The atypical Rho family GTPase Wrch-1 regulates focal adhesion formation and cell migration.

Wrch-1 (Wnt-regulated Cdc42 homolog) is a new member of the Rho family that was identified as a gene transcriptionally upregulated by Wnt-1. Wrch-1 has no detectable GTPase activity and displays very high intrinsic guanine nucleotide exchange, implying that it is constitutively GTP-bound. The biological functions of Wrch-1 largely remain to be characterized. Here, we report that Wrch-1 prominently localizes to focal adhesions. Depletion of Wrch-1 by small interfering RNA increases focal adhesion formation, whereas Wrch-1 overexpression disassembles focal adhesions. Wrch-1 depletion inhibits myosin-light-chain phosphorylation, which in turn leads to an increase in the number of focal adhesions and inhibits cell migration in response to wound healing. Depletion of Wrch-1 also inhibits Akt and JNK activation. Although pharmacological inhibitors of Akt and JNK inhibit cell migration, they do not affect focal adhesions. Thus, our data suggest that Wrch-1 regulates cell migration by multiple mechanisms: on the one hand Wrch-1 controls focal adhesions by regulating myosin light chain and on the other hand Wrch-1 stimulates the activation of Akt and JNK.


Introduction
Rho family GTPases are crucial regulators of diverse functions, including actin dynamics, cell migration, vesicular trafficking and cell proliferation (Jaffe and Hall, 2005). Wrch-1 (Wntregulated Cdc42 homolog) is a novel member of the Rho family that was originally identified as a gene transcriptionally upregulated by Wnt-1 (Tao et al., 2001).
Several features distinguish Wrch-1 from most other Rho proteins. Wrch-1 has no detectable GTPase activity but its intrinsic guanine nucleotide exchange activity is extremely high, implying that it is constitutively in the GTP-bound, active state Shutes et al., 2004). Wrch-1 displays extensive sequence homology to Cdc42-related proteins and has both N-and C-terminal extensions. The N terminus has three PxxP motifs that can serve as docking sites for SH3domain-containing proteins. The N-terminal domain negatively regulates Wrch-1 function and this internal inhibition can be released by binding of the N-terminus to SH3-domain-containing proteins such as Grb2 (Shutes et al., 2004). Wrch-1 also is unique in its post-translational modifications and mode of membrane anchorage, being palmitoylated rather than prenylated (Berzat et al., 2005).
The biological functions of Wrch-1 largely remain to be characterized. Like Cdc42, activation of Wrch-1 stimulates the formation of filipodia  and, similar to other Rho family proteins including Cdc42, Rac and Rho, overexpression of Wrch-1 promotes cell transformation (Shutes et al., 2004). In this work we employ RNA interference (RNAi) to study Wrch-1 function. We show that small interfering RNA (siRNA)-mediated depletion of Wrch-1 increases focal adhesion and stress fiber formation, and inhibits cell migration. We also examine the molecular mechanisms that underlie the role of Wrch-1 in the regulation of focal adhesions and cell migration. This is the first study that investigates the cellular functions of endogenous Wrch-1.

Wrch-1 regulates focal adhesions
The high similarity between Wrch-1 on the one hand and Cdc42 and Rac proteins on the other suggested that Wrch-1 may also function in the organization of the actin cytoskeleton. To examine this, we first determined the intracellular localization of Wrch-1. Although we have generated a polyclonal antibody that specifically recognizes recombinant Wrch-1, it failed to detect the endogenous protein in a number of cell lines we had examined, probably because of the low expression levels of endogenous Wrch-1. We therefore used a GFP-Wrch-1 fusion protein and expressed it in HeLaS3 cells, which express relatively high levels of endogenous Wrch-1 mRNA (Kirikoshi and Katoh, 2002). Interestingly, in cells that express low levels of GFP-Wrch-1, we observed that this protein is strongly enriched in focal adhesions, as indicated by Wrch-1 (Wnt-regulated Cdc42 homolog) is a new member of the Rho family that was identified as a gene transcriptionally upregulated by Wnt-1. Wrch-1 has no detectable GTPase activity and displays very high intrinsic guanine nucleotide exchange, implying that it is constitutively GTP-bound. The biological functions of Wrch-1 largely remain to be characterized. Here, we report that Wrch-1 prominently localizes to focal adhesions. Depletion of Wrch-1 by small interfering RNA increases focal adhesion formation, whereas Wrch-1 overexpression disassembles focal adhesions. Wrch-1 depletion inhibits myosin-light-chain phosphorylation, which in turn leads to an increase in the number of focal adhesions and inhibits cell migration in response to wound healing. Depletion of Wrch-1 also inhibits Akt and JNK activation. Although pharmacological inhibitors of Akt and JNK inhibit cell migration, they do not affect focal adhesions. Thus, our data suggest that Wrch-1 regulates cell migration by multiple mechanisms: on the one hand Wrch-1 controls focal adhesions by regulating myosin light chain and on the other hand Wrch-1 stimulates the activation of Akt and JNK.
