catenin activation in colon cancer cells - Nature

Oncogene (2012) 31, 1001–1012

& 2012 Macmillan Publishers Limited All rights reserved 0950-9232/12 www.nature.com/onc

ORIGINAL ARTICLE

A Rac1/PAK1 cascade controls b-catenin activation in colon cancer cells G Zhu1, Y Wang1, B Huang1, J Liang1, Y Ding1, A Xu2 and W Wu1 1

Protein Science Laboratory of the Ministry of Education, School of Life Sciences, Tsinghua University, Beijing, China and Beijing Friendship Hospital, Beijing, China

2

P21-activated kinase 1 (PAK1) is associated with colon cancer progression and metastasis, whereas the molecular mechanism remains elusive. Here, we show that downregulation of PAK1 in colon cancer cells reduces total b-catenin level, as well as cell proliferation. Mechanistically, PAK1 directly phosphorylates b-catenin proteins at Ser675 site and this leads to more stable and transcriptional active b-catenin. Corroborating these results, PAK1 is required for full Wnt signaling, and superactivation of b-catenin is achieved by simultaneous knockdown of adenomatous polyposis coli protein and activation of PAK1. Moreover, we show that Rac1 functions upstream of PAK1 in colon cancer cells and contributes to b-catenin phosphorylation and accumulation. We conclude that a Rac1/PAK1 cascade controls b-catenin S675 phosphorylation and full activation in colon cancer cells. Supporting this conclusion, overexpression of PAK1 is observed in 70% of colon cancer samples and is correlated with massive b-catenin accumulation. Oncogene (2012) 31, 1001–1012; doi:10.1038/onc.2011.294; published online 8 August 2011 Keywords: colon phosphorylation

cancer;

b-catenin;

PAK1;

Rac1;

Introduction Wnt/b-catenin signaling controls the fundamental cellular processes, including proliferation, during tissue homeostasis and its aberrant activation is implicated with a wide range of human cancers (Polakis, 2000; Logan and Nusse, 2004; Moon et al., 2004; Clevers, 2006; MacDonald et al., 2009). b-catenin is the major cellular effector of the Wnt signaling and is normally captured by the destroy complex made of axin, APC (adenomatous polyposis coli protein), glycogen synthase kinase-3b and casein kinase 1a and promoted for proteasome degradation after phosphorylation and ubiquitination (Polakis, 2002; MacDonald et al., 2009). In this complex, axin serves as a scaffold facilitating multiple protein–protein interactions. Casein kinase 1a and Correspondence: Professor W Wu, Protein Science Laboratory of the Ministry of Education, School of Life Sciences, Tsinghua University, Qinghuayuan 1, Beijing 100084, China. E-mail: [email protected] Received 12 March 2011; revised 23 May 2011; accepted 7 June 2011; published online 8 August 2011

glycogen synthase kinase-3b sequentially phosphorylate N-terminal residuals of b-catenin and this creates the recognizing motif for the E3-ligase b-Trcp. APC is also crucial for capture of b-catenin in the degradation complex and it binds b-catenin directly via its centrally localized armadillo repeats (Xing et al., 2004). Upon Wnt activation, accumulated b-catenin enters the nucleus and binds to the TCF/LEF family transcriptional factors and induces the expression of its target genes (Moon et al., 2004; Mosimann et al., 2009). The key event of both Wnt signaling transduction and cancerous cell proliferation is the regulation of b-catenin stability and activity. In addition to its N-terminal residues, Ser552 (Fang et al., 2007) and Ser675 (Hino et al., 2005; Taurin et al., 2006) of b-catenin have also been identified as potential phosphorylation sites and may affect its signaling activity or protein stability. Colorectal cancer is the most extensively investigated Wnt-related cancer, as mutations on APC or b-catenin account for the vast majority of cancer samples (Ilyas et al., 1997; Giles et al., 2003). In either case, elevated b-catenin was observed and demonstrated as being responsible for cancer progression (Bienz and Clevers, 2000; Valentini et al., 2003; Phelps et al., 2009). However, it was argued recently that APC mutation alone probably is not sufficient to promote full b-catenin signaling (Anderson et al., 2002; Phelps et al., 2009). Mutation of K-Ras was often observed in more advanced colon cancer samples and it was suggested that a K-Ras/Rac1/ JNK2 cascade may further enhance b-catenin nuclear accumulation by JNK2-mediated phosphorylation of b-catenin at Ser191 and Ser605 (Wu et al., 2008; Phelps et al., 2009). However, Rac1 was also found being able to elevate cytosolic, as well as nuclear levels of b-catenin (Esufali and Bapat, 2004). We thought to investigate whether another downstream effector of Rac1, P21activated kinase (PAK), was involved. In supporting this hypothesis, both Rac1 and PAK1 were found overactivated in colon cancer samples (Esufali and Bapat, 2004; Kumar et al., 2006; Dummler et al., 2009). PAKs are serine and threonine protein kinases, which are important for cell motility as well as survival (Bokoch, 2003; Arias-Romero and Chernoff, 2008). Group I PAKs (PAK1, PAK2 and PAK3) were initially identified as effectors of the small GTPases Cdc42 and Rac1 (Manser et al., 1994), but could also be activated by diverse pathways (Bokoch, 2003; Kumar et al., 2006), whereas Group II PAKs (PAK4, PAK5 and PAK6) were identified more recently with their regulation

Rac1/PAK1 and b-catenin in colon cancer G Zhu et al

1002

less understood (Wells and Jones, 2010). PAKs have oncogenic functions in a broad range of cancers and their over-activation have been well documented (Kumar et al., 2006; Dummler et al., 2009; Molli et al., 2009). For example, PAK1 phosphorylates Merlin at Ser518 and inhibits its activity in neurofibromatosis (Xiao et al., 2002; Rong et al., 2004). PAK1 also phosphorylates estrogen receptor a at Ser305 to stimulate target genes transcription and mammary gland hyperplasia (Wang et al., 2002). In colon cancer tissues, elevated PAK1 expression was correlated with cancer progression and lymph node metastases (Carter et al., 2004). However, little was known about the molecular mechanism by which PAK1 promotes colon cancer progression. There were evidences showing that PAK1 was associated with b-catenin in gastric epithelial cell line (He et al., 2008) and that PAK1 downregulation blocked insulin-stimulated nuclear b-catenin accumulation (Sun et al., 2009). These results promoted us to address whether PAK1 drives colon cancer progression via Wnt/b-catenin signaling.

