If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Supported by National Cancer Institute, Grant CA83182. To whom correspondence should be addressed: Cold Spring Harbor Laboratory, P.O. Box 100, Cold Spring Harbor, NY 11724. Tel.: 516-367-8838; Fax: 516-367-8815
* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Cellular transformation by v-Src is believed to be caused by aberrant activation of signaling pathways that are normally regulated by cellular Src. Using normal rat kidney cells expressing a temperature-sensitive mutant of v-Src, we examined the role of the Raf/MEK/ERK, phosphatidylinositol 3-kinase/Akt, and Rho pathways in morphological transformation and cytoskeletal changes induced by v-Src. Activation of v-Src elicited a loss of actin stress fibers and focal contacts. A decrease in the phosphorylation level of cofilin was detected upon v-Src activation, which is indicative of attenuated Rho function. Inhibition of MEK using U0126 prevented v-Src-induced disruption of the cytoskeleton as well as dephosphorylation of cofilin, whereas treatment with a phosphatidylinositol 3-kinase inhibitor had no protective effect. In normal rat kidney cells stably transformed by v-Src, we found that the chronic activation of MEK induces down-regulation of ROCK expression, thereby uncoupling Rho from stress fiber formation. Taken together, these results establish MEK as an effector of v-Src-induced cytoskeleton disruption, participating in v-Src-induced antagonism of the cellular function of Rho.
extracellular regulated protein kinase
mitogen-activated protein kinase/extracellular signal-regulated kinase kinase
normal rat kidney
temperature-sensitive normal rat kidney
green fluorescence protein
Organization of actin filaments is controlled by the Rho family of small GTPases, Rho, Rac, and Cdc42 (
). ROCK increases the phosphorylation of myosin light chains by directly phosphorylating myosin light chains and negatively regulating myosin phosphatase and thus increases actomyosin-based contractility (
In cells transformed by the product of the Rous sarcoma virus, the v-Src tyrosine kinase, the shape and cytoskeletal architecture are dramatically altered, with a disruption of actin stress fibers and focal contacts and the formation of dot-like actin-associated adhesions called podosomes. This disruption of the cytoskeleton is believed to contribute to several aspects of the transformed phenotype, including adhesion-independent cell growth and increased migration abilities (
). However, the mechanism by which v-Src disrupts actin stress fibers and focal contacts is not clear. Stimulation of a Rho-GAP (GTPase-activating protein) activity was reported upon v-Src activation in chicken embryo fibroblasts (
), suggesting that v-Src-induced cytoskeleton disruption is linked to inhibition of Rho activity. In addition, an active mutant form of Rho was shown to restore stress fibers and adhesion plaques in v-Src-transformed fibroblasts (
). Despite these experiments, the relationship between Rho-mediated stress fiber assembly and v-Src-induced morphological transformation, as well as the signaling pathways mediating v-Src effects on the cytoskeleton, have not been clearly identified.
Expression of v-Src results in tyrosine phosphorylation of numerous substrates and simultaneous activation of several signaling pathways, including the Ras/MEK and PI3K pathways (
). Both pathways have been implicated in cytoskeleton regulation in normal cells. Activation of PI3K is necessary for Rac- and Rho-mediated rearrangements of the actin cytoskeleton induced by platelet-derived growth factor and other growth factors (
). ERK activity has been implicated in cell migration and was shown to directly phosphorylate myosin light chain kinase, suggesting a role for ERK in the regulation of cytoskeleton and focal contact dynamics (
), raising the possibility that both pathways participate in cellular adhesion dynamic during transformation.
We report here that activation of MEK, but not PI3K, is required for the morphological changes and reorganization of the actin cytoskeleton induced by v-Src expression in rat fibroblasts. We also demonstrate that v-Src-induced cytoskeleton disruption results from inactivation of Rho-mediated stress fiber assembly, and we provide evidence that MEK contributes to this inactivation.
