RhoA- and RhoD-dependent regulatory switch of ... - The FASEB Journal

RhoA- and RhoD-dependent regulatory switch of G␣ subunit signaling by PAR-1 receptors in cellular invasion QUANG-DE´ NGUYEN, SANDRINE FAIVRE, ERIK BRUYNEEL,* CHRISTINE RIVAT, MINORU SETO,† TAKESHI ENDO,‡ MARC MAREEL,* SHAHIN EMAMI, AND CHRISTIAN GESPACH1 INSERM U482, Signal Transduction and Cellular Functions in Diabetes and Digestive Cancers, Hoˆpital Saint-Antoine, 75571 Paris Cedex 12, France; *The Laboratory of Experimental Cancerology, Ghent University Hospital, B-9000 Ghent, Belgium; †Asahi Chemical Industry Company, Samejima, Fuji, Shizuoka 416-0934, Japan; and ‡Chiba University, Department of Biology, 1-33 Yayoicho, Inageku, Chiba, Chiba 263-8522, Japan Thrombin and proteinase-activated receptors (PAR) specifically regulate several functions that markedly enhance the transformation phenotype such as inflammation, cell proliferation, tumor growth, and metastasis. We recently reported that thrombin inhibits cellular invasion induced by src, hepatocyte growth factor (HGF), and leptin in kidney and colonic epithelial cells via predominant activation of the pertussis toxin (PTx) -sensitive G-proteins G␣o/G␣i. We provide pharmacological and biochemical evidence that in the presence of PTx, PAR-1 induced cellular invasion through G␣12/G␣13- and RhoA/Rho kinase (ROCK) -dependent signaling. However, inhibition of the endogenous small GTPase RhoA by the C3 exoenzyme, dominant-negative N19-RhoA, activated G26VRhoD, and activators of the nitric oxide/cGMP pathways conferred invasive activity to PAR-1 via a signaling cascade using G␣q, phospholipase C (PLC), Ca2ⴙ/ calmodulin myosin light chain kinase (CaM-MLCK), and phosphorylation of MLC. We found that cellular invasion induced by the src oncogene is abrogated by inhibitors of the RhoA/ROCK pathway and is independent of PLC/CaM-MLCK signaling. Our data demonstrate that the RhoA and RhoD small GTPases are acting as a molecular switch of cellular invasion and reveal a novel critical mechanism by which PAR-1 bypass G␣o/i and RhoA inhibition via differential coupling to heterotrimeric G-proteins linked to divergent or convergent biological responses. Our data also indicate that Rho GTPases and ROCK mediate a srcdependent invasion signal in kidney and colonic cancer cells. We conclude that dynamic regulation of Rho GTPases activation and inactivation by oncogenes, growth factors, cGMP-inducing agents, and adhesion molecules can initiate convergent invasion signals controlled by the thrombin PAR-1 in cancer cells.— Nguyen, Q.-D., Faivre, S., Bruyneel, E., Rivat, C., Seto, M., Endo, T., Mareel, M., Emami, S., Gespach, C. RhoA- and RhoD-dependent regulatory switch of G␣ subunit signaling by PAR-1 receptors in cellular invasion. FASEB J. 16, 565–576 (2002) ABSTRACT

0892-6638/02/0016-0565 © FASEB

Key Words: colon cancer progression 䡠 metastatic potential 䡠 guanylate cyclases 䡠 src 䡠 rac 䡠 myosin light chain kinase 䡠 guanylin 䡠 cGMP

The serine protease thrombin elicits critical biological responses in the central nervous system and peripheral tissues by acting on a wide variety of epithelial and nonepithelial cells, including mesenchymal and vascular smooth muscle cells, bone marrow, and platelets (1, 2). Thrombin plays a pivotal role in many biological functions, including embryonic development, inflammation, and proliferative responses associated with angiogenesis and cancer progression. Thrombin’s actions are primarily mediated through its interaction with two G-protein-coupled receptors (1): the proteinase-activated receptors (PAR-1 and -3). PAR-2, a possible trypsin receptor, is not thrombin activated. PAR-4 lacks a thrombin binding hirudin sequence and is activated by proteases other than thrombin, such as cathepsin G in human platelets. Thrombin regulates cell proliferation and tumor growth in human solid tumors by its interaction with PAR-1, a serpentine transmembrane receptor coupled to several G␣-proteins, including the PTx-sensitive G␣o/i subunits (and coupled to the PTx-insensitive G␣12/G␣13 and G␣q subunits) (1, 3–5). We recently reported that PAR-1 inhibits cellular invasion induced by src, hepatocyte growth factor (HGF), and leptin in kidney and colonic epithelial cells via predominant activation of pertussis toxin (PTx) -sensitive G-proteins G␣o/G␣i (6). In contrast, recent data indicate that thrombin receptors are preferentially expressed in highly metastatic human breast tumors and promote Matrigel invasion by mammary cancer cell lines (3). Accordingly, we presented evidence that the G␤␥ dimers liberated from activated heterotrimeric 1 Correspondence: INSERM Unit U482, Hoˆpital Saint-Antoine, 75571 Paris Cedex 12, France. E-mail: [email protected]

565

G-proteins downstream the platelet-activating factor receptor (PAF-R) and thrombin PAR-1 receptors are invasion promoters (6, 7). Mechanisms underlying the differential regulation of cellular invasion through activation of these PTx-sensitive and -insensitive G␣ and G␤␥ subunits by activated PAR-1 thrombin receptors are largely unknown. Thrombin receptors and the heterotrimeric G-proteins G␣12 and G␣13 participate in cell transformation through stimulation of Rho guanine nucleotide exchange factors and downstream activation of the small GTPase RhoA (8 –13). We therefore tested whether Rho GTPase inhibitors, including C3T exoenzyme transferase (14), a dominant-negative form of RhoA, and an activated form of the RhoA inhibitor RhoD (15, 16), might alter the regulation of cellular invasion by thrombin. The molecular identity of the thrombin receptor involved in regulating cellular invasion in kidney and colonic epithelial cells transformed by the src oncogene was characterized, using specific agonists of the PAR subtypes. The consequences of stable expression of an antisense PAR-1 cDNA vector (3) on invasiveness of kidney MDCKts.src cells were also analyzed. Results were compared to the negative invasion pathways monitored by PAF-R, a G-protein-coupled receptor linked to PTx-sensitive responses (6, 7). We found that Rho inhibition relieves the G␣o/i-mediated inhibition of cellular invasion by PAR-1. We used pharmacological and biochemical methods to help determine the invasion pathways connected with PAR-1 and its associated PTx-insensitive G-protein subunits G␣12/13 and G␣q. We next attempted to identify the elements that are differentially activated downstream of PAR-1 and G␣12/13 (the RhoA small GTPase and its effector, Rho kinase [ROCK]) or downstream of PAR-1 and G␣q (phospholipase C-␤ and the Ca2⫹/calmodulin-dependent myosin light chain kinase [CaM-MLCK]) in kidney and colonic epithelial cells (17–23). Our data strongly suggest that these two separable proinvasive pathways induced by PAR-1 converge on MLC phosphorylation through differential commutation to G␣12/G␣13 and G␣q signaling. This commutation of G␣ subunit signaling is generated by functional neutralization of the RhoA small GTPase by pharmacological and biochemical agents (including C3T exoenzyme, dominant-negative form of RhoA) and by several signaling pathways controlled by physiological regulators: the RhoA inhibitor RhoD and agonists of the nitric oxide/cGMP/protein kinase G (PKG) cascades that use soluble and particulate guanylate cyclases.