Results PAK1 is required for proliferation of colon cancer cells In order to clarify the role of PAK1 in colorectal cancer cells, we applied several strategies to inhibit endogenous PAK1 activity in SW480 or HCT116 cells and then measured cell proliferation using colony formation assay and flow cytometry. First, stable expression of a dominant negative (DN) PAK1 (GFP–PAK1–K299A) in SW480 cells significantly inhibited colony formation and caused moderate G1 phase arrest (Figures 1a–c). Second, knocking down of PAK1 in SW480 cells by expressing short hairpin RNA (shRNA) resulted in similar effects (Figures 1d–f). Third, IPA-3, a recently identified allosteric PAK inhibitor (Deacon et al., 2008), was able to dramatically inhibit colony formation of HCT116 and SW480 cells in a concentration-dependent manner (Figures 1g–i). Taken together, these results indicate that downregulation of PAK1 protein or its activity inhibits cell proliferation, suggesting that PAK1 has a pivotal role in promoting proliferation of colon cancer cells.

Figure 1 PAK1 promotes cell proliferation in colon cancer cells. (a) Western blot of SW480 cells stably expressing pEGFP-C1 or GFP-PAK1 K299A. (b) SW480 cells stably expressing GFP or GFP–PAK1 K299A were subjected to soft agar assay. The graph illustrates the number of colonies in one well (mean±s.e.m.; n ¼ 3, *Po0.05). (c) Cell cycle analysis of SW480 cells stably expressing GFP or GFP–PAK1 K299A. (d) Western blots showing that the endogenous PAK1 protein was downregulated by PAK1 shRNA. (e) As in (b), SW480 cells stably expressing control or PAK1 shRNA were subjected to soft agar assay (mean±s.e.m.; n ¼ 3, **Po0.001). (f) As in (c), cell cycle analysis of SW480 cells stably expressing control shRNA or PAK1 shRNA. (g) Colony formation of SW480 and HCT116 cells with or without indicated concentrations of IPA-3. (h, i) The colony numbers of the experiment shown in (g). Mean±s.e.m.; n ¼ 3, *Po0.05, **Po0.001. Oncogene


1003

PAK1 controls b-catenin level in colon cancer cells Ectopically accumulated b-catenin is considered as the major driving force for colon cancer cell proliferation and complete blockage of b-catenin/TCF signaling has been reported to arrest cell proliferation and induce cell differentiation (van de Wetering et al., 2002; Moon et al., 2004). Consistently, knocking-down of b-catenin using shRNA in SW480 cells caused G1 phase arrest and inhibition of colony formation, accompanied with a significant reduction of cyclin D1 level (Supplementary Figures 1A–E). The fact that knockdown of PAK1 or b-catenin caused similar effects suggested that PAK1 might be promoting cell proliferation, at least, in part, through b-catenin signaling. To test this possibility, we stained total b-catenin in SW480 cells transiently transfected with GFP-tagged PAK1–K299A, a DN-PAK1, and observed drastic reduction of b-catenin signal (Supplementary Figure 2A). To ascertain this result, we monitored b-catenin levels in stable SW480 cell lines expressing PAK1 K299A or PAK1 shRNA.

Consistently, b-catenin staining in these cells was significantly weaker (Figures 2a and c) and total b-catenin protein level was markedly reduced, together with lower cyclin D1 (Figures 2b and d). The reduction of b-catenin protein was transcription independent, as its mRNA level was unchanged in DN-PAK1 expressing cells (Supplementary Figure 2B). Next, we examined whether IPA-3 was able to reduce b-catenin level in colon cancer cells. IPA-3 of 30 mM was added into the SW480 culture and the cells were harvested at different time points. Upon IPA-3 treatment, the b-catenin level became lower in 6 h and cyclin D1 was also gradually downregulated (Figure 2e). In another experiment, SW480 cells were treated with increasing concentrations of IPA-3 for 12 h. As shown in Figure 2f, the levels of total b-catenin, as well as cyclin D1, were significantly reduced by 20 mM or higher concentrations of IPA-3. An inactive form of IPA-3, PIR3.5, was applied to SW480 culture for 12 h and no b-catenin reduction was observed (Supplementary Figure 2C).

Figure 2 PAK1 is required for b-catenin accumulation in colon cancer cells. (a) Immunofluorence confocal images of SW480 cells stably expressing GFP or GFP–PAK1 K299A. b-catenin and cell nuclei were exhibited as red and blue (DAPI), respectively. (b) Western blots showing that the levels of b-catenin and cyclin D1 were reduced in GFP–PAK1 K299A-expressing SW480 cells (GFP–PAK1DN). (c) Immunofluorence confocal images of SW480 cells stably expressing Control shRNA or PAK1 shRNA. b-catenin and cell nuclei were exhibited as red and blue (DAPI), respectively. (d) Western blots showing that the levels of b-catenin, PAK1 and cyclin D1 were reduced in PAK1 shRNA-expressing SW480 cells. (e, f) Western blots showing that the levels of b-catenin and cyclin D1 were gradually reduced when SW480 cells were treated with IPA-3, either for increasing time periods with a fixed concentration (e) or with increasing concentrations in a fixed time period (f). Oncogene


1004

Immunostainning on IPA-3-treated SW480 cells also demonstrated dramatic reduction of b-catenin (Supplementary Figure 2D). All together, these results indicate that endogenous PAK1 is required to maintain high level of b-catenin in these colon cancer cells and suggest that PAK1 promotes cell proliferation likely via b-catenin. PAK1 interacts with and phosphorylates b-catenin at S675 To address the mechanism by which PAK1 regulates the protein level of b-catenin, we first performed coimmunoprecipitation experiments and found that endogenous b-catenin was specifically co-precipitated with PAK1 from SW480 cells (Figure 3a). Co-immunoprecipitation after cell fractionation further demonstrated that the interaction between PAK1 and b-catenin mainly occurred in the cytoplasm (Supplementary Figure 2E). Next, we scanned b-catenin protein sequences from several species and identified two conserved serine sites (S552 and S675) as potential PAK1 targets according to the optimal recognizing sequence of PAK1 (RRRRRSWYFS) (Rennefahrt et al., 2007; AriasRomero and Chernoff, 2008). We then co-transfected Flag-b-catenin together with various PAK1 or Rac1 constructs and verified whether any of them was capable of promoting b-catenin phosphorylation, using commercially available antibodies against Phos-S552 or Phos-S675. Rac1 was included, as it functions upstream of PAK1. Forskolin, a potent activator of PKA, was used as a positive control (Hino et al., 2005; Fang et al., 2007). As expected, both S552 and S675 were heavily phosphorylated upon forskolin treatment (Figure 3b). To a similar extent, wild type (WT) and constitutive active (CA) forms of PAK1 and Rac1 were all able to induce b-catenin phosphorylation specifically on S675, but not S552, whereas DN forms were not (Figure 3b). S675 phosphorylation by PAK1 or Rac1 was specific, as PAK2 was non-effective (Supplementary Figure 2F). The CA PAK1 and Rac1 also enhanced S675 phosphorylation on b-catenin S37A, a stabilized mutant, and again, S552 was not affected (Figure 3c). To determine whether PAK1 could directly phosphorylate b-catenin, we performed in vitro kinase assay in which immunoprecipitated PAK1 proteins from transfected HEK293T cells were incubated with GST-b-catenin purified from E. coli. S675 was highly phosphorylated by CA-PAK1, weakly by WT-PAK1, but not by DN-PAK1, whereas S552 was not phosphorylated at all (Figure 3d). Hence, PAK1 is able to specifically phosphorylate Ser675 of b-catenin. We next investigated whether PAK1 activity contributes to S675 phosphorylation in colon cancer cells. As shown in Figures 3e and f, in SW480 cells expressing DN-PAK1 or shRNA against PAK1, the level of S675-phosphorylated b-catenin was drastically diminished, accompanied with reduced total b-catenin level. Importantly, the relative ratio of Phos-S675 against total b-catenin was clearly reduced, indicating that PAK1 is required for S675 phosphorylation in colon cancer cells. By contraries, CA PAK1–T423E, when Oncogene