Antibodies and Reagents
The following antibodies were used: mAb anti-vinculin (V-9131; Sigma), mAb anti-β-actin (A-5441; Sigma), mAb anti-c-Myc (M-5546; Sigma), mAb anti-phospho-ERK1/2 (sc-7383; Santa Cruz Biotechnology, Inc.), polyclonal anti-phospho-Akt (9271; Cell Signaling Technology), mAb anti-Src (05-183; Upstate Biotechnology, Inc., Lake Placid, NY), polyclonal anti-phospho-Src (07-020; Upstate Biotechnology), mAbs anti-ROCKI and anti-ROCKII (R81520 and R54520; Transduction Laboratories), polyclonal anti-cofilin (ACFL02; Cytoskeleton), rabbit serum recognizing cofilin when phosphorylated by LIMK (a gift from J. R. Bamburg, Colorado State University, Fort Collins, CO), and mAb anti-hemagglutinin tag (12CA5; produced and purified by the Antibody Facility of Cold Spring Harbor Laboratory). Secondary antibodies Cy3-conjugated goat anti-mouse and goat anti-rabbit IgG were purchased from Jackson ImmunoResearch Laboratories. Oregon Green-conjugated phalloidin was from Molecular Probes, Inc. (Eugene, OR). All chemicals and reagents were obtained from Sigma, unless otherwise indicated, and all tissue culture reagents were from Life Technologies.
Cell Culture and Drug Treatments
Normal NRK cells were from ATCC (NRK ATCC CRL 1570). NRK/4.435 are Rous sarcoma virus-transformed NRK cells (
). All cells were maintained in Dulbecco's modified Eagle's medium containing 5% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin in a humidified air, 5% CO2 atmosphere. Cultures of tsNRK/src cells were started at 37 °C for 24 h before transfer to the experimental temperature of 40 or 33 °C for restrictive or permissive growth, respectively.
U0126 (25 μm), PD98059 (50 μm), and LY294002 (10 μm) (all from Calbiochem) resuspended in Me2SO or Me2SO alone as a control were used to pretreat tsNRK/src cells for 1 h before transferring from restrictive to permissive temperature or directly added to NRK or NRK/4.435 cells. After various times, cells were fixed for immunofluorescence or lysed for Western blotting.
Expression Vectors and Transient Transfections
Constitutively active RhoV14 was described previously (
). The pEF-Bos-Myc-Rho-kinase/CAT construct, encoding Myc-tagged activated Rho-kinase, and the pCAG-Myc-160-ROCK construct, encoding wild-type ROCK, were kindly provided by K. Kaibuchi (Nara Institute of Science and Technology, Nara, Japan) and S. Narumyia (Kyoto University Faculty of Medicine, Kyoto, Japan), respectively. Wild-type LIMK1 (GFP-LIMK) and constitutively active LIMK1 (GFP-LIMK-KL) constructs were a generous gift of O. Bernard (Walter and Eliza Hall Institute of Medical Research, Victoria, Australia). Cells plated onto glass coverslips were transfected with a total of 2 μg of expression plasmid per well, using the LipofectAMINE PLUS reagent (Invitrogen) according to the manufacturer's instructions. Coverslips were harvested 24 h post-transfection and processed for immunofluorescence.
Cells grown on glass coverslips were fixed with 3% paraformaldehyde, permeabilized with 0.2% Triton X-100, and then blocked with 1% bovine serum albumin at room temperature. Incubations with primary antibodies against vinculin (1:400), hemagglutinin tag (1:800), or Myc tag (1:500) were conducted at room temperature for 1 h, followed by incubation with Cy3-conjugated secondary antibodies (1:500) for 45 min. To stain actin, fixed cells were incubated with Oregon Green- or Texas Red-conjugated phalloidin. Cells were finally stained with 4′6-diamidino-2-phenylindole and coverslips were mounted using Prolong Antifade (Molecular Probes). Samples were examined, and pictures were acquired on a Zeiss Axiophot microscope equipped with a Photometrics SenSys cooled CCD camera using Image 2.0.5 software (Oncor). All photographs were taken at the same magnification.
Western Blot Analysis
Cells were washed with ice-cold PBS containing 1 mm sodium orthovanadate before direct extraction in 2% SDS Laemmli sample buffer. Lysates were clarified by centrifugation (16,000 × g, 15 min at 4 °C), and protein concentrations were measured by bicinchoninic acid protein assay (Bio-Rad). Equal amounts of proteins were resolved by SDS-PAGE and transferred to nitrocellulose membrane (Schleicher & Schuell). After incubation with primary antibodies, followed by appropriate horseradish peroxidase-conjugated secondary antibodies, the immunoreactive bands were detected by chemiluminescence (PerkinElmer Life Sciences) according to the manufacturer's instructions.
Rho-GTP Pull-down Assay
Measurement of GTP-bound Rho was performed using the Rho activation assay kit (Upstate Biotechnology), following the manufacturer's instructions. Briefly, the RhoA-binding domain of rhotekin expressed as a glutathione S-transferase fusion protein was used to affinity-precipitate GTP-bound Rho from cells lysed in 50 mm Tris, pH 7.2, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 500 mm NaCl, 10 mm MgCl2, and a mixture of protease inhibitors (Roche Molecular Biochemicals). Precipitated Rho-GTP was then detected by immunoblot analysis, using a polyclonal anti-Rho (-A, -B, -C) antibody (Upstate Biotechnology).
Inactivation of Rho-dependent Stress Fiber Formation in NRK Cells Transformed by v-Src
NRK cells infected with a temperature-sensitive mutant of v-Src (tsNRK/src cells) (
) were used to study the modifications in cytoskeleton organization that occur during transformation of fibroblasts by the v-Src oncogene. Cells were shifted from the restrictive temperature of 40 °C to the permissive temperature, 33 °C, at which the v-Src protein is stabilized and activated, and their morphology was examined after 24 h by indirect immunofluorescence (Fig.1A). tsNRK/src cells maintained at 40 °C exhibited a characteristic fibroblastic morphology, with well developed actin stress fibers and vinculin-containing focal adhesions (Fig. 1A). A striking loss of these structures was observed upon v-Src activation, 24 h after the shift at 33 °C. Actin filaments accumulated in patches at the cell periphery, whereas vinculin was distributed throughout the cytoplasm (Fig. 1A).
Stress fiber formation is under the control of the Rho family of small GTPases. More specifically, activation of the Rho-ROCK-LIMK pathway leads to stress fiber assembly by both activation of actomyosin contractility and suppression of the actin-severing activity of cofilin (
). A Rho-GTP pull-down assay was performed to measure the activity of endogenous Rho. There was little difference between Rho-GTP levels in tsNRK/src cells maintained at 40 °C, versus cells shifted at 33 °C (Fig. 1B), suggesting that the loss of stress fibers induced by v-Src was not associated with a direct down-regulation of Rho activity. Next, we assessed the activation status of the Rho-ROCK-LIMK pathway, by measuring the level of phosphorylation of cofilin, which is the terminal effector of this cascade (
). Using a phosphospecific antibody, we found that cofilin phosphorylation decreased in a time-dependent manner in cells shifted to 33 °C, compared with cells maintained at 40 °C (Fig. 1B). Taken together, these results suggest that activity of the Rho-ROCK-LIMK pathway was decreased upon v-Src activation, downstream of Rho.
To further address whether cytoskeleton disruption induced by v-Src involves inactivation of Rho-mediated stress fiber assembly, we transiently transfected tsNRK/src cells with constitutively active mutants of Rho, ROCK and LIMK. All of these constructs further increased stress fiber assembly in cells maintained at 40 °C (not shown). In cells transformed by v-Src (i.e. shifted to 33 °C), expression of these mutants overcame the effects of v-Src on the cytoskeleton (Fig. 2). RhoV14 induced a thin network of actin filaments, whereas cells transfected with Rho-kinase/CAT (
), our data provide the first demonstration that expression of active mutants of ROCK and LIMK suppresses v-Src-induced loss of stress fibers. Taken together, these results indicate that in NRK cells, the activated v-Src tyrosine kinase antagonizes the activity of the Rho-ROCK-LIMK pathway and perturbs the actin assembly/disassembly cycle as a consequence.
Activation of MEK Is Required for v-Src-induced Morphological Transformation and Inactivation of the Rho-ROCK-LIMK Pathway
Activation of v-Src induces several signaling pathways, among which are the MEK/ERK and PI3K/Akt pathways. To determine whether these pathways were implicated in v-Src-induced morphological transformation, tsNRK/src cells were shifted from the restrictive temperature of 40 °C to the permissive temperature, 33 °C, in the presence of 25 μm U0126 (a MEK inhibitor), 10 μm LY294002 (a PI3K inhibitor), or Me2SO alone as a control. Upon shifting to 33 °C, both ERK and Akt were phosphorylated, demonstrating that the MEK/ERK and PI3K/Akt pathways were activated in tsNRK/src cells (Fig.3A). In the presence of U0126, ERK phosphorylation was inhibited, whereas Akt activation was unaffected. Conversely, treatment with LY294002 led to a complete inhibition of Akt phosphorylation, whereas the ERK pathway remained activated. As shown in Fig. 3A, neither the level of expression of v-Src nor its activation, as assessed by Western blotting with an antiserum specific for the autophosphorylated Tyr416 of v-Src, were significantly altered by U0126 or LY294002. These results demonstrate that U0126 and LY294002 can be used to specifically inhibit the MEK/ERK and PI3K pathways, respectively, without any major effects on v-Src expression or activity.
Morphology of the cells was examined by indirect immunofluorescence after 24 h of treatment with the different inhibitors. As described for Fig. 1A, tsNRK/src cells maintained at 40 °C exhibited well developed actin stress fibers and focal adhesions, whereas cells shifted at 33 °C were devoid of these structures (Fig. 3B). Inhibition of MEK by U0126, before and throughout the shift to permissive temperature, prevented the disruption of stress fibers and focal contacts (Fig. 3B). To eliminate the possibility that U0126 may protect actin stress fibers and focal contacts by acting on MKK5 rather than MEK1/2, we used the structurally unrelated MEK inhibitor PD098059, which was shown to have no inhibitory effects toward MKK5 below 100 μm (
). Consistent with the effects being MEK-dependent, treatment of tsNRK/src cells with 50 μm PD098059 led to effects similar to those of U0126 (data not shown). Both compounds were used throughout this study, and only the results obtained with U0126 will be described hereafter. By contrast, inhibition of PI3K by a 10 μm concentration of the LY294002 compound had no protective effect, and cells still exhibited a transformed morphology (Fig. 3B). Taken together, these results indicated that morphological transformation induced by v-Src is mediated by a MEK-dependent pathway but does not require activation of PI3K.
The data presented in Fig. 3A demonstrated that the MEK inhibitor did not inactivate the v-Src protein itself and rather suggested that it interfered with a v-Src-induced signaling pathway, leading to cytoskeleton disruption. Therefore, we tested whether MEK was implicated in v-Src-induced inactivation of Rho-dependent signaling. The level of cofilin phosphorylation was measured in cells shifted to 33 °C in the presence of U0126. Dephosphorylation of cofilin was significantly prevented by treatment with U0126 (Fig.4), suggesting that the ability of this compound to prevent the loss of stress fibers is correlated with its ability to prevent inactivation of the Rho-ROCK-LIMK pathway upon v-Src activation.
Similar results were obtained with an established v-Src-transformed cell line (NRK/4.435), when the inhibitors were added after transformation had become established. Immunofluorescence labeling showed that the punctated actin staining characteristic of NRK/4.435 cells was reorganized into thick bundles of actin stress fibers, associated with vinculin-containing focal contacts, upon treatment with 25 μm U0126 (Fig.5A). By contrast, cells incubated with 20 μm LY294002 still exhibited a disrupted actin cytoskeleton, despite some accumulation of actin at the cell-cell boundaries. Restoration of the actin cytoskeleton upon MEK inhibition was correlated with induction of cofilin phosphorylation (Fig.5B), suggesting that the ROCK-LIMK-cofilin pathway was functionally restored. Taken together, these results establish MEK as an effector of v-Src-induced morphological transformation, participating in v-Src-mediated inactivation of the Rho-ROCK-LIMK pathway.
Activation of MEK Uncouples Rho from Its Downstream Effector, ROCK
To determine how MEK participates in Rho signaling inactivation upon v-Src induction, we assessed the level of expression of the different components of the cascade by Western blot. We found that expression of both ROCKI and ROCKII was greatly decreased in the established NRK/4.435 cells compared with untransformed NRK cells (Fig.6), thus interrupting the Rho-ROCK-LIMK signaling cascade. Upon treatment with U0126, ROCK expression returned to a level similar to that seen in parental NRK cells (Fig. 6), consistent with the ability of this compound to induce cofilin phosphorylation and to restore stress fiber assembly (Fig. 5). To demonstrate the requirement for ROCK activity in the MEK inhibitor effects, we treated NRK/4.435 cells with U0126 in the presence of Y27632, an inhibitor of ROCK. Under these conditions, treatment with U0126 failed to restore stress fibers, thus demonstrating that its effects are mediated by the ROCK pathway (not shown). These results demonstrate that chronic activation of the MEK pathway in v-Src-transformed cells uncouples Rho from stress fiber formation by down-regulating expression of ROCK, a downstream component of the cascade. In tsNRK/src cells, no significant differences were detected in the levels of expression of the different effectors of Rho, between cells maintained at 40 °C or shifted to 33 °C for 24h or 48h (not shown). However, we found that overexpression of wild-type ROCK or wild-type LIMK led to stress fiber formation in cells maintained at 40 °C as well as in cells shifted to 33 °C (Fig.7). In addition to the results presented in Fig. 1, this experiment further shows that Rho is still active in the v-Src-transformed cells (i.e. at 33 °C) but uncoupled from its downstream effectors.
The v-Src oncoprotein severely perturbs the dynamic assembly and disassembly of actin stress fibers and focal adhesions by mechanisms that are not yet fully understood. The regulation of stress fiber and focal adhesion formation is of particular interest due to the putative role of these structures in the regulation of cell growth (
). The conditional temperature-sensitive mutants of v-Src are therefore useful tools to study the early events associated with disruption of the organized actin cytoskeleton. The findings reported here indicate that activation of the MEK pathway, together with inactivation of Rho-dependent signaling, is required for the loss of stress fibers and dissolution of focal contacts induced by v-Src. Furthermore, we demonstrate that activated MEK mediates inactivation of Rho-dependent signaling.
MEK/ERK are targeted to focal adhesions upon v-Src activation (
) and are therefore likely mediators of v-Src effects on the actin cytoskeleton. In fact, we found that upon v-Src activation, MEK inhibitor-treated tsNRK/src cells did not undergo the complete reorganization of actin filaments characteristic of v-Src-transformed cells. They retained stress fibers and focal contacts and did not form the numerous dotlike podosomes seen in control cells shifted to 33 °C in the absence of inhibitors. In this study, we also demonstrate that after inhibition of MEK in NRK cells already transformed by v-Src (NRK/4.435), cells undergo a phenotypic reversion, during which the focal adhesion-associated protein vinculin relocated from cytoplasm to cell-extracellular matrix contacts, whereas actin was assembled into bundles of stress fibers. These results show that MEK is implicated in both induction of the v-Src-transformed phenotype and maintenance of this phenotype once transformation is established. Our results contrast with what has been observed in chicken embryo fibroblasts, in which morphological transformation by v-Src was not blocked by inhibition of either PI3K or Ras/ERK activity (
), our indirect immunofluorescence analysis reveal more profound modifications in cell morphology and cytoskeleton organization and demonstrate implication of MEK in this process.
Modifications in cytoskeleton organization are often indicative of altered Rho function, and in fact, inactivation of the cellular function of Rho has been previously suggested as a mechanism leading to cytoskeleton disorganization upon v-Src expression (
). Even if the involvement of mDia cannot be excluded, our transfection experiments clearly demonstrate that activation of the Rho-ROCK-LIMK pathway is sufficient to restore stress fiber assembly in v-Src-transformed cells, since expression of active mutants of ROCK or LIMK suppresses v-Src-induced loss of stress fibers (Fig. 3). To assess the activation status of the Rho pathway, we measured the level of phosphorylation of cofilin. Activation of ROCK by Rho leads to activation of the kinase activity of LIMK and thereby phosphorylation of cofilin (
). Our Western blot analysis of the cofilin phosphorylation level therefore provides a relevant measure of the activation status of the Rho-ROCK-LIMK pathway and demonstrates inactivation of Rho-dependent signaling upon v-Src expression in NRK cells. At the same time, no decrease in the level of Rho-GTP was found in cells transformed by v-Src (i.e. at 33 °C), thus demonstrating that the Rho-ROCK-LIMK pathway was interrupted downstream of Rho. Taken together, these results demonstrate that v-Src-induced inactivation of the Rho-ROCK-LIMK pathway is required for cytoskeleton disruption in NRK cells.
Direct phosphorylation of cytoskeletal proteins could also be a mechanism for v-Src-induced cytoskeleton disruption. Indeed, several proteins localized in focal contacts are tyrosine-phosphorylated in v-Src-transformed cells, including vinculin and focal adhesion kinase (
). However, we demonstrated that MEK inhibitors are able to prevent the alterations of the cytoskeleton induced by v-Src without affecting the kinase activity of v-Src (Fig. 3A). From these data, we reasoned that the morphological changes induced by v-Src were more likely to be mediated by the Rho pathway and that inhibition of MEK was interrupting the signal from v-Src to Rho. Indeed, we demonstrated that MEK participates in v-Src-induced inactivation of the Rho-ROCK-LIMK pathway, since inhibition of MEK during v-Src activation prevented dephosphorylation of cofilin, consistent with the ability of this compound to prevent the loss of stress fibers. In cells stably transformed by v-Src (NRK/4.435), we identified the mechanism by which MEK mediates v-Src-induced inactivation of Rho signaling. We found that constitutive activation of MEK leads to down-regulation of both ROCKI and ROCKII expression, thus interrupting the Rho-ROCK-LIMK cascade. We have previously shown that the very same mechanism leads to cytoskeleton disruption in response to oncogenic Ras (
). But it is likely that the decreased expression of ROCK is the result of a selection in response to high levels of ERK signaling, during the establishment of the cell lines, since the level of expression of ROCK did not change during acute transformation of tsNRK/src cells. However, transfection of wild-type ROCK induced stress fiber formation in tsNRK/src cells, even at 33 °C, suggesting that Rho is uncoupled from its downstream effector, even if the level of expression of ROCK did not change. This result raises the possibility that the endogenous ROCK may have been relocated upon v-Src activation, thereby becoming inaccessible to Rho. Such a mechanism has been recently demonstrated in Ras-transformed fibroblasts (
In conclusion, these results establish MEK as an effector of v-Src-induced morphological transformation, participating in v-Src-mediated inactivation of the Rho-ROCK-LIMK pathway. This work, together with other studies (
), demonstrates that two functionally different oncogenes, namely the small GTPase Ras and the tyrosine kinase v-Src, use the same signaling pathway to achieve morphological transformation of fibroblasts, i.e.inactivation of Rho-mediated stress fiber assembly through activation of MEK (Fig. 8). Further characterization of the pathway linking MEK to Rho-dependent signaling is therefore a crucial step toward understanding the morphological effects of these oncogenes, and experiments are in progress to elucidate this link, both in Ras-transformed and v-Src-transformed cells.
We thank Dr. J. R. Bamburg, Dr. T. Miller, Dr. O. Bernard, and Dr. S. Narumiya and Dr. K. Kaibuchi for generous gifts of phosphocofilin antibody, tsNRK/src cells, LIMK constructs, and ROCK constructs, respectively. We thank E. Julien for critical comments and helpful suggestions regarding this manuscript.