Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Loss of p53 promotes RhoA–ROCK-dependent cell migration and invasion in 3D matrices

Loss of p53 promotes RhoA–ROCK-dependent cell migration and invasion in 3D matrices JCB: REPORT Loss of p53 promotes RhoA–ROCK-dependent cell migration and invasion in 3D matrices Gilles Gadea, Marion de Toledo, Christelle Anguille, and Pierre Roux Centre de Recherche en Biochimie Macromoléculaire, Centre National de la Recherche Scientifi que, Universite Mixte de Recherche 5237, Institut Federatif de Recherche 122, F-34293 Montpellier, Cedex 5, France n addition to its role in controlling cell cycle progression, movement and have considerably increased invasive prop- the tumor suppressor protein p53 can also affect other erties. The morphological transition requires the RhoA– I cellular functions such as cell migration. In this study, we ROCK (Rho-associated coil-containing protein kinase) show that p53 defi ciency in mouse embryonic fi broblasts pathway and is prevented by RhoE. A similar p53-mediated cultured in three-dimensional matrices induces a switch from transition is observed in melanoma A375P cancer cells. an elongated spindle morphology to a markedly spherical Our data suggest that genetic alterations of p53 in tumors and fl exible one associated with highly dynamic membrane are suffi cient to promote motility and invasion, thereby con- blebs. These rounded, motile cells exhibit amoeboid-like tributing to metastasis. Introduction and its main effector, Rho-associated coil-containing protein A critical event during tumorigenesis is the conversion from a kinase (ROCK), and resembles the amoeboid movement. This static primary tumor to an invasive, disseminating metastasis. involves a rounded bleb-associated movement that generates Moreover, tumor cells show an increased capacity to migrate. propulsive motion through the matrix independently of proteol- Numerous intracellular signaling molecules have been impli- ysis (Sahai and Marshall, 2003). cated in migratory processes. Among them, the Rho GTPase Curiously, genes encoding Rho GTPases are rarely found family plays a pivotal role in regulating the biochemical and cyto- mutated in human cancers, in which only their functional ac- skeletal pathways relevant to cell migration. The Rho GTPases tivities seem to be deregulated (Nakamoto et al., 2001; Rihet Rac, Cdc42, and Rho control cell protrusions during migration. et al., 2001). This suggests that alterations in others genes Aberrant regulation of Rho proteins is believed to associate with account for functional modifications of Rho GTPases ac- metastasis by promoting cell motility (Clark et al., 2000; Evers companying actin cytoskeleton remodeling during metastasis. et al., 2000; Jaffe and Hall, 2002; Steeg, 2003). We hypothesize that this set of genes controls the cell cycle. The bona fi de environment for migrating cells is the extra- Their mutation during the initiation phase in cancer would cellular matrix, which permits movement in a 3D scaffold that modify the behavior of proteins involved in actin cytoskeleton is biochemically complex and shows dynamic fl exibility. 3D dynamics, such as Rho GTPases, leading to a migratory and tissue culture models reconstitute an environment that resem- invasive phenotype. bles the in vivo situation in regard to cell shape and movement. Beyond these genes is the tumor suppressor p53, whose This model provides important insights into the mechanisms of mutations occurs in >50% of human tumors (Hollstein et al., cell motion during carcinogenesis. Recent advances have iden- 1994). p53 protects cells from malignant transformation by tifi ed two modes of cell motility in 3D matrices. The elongated regulating cell cycle arrest or by promoting apoptosis (Levine, mode of migration is a consequence of Rac activity and gener- 1997; Giaccia and Kastan, 1998). We and others have recently ates membrane protrusions, the lamellipodia that drive motility. shown that p53 modulates cell migration: p53 negatively In contrast, a novel rounded mode of motility depends on RhoA modulates Rho GTPases and regulates cell polarization and migration (Gadea et al., 2002, 2004; Guo et al., 2003). How- Correspondence to Pierre Roux: [email protected] ever, little is known about the role of p53 in cells moving in G. Gadea’s present address is Institute of Cancer Research, London SW3 6JB, UK. a 3D matrix that mimics the in vivo microenvironment of Abbreviations used in this study: DIC, differential interference contrast; F-actin, tumor cells. fi lamentous actin; MEF, mouse embryonic fi broblast; ROCK, Rho-associated coil- Identifying the mechanisms by which p53 modulates cell containing protein kinase; wt, wild type. The online version of this article contains supplemental material. migration is important to understand how invasive cells arise. © The Rockefeller University Press $15.00 The Journal of Cell Biology, Vol. 178, No. 1, July 2, 2007 23–30 http://www.jcb.org/cgi/doi/10.1083/jcb.200701120 JCB 23 THE JOURNAL OF CELL BIOLOGY In this study, we show that the elongated spindle-shaped fi bro- at http://www.jcb.org/cgi/content/full/jcb.200701120/DC1). blastoid mode of motility can be converted to a rounded bleb- wt MEF maintained a spindle-shaped elongated morphology, −1 bing movement by blocking p53 function. This amoeboid mode generated extensions, and moved slowly at 2 ± 1 μm/h , −/− of motility requires RhoA–ROCK signaling and confers higher whereas p53 MEF moved as rounded cells using a transla- velocity and invasive properties to p53-defi cient cells. Thus, the tory motion in the direction of the serum growth factor gradient. range of p53 tumor suppressor activity extends to the control of They show many dynamic bleblike structures, alternating rapid the mode of migration of invasive cells. cycles of squeezing and expansion, and considerable deform- −/− −1 ability. p53 cells moved at 12 ± 5 μm/h , maintaining their rounded morphology. Expression of the GFP-tagged mu- Results and discussion tant p53 H273, a dominant-negative form of p53, converts −/− p53 fi broblasts show rounded the elongated morphology of MEF to a marked spherical and blebbing movement fl exible one associated with highly dynamic membrane blebs −/− Initially, we compared the movement of p53-defi cient primary similar to those observed in p53 MEFs (Fig. 1 B and Video 3). −/− −/− mouse embryonic fi broblasts (MEFs; p53 MEFs) with that The expression of GFP-tagged wt p53 in p53 MEFs inhibits of wild-type (wt) MEFs cultured in 3D Matrigel matrix using dynamic blebbing and leads to an elongated morphology (Fig. time-lapse video microscopy. The migration of the two cell types 1 B and Video 4), whereas that of GFP-tagged p53 H273 has no is strikingly different (Fig. 1 A and Videos 1 and 2, available effect. Thus, the loss of wt p53 activity is suffi cient to drive Figure 1. Rounded blebbing movement is provoked by p53 defi ciency. (A) Time-lapse analysis of MEFs −/− and p53 MEFs moving within Matrigel. Still DIC time-lapse images from accompanying videos of −/− wt MEFs (Video 1) and p53 MEFs (Video 2, avail- able at http://www.jcb.org/cgi/content/full/jcb .200701120/DC1). Cells were plated into a thick layer of Matrigel, and a gradient of serum growth fac- tors was established. Frames show the cell migration at the indicated times. Arrows indicate the direction of cell movement. Dotted circles indicate the initial situa- tion of the cell when the video starts. (B) Time-lapse analysis of the different modes of cell motility within Matrigel. DIC images of cells transfected with GFP- tagged plasmids (to enable the identifi cation of trans- fected cells as visualized in green) expressing either p53 H273 in MEFs (still images from Video 3), p53 wt −/− in p53 MEFs (still images from Video 4), or p53 −/− H273 in p53 MEFs as control (right). Cells were plated within Matrigel as in A. (C) p53 regulates sub- cellular distributions of integrin β1 and ezrin in rounded migrating cells. Representative integrin β1 and ezrin −/− fl uorescent staining of MEFs and p53 MEFs trans- fected with p53 wt, p53 H273, siRNA p53, or plas- mid EGFP. Insets show the covisualization of F-actin (red), GFP (green), and ezrin or integrin β1 (blue) as indicated. Cells were plated in Matrigel, and a gradient of serum growth factors was established. Arrows show the direction of cell movement in Matrigel. Bars, 10 μm. 24 JCB • VOLUME 178 • NUMBER 1 • 2007 the switch from an elongated fi broblast-like to a rounded type in the newly formed long extensions (Fig. 1 C). Thus, p53 con- of migration. trols the clustering of integrin β1 and ezrin in rounded cells. In Matrigel, cells moving with a rounded or elongated mor- Culture of fi broblastic cells in monolayers imposes an arti- phology show a distinct subcellular distribution of integrin β1 fi cial polarity between the lower and upper surface of these apo- and ezrin that helps in establishing the direction of movement lar cells. Their morphology and migration change in a 3D matrix, (Sahai and Marshall, 2003). In Matrigel, rounded blebbing suggesting that the spatial constraints and surrounding biochem- MEFs that arise from p53 inactivation show integrin β1 and ezrin ical microenvironment are a better approximation of the in vivo relocalization to the moving front of the cell. A similar clustering situation (Friedl and Brocker, 2000; Cukierman et al., 2001). −/− was also observed