transfected into SW480 cells, enhanced S675 phosphorylation as well as total b-catenin level, accompanied with a higher cyclin D1 expression (Figure 3g). Immunofluorescence assay demonstrated that the staining of phos-S675 b-catenin was very strong in SW480 cells (Supplementary Figure 2G), indicating that S675 of b-catenin had been heavily phosphorylated. Over all, these results indicate that PAK1 contributes to b-catenin S675 phosphorylation in colon cancer cells. Our gain- and loss-of-function results all indicated that PAK1 activity controls cyclin D1 level, consistent with previous reports (Balasenthil et al., 2004). Next, we asked whether it controls cyclin D1 expression through b-catenin signaling. A cyclin D1 promoter-driven luciferase reporter was used to monitor the regulation of cyclin D1 expression in HCT116 cells. As shown in Figure 3h, CA-PAK1 was able to activate cyclin D1 promoter and this activity relied on endogenous b-catenin signaling, as it was blocked by DN-TCF4. Taken together, these results suggest that PAK1 enhances b-catenin level/activity through S675 phosphorylation in colon cancer cells. S675-phosphorylated b-catenin is more stable and transcriptionally more active To verify the functional relevance of S675 phosphorylation, we created S675A (serine to alanine) and S675D (serine to aspartic acid) mutants, which mimicked unphosphorylated and phosphorylated forms, respectively. Immunofluorescence assay indicated that the mutant b-catenin proteins retained the same cellular localization as the WT (Supplementary Figure 3A), excluding the possibility that phosphorylation of S675 might alter its subcellular localization. Protein stability was directly tested in cycloheximide-treated cells and the results indicated that S675D is indeed more stable than WT or S675A mutant (Supplementary Figure 3B). Consistently, ubiquitination of S675D mutant was clearly much weaker than that of WT- and S675Ab-catenin (Supplementary Figure 3C). In order to confirm that PAK1 regulate b-catenin level via S675phosphorylation-mediated stabilization, we tested whether PAK1 is capable of modulating b-catenin ubiquitination. As shown in Figure 4a, the basal ubiquitination of b-catenin in the presence of MG132, a proteasome inhibitor, was significantly reduced by co-transfected WT- and CA-PAK1, but enhanced by DN-PAK1. As certain portions of b-catenin were phosphorylated at S675 in SW480 cells (for example, Figures 3e–g), we examined whether these b-catenin proteins are free from ubiquitination and degradation. Total or S675-phosphorylated b-catenin was immunoprecipitated from SW480 cells using specific antibodies and normalized to equal amounts of total b-catenin. As shown in Figure 4b, the phos-S675 b-catenin was almost free from ubiquitination and this result provided compelling evidence that S675-phosphorylated b-catenin is more stable than non-phosphorylated ones. As b-catenin is ubiquitinated via b-Trcp E3 ligase, we also tested its interaction with WT-, S675A- and S675D-b-catenin.


1005

Figure 3 PAK1 interacts with b-catenin and phosphorylates it at Ser675. (a) Western blots of initial lysates or immunoprecipitates (IP) by the indicated antibodies from SW480 cells. IB, immunoblotting. (b) HEK293T cells were transfected with the indicated plasmids, including Flag-b-catenin, myc-PAK1 (WT-PAK1), myc-PAK1 K299A (DN-PAK1), myc-PAK1 T423E (CA-PAK1), mycRac1, myc-Rac1 N17, myc-Rac1 V12 and empty vector. Cell lysates were immunoprecipitated with anti-Flag (M2) antibody. Total lysates and IP were subjected to IB with the indicated antibodies. Forskolin was used as a positive control. (c) HEK293T cells were transfected with Flag-b-catenin S37A, myc-PAK1 T423E, myc-Rac1 V12 or empty vector as indicated. Total cell lysates and IP by Flag (M2) beads were subjected to IB with the indicated antibodies. (d) Immunoblots of protein samples from in vitro kinase assay. Purified myc-PAK1 (WT-PAK1), myc-PAK1 K299A (DN-PAK1), myc-PAK1 T423E (CA-PAK1) were incubated with purified GSTb-catenin and the samples were subjected to IB with the indicated antibodies. (e) Lysates from SW480 cells stably expressing GFP or GFP–PAK1 K299A (GFP–PAK1 DN) were subjected to IB with the indicated antibodies. (f) Lysates from SW480 cells stably expressing control (Co) or PAK1 shRNA were subjected to IB with the indicated antibodies. (g) SW480 cells were transiently transfected with GFP or GFP–PAK1 T423E (CA-PAK1). GFP positive cells were sorted out 48 h later by a cell sorter and cell lysates were subjected to IB with the indicated antibodies. The relative amount of b-catenin in GFP-transfected cells is designated as 1.0. (h) Luciferase reporter assay in HCT116 cells using a cyclin D1 promoter-driven reporter construct. The cells were transiently transfected with myc-PAK1 T423E (CA-PAK1) and/or Flag-dominant negative-TCF4 (DN-TCF4).