MATERIALS AND METHODS Cell culture, expression vectors, and generation of cell lines Human colonic epithelial cells PCmsrc and HCT8/S11, canine kidney epithelial cells MDCK, MDCKp110*, MDCKts.src, and MDCKT23 cells expressing the mutant small GTPases encoding the dominant-negative form of RhoA(N19) and Rac1(N17) under the tetracycline-repressible trans-activator 566

Vol. 16

April 2002

have been described (7, 15, 24). Expression of these Rho-like GTPases was induced by removing doxycycline (DOX) from the culture media (Sigma, Saint-Quentin Fallavier, France) for 24 h (N17Rac1) or 40 h (N19RhoA), as described (15). The PAR-1 antisense cDNA in pcDNA3.1 and the pCMVmyc vectors encoding the constitutive active (G26V) or negative (T31K) forms of the small GTPase RhoD (DCRhoD and DNRhoD) have been described (3, 16). The GTPase-deficient forms of rat G␣q-Q209L (AG␣q), G␣12(Q229L), and G␣13(Q226L) were cloned into the pcDNA3.1 expression vector (25). MDCK and HCT8/S11 cell lines were stably transfected using 3 ␮g of expression plasmids and the LipofectAMINE Plus reagent, as described previously (24). Control transfections were performed using corresponding empty vectors. Cells were selected for 2 wk in 1 mg/ml neomycin and individual colonies were ring-cloned as individual clones or pooled for analysis of ectopic expression of the transgenes (24). Antibodies, immunoblotting analysis, and compounds Cultured cells were lysed in cold Nonidet P-40 in buffer with protease/phosphatase inhibitors (24). From 50 to 200 ␮g protein was subjected to immunoblotting using the following primary antibodies: the anti-thrombin receptor monoclonal antibody (1:4000) from Biodesign Int. (Kennebunk, ME) and the anti-mouse IgG-HRP conjugate (1:3000) from Amersham Pharmacia Biotech (Orsay, France). The polyclonal antibody pAb specific for G␣13 (1:1000) was from Dr. P. Sternweis (University of Texas, Dallas, TX); pAbs raised against G␣12 (S-20, 1:100) and G␣q (E-17, 1:200) were from Santa Cruz Biotechnologies (TEBU, Le Perray-en-Yvelines, France). For MLC phosphorylation assay, cultured MDCKts.src cells were deprived of serum overnight in the presence or absence of C3T and incubated with TRAP. Cultured cells were then homogenized at 4°C in RIPA buffer containing 0.1 mg/ml phenylmethylsulfonyl fluoride, 100 ␮M benzamine, and 100 mM Na3VO4 as protease and phosphatase inhibitors. Proteins were resolved in 15% SDS-PAGE gels and transferred to PVDF membranes (Amersham). Membranes were blocked overnight in Tris-buffered saline (20 mM Tris-HCl, pH 8, 150 mM NaCl) containing 5% dried skimmed milk. The blots were probed for 4 h at room temperature with the anti ser-P MLC mAb (1:100) as described (23). Densitometry of the autoradiography was performed using a Biocom software. The PAR-2 and PAR-4 agonists were from Neosystem (Strasbourg, France); TRAP (SFLLRN) was from Bachem Biochemie (Voisins-le-Bretonneux, France); thrombin, PAF, LPA, PTx, and SNP were from Sigma; cathepsin G, the PI3⬘-kinase inhibitor wortmannin, KT5926, and U73122 were from Calbiochem (Meudon, France); the PAR-1 agonist TR1– 41 was from Dr. Marc Barnard (University of Massachusetts Medical Center (Worcester, MA); and the ROCK inhibitor Y27632 was kindly provided by Yoshitomi Pharmaceutical Industries Ltd. (Osaka, Japan). Clostridium botulinum exoenzyme C3 transferase (C3T), which ADP-ribosylates and inactivates the small GTPase RhoA, and the Rho activator CNF1 (Escherichia coli cytotoxic necrotizing factor 1) that permanently activates the p21 protein (26) were a generous gift from Dr. Gilles Flatau (INSERM U452, Nice, France). Human guanylin was from Alexis Biochemicals (COGER, Paris, France). Invasion assays Collagen invasion assay was performed as described (27). Colonic and kidney epithelial cells were cultured for 24 h atop collagen gels in the presence of effectors. Kidney

The FASEB Journal

NGUYEN ET AL.

MDCKts.src cells display an invasive phenotype at the permissive temperature of 35°C for src activity and are not invasive at the restrictive temperature of 40°C. After transfer of the activated c-src oncogene into the premalignant human colonic PC/AA/C1 cell line, PCmsrc cells became tumorigenic in the athymic nude mice and were invasive upon addition of HGF (6). Unless otherwise indicated, values presented are mean ⫾ se of at least three experiments. Where appropriate, data were analyzed by the unpaired Student’s t test.

RESULTS We set out to determine the relative contribution of thrombin and PAR agonists to the inhibition of cellular invasion induced by src and HGF in kidney MDCKts.src and colonic PCmsrc epithelial cells (6). Within the range of physiological concentrations, thrombin dosedependently (0.5–5 nM) inhibited invasion of collagen gels in the two models (Fig. 1A). This effect was mimicked by the PAR-1 agonists TRAP (TR 42–55) and TR 1– 41 at concentrations ranging from 1 to 10 ␮M (28). The inhibitory potency of TRAP was observed at the IC50 ⫽ 2–3 ␮M TRAP. In contrast, the PAR-2

agonist SLIGRL and PAR-4 agonist GYPGQV were ineffective for up to 1 mM. As expected, stable expression of the PAR-1 antisense cDNA (PAR-1/AS in clones 2 and 5) considerably reduced receptor levels detected in control MDCKts.src cells transfected by the empty vector pcDNA3.1 (Fig. 1B) and prevented (at similar levels) the TRAP-induced inhibition of invasion in src-transformed cells. This PAR-1/AS cDNA had no effect on the negative control exerted by PAF in MDCKts.src cells activated by the src oncogene (Fig. 1B). The pharmacological specificity of the PAR-1 receptor involved in the negative regulation of invasion was further demonstrated by its sensitivity to the neutrophil granule serine protease cathepsin G that is liberated at sites of inflammation (4). Cathepsin G (0.4 ␮M) blocked thrombin inhibition of HGF-induced invasiveness in PCmsrc cells, but had no effect on the positive or negative invasion pathways controlled by the Met receptor agonist HGF and PAF-R (not shown). Several authors have demonstrated that PAR-1 couple to several G-proteins, including the PTx-sensitive G␣o/G␣i subunits, and to the PTx-insensitive G␣12/ G␣13 and G␣q subunits (1, 4). Because the heterotri-

Figure 1. Inhibition of cellular invasion by PAR-1 in kidney and colonic epithelial cells MDCKts.src and PCmsrc transformed by the src oncogene. A) Effects of the PAR-1 agonists thrombin, TRAP, and TR1– 41, and the PAR-2 and PAR-4 agonists (SLIGRL and GYPGQV, respectively) on cellular invasion induced by src in MDCKts.src cells and HGF (10 U/ml) in PCmsrc cells cultured for 24 h atop collagen type I gels. B) Effects of stably transfected PAR-1 antisense cDNA (PAR-1/AS) on TRAP- and PAF-induced inhibition of cellular invasion in MDCKts.src cells cultured for 24 h at the permissive temperature 35°C for src activation. Cells were treated by the indicated concentrations of TRAP (upper panel) or 0.1 ␮M PAF (lower panel). Western blot analysis showed 80 –90% reduction of PAR-1 protein expression (90 kDa band) in clones 2 and 5 (C2 and C5) vs. sham-transfected MDCKts.src cells (Cont), using the empty vector pcDNA3.1. Data are means ⫾ se from 3– 4 experiments. Rho-DEPENDENT ACTIVITY OF THE G-PROTEIN RECEPTOR PAR-1