in rounded p53 MEFs. In contrast, conver- This is particularly striking for p53-defi cient MEFs that become −/− sion of p53 MEFs to elongated morphology by the expression blebbing spherical cells once suspended in 3D matrices, losing of p53 wt led to their relocalization throughout the cell, notably their initial fi broblastoid morphology observed in 2D culture. −/− Figure 2. p53 regulates RhoA activation. (A) GTPase activities in wt MEFs and p53 MEFs. Cells were lysed 24 h after plating, and the levels of GTP- −/− bound RhoA, Rac1, and Cdc42 were assayed. (B) RhoA activity in transfected MEFs and p53 MEFs. Cells transfected with p53 wt and p53 H273 were assayed for RhoA activity. (A and B) The values are the mean ± SD (error bars) of three independent experiments. (C–F) Representative F-actin and RhoA −/− fl uorescent staining of wt MEFs and p53 MEFs untreated (C), treated with TAT-C3 (F), or transfected with plasmids expressing either p53 wt (D) or p53 H273 (E). Panels show phalloidin-stained F-actin (red) and RhoA labeling (green). Arrows point to blebbing structures. Bars, 10 μm. P53 DRIVES THE MODE OF CELLULAR MIGRATION • GADEA ET AL. 25 −/− −/− RhoA activity is enhanced in p53 MEFs actin (F-actin) in p53 and wt MEFs. In wt MEFs, RhoA −/− This rounded blebbing movement observed in p53 MEFs is exhibits a punctuate distribution throughout the cytoplasm with strikingly similar to amoeboid-like motility, which is dependent a marked concentration in the perinuclear region. F-actin stress on Rho–ROCK signaling in a 3D matrix (Sahai and Marshall, fi bers cross the cytoplasm and accumulate slightly around the −/− −/− 2003). We compared the level of GTP-bound RhoA in p53 edge of the cell (Fig. 2 C, top). In p53 MEFs, RhoA shows −/− and wt MEFs. GTP-RhoA was increased 3.4-fold in p53 its punctuate cytoplasmic localization but is excluded from the MEFs, whereas GTP-Rac and Cdc42 were barely affected (Fig. perinuclear region (Fig. 2 C, bottom). Instead, RhoA colocal- 2 A). This change was functional because protein expression izes with large patches of polymerized actin in bleblike globular levels were unaltered by p53 deletion. Reintroduction of wt p53 structures that are distributed over and protruding from the cell −/− in p53 MEFs lowered the level of GTP-RhoA to that of wt surface (Fig. 2 C, arrows). In addition, MEFs lacking p53 activ- MEFs (Fig. 2 B). Conversely, p53 H273 did not affect the over- ity show more peripheral microspikes, which were previously −/− activation of RhoA in p53 MEFs. Finally, blocking p53 in characterized as fi lopodia (Gadea et al., 2002). They also show MEFs using p53 H273 or siRNA-mediated knockdown of p53 an accumulation of peripheral polymerized actin bundles (i.e., a was suffi cient to increase GTP-RhoA (Fig. 2 B). These results redistribution of F-actin stress fi bers; Fig. 2 C). The expression show that wt p53 inhibits RhoA activation and has a role in reg- of wt p53 led to loss of the bleblike protrusions and recovery of ulating RhoA signaling pathways. the punctuate RhoA localization in the perinuclear region. This was not seen with p53 H273, indicating that wt p53 activity is −/− Distinct localization of RhoA in p53 MEFs required for RhoA localization (Fig. 2, compare D with E). −/− Translocation from the cytoplasm to the membrane area is essen- Treatment of p53 MEFs with TAT-C3, which inactivates tial for both the activation and function of RhoA (Takaishi et al., RhoA, RhoB, and RhoC (Coleman et al., 2001), abolished bleb- 1995). We compared the localization of RhoA and fi lamentous like structures and led to RhoA redistribution in the perinuclear Figure 3. ROCK activity is required for rounded blebbing morphology driven by p53 defi ciency. (A) Still DIC time-lapse images from Video 5 (available at http://www.jcb.org/cgi/content/ −/− full/jcb.200701120/DC1) of p53 MEFs treated with Y27632. Frames show the spread- −/− ing of a p53 MEF at the indicated times. (B) DIC images show wt MEFs cotransfected with GFP to identify transfected cells (insets) and ROCK-∆1 (still images from Video 6; Fig. S1) or with GFP alone as a control. (C) Represen- tative F-actin and RhoA fl uorescent staining of −/− wt MEFs and p53 MEFs treated either with Y27632 or H1152. (D) Time-lapse analysis of different modes of cell motility within Matrigel. Still DIC images of MEFs transfected either with ROCK-∆1 (still DIC images from Video 7) or GFP alone as a control. (bottom) Still DIC −/− images of p53 MEFs treated with Y27632 or H1152 and transfected with p53 H273 (from Video 8 and Video 9, respectively). The transfected cells expressing either ROCK-∆1 or p53 H273 are visualized using GFP. Cells were plated within Matrigel as in Fig. 1. Bars, 10 μm. 26 JCB • VOLUME 178 • NUMBER 1 • 2007 Invasion of p53-defi cient cells requires region (Fig. 2 F). Thus, RhoA activation is required for bleb −/− ROCK activity and is prevented by RhoE formation in p53 cells. RhoA–ROCK signaling also plays a key role in invasion by p160 ROCK controls morphological blebbing cells moving in 3D matrix (Sahai and Marshall, 2003). −/− −/− MEFs, which have an overactivation of this changes observed in p53 fi broblasts Similarly, p53 RhoA-driven actin reorganization largely depends on the serine– pathway, showed substantially higher invasiveness than wt threonine protein kinase ROCK (Amano et al., 1997; Ishizaki MEFs in Matrigel (Fig. 4 A). Inhibition of Rho or ROCK by et al., 1997). To determine whether ROCK mediates mem- −/− brane blebbing in p53 MEFs cultured in monolayer, we used Y27632 and H1152, which are two distinct, structurally un- −/− related pharmacological inhibitors of ROCK. p53 MEFs promptly lost their extensive dynamic bleblike activities and their rounded morphology within 40 min, adopting a fl attened, spread shape when treated with Y27632 (Fig. 3 A and Video 5, avail- able at http://www.jcb.org/cgi/content/full/jcb.200701120/DC1) or H1152 (not depicted). Consistent with this, the expression of ROCK-∆1, an active mutant, in wt MEF promoted a rounded morphology accompanied by numerous dynamic bleblike ex- −/− tensions similar to p53 MEF (Fig. 3 B and Video 6). GFP −/− alone had no effect (Fig. 3 B). The spreading of p53 MEFs induced by Y27632 or H1152 was accompanied by the reorga- nization of F-actin from cortical bundles into scattered cyto- plasmic dotlike structures concomitant with the loss of fi lopo dia (Fig. 3 C). Similar to other studies (Ishizaki et al., 2000; Totsukawa et al., 2000), ROCK inhibitors strongly reduced −/− the number of stress fi bers in both wt and p53 MEFs. In −/− addition, ROCK inhibition in p53 MEFs led to RhoA relocal- ization to the perinuclear region (compare Fig. 3 C with Fig. 2 C). Thus, ROCK inhibition causes RhoA to lose its mem- brane localization. We tested whether ROCK activity is required for the −/− rounded blebbing movement of p53 MEFs cultured in Matrigel. Transfection of ROCK-∆1 converts MEFs from an elongated to a rounded blebbing morphology; GFP alone had no effect (Fig. 3 D and Video 7, available at http://www.jcb.org/cgi/content/ −/− full/jcb.200701120/DC1). In contrast, treatment of p53 MEFs with Y27632 or H1152 inhibits dynamic blebbing and leads to an elongated morphology even upon the expression of p53 H273 (Fig. 3 D and Videos 8 and 9). This transition is ac- companied by the redistribution of integrin β1 and ezrin to the moving front of the cell (Fig. S1). Thus, ROCK activity is required for the amoeboid-like rounded morphology driven by p53 defi ciency. Collectively, our results indicate that characteristics of the fi broblast type of motility (i.e., that using elongation and trac- tion) are lost when p53 function is abrogated using either p53 H273 or siRNA. Cells adopt a rounded bleb morphology similar Figure 4. p53 controls cell invasion through the RhoA–ROCK pathway. −/− −/− to that observed in p53 MEFs cultured in 3D matrices. A p53 (A) Matrigel invasion of wt MEFs and p53 MEFs. Cells were treated with either TAT-C3, Y27632, H1152, or protease inhibitors or were transfected defect leads to the aberrant overactivation of RhoA, which con- with plasmids encoding either p53 wt, p53 H273, siRNA p53, ROCK-∆1, sequently translocates to specifi c membrane blebbing structures. or RhoA-V14 as indicated. The changes in Matrigel invasion were ana- Importantly, we show that the RhoA–ROCK signaling pathway lyzed as described in Materials and methods. For each data point, GFP- positive cells were scored. (B) Representative RhoE and actin fl uorescent is involved in p53-dependent morphological changes by modu- −/− staining of p53 MEFs transfected with GFP-RhoE. (C) DIC images of lating round blebbing. It also appears to promote blebbing in −/− p53 MEFs transfected with GFP-tagged RhoE (still images from Video 10, −/− p53 cells. Notably, RhoA signaling through ROCK was pre- available at http://www.jcb.org/cgi/content/full/jcb.200701120/DC1). Cells were plated in Matrigel as in Fig. 1 A. (D) Matrigel invasiveness of viously shown to be important in the rounded blebbing move- −/− p53 MEFs transfected either with vector alone (control) or with RhoE. ment but not for the elongated protrusive movement of cancer Quantifi cation was performed as in A. (A and D) Values are the means ± SD cells cultured in 3D matrices (Sahai and Marshall, 2003). (error bars) of three independent experiments. Bars, 10 μm. P53 DRIVES THE MODE OF CELLULAR MIGRATION • GADEA ET AL. 27 treatment