The results indicated that S675D-b-catenin bound less b-Trcp (Figure 4c), consistent with it being less ubiquitinated. Whereas the interactions between b-catenin mutations and APC, glycogen synthase kinase-3b or axin were with no significant difference (Supplementary Figures 3D–F). These data strongly suggest that S675

phosphorylation has an important role in enhancing b-catenin stability and PAK1 stabilizes b-catenin through S675 phosphorylation in colon cancer cells. In addition to protein stability, S675 of b-catenin has been reported to be associated with transcriptional regulation (Taurin et al., 2006). However, more detailed Oncogene


1006

Figure 4 Phosphorylation at S675 ensures prolonged stabilization and transcriptional activity of b-catenin. (a) Flag-b-catenin was cotransfected with the indicated PAK1 constructs into HEK293T cells and 36 h later, the cells were treated with dimethyl sulfoxide or MG132 as indicated. The total lysates or the immunoprecipitates (IP) were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis and immunoblotting (IB) with the indicated antibodies. (b) Cell lysates from SW480 cells treated for 12 h with dimethyl sulfoxide or MG132 were immunoprecipitated with antibody against total or S675-phosphorylated b-catenin. The IP were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis and IB with the indicated antibodies. Please note that much more IP by S675-phos antibody were loaded to obtain equal total b-catenin. (c) Cell lysates from HEK293T cells co-transfected with Flag-b-Trcp and myc-b-catenin plasmids were subjected to immunoprecipitation with anti-myc agarose beads. Total lysates and IP were subjected to IB with the indicated antibodies. Cells were treated with ALLN (50 mM) for 4 h before lysis. (d) Wnt-responsive TOPflash reporter assay in HEK293T cells co-transfected with Flag-b-catenin WT, S675A or S675D plasmids. Equal cell lysates were analyzed by IB with anti-Flag antibody. (e) Cell lysates from HEK293T cells co-transfected with HA-HDAC1 and Flag-b-catenin WT, S675A or S675D plasmids were subjected to immunoprecipitation with anti-Flag M2 agarose beads. Total lysates and IPs were subjected to IB with the indicated antibodies. (f) Cell lysates from HEK293T cells co-transfected with HA-HDAC1, Flag-b-catenin, myc-CA-PAK1 (T423E) or myc-DN-PAK1 (K299A) were subjected to immunoprecipitation with anti-Flag M2 agarose beads. Total lysates and IPs were subjected to IB with the indicated antibodies.

information about how S675 regulates transcriptional activity of b-catenin remains elusive. In the Wnt reporter assay, S675D is much stronger, whereas S675A is weaker than the WT-b-catenin, when they were expressed at similar levels (Figure 4d). Also, S675D is more active in promoting reporter gene expression under the control of cyclin D1 promoter (Supplementary Figure 4A). To explore the underlying mechanism, we verified whether S675-phosphorylation influenced the formation of b-catenin-dependent transcriptional complexes. Immunoprecipitation experiments indicated that all three types of b-catenin bound to TCF4, as well as LEF-1, to a similar extent (Supplementary Figures 4B and C). However, HDAC1, a general transcriptional repressor involved in canonical Wnt signaling (Billin et al., 2000), exhibited significantly reduced interaction with S675D, but a slightly enhanced binding with S675A b-catenin (Figure 4e). Consistent with this result, CA-PAK1, but not the DN-PAK1, was able to almost completely block b-catenin/HDAC1 interaction (Figure 4f). Together, these results indicate that S675-phosphorylated b-catenin is more stable due to less ubiquitination and is more active in transcription due to enhanced CBP (Taurin et al., 2006), but reduced HADC interactions. These results suggest that PAK1, via phosphorylating S675, strengthens b-catenin in transducing Wnt signaling. Oncogene

Rac1 controls b-catenin level and phosphorylation through PAK1 Rac1 is able to enhance b-catenin phosphorylation on S675 (Figure 3b), we therefore addressed whether it controls b-catenin level in SW480 cells. Similarly, Rac1 was silenced by specific small interfering RNA (siRNA) and total b-catenin level was monitored by immunostainning and western blot. The results indicated that total b-catenin level was clearly reduced (Figures 5a and b). In comparison with the PAK1-knockdown cells, nuclear b-catenin staining was further reduced in Rac1-knockdown cells (compare Figures 5a and 2c) and this is consistent with the previous report that Rac1/JNK2 controls b-catenin nuclear localization (Wu et al., 2008; Phelps et al., 2009). In addition, NSC23766, a Rac1 inhibitor, was also able to reduce total b-catenin in SW480 cells (Figures 5c and d). As expected, the levels of S675-phosphorylated b-catenin and cyclin D1 were also reduced (Figures 5b and d). These results suggest that Rac1 is also required to keep high level of b-catenin accumulation in SW480 cells, in addition to promoting its nuclear localization. To further demonstrate that activated Rac1 is indeed able to enhance b-catenin accumulation, we transfected CA Rac1 (GFP–Rac1 V12) into SW480 cells and examined its effects on b-catenin. As shown in


1007

Figure 5 Rac1 controls the protein level and S675 phosphorylation of b-catenin via PAK1. (a) Immunofluoresence confocal images of SW480 cells transfected with control and Rac1 siRNA for 72 h. b-catenin and cell nuclei were exhibited as red and blue (DAPI), respectively. (b) Western blots of total lysates from SW480 cells transfected with control (Co) or Rac1 siRNA for 72 h. (c) Immunofluoresence confocal images of SW480 cells treated NSC23766 (a Rac1 inhibitor) for 24 h. b-catenin and cell nuclei were exhibited as red and blue (DAPI), respectively. (d) Western blots of total lysates from SW480 cells treated with NSC23766 for 24 h. (e) SW480 cells were transiently transfected with GFP or GFP–Rac1 V12 plasmids. GFP positive cells were sorted 48 h later with a cell sorter and the cell lysates were subjected to immunoblotting with the indicated antibodies. The relative amount of b-catenin in GFPtransfected cells is designated as 1.0. (f) SW480 cells were transfected with control (Co) or PAK1 siRNA and GFP–Rac1 V12 plasmids as indicated. After 48 h, total cell lysates were prepared and subjected to immunoblotting with the indicated antibodies.

Figure 5e, the S675-phosphorylated and total b-catenin were both upregulated. Epistatically, Rac1 modulate b-catenin via PAK1, as siRNA against PAK1 was able to block Rac1-induced S675-phosphorylation on b-catenin (Figure 5f). Together, these results suggest that a Rac1/PAK1 cascade is required in maintaining high level of b-catenin accumulation in SW480 cells.

Rac1/PAK1 activation and b-catenin accumulation Mutations or truncations of APC are frequent in colon cancer (Korinek et al., 1997; Morin et al., 1997), which is the major genetic abnormality leading to ectopic b-catenin accumulation. To further demonstrate that Rac1/PAK1 controls b-catenin stability, we thought to reconstitute this cooperative activity between loss of Oncogene


1008

Figure 6 b-catenin accumulation is correlated with PAK1 activation. (a) Wnt-responsive TOPflash reporter assay in HEK293T cells transfected with plasmids or siRNAs as indicated. Inset showing the knocking-down efficiency of APC. CA-PAK1 represents myc-CAPAK1 (T423E). Co, control. (b) Western blots of total lysates, nuclear or cytosol fractions from HEK293T cells transfected with the indicated plasmids or siRNAs. CA-PAK1 represents myc-PAK1 T423E. Co, control. (c) Wnt-responsive TOPflash reporter assay in HEK293T cells transfected with indicated plasmids or siRNAs. Rac1 V12 represents myc-Rac1 V12. DN-PAK1 represents myc-DNPAK1 (K299A). Co, control. (d) Representative immunohistochemical images of human colon cancer samples (lower panels) and paracancerous tissue (upper panels). The expression of PAK1 (left) and b-catenin (right) were detected with specific antibodies from the same tissue. The images were captured under the same microscopy settings.