567

meric G-proteins G␣12 and G␣13 participate in cell transformation through downstream activation of the small GTPase RhoA (8 –13), we tested whether the Rho GTPase inhibitor C3T might alter regulation of cellular invasion by PAR-1. Rho activation promotes actin stress fiber assembly and focal adhesion formation, common responses to cell adhesion and migration (29, 30). This action of Rho is mediated at least in part by the Rho-associated kinase ROCK that phosphorylates and inactivates myosin phosphatase (18, 21), leading to increased contractility and stress fiber formation. We found that C3T had no effect on cellular invasion induced by the activated form of PI3⬘-kinase in MDCKp110* cells (7), but completely reversed the stimulation produced by the src oncogene in MDCKts.src cells incubated at 35°C (Fig. 2A). Such data are consistent with the notion that RhoA activation is necessary for induction of the invasive phenotype monitored by src. Most important, association of the two invasion inhibitors TRAP and C3T produced an inverse effect by which TRAP, but not PAF, induced invasion when Rho activity was neutralized by C3T. Similar results were obtained in PCmsrc cells challenged with the association of C3T plus thrombin in the presence or absence of HGF (Fig. 2A). In this case, both thrombin and C3T inhibited cellular invasion induced by HGF in PCmsrc cells, whereas their combination (thrombin⫹C3T) reversed the inhibition. Similar induction was observed in PCmsrc cells treated with TRAP and C3T (not shown). Furthermore, induction of cellular invasion by thrombin in PCmsrc cells incubated in the presence of HGF and C3T was PTx insensitive, suggesting that PAR-1 is acting through G␣q signaling when the Rho GTPases are inactivated by C3T. Consistent with this possibility, we observed that in the presence of the Rho inhibitor C3T, thrombin dose-dependently stimulated invasion in PCmsrc cells (Fig. 2B, upper panel) within the same concentration range (0.5–5 nM) that was effective in inhibiting src and HGF/Met invasiveness in MDCKts.src and PCmsrc cells in the absence of C3T (Fig. 1A). We next sought to determine the biochemical relevance of our data using MDCK epithelial cells transfected by the dominant-negative form of RhoA, the N19RhoA mutant (15). Thrombin (0.5–5 nM) similarly induced invasion in MDCKT23-DNRhoA cells cultured for 40 h in the absence of DOX (Fig. 2B, lower panel). The proinvasive activity of TRAP was not induced in MDCKT23-DNRhoA cells grown in the continuous presence of 20 ng/ml DOX (invasion index⫽0.5%), whereas the invasion promoter leptin was effective under these conditions (invasion index⫽7.5%), in agreement with our previous studies (31). We next examined the possible contribution of the Rac1 GTPase in cellular invasion promoted by PAR-1 in MDCKT23-DNRac cells exposed to TRAP and C3T (Fig. 2C, upper panel). After induction of the DNRac1(N17) mutant in MDCKT23-DNRac cells grown in the absence of DOX for 24 h, cellular invasion induced by leptin was abolished but persisted in 568

Vol. 16

April 2002

the presence of TRAP combined with C3T. Furthermore, cellular invasion was not induced by TRAP in MDCKT23-DNRac cells expressing the dominant-negative Rac1(N17) mutant. As control, MDCKT23-DNRac cells were continuously grown in the presence of the DOX repressor (20 ng/ml) to block expression of the dominant-negative form of Rac1. Under these conditions (Fig. 2C, upper panel), cellular invasion was induced in MDCKT23-DNRac cells cultured for 24 h in the continuous presence of DOX and leptin (100 ng/ml) or TRAP (10 ␮M) combined with C3T (3 ␮g/ml). Using the same approach, we determined whether the endogenous RhoA inhibitor RhoD (16) can interfere with PAR-1 signaling. Introduction of the constitutively active form of G26VRhoD in MDCKts.srcDCRhoD cells (Pools 1 and 2) induced the functional competence of PAR-1 on cellular invasion (Fig. 2C, lower panel). In contrast, antagonism of RhoA by DCRhoD abolished cellular invasion induced by the RhoA-dependent signaling pathways controlled by src, HGF, and intestinal trefoil factor ITF (24) and did not exert any permissive action for the PAF-R (Fig. 2C, lower panel). We noticed that, as a control, the dominant-negative form of RhoD (T31KRhoD) did not induce the invasive potential of PAR-1 in MDCKts.src cells incubated in the presence of TRAP (not shown). PAR-1 are now suspected to signal through the G␣q subunit in kidney and colonic epithelial cells exposed to C3T or stably transfected with DNRhoA and DCRhoD. We next examined the invasive potential of this G␣-protein subunit toward the action of pharmacological inhibitors acting on downstream signaling elements of either G␣q (phospholipase C (PLC-␤) and Ca2⫹/calmodulin-dependent myosin light chain kinase (CaM-MLCK) or G␣12/G␣13 (Rho and Rho-associated kinase p160(ROCK) (17–23). It is held that increased intracellular [Ca2⫹]i levels result in the activation of calmodulin, which activates MLCK and consequent phosphorylation of the regulatory light chain of myosin II, RMLC (21). Phosphorylation of RMLC at Ser-19 by MLCK in turn regulates the activity of actin-activated myosin ATPase and myosin filament formation that is believed to promote the stability and contractility of the actin cytoskeleton (23, 32). Thus, Ca2⫹-dependent RMLC phosphorylation plays an important role in epithelial cell contraction and locomotion associated with wound healing and local invasion in normal and pathological processes, such as tumor progression. As shown in Fig. 3A, cellular invasion promoted by the constitutively activated form of G␣q (AG␣q-Q209L) or TRAP combined with C3T (not shown) was fully reversed by the inhibitors of PLC-␤ (U73122) and CaM-MLCK (KT5926), but was resistant to C3 exoenzyme and ROCK inhibition by the pyridine derivative Y27632 (22, 33). Similar responses and pharmacological profile were obtained in human colonic epithelial cells HCT8/S11 stably transfected by the RhoA antagonist DCRhoD and challenged with TRAP (not shown). These invasion pathways were still blocked by TRAP

The FASEB Journal

NGUYEN ET AL.

Figure 2. Evidence for dual effects of Rho inhibition and PAR-1 activation on cellular invasion induced by oncogenes and epigenetic factors in kidney and colonic epithelial cells. A) Effects of the Rho inhibitor C3 exoenzyme (C3T, 3 ␮g/ml) alone or combined with TRAP (10 ␮M), thrombin (10 nM), PAF (0.1 ␮M), and PTx (200 ng/ml) on invasiveness caused by activated p110 PI3⬘-K␣, src, and HGF/Met in MDCKp110* cells, MDCKts.src, and PCmsrc cells incubated for 24 h. B) Activation of invasion by thrombin in colonic PCmsrc cells cultured for 24 h in the presence or absence of 3 ␮g/ml C3T (upper panel) and kidney MDCKT23-DNRhoA cells expressing a dominantnegative form of RhoA GTPase (lower panel) under control of the tetracycline repressible trans-activator. After removal of the doxycycline repressor (⫺DOX) for 40 h, thrombin dose-dependently increased cellular invasion in MDCKT23-DNRhoA cells (lower panel). As controls, we used the same cell population grown continuously in the presence of 20 ng/ml doxycycline (DOX) to repress DNRhoA expression; thrombin was ineffective in inducing invasion of collagen type I gels. C) Upper panel: effects of leptin (100 ng/ml), 10 ␮M TRAP alone, or combined with 3 ␮g/ml C3T on the 24 h invasion assay after induction of the DNRac1(N17) mutant in MDCKT23-DNRac cells grown in the absence of doxycycline for 24 h (⫺DOX). As controls, MDCKT23-DNRac cells were continuously grown in the presence of the DOX repressor (20 ng/ml) to block expression of the dominant-negative form of Rac1 (⫹DOX). Lower panel: induction of invasion by TRAP (10 ␮M) in stably transfected MDCKts.src-DCRhoD cells expressing the constitutively activated form of the RhoA inhibitor RhoD harboring a myc epitope. Pools 1 and 2 (P1/P2) were selected by immunoblotting using the mAb against the myc epitope. DCRhoD-transfected cells were stimulated for 24 h with HGF (10 u/ml), intestinal trefoil factor (ITF, 100 nM), TRAP, and PAF (0.1 ␮M) and incubated at the nonpermissive (40°C) or permissive temperature for src activation (35°C). Data are means ⫾ se from 3– 4 experiments.

Rho-DEPENDENT ACTIVITY OF THE G-PROTEIN RECEPTOR PAR-1

569

Figure 3. Determination of cellular invasion by activated forms of the G␣q and G␣12/13 subunits, Rho activation by CNF1, and stimulation of PAR-1 signaling in kidney epithelial cells exposed to pertussis toxin or C3 exoenzyme. A) Introduction of the constitutively activated form of the G-protein AG␣q in MDCKts.srcAG␣q cells cultured for 24 h at 40°C is associated with acquisition of invasiveness via a PLC/CaM-MLCK-dependent and Rho/ROCK-independent pathway. B) Introduction of the constitutively activated form of the G-protein AG␣13 in MDCKts.src-AG␣13 cells (upper panel) cultured for 24 h at 40°C is associated with acquisition of invasiveness via a Rho/ROCK-dependent and PLC/CaM-MLCK-independent pathway. Similar data were obtained in MDCKts.src cells treated with the Rho activator CNF1 (1 nM) for 24 h at 40°C (lower panel). C) Upper panel: activation of cellular invasion by TRAP (10 ␮M) through PAR-1, G␣12/G␣13, Rho GTPases, and ROCK is revealed in MDCK cells after inhibition of G␣o/i signaling by PTx (200 ng/ml). Lower panel: activation of cellular invasion by src and the combination C3T ⫹ TRAP in MDCKts.src cells cultured for 24 h at 35°C. Invasion caused by either AG␣q, AG␣13, CNF1, src, or PAR-1 activation is determined in the presence or absence of the pharmacological inhibitors of Rho (C3T, 3 ␮g/ml), ROCK (Y27632, 10 ␮M), phospholipase C-␤ (U73122, 1 ␮M), CaM-MLCK (KT5926, 20 nM), and the G-protein receptor agonists TRAP (10 ␮M) or PAF (0.1 ␮M). Inserts: positive MDCKts.src clones expressing either AG␣q (C2 and C3) or AG␣13 (C2 and C3) subunits (41 kDa) by immunoblot analysis (WB: Western blot). Control MDCKts.src cells (Cont) transfected by the control empty vector pcDNA3 are not invasive at the nonpermissive temperature 40°C for src activation. Data are means ⫾ se from 3– 4 experiments.

and PAF in MDCKts.src-AG␣q cells, suggesting that PAF-R and PAR-1 exert a dominant-negative control on invasiveness induced by the G␣q subunit when endogenous RhoA is not inhibited. 570

Vol. 16

April 2002

In contrast, stable expression of the activated, GTPase-deficient form Q229L of G␣13 (AG␣13) induced constitutive invasion (Fig. 3B, upper panel), which was fully reversed by the RhoA and ROCK inhibitors (C3T

The FASEB Journal

NGUYEN ET AL.

and Y27632) but insensitive to the PLC-␤ and CaMMLCK inhibitors (U-73122 and KT5926). Similar pharmacological control was obtained in MDCKts.src cells stably transfected with the GTPase-deficient form (Q226L) of G␣12 (not shown). Therefore, we next examined the role of the Rho GTPase-activating protein CNF1 that permanently activates p21 Rho by deamidation of glutamine 63 (26). By activating Rho and ROCK, CNF1 promotes the assembly of the actin stress fibers, focal adhesions, phosphorylation of focal adhesion kinase, and cell contractility (26). As shown in Fig. 3B (lower panel), PAF and TRAP impaired CNF1induced cellular invasion in MDCKts.src cells cultured for 24 h at 40°C, further indicating that PAF-R and PAR-1 exert a dominant-negative control on cellular invasion induced by G␣12/13 and the Rho GTPases. We observed that CNF1-induced invasion was abrogated by the ROCK inhibitor Y27632 via a PLC-␤ and CaM-MLCK-independent signaling pathway. Similar data were obtained in the parental MDCK cells and PCmsrc cells (not shown). The data support our starting hypothesis (6) that PAR-1 inhibit cellular invasion through activation of the PTx-sensitive G␣o/i subunits in kidney and colonic cancer cells, as confirmed by other recent studies in breast cancer cells (34). Accordingly, inhibition of the G␣o/G␣i subunits by PTx in parental MDCK cells resulted in promotion of cellular invasion in the presence of TRAP, but not PAF (Fig. 3C, upper panel). Cells stimulated with TRAP in the presence of PTx were blocked by the inhibitors of the G␣12/G␣13 signaling pathway (C3T, Y27632), but not by KT5926, acting on the downstream kinase of the G␣q/PLC-␤/CaM-MLCK cascade. Similar invasion induction was observed in MDCKts.src cells, PCmsrc, and HCT8/S11 cells treated with TRAP in the presence of PTx (not shown). We checked that PTx cannot reverse the inhibition of cellular invasion induced by the constitutively activated, GTPase-deficient form of the PTx-sensitive G-protein G␣i3(Q204L) subunit in stably transfected MDCKts.srcAG␣i3 cells (6) treated with the invasion promoter HGF (not shown). It becomes apparent that in the presence of PTx, the PAR-1 signal through G␣12/G␣13 mediated activation of ROCK to induce cellular invasion. Deactivation of the G␣o/i proteins by PTx in the present study might be induced physiologically by the regulators of G-protein signaling (RGS) that are GTPase-activating proteins (GAP) for these G␣ subunits (35). Thus, some RGS such as RGS4 that are G␣q- and G␣o/i-GAP can determine convergent PAR-1 signaling through G␣12/G␣13. Because src is a downstream target of the thrombin receptor PAR-1, we next investigated the possible contribution of the PLC-␤/CaM-MLCK and Rho/ROCK pathways in cellular invasion induced by the src oncogene in MDCKts.src cells incubated at the permissive temperature 35°C. As shown in Fig. 3C (lower panel), src-induced invasion of collagen gels was not altered by inhibition of the downstream elements of the G␣q subunit cascade by U73122 (PLC-␤) and KT5926 (CaMRho-DEPENDENT ACTIVITY OF THE G-PROTEIN RECEPTOR PAR-1

MLCK), but was abolished by the ROCK inhibitor Y27632. In contrast, cellular invasion induced by activated PAR-1 in MDCKts.src cells incubated at 35°C in the presence of the C3 exoenzyme was blocked by U73122 and KT5926, but not by the ROCK inhibitor Y27632. This profile agrees with neutralization of the src oncogenic pathway by C3T and with mobilization of a RhoA/ROCK-independent invasion pathway using the PLC-␤ and CaM-MLCK cascade in kidney and colonic epithelial cells treated by TRAP ⫹ C3T (and other Rho inhibitors), as shown in Fig. 2 and below. Collectively, the present data indicate that neutralization of RhoA by C3T, DNRhoA, and DCRhoD initiates a new PAR-1 invasion pathway that circumvents G␣o/i and Rho inhibition by using the KT5926-sensitive G␣q/ PLC-␤/CaM-MLCK cascade (Fig. 4). On the other hand, suppression of the negative control exerted by the G␣o/i subunits by PTx is sufficient to reveal the invasion potential of PAR-1 through G␣12/G␣13 signaling (Fig. 4). Moreover, activation of RhoA through PAR-1, G␣12/13, and CNF1 led to the inhibition of MLC phosphatase (Fig. 4) via the action of ROCK (18), suggesting that this mechanism contributes to RMLC phosphorylation. Several signaling elements downstream ROCK, including ezrin, adducin, and LIMkinase, generate cellular functions involved in cellular adherence, motility, and invasion (36). This proposed model constitutes a key link to the signaling competence of PAR-1 in tumor cell invasion according to the data presented by Even-Ram et al., where PAR-1 is presented as a positive mediator of cellular invasion in human breast cancer cells (3), and our data on intestinal and kidney epithelial cells, where PAR-1 is dominantly coupled to negative invasion pathways, using PTx-sensitive G-proteins G␣o/i (6), in agreement with recent data by Kamath et al. (34). To confirm the validity of this proposed model (Fig. 4), we first explored the interactions between PAR-1 signaling and the cyclic GMP-dependent protein kinase pathway (cGMP-PKGI␣) that inactivates RhoA through phosphorylation on Ser-188 and inhibition of membrane attachment of this small GTPase (37, 38). Consistent with our data presented in parental and srctransformed MDCK and PC/AA/C1 cells (Fig. 2 and Fig. 3C), we checked that the combination between the C3 exoenzyme (3 ␮g/ml) and TRAP (10 ␮M) induced a proinvasive activity in HCT8/S11 cells established from a human sporadic colon cancer, as shown in Fig. 5A (invasion index⫽8.7⫾1.2%, n⫽4 experiments). At concentrations ranging from 0.5 to 50 ␮M, TRAP alone was ineffective. Invasiveness of HCT8/S11 cells induced by PAR-1 in the presence of the Rho inhibitor C3T was fully reversed by the PI3⬘-kinase inhibitor wortmannin at 10 nM (not shown). This strengthens the notion that PI3⬘-kinases play a critical role in cellular invasion and tumor progression (7, 24, 25, 31, 41, 42). As shown in Fig. 5A, activation of the soluble guanylate cyclase sGC (39) by the nitric oxide (NO) donor sodium nitroprussiate SNP at 10 ␮M (40) revealed the invasion potential of PAR-1 in human colonic HCT8/ 571

Figure 4. Proposed model depicting the negative (⫺) and positive (⫹) invasion pathways induced by PAR-1 and controlled by the PTx-sensitive G-proteins G␣o/i and the RhoA/RhoD small GTPAses. The thrombin PAR-1 receptors exert dominant-negative control on cellular invasion through the PTx-sensitive G-proteins G␣o/i (⫺). Neutralization of the G␣o/i subunits by PTx leads to commutation of PAR-1 signaling with the positive invasion pathway using G␣12/13 subunits, Rho, and ROCK (commutation 1 via green arrow 1). Neutralization of the RhoA GTPase (by C3T, a dominant-negative form of RhoA; DNRhoA, the RhoA inhibitor RhoD; and signaling elements activating the NO/cGMP/PKG cascades) bypasses Rho inhibition via an alternative positive invasion pathway, using G␣q/PLC-␤/CaM-MLCK (commutation 2 via green arrows 1⫹2). These two positive PTx-insensitive PAR-1 invasion pathways converge on RMLC phosphorylation and regulation of the actin cytoskeleton. This signaling network is valid in premalignant and srctransformed kidney and colonic cancer cells (MDCK, PC/AA/C1, MDCKts.src, PCmsrc) and in colon cancer cells HCT8/S11 established from a human sporadic tumor.

S11 cells according to a similar dose-response relationship for G-protein-coupled receptor activation (0.5–50 ␮M TRAP, EC50⫽4 ␮M TRAP) through a ROCKindependent and PLC-␤/CaM-MLCK-dependent signaling pathway. At concentrations ranging from 0.1 to 10 ␮M, the NO donor SNP alone was ineffective. Acting on its own receptor, the membrane-bound GC-C, and downstream through PKG activation (43), the cGMPinducing agent guanylin transmits a PAR-1-dependent positive invasion signal via the G␣q/CaM-MLCK cascade in HCT8/S11 cells (Fig. 5B). Therefore, the endogenous intestinal factor guanylin (0.01–50 ␮M) is proficient at converting negative invasion pathways mediated by PAR-1 and G␣o/i subunits into positive signals (EC50⫽0.2 ␮M guanylin determined in the presence of 10 ␮M TRAP), confirming our results on Rho inhibition by C3 exoenzyme, RhoD, and the NO donor SNP. In the presence of 50 ␮M guanylin, TRAP dose-dependently (0.5–50 ␮M TRAP) stimulated invasiveness of HCT8/S11 cells according to potency EC50 ⫽ 1.7 ⫾ 0.5 ␮M TRAP (n⫽3 experiments). At concentrations ranging from 0.01 to 50 ␮M, guanylin alone had no effect on cellular invasion in HCT8/S11 cells (not shown). Our data are consistent with the idea that cGMP kinase blocks PAR-1 signaling through phosphorylation and neutralization of the PTx-sensitive G␣o/i subunits (43) involved in the negative control of cellular invasion (6). Guanylin and uroguanylin are endogenous ligands and activators of the colonic epithelial cell GC-C, the intestinal receptor of the heatstable enterotoxins, with similar potency (EC50⫽0.3 ␮M guanylin) for stimulation of GC-C receptors and cGMP production in the human colonic cancer cell line T84 (39). Second, we compared the ability of the PAR-1 agonist TRAP to induce phosphorylation of RMLC at ser-19 572

Vol. 16

April 2002

(RMLC-P) in serum-starved MDCKts.src cells cultured overnight in the presence or absence of C3 exoenzyme (Fig. 5C). Under control conditions, PAR-1 activation promoted rapid and transient MLC phosphorylation (observed 2 min after the addition of TRAP) and declined thereafter to basal levels at 15 min. In the presence of C3T, TRAP-induced MLC phosphorylation was preserved, but a delayed response was seen. There was a lag period of ⬃10 min after the addition of TRAP before the detection of similar changes in RMLC-P, thereby indicating differential kinetics of CaM-MLCK activation by PAR-1 in the presence or absence of the Rho inhibitor. This observation is compatible with the G␣o/i- and Rho-dependent bifurcation of three separate signaling pathways from activated PAR-1, as shown in Fig. 4. Finally, this proposed model is consistent with the presence of two regulatory systems implicated in immediate and sustained responses for activation of the actomyosin system in human foreskin fibroblasts, using either CaM-MLCK or Rho/ROCK cascades or both (44). Delayed responses controlled by these positive and negative invasion pathways are implicated in controlling expression of target genes involved in cancer progression through nuclear targeting of specific transcription factors (45, 46).

DISCUSSION Acquisition of the invasive phenotype is the hallmark of adenoma-adenocarcinoma conversion during the progression of colonic tumors toward local invasion and metastasis. Identification of positive and negative invasion pathways controlled by oncogenes and epigenetic factors constitute a challenge for the development of efficient therapeutic strategies against cancer progres-

The FASEB Journal

NGUYEN ET AL.

Figure 5. Determination of PAR-1 signaling through G␣q is governed by inhibition of Rho-like GTPases and activation of the NO/cGMP/PKG cascades in kidney and colonic epithelial cells. A) Induction of cellular invasion induced by TRAP (0.5–50 ␮M) in human colonic epithelial cells HCT8/S11 incubated for 24 h in the presence of the nitric oxide donor SNP (10 ␮M) is reversed by the PLC-␤ and CaM-MLCK inhibitors U73122 (1 ␮M) and KT5926 (20 nM), but unaffected by the ROCK inhibitor Y27632 (10 ␮M). Results are compared to cellular invasion measured in the presence of TRAP alone (10 ␮M) or combined with C3T (3 ␮g/ml) in our 24 h invasion assay. As control, we checked that TRAP alone, at concentrations ranging from 0.5 to 50 ␮M, had no effect on cellular invasion in HCT8/S11 cells. Stimulation of cellular invasion is induced in HCT8/S11 cells preincubated overnight in the presence of C3T and challenged for 24 h with the PAR-1 agonist TRAP (10 ␮M): invasion index ⫽ 7.6 ⫾ 0.4% compared with simultaneous addition of 3 ␮g/ml C3T ⫹ 10 ␮M TRAP (10.6%⫾0.4%) in 24 h invasion assay. Data are means ⫾ se from 3– 4 experiments. B) Induction of cellular invasion by TRAP in HCT8/S11 cells exposed for 24 h to various concentrations of the cGMP-inducing agent guanylin (GUA: 0.01–50 ␮M) acting through the intestinal receptor– guanylate cyclase complex GC-C. Cotreatment by 10 ␮M TRAP and 50 ␮M guanylin is insensitive to the ROCK inhibitor Y27632 (10 ␮M). Conversely, invasion induced by TRAP and guanylin is completely reversed by the PLC-␤ and CaM-MLCK inhibitors: 1 ␮M U73122 and 20 nM KT5926. At 50 ␮M, GUA alone was ineffective on cellular activity. Data are means ⫾ se from 3– 4 experiments. C) Kinetics of phosphorylation of the regulatory light chain of myosin II (RMLC) induced by PAR-1 in serum-starved MDCKts.src cells incubated overnight at 40°C in the presence or absence of C3 exoenzyme (C3T, 3 ␮g/ml). Cells were incubated under control conditions (open bars) and stimulated with TRAP (10 ␮M) alone or combined with C3T (filled bars). Equal amounts of cell lysate were analyzed by Western blotting (upper panel) and densitometric analysis (lower panel). Immunoblots show the amount of Ser-19-phosphorylated RMLC (20 kDa) (23). Blots were stripped and reprobed with a total MLC antiserum (Sigma) to check for equivalent expression of MLC. Data are representative of 4 other independent experiments.

sion. We presented evidence here that PAR-1 can be differentially coupled to heterotrimeric G-proteins linked to opposite or convergent biological responses associated with the invasive potential of solid tumors according to the neutralization of the PTx-sensitive G-proteins G␣o/i or inhibition of the RhoA small GTPase, respectively. We identify a new function of the RhoA and RhoD Rho-DEPENDENT ACTIVITY OF THE G-PROTEIN RECEPTOR PAR-1

GTPases acting as molecular switch between the negative and positive signaling activity of PAR-1 in cellular invasion. Commutation of G␣o/i-dependent negative signals (6) into G␣q-dependent positive signals by RhoA inhibition and RhoD activation was found. Thus, Rho-like small GTPases RhoA and RhoD govern the functional activity of PTx-sensitive and -insensitive heterotrimeric G-proteins linked to the serpentine PAR-1, 573

playing a key role in tumor cell growth and metastasis. Most important, we demonstrate that several signaling pathways converge to RhoA inhibition and are potentially involved in such commutation. In their interesting paper, Gilchrist et al. (46) have presented the biological activity of G␣ minigene vectors expressing the carboxyl-terminal sequences of G␣q or G␣13, which are supposed to act specifically as dominant/negative-interfering fragments of G␣ subunits. It is apparent, however, that the G␣q minigene vector exerts only partial inhibition of calcium fluxes generated by thrombin and that all the carboxyl-terminal minigenes encoding G␣i, G␣q, and G␣12/G␣13 fragments inhibit thrombinmediated, mitogen-activated protein kinase activity. Direct demonstration of the involvement of the G␣q subunit or its downstream elements in RhoA/RhoDdependent PAR-1 signaling and tumor cell invasion presented in Fig. 4 might be accomplished after invalidation of the G␣q/G␣11 genes, antisense strategies, or other approaches using specific G␣q/G␣11- and PLC␤-dependent nuclear translocation of tubby (47). Tubby proteins are transcription factors expressed in brain and large intestine for which loss of function leads to obesity (47, 48). Our findings indicate that upstream effectors of the RhoD and NO/cGMP-dependent PKG pathways (16, 37–39), such as guanylin and activators of the NO synthases, constitute potential components of the regulatory switch reported here. Moreover, the E-cadherin-interacting protein p120 catenin was recently shown to disrupt stress fibers and focal adhesion through inhibition of Vav2 (49 –51), a guanine exchange factor with activity for RhoA activation. At junctional complexes, p120 interacts indirectly with the actin cytoskeleton through multimolecular associations comprising the E-cadherin undercoat proteins ␣-, ␤-, and ␥-catenins. Increased cytoplasmic levels of p120 decrease Rho activity, leading to activation of the Rho-like GTPases Cdc42 and Rac, and promote cell migration. Other recent studies have demonstrated that certain extracellular matrix molecules binding fibronectin, such as tenascin-C, may suppress RhoA activation (52), cellular adhesion, spreading and migration of human mammary carcinoma cells. The glycoprotein tenascin-C is present in limited levels in adult animals, but is induced during embryonic development, wound healing, angiogenesis, and inflammatory and neoplastic diseases (53). Stromal expression of tenascin-C is associated with a poor prognosis in colorectal cancer and is a marker for epithelial malignancy in the mammary gland and osteosarcomas (53–55). This suggests the interesting hypothesis that cell– cell and cell–matrix adhesions determine the activation status of PAR-1 through the RhoA inhibitors p120 and tenascin-C, and might contribute to the regulation of complex mechanisms involved in the RhoA/RhoDdependent conversion between the anti- and proinvasive activity of PAR-1. Remodeling of the matrix-substratum at sites of injury and inflammation and during tumor progression contribute to the redistribution and 574

Vol. 16

April 2002

differential activation of signaling molecules involved in cellular migration and invasiveness. Membrane-anchored integrins at adhesion sites are regulating the reorganization of the actin cytoskeleton driven by the Rho family GTPases (56 –59). Cell–substrate adhesion via integrins (41) and PI3⬘-kinases (6, 42) plays a major role in tumor cell growth, motility, and invasiveness. Recent data demonstrate that the immediate cellular responses to integrin engagement lead to RhoA inhibition in an src-dependent manner (60). Thus, a delicate balance between activation and inhibition of the RhoA GTPase can initiate distinct and convergent invasion pathways in cancer cells. Dysregulation by oncogenic signals in the expression, turnover, or structure of the molecular components of the basement membrane allowing cell–substratum interactions (e.g., collagen types I and IV, nidogen, laminin, BM-40/SPARC, tenascin C, fibronectin, and experimental matrix components) therefore play a potential role in cellular invasion, survival, and angiogenesis, crucial components of the metastatic cascade (3, 61– 63). Upstream regulation of Rho GTPases by the Dbl family of guanine nucleotide exchange factors (GEFs) such as p115 and p190 RhoGEFs (10 –13, 64), Rho-GAPs, or G␣q (65) may directly or indirectly influence positive and negative invasion signals (Fig. 4). In light of our data, it will be important to reconsider therapeutic approaches based on pharmacological inhibitors of the Rho GTPases in the adult population with inherited predisposition to cancer or with early colorectal, breast, and prostatic cancers detected by systematic screening. Conversely, our data suggest that pharmacological blockade of the PAR-1 and src proinvasive pathways by the ROCK inhibitor Y27632 can exert a therapeutical role in the progression of breast and colon cancers (66). Src activation is an early and frequent event in human colon cancers, and a mutation in c-src that results in carboxy-terminal truncation of this nonreceptor tyrosine kinase has transforming activities (65). In concert with the actin cytoskeleton, individual RhoA and Rac GTPAses participate in the peripheral localization of src in discrete subcompartments of the cell surface microdomains, at adhesion sites, and in clusters of several signaling molecules involved in oncogenesis (67). We conclude that the RhoA/RhoD-dependent commutation of G␣ subunit signaling by PAR-1 has the potential to initiate adverse effects on the progression of human solid tumors in breast and colon, probably through autocrine and paracrine modulation by upstream regulators of the RhoA/RhoD small GTPases, cellular adhesion factors, and cGMP-dependent signaling pathways. Other RhoA inhibitors such as tenascin (overexpressed in breast and colon cancers) and p120 might contribute to the proinvasive potential of PAR-1 through G␣q signaling, as well as inhibitors of the PTx-sensitive G␣o/i subunits such as the RGS proteins (35). Thus, a complex view of the signaling competence of PAR-1 in cellular invasion is beginning to emerge. The thrombin receptor PAR-1 is now clearly

The FASEB Journal

NGUYEN ET AL.

implicated in cellular invasion and the metastatic potential of solid tumors.

16.

This study was supported by INSERM, GOA (no. 1205126.00), Belgische Federatie tegen Kanker, and the Fortis Bank, Verzekeringen (Brussels, Belgium). We thank Drs. E. Peralta (Harvard University, Cambridge, MA), N. Dhanasekaran (Fels Institute for Cancer Research, Philadelphia, PA), R. Bar-Shavit (Hadassah-Hebrew University Hospital, Jerusalem, Israel), M. Furman (University of Massachusetts Medical Center, Worcester, MA), G. Flatau (INSERM U452, UFR de Me´ decine, Nice, France), and H. Sueoka (Welfide Corporation, Fukuoka, Japan) for providing reagents.

17.

18.

19.

20.

REFERENCES 1. 2. 3.

4. 5.

6.

7.

8.

9. 10.

11.

12.

13.

14. 15.