with TAT-C3, Y27632, or H1152 diminished the abil- (Fig. 4 D), suggesting that RhoE helps mediate the control of −/− ity of p53 MEFs to invade Matrigel by around 60%. Expres- cell morphology and invasiveness by p53. sion of wt p53 but not p53 H273 blocked this behavior. The ability of MEFs to invade Matrigel was increased p53 inactivation promotes rounded when p53 activity was blocked either by expression of the H273 blebbing movement in A375P melanoma p53 mutant or using siRNA to p53 in MEFs. Similarly, invasive- cancer cells ness was enhanced by an active mutant of RhoA (RhoA-V14) We sought to extend these observations to cancer cells. We chose and ROCK-∆1. Invasiveness of blebbing cells is independent of A375P melanoma cells, which have a wt form of p53 and show an protease activities (Sahai and Marshall, 2003); accordingly, the elongated mode of migration in 3D matrix (Sahai and Marshall, −/− invasive activity of p53 MEFs was unaffected by the inhibi- 2003). Cell morphology was visualized using fl uorescent wheat tion of proteases. Thus, the Rho–ROCK signaling module is germ agglutinin, which provides highly selective labeling of the −/− necessary for the invasive behavior of p53 cells in Matrigel. external plasma membrane. The expression of GFP-tagged p53 RhoE, a member of the Rho family that blocks actin stress H273 confers a rounded blebbing morphology to A375P cells, fi bers (Guasch et al., 1998; Nobes et al., 1998), is a transcrip- which associates with membrane blebs similar to those observed −/− tional target of p53 in response to genotoxic stress; RhoE pro- in p53 MEFs (Fig. 5 A). In contrast, the elongated morphology motes cell survival through the inhibition of ROCK1-mediated of A375P was unchanged when p53 accumulation was induced apoptosis (Ongusaha et al., 2006). The overexpression of RhoE by etoposide, which was verifi ed using the induction of p21WAF, clearly prevented actin polymerization and abolished the bleb- a transcriptional target of p53 (Fig. 5 B). These morphologic −/− like protrusions in p53 MEFs, converting them to an elon- changes correlated with the level of GTP-RhoA, which was gated morphology (Fig. 4, B and C; and Video 10, available at 4.6-fold higher upon the expression of p53 H273 (Fig. 5 C). Thus, http://www.jcb.org/cgi/content/full/jcb.200701120/DC1). RhoE the amoeboid-like motility of A375P melanoma cells induced by −/− also considerably reduced the invasiveness of p53 MEFs the loss of p53 activity is dependent on RhoA activation. Figure 5. A375 melanoma cell invasiveness is regulated by p53 activity. (A) Confocal images of A375P cells expressing plasmid EGFP alone or p53 H273. Cells were plated in Matrigel, and a gradient of serum growth factors was established to observe the mode of migration. The transfected cells are visualized using GFP, and plasma membranes were labeled with fl uorescent WGA (wheat germ agglutinin; red). (B) p21 induction in A375P cells. A375P cells were untreated or treated with etoposide. Total protein lysates were analyzed by SDS-PAGE followed by immunoblot analysis using antibodies to p21. Normalization was performed with an anti–α-tubulin antibody. (C) RhoA activity in transfected A375P. Cells expressing GFP-tagged p53 H273 or plasmid EGFP (control) were compared for GTP-RhoA levels as in Fig. 2. The image is representative of three independent experiments. The values are the mean ± SD (error bars). (D) Invasiveness in 3D matrigel matrix of A375P-expressing p53 wt or p53 H273 treated with TAT-C3, Y27632, or H1152 as indicated. Results are expressed as in Fig. 4. 28 JCB • VOLUME 178 • NUMBER 1 • 2007 Amoeboid cancer cell motility is particularly effi cient in were then fi lled with serum-free DME and DME with 10% FCS, respectively, thus establishing a soluble gradient of chemoattractant that permits cell in- promoting the invasiveness of metastatic cells in 3D environ- vasion throughout the Matrigel. Inhibitors were added immediately after ments. Therefore, we performed invasion assays in Matrigel us- cell plating at the aforementioned concentrations. Cells were allowed to in- −/− ing this model. The expression of p53 H273 strongly increased vade at 37°C and 5% CO through the gel for 36 h (for MEF and p53 MEF) or for 24 h (for A375P) before fi xing for 15 min in 3.7% formalde- the invasiveness of A375P, whereas wt p53 inhibited it (Fig. 5 D). hyde. Cells that had invaded through the Matrigel were detected on the TAT-C3, Y27632, or H1152 treatment largely reduced this re- lower side of the fi lter by GFP fl uorescence and counted for cell number. sponse in p53 H273–expressing A375P cells, arguing that mu- Six fi elds per fi lter were counted, and each assay was performed twice in triplicate for each cell line. tated p53 drives this behavior through RhoA–ROCK signaling. In contrast, protease inhibition did not alter p53 H273–driven Immunofl uorescence invasiveness (Fig. 5 D). This stresses the role of the amoeboid- Cells were transfected and/or treated with TAT-C3, Y27632, or H1152 on coverslips at a confl uence of 30% before fi xation in 3.7% formaldehyde like rounded morphology of p53-defi cient A375P cells in their in PBS followed by a 5-min permeabilization in 0.1% Triton X-100 in PBS invasive potential. and were incubated in PBS containing 0.1% BSA before staining with pri- Our results show that blocking p53 function is suffi cient to mary antibodies as follows: RhoA (Santa Cruz Biotechnology, Inc.), anti- ezrin polyclonal antibody (provided by P. Mangeat, Centre de Recherche en convert melanoma cells to a rounded locomotion, indicating that Biochimie Macromoléculaire, Montpellier, France), and integrin β1 (Santa cancer cells may also be subject to this regulation. This switch Cruz Biotechnology, Inc.). All of the antibodies were revealed with either would favor the illegitimate dissemination of cancer cells con- an AlexaFluor546- or -488–conjugated goat anti–mouse or anti–rabbit antibody (Invitrogen and Interchim). Cells were stained for F-actin using taining mutant p53, possibly increasing their invasiveness. TRITC-conjugated phalloidin (Sigma-Aldrich) or for membrane morphology The tumor suppressor p53 is mutated in >50% of human using cell-impermeant AlexaFluor594 wheat germ agglutinin, which binds cancers (Hollstein et al., 1994) and, thus, constitutes an appeal- selectively to N-acetylglucosamin and N-acetylneuraminic (sialic) acid resi- dues and provides highly selective labeling of the plasma membrane (Invit- ing target for anticancer therapy. Diverse therapeutic strategies rogen). Cells were prepared as described previously (Gadea et al., 2002). already exist that attempt to restore wt p53 function in cancer Stacks of 16-bit fl uorescent images (z step of 0.1 μm) were captured with cells, including the rescue of mutant p53 function and reactiva- a MetaMorph-driven microscope (DMRB; Leica) using a 63× NA 1.4 Apo- chromat oil immersion objective (Leica), a piezo stepper (E662; Physik In- tion of wt p53 (Blagosklonny, 2002). Given that there is no struments), and a camera (CoolSNAP HQ2; Photometrics). Epifl uorescence obvious selection for a metastatic phenotype, Bernards and of all images (in 2D and 3D) were restored and deconvolved with Huygens Weinberg (2002) proposed that classic oncogenes and tumor (Scientifi c Volume Imaging) using the maximum likelihood estimation algo- rithm. Restored stacks were processed with Imaris (Bitplane) for visualiza- suppressor genes implicated in the early stages of tumorigenesis tion. The restored images were saved as tiff fi les that were mounted using may also play a role in metastasis. Our studies indicate that p53 Photoshop and Illustrator (Adobe). is an ideal candidate, as it is involved both in the control of tumor cell apoptosis and tumor cell invasion. This extends the Immunofl uorescence in 3D matrices For immunofl uorescence of cells in 3D Matrigel network, cultures were applications for anticancer agents aimed at restoring p53 wt transfected and treated with TAT-C3, Y27632, or H1152 as monolayer fol- functions: such agents may combine antiproliferative and anti- lowed by trypsinization and embedding into Matrigel in transwell cell cul- invasive activities. ture chambers (as previously described in the Invasion assay section except that cells were allowed to invade through the gel for 6 h). Controls were left untreated. 50-μm cryosections were then cut at –20°C before fi xation in 4% PFA in PBS followed by a 2-min permeabilization in 0.5% Triton Materials and methods X-100. Immunolabeling with various antibodies was performed as in the pre- vious section. Cells were analyzed using a microscope (LSM510 Meta; Cell culture, transfection, plasmids, reagents, and Rho GTPase Carl Zeiss MicroImaging, Inc.) with a 63× NA 1.32 Apochromat water activity assays −/− immersion objective (Carl Zeiss MicroImaging, Inc.) and a photomultiplica- wt and p53 MEFs were generated, cultured, and transfected as previ- tor. Stacks of images were captured with LSM510 expert mode and were ously described (Gadea et al., 2002). A375P cells were cultured in DME restored as described in the previous section. with 10% FCS and transfected using electroporation (Nucleofector; Amaxa). In brief, 10 cells were suspended in 100 μl of solution R and mixed with 3 μg DNA for electroporation (program T-016). The primary Time-lapse imaging antibodies used were anti-p21 (C-19; Santa Cruz Biotechnology, Inc.), Time-lapse microscopy was performed on an inverted microscope (DL IRBE; anti-p53 (DO-1; Novocastra), and anti–α-tubulin (DM1A; Sigma-Aldrich). Leica) equipped with differential interference contrast (DIC) and GFP optics The inhibitors used were 10 μM Y27632 (Calbiochem) and 5 μM H1152 using a 63× NA 1.3 oil-immersion objective (Leica), sample heater (37°C), −1 (Calbiochem). Etoposide (Sigma-Aldrich) was used at 25 μg/ml . GFP- and a CO incubation chamber (Leica). Images were captured with a CCD tagged p53 wt and p53 H273 were previously described (Gadea et al., camera (MicroMax 1300; Princeton Instruments) using MetaMorph 6 soft- 2002). pSuper-siRNA p53 and plasmid EGFPC1-RhoE were gifts from ware (Molecular Devices), converted to tiff fi les that were edited with J.