APC and activation of Rac1/PAK1 in HEK293T cells. Wnt signaling was slightly activated upon knockdown of APC or overexpression of CA-PAK1; however, a much stronger signaling was observed when both were applied (Figure 6a). Consistently, CA-PAK1 further elevated protein levels of b-catenin in both cytosolic and nuclear fractions after APC knockdown (Figure 6b). We further tested CA-Rac1 in the Wnt reporter assay in APC-siRNA-transfected cells and observed similar cooperation (Figure 6c). Importantly, DN-PAK1 could significantly block Rac1-induced Wnt signaling in APC knockdown cells (Figure 6c). These results demonstrate that the Rac1/PAK1 cascade is able to enhance Wnt signaling in APC mutant cells, such as in colon cancer tissue. In order to establish a link between PAK1 activation and b-catenin accumulation in human colon cancer, we scanned 30 samples of colon cancer tissue and corresponding paracancerous tissue, with their expression and localization of PAK1 and total b-catenin, using immunohistochemistry. More than 70% of tumor samples (22/30) showed significant overexpression of PAK1, compared with corresponding paracancerous tissue (Figure 6d and Supplementary Table S1), which was consistent with previous report (Carter et al., 2004). In paracancerous tissue, b-catenin was only detected at Oncogene

the cell membrane, whereas in more than 90% (28/30) cancer samples, cytosolic and nuclear b-catenin staining were evident. Among these cancer samples, a significant correlation was observed between PAK1 overexpression and massive b-catenin accumulation (Supplementary Table S1). These results support our conclusion that overexpressed PAK1 contributes to b-catenin accumulation in colon cancer cells. PAK1 activity is required for full Wnt signaling We next addressed whether PAK1 is required for Wntinduced b-catenin signaling. As Wnt ligand is able to activate Rac1 (Habas et al., 2003; Bikkavilli et al., 2008; Wu et al., 2008), it is reasonable to speculate that PAK1 could also be activated, via Rac1. As expected, T423phosphorylated PAK1, an activated form, was induced upon treatment of Wnt3a conditioned medium (CM) (Figure 7a) or recombinant Wnt3a protein (Figure 7b) in HEK293T cells. When endogenous PAK1 was silenced using specific siRNA (Figure 7c), Wnt3a CM- or Wnt1induced TOPflash reporter expression was significantly reduced (Figures 7d and e). Moreover, DN-PAK1 also inhibited Wnt3a or Wnt1 signaling (Figures 7f and g), and PAK1 autoinhibitory domain, another DN form of PAK1, was also inhibitory to Wnt1 signaling (Figure 7g).


1009

Figure 7 PAK1 is required for canonical Wnt signaling. (a) Western blots of total lysates from HEK293T cells treated with L or Wnt3a conditioned medium (CM) for indicated time periods. P-PAK T423 recognizes the T423-phosphorylated and, therefore, activated PAK1 and PAK2. (b) Western blots of total lysates from HEK293T cells treated with purified recombinant Wnt3a protein (rWnt3a) at 100 ng/ml for indicated time periods. (c) Western blots showing reduced PAK1 protein in HEK293T cells transfected with PAK1 siRNA for 48 h. Co, control siRNA. (d–g) Wnt-responsive TOPflash reporter assays in HEK293T cells transfected with siRNA and/or plasmids as indicated. In (d), L or Wnt3a CM was applied to stimulate the signaling. In (e), mouse Wnt1 plasmid was co-transfected. DN-PAK1 represents myc-DN-PAK1 (K299A) and PID represents myc-PID (PAK1 autoinhibitory domain). (h) Western blots of total lysates from mouse L-cells transfected with the indicated siRNAs and treated with L or Wnt3a CM for 2 h. (i) Western blots of total lysates from mouse L-cells treated with L or Wnt3a CM, together with or without IPA-3 at indicated concentrations for 2 h. (j–l) Wnt-responsive TOPflash reporter assays in HEK293T cells transfected with the indicated plasmids. CA-PAK1 represents myc-CA-PAK1 (T423E). (m) Western blots of cytosolic fractions from HEK293T cells transfected with or without myc-CA-PAK1 (T423E), following 2 h treatment of L or Wnt3a CM.

In addition, we also found that Dvl2, b-catenin and even S37A b-catenin-induced Wnt signaling were dependent on endogenous PAK1 (Supplementary Figures 5A and B). To further confirm if PAK1 is required for b-catenin accumulation, we treated mouse L-cells with controlor Wnt3a-CM and detected total b-catenin levels. As shown in Figure 7h, knockdown of PAK1 blocked Wnt-induced b-catenin accumulation. Moreover, IPA-3 could also block Wnt3a-CM-induced b-catenin elevation (Figure 7i). Together, these results indicate that PAK1 activity is required for full b-catenin accumulation upon Wnt stimulation. Supporting this conclusion,

CA-PAK1 was able to enhance signaling from Wnt1, b-catenin, as well as b-catenin S37A (Figures 7j–l). Western blot further exhibited that CA-PAK1 could enhance Wnt3a-induced cytosolic b-catenin accumulation in HEK293T cells (Figure 7m). Finally, we found, using immunofluorescence assay, that a strong phosS675 b-catenin staining was accompanied with the elevation of total b-catenin in L-cells upon Wnt3a stimulation (Supplementary Figure 5C). In conclusion, these data suggest that Wnt stimulation is able to activate PAK1 and PAK1 activity is required for the full activation of b-catenin. Oncogene


