Overexpression of a Novel Activator of PAK4, the CDK5 Kinase–Associated Protein CDK5RAP3, Promotes Hepatocellular Carcinoma Metastasis

Published OnlineFirst March 8, 2011; DOI: 10.1158/0008-5472.CAN-10-4046

Cancer Molecular and Cellular Pathobiology Research

Grace Wing-Yan Mak1, Mandy Man-Lok Chan1, Veronica Yee-Law Leong1, Joyce Man-Fong Lee2,3, Tai-On Yau2, Irene Oi-Lin Ng2,3, and Yick-Pang Ching1,3

Abstract The CDK5 kinase regulatory subunit-associated protein 3 (CDK5RAP3 or C53/LZAP) regulates apoptosis induced by genotoxic stress. Although CDK5RAP3 has been implicated in cancer progression, its exact role in carcinogenesis is not well established. In this article, we report that CDK5RAP3 has an important prometastatic function in hepatocarcinogenesis. An examination of human hepatocellular carcinoma (HCC) samples revealed at least twofold overexpression of CDK5RAP3 transcripts in 58% (39/67) of HCC specimens when compared with corresponding nontumorous livers. CDK5RAP3 overexpression was associated with more aggressive biological behavior. In HCC cell lines, stable overexpression of CDK5RAP3 promoted, and small interfering RNA–mediated knockdown inhibited, tumorigenic activity and metastatic potential. We found that overexpression of CDK5RAP3 and p21-activated protein kinase 4 (PAK4) correlated in human HCCs, and that CDK5RAP3 was a novel binding partner of PAK4, and this binding enhanced PAK4 activity. siRNA-mediated knockdown of PAK4 in CDK5RAP3-expressing HCC cells reversed the enhanced cell invasiveness mediated by CDK5RAP3 overexpression, implying that PAK4 is essential for CDK5RAP3 function. Taken together, our findings reveal that CDK5RAP3 is widely overexpressed in HCC and that overexpression of CDK5RAP3 promotes HCC metastasis through PAK4 activation. Cancer Res; 71(8); 2949–58. 2011 AACR.

Introduction suggests that CDK5RAP3 may function as a tumor suppressor. On the contrary, stable overexpression of the CDK5RAP3 The CDK5 kinase regulatory subunit-associated protein 3 isoform has been shown to promote hepatocellular carcinoma (CDK5RAP3, also called C53/LZAP) was first identified as a (HCC) and cardiac cell proliferation (6, 7), which indicates, binding partner of cyclin-dependent kinase 5 activator, furthermore, that CDK5RAP3 may enhance cell growth. p35nck5a, in yeast 2-hybrid screening (1). Northern analysis CDK5RAP3 is located at chromosome region 17q21.32, indicated that CDK5RAP3 is widely expressed in human which has been reported to be amplified in HCC; however, tissues, and the expression level is relatively constant in the the role of CDK5RAP3 in HCC has not been explored so far (8). heart, brain, skeletal muscle, placenta, lung, liver, kidney, and In this study, we found that the expression of CDK5RAP3 was pancreas (2). Overexpression of CDK5RAP3 has been shown to frequently upregulated in human HCCs at both transcript and sensitize cells to apoptosis induced by genotoxic stress (3). protein levels. More importantly, we detected a remarkable CDK5RAP3 can interact with a well-known tumor suppressor, enhancement of CDK5RAP3 expression in metastatic HCC. namely, the alternate reading frame (ARF; p14ARF), by which it Although little information is available on how CDK5RAP3 stabilizes and promotes the transcription activity of p53 (4). regulates cancer metastasis, we found that CDK5RAP3 is a More recently, CDK5RAP3 has been found to be underex- novel activator of p21-activated protein kinase 4 (PAK4) and pressed in head and neck cancers, and forced expression of activation of PAK4 can promote HCC cell migration. There- CDK5RAP3 can negatively regulate NF-kB activity (5), which fore, we provided, in this study, a novel mechanism by which CDK5RAP3 contributes to the metastasis of HCC by activation Authors' Affiliations: Departments of 1Anatomy and 2Pathology, Li Ka of PAK4. Shing Faculty of Medicine, and 3State Key Laboratory for Liver Research, The University of Hong Kong, Hong Kong, China Materials and Methods Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/). Cell culture Corresponding Author: Yick-Pang Ching, Department of Anatomy, The University of Hong Kong, Room L1-43, Laboratory Block, Faculty of Human hepatoma cell lines PLC/PRF/5 and HepG2 were Medicine Building, Hong Kong. Phone: 852-28199434; Fax: 852- purchased from the American Type Culture Collection. The 28170857; E-mail: [email protected] authentication of these cell lines was ensured by the provider doi: 10.1158/0008-5472.CAN-10-4046 through cytogenetic analysis. No additional test was conducted 2011 American Association for Cancer Research. specifically for this study. The human HCC cell line

www.aacrjournals.org 2949

Mak et al.