its colocalization with vinculin ( Fig. 1). Punctate staining in the center of the cell can also be discerned, in line with the earlier documented localization of Wrch-1 to an endosomal compartment (Berzat et al., 2005). A Myc-epitope-tagged version of Wrch-1 displays a similar localization pattern compared with that of GFP-Wrch-1 (data not shown). Taken together, these data strongly suggest that Wrch-1 resides at focal adhesions.
The prominent localization of Wrch-1 to focal adhesions suggests that Wrch-1 can regulate the dynamics of these structures. To examine this, we knocked down endogenous Wrch-1 using siRNA and determined the effect of depleting Wrch-1 on focal adhesions. Wrch-1 knockdown modestly, but significantly, increases the number of focal adhesions per cell ( Fig. 2A,D). Wrch-1 knockdown was confirmed by quantitative RT-PCR (Fig. 2B). Similar results were obtained with two independent oligos, indicating that the siRNAs act in a specific fashion to interfere with the generation of Wrch-1 protein.
We also note that depletion of Wrch-1 does not significantly alter the length or brightness of focal adhesions, indicating that the increase in the number of focal adhesions in Wrch-1depleted cells is not compensated for by a decrease in focal adhesion size or density of focal adhesion proteins. Wrch-1 depletion also slightly, but significantly, decreases the steadystate cellular spread area (Fig. 2C), indicating that the increase in the number of focal adhesions per cell is not due to an increase in cell spreading. Remarkably, whereas the focal adhesions in cells transfected with control oligonucleotides directed against luciferase tend to be localized in the periphery of the cell, cells depleted of Wrch-1 display a marked increase in the number of centrally localized focal adhesions. Concomitant with the increase in focal adhesions, Wrch-1depleted cells also show an increase in the formation of stress fibers.
We also examined the effect of overexpressing Wrch-1 on focal adhesions. Cells that express medium to high levels of GFP-Wrch-1 (see Materials and Methods, Immunofluorescence) Journal of Cell Science 120 (11) Fig. 1. Wrch-1 localizes to focal adhesions. HelaS3 cells were plated and transfected with GFP-Wrch-1 as described in Materials and Methods. Sixteen hours after transfection, cells were plated on laminin-coated coverslips for 24 hours, fixed and processed for indirect immunofluorescence using a vinculin antibody and Texas-Red-labeled secondary antibody. Micrographs depict cells with low GFP-Wrch-1 expression level. Punctate staining in the center of the cell that does not correspond to focal adhesions is indicated with an asterisk. Bar, 10 m. Cells were transfected with siRNA targeting luciferase (control) or Wrch-1 (W-1 and W-2) as described in Materials and Methods. Twenty-four hours after transfection, cells were re-plated onto laminin-coated coverslips for another 48 hours. Subsequently, cells were fixed and processed for indirect immunofluorescence using a anti-vinculin antibody and a Texas-Redlabeled secondary antibody, and co-stained for F-actin (FITC-phalloidin). Bar, 10 m. (B) Quantitative PCR analysis of Wrch-1 knockdown. Cells were transfected as described for A. Seventy-two hours after transfection, total RNA was isolated and analyzed for Wrch-1 mRNA levels. Shown is the average of seven independent experiments ± s.e.m. (C) Wrch-1 depletion diminishes cell spreading. Cells were processed as described for A. For each treatment, the average cell area and s.e.m. was determined for 20 to 25 cells using Esee software (Inovision). Data shown are representative of three independent experiments. (D-F) Quantification of focal adhesions (D), average focal adhesion length (E) and focal adhesion brightness (F) was performed as described in Materials and Methods. For each treatment, the average number of focal adhesions per cell, the average focal adhesion length and brightness (± s.e.m.) was determined for eight cells, randomly selected. Data shown are representative of three independent experiments. *P<0.05, two-tailed t-test. display a strong decrease in the number of focal adhesions (Fig.  3A,B) and cell rounding (Fig. 3A,C). Similar findings have been described previously . Most of the staining in these cells accumulates towards the center of the cell and does not colocalize with focal adhesions. We note that completely rounded cells, most of which displayed condensed chromatin, were excluded from the quantitative analysis. Focal adhesions in cells that express high levels of GFP itself are not affected, however (Fig. 3A). Thus, the results from both Wrch-1 depletion and overexpression studies strongly indicate a role of Wrch-1 in the control of focal adhesions.