1010

Discussion Over-simplified model suggested that Wnt stimulation blocks the N-terminal phosphorylation of b-catenin and this stabilized form enters the nucleus and activate transcription of target genes. However, there are evidences suggesting that different forms of b-catenin may have distinct transcriptional activities (Lu and Hunter, 2004; Hendriksen et al., 2008), arguing for further modifications on b-catenin protein. Subsequently, S191 and S605 of b-catenin were found to be phosphorylated by JNK2 and required for its nuclear localization (Wu et al., 2008). In this study, we demonstrate that PAK1 is capable of phosphorylating b-catenin at its S675 residue and this modification ensures further stability and transcriptional activity. These additional modifications on b-catenin could either be triggered by Wnt stimulation or by other signaling pathways. For example, S191 and S605 phosphorylation could be induced by Wnt stimulation in normal cells, whereas in colon cancer cells, K-Ras mutation may mimic this effect. Similarly, we found that Wnt protein is able to induce PAK1 activation in HEK293T cells and endogenous PAK1 activity is partially required for full b-catenin accumulation in both HEK293T and mouse L-cells upon Wnt stimulation. In colon cancer cells, signaling from K-Ras/Rac1 (Lee et al., 2007; Wang et al., 2010b), IGF (Sun et al., 2009) or EGF (Galisteo et al., 1996; Lu et al., 1997) pathways all may induce PAK1 activation, which maintains the high phosphorylation state of b-catenin S675. The oncogenic activity of PAK1 is quite certain; however, its cellular mechanisms are complicated, with diverse signaling pathways implicated. A recent study by Huynh et al. (2010) suggested that PAK1 stimulated colon cancer cell proliferation through AKT and ERK activation. However, we found that b-catenin level and Wnt signaling are heavily dependent on PAK1 activity in colon cancer cells. Our results suggest that activated Rac1/PAK1 strengthen b-catenin protein through S675 phosphorylation, which makes it more stable as well as more active in transcription. These results confirm that Rac1/PAK1 cascade can enhance cyclin D1 expression and cell proliferation, but emphasize that the activation is at least partially via b-catenin signaling. A K-Ras/Rac1/JNK2 cascade has been proposed to control b-catenin nuclear localization in response to Wnt stimulation, as well as in colon cancer progression (Esufali and Bapat, 2004; Wu et al., 2008; Phelps et al., 2009). As another downstream effector of Rac1, PAK1 would likely be activated through K-Ras/Rac1, suggesting that a potential K-Ras/Rac1/PAK1 cascade may exist in colon cancer cells. This is consistent with the previous finding that PAK1 functions downstream of Ras in promoting cell proliferation and transformation (Tang et al., 1997; Nheu et al., 2004). In our Rac1 knockdown cells, nuclear clearance of b-catenin staining (the staining intensity in the nucleus and cytosol became the same) was observed, confirming the previous report (Phelps et al., 2009). However, a reduction of total b-catenin level was also evident Oncogene

both in immunostainning and western blot experiments. In contrast to Rac1 knockdown cells, only total reduction of b-catenin was observed when PAK1 was downregulated. In these cells, a brighter nucleus was always detected, suggesting that the PAK1 controls the total level, but not the distribution of b-catenin in colon cancer cells. These results suggest that there are probably two signal branches that are activated by K-Ras/Rac1, of which one is mediated by JNK2 that promote b-catenin nuclear transportation and another is mediated by PAK1 that controls b-catenin stability and transcriptional activity. A link between Rac1/PAK1 cascade and Wnt signaling further exemplify the complex signaling network underlining tumorigenesis. Targeting PAK1 could be another potential therapeutic strategy for colon cancer treatment. IPA-3, a recently identified allosteric inhibitor for group I PAK, was very effective in blocking colon cancer cell proliferation in vitro (Figures 1g–i). The administration of IPA-3 in a colon cancer animal model, if possible, will be very interesting. We believe that it is of great importance to further design or screen specific inhibitors targeting PAK1 activation or overexpression for cancer therapy.

Materials and methods Cell culture HEK293T, HCT116 and mouse L-cells were cultured in Dulbecco’s modified Eagle’s medium. SW480 cells were cultured in L-15 (Leibovitz) medium. All culture medium were supplemented with 10% fetal bovine serum, penicillin and streptomycin. Luciferase reporter assay Luciferase reporter assays were carried out in 96-well plates, with triplicates as described previously (Wang et al., 2010a). DNA per well was: PAK1, 50 ng; Rac1 and its mutants, 50 ng; cyclin D1–luciferase reporter construct, 50 ng. Transfection and RNA interference Transfections were performed using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA, for L-cells or SW480 cells) or VigoFect transfection reagent (Vigorous Biotech, Beijing, China, for HEK293T and HCT116 cells) according to the manufacturer’s instructions. APC siRNA (Segditsas et al., 2008), PAK1 shRNA (Jacobs et al., 2007) and b-catenin shRNA (van de Wetering et al., 2003) were used as described previously. All siRNAs were synthesized by GenePharma (Shanghai, China). shRNA sequences for PAK1, b-catenin and control were as follows: human PAK1 shRNA (corresponding nucleotide sites 1412–1430 of PAK1 cDNA), sense 50 – 30 : gatccccgGAGCCTTGTACCTCATTGCttcaagagaGCAAT GAGGTACAAGGCTCctttttc; antisense 50 –30 : tcgagaaaaag GAGCCTTGTACCTCATTGCtctcttgaaGCAATGAGGTAC AAGGCTCcggg. Human b-catenin shRNA (corresponding nucleotide sites 1321–1339 of b-catenin cDNA), sense 50 –30 : gatccccgGTGGGTGGTATAGAGGCTCttcaagagaGAGCCT CTATACCACCCACctttttc; antisense 50 –30 : tcgagaaaaag GTGGGTGGTATAGAGGCTCtctcttgaaGAGCCTCTATA CCACCCACcggg. Control shRNA: sense 50 –30 : gatccccg GATGGATCGATATAGTGAGttcaagagaCTCACTATATC


1011 GATCCATCctttttc; antisense 50 –30 : tcgagaaaaagGATGG ATCGATATAGTGAGtctcttgaaCTCACTATATCGATCCA TCcggg. siRNA sequences for PAK1 and Rac1 were as follows: human PAK1-1: 50 -GAGTGTGGGCGATCCTA AGAAGAAA-30 ; PAK1-2: 50 -GCATTCGAACCAGGTCA TT-30 . Murine PAK1: 50 -GCTTCAGGCACAGTGTATA-30 . Human Rac1-1: 50 -GATAAAGACACGATCGAGA-30 ; Rac1-2: 50 -GCAAACAGATGTGTTCTTA-30 . Protein stability assay, co-immunoprecipitation assay and western blot Protein stability assay, western blot and co-immunoprecipitation experiments were performed as described previously (Wang et al., 2010a). To detect the interaction between endogenous PAK1 and b-catenin, SW480 cells were lysed with RIPA buffer (50 mM Tris–HCl pH 7.4, 150 mM NaCl, 1% NP40, 0.25% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 1 mM dithiothreitol, supplemented with protease inhibitor cocktail, 2 mM Na3VO4 and 25 mM NaF) and lysates were immunoprecipitated with anti-PAK1 antibody. The immunoprecipitates were washed with RIPA buffer for three times and subjected for sodium dodecyl sulfate–polyacrylamide gel electrophoresis and western blot. For ubiquitination assay in Figures 4a and b and Supplementary Figure 3C, HEK293T cells expressing exogenous proteins or SW480 cells were lysed with RIPA buffer and lysates were immunoprecipitated with the indicated antibodies for 4 h at 4 1C. Finally, immunoprecipitates were washed four times with RIPA buffer and subjected to western blot with anti-ubiquitin antibody. Immunofluorescence assay Cells on coverslips were washed with phosphate-buffered saline (PBS), fixed in 4% paraformaldehyde/PBS for 20 min at room temperature and then permeabilized in 0.2% Triton/PBS for 10 min. After 30 min blocking in 2% bovine serum albumin/PBS at room temperature, cells were incubated with following primary antibodies (b-catenin, diluted 1:250 in PBS; P-b-catenin S675, diluted 1:100; c-myc, diluted 1:100) for 1 h at room temperature or overnight at 4 1C. Then, cells were washed three times with PBS for 10 min each and incubated