SMMC-7721 was a gift from the Shanghai Institute of Biochem- Purification of protein istry and Cell Biology, Chinese Academy of Sciences. Cells were GST fusion proteins were purified using reduced glu- maintained in Dulbecco's modified Eagle medium (DMEM)- tathione (GSH)-sepharose beads by a procedure described high glucose (Life Technologies) supplemented with 1 mmol/L previously (1). His-tagged proteins were purified using Nickel- sodium pyruvate and 10% heat-inactivated FBS (JRH Bios- NTA agarose beads (Qiagen). Bacterial cells were lysed in a ciences). Cells were transfected with DNA constructs, using lysis buffer containing 20 mmol/L Tris (pH 7.5), 500 mmol/L Lipofectamine 2000 (Invitrogen) according to the man- NaCl, 5 mmol/L imidazole, 1 mmol/L dithiothreitol (DTT), ufacturer's protocol. For constructing the stable clones, cells 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L leupep- were transfected with expression constructs (Myc-CDK5RAP3 tin, and 10 mg/mL lysozyme. The protein was eluted with or shCDK5RAP3) and were selected with the correspond- 100 mmol/L imidazole in Tris buffer. ing antibiotic puromycin at 0.4 mg/mL (Sigma) or G418 at 0.8 mg/mL (Calbiochem), 48 hours after transfection. PAK4 kinase assay An in vitro gel kinase assay was conducted as described CDK5RAP3 antibody previously (12). GST-CDK5RAP3 and its mutants were incu- Rabbit anti-CDK5RAP3 polyclonal antibody was generated bated with His-PAK4 and GST-PAK4 peptide substrate [amino using purified glutathione S-transferase (GST)-CDK5RAP3 acid (aa) sequence AARRLSVASAK, named PAK4tide, designed fusion protein as antigen (Genscript Company). Then, the on the basis of the published result (13)] in PAK4 kinase buffer antibody was affinity purified by (CNBr)-GST-CDK5RAP3 and [50 mmol/L HEPES (pH 7.5), 5 mmol/L MgCl2, 100 mmol/L GST column. NaCl, and 1 mmol/L DTT] containing 10 mCi [g-32P]ATP. For peptide-kinase assay, GST and GST-CDK5RAP3 were incu- Immunohistochemistry bated with His-PAK4 in the presence of 1 mmol/L PAK4tide Immunohistochemical staining for CDK5RAP3 and phos- and 10 mCi [g-32P]ATP at 30C for 10 minutes (14). Peptides pho-PAK4 (p-PAK4) was done as described previously (9). with single mutation (AARRLAVASAK) were included as a Purified anti-CDK5RAP3 antibody and anti-p-PAK4 (Ser474) negative control. antibody (Cell Signaling Technology) were used at 1:100 and 1:50 dilutions, respectively. Scoring of the protein expression GST affinity pull-down assay was assessed by an experienced pathologist. GST and GST-CDK5RAP3 immobilized by GSH-sepharose beads (GE Healthcare) were incubated with His-PAK4 for 2 Real-time quantitative reverse transcriptase PCR hours at 4C, followed by washing 3 times with NETN buffer Quantitative PCR (qPCR) was done as described previously [50 mmol/L NaCl, 5 mmol/L EDTA, 50 mmol/L Tris (pH 8.0), (9). The sequence of TaqMan probes (Applied Biosystems) for 1% NP40]. The bound proteins were then visualized by CDK5RAP3 and PAK4 are 50-AGGAAAGATGGAGGACCAT- Western blotting. CAGCAC-30 and 50-GCGGCGCCGAGCCGATGAGTAACCC-30, respectively; cellular 18S rRNA was used as an internal control. Coimmunoprecipitation HEK293T cells, which were ectopically expressed with GFP- Colony formation assay PAK4 and Myc-CDK5RAP3, were lysed with NETN buffer sup- Cells were transiently transfected with plasmids expressing plemented with freshly prepared protease inhibitors (1 mmol/L CDK5RAP3 and shCDK5RAP3 and were selected with corre- phenylmethylsulfonyl fluoride, 1 mg/mL leupeptin, 2 mg/mL sponding antibiotics for 2 weeks. The experiments were done aprotinin, and 1 mmol/L DTT). Antibodies used for immuno- 3 times independently. precipitation include anti-Myc and rabbit anti-PAK4, and antibodies used for immunoblotting are rabbit anti-Myc, rabbit anti- Cell proliferation, soft agar growth, cell migration, and GFP, and rabbit anti-CDK5RAP3. invasion assays The procedure described by Leung and colleagues (10) and Confocal microscopy Wong and colleagues (11) was adopted. For soft agar assay, Cells were fixed with 4% paraformaldehyde and permeabi- colonies with diameter greater than 50 mmin10fieldswere lized with 0.2% Triton X-100 (9). Images were captured by a counted and data were shown as average number of colonies per confocal laser scanning microscope LSM510 (Carl Zeiss). field. The experiments were carried out 3 times independently. Statistical analysis Nude mouse xenograft assay The Student t test and the Mann–Whitney test were used CDK5RAP3 stable knockdown PLC/PRF/5 cells and control for statistical analysis of data. Tests were considered signifi- cells (5 106) were administered by s.c. injection into the right cant with P < 0.05. and left flanks of 6-week-old male nude mice, respectively. The larger (a) and smaller (b) diameters of a tumor were measured Results weekly. Tumor volume was estimated according to the formula: volume ¼ 1/2 a b2; and it was measured from week 2 to CDK5RAP3 was overexpressed in human HCCs week 5. After 5 weeks, the mice were sacrificed in accordance To elucidate the role of CDK5RAP3 in human HCCs, we with institutional regulations for animal experiments. examined CDK5RAP3 transcripts in human HCCs by using

2950 Cancer Res; 71(8) April 15, 2011 Cancer Research

CDK5RAP3 Activates PAK4 in HCC

Figure 1. Overexpression of ABNontumor liver HCC CDK5RAP3 in HCC. A, 10 Case

CDK5RAP3 transcripts were 6 250 measured in 67 paired human HCCs and corresponding nontumorous liver tissues by 1 qPCR. The horizontal lines C indicate the median of mRNA expression; P < 0.001, Mann– Nontumor liver HCC HCC metastasis Whitney test; T, tumor; NT, 0.1 nontumor. B, representative immunohistochemical staining for CDK5RAP3 in a case of HCC and nfumber/10 Relative copy its corresponding nontumorous 0.01 CDK5RAP3 NT T liver tissues. C, representative Case 9 Nontumor liver HCC HCC metastasis immunohistochemical staining (magnification 20) of CDK5RAP3 and p-PAK4 (Ser474) in nontumorous liver, primary HCC, and extrahepatic