Depletion of Wrch-1 inhibits cell migration
The importance of Wrch-1 in regulating focal adhesions suggested that Wrch-1 can also regulate cell migration. We therefore examined the role of Wrch-1 in cell migration using the monolayer wound healing assay (Valster et al., 2005). We observed that depletion of endogenous Wrch-1 significantly slows down cell migration (Fig. 4A).
The level of activated Wrch-1 is thought to be primarily controlled at the level of transcription rather than GTP loading, because its intrinsic guanine nucleotide exchange activity is very high Shutes et al., 2004). We therefore examined the level of Wrch-1 mRNA during wound healing. Wrch-1 mRNA levels increase significantly and very rapidly after wounding, reaching half-maximum levels within 5 minutes of wounding (Fig. 4B). The increase in Wrch-1 mRNA is sustained for at least two hours. Although the increase in wounding-induced Wrch-1 mRNA we were able to detect was only approximately twofold, this is likely to be a strong underestimate, because we expect that only a relatively small fraction of the monolayer cells will be close enough to a wound edge to be able to respond to wounding.
Wrch-1 mediates phosphorylation of myosin regulatory light chain Actomyosin contractility plays a crucial role in the regulation of focal adhesion assembly and turnover (Burridge and Chrzanowska-Wodnicka, 1996;Webb et al., 2002). Myosin II activity is controlled by the phosphorylation status of its regulatory light chain (MLC). We therefore examined the effect of depleting Wrch-1 on MLC phosphorylation using an antibody specific for MLC phosphorylated at residues Thr18 and Ser19. Wrch-1 depletion strongly inhibits MLC phosphorylation in HeLaS3 cells growing in the presence of serum (Fig. 5A). Levels of total MLC are not affected by Wrch-1 depletion (Fig. 5A). These results indicate that Wrch-1 is necessary for maintaining myosin II activity. Rho subfamily members play key roles in MLC phosphorylation (Burridge and Wennerberg, 2004;Jaffe and Hall, 2005). Thus, one possible mechanism for the role of Wrch-1 in MLC phosphorylation is that Wrch-1 regulates the activity of Rho proteins. However, we did not find any significant effect on Rho activation upon depleting Wrch-1. Moreover, the ROCK inhibitor Y27632 stimulates -rather than inhibits -cell migration in HeLaS3 cells (data not shown).
