFP-CAAX but not between the IgGcoated beads and the cells (Fig. 7A). Co-expression of Flag-FRG with GFP-Rap1GAP did not affect the localization of Flag-FRG at the contact sites between the Nef-3-coated beads and nectin-1-L cells (Fig. 7B). Nectin-1-L cells overexpressing Flag-FRG and c-Src with GFP-Rap1GAP, GFP-V12Rap1, or GFP as a control were incubated with clustered Nef-3 or IgG and subjected to the immunoprecipitation assay using the anti-FLAG mAb followed by Western blotting with the anti-phosphoty-rosine mAb. The tyrosine phosphorylation of FRG was induced by Nef-3, which was not affected by Rap1GAP or V12Rap1 (Fig.  7C). These results indicate that Rap1 is not necessary for the recruitment of FRG to the nectin-based cell-cell adhesion or the c-Src-mediated tyrosine phosphorylation of FRG.

Requirement of Crk, C3G, and Rap1 for the Nectin-induced Formation of the E-cadherin-based AJs-We have previously demonstrated that the nectin-induced activation of c-Src and
Cdc42 is involved in the formation of the E-cadherin-based AJs in MDCK cells (23,26). In the last set of experiments, we examined whether Crk, C3G, and Rap1 are indeed involved in the formation of the E-cadherin-based AJs in nectin-1-MDCK cells. When nectin-1-MDCK cells were cultured at 2 M Ca 2ϩ , the immunofluorescence signal for E-cadherin disappeared at any site along the cell-cell adhesion sites as described previously (38, 39) (data not shown). However, when these cells precultured at 2 M Ca 2ϩ were recultured at 2 mM Ca 2ϩ , the signal for E-cadherin was restored at the cell-cell contact sites as described previously (38, 39) (data not shown). When nectin-1-MDCK cells transiently expressing CrkI-W169L, C3G-dCD, or GFP as a control were precultured at 2 M Ca 2ϩ and recul- tured at 2 mM Ca 2ϩ , the signal for E-cadherin was not observed at the cell-cell contact sites in the cells expressing CrkI-W169L or C3G-dCD, whereas it was reconcentrated there in the cells expressing GFP (Fig. 8A). Moreover, when nectin-1-MDCK cells expressing Myc-Rap1GAP precultured at 2 M Ca 2ϩ were recultured at 2 mM Ca 2ϩ , the signal for E-cadherin was not observed at cell-cell contact sites, and this disappearance of the signal for E-cadherin was rescued by co-expression with GFP-V12Cdc42 (Fig. 8B). These results indicate that Crk, C3G, and Rap1 are required for the nectin-induced, Cdc42-enhanced formation of the E-cadherin-based AJs. DISCUSSION Evidence is accumulating that Rap1 is involved in many cell functions including the formation of AJs (11,40), but it remains largely unknown how Rap1 is activated in response to extracellular signals or how Rap1 regulates cell functions. Cdc42 and Rac have also been shown to play important roles in many cell functions including the formation of AJs and TJs (17,19,20,41). It remained unknown how Cdc42 and Rac are activated during the formation of the cell-cell junctions. It has recently been reported that Rac, but not Cdc42, is activated by the action of cadherins (42), and we have found that both Cdc42 and Rac are activated by the action of nectins (23,33,43). We have subsequently shown that nectins first recruit and activate c-Src at the nectin-based cell-cell contact sites and that c-Src activated in this way then phosphorylates and activates FRG, inducing the activation of Cdc42 (23).
Extending these earlier observations, the present study showed that Rap1 is additionally involved in the nectin-induced, c-Src-and FRG-mediated activation of Cdc42 and formation of AJs. Nectins first recruit and activate c-Src at the nectin-based cell-cell contact sites. On one hand, c-Src activated in this way then recruits and phosphorylates FRG there, although it does not activate FRG. On the other hand, c-Src induces the activation of Rap1 through the Crk-C3G complex at the same area. Rap1 activated in this way then induces the activation of tyrosine-phosphorylated FRG, eventually inducing the activation of Cdc42. It may be noted that the c-Src- A and B, inhibition by CrkI-W169L, C3G-dCD, or Rap1GAP of the nectin-induced formation of the E-cadherin-based AJs in MDCK cells. A, nectin-1-MDCK cells overexpressing Flag-CrkI-W169L, C3G-dCD, or GFP as a control were incubated at 2 M Ca 2ϩ for 2 h and then incubated at 2 mM Ca 2ϩ for 2 h. After the incubation, the cells were immunostained with the anti-E-cadherin mAb, anti-FLAG mAb, and anti-C3G pAb. Scale bars, 10 m. Arrowheads, the absence of the signal for E-cadherin at the cell-cell contact sites. B, nectin-1-MDCK cells expressing the combination of Myc-Rap1GAP, GFP-V12Cdc42, GFP, or an empty vector were incubated as described in A. After the incubation, the cells were immunostained with the E-cadherin mAb and anti-Myc mAb. Arrowheads, the absence of the signal for E-cadherin at the cell-cell contact sites; arrows, the restoration of the signal for E-cadherin at the cell-cell contact sites. Bars in the graphs of A and B represent the E-cadherin recruitment index, which is defined as the percentage of cell-cell contacts with the signal for E-cadherin of the total 50 cell-cell contacts counted and are expressed as means Ϯ S.E. of the three independent experiments. mediated activation of Rap1 or the phosphorylation of FRG alone is insufficient for and both are essential for the activation of FRG. We have recently revealed that both c-Src and Cdc42 are necessary for the nectin-induced activation of Rac through Vav2, a Rac-GEF (54). We have also shown here that Rap1 is necessary for the activation of Rac presumably through FRG and Cdc42. Taken together, we would like to propose a model for the nectin-induced activation of Cdc42 and Rac as shown in Fig. 9. Cdc42 is likely to induce formation of filopodia and to increase the number of cell-cell contact sites at the initial stage of the formation of AJs. On the other hand, Rac induces formation of lamellipodia that efficiently expands the cell-cell adhesion between filopodia, acting like a zipper (18 -20, 44 -47). Consistently we demonstrated here that Crk, C3G, and Rap1 are involved in the nectin-induced formation of the E-cadherinbased AJs and that the inhibitory effect of Rap1GAP on the formation of AJs are rescued by a constitutively active mutant of Cdc42. Nectins furthermore recruit first junctional adhesion molecules and then claudins to the apical side of AJs, resulting in the formation of TJs in epithelial cells. Thus, after the formation of AJs, nectins play important roles in both the formation of the cell-cell junctions and the initiation of cell polarization in mammals, and the small G protein pathway of Rap1 controlling Cdc42 is functionally conserved from yeast to mammals and is essential for the formation of cell-cell junctions and cell polarization as described for the budding process in the yeast (15).
Here we have attempted to clarify how Rap1 induces the activation of FRG tyrosine-phosphorylated by c-Src. Several molecules that function downstream of Rap1 have been identified (9,48). These include phosphatidylinositol 3-kinase, afadin, B-Raf, and RalGDS. B-Raf activates the MEK-ERK signaling pathway (7), whereas RalGDS activates Ral as a GEF (49). We have previously shown that wortmannin, a phosphatidylinositol 3-kinase inhibitor, does not inhibit the Nef-3-induced formation of filopodia or lamellipodia in nectin-1-L cells (50) and that afadin is not necessary for the Nef-3-induced activation of Cdc42 or Rac in nectin-1-L cells (33). U0126, an MEK inhibitor, does not inhibit the Nef-3-induced formation of filopodia or lamellipodia in nectin-1-L cells (data not shown). A dominant negative mutant of RalA does not inhibit the Nef-3induced formation of filopodia or lamellipodia in nectin-1-L cells (data not shown). These results indicate that at least any of these known downstream targets of Rap1 is not involved in the Rap1-induced activation of FRG and suggest that an unidentified molecule(s) may relay the signaling from Rap1 to FRG.
On the basis of the observations thus far available, we finally discuss the physiological role of nectins in the formation of cell-cell junctions and cell polarization. Formation of cell-cell junctions and establishment of cell polarity is a multistep process that is regulated by internal cues or extracellular signals (15,51). It requires the selection of a specific site at the cell cortex followed by the recruitment and activation of the components involved in actin or microtubule polymerization at that site. These cytoskeletal rearrangements lead to various forms of polarized growth, including asymmetric cell division and the formation of filopodia, lamellipodia, stress fibers, and epithelial cell polarity. However, it remains unclear how epithelial cell polarity is established. During the establishment of epithelial cell polarity, nectins are likely to form initial cell-cell contacts and recruit E-cadherin to the nectin-based cell-cell contact sites (17). Furthermore, at the cell-cell contact sites formed by the trans-interaction of E-cadherin, the nectins, which do not trans-interact with other nectins (non-trans-interacting nectins), inhibit the E-cadherin-induced activation of Rac and the subsequent formation of AJs until eventually the nontrans-interacting nectins trans-interact with other nectin molecules and induce the activation of Cdc42 (43). Therefore, nectins, but not cadherins, are likely a good candidate for the cue to select initial cell-cell contact sites. The transinteracting nectins at newly formed cell-cell contact sites may organize some signaling microdomains and transduce the positional signal through the c-Src-Crk-C3G-Rap1-FRG-Cdc42-Rac pathway to initiate cell-cell contact, resulting in the formation of cell-cell junctions and establishment of cell polarity in epithelial cells.
Two other groups have recently reported the role of Rap1 in the formation of the E-cadherin-based AJs in epithelial cells (52,53). One has demonstrated that Ras-transformed MDCK cells, which show the mesenchymal phenotype, form the Ecadherin-based AJs when they are transfected with the plasmid of V12Rap1, and that in contrast overexpression of Rap1GAP disrupts the epithelial cell-cell contacts (52). However, it was not studied in this report how Rap1 is activated. The other one has reported that the trans-interaction of Ecadherin induces the activation of Rap1 through C3G, which binds to the cytoplasmic region of E-cadherin, and that the activation of Rap1 is necessary for the maturation of cell-cell contacts in epithelial cells (53). Our present results are consistent with the results of these two reports in the sense that Rap1 is involved in the formation of AJs.