metastasis. (Ser474) p-PAK4

qPCR. We found that CDK5RAP3 transcripts were frequently HCCs than in the corresponding nontumorous livers. Among (58%, 39/67) overexpressed (2-fold) in the HCCs as com- the 11 cases with overexpression of CDK5RAP3, 55% (6/11) pared with their corresponding nontumorous livers (Fig. 1A). had higher levels of CDK5RAP3 in tumor metastases than in Moreover, the transcript levels of CDK5RAP3 in the tumor the corresponding primary HCCs (Fig. 1C), indicating upre- samples were significantly higher than those of the nontu- gulation of CDK5RAP3 during HCC progression. morous livers (P < 0.001, Mann–Whitney test). To understand the clinicopathologic significance of CDK5RAP3 in HCC, we CDK5RAP3 enhanced cell proliferation and correlated the overexpression of CDK5RAP3 mRNA with the tumorigenicity of HCC cells patients’ clinicopathologic features (Table 1). Overexpression The expression of the human CDK5RAP3 isoform (IC53-2) of CDK5RAP3 was found to be significantly associated with a has previously been shown to promote the proliferation of more aggressive phenotype, namely, the presence of tumor HCC cells (7). To study the effect of the full-length (FL) form microsatellite formation (P ¼ 0.024) and poorer cellular dif- of CDK5RAP3 on growth properties of HCC cells, we con- ferentiation (P ¼ 0.023). However, for the survival analysis, ducted a colony formation assay in human HCC cell lines overexpression of CDK5RAP3 was not associated with the with transient overexpression and knockdown of overall and disease-free survival rates among patients CDK5RAP3. PLC/PRF/5 cells transiently transfected with (Table 1). CDK5RAP3 expression plasmid formed more colonies than To examine the expression level of the protein CDK5RAP3 the vector control (P ¼ 0.005; Fig. 2A), whereas transient in human HCCs, we generated an antibody that specifically knockdown of CDK5RAP3 using short hairpin RNA (shRNA) recognized the CDK5RAP3 protein (Supplementary Fig. S1A). specific to CDK5RAP3 suppressed colony formation in PLC/ Using the purified antibody, we showed that CDK5RAP3 was PRF/5 cells (P ¼ 0.005; Fig. 2B), indicating that CDK5RAP3 ubiquitously expressed in a panel of HCC cell lines (Supple- promotes HCC cell growth. To further examine the effect of mentary Fig. S1B). To confirm that CDK5RAP3 was upregu- CDK5RAP3 in HCC tumorigenesis, we established 2 lated in human HCCs, immunohistochemical staining was CDK5RAP3 stably overexpressing HepG2 cells. The carried out. As shown in Fig. 1B, there was a strong cytoplas- CDK5RAP3 stable overexpression was confirmed by immu- mic staining of CDK5RAP3 in the tumor cells but not in the noblotting (Fig. 2C). A cell proliferation assay using the corresponding nontumorous hepatocytes, confirming that stable clones showed that the doubling time of both CDK5RAP3 was overexpressed in HCCs. Because overexpres- CDK5RAP3 stably expressing clones 1 and 2 (CDK5RAP3#1, sion of CDK5RAP3 mRNA in HCC samples was associated with 29.28 hours; CDK5RAP3#2, 28.56 hours) was remarkably a more metastatic phenotype, we examined the expression of shorter than the vector control (31.6 hours), indicating that CDK5RAP3 in patients with extrahepatic metastatic HCC. A stable clones grew faster than the vector (Fig. 2C). Because tissue microarray consisting of 25 cases each, containing the there is little information about the tumorigenic activity of primary HCC, extrahepatic metastasis, and nontumorous liver CDK5RAP3, a soft agar growth assay was conducted to from the same patient, was examined for CDK5RAP3 expres- evaluate the change in the anchorage-independent growth sion. Quantification of CDK5RAP3 staining was scored by an property of the CDK5RAP3 stably expressing HCC cells. The experienced pathologist (I. Oi-Lin Ng). Eleven of the 25 (44%) results showed that 3-fold more colonies were formed in cases had higher expression levels of CDK5RAP3 in primary stable clones CDK5RAP3#1 (P ¼ 0.027) and CDK5RAP3#2

www.aacrjournals.org Cancer Res; 71(8) April 15, 2011 2951

Mak et al.

Table 1. Association of overexpression of CDK5RAP3 mRNA with clinicopathologic features in human HCCs

Parameters Category CDK5RAP3 not CDK5RAP3 P value overexpressed overexpressed

Overall survival 23 37 0.264 Disease-free survival 23 37 0.701 Sex Male 16 27 0.026a Female 10 4 Age, y 40 2 4 0.678 >40 24 27 HBV surface antigen Absent 9 8 0.267 Present 18 30 HBV core antigen Absent 14 18 0.582 Present 2 1 Liver cirrhosis Absent 8 16 0.139 Present 18 16 Resection margin Absent 25 37 0.642 Present 3 2 Venous invasion Absent 17 18 0.239 Present 11 21 Tumor microsatellite Absent 20 17 0.024a Present 8 22 Liver invasion Absent 16 20 0.192 Present 7 18 Tumor encapsulation Absent 16 27 0.322 Present 11 11 Cellular differentiation I–III 19 15 0.023a (Edmondson's grading) IV–VI 9 23 Tumor size (cm) 5 9 17 0.159 >51715 Tumor stage I–II 12 13 0.672 III–V14 19 Number of tumor nodules Single 22 27 1.000 Multiple 4 5

aStatistically significant.

(P ¼ 0.002) than in the vector control (Fig. 2D). To address CDK5RAP3 stably overexpressing HepG2 cells, these results the loss-of-function effect of CDK5RAP3 in HCC cells, we consistently suggest that CDK5RAP3 enhanced the tumor- established CDK5RAP3 stable knockdown clones in PLC/ igenicity of HCC cells and the effect was not cell-line specific. PRF/5 and the highly motile SMMC-7721 HCC cells by using To further show the tumorigenic properties of CDK5RAP3 shRNA (3). Thus, 2 different cell lines were used for estab- in vivo, a nude mouse xenograft assay was conducted using lishing stable clones to eliminate cell-line–specific effect. CDK5RAP3 stable knockdown PLC/PRF/5 cells. Although the Two CDK5RAP3 stable knockdown clones were selected tumor incidences for the PLC/PRF/5-shCDK5RAP3#2 (4/5) (shCDK5RAP3#1 and shCDK5RAP3#2) in each cell line, and vector control (5/5) were similar, tumors generated from and the knockdown of CDK5RAP3 was confirmed by Wes- PLC/PRF/5-shCDK5RAP3#2 had remarkably lower weights tern blotting (Fig. 2E and Supplementary Fig. S2A). As and slower growth rates than the control (Fig. 2G), suggesting compared with vector control cells, the loss of CDK5RAP3 that loss of CDK5RAP3 inhibited tumor growth in vivo. significantly reduced both the cell-proliferation rate (Fig. 2E) and anchorage-independent growth of PLC/PRF/5 cells Knockdown of CDK5RAP3 inhibited cell migration and (Fig. 2F, P < 0.005). Furthermore, a similar result was invasiveness of HCC cells observed in CDK5RAP3 stable knockdown SMMC-7721 Clinical data for HCCs in this study reveal that clones (Supplementary Fig. S2). Together with the data in CDK5RAP3 protein levels increased with HCC progression

2952 Cancer Res; 71(8) April 15, 2011 Cancer Research

CDK5RAP3 Activates PAK4 in HCC

Tumor incidence ACEMock G Vector shCDK5RAP3#2 5/5 4/5 Vector arental Vector shCDK5RAP3#2 HepG2vector parentalCDK5RAP3#1CDK5RAP3#2Myc-CDK5RAP3 P vector#1vector#2shCDK5RAP3#1shCDK5RAP3#2 PLC/PRF/5 Endogenous CDK5RAP3 CDK5RAP3 CDK5RAP3 β-Actin β 80 HepG2 parental -Actin 250 Vector#1 45 Vector 70 Vector#2 1,250 Vector