-C. Bourdon (University of Dundee, Dundee, Scotland, UK) and Ph. Fort (Cen- ImageJ (National Institutes of Health), and compiled into QuickTime videos tre National de la Recherche Scientifi que, Montpellier, France), respectively. (Macintosh). The exposure time was set to 50 ms. All videos were recorded The Rho GTPase activities were processed as previously described (Gadea at the frequency of one image every 5 s. et al., 2002). For 3D videos, we devised home-made bicompartment chambers comprised of an 8-mm metal ring placed in the center of a 30-mm Petri Invasion assay dish. The cells were mixed with serum-free Matrigel, and the resultant sus- The quantifi cation of cell invasion was performed in transwell cell culture pension was poured into the well formed by the metal ring. The assembly chambers containing fl uorescence-blocking polycarbonate porous mem- was placed in an incubator for 30 min while the Matrigel solidifi ed, after brane inserts (pore size of 8 μm; Fluoroblock; BD Biosciences). 100 μl which the metal ring was removed. The resultant disc of Matrigel was sur- Matrigel with reduced growth factors (a commercially prepared reconsti- rounded with 10% serum-complemented medium, taking care not to allow tuted basement membrane from Englebreth-Holm-Swarm tumors; BD Bio- any medium to fl ow on top, thus exposing the cells to a lateral serum gra- sciences) were prepared in a transwell. Cells were transfected and treated dient. Cells close to the periphery of the Matrigel disc were time lapsed to with TAT-C3 for 16 h and Y27632 or H1152 for 2 h as monolayer before monitor their ability to migrate along the serum gradient. Z sections re- trypsinization and plating (5 × 10 ) in serum-free media on top of a thick corded over a series of time points were combined into QuickTime videos. layer (around 500 μm) of Matrigel contained within the upper chamber of Cell velocity (Fig. 1 A) was determined by measuring the speed of 10–12 a transwell. Controls were left untreated. The upper and lower chambers cells, and the values are the means of three independent experiments. P53 DRIVES THE MODE OF CELLULAR MIGRATION • GADEA ET AL. 29 Levine, A.J. 1997. p53, the cellular gatekeeper for growth and division. Cell. Online supplemental material 88:323–331. Video 1 shows MEFs move using an elongated mode of motility in 3D matrix. −/− Video 2 shows p53 MEFs move using a rounded mode of motility Nakamoto, M., H. Teramoto, S. Matsumoto, T. Igishi, and E. Shimizu. 2001. K-ras and rho A mutations in malignant pleural effusion. Int. J. Oncol. in 3D matrix. Video 3 shows that rounded blebbing movement is pro- 19:971–976. voked by p53 defi ciency in 3D matrix. Video 4 shows that elongated Nobes, C.D., I. Lauritzen, M.G. Mattei, S. Paris, A. Hall, and P. Chardin. 1998. movement is driven by p53 activity in 3D matrix. Video 5 shows the −/− A new member of the Rho family, Rnd1, promotes disassembly of actin spreading of a p53 MEF treated with Y27632. Video 6 shows that fi lament structures and loss of cell adhesion. J. Cell Biol. 141:187–197. ROCK-∆1 promotes a rounded morphology. Video 7 shows that rounded Ongusaha, P.P., H.G. Kim, S.A. Boswell, A.J. Ridley, C.J. Der, G.P. Dotto, movement is promoted by ROCK-∆1 in 3D matrix. Videos 8 and 9 show Y.B. Kim, S.A. Aaronson, and S.W. Lee. 2006. RhoE is a pro-survival that p53 defi ciency–driven rounded morphology depends on ROCK p53 target gene that inhibits ROCK I-mediated apoptosis in response to activity in 3D matrix. Video 10 shows that p53 defi ciency–driven genotoxic stress. Curr. Biol. 16:2466–2472. rounded morphology is prevented by RhoE in 3D matrix. Fig. S1 shows Rihet, S., P. Vielh, J. Camonis, B. Goud, S. Chevillard, and J. de Gunzburg. 2001. that ROCK mediates p53-dependent regulation of the subcellular distribu- Mutation status of genes encoding RhoA, Rac1, and Cdc42 GTPases in tion of inte grin β1 and ezrin in rounded migrating cells cultured in 3D a panel of invasive human colorectal and breast tumors. J. Cancer Res. Matrigel. Online supplemental material is available at http://www.jcb Clin. Oncol. 127:733–738. .org/cgi/content/full/jcb.200701120/DC1. Sahai, E., and C.J. Marshall. 2003. Differing modes of tumour cell invasion have distinct requirements for Rho/ROCK signalling and extracellular proteo- We are grateful to Montpellier Rio Imaging for constructive microscopy and to lysis. Nat. Cell Biol. 5:711–719. R. Hipskind, C. Marshall, and A. Self for critical comments on the manuscript. Steeg, P.S. 2003. Metastasis suppressors alter the signal transduction of cancer We thank Dr. Ph. Fort and Dr. J.-C. Bourdon for providing plasmids. cells. Nat. Rev. Cancer. 3:55–63. This work was supported by the Fondation de France, Association pour la Recherche contre le Cancer (contrat 4028), Institut National de la Santé et Takaishi, K., T. Sasaki, T. Kameyama, S. Tsukita, and Y. Takai. 1995. Trans- location of activated Rho from the cytoplasm to membrane ruffl ing area, de la Recherche Médicale, and Centre National de la Recherche Scientifi que. cell-cell adhesion sites and cleavage furrows. Oncogene. 11:39–48. Totsukawa, G., Y. Yamakita, S. Yamashiro, D.J. Hartshorne, Y. Sasaki, and Submitted: 23 January 2007 F. Matsumura. 2000. Distinct roles of ROCK (Rho-kinase) and MLCK in Accepted: 31 May 2007 spatial regulation of MLC phosphorylation for assembly of stress fi bers and focal adhesions in 3T3 fi broblasts. J. Cell Biol. 150:797–806. References Amano, M., K. Chihara, K. Kimura, Y. Fukata, N. Nakamura, Y. Matsuura, and K. Kaibuchi. 1997. Formation of actin stress fi bers and focal adhesions enhanced by Rho-kinase. Science. 275:1308–1311. Bernards, R., and R.A. Weinberg. 2002. A progression puzzle. Nature. 418:823. Blagosklonny, M.V. 2002. P53: an ubiquitous target of anticancer drugs. Int. J. Cancer. 98:161–166. Clark, E.A., T.R. Golub, E.S. Lander, and R.O. Hynes. 2000. Genomic analysis of metastasis reveals an essential role for RhoC. Nature. 406:532–535. Coleman, M.L., E.A. Sahai, M. Yeo, M. Bosch, A. Dewar, and M.F. Olson. 2001. Membrane blebbing during apoptosis results from caspase-mediated acti- vation of ROCK I. Nat. Cell Biol. 3:339–345. Cukierman, E., R. Pankov, D.R. Stevens, and K.M. Yamada. 2001. Taking cell- matrix adhesions to the third dimension. Science. 294:1708–1712. Evers, E.E., R.A. van der Kammen, J.P. ten Klooster, and J.G. Collard. 2000. Rho- like GTPases in tumor cell invasion. Methods Enzymol. 325:403–415. Friedl, P., and E.B. Brocker. 2000. The biology of cell locomotion within three- dimensional extracellular matrix. Cell. Mol. Life Sci. 57:41–64. Gadea, G., L. Lapasset, C. Gauthier-Rouviere, and P. Roux. 2002. Regulation of Cdc42-mediated morphological effects: a novel function for p53. EMBO J. 21:2373–2382. Gadea, G., L. Roger, C. Anguille, M. de Toledo, V. Gire, and P. Roux. 2004. TNF{alpha} induces sequential activation of Cdc42- and p38/p53-dependent pathways that antagonistically regulate fi lopodia formation. J. Cell Sci. 117:6355–6364. Giaccia, A.J., and M.B. Kastan. 1998. The complexity of p53 modulation: emerg- ing patterns from divergent signals. Genes Dev. 12:2973–2983. Guasch, R.M., P. Scambler, G.E. Jones, and A.J. Ridley. 1998. RhoE regu- lates actin cytoskeleton organization and cell migration. Mol. Cell. Biol. 18:4761–4771. Guo, F., Y. Gao, L. Wang, and Y. Zheng. 2003. p19Arf-p53 tumor suppressor pathway regulates cell motility by suppression of phosphoinositide 3-kinase and Rac1 GTPase activities. J. Biol. Chem. 278:14414–14419. Hollstein, M., K. Rice, M.S. Greenblatt, T. Soussi, R. Fuchs, T. Sorlie, E. Hovig, B. Smith-Sorensen, R. Montesano, and C.C. Harris. 1994. Database of p53 gene somatic mutations in human tumors and cell lines. Nucleic Acids Res. 22:3551–3555. Ishizaki, T., M. Naito, K. Fujisawa, M. Maekawa, N. Watanabe, Y. Saito, and S. Narumiya. 1997. p160ROCK, a Rho-associated coiled-coil forming protein kinase, works downstream of Rho and induces focal adhesions. FEBS Lett. 404:118–124. Ishizaki, T., M. Uehata, I. Tamechika, J. Keel, K. Nonomura, M. Maekawa, and S. Narumiya. 2000. Pharmacological properties of Y-27632, a specifi c inhibitor of rho-associated kinases. Mol. Pharmacol. 57:976–983. Jaffe, A.B., and A. Hall. 2002. Rho GTPases in transformation and metastasis. Adv. Cancer Res. 84:57–80. 30 JCB • VOLUME 178 • NUMBER 1 • 2007 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journal of Cell Biology Pubmed Central

Loss of p53 promotes RhoA–ROCK-dependent cell migration and invasion in 3D matrices

Gadea, Gilles; de Toledo, Marion; Anguille, Christelle; Roux, Pierre
The Journal of Cell Biology , Volume 178 (1) – Jul 2, 2007
Download PDF
8 pages

Loading next page...
 
/lp/pubmed-central/loss-of-p53-promotes-rhoa-rock-dependent-cell-migration-and-invasion-Bj4Lb5lIin

References (33)

Publisher
Pubmed Central
Copyright
Copyright © 2007, The Rockefeller University Press
ISSN
0021-9525
eISSN
1540-8140
DOI
10.1083/jcb.200701120
Publisher site
See Article on Publisher Site

Abstract

JCB: REPORT Loss of p53 promotes RhoA–ROCK-dependent cell migration and invasion in 3D matrices Gilles Gadea, Marion de Toledo, Christelle Anguille, and Pierre Roux Centre de Recherche en Biochimie Macromoléculaire, Centre National de la Recherche Scientifi que, Universite Mixte de Recherche 5237, Institut Federatif de Recherche 122, F-34293 Montpellier, Cedex 5, France n addition to its role in controlling cell cycle progression, movement and have considerably increased invasive prop- the tumor suppressor protein p53 can also affect other erties. The morphological transition requires the RhoA– I cellular functions such as cell migration. In this study, we ROCK (Rho-associated coil-containing protein kinase) show that p53 defi ciency in mouse embryonic fi broblasts pathway and is prevented by RhoE. A similar p53-mediated cultured in three-dimensional matrices induces a switch from transition is observed in melanoma A375P cancer cells. an elongated spindle morphology to a markedly spherical Our data suggest that genetic alterations of p53 in tumors and fl exible one associated with highly dynamic membrane are suffi cient to promote motility and invasion, thereby con- blebs. These rounded, motile cells exhibit amoeboid-like tributing to metastasis. Introduction and its main effector, Rho-associated coil-containing protein A critical event during tumorigenesis is the conversion from a kinase (ROCK), and resembles the amoeboid movement. This static primary tumor to an invasive, disseminating metastasis. involves a rounded bleb-associated movement that generates Moreover, tumor cells show an increased capacity to migrate. propulsive motion through the matrix independently of proteol- Numerous intracellular signaling molecules have been impli- ysis (Sahai and Marshall, 2003). cated in migratory processes. Among them, the Rho GTPase Curiously, genes encoding Rho GTPases are rarely found family plays a pivotal role in regulating the biochemical and cyto- mutated in human cancers, in which only their functional ac- skeletal pathways relevant to cell migration. The Rho GTPases tivities seem to be deregulated (Nakamoto et al., 2001; Rihet Rac, Cdc42, and Rho control cell protrusions during migration. et al., 2001). This suggests that alterations in others genes Aberrant regulation of Rho proteins is believed to associate with account for functional modifications of Rho GTPases ac- metastasis by promoting cell motility (Clark et al., 2000; Evers companying actin cytoskeleton remodeling during metastasis. et al., 2000; Jaffe and Hall, 2002; Steeg, 2003). We hypothesize that this set of genes controls the cell cycle. The bona fi de environment for migrating cells is the extra- Their mutation during the initiation phase in cancer would cellular matrix, which permits movement in a 3D scaffold that modify the behavior of proteins involved in actin cytoskeleton is biochemically complex and shows dynamic fl exibility. 3D dynamics, such as Rho GTPases, leading to a migratory and tissue culture models reconstitute an environment that resem- invasive phenotype. bles the in vivo situation in regard to cell shape and movement. Beyond these genes is the tumor suppressor p53, whose This model provides important insights into the mechanisms of mutations occurs in >50% of human tumors (Hollstein et al., cell motion during carcinogenesis. Recent advances have iden- 1994). p53 protects cells from malignant transformation by tifi ed two modes of cell motility in 3D matrices. The elongated regulating cell cycle arrest or by promoting apoptosis (Levine, mode of migration is a consequence of Rac activity and gener- 1997; Giaccia and Kastan, 1998). We and others have recently ates membrane protrusions, the lamellipodia that drive motility. shown that p53 modulates cell migration: p53 negatively In contrast, a novel rounded mode of motility depends on RhoA modulates Rho GTPases and regulates cell polarization and migration (Gadea et al., 2002, 2004; Guo et al., 2003). How- Correspondence to Pierre Roux: [email protected] ever, little is known about the role of p53 in cells moving in G. Gadea’s present address is Institute of Cancer Research, London SW3 6JB, UK. a 3D matrix that mimics the in vivo microenvironment of Abbreviations used in this study: DIC, differential interference contrast; F-actin, tumor cells. fi lamentous actin; MEF, mouse embryonic fi broblast; ROCK, Rho-associated coil- Identifying the mechanisms by which p53 modulates cell containing protein kinase; wt, wild type. The online version of this article contains supplemental material. migration is important to understand how invasive cells arise. © The Rockefeller University Press $15.00 The Journal of Cell Biology, Vol. 178, No. 1, July 2, 2007 23–30 http://www.jcb.org/cgi/doi/10.1083/jcb.200701120 JCB 23 THE JOURNAL OF CELL BIOLOGY In this study, we show that the elongated spindle-shaped fi bro- at http://www.jcb.org/cgi/content/full/jcb.200701120/DC1). blastoid mode of motility can be converted to a rounded bleb- wt MEF maintained a spindle-shaped elongated morphology, −1 bing movement by blocking p53 function. This amoeboid mode generated extensions, and moved slowly at 2 ± 1 μm/h , −/− of motility requires RhoA–ROCK signaling and confers higher whereas p53 MEF moved as rounded cells using a transla- velocity and invasive properties to p53-defi cient cells. Thus, the tory motion in the direction of the serum growth factor gradient. range of p53 tumor suppressor activity extends to the control of They show many dynamic bleblike structures, alternating rapid the mode of migration of invasive cells. cycles of squeezing and expansion, and considerable deform- −/− −1 ability. p53 cells moved at 12 ± 5 μm/h , maintaining their rounded morphology. Expression of the GFP-tagged mu- Results and discussion tant p53 H273, a dominant-negative form of p53, converts −/− p53 fi broblasts show rounded the elongated morphology of MEF to a marked spherical and blebbing movement fl exible one associated with highly dynamic membrane blebs −/− Initially, we compared the movement of p53-defi cient primary similar to those observed in p53 MEFs (Fig. 1 B and Video 3). −/− −/− mouse embryonic fi broblasts (MEFs; p53 MEFs) with that The expression of GFP-tagged wt p53 in p53 MEFs inhibits of wild-type (wt) MEFs cultured in 3D Matrigel matrix using dynamic blebbing and leads to an elongated morphology (Fig. time-lapse video microscopy. The migration of the two cell types 1 B and Video 4), whereas that of GFP-tagged p53 H273 has no is strikingly different (Fig. 1 A and Videos 1 and 2, available effect. Thus, the loss of wt p53 activity is suffi cient to drive Figure 1. Rounded blebbing movement is provoked by p53 defi ciency. (A) Time-lapse analysis of MEFs −/− and p53 MEFs moving within Matrigel. Still DIC time-lapse images from accompanying videos of −/− wt MEFs (Video 1) and p53 MEFs (Video 2, avail- able at http://www.jcb.org/cgi/content/full/jcb .200701120/DC1). Cells were plated into a thick layer of Matrigel, and a gradient of serum growth fac- tors was established. Frames show the cell migration at the indicated times. Arrows indicate the direction of cell movement. Dotted circles indicate the initial situa- tion of the cell when the video starts. (B) Time-lapse analysis of the different modes of cell motility within Matrigel. DIC images of cells transfected with GFP- tagged plasmids (to enable the identifi cation of trans- fected cells as visualized in green) expressing either p53 H273 in MEFs (still images from Video 3), p53 wt −/− in p53 MEFs (still images from Video 4), or p53 −/− H273 in p53 MEFs as control (right). Cells were plated within Matrigel as in A. (C) p53 regulates sub- cellular distributions of integrin β1 and ezrin in rounded migrating cells. Representative integrin β1 and ezrin −/− fl uorescent staining of MEFs and p53 MEFs trans- fected with p53 wt, p53 H273, siRNA p53, or plas- mid EGFP. Insets show the covisualization of F-actin (red), GFP (green), and ezrin or integrin β1 (blue) as indicated. Cells were plated in Matrigel, and a gradient of serum growth factors was established. Arrows show the direction of cell movement in Matrigel. Bars, 10 μm. 24 JCB • VOLUME 178 • NUMBER 1 • 2007 the switch from an elongated fi broblast-like to a rounded type in the newly formed long extensions (Fig. 1 C). Thus, p53 con- of migration. trols the clustering of integrin β1 and ezrin in rounded cells. In Matrigel, cells moving with a rounded or elongated mor- Culture of fi broblastic cells in monolayers imposes an arti- phology show a distinct subcellular distribution of integrin β1 fi cial polarity between the lower and upper surface of these apo- and ezrin that helps in establishing the direction of movement lar cells. Their morphology and migration change in a 3D matrix, (Sahai and Marshall, 2003). In Matrigel, rounded blebbing suggesting that the spatial constraints and surrounding biochem- MEFs that arise from p53 inactivation show integrin β1 and ezrin ical microenvironment are a better approximation of the in vivo relocalization to the moving front of the cell. A similar clustering situation (Friedl and Brocker, 2000; Cukierman et al., 2001). −/− was also observed in rounded p53 MEFs. In contrast, conver- This is particularly striking for p53-defi cient MEFs that become −/− sion of p53 MEFs to elongated morphology by the expression blebbing spherical cells once suspended in 3D matrices, losing of p53 wt led to their relocalization throughout the cell, notably their initial fi broblastoid morphology observed in 2D culture. −/− Figure 2. p53 regulates RhoA activation. (A) GTPase activities in wt MEFs and p53 MEFs. Cells were lysed 24 h after plating, and the levels of GTP- −/− bound RhoA, Rac1, and Cdc42 were assayed. (B) RhoA activity in transfected MEFs and p53 MEFs. Cells transfected with p53 wt and p53 H273 were assayed for RhoA activity. (A and B) The values are the mean ± SD (error bars) of three independent experiments. (C–F) Representative F-actin and RhoA −/− fl uorescent staining of wt MEFs and p53 MEFs untreated (C), treated with TAT-C3 (F), or transfected with plasmids expressing either p53 wt (D) or p53 H273 (E). Panels show phalloidin-stained F-actin (red) and RhoA labeling (green). Arrows point to blebbing structures. Bars, 10 μm. P53 DRIVES THE MODE OF CELLULAR MIGRATION • GADEA ET AL. 25 −/− −/− RhoA activity is enhanced in p53 MEFs actin (F-actin) in p53 and wt MEFs. In wt MEFs, RhoA −/− This rounded blebbing movement observed in p53 MEFs is exhibits a punctuate distribution throughout the cytoplasm with strikingly similar to amoeboid-like motility, which is dependent a marked concentration in the perinuclear region. F-actin stress on Rho–ROCK signaling in a 3D matrix (Sahai and Marshall, fi bers cross the cytoplasm and accumulate slightly around the −/− −/− 2003). We compared the level of GTP-bound RhoA in p53 edge of the cell (Fig. 2 C, top). In p53 MEFs, RhoA shows −/− and wt MEFs. GTP-RhoA was increased 3.4-fold in p53 its punctuate cytoplasmic localization but is excluded from the MEFs, whereas GTP-Rac and Cdc42 were barely affected (Fig. perinuclear region (Fig. 2 C, bottom). Instead, RhoA colocal- 2 A). This change was functional because protein expression izes with large patches of polymerized actin in bleblike globular levels were unaltered by p53 deletion. Reintroduction of wt p53 structures that are distributed over and protruding from the cell −/− in p53 MEFs lowered the level of GTP-RhoA to that of wt surface (Fig. 2 C, arrows). In addition, MEFs lacking p53 activ- MEFs (Fig. 2 B). Conversely, p53 H273 did not affect the over- ity show more peripheral microspikes, which were previously −/− activation of RhoA in p53 MEFs. Finally, blocking p53 in characterized as fi lopodia (Gadea et al., 2002). They also show MEFs using p53 H273 or siRNA-mediated knockdown of p53 an accumulation of peripheral polymerized actin bundles (i.e., a was suffi cient to increase GTP-RhoA (Fig. 2 B). These results redistribution of F-actin stress fi bers; Fig. 2 C). The expression show that wt p53 inhibits RhoA activation and has a role in reg- of wt p53 led to loss of the bleblike protrusions and recovery of ulating RhoA signaling pathways. the punctuate RhoA localization in the perinuclear region. This was not seen with p53 H273, indicating that wt p53 activity is −/− Distinct localization of RhoA in p53 MEFs required for RhoA localization (Fig. 2, compare D with E). −/− Translocation from the cytoplasm to the membrane area is essen- Treatment of p53 MEFs with TAT-C3, which inactivates tial for both the activation and function of RhoA (Takaishi et al., RhoA, RhoB, and RhoC (Coleman et al., 2001), abolished bleb- 1995). We compared the localization of RhoA and fi lamentous like structures and led to RhoA redistribution in the perinuclear Figure 3. ROCK activity is required for rounded blebbing morphology driven by p53 defi ciency. (A) Still DIC time-lapse images from Video 5 (available at http://www.jcb.org/cgi/content/ −/− full/jcb.200701120/DC1) of p53 MEFs treated with Y27632. Frames show the spread- −/− ing of a p53 MEF at the indicated times. (B) DIC images show wt MEFs cotransfected with GFP to identify transfected cells (insets) and ROCK-∆1 (still images from Video 6; Fig. S1) or with GFP alone as a control. (C) Represen- tative F-actin and RhoA fl uorescent staining of −/− wt MEFs and p53 MEFs treated either with Y27632 or H1152. (D) Time-lapse analysis of different modes of cell motility within Matrigel. Still DIC images of MEFs transfected either with ROCK-∆1 (still DIC images from Video 7) or GFP alone as a control. (bottom) Still DIC −/− images of p53 MEFs treated with Y27632 or H1152 and transfected with p53 H273 (from Video 8 and Video 9, respectively). The transfected cells expressing either ROCK-∆1 or p53 H273 are visualized using GFP. Cells were plated within Matrigel as in Fig. 1. Bars, 10 μm. 26 JCB • VOLUME 178 • NUMBER 1 • 2007 Invasion of p53-defi cient cells requires region (Fig. 2 F). Thus, RhoA activation is required for bleb −/− ROCK activity and is prevented by RhoE formation in p53 cells. RhoA–ROCK signaling also plays a key role in invasion by p160 ROCK controls morphological blebbing cells moving in 3D matrix (Sahai and Marshall, 2003). −/− −/− MEFs, which have an overactivation of this changes observed in p53 fi broblasts Similarly, p53 RhoA-driven actin reorganization largely depends on the serine– pathway, showed substantially higher invasiveness than wt threonine protein kinase ROCK (Amano et al., 1997; Ishizaki MEFs in Matrigel (Fig. 4 A). Inhibition of Rho or ROCK by et al., 1997). To determine whether ROCK mediates mem- −/− brane blebbing in p53 MEFs cultured in monolayer, we used Y27632 and H1152, which are two distinct, structurally un- −/− related pharmacological inhibitors of ROCK. p53 MEFs promptly lost their extensive dynamic bleblike activities and their rounded morphology within 40 min, adopting a fl attened, spread shape when treated with Y27632 (Fig. 3 A and Video 5, avail- able at http://www.jcb.org/cgi/content/full/jcb.200701120/DC1) or H1152 (not depicted). Consistent with this, the expression of ROCK-∆1, an active mutant, in wt MEF promoted a rounded morphology accompanied by numerous dynamic bleblike ex- −/− tensions similar to p53 MEF (Fig. 3 B and Video 6). GFP −/− alone had no effect (Fig. 3 B). The spreading of p53 MEFs induced by Y27632 or H1152 was accompanied by the reorga- nization of F-actin from cortical bundles into scattered cyto- plasmic dotlike structures concomitant with the loss of fi lopo dia (Fig. 3 C). Similar to other studies (Ishizaki et al., 2000; Totsukawa et al., 2000), ROCK inhibitors strongly reduced −/− the number of stress fi bers in both wt and p53 MEFs. In −/− addition, ROCK inhibition in p53 MEFs led to RhoA relocal- ization to the perinuclear region (compare Fig. 3 C with Fig. 2 C). Thus, ROCK inhibition causes RhoA to lose its mem- brane localization. We tested whether ROCK activity is required for the −/− rounded blebbing movement of p53 MEFs cultured in Matrigel. Transfection of ROCK-∆1 converts MEFs from an elongated to a rounded blebbing morphology; GFP alone had no effect (Fig. 3 D and Video 7, available at http://www.jcb.org/cgi/content/ −/− full/jcb.200701120/DC1). In contrast, treatment of p53 MEFs with Y27632 or H1152 inhibits dynamic blebbing and leads to an elongated morphology even upon the expression of p53 H273 (Fig. 3 D and Videos 8 and 9). This transition is ac- companied by the redistribution of integrin β1 and ezrin to the moving front of the cell (Fig. S1). Thus, ROCK activity is required for the amoeboid-like rounded morphology driven by p53 defi ciency. Collectively, our results indicate that characteristics of the fi broblast type of motility (i.e., that using elongation and trac- tion) are lost when p53 function is abrogated using either p53 H273 or siRNA. Cells adopt a rounded bleb morphology similar Figure 4. p53 controls cell invasion through the RhoA–ROCK pathway. −/− −/− to that observed in p53 MEFs cultured in 3D matrices. A p53 (A) Matrigel invasion of wt MEFs and p53 MEFs. Cells were treated with either TAT-C3, Y27632, H1152, or protease inhibitors or were transfected defect leads to the aberrant overactivation of RhoA, which con- with plasmids encoding either p53 wt, p53 H273, siRNA p53, ROCK-∆1, sequently translocates to specifi c membrane blebbing structures. or RhoA-V14 as indicated. The changes in Matrigel invasion were ana- Importantly, we show that the RhoA–ROCK signaling pathway lyzed as described in Materials and methods. For each data point, GFP- positive cells were scored. (B) Representative RhoE and actin fl uorescent is involved in p53-dependent morphological changes by modu- −/− staining of p53 MEFs transfected with GFP-RhoE. (C) DIC images of lating round blebbing. It also appears to promote blebbing in −/− p53 MEFs transfected with GFP-tagged RhoE (still images from Video 10, −/− p53 cells. Notably, RhoA signaling through ROCK was pre- available at http://www.jcb.org/cgi/content/full/jcb.200701120/DC1). Cells were plated in Matrigel as in Fig. 1 A. (D) Matrigel invasiveness of viously shown to be important in the rounded blebbing move- −/− p53 MEFs transfected either with vector alone (control) or with RhoE. ment but not for the elongated protrusive movement of cancer Quantifi cation was performed as in A. (A and D) Values are the means ± SD cells cultured in 3D matrices (Sahai and Marshall, 2003). (error bars) of three independent experiments. Bars, 10 μm. P53 DRIVES THE MODE OF CELLULAR MIGRATION • GADEA ET AL. 27 treatment with TAT-C3, Y27632, or H1152 diminished the abil- (Fig. 4 D), suggesting that RhoE helps mediate the control of −/− ity of p53 MEFs to invade Matrigel by around 60%. Expres- cell morphology and invasiveness by p53. sion of wt p53 but not p53 H273 blocked this behavior. The ability of MEFs to invade Matrigel was increased p53 inactivation promotes rounded when p53 activity was blocked either by expression of the H273 blebbing movement in A375P melanoma p53 mutant or using siRNA to p53 in MEFs. Similarly, invasive- cancer cells ness was enhanced by an active mutant of RhoA (RhoA-V14) We sought to extend these observations to cancer cells. We chose and ROCK-∆1. Invasiveness of blebbing cells is independent of A375P melanoma cells, which have a wt form of p53 and show an protease activities (Sahai and Marshall, 2003); accordingly, the elongated mode of migration in 3D matrix (Sahai and Marshall, −/− invasive activity of p53 MEFs was unaffected by the inhibi- 2003). Cell morphology was visualized using fl uorescent wheat tion of proteases. Thus, the Rho–ROCK signaling module is germ agglutinin, which provides highly selective labeling of the −/− necessary for the invasive behavior of p53 cells in Matrigel. external plasma membrane. The expression of GFP-tagged p53 RhoE, a member of the Rho family that blocks actin stress H273 confers a rounded blebbing morphology to A375P cells, fi bers (Guasch et al., 1998; Nobes et al., 1998), is a transcrip- which associates with membrane blebs similar to those observed −/− tional target of p53 in response to genotoxic stress; RhoE pro- in p53 MEFs (Fig. 5 A). In contrast, the elongated morphology motes cell survival through the inhibition of ROCK1-mediated of A375P was unchanged when p53 accumulation was induced apoptosis (Ongusaha et al., 2006). The overexpression of RhoE by etoposide, which was verifi ed using the induction of p21WAF, clearly prevented actin polymerization and abolished the bleb- a transcriptional target of p53 (Fig. 5 B). These morphologic −/− like protrusions in p53 MEFs, converting them to an elon- changes correlated with the level of GTP-RhoA, which was gated morphology (Fig. 4, B and C; and Video 10, available at 4.6-fold higher upon the expression of p53 H273 (Fig. 5 C). Thus, http://www.jcb.org/cgi/content/full/jcb.200701120/DC1). RhoE the amoeboid-like motility of A375P melanoma cells induced by −/− also considerably reduced the invasiveness of p53 MEFs the loss of p53 activity is dependent on RhoA activation. Figure 5. A375 melanoma cell invasiveness is regulated by p53 activity. (A) Confocal images of A375P cells expressing plasmid EGFP alone or p53 H273. Cells were plated in Matrigel, and a gradient of serum growth factors was established to observe the mode of migration. The transfected cells are visualized using GFP, and plasma membranes were labeled with fl uorescent WGA (wheat germ agglutinin; red). (B) p21 induction in A375P cells. A375P cells were untreated or treated with etoposide. Total protein lysates were analyzed by SDS-PAGE followed by immunoblot analysis using antibodies to p21. Normalization was performed with an anti–α-tubulin antibody. (C) RhoA activity in transfected A375P. Cells expressing GFP-tagged p53 H273 or plasmid EGFP (control) were compared for GTP-RhoA levels as in Fig. 2. The image is representative of three independent experiments. The values are the mean ± SD (error bars). (D) Invasiveness in 3D matrigel matrix of A375P-expressing p53 wt or p53 H273 treated with TAT-C3, Y27632, or H1152 as indicated. Results are expressed as in Fig. 4. 28 JCB • VOLUME 178 • NUMBER 1 • 2007 Amoeboid cancer cell motility is particularly effi cient in were then fi lled with serum-free DME and DME with 10% FCS, respectively, thus establishing a soluble gradient of chemoattractant that permits cell in- promoting the invasiveness of metastatic cells in 3D environ- vasion throughout the Matrigel. Inhibitors were added immediately after ments. Therefore, we performed invasion assays in Matrigel us- cell plating at the aforementioned concentrations. Cells were allowed to in- −/− ing this model. The expression of p53 H273 strongly increased vade at 37°C and 5% CO through the gel for 36 h (for MEF and p53 MEF) or for 24 h (for A375P) before fi xing for 15 min in 3.7% formalde- the invasiveness of A375P, whereas wt p53 inhibited it (Fig. 5 D). hyde. Cells that had invaded through the Matrigel were detected on the TAT-C3, Y27632, or H1152 treatment largely reduced this re- lower side of the fi lter by GFP fl uorescence and counted for cell number. sponse in p53 H273–expressing A375P cells, arguing that mu- Six fi elds per fi lter were counted, and each assay was performed twice in triplicate for each cell line. tated p53 drives this behavior through RhoA–ROCK signaling. In contrast, protease inhibition did not alter p53 H273–driven Immunofl uorescence invasiveness (Fig. 5 D). This stresses the role of the amoeboid- Cells were transfected and/or treated with TAT-C3, Y27632, or H1152 on coverslips at a confl uence of 30% before fi xation in 3.7% formaldehyde like rounded morphology of p53-defi cient A375P cells in their in PBS followed by a 5-min permeabilization in 0.1% Triton X-100 in PBS invasive potential. and were incubated in PBS containing 0.1% BSA before staining with pri- Our results show that blocking p53 function is suffi cient to mary antibodies as follows: RhoA (Santa Cruz Biotechnology, Inc.), anti- ezrin polyclonal antibody (provided by P. Mangeat, Centre de Recherche en convert melanoma cells to a rounded locomotion, indicating that Biochimie Macromoléculaire, Montpellier, France), and integrin β1 (Santa cancer cells may also be subject to this regulation. This switch Cruz Biotechnology, Inc.). All of the antibodies were revealed with either would favor the illegitimate dissemination of cancer cells con- an AlexaFluor546- or -488–conjugated goat anti–mouse or anti–rabbit antibody (Invitrogen and Interchim). Cells were stained for F-actin using taining mutant p53, possibly increasing their invasiveness. TRITC-conjugated phalloidin (Sigma-Aldrich) or for membrane morphology The tumor suppressor p53 is mutated in >50% of human using cell-impermeant AlexaFluor594 wheat germ agglutinin, which binds cancers (Hollstein et al., 1994) and, thus, constitutes an appeal- selectively to N-acetylglucosamin and N-acetylneuraminic (sialic) acid resi- dues and provides highly selective labeling of the plasma membrane (Invit- ing target for anticancer therapy. Diverse therapeutic strategies rogen). Cells were prepared as described previously (Gadea et al., 2002). already exist that attempt to restore wt p53 function in cancer Stacks of 16-bit fl uorescent images (z step of 0.1 μm) were captured with cells, including the rescue of mutant p53 function and reactiva- a MetaMorph-driven microscope (DMRB; Leica) using a 63× NA 1.4 Apo- chromat oil immersion objective (Leica), a piezo stepper (E662; Physik In- tion of wt p53 (Blagosklonny, 2002). Given that there is no struments), and a camera (CoolSNAP HQ2; Photometrics). Epifl uorescence obvious selection for a metastatic phenotype, Bernards and of all images (in 2D and 3D) were restored and deconvolved with Huygens Weinberg (2002) proposed that classic oncogenes and tumor (Scientifi c Volume Imaging) using the maximum likelihood estimation algo- rithm. Restored stacks were processed with Imaris (Bitplane) for visualiza- suppressor genes implicated in the early stages of tumorigenesis tion. The restored images were saved as tiff fi les that were mounted using may also play a role in metastasis. Our studies indicate that p53 Photoshop and Illustrator (Adobe). is an ideal candidate, as it is involved both in the control of tumor cell apoptosis and tumor cell invasion. This extends the Immunofl uorescence in 3D matrices For immunofl uorescence of cells in 3D Matrigel network, cultures were applications for anticancer agents aimed at restoring p53 wt transfected and treated with TAT-C3, Y27632, or H1152 as monolayer fol- functions: such agents may combine antiproliferative and anti- lowed by trypsinization and embedding into Matrigel in transwell cell cul- invasive activities. ture chambers (as previously described in the Invasion assay section except that cells were allowed to invade through the gel for 6 h). Controls were left untreated. 50-μm cryosections were then cut at –20°C before fi xation in 4% PFA in PBS followed by a 2-min permeabilization in 0.5% Triton Materials and methods X-100. Immunolabeling with various antibodies was performed as in the pre- vious section. Cells were analyzed using a microscope (LSM510 Meta; Cell culture, transfection, plasmids, reagents, and Rho GTPase Carl Zeiss MicroImaging, Inc.) with a 63× NA 1.32 Apochromat water activity assays −/− immersion objective (Carl Zeiss MicroImaging, Inc.) and a photomultiplica- wt and p53 MEFs were generated, cultured, and transfected as previ- tor. Stacks of images were captured with LSM510 expert mode and were ously described (Gadea et al., 2002). A375P cells were cultured in DME restored as described in the previous section. with 10% FCS and transfected using electroporation (Nucleofector; Amaxa). In brief, 10 cells were suspended in 100 μl of solution R and mixed with 3 μg DNA for electroporation (program T-016). The primary Time-lapse imaging antibodies used were anti-p21 (C-19; Santa Cruz Biotechnology, Inc.), Time-lapse microscopy was performed on an inverted microscope (DL IRBE; anti-p53 (DO-1; Novocastra), and anti–α-tubulin (DM1A; Sigma-Aldrich). Leica) equipped with differential interference contrast (DIC) and GFP optics The inhibitors used were 10 μM Y27632 (Calbiochem) and 5 μM H1152 using a 63× NA 1.3 oil-immersion objective (Leica), sample heater (37°C), −1 (Calbiochem). Etoposide (Sigma-Aldrich) was used at 25 μg/ml . GFP- and a CO incubation chamber (Leica). Images were captured with a CCD tagged p53 wt and p53 H273 were previously described (Gadea et al., camera (MicroMax 1300; Princeton Instruments) using MetaMorph 6 soft- 2002). pSuper-siRNA p53 and plasmid EGFPC1-RhoE were gifts from ware (Molecular Devices), converted to tiff fi les that were edited with J.-C. Bourdon (University of Dundee, Dundee, Scotland, UK) and Ph. Fort (Cen- ImageJ (National Institutes of Health), and compiled into QuickTime videos tre National de la Recherche Scientifi que, Montpellier, France), respectively. (Macintosh). The exposure time was set to 50 ms. All videos were recorded The Rho GTPase activities were processed as previously described (Gadea at the frequency of one image every 5 s. et al., 2002). For 3D videos, we devised home-made bicompartment chambers comprised of an 8-mm metal ring placed in the center of a 30-mm Petri Invasion assay dish. The cells were mixed with serum-free Matrigel, and the resultant sus- The quantifi cation of cell invasion was performed in transwell cell culture pension was poured into the well formed by the metal ring. The assembly chambers containing fl uorescence-blocking polycarbonate porous mem- was placed in an incubator for 30 min while the Matrigel solidifi ed, after brane inserts (pore size of 8 μm; Fluoroblock; BD Biosciences). 100 μl which the metal ring was removed. The resultant disc of Matrigel was sur- Matrigel with reduced growth factors (a commercially prepared reconsti- rounded with 10% serum-complemented medium, taking care not to allow tuted basement membrane from Englebreth-Holm-Swarm tumors; BD Bio- any medium to fl ow on top, thus exposing the cells to a lateral serum gra- sciences) were prepared in a transwell. Cells were transfected and treated dient. Cells close to the periphery of the Matrigel disc were time lapsed to with TAT-C3 for 16 h and Y27632 or H1152 for 2 h as monolayer before monitor their ability to migrate along the serum gradient. Z sections re- trypsinization and plating (5 × 10 ) in serum-free media on top of a thick corded over a series of time points were combined into QuickTime videos. layer (around 500 μm) of Matrigel contained within the upper chamber of Cell velocity (Fig. 1 A) was determined by measuring the speed of 10–12 a transwell. Controls were left untreated. The upper and lower chambers cells, and the values are the means of three independent experiments. P53 DRIVES THE MODE OF CELLULAR MIGRATION • GADEA ET AL. 29 Levine, A.J. 1997. p53, the cellular gatekeeper for growth and division. Cell. Online supplemental material 88:323–331. Video 1 shows MEFs move using an elongated mode of motility in 3D matrix. −/− Video 2 shows p53 MEFs move using a rounded mode of motility Nakamoto, M., H. Teramoto, S. Matsumoto, T. Igishi, and E. Shimizu. 2001. K-ras and rho A mutations in malignant pleural effusion. Int. J. Oncol. in 3D matrix. Video 3 shows that rounded blebbing movement is pro- 19:971–976. voked by p53 defi ciency in 3D matrix. Video 4 shows that elongated Nobes, C.D., I. Lauritzen, M.G. Mattei, S. Paris, A. Hall, and P. Chardin. 1998. movement is driven by p53 activity in 3D matrix. Video 5 shows the −/− A new member of the Rho family, Rnd1, promotes disassembly of actin spreading of a p53 MEF treated with Y27632. Video 6 shows that fi lament structures and loss of cell adhesion. J. Cell Biol. 141:187–197. ROCK-∆1 promotes a rounded morphology. Video 7 shows that rounded Ongusaha, P.P., H.G. Kim, S.A. Boswell, A.J. Ridley, C.J. Der, G.P. Dotto, movement is promoted by ROCK-∆1 in 3D matrix. Videos 8 and 9 show Y.B. Kim, S.A. Aaronson, and S.W. Lee. 2006. RhoE is a pro-survival that p53 defi ciency–driven rounded morphology depends on ROCK p53 target gene that inhibits ROCK I-mediated apoptosis in response to activity in 3D matrix. Video 10 shows that p53 defi ciency–driven genotoxic stress. Curr. Biol. 16:2466–2472. rounded morphology is prevented by RhoE in 3D matrix. Fig. S1 shows Rihet, S., P. Vielh, J. Camonis, B. Goud, S. Chevillard, and J. de Gunzburg. 2001. that ROCK mediates p53-dependent regulation of the subcellular distribu- Mutation status of genes encoding RhoA, Rac1, and Cdc42 GTPases in tion of inte grin β1 and ezrin in rounded migrating cells cultured in 3D a panel of invasive human colorectal and breast tumors. J. Cancer Res. Matrigel. Online supplemental material is available at http://www.jcb Clin. Oncol. 127:733–738. .org/cgi/content/full/jcb.200701120/DC1. Sahai, E., and C.J. Marshall. 2003. Differing modes of tumour cell invasion have distinct requirements for Rho/ROCK signalling and extracellular proteo- We are grateful to Montpellier Rio Imaging for constructive microscopy and to lysis. Nat. Cell Biol. 5:711–719. R. Hipskind, C. Marshall, and A. Self for critical comments on the manuscript. Steeg, P.S. 2003. Metastasis suppressors alter the signal transduction of cancer We thank Dr. Ph. Fort and Dr. J.-C. Bourdon for providing plasmids. cells. Nat. Rev. Cancer. 3:55–63. This work was supported by the Fondation de France, Association pour la Recherche contre le Cancer (contrat 4028), Institut National de la Santé et Takaishi, K., T. Sasaki, T. Kameyama, S. Tsukita, and Y. Takai. 1995. Trans- location of activated Rho from the cytoplasm to membrane ruffl ing area, de la Recherche Médicale, and Centre National de la Recherche Scientifi que. cell-cell adhesion sites and cleavage furrows. Oncogene. 11:39–48. Totsukawa, G., Y. Yamakita, S. Yamashiro, D.J. Hartshorne, Y. Sasaki, and Submitted: 23 January 2007 F. Matsumura. 2000. Distinct roles of ROCK (Rho-kinase) and MLCK in Accepted: 31 May 2007 spatial regulation of MLC phosphorylation for assembly of stress fi bers and focal adhesions in 3T3 fi broblasts. J. Cell Biol. 150:797–806. References Amano, M., K. Chihara, K. Kimura, Y. Fukata, N. Nakamura, Y. Matsuura, and K. Kaibuchi. 1997. Formation of actin stress fi bers and focal adhesions enhanced by Rho-kinase. Science. 275:1308–1311. Bernards, R., and R.A. Weinberg. 2002. A progression puzzle. Nature. 418:823. Blagosklonny, M.V. 2002. P53: an ubiquitous target of anticancer drugs. Int. J. Cancer. 98:161–166. Clark, E.A., T.R. Golub, E.S. Lander, and R.O. Hynes. 2000. Genomic analysis of metastasis reveals an essential role for RhoC. Nature. 406:532–535. Coleman, M.L., E.A. Sahai, M. Yeo, M. Bosch, A. Dewar, and M.F. Olson. 2001. Membrane blebbing during apoptosis results from caspase-mediated acti- vation of ROCK I. Nat. Cell Biol. 3:339–345. Cukierman, E., R. Pankov, D.R. Stevens, and K.M. Yamada. 2001. Taking cell- matrix adhesions to the third dimension. Science. 294:1708–1712. Evers, E.E., R.A. van der Kammen, J.P. ten Klooster, and J.G. Collard. 2000. Rho- like GTPases in tumor cell invasion. Methods Enzymol. 325:403–415. Friedl, P., and E.B. Brocker. 2000. The biology of cell locomotion within three- dimensional extracellular matrix. Cell. Mol. Life Sci. 57:41–64. Gadea, G., L. Lapasset, C. Gauthier-Rouviere, and P. Roux. 2002. Regulation of Cdc42-mediated morphological effects: a novel function for p53. EMBO J. 21:2373–2382. Gadea, G., L. Roger, C. Anguille, M. de Toledo, V. Gire, and P. Roux. 2004. TNF{alpha} induces sequential activation of Cdc42- and p38/p53-dependent pathways that antagonistically regulate fi lopodia formation. J. Cell Sci. 117:6355–6364. Giaccia, A.J., and M.B. Kastan. 1998. The complexity of p53 modulation: emerg- ing patterns from divergent signals. Genes Dev. 12:2973–2983. Guasch, R.M., P. Scambler, G.E. Jones, and A.J. Ridley. 1998. RhoE regu- lates actin cytoskeleton organization and cell migration. Mol. Cell. Biol. 18:4761–4771. Guo, F., Y. Gao, L. Wang, and Y. Zheng. 2003. p19Arf-p53 tumor suppressor pathway regulates cell motility by suppression of phosphoinositide 3-kinase and Rac1 GTPase activities. J. Biol. Chem. 278:14414–14419. Hollstein, M., K. Rice, M.S. Greenblatt, T. Soussi, R. Fuchs, T. Sorlie, E. Hovig, B. Smith-Sorensen, R. Montesano, and C.C. Harris. 1994. Database of p53 gene somatic mutations in human tumors and cell lines. Nucleic Acids Res. 22:3551–3555. Ishizaki, T., M. Naito, K. Fujisawa, M. Maekawa, N. Watanabe, Y. Saito, and S. Narumiya. 1997. p160ROCK, a Rho-associated coiled-coil forming protein kinase, works downstream of Rho and induces focal adhesions. FEBS Lett. 404:118–124. Ishizaki, T., M. Uehata, I. Tamechika, J. Keel, K. Nonomura, M. Maekawa, and S. Narumiya. 2000. Pharmacological properties of Y-27632, a specifi c inhibitor of rho-associated kinases. Mol. Pharmacol. 57:976–983. Jaffe, A.B., and A. Hall. 2002. Rho GTPases in transformation and metastasis. Adv. Cancer Res. 84:57–80. 30 JCB • VOLUME 178 • NUMBER 1 • 2007

Journal

The Journal of Cell Biology – Pubmed Central

Published: Jul 2, 2007

There are no references for this article.