Coughlin, S. R. (2000) Thrombin signalling and proteaseactivated receptors. Nature (London) 407, 258 –264 Donovan, F. M., and Cunningham, D. D. (1998) Signaling pathways involved in thrombin-induced cell protection. J. Biol. Chem. 273, 12746 –12752 Even-Ram, S., Uziely, B., Cohen, P., Grisaru-Granovsky, S., Maoz, M., Ginzburg, Y., Reich, R., Vlodavsky, I., and Bar-Shavit, R. (1998) Thrombin receptor overexpression in malignant and physiological invasion processes. Nature Med. 4, 909 –914 De´ ry, O., Corvera, C. U., Steinhoff, M., and Bunnett, N. W. (1998) Proteinase-activated receptors: novel mechanisms of signaling by serine proteases. Am. J. Physiol. 274, C1429 –C1452 Wojtukiewicz, M. Z., Tang, D. G., Ben-Josef, E., Renaud, C., Walz, D. A., and Honn, K. V. (1995) Solid tumor cells express functional ‘tethered ligand’ thrombin receptor. Cancer Res. 55, 698 –704 Faivre, S., Re´ gnauld, K., Bruyneel, E., Nguyen, Q.-D., Mareel, M., Emami, S., and Gespach, C. (2001) Suppression of cellular invasion by activated G-proteins subunits G␣o/G␣i1–2-3 and sequestration of G␤␥. Mol. Pharmacol. 60, 363–372 Kotelevets, L., Noe¨ , V., Bruyneel, E., Myssiakine, E., Chastre, E., Mareel, M., and Gespach, C. (1998) Inhibition by plateletactivating factor of Src- and hepatocyte growth factor-dependent invasiveness of intestinal and kidney epithelial cells. Phosphatidylinositol 3⬘-kinase is a critical mediator of tumor invasion. J. Biol. Chem. 273, 14138 –14145 Buhl, A. M., Johnson, N. L., Dhanasekaran, N., and Johnson, G. L. (1995) G␣12 and G␣13 stimulate rho-dependent stress fiber formation and focal adhesion assembly. J. Biol. Chem. 270, 24631–24634 Offermanns, S., Mancino, V., Revel, J. P., and Simon, M. I. (1997) Vascular system defects and impaired cell chemokinesis as a result of G␣13 deficiency. Science 275, 533–536 Hart, M. J., Jiang, X., Kozasa, T., Roscoe, W., Singer, W. D., Gilman, A. G., Sternweis, P. C., and Bollag, G. (1998) Direct stimulation of the guanine nucleotide exchange activity of p115 RhoGEF by G␣13. Science 280, 2112–2114 Mao, J., Yuan, H., Xie, W., and Wu, D. (1998) Guanine nucleotide exchange factor GEF115 specifically mediates activation of Rho and serum response factor by the G protein ␣ subunit G␣13. Proc. Natl. Acad. Sci. USA 95, 12973–12976 Majumdar, M., Seasholtz, T. M., Buckmaster, C., Toksoz, D., and Brown, J. H. (1999) A Rho exchange factor mediates thrombin and G␣12-induced cytoskeletal responses. J. Biol. Chem. 274, 26815–26821 Bhattacharyya, R., and Wedegaertner, P. B. (2000) G␣13 requires palmitoylation for plasma membrane localization, Rhodependent signaling, and promotion of p115-RhoGEF membrane binding. J. Biol. Chem. 275, 14992–14999 Wilde, C., Genth, H., Aktories, K., and Just, I. (2000) Recognition of RhoA by Clostridium botulinum C3 exoenzyme. J. Biol. Chem. 275, 16478 –16483 Jou, T. S., and Nelson, W. J. (1998) Effects of regulated expression of mutant RhoA and Rac1 small GTPases on the development of epithelial (MDCK) cell polarity. J. Cell Biol. 142, 85–100

Rho-DEPENDENT ACTIVITY OF THE G-PROTEIN RECEPTOR PAR-1

21.

22.

23. 24.

25.

26.

27.

28.

29. 30. 31.

32.

33.

34.

Tsubakimoto, K., Matsumoto, K., Abe, H., Ishii, J., Amano, M., Kaibuchi, K., and Endo, T. (1999) Small GTPase RhoD suppresses cell migration and cytokinesis. Oncogene 18, 2431– 2440 Amano, M., Ito, M., Kimura, K., Fukata, Y., Chihara, K., Nakano, T., Matsuura, Y., and Kaibuchi, K. (1996) Phosphorylation and activation of myosin by Rho-associated kinase (Rho-kinase). J. Biol. Chem. 271, 20246 –20249 Kimura, K., Ito, M., Amano, M., Chihara, K., Fukata, Y., Nakafuku, M., Yamamori, B., Feng, J., Nakano, T., Okawa, K., Iwamatsu, A., and Kaibuchi, K. (1996) Regulation of myosin phosphatase by Rho and Rho-associated kinase (Rho-kinase). Science 273, 245–248 Amano, M., Chihara, K., Kimura, K., Fukata, Y., Nakamura, N., Matsuura, Y., and Kaibuchi, K. (1997) Formation of actin stress fibers and focal adhesions enhanced by Rho-kinase. Science 275, 1308 –1311 Kureishi, Y., Kobayashi, S., Amano, M., Kimura, K., Kanaide, H., Nakano, T., Kaibuchi, K., and Ito, M. (1997) Rho-associated kinase directly induces smooth muscle contraction through myosin light chain phosphorylation. J. Biol. Chem. 272, 12257– 12260 Essler, M., Amano, M., Kruse, H. J., Kaibuchi, K., Weber, P. C., and Aepfelbacher, M. (1998) Thrombin inactivates myosin light chain phosphatase via Rho and its target Rho kinase in human endothelial cells. J. Biol. Chem. 273, 21867–21874 Itoh, K., Yoshioka, K., Akedo, H., Uehata, M., Ishizaki, T., and Narumiya, S. (1999) An essential part for Rho-associated kinase in the transcellular invasion of tumor cells. Nature Med. 5, 221–225 Sakurada, K., Seto, M., and Sasaki, Y. (1998) Dynamics of myosin light chain phosphorylation at Ser19 and Thr18/Ser19 in smooth muscle cells in culture. Am. J. Physiol. 274, C1563–C1572 Emami, S., Le Floch, N., Bruyneel, E., Thim, L., May, F., Westley, B., Rio, M. C., Mareel, M., and Gespach, C. (2001) Induction of scattering and cellular invasion by trefoil peptides in src- and RhoA-transformed kidney and colonic epithelial cells. FASEB J. 15, 351–361 Rodrigues, S., Nguyen, Q. D., Faivre, S., Bruyneel, E., Thim, L., Westley, B., May, F., Flatau, G., Mareel, M., Gespach, C., and Emami, S. (2001) Activation of cellular invasion by trefoil peptides and src is mediated by cyclooxygenase- and thromboxane A2 receptor-dependent signaling pathways. FASEB J. 15, 1517–1528 Fiorentini, C., Fabbri, A., Flateau, G., Donelli, G., Matarrese, P., Lemichez, E., Falzano, L., and Boquet, P. (1997) Escherichia coli cytotoxic necrotizing factor 1 (CNF1), a toxin that activates the Rho GTPase. J. Biol. Chem. 272, 19532–19537 Vleminckx, K., Vakaet, L., Jr., Mareel, M., Fiers, W., and Van Roy, F. (1991) Genetic manipulation of E-cadherin expression by epithelial tumor cells reveals an invasion suppressor role. Cell 66, 107–119 Furman, M. I., Liu, L., Benoit, S. E., Becker, R. C., Barnard, M. R., and Michelson, A. D. (1998) The cleaved peptide of the thrombin receptor is a strong platelet agonist. Proc. Natl. Acad. Sci. USA 95, 3082–3087 Chrzanowska-Wodnicka, M., and Burridge, K. (1996) Rhostimulated contractility drives the formation of stress fibers and focal adhesions. J. Cell Biol. 133, 1403–1415 Hall, A. (1998) Rho GTPases and the actin cytoskeleton. Science 279, 509 –514 Attoub, S., Noe, V., Pirola, L., Bruynel, E., Chastre, E., Mareel, M., Wymann, M., and Gespach, C. (2000) Leptin promotes invasiveness of kidney and colonic epithelial cells via phosphoinositide 3-kinase-, Rho-, and Rac-dependent signaling pathways. FASEB J. 14, 2329 –2338 Matsumura, F., Ono, S., Yamakita, Y., Totsukawa, G., and Yamashiro, S. (1998) Specific localisation of serine 19 phosphorylated myosin II during cell locomotion and mitosis of cultured cells. J. Cell Biol. 140, 119 –129 Yoshioka, K., Matsumura, F., Akedo, H., and Itoh, K. (1998) Small GTP-binding protein Rho stimulates the actomyosin system, leading to invasion of tumor cells. J. Biol. Chem. 273, 5146 –5154 Kamath, L., Meydani, A., Foss, F., and Kuliopulos, A. (2001) Signaling from protease-activated receptor-1 inhibits migration and invasion of breast cancer cells. Cancer Res. 61, 5933–5940

575

35. 36. 37.

38.

39.

40.

41.

42.

43.

44. 45. 46.

47. 48. 49. 50.