CDK5RAP3#1 3

40 ) CDK5RAP3#2 ) shCDK5RAP3#1 4 4 200 shCDK5RAP3#2 shCDK5RAP3#2 35 60 1,000 30 50 25 150 750 20 40 15 30 100 500

Number of colonies 10 Tumor volume (mm ) Cell number (×10 5 Cell number (×10 20 0 50 250 Mock Vector CDK5RAP3 10 Myc-CD5RAP3 0 0 0 0 24 48 72 96 120 140 168 0 24 48 72 96 120 144 β-Actin 0123456 Time (h) Time (h) Week after injection 80 BD25 F50 1.00 70 45 20 60 40 0.75 50 15 35 40 30 10 30 25 0.50 20 20

Number of colonies 5

10 15 weight (g) Tumor 0.25 Number of colonies >50 μ m 0 0 10 Vector shCDK5RAP3 Number of colonies >50 μ m 5 CDK5RAP3 ector 0.00 V 0 Vector shCDK5RAP3#2 β-Actin

CDK5RAP3#1CDK5RAP3#2 ector#1 ector#2 V V

CDK5RAP3#1CDK5RAP3#2

Figure 2. CDK5RAP3-enhanced tumorigenicity of HCC cells. A, the construct expressing Myc-tagged CDK5RAP3 was transfected into PLC/PRF/5 cells for colony formation assay. The expression of CDK5RAP3 was confirmed by immunoblotting. Error bars, mean SD; *, P < 0.05 compared with vector control (Student t test). Empty vector (vector) and untransfected cells (mock) were used as controls. B, same as (A), but endogenous CDK5RAP3 expression was silenced by shCDK5RAP3. C, CDK5RAP3 stably expressing HepG2 cells (CDK5RAP3#1 and CDK5RAP3#2), vector control, and parental cells were used for proliferation assay. Top, immunoblotting; bottom, proliferation curve. D, cells in (C) were used for soft agar growth assay. *, P < 0.05; **, P < 0.005, compared with HepG2–vector control cells (Student t test). Error bars, mean SD. E, CDK5RAP3 stable knockdown PLC/PRF/5 clones were used for proliferation assay. Top, immunoblotting; bottom, proliferation curve. F, 2 CDK5RAP3 stable knockdown PLC/PRF/5 clones were used for soft agar growth assay; *, P < 0.005 compared with PLC-vector 1 (Student t test). Error bars, mean SD. G, CDK5RAP3 stable knockdown PLC/PRF/5 clones and vector control were used for the nude mouse xenograft model. Tumor volume (error bars, mean SEM); tumor weight (error bars, mean SD).

and that overexpression of CDK5RAP3 was statistically cor- Association analysis and physical correlation between related with clinicopathologic features of a more metastatic CDK5RAP3 and PAK4 in HCCs phenotype. To examine whether the loss of CDK5RAP3 inhi- From our functional assays and clinicopathologic correla- bits cell motility, Transwell migration assay was conducted tion data, we inferred that CDK5RAP3 was closely associated using SMMC-7721 cells treated with small interfering RNA with HCC cell invasiveness. We then analyzed the correlation (siRNA) targeting CDK5RAP3 (siCDK5RAP3). The number of of the overexpression of CDK5RAP3 transcripts with the migrated SMMC-7721 cells transfected with siCDK5RAP3 was expression levels of metastasis-related genes (e.g., PAK4, reduced by 36% as compared with the cells transfected with PAK1, NF-kB, and p53) in our database (9). Interestingly, we siRNA control (Fig. 3A). Furthermore, we then queried found that CDK5RAP3 overexpression was significantly corre- whether loss of CDK5RAP3 suppressed the invasiveness of lated with PAK4 overexpression in 63 cases of HCC (P < 0.001, HCC cells. Compared with parental and siRNA-control trans- Pearson correlation; Fig. 4A). fected cells, the number of invaded cells was reduced by 43% in To delineate the physiologic implication of this correla- siCDK5RAP3 transfected SMMC-7721 cells (Fig. 3B). In con- tion, we first examined the interaction of PAK4 and trast, the CDK5RAP3 stably overexpressing HepG2 cells CDK5RAP3 by coimmunoprecipitation assay. GFP-PAK4, showed an increase in migration rate by 37% in the Transwell but not GFP-PAK1, was coprecipitated with Myc-CDK5RAP3 assay (Fig. 3C). Thus, both loss- and gain-of-function assays in transfected HEK293T cells, indicating that CDK5RAP3 is indicated that CDK5RAP3 promoted the cell migration and specificallyassociatedwithPAK4(Fig.4B).Furthermore,the invasiveness of HCC cells. endogenous CDK5RAP3 and PAK4proteinswereshown,by

www.aacrjournals.org Cancer Res; 71(8) April 15, 2011 2953

Mak et al.

A B Mock siControl siCDK5RAP3

Mock siControl siCDK5RAP3 600 * CDK5RAP3 500 Figure 3. Knockdown of β-Actin CDK5RAP3 suppressed HCC cell 400 migration and invasiveness. Mock siControl siCDK5RAP3 300 SMMC-7721 cells were transfected with siCDK5RAP3 and 200 siRNA control, and nontransfected cells were 100 subjected to (A) the Transwell 350 * Number of invaded cells Number of invaded 0 migration assay and (B) the 300 Mock siControl siCDK5RAP3 invasion assay. *, P ¼ 0.05 compared with siControl (Student 250 C 500 t test). Immunoblotting showed 200 the CDK5RAP3 expression. 400 Representative photomicrographs 150 are shown. C, CDK5RAP3 stably 300 100 overexpressing HepG2 cells were used for the Transwell migration 200 50 assay. *, P < 0.05 compared with HepG2–vector control cells Number of immigrated cells 0 100 (Student t test). Error bars, mean Mock siControl siCDK5RAP3 0 SD. Number of immigrated cells