with secondary antibody (donkey-anti-mouse Alexa 568 or goat-anti-rabbit Alexa 488, Invitrogen; diluted 1:500 in PBS containing 2% bovine serum albumin) for 1 h at room temperature. After washing with PBS, cells were incubated with 40 ,6-diamidino-2-phenylindole (DAPI, 1 mg/ml; Roche, Indianapolis, IN, USA) for 5 min and then washed three times with PBS and rinsed with ddH2O once. Finally, coverslips were sealed with Prolong Gold antifade reagent (Invitrogen). Samples were processed by Zeiss LSM 710 confocal microscope (Carl Zeiss MicroImaging GmbH, Munich, Germany). All immunofluoresence images within each experiment were captured under the same settings. Scale bar, 20 mm. Immunohistochemical analysis Immunohistochemistry was performed with human colon carcinoma (grades I–III with normal controls) tissue array (BC05118) by Cybrdi Inc (Xi’an, China). Anti-PAK1 and antib-catenin antibodies were used at 1:25 and 1:200 dilutions, respectively. A semi-quantitative analysis of PAK1 and bcatenin expression in stained sections was performed by an independent pathologist.

Conflict of interest The authors declare no conflict of interest.

Acknowledgements We thank Drs Ye-Guang Chen, Jonathan Chernoff, Raymond Habas, Alan Hall, Tianhui Hu, Edward Leof, Chaojun Li, Randy Moon, Christof Niehrs, Roel Nusse and Thomas Rudel for reagents. This work was supported by grants to WW from the National Natural Science Foundation of China (No. 30730048, 30921004), the Major Science Programs of China (2006CB943402, 2011CB943803), Tsinghua University Initiative Scientific Research Program (2010THZ0) and Bayer Healthcare.

References Anderson CB, Neufeld KL, White RL. (2002). Subcellular distribution of Wnt pathway proteins in normal and neoplastic colon. Proc Natl Acad Sci USA 99: 8683–8688. Arias-Romero LE, Chernoff J. (2008). A tale of two Paks. Biol Cell 100: 97–108. Balasenthil S, Sahin AA, Barnes CJ, Wang RA, Pestell RG, Vadlamudi RK et al. (2004). p21-activated kinase-1 signaling mediates cyclin D1 expression in mammary epithelial and cancer cells. J Biol Chem 279: 1422–1428. Bienz M, Clevers H. (2000). Linking colorectal cancer to Wnt signaling. Cell 103: 311–320. Bikkavilli RK, Feigin ME, Malbon CC. (2008). G alpha o mediates WNT-JNK signaling through dishevelled 1 and 3, RhoA family members, and MEKK 1 and 4 in mammalian cells. J Cell Sci 121: 234–245. Billin AN, Thirlwell H, Ayer DE. (2000). Beta-catenin-histone deacetylase interactions regulate the transition of LEF1 from a transcriptional repressor to an activator. Mol Cell Biol 20: 6882–6890. Bokoch GM. (2003). Biology of the p21-activated kinases. Annu Rev Biochem 72: 743–781. Carter JH, Douglass LE, Deddens JA, Colligan BM, Bhatt TR, Pemberton JO et al. (2004). Pak-1 expression increases with

progression of colorectal carcinomas to metastasis. Clin Cancer Res 10: 3448–3456. Clevers H. (2006). Wnt/beta-catenin signaling in development and disease. Cell 127: 469–480. Deacon SW, Beeser A, Fukui JA, Rennefahrt UE, Myers C, Chernoff J et al. (2008). An isoform-selective, small-molecule inhibitor targets the autoregulatory mechanism of p21-activated kinase. Chem Biol 15: 322–331. Dummler B, Ohshiro K, Kumar R, Field J. (2009). Pak protein kinases and their role in cancer. Cancer Metastasis Rev 28: 51–63. Esufali S, Bapat B. (2004). Cross-talk between Rac1 GTPase and dysregulated Wnt signaling pathway leads to cellular redistribution of beta-catenin and TCF/LEF-mediated transcriptional activation. Oncogene 23: 8260–8271. Fang D, Hawke D, Zheng Y, Xia Y, Meisenhelder J, Nika H et al. (2007). Phosphorylation of beta-catenin by AKT promotes betacatenin transcriptional activity. J Biol Chem 282: 11221–11229. Galisteo ML, Chernoff J, Su YC, Skolnik EY, Schlessinger J. (1996). The adaptor protein Nck links receptor tyrosine kinases with the serine-threonine kinase Pak1. J Biol Chem 271: 20997–21000. Giles RH, van Es JH, Clevers H. (2003). Caught up in a Wnt storm: Wnt signaling in cancer. Biochim Biophys Acta 1653: 1–24. Oncogene