The phosphorylation status of MLC is controlled by a number of MLC kinases and the myosin phosphatase (Matsumura, 2005). Of these kinases, myosin light-chain kinase (MLCK) appears to play the key role in regulating focal adhesion turnover, at least in fibroblasts (Totsukawa et al., 2004;Webb et al., 2004). To examine the hypothesis that MLCK mediates the function of Wrch-1 in the regulation of focal adhesions in HeLaS3 cells, we used ML-7, a small molecule that inhibits both MLCK and zipper-interacting protein kinase (ZIPK) (Saitoh et al., 1987;Komatsu and Ikebe, 2004). We found that, similar to depletion of Wrch-1, low concentrations of ML-7 cause an increase in the formation of focal adhesions and stress fibers (Fig. 5B,C), supporting the view that Wrch-1 stimulates the disassembly of focal adhesions by regulating an MLCK-related kinase. Inhibition of MLCK and/or ZIPK by ML-7 also slows down cell migration (Fig. 5D), consistent with the role of focal adhesion site turnover in cell migration (Totsukawa et al., 2004). Importantly, addition of ML-7 to Wrch-1-depleted cells does not further decrease cell migration (Fig. 5D), suggesting that Wrch-1 acts on the same pathway as MLCK and/or ZIPK in the regulation of cell migration.
The Wrch-1 effector PAK regulates focal adhesions and cell migration Currently, the only known effector of Wrch-1 is the serinethreonine kinase PAK (p21-activated protein kinase) (Tao et al., 2001;Saras et al., 2004;Aspenstrom et al., 2004). We therefore also examined the effect depletion of PAK kinases has on focal adhesions and cell migration. Combined depletion of PAK1 and PAK2 significantly increases the number of focal adhesions per cell (Fig. 6A,B). Depletion of PAK1/2 also inhibits MLC phosphorylation and cell migration (Fig. 6C,D). Moreover, addition of ML-7 to PAK1/2-depleted cells does not further decrease cell migration (Fig. 6D), similar to the lack of any effect of ML-7 on Wrch-1-depleted cells (Fig. 5D). Thus, these results support the notion that Wrch-1 acts upstream of PAK1 and MLC to regulate focal adhesions and cell migration.

Wrch-1 regulates JNK activation
Overexpression of Wrch-1 has been shown to activate Jun Nterminal kinase (JNK) (Tao et al., 2001). Furthermore, JNK phosphorylates paxillin, a protein that is associated with focal adhesions, and this phosphorylation event is thought to contribute to focal adhesion turnover and cell migration (Huang et al., 2003). To examine whether Wrch-1 regulates focal adhesions and cell migration via JNK, we first determined the effect of Wrch-1 depletion on JNK activation induced during the wound healing response of HeLaS3 cell cultures. As observed for airway epithelial cells (White et al., 2005), JNK is activated within 30 minutes after wound healing, and depletion of Wrch-1 inhibits JNK activation to below baseline levels (Fig. 7A). Interestingly, although inhibition of JNK activity using the specific small molecule inhibitor SP600125 does not alter the number of focal adhesions (Fig.  7C), it does significantly inhibit cell migration in control cells, but not in Wrch-1-depleted cells (Fig. 7B). These data suggest that Wrch-1-mediated JNK activation plays a role in cell migration, and that this role is independent from the role of Wrch-1 in focal adhesions.
Wrch-1 binds to the regulatory p85 subunit of PI3K and regulates Akt We also examined the effect of Wrch-1 depletion on additional signaling elements that have been implicated in cell migration. Whereas depletion of Wrch-1 does not have any significant effect on ERK phosphorylation in wounded monolayers (data not shown), it significantly inhibits wound-induced Akt phosphorylation (Fig. 8A). Akt has been shown to play a crucial role in HeLaS3 cell migration (Li et al., 2005). We Journal of Cell Science 120 (11) Cells were transfected with siRNA targeting luciferase (control) or Wrch-1 (W-1) as described in Materials and Methods. Ninety hours after transfection, cells were lysed directly into 2ϫ sample buffer and phosphorylated MLC was analyzed by western blotting. Subsequently, blots were stripped and re-blotted with anti-MLC antibody. Tubulin blot was used as loading control. Levels of phosphorylated MLC from seven independent experiments were quantified as detailed in Materials and Methods and normalized to levels of either MLC or tubulin. *P=10 -4 , two-tailed t-test. (B) treatment with ML-7 increases stress fiber formation and the number of focal adhesions. Cells were plated onto laminin-coated coverslips for 1 hour and subsequently incubated in the presence of 5 M ML-7, or DMSO as control. After 48 hours, cells were fixed and processed as described in Fig. 2A. Bars, 10 m. (C) Quantification of focal adhesions was performed as in Fig. 2D. Data shown are representative of two independent experiments. *P<0.0005, two-tailed t-test. (D) ML-7 inhibits cell migration. HelaS3 cells were transfected with siRNA targeting luciferase (control) or Wrch-1 (W-1) for 72 hours, plated as a confluent monolayer for several hours and subsequently pretreated overnight with either ML-7 (5 M) or DMSO, and treated with ML-7 or DMSO upon wounding. Migration was quantified 7 hours after wounding, as described in Fig. 4A. Shown is the average ± s.e.m. of eight sites. *P<10 -5 , two-tailed t-test. Data shown are representative of four independent experiments. confirmed this in HeLaS3 cells using an inhibitor of phosphatidylinositol 3-kinase (PI3K), a signaling element that acts directly upstream of Akt (Fig. 8B). Notably, the PI3K inhibitor LY294002 does not further decrease the migration of Wrch-1-depleted cells, suggesting that Wrch-1 acts upstream of PI3K, and probably Akt, in the regulation of cell migration. LY294002 does not significantly affect the organization of the actin cytoskeleton, however, and does not alter the number of focal adhesions (supplementary material Fig. S1), indicating that the roles of PI3K and Akt in cell migration do not involve the regulation of focal adhesions.
Wrch-1 displays a high degree of sequence homology with Cdc42 and Rac (Tao et al., 2001), both of which can activate PI3K by binding to its p85 regulatory subunit (Zheng et al., 1994;Tolias et al., 1995). A GSTfusion protein of the N-terminal domain (a.a. 1-333) of p85 (Hill et al., 2001) efficiently pulls down full-length Wrch-1 from cell lysates (Fig. 8C). p85N also interacts with a Wrch-1 deletion mutant that misses the N-terminal proline-rich domain, but in a less efficient manner. These observations suggest that Wrch-1 interacts with p85 both via its effector domainpresumably with the Bcr domain of p85 (Zheng et al., 1994) -and its proline-rich domain (probably with the SH3 domain of p85).

Discussion
Here, we used RNAi to examine the role of Wrch-1 in the organization of the actin cytoskeleton and cell migration. Our data suggest that Wrch-1 regulates cell migration by multiple mechanisms, by stimulating Akt and JNK activation on the one hand and, on the other hand, by stimulating focal adhesion disassembly in a manner that is independent of Akt and JNK. This is the first study that analyzes the cellular functions of endogenous Wrch-1.
Wrch-1 is unique among the Rho family of GTPases in several respects, including its very high intrinsic guanine nucleotide exchange activity and its post-translational modification by a palmitoyl group Shutes et al., 2004;Berzat et al., 2005). Adding to these unique features, we found that Wrch-1 prominently localizes to focal adhesions, although vesicular perinuclear staining is also  visible, especially at higher levels of expression. Localization of Wrch-1 to focal adhesions is probably mediated by palmitoylation (Berzat et al., 2005), although binding to focaladhesion-localized binding partners may also contribute. In line with the localization of Wrch-1 to focal adhesions, we also found that Wrch-1 plays a crucial role in regulating these structures. We showed that knockdown of endogenous Wrch-1 significantly increases the number of focal adhesions per cell, even though the spread area of the cell is diminished. Conversely, overexpression of Wrch-1 causes focal adhesion disassembly, as has been observed previously . A possible mechanism for the increase in the number of focal adhesions in Wrch-1-depleted cells is that Wrch-1 stimulates focal adhesion turnover. Interestingly, both depletion and overexpression of Wrch-1 inhibit cell spreading. Although it is possible that cell rounding caused by high levels of ectopic Wrch-1 is an artefact of overexpression, these results would suggest that optimal cell spreading, at least in HeLaS3 cells, depends on a set level of Wrch-1 expression, through yetto-be-determined mechanisms.
We also started to investigate the molecular mechanisms that mediate the role of Wrch-1 in the regulation of focal adhesions. We found that the effects of depleting Wrch-1 on focal adhesions and cell migration are mimicked by the concomitant depletion of PAK1 and PAK2, the only effectors of Wrch-1 that have been documented thus far (Tao et al., 2001;Saras et al., 2004;Aspenstrom et al., 2004). Moreover, we showed that, like depletion of Wrch-1, depletion of PAK1/2 also inhibits the phosphorylation of MLC, suggesting that PAK1/2 acts downstream of Wrch-1 to regulate MLC, and subsequently focal adhesions and cell migration. Our results that PAKs are necessary for MLC phosphorylation are consistent with previous findings, describing roles for PAKs in cell migration and endothelial permeability (Sells et al., 1999;Stockton et al., 2004), and that PAK2 can directly phosphorylate MLC (Chew et al., 1998).
We also found that ML-7, a small molecule inhibitor of both MLCK and ZIPK, causes an increase in the total number of focal adhesions per cell, a phenotype that is similar to the one caused by Wrch-1 depletion. Moreover, both Wrch-1 depletion and ML-7 markedly increase focal adhesion formation in the center of the cell. MLCK is thought to preferentially act on myosin in the cell periphery, whereas ZIPK appears to preferentially act in the cell center, at least in fibroblasts (Totsukawa et al., 2004;Komatsu and Ikebe, 2004). These considerations would suggest that Wrch-1 regulates focal adhesions via ZIPK or a related kinase, rather than via MLCK.
Since Wrch-1 can activate JNK (Tao et al., 2001), we also examined the possibility that Wrch-1-mediated JNK activation contributes to the regulation of focal adhesions. Indeed, both JNK itself and JSAP1/JIP3, a scaffold protein that coordinates JNK activation, localize to focal adhesions (Almeida et al., 2000;Takino et al., 2005). Furthermore, JNK phosphorylates paxillin, and overexpression of a paxillin mutant that is not phosphorylatable by JNK enhances the formation of focal adhesions and slows down cell migration (Huang et al., 2003). We found that, although Wrch-1 depletion inhibits JNK activation induced by wound healing, chemical inhibition of JNK using SP600125 does not cause accumulation of focal adhesions in HeLaS3 cells.
SP600125 strongly inhibits the migration of control HeLaS3 cells but not of Wrch-1-depleted cells, suggesting that Wrch-1-mediated JNK activation does play a role in the migratory behavior of these cells, independently from its function in the regulation of focal adhesions. JNK activation is crucial for migration in a large number of different cell systems (Huang et al., 2004;Xia and Karin, 2004). In addition to paxillin, relevant JNK substrates include the actin-nucleating protein Spir and a number of microtubule-associated proteins (Huang et al., 2004;Xia and Karin, 2004). Another likely mechanism for the role of JNK in cell migration is activation of the AP-1 transcription factor (Malliri et al., 1998;Huang et al., 2004;Xia and Karin, 2004).
We also found that depletion of Wrch-1 significantly inhibits wounding-induced Akt phosphorylation, and that inhibition of Akt by blocking its upstream activator PI3K suppresses the Journal of Cell Science 120 (11) Fig. 8. Wrch-1 binds to the regulatory p85 subunit of PI3K and is necessary for Akt activation. (A) Wrch-1 depletion inhibits wound-induced Akt phosphorylation. Lysates were prepared as in Fig. 7A. Akt phosphorylation was visualized using an against antibody phosphorylated Akt (Ser473). An antibody against dynamin 2 was used to demonstrate equal loading. Levels of phosphorylated Akt from three independent experiments were quantified and normalized to levels of dynamin. *P<0.001, two-tailed t-test. (B) PI3K is necessary for cell migration. PI3K activity was inhibited by LY294002 (4 M) (with overnight pre-treatment). Other conditions are as described in Fig. 5D. *P<0.0005 and **P<5ϫ10 -6 , two-tailed t-test. Data are representative of two independent experiments. (C) Wrch-1 binds to p85. Wrch-1 proteins were expressed in HelaS3 cells and pulled down using GST-p85 beads, as described in Materials and Methods. migration of control HeLaS3 cells but not of Wrch-1-depleted cells. Furthermore, we showed that Wrch-1 directly interacts with the p85 regulatory subunit of PI3K. Together, these observations strongly suggest that Wrch-1 acts upstream of PI3K and Akt in the regulation of cell migration. Moreover, the PI3K-Akt branch of Wrch-1 signaling does not appear to be involved in regulating focal adhesions. Since Akt-1 and Akt-2 have been shown to play negative and positive roles, respectively, in cell migration (Yoeli-Lerner et al., 2005;Irie et al., 2005), it will be of great interest to examine whether Wrch-1 selectively mediates the activation of Akt2 versus Akt1.
Overexpression of activated versions of Wrch-1, generated either by a point mutation that interferes with GTP hydrolysis, by truncation of the N-terminal domain or by both, has been shown to transform NIH 3T3 cells using anchorageindependent growth as a read-out (Shutes et al., 2004). We therefore also examined the potential role of Wrch-1 in cell proliferation. However, we did not find any effect on the proliferation of HeLaS3 cells in limiting serum concentrations upon depletion of Wrch-1 (supplementary material Fig. S2), arguing for a specific role for Wrch-1 in the migratory behavior of these cells.
Several Wnt family members, including Wnt1, Wnt3a, Wnt4 and Wnt11 have been shown to stimulate cell migration (De Calisto et al., 2005;Endo et al., 2005;Qiang et al., 2005;Kuphal et al., 2006), suggesting that induction of Wrch-1 contributes to Wnt-regulated cell migration. However, whether Wnt signaling is involved in wound healing in HeLaS3 monolayers remains to be determined. Notably, we did indeed observe that Wrch-1 message is significantly induced immediately after wounding, although the very fast kinetics of this increase appears to argue against the involvement of Wnt release and subsequent Wnt-mediated induction of Wrch-1. Thus, it is possible, if not likely, that Wrch-1 message levels can also be controlled in a Wnt-independent fashion.
In conclusion, we describe a new function of Wrch-1 in cell migration that is likely to be mediated by the concerted action of multiple Wrch-1-regulated signaling mechanisms respectively controlling focal adhesions, and the activation of Akt and JNK.

Cell culture and transfection
HelaS3 cells were grown at 37°C in Dulbecco's modified Eagle's medium (DMEM) with L-glutamine, supplemented with 5% fetal bovine serum, 5% bovine growth serum (Hyclone) and penicillin/streptomycin. Transient transfection of siRNA was carried out using Lipofectamine 2000 (Invitrogen). Cells were plated in a 24-well plate. As soon as cells were fully spread, medium was changed to 500 l of DMEM without antibiotics. Twenty nM of siRNA were transfected using the protocol provided by manufacturer. Wrch-1 levels were determined by quantitative PCR on the day of the assay. Transient transfection of plasmids was carried out using Effectene (Qiagen). Cells were plated at 30% confluency in a 24-well plate with growth medium. Once fully spread (approximately 4 hours after plating), cells were transfected following the protocol supplied by the manufacturer (0.4 g of plasmid, 3.2 l of enhancer and 4 l of Effectene).

GST pull-down assays
Plasmids were transfected into HelaS3 cells with Effectene (Qiagen) as described above for 16 hours. Cells were lysed in RIPA buffer with Protease Inhibitor Cocktail without EDTA (Roche) and total protein concentration was determinated with Protein Assay reagent (Bio-Rad). Equal amounts of purified GST-p85N (Hill et al., 2001) were incubated with equal concentrations of cell lysates containing either ⌬NWrch-1Q107L or Wrch-1Q107L at 4°C for 16 hours. 50 l of glutathione beads (GE) were added at 4°C for another 2 hours. Subsequently, the beads were washed five times and eluted with 80 l of 2ϫ sample buffer. Samples were run in either 10% LongLife gels (Life Therapeutics) or 4-20% Precise gels (Pierce) and transferred to PVDF membrane (Bio-Rad). Proteins were detected with antibodies against HA (Wrch-1) or GST (GST-p85N).

Wound healing assay
Wound healing assays were performed essentially as described in (Valster et al., 2005). In brief, a confluent monolayer of HelaS3 cells was wounded with a P1000 tip in 1ϫ PBS without Ca 2+ and Mg 2+ and subsequently returned to full growth medium. Wounds were imaged at various times at the same spots using an IX70 Olympus inverted microscope equipped with a 10ϫ, NA=0.3 objective, an Orca II cooled charge coupled device (CCD) camera (Hamamatsu, Bridgewater, NJ) and Esee image analysis software (Inovision, Raleigh, NC). Wound widths were measured for at least eight spots. Distance of cell migration was calculated as half of the wound width at a given time point and by subtracting this value from the initial half-width.

Immunofluorescence and quantification of focal adhesions
Cells were plated sparsely on laminin-coated coverslips (BD Falcon) for 24 or 48 hours. Subsequently, cells were fixed in 4% formaldehyde/PBS, permeabilized with 0.1% Triton X-100 dissolved in PBS, followed by staining with FITC-conjugated phalloidin and indirect immunofluorescence using an anti-vinculin antibody. Images were collected with the set-up described under Wound healing assay, using a 60ϫ, NA=1.4 objective.
Focal adhesions were manually traced using Esee image analysis software. The average length of focal adhesion was calculated by dividing the total length of all focal adhesions in a given cell by the number of focal adhesions in that cell. The average focal adhesion brightness was calculated by first determining the total focal adhesion brightness (approximated by the integrated pixel intensity along all focal adhesions), which was then normalized against the brightness of the whole cell (determined at a plane of focus 8 m higher than the ventral plane). The average focal adhesion brightness was then calculated as the total focal adhesion brightness divided by the number of focal adhesions.
To determine the effect of overexpressing Wrch-1 on focal adhesions, we chose cells that expressed medium to high levels of GFP-Wrch-1, including cells that displayed GFP-Wrch-1 at levels visible through the microscope ocular, in contrast to low expression levels that were only detectable by the CCD camera.

Multiple wounding assay
This method was modified from a previously described protocol (White et al., 2005). Confluent monolayers in 60-mm dishes were wounded in four directions using a metal nit comb (Bayer), with a tip width of 0.35-0.40 mm and tip to tip distance of 0.6-0.7 mm. Monolayers were wounded in PBS without Ca 2+ and Mg 2+ . Wounded cells were returned to growth medium at 37°C. JNK activation was determined using a kit and accompanying protocol from Cell Signaling. Cell lysates were collected in 1ϫ cell lysis buffer plus Protease Inhibitor Cocktail without EDTA (Roche), followed by sonication and clarification. For the JNK activation assays, equal amounts of protein were immunoprecipitated with glutathione-Sepharose beads (Amersham) loaded with 2 g of GST-Jun protein. Beads were washed twice with cell lysis buffer and twice with kinase buffer (25 mM Tris-HCl, pH 7.5, 5 mM ␤glycerophosphate, 2 mM dithiothreitol, 0.1 mM Na 3 VO 4 and 10 mM MgCl 2 ). JNK kinase assay was performed in 50 l of kinase buffer supplemented with 200 M ATP for 30 minutes at 30°C. Fifty l of 2ϫ SDS sample buffer was added to terminate reactions after spinning down beads and removal of kinase buffer. Samples were run in 10% LongLife Gels (Life Therapeutics), transferred onto PVDF membrane (Bio-Rad) and detected with anti-phosphorylated Jun(Ser63). For the detection of activated Akt, 5ϫ sample buffer was added to the cell lysates and phosphorylated Akt was detected by western blotting. For the Wrch-1 mRNA induction experiments, at various times after wounding, RNA was extracted using RNeasy (Qiagen) and Wrch-1 mRNA levels were determined by quantitative PCR.

Quantification of western blots
Western blots were scanned using ScanMaker i900 (Microtek) and analyzed using ImageJ software (NIH).

Cell proliferation assay
Cell proliferation was determined using the sulforhodamine B colorimetric assay (Skehan et al., 1990). 4ϫ10 3 siRNA-transfected cells were plated into six wells of a 96-well plate in growth medium supplemented with 1% serum. Plates were processed as previously described .