576

Hepler, J. R. (1999) Emerging roles for RGS proteins in cell signalling. TiPS 20, 376 –382 Fukata, Y., Amano, M., and Kaibuchi, K. (2001) Rho–Rho-kinase pathway in smooth muscle contraction and cytoskeletal reorganization of non-muscle cells. TiPS 22, 32–39 Sauzeau, V., Le Jeune, H., Cario-Toumaniantz, C., Smolenski, A., Lohmann, S. M., Bertoglio, J., Chardin, P., Pacaud, P., and Loirand, G. (2000) Cyclic GMP-dependent protein kinase signaling pathway inhibits RhoA-induced Ca2⫹ sensitization of contraction in vascular smooth muscle. J. Biol. Chem. 275, 21722–21729 Sawada, N., Itoh, H., Yamashita, J., Doi, K., Inoue, M., Masatsugu, K., Fukunaga, Y., Sakaguchi, S., Sone, M., Yamahara, K. I., Yurugi, T., and Nakao, K. (2001) cGMP-dependent protein kinase phosphorylates and inactivates RhoA. Biochem. Biophys. Res. Commun. 280, 798 – 805 Lucas, K. A., Pitari, G. M., Kazerounian, S., Ruiz-Stewart, I., Park, J., Schulz, S., Chepenik, K. P., and Waldman, S. A. (2000) Guanylyl cyclases and signaling by cyclic GMP. Pharmacol. Rev. 52, 375– 413 Launay, J. M., Bondoux, D., Oset-Gasque, M. J., Emami, S., Mutel, V., Haimart, M., and Gespach, C. (1994) Increase of human platelet serotonin uptake by atypical histamine receptors. Am. J. Physiol. 35, R526 –R536 Shaw, L. M., Rabinovitz, I., Wang, H. H. F., Toker, A., and Mercurio, A. M. (1997) Activation of phosphoinositide 3-OH kinase by the ␣6␤4 integrin promotes carcinoma invasion. Cell 91, 949 –960 Barbier, M., Attoub, S., Calvez, R., Laffargue, M., Jarry, A., Mareel, M., Altruda, F., Gespach, C., Wu, D., Lu, B., Hirsch, E., and Wymann, M. P. (2001) Weakening link to colorectal cancer? Nature (London) 413, 796 Pfeifer, A., Nu¨ rnberg, B., Kamm, S., Uhde, M., Schultz, G., Ruth, P., and Hofmann, F. (1995) Cyclic GMP-dependent protein kinase blocks pertussis toxin-sensitive hormone receptor signaling pathways in Chinese hamster ovary cells. J. Biol. Chem. 270, 9052–9059 Katoh, K., Kano, Y., Amano, M., Onishi, H., Kaibuchi, K., and Fujiwara, K. (2001) Rho-kinase-mediated contraction of isolated stress fibers. J. Cell Biol. 153, 569 –583 Marinissen, M. J., and Gutkind, J. S. (2001) G-protein-coupled receptors and signaling networks: emerging paradigms. TiPS 22, 368 –376 Gilchrist, A., Vanhauwe, JF., Li, A., Thomas, T. O., VoynoYasenetskaya, T., and Hamm, H. E. (2001) G␣ minigenes expressing C-terminal peptides serve as specific inhibitors of thrombin-mediated endothelial activation. J. Biol. Chem. 276, 25672–25679 Santaga, S., Boggon, T. J., Baird, C. L., Gomez, C. A., Zhao, J., Shan, W. S., Myszka, D. G., and Shapiro, L. (2001) G-protein signaling through tubby proteins. Science 292, 2041–2050 Noben-Trauth, K., Naggert, J. K., North, M. A., and Nishina, P. M. (1996) A candidate gene for the mouse mutation tubby. Nature (London) 380, 534 –538 Anastasiadis, P. Z., Moon, S. Y., Thoreson, M. A., Mariner, D. J., Crawford, H. C., Zheng, Y., and Reynolds, A. B. (2000) Inhibition of RhoA by p120 catenin. Nature Cell Biol. 2, 637– 644 Grosheva, I., Shtutman, M., Elbaum, M., and Bershadsky, A. D. (2000) p120 catenin affects cell motility via modulation of activity of Rho-family GTPases: a link between cell-cell contact formation and regulation of cell locomotion. J. Cell Sci. 114, 695–707

Vol. 16

April 2002

51. 52. 53.

54. 55.

56. 57. 58.

59. 60. 61.

62.

63. 64.

65.

66. 67.

Noren, N. K., Liu, B. P., Burridge, K., and Kreft, B. (2000) p120 catenin regulates the actin cytoskeleton via Rho family GTPases. J. Cell Biol. 150, 567–579 Wenk, M. B., Midwood, K. S., and Schwarzbauer, J. E. (2000) Tenascin-C suppresses Rho activation. J. Cell Biol. 150, 913–919 Mackie, E. J., Chiquet-Ehrismann, R., Pearson, C. A., Inaguma, Y., Taya, K., Kawarada, Y., and Sakakura, T. (1987) Tenascin is a stromal marker for epithelial malignancy in the mammary gland. Proc. Natl. Acad. Sci. USA 84, 4621– 4625 Mackie, E. J., Halfter, W., and Liverani, D. (1988) Induction of tenascin in healing wounds. J. Cell Biol. 107, 2757–2767 Tanaka, M., Yamakasi, T., Araki, N., Yoshikawa, H., Yoshida, T., Sakakura, T., and Uchida, A. (2000) Clinical significance of tenascin-C expression in osteosarcoma: tenascin-C promotes distant metastases of osteosarcoma. Int. J. Mol. Med. 5, 505–510 Clark, E. A., King, W. G., Brugge, J. S., Symons, M., and Hynes, R. O. (1998) Integrin-mediated signals regulated by members of the Rho family of GTPases. J. Cell Biol. 142, 573–586 Keely, P., Parise, L., and Juliano, R. (1998) Integrins and GTPases in tumour cell growth, motility and invasion. TiCB 8, 101–106 O’Connor, K. L., Nguyen, B. K., and Mercurio, A. M. (2000) RhoA function in lamellae formation and migration is regulated by the ␣6␤4 integrin and cAMP metabolism. J. Cell Biol. 148, 253–258 Ridley, A. (2000) Rho GTPases: integrating integrin signaling. J. Cell Biol. 150, F107–F109 Arthur, W. T., Petch, L. A., and Burridge, K. (2000) Integrin engagement suppress RhoA activity via a c-src-dependent mechanism. Curr. Biol. 10, 719 –722 Chastre, E., Empereur, S., Di Giola, Y., El Mahdani, N., Mareel, M., Vleminckx, K., Van Roy, F., Bex, V., Emami, S., Spandidos, D. A., and Gespach, C. (1993) Neoplastic progression of human and rat intestinal cell lines after transfer of the ras and polyoma middle T oncogenes. Gastroenterology 105, 1776 –1789 Porte, H., Chastre, E., Prevot, S., Nordlinger, B., Empereur, S., Basset, P., Chambon, P., and Gespach, C. (1995) Neoplastic progression of human colorectal cancer is associated with overexpression of the stromelysin-3 and BM-40/SPARC genes. Int. J. Cancer 64, 70 –75 Bradshaw, A. D., and Sage, E. H. (2001) SPARC, a matricellular protein that functions in cellular differentiation and tissue response to injury. J. Clin. Invest. 107, 1049 –1054 Van Horck, F. P. G., Ahmadian, M. R., Haeusler, L. C., Moolenaar, W. H., and Kranenburg, O. (2001) Characterization of p190RhoGEF, a RhoA-specific guanine nucleotide exchange factor that interacts with microtubules. J. Biol. Chem. 276, 4948 – 4956 Sagi, S. A., Seasholtz, T. M., Kobiashvili, M., Wilson, B. A., Toksoz, D., and Brown, J. H. (2001) Physical and functional interactions of G␣q with Rho and its exchange factors. J. Biol. Chem. 276, 15445–15452 Irby, R. B., and Yeatman, T. J. (2000) Role of src expression and activation in human cancer. Oncogene 19, 5636 –5642 Fincham, V. J., Unlu, M., Brunton, V. G., Pitts, J. D., Wyke, J. A., and Frame, M. C. (1996) Translocation of src kinase to the cell periphery is mediated by the actin cytoskeleton under the control of the Rho family of small G proteins. J. Cell Biol. 135, 1551–1564

The FASEB Journal

Received for publication July 17, 2001. Revised for publication October 4, 2001.

NGUYEN ET AL.