1012 Habas R, Dawid IB, He X. (2003). Coactivation of Rac and Rho by Wnt/frizzled signaling is required for vertebrate gastrulation. Genes Dev 17: 295–309. He H, Shulkes A, Baldwin GS. (2008). PAK1 interacts with betacatenin and is required for the regulation of the beta-catenin signalling pathway by gastrins. Biochim Biophys Acta 1783: 1943–1954. Hendriksen J, Jansen M, Brown CM, van der Velde H, van Ham M, Galjart N et al. (2008). Plasma membrane recruitment of dephosphorylated beta-catenin upon activation of the Wnt pathway. J Cell Sci 121: 1793–1802. Hino S, Tanji C, Nakayama KI, Kikuchi A. (2005). Phosphorylation of beta-catenin by cyclic AMP-dependent protein kinase stabilizes beta-catenin through inhibition of its ubiquitination. Mol Cell Biol 25: 9063–9072. Huynh N, Liu KH, Baldwin GS, He H. (2010). P21-activated kinase 1 stimulates colon cancer cell growth and migration/invasion via ERK- and AKT-dependent pathways. Biochim Biophys Acta 1803: 1106–1113. Ilyas M, Tomlinson IP, Rowan A, Pignatelli M, Bodmer WF. (1997). Beta-catenin mutations in cell lines established from human colorectal cancers. Proc Natl Acad Sci USA 94: 10330–10334. Jacobs T, Causeret F, Nishimura YV, Terao M, Norman A, Hoshino M et al. (2007). Localized activation of p21-activated kinase controls neuronal polarity and morphology. J Neurosci 27: 8604–8615. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Kinzler KW et al. (1997). Constitutive transcriptional activation by a beta-catenin-Tcf complex in APC/ colon carcinoma. Science 275: 1784–1787. Kumar R, Gururaj AE, Barnes CJ. (2006). p21-activated kinases in cancer. Nat Rev Cancer 6: 459–471. Lee SH, Kunz J, Lin SH, Yu-Lee LY. (2007). 16-kDa prolactin inhibits endothelial cell migration by down-regulating the RasTiam1-Rac1-Pak1 signaling pathway. Cancer Res 67: 11045–11053. Logan CY, Nusse R. (2004). The Wnt signaling pathway in development and disease. Annu Rev Cell Dev Biol 20: 781–810. Lu W, Katz S, Gupta R, Mayer BJ. (1997). Activation of Pak by membrane localization mediated by an SH3 domain from the adaptor protein Nck. Curr Biol 7: 85–94. Lu Z, Hunter T. (2004). Wnt-independent beta-catenin transactivation in tumor development. Cell Cycle 3: 571–573. MacDonald BT, Tamai K, He X. (2009). Wnt/beta-catenin signaling: components, mechanisms, and diseases. Dev Cell 17: 9–26. Manser E, Leung T, Salihuddin H, Zhao ZS, Lim L. (1994). A brain serine/threonine protein kinase activated by Cdc42 and Rac1. Nature 367: 40–46. Molli PR, Li DQ, Murray BW, Rayala SK, Kumar R. (2009). PAK signaling in oncogenesis. Oncogene 28: 2545–2555. Moon RT, Kohn AD, De Ferrari GV, Kaykas A. (2004). WNT and beta-catenin signalling: diseases and therapies. Nat Rev Genet 5: 691–701. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B et al. (1997). Activation of beta-catenin-Tcf signaling in colon cancer by mutations in beta-catenin or APC. Science 275: 1787–1790. Mosimann C, Hausmann G, Basler K. (2009). Beta-catenin hits chromatin: regulation of Wnt target gene activation. Nat Rev Mol Cell Biol 10: 276–286. Nheu T, He H, Hirokawa Y, Walker F, Wood J, Maruta H. (2004). PAK is essential for RAS-induced upregulation of cyclin D1 during the G1 to S transition. Cell Cycle 3: 71–74.

Phelps RA, Chidester S, Dehghanizadeh S, Phelps J, Sandoval IT, Rai K et al. (2009). A two-step model for colon adenoma initiation and progression caused by APC loss. Cell 137: 623–634. Polakis P. (2000). Wnt signaling and cancer. Genes Dev 14: 1837–1851. Polakis P. (2002). Casein kinase 1: a Wnt’er of disconnect. Curr Biol 12: R499–R501. Rennefahrt UE, Deacon SW, Parker SA, Devarajan K, Beeser A, Chernoff J et al. (2007). Specificity profiling of Pak kinases allows identification of novel phosphorylation sites. J Biol Chem 282: 15667–15678. Rong R, Surace EI, Haipek CA, Gutmann DH, Ye K. (2004). Serine 518 phosphorylation modulates merlin intramolecular association and binding to critical effectors important for NF2 growth suppression. Oncogene 23: 8447–8454. Segditsas S, Sieber OM, Rowan A, Setien F, Neale K, Phillips RK et al. (2008). Promoter hypermethylation leads to decreased APC mRNA expression in familial polyposis and sporadic colorectal tumours, but does not substitute for truncating mutations. Exp Mol Pathol 85: 201–206. Sun J, Khalid S, Rozakis-Adcock M, Fantus IG, Jin T. (2009). P-21-activated protein kinase-1 functions as a linker between insulin and Wnt signaling pathways in the intestine. Oncogene 28: 3132–3144. Tang Y, Chen Z, Ambrose D, Liu J, Gibbs JB, Chernoff J et al. (1997). Kinase-deficient Pak1 mutants inhibit Ras transformation of Rat-1 fibroblasts. Mol Cell Biol 17: 4454–4464. Taurin S, Sandbo N, Qin Y, Browning D, Dulin NO. (2006). Phosphorylation of beta-catenin by cyclic AMP-dependent protein kinase. J Biol Chem 281: 9971–9976. Valentini AM, Pirrelli M, Renna L, Armentano R, Caruso ML. (2003). P53 and beta-catenin in colorectal cancer progression. Curr Pharm Des 9: 1932–1936. van de Wetering M, Oving I, Muncan V, Pon Fong MT, Brantjes H, van Leenen D et al. (2003). Specific inhibition of gene expression using a stably integrated, inducible small-interfering-RNA vector. EMBO Rep 4: 609–615. van de Wetering M, Sancho E, Verweij C, de Lau W, Oving I, Hurlstone A et al. (2002). The beta-catenin/TCF-4 complex imposes a crypt progenitor phenotype on colorectal cancer cells. Cell 111: 241–250. Wang RA, Mazumdar A, Vadlamudi RK, Kumar R. (2002). P21-activated kinase-1 phosphorylates and transactivates estrogen receptor-alpha and promotes hyperplasia in mammary epithelium. EMBO J 21: 5437–5447. Wang Y, Fu Y, Gao L, Zhu G, Liang J, Gao C et al. (2010a). Xenopus skip modulates Wnt/beta-catenin signaling and functions in neural crest induction. J Biol Chem 285: 10890–10901. Wang Z, Pedersen E, Basse A, Lefever T, Peyrollier K, Kapoor S et al. (2010b). Rac1 is crucial for Ras-dependent skin tumor formation by controlling Pak1-Mek-Erk hyperactivation and hyperproliferation in vivo. Oncogene 29: 3362–3373. Wells CM, Jones GE. (2010). The emerging importance of group II PAKs. Biochem J 425: 465–473. Wu X, Tu X, Joeng KS, Hilton MJ, Williams DA, Long F. (2008). Rac1 activation controls nuclear localization of beta-catenin during canonical Wnt signaling. Cell 133: 340–353. Xiao GH, Beeser A, Chernoff J, Testa JR. (2002). p21-activated kinase links Rac/Cdc42 signaling to merlin. J Biol Chem 277: 883–886. Xing Y, Clements WK, Le Trong I, Hinds TR, Stenkamp R, Kimelman D et al. (2004). Crystal structure of a beta-catenin/ APC complex reveals a critical role for APC phosphorylation in APC function. Mol Cell 15: 523–533.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene