Calmodulin Binds to K-Ras, but Not to H- or N-Ras, and Modulates Its Downstream Signaling

Abstract

Activation of Ras induces a variety of cellular responses depending on the specific effector activated and the intensity and amplitude of this activation. We have previously shown that calmodulin is an essential molecule in the down-regulation of the Ras/Raf/MEK/extracellularly regulated kinase (ERK) pathway in cultured fibroblasts and that this is due at least in part to an inhibitory effect of calmodulin on Ras activation. Here we show that inhibition of calmodulin synergizes with diverse stimuli (epidermal growth factor, platelet-derived growth factor, bombesin, or fetal bovine serum) to induce ERK activation. Moreover, even in the absence of any added stimuli, activation of Ras by calmodulin inhibition was observed. To identify the calmodulin-binding protein involved in this process, calmodulin affinity chromatography was performed. We show that Ras and Raf from cellular lysates were able to bind to calmodulin. Furthermore, Ras binding to calmodulin was favored in lysates with large amounts of GTP-bound Ras, and it was Raf independent. Interestingly, only one of the Ras isoforms, K-RasB, was able to bind to calmodulin. Furthermore, calmodulin inhibition preferentially activated K-Ras. Interaction between calmodulin and K-RasB is direct and is inhibited by the calmodulin kinase II calmodulin-binding domain. Thus, GTP-bound K-RasB is a calmodulin-binding protein, and we suggest that this binding may be a key element in the modulation of Ras signaling.

Small GTPases of the Ras superfamily are key regulators of mammalian cell signaling pathways. Among these proteins, the prototypical Ras family members H-, N-, and K-Ras are major players in most extracellular signal-regulated cell decisions, including proliferation, differentiation, survival, and apoptosis (15, 30, 45). Their role in cell transformation and oncogenesis is highlighted by the fact that more than 10% of human cancers harbor point mutations in Ras proteins: K-Ras in the case of colon and pancreatic carcinomas and N-Ras in the case of lymphomas (4, 6). The molecular basis for such a great variety of cell responses controlled by Ras proteins relies on the fact that Ras is able to transduce signals from different extracellular stimuli, including growth factors, hormones, and cell-extracellular matrix contacts, to many downstream effectors (29). These include the serine/threonine kinase Raf, which leads to the activation of the extracellularly regulated kinase (ERK) pathway that enables transcription of many mitogenically regulated genes involved in cell cycle progression (33, 36, 39, 48); the lipid kinase phosphatidylinositol-3-kinase (PI3K), which in turn activates through its second-messenger products protein kinase B (PKB) (also called Akt), a pathway that supplies a survival signal in many cell systems (2, 11, 49); and the nucleotide exchange factors for Ral GTPase, RalGDS, Rlf, and Rlg, which have been suggested to connect Ras with the Rho family member Cdc42 GTPase and thereby to the actin cytoskeleton and the control of cell morphology (59). Other proteins have been described as binding directly to Ras in its GTP-bound active form and may be considered effectors contributing to Ras signaling (41). The high degree of homology between the different Ras isoforms suggested that they would be functionally identical, but evidence pointing to a preferential activation of specific effectors by the different Ras isoforms is accumulating (61). The fact that the diverse Ras isoforms are also located at different membrane microdomains enforces the idea of a distinct functionality and regulation of these proteins (50). Furthermore, experiments with mice knocked out selectively for each one of the Ras isoforms showed that K-Ras, but not H-Ras or N-Ras, is essential for development (27, 58).

As a molecular switch, Ras cycles between a GTP-bound active state and an inactive state when GTP is hydrolyzed to GDP. Many molecules have been described as influencing the Ras GTP-GDP cycle, mainly through two distinct biochemical activities: the guanine nucleotide exchange factors (GEFs), which regulate the replacement of the nucleotide bound to Ras, favoring the GTP-bound active state, and the GTPase-activating proteins (GAPs), which increase Ras's low intrinsic GTPase activity and thereby promote the inactivation of Ras proteins. The present model for Ras activation following extracellular stimulation is based on the recruitment of GEFs to the plasma membrane, where Ras is located, through binding of these proteins to a set of molecular adapters and induction of transient Ras-GTP complexes (5, 14).

Although there has been much effort to understand the mechanisms that lead to Ras activation and the downstream effectors that mediate Ras functions, our present understanding of the molecular mechanisms leading to Ras inactivation following stimulation is modest. However, there must be a correct balance between activation and inhibition to ensure an appropriate signaling output, and many effects relating to the timing and strength of Ras signaling have been described (40). For instance, sustained, high activation of the ERK pathway induces cell cycle arrest in some cell lines and drives cell differentiation in others, while transient activation followed by a sustained but lower level of ERK activity is a common feature of cell proliferation in many systems (28, 44). This dual effect on cell behavior has been shown to be dependent on the levels of p21^cip1, a cyclin-dependent kinase inhibitor that is induced transcriptionally by the ERK pathway, an induction that is dependent on the duration and intensity of ERK signaling (53, 60). Inactivation of ERKs by specific phosphatases which at the same time act as nuclear anchor proteins for ERK1/2 and the regulation of the levels of those phosphatases are now well documented (9, 35, 55), but down-regulation of upstream elements of the pathway is not as well understood. Thus, it is important to achieve comprehensive knowledge of Ras activation, including not only the mechanisms that couple extracellular signals to Ras activation and hence to Ras effector pathways but also the signaling network that tightly regulates the timing of Ras activation and thus the specificity of the signal itself.

The Ca²⁺-binding protein calmodulin (CaM) acts as a second messenger in cellular signal transduction pathways and regulates cell proliferation (25, 32, 38, 52). CaM functions are mediated by its association with specific target proteins called CaM-binding proteins (CaMBPs) whose activity is regulated upon CaM binding (1, 3). CaMBPs include a great variety of proteins, such as CaM-dependent kinase II (CaMKII) and CaMKIV (52), calcineurin (31), spectrin (22), hnRNP A2 (8), and p21^cip1 (57). CaM regulates these proteins and thus a variety of cellular processes such as gene expression, protein translation, and protein phosphorylation. A role of CaM in ERK activation regulation at different levels of the pathway and with different consequences depending on the cellular type has been described. The epidermal growth factor (EGF) receptor is able to bind to CaM, although the function of this interaction is not yet well understood (42, 51). Two Ras GEFs, Ras-GRF and Ras-GRF2, which are expressed mainly in cortical neurons, have been shown to contain IQ motifs that allow their binding to CaM and its activation by Ca²⁺ (20, 21). In PC12 cells, Ca²⁺ and CaM are both necessary for the acute activation of ERKs after TrkA or EGF receptor stimulation. In this case CaM antagonists completely block the initial Raf-1 activation without affecting Ras-GTP levels (16–18). CaM-dependent kinases have been involved in the ERK activation pathway in NG108 cells and in rabbit aortic smooth muscle cells (19, 43). In contrast, we have shown that in cultured fibroblasts, Ca²⁺ and CaM are important for the inactivation of the Ras/Raf/MEK/ERK pathway (7). Inactivation of CaM in serum-starved NIH 3T3 and NRK cells induces activation of the Ras/Raf/MEK/ERK pathway. Furthermore, we have also proved that CaM is essential to inhibit the sustained activation of ERK1/2 after stimulation of these cells by growth factors and thus to attenuate p21^cip1 levels. Thus, CaM could be necessary to allow a proliferative effect of the Ras/Raf/MEK/ERK pathway. In agreement with our results, it has recently been shown that chelation of basal intracellular Ca²⁺ induces an increased and prolonged ERK1/2 activation in mouse embryonic fibroblasts (26). Furthermore, expression of one of the ERK1/2-inactivating phosphatases, MKP1, is Ca²⁺ dependent (12).

Here we provide new data on the contribution of CaM to Ras regulation. We show that inactivation of CaM, even in the absence of any other stimuli, is able to induce Ras activation. Thus, CaM is an important element in the down-regulation of the Ras/Raf/MEK/ERK pathway. We also analyzed the binding of diverse regulatory proteins of this pathway to CaM and showed an interaction of Ras and Raf with CaM. In the case of Ras, this binding was direct and specific for GTP-Ras. Our data also provide evidence for a differential down-regulation of Ras isoforms, since K-RasB was the only Ras isoform able to bind to CaM.

MATERIALS AND METHODS

Cell culture.

NIH 3T3 cells were grown in Dulbecco's minimum essential medium supplemented with 10% donor calf serum and made quiescent by being cultured for 24 h with medium containing 0.5% fetal bovine serum (FBS). Swiss 3T3 cells were maintained in Dulbecco's minimum essential medium supplemented with 10% FBS and made quiescent by incubating 10⁴ cells/cm² until confluence (6 to 8 days) and keeping them in 0.5% FBS medium overnight during the last day and in serum-free medium for the last 3 h. Purified growth factors (EGF, platelet-derived growth factor [PDGF], or bombesin), 10% FBS, or drugs (W12, W13, W7, trifluoroperazine, or geldanamycin) were added directly to the medium, and cells were harvested at the time points indicated in Results.

Gel electrophoresis and immunoblotting.

Cells were lysed in a buffer containing 2% sodium dodecyl sulfate (SDS), 67 mM Tris-HCl (