Vector

CDK5RAP3#1CDK5RAP3#2

coimmunoprecipitation assay, to interact with each other in of CDK5RAP3 were associated with N-terminal region of PAK4, HepG2 cells (Fig. 4C). Moreover, a GST affinity pull-down assay which contained the PBD domain but not the C-terminal showed a direct interaction between CDK5RAP3 and PAK4 kinase domain (Fig. 5B). To further examine whether (Fig. 5A). By confocal microscopy, CDK5RAP3 was found to be CDK5RAP3 regulated PAK4 activity, a PAK4 kinase assay significantly colocalized with PAK4 (Fig. 4D). The subcellular was conducted. The result revealed that incubation of FL, localization of CDK5RAP3, which was mainly nuclear, was M2, and M3 of CDK5RAP3 with PAK4 not only promoted altered by coexpression with PAK4 and became more concen- the PAK4 activity measured by a specific GST-PAK4tide sub- trated at peripheral regions (Fig. 4D). strate but also promoted the autophosphorylation of PAK4 (Fig. 5C, lanes 2–4). In contrast, the 434-506 mutant (M4) of CDK5RAP3-regulated PAK4 kinase activity CDK5RAP3 showed no PAK4 activation, similar to the GST PAK4 has been shown to induce tumorigenesis and cell control (Fig. 5C, lane 5). This result indicated that the aa adhesion by phosphorylating its substrates (15), and, therefore, sequence 255–436 of CDK5RAP3 contains an activation we wondered whether CDK5RAP3 promotes HCC formation by domain of PAK4. Moreover, as revealed by kinase assay, regulating PAK4 activity. To evaluate this, we constructed a CDK5RAP3 was a good substrate for PAK4 (Fig. 5C). Using a panel of CDK5RAP3 truncated mutants (Fig. 5A) and assessed PAK4 peptide kinase assay, we further illustrated that the FL their interaction with PAK4 as well as their effect on PAK4 and M2 of CDK5RAP3 significantly enhanced the activity of activity. The results of the GST affinity pull-down assay PAK4, but not the 1-304 mutant (M1), M4, and GST control, in a revealed that the GST control, 434-506 and 1-304 mutants of dose-dependent manner (Fig. 5D). Together with the interac- CDK5RAP3, did not bind to PAK4. However, FL and 255-436 tion data, our results implied that the direct binding of mutant (M2) showed strong binding affinity to PAK4, whereas CDK5RAP3 to PAK4 elicited the activation of PAK4. Further- the 256-506 mutant (M3) displayed a weaker binding affinity to more, the result from Fig. 5B indicated that CDK5RAP3 may PAK4 (Fig. 5A). It is possible that the additional aa 437-506 may interact with the N-terminal domain, possibly the PBD, thus have attenuated the binding of M3 to PAK4. To further inves- leading to the activation of PAK4. tigate how CDK5RAP3 interacted with PAK4, we constructed 3 deletion mutants of PAK4: 2 of them (aa 1–201 and 1–325) Knockdown of PAK4 abolished the promotion of cell contained the N-terminal p21-binding domain (PBD) and one invasiveness in CDK5RAP3 overexpression cells contained the C-terminal kinase domain (aa 325–591). Using a To investigate whether the overexpression of CDK5RAP3 GST affinity pull-down assay, we found that both the FL and M2 activated PAK4 in HCCs, the correlation between CDK5RAP3

2954 Cancer Res; 71(8) April 15, 2011 Cancer Research

CDK5RAP3 Activates PAK4 in HCC

A BCIP 100 Figure 4. CDK5RAP3 interacted GFP-PAK1 GFP-PAK4 Vector with PAK4. A, CDK5RAP3 and Mock ) 10 PAK4 transcripts were measured 10 α-GFP Input in 63 paired HCC samples by Input No Ab control α -PAK4 α-Myc qPCR. The line from the origin kDa 1 α-GFP IP: α-Myc 85 indicates the positive correlation; (log PAK4 WB: α-PAK4 P < 0.001 (Pearson correlation). B, HEK293T cells were 0.1 cotransfected with GFP-PAK4 0.1 1 10 100 and Myc-CDK5RAP3 in the CDK5RAP3 (log10) 50 coimmunoprecipitation assay. GFP-PAK4 Myc -CDK5RAP3 D 85 GFP-PAK1 was included as a WB: α-CDK5RAP3 control for specific interaction. C, endogenous CDK5RAP3 and PAK4 in HepG2 cells were WB: lgG 50 coimmunoprecipitated with the indicated antibodies. D, confocal microscopy with Myc-CDK5RAP3 and GFP-PAK4 cotransfected in HepG2 cells. Myc-CDK5RAP3 GFP-PAK4 Myc -CDK5RAP3 Overlay was visualized by anti-Myc antibody (Ab) and secondary antibody conjugated with Texas Red. IP, immunoprecipitation; WB, Western blotting. Cotransfection Single transfection

and p-PAK4 expression levels in the same set of tissue The result showed that loss of PAK4, as compared with the microarray was assessed by immunohistochemical staining. vector control cells, significantly reduced the invasiveness of Phosphorylated PAK4 showed strong nuclear staining, and the both CDK5RAP3 stably expressing clones (Fig. 6C) and, thus, staining was higher in the primary HCCs than in the corre- strongly suggests that CDK5RAP3 enhanced HCC-cell inva- sponding nontumorous liver tissues (10/20 cases); however, siveness through activation by PAK4. the expression levels of CDK5RAP3 and p-PAK4 positively correlated with one another (Fig. 1C; P ¼ 0.024, Pearson Discussion correlation). In addition, 3 of 5 cases with higher CDK5RAP3 expression in the metastases than in their primary HCCs Several studies have investigated the roles of CDK5RAP3 in showed a higher level of p-PAK4 in metastases than in primary carcinogenesis by using various cancer models; however, so HCCs (Fig. 1C), suggesting a positive correlation of CDK5RAP3 far, its definite roles remain contradictory. CDK5RAP3 has and PAK4 activity during HCC progression. In addition, PAK4 been found to promote apoptosis induced by genotoxic stress activity, as revealed by p-PAK4 staining, was quantitatively in HeLa cells by triggering G2/M arrest (3). In addition, reduced in the tumor tissues obtained from the CDK5RAP3- CDK5RAP3 has been proposed to be a tumor suppressor knockdown, nude mouse xenograft (Supplementary Fig. S3). because CDK5RAP3 inhibits the NF-kB cell-survival pathway To further confirm the specific activation of PAK4 by and its protein level is significantly reduced in head and neck CDK5RAP3 within cells, PAK4, in combination with increasing squamous cell carcinomas (5). However, a similar NF-kB– doses of CDK5RAP3, was cotransfected into HepG2 cells and suppressive effect of CDK5RAP3 was not observed in our study p-PAK4 was determined by immunoblotting. As shown in using HCC cells, and this could probably be attributed to some Fig. 6A, increased expression of CDK5RAP3 promoted the total tissue-specific effects. In contrast, overexpression of the protein and phosphorylation levels of PAK4 in HepG2 cells in a CDK5RAP3 isoform has been reported to promote cell pro- dose-dependent manner. The upregulation of phosphorylated liferation of HCC cells (7). Recently, CDK5RAP3 has been PAK4 was functionally active, as indicated by the increase in shown to be overexpressed in lung adenocarcinoma (16). In PAK4 activity in CDK5RAP3 cotransfected cells (Fig. 6A). the present study, using qPCR and immunohistochemical Consistently, the phosphorylation of PAK4 by CDK5RAP3 staining assays, we showed that both transcripts and protein was observed in CDK5RAP3 stably overexpressing HepG2 cells of CDK5RAP3 were frequently and significantly overexpressed (Fig. 6B). To examine whether CDK5RAP3 promoted HCC-cell in human HCCs (Fig. 1). The mechanism of such overexpres- invasiveness through activation of PAK4, we used siRNA to sion is currently unclear, but it has been reported that the specifically knock down PAK4 in CDK5RAP3 stably overex- chromosomal region 17q, which contains the CDK5RAP3,is pressing HepG2 cells and conducted the cell-invasion assay. frequently amplified in HCCs (8). Although CDK5RAP3 has

www.aacrjournals.org Cancer Res; 71(8) April 15, 2011 2955

Mak et al.

ABaa NH2 357 385 COOH aa 11 47 326 572 Leucine PAK4 (1-591) Kinase domain 591 FL: CDK5RAP3 (1–506) zipper 506 M1: CDK5RAP3 (1–304) PAK4 (1-325) PAK4 (1-201) M1: CDK5RAP3 (255–436) PAK4-KD (325-591) M1: CDK5RAP3 (256–506) Kinase domain M1: CDK5RAP3 (434–506) GST-CDK5RAP3 Input GST FL M2 Pull down

GST-CDK5RAP3 His-PAK4 1–591 1–325 1–301 KD 1–591 1–325 1–301 KD 1–591 1–325 1–301 KD 1–591 1–325 1–301 KD (kDa) WB: Anti-His His-PAK4 GST FL M1 M2 M3 M4 100 His-PAK4 (1–591) 71 WB: Anti-His His-PAK4 50 His-PAK4 (1–325) WB: Anti-GST 35 (KDa) GST-FL His-PAK4 (1–201) 100 71 GST-M1 GST-M3 WB: Anti-GST GST-M2 (kDa) 100 GST-FL 50 71 * 50 GST-M2

GST-M4 35 35 GST GST

Lane 1 2345 CD300,000 GST-M2 His-PAK4 +++++ +++++ GST-PAK4tide +++++ +++++ 280,000 GST-FL GST + ––– – + ––– – 260,000 GST-CDK5RAP3 – FL M2 M3 M4 – FL M2 M3 M4 240,000 (kDa)100 GST-FL 220,000 71 His-PAK4 GST-M3 200,000 GST-M2 50 180,000 GST-M1

35 Count per min GST-M4 160,000 GST-PAK4tide 140,000 GST GST 120,000 GST-M4 Autoradiograph Coomassie stained His-PAK4 index 1 1.4 1.3 1.1 0.9 100,000 210 43 PAK4 substrate index 1 1.2 1.2 1.1 0.9 CDK5RAP3 (μg)

Figure 5. CDK5RAP3-regulated PAK4 kinase activity. A (top), schematic diagram of CDK5RAP3 deletion mutants. The numbers represent the corresponding amino acid residues of the sequence; bottom, result of GST pull-down assay. B (top), schematic diagram of PAK4 deletion mutants. PBD, p21-binding domain; KD, kinase domain; bottom, CDK5RAP3 FL and 255-436 mutant (M2) were mixed with PAK4 FL 1-591, 1-201, and 1-325 mutants for GST pull-down assay. C, CDK5RAP3 FL and mutants were incubated with His-PAK4, and in vitro gel kinase assay was conducted using GST-PAK4tide as substrate; GST was included as a negative control; index of relative band intensity is shown. D, different amounts of GST-CDK5RAP3 proteins were incubated with His-PAK4, and peptide kinase assay was conducted using PAK4tide as a substrate; the incorporation rate of radioactivity (counts per minute) was plotted against the amount of GST-CDK5RAP3. WB, Western blotting.

been hypothesized to possess tumor-suppressor activity, sev- overexpressing HCC cells showed an increase in invasion rate eral lines of evidence here indicate that overexpression of (Fig. 3C) whereas knockdown of CDK5RAP3 suppressed the CDK5RAP3 is causally associated with HCC tumorigenicity. invasion (Fig. 3B). Taken together, these data strongly sug- We have used 3 different HCC cell lines, specifically PLC/PRF/ gested that CDK5RAP3 is involved in cancer metastasis. 5, SMMC-7721, and HepG2, to show the transforming ability of Although a previous report has shown that overexpression CDK5RAP3. Results from both gain-of-function and loss-of- of CDK5RAP3 suppressed cancer invasion in head and neck function approaches concurred to indicate that CDK5RAP3 cancers (5), our results have suggested that loss of CDK5RAP3 enhanced cell-proliferation rate, anchorage-independent suppressed HCC-cell invasion. The reason for the discrepancy growth, cell migration, and cell invasiveness, suggesting that remains unclear, but the possibility that the effect of CDK5RAP3 plays an oncogenic role in hepatocarcinogenesis. CDK5RAP3 is tissue specific cannot be completely ruled out. In our clinicopathologic analysis, overexpression of The mechanism by which CDK5RAP3 enhances cancer CDK5RAP3 was significantly correlated with more aggressive metastasis is not completely understood, but one of the tumor behavior in terms of the presence of tumor micro- possible mechanisms is through the activation of PAK4. satellite formation and poorer HCC differentiation (Table 1). Previous studies have shown that the constitutively active Consistently, our immunohistochemical staining in clinical form of PAK4 induces anchorage-independent growth, cell- HCCs indicated that CDK5RAP3 is highly expressed in meta- rounding phenotype, and defect in cell spreading onto static HCC cells (Fig. 1C). Furthermore, CDK5RAP3 stably a fibronectin-coated surface, all of which are related to

2956 Cancer Res; 71(8) April 15, 2011 Cancer Research

CDK5RAP3 Activates PAK4 in HCC

GFP-PAK4 Vector CDK5RAP3#1 CDK5RAP3#1 Figure 6. CDK5RAP3 promoted A 1μg 2μg C HCC-cell invasiveness mediated Vector Myc-CDK5RAP3 through activation by PAK4 kinase α-CDK5RAP3 activity. A, increasing amounts of α-p-PAK4 (ser474) Mock siControl siPAK4 Mock siControl siPAK4 Mock siControl siPAK4 CDK5RAP3 were transiently PAK4 α-GFP myc-CDK5RAP3 cotransfected with PAK4 in Endogenous CDK5RAP3 α-β-Actin HepG2 cells; expression of GFP- β-Actin PAK4 and Myc-CDK5RAP3 and phosphorylation of PAK4 were 30,000 Negative substrate 180 siControl determined by immunoblotting; PAK4tide 25,000 160 siPAK4 bottom, graph showing PAK4 20,000 140 kinase activity in cell lysates. B, 120 cell lysates from vector and 15,000 100 CDK5RAP3 stably expressing 10,000 80 HepG2 cells (CDK5RAP3#1 and Counts per min 5,000 60 CDK5RAP3#2) were analyzed by 0 Western blotting. C, CDK5RAP3 40 GFP-PAK4 – + + + 20 stably overexpressing HepG2 Number of invaded cells μ μ cells were transfected with Myc-CDK5RAP3 –– 1 g 2 g 0 #1 or 3 3#2 siPAK4, as indicated for invasion ect P P V A assay; top, Western blotting; #1 R B 3 K5 bottom, graph showing number AP r R CD CDK5RA 5RAP3#2 of invaded cells after transfection K5 K of PAK4 and control siRNA. VectoCD CD P ¼ t CDK5RAP3 * 0.005 (Student test); p-PAK4 (ser474) representative photomicrographs PAK4 are shown. β-Actin

cancer-cell migration (17, 18). With regard to the upstream requires further investigation. However, it is worth noting signaling, PAK4 is found to be activated by hepatocyte and that CDK5RAP3 is a good substrate of PAK4 (Fig. 5C); the keratinocyte growth factors during cell migration (19, 20). Our possibility that PAK4 can modulate the phosphorylation, and, data indicate that overexpression of CDK5RAP3 in HCC cells therefore, the translocation and function of CDK5RAP3, is was significantly associated with enhanced expression and currently under investigation. activity of PAK4 through Ser474 phosphorylation (Figs. 1C, 4A, With our in vitro kinase assay, we showed that CDK5RAP3 6A and B). With regard to the mechanism by which the significantly and directly promoted the autophosphorylation overexpression of CDK5RAP3 could increase the expression and activity of PAK4 (Fig. 5C). In addition, we showed that of PAK4 in HCCs, we hypothesized that CDK5RAP3 might increased expression of CDK5RAP3 remarkably enhanced the stabilize PAK4 through inhibition of PAK4 ubiquitination. phosphorylation and activity of PAK4, both in CDK5RAP3 Furthermore, recent reports have shown that CDK5RAP3 transiently transfected cells and in stably expressing HCC cells mutually stabilizes its binding partner, KIAA0776, by inhibit- (Fig. 6A and B), thus strongly indicating that CDK5RAP3 is a ing the ubiquitination of KIAA0776 (21). Thus, it is possible cellular activator of PAK4. With regard to the mechanism by that CDK5RAP3 not only activates PAK4 but also stabilizes which CDK5RAP3 activates PAK4, we mapped the central PAK4, making it a potent regulator of PAK4 activity. region of CDK5RAP3, which contains a leucine zipper domain, We have shown that CDK5RAP3 is a PAK4 cellular binding required for the activation of PAK4. In our GST affinity pull- partner, using coimmunoprecipitation, coimmunostaining, down experiment, we found that CDK5RAP3 interacted with and GST affinity pull-down assays. Although we have not the N-terminal region of PAK4, which contains the PBD, and examined the interaction of CDK5RAP3 with all other mem- Figure 5A shows that the CDK5RAP3 mutant (aa 255–436) is, bers of the PAK family, the coimmunoprecipitation data indeed, the activator for PAK4, as compared with mutants that clearly show that CDK5RAP3 only binds to PAK4 but not showed lower or absent binding ability. This result led us to to PAK1, indicating that the binding of CDK5RAP3 is relatively speculate that, similar to small Rho GTPase, Cdc42, and Rac1, specific for PAK4. In addition, the analysis in human HCCs the CDK5RAP3 activates PAK4 through direct protein–protein showed no significant correlation of CDK5RAP3 and PAK1 interaction for which PBD is a requisite factor. overexpression. Interestingly, our confocal immunofluores- PAK4 has been shown to play important roles in cell cence staining data revealed that ectopic expression of migration and adhesion by regulating the actin-cytoskeleton CDK5RAP3 with PAK4 altered the localization of CDK5RAP3, organization and integrin signaling (22). Furthermore, studies which was mainly nuclear and cytoplasmic, to be concen- have shown that PAK4 is an important mediator for hepato- trated at the membrane peripheral (Fig. 4D). Whether the cyte growth factor–stimulated cell migration in epithelial peripheral localization of CDK5RAP3 plays a role in the (20) and cancer cells (23). Although the role of PAK4 in activation of PAK4 and the regulation of cell migration HCC has not been characterized, a recent report has shown

www.aacrjournals.org Cancer Res; 71(8) April 15, 2011 2957

Mak et al.

that overexpression of PAK4 in prostate cancer promotes cell Disclosure of Potential Conflicts of Interest migration and invasion (15). Thus, we hypothesized that the activation of PAK4 by CDK5RAP3 may promote HCC-cell No potential conflicts of interest were disclosed. invasion. For this purpose, we used siRNA to specifically knock down PAK4 in CDK5RAP3 stably overexpressing HepG2 Acknowledgments

cells and showed that the loss of PAK4 attenuated the inva- We thank Dr. Honglin Li for providing the plasmids and Dr. Abel Chun for siveness of stable clones back to similar level as the vector critical reading of manuscript. Irene O.L. Ng is the Loke Yew Professor in control, indicating that PAK4 plays a key role in CDK5RAP3- Pathology. mediated HCC cancer metastasis (Fig. 6C). Our findings have established that upregulation of CDK5RAP3 may occur in HCC Grant Support progression and metastasis through the regulation of PAK4. In summary, we provide the first evidence that CDK5RAP3 The Hong Kong Research Grant Council (N_HKU715/08, HKU 1/06C and 7/CRF/09) and The University of Hong Kong, Seed Funding Pro- is overexpressed in human HCCs and that overexpression of gramme (200711159100; to Y.P. Ching), supported this research study. CDK5RAP3 promotes metastasis of HCC. We also provide The costs of publication of this article were defrayed in part by the payment of page charges. This article must, therefore, be hereby marked evidence that supports a single mechanism wherein advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this CDK5RAP3 enhances HCC metastasis through the activation fact. of PAK4. Therefore, inhibition of CDK5RAP3 can potentially be used to suppress HCC formation, providing a new molecular Received November 11, 2010; revised February 1, 2011; accepted February 22, target for therapeutic intervention in HCC. 2011; published OnlineFirst March 8, 2011.

References 1. Ching YP, Qi Z, Wang JH. Cloning of three novel neuronal Cdk5 12. Ching YP, Leong VYL, Wong CM, Kung HF. Identification of an activator binding proteins. Gene 2000;242:285–94. autoinhibitory domain of p21-activated protein kinase 5. J Biol 2. Yin X, Warner DR, Roberts EA, Pisano MM, Greene RM. Novel Chem2003;278:33621–4. interaction between nuclear co-activator CBP and the CDK5 activator 13. Rennefahrt UE, Deacon SW, Parker SA, Devarajan K, Beeser A, binding protein—C53. Int J Mol Med 2005;16:251–6. Chernoff J, et al. Specificity profiling of Pak kinases allows identifica- 3. Jiang H, Luo S, Li H. Cdk5 activator-binding protein C53 regulates tion of novel phosphorylation sites. J Biol Chem 2007;282:15667–78. apoptosis induced by genotoxic stress via modulating the G2/M DNA 14. Lew J, Beaudette K, Litwin CM, Wang JH. Purification and character- damage checkpoint. J Biol Chem 2005;280:20651–9. ization of a novel proline-directed protein kinase from bovine brain. 4. Wang J, He X, Luo Y, Yarbrough WG. A novel ARF-binding protein J Biol Chem 1992;267:13383–90. (LZAP) alters ARF regulation of HDM2. Biochem J 2006;393:489– 15. Wells CM, Whale AD, Parsons M, Masters JR, Jones GE. PAK4: a 501. pluripotent kinase that regulates prostate cancer cell adhesion. J Cell 5. Wang J, An H, Mayo MW, Baldwin AS, Yarbrough WG. LZAP, a Sci 2010;123:1663–73. putative tumor suppressor, selectively inhibits NF-kappaB. Cancer 16. Stav D, Bar I, Sandbank J. Usefulness of CDK5RAP3, CCNB2, and Cell 2007;12:239–51. RAGE genes for the diagnosis of lung adenocarcinoma. Int J Biol 6. Chen J, Liu B, Liu Y, Han Y, Yu H, Zhang Y, et al. A novel gene IC53 Markers 2007;22:108–13. stimulates ECV304 cell proliferation and is upregulated in failing heart. 17. Callow MG, Clairvoyant F, Zhu S, Schryver B, Whyte DB, Bischoff JR, Biochem Biophys Res Commun 2002;294:161–6. et al. Requirement for PAK4 in the anchorage-independent growth of 7. Xie YH, He XH, Tang YT, Li JJ, Pan ZM, Qin WX, et al. Cloning and human cancer cell lines. J Biol Chem 2002;277:550–8. characterization of human IC53-2, a novel CDK5 activator binding 18. Dan C, Kelly A, Bernard O, Minden A. Cytoskeletal changes regulated protein. Cell Res 2003;13:83–91. by the PAK4 serine/threonine kinase are mediated by LIM kinase 1 and 8. Raidl M, Pirker C, Schulte-Hermann R, Aubele M, Kandioler-Eckers- cofilin. J Biol Chem 2001;276:32115–21. berger D, Wrba F, et al. Multiple chromosomal abnormalities in human 19. Lu Y, Pan ZZ, Devaux Y, Ray P. p21-activated protein kinase 4 (PAK4) liver (pre)neoplasia. J Hepatol 2004;40:660–8. interacts with the keratinocyte growth factor receptor and participates 9. Ching YP, Leong VY, Lee MF, Xu HT, Jin DY, Ng IO. p21-activated in keratinocyte growth factor-mediated inhibition of oxidant-induced protein kinase is overexpressed in hepatocellular carcinoma and cell death. J Biol Chem 2003;278:10374–80.

enhances cancer metastasis involving c-Jun NH2-terminal kinase 20. Wells CM, Abo A, Ridley AJ. PAK4 is activated via PI3K in HGF- activation and paxillin phosphorylation. Cancer Res 2007;67:3601– stimulated epithelial cells. J Cell Sci 2002;115:3947–56. 8. 21. Kwon J, Cho HJ, Han SH, No JG, Kwon JY, Kim H. A novel LZAP- 10. Leung TH, Ching YP, Yam JW, Wong CM, Yau TO, Jin DY, et al. binding protein, NLBP, inhibits cell invasion. J Biol Chem 2010;285: Deleted in liver cancer 2 (DLC2) suppresses cell transformation by 12232–40. means of inhibition of RhoA activity. Proc Natl Acad Sci U S A 22. Jaffer ZM, Chernoff J. p21-activated kinases: three more join the Pak. 2005;102:15207–12. Int J Biochem Cell Biol 2002;34:713–7. 11. Wong CC, Wong CM, Tung EK, Man K, Ng IO. Rho-kinase 2 is 23. Paliouras GN, Naujokas MA, Park M. Pak4, a novel Gab1 binding frequently overexpressed in hepatocellular carcinoma and involved partner, modulates cell migration and invasion by the Met receptor. in tumor invasion. Hepatology 2009;49:1583–94. Mol Cell Biol 2009;29:3018–32.

2958 Cancer Res; 71(8) April 15, 2011 Cancer Research

Overexpression of a Novel Activator of PAK4, the CDK5 Kinase−Associated Protein CDK5RAP3, Promotes Hepatocellular Carcinoma Metastasis

Grace Wing-Yan Mak, Mandy Man-Lok Chan, Veronica Yee-Law Leong, et al.

Cancer Res 2011;71:2949-2958. Published OnlineFirst March 8, 2011.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-10-4046

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2011/03/07/0008-5472.CAN-10-4046.DC1

Cited articles This article cites 22 articles, 12 of which you can access for free at: http://cancerres.aacrjournals.org/content/71/8/2949.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/71/8/2949.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/71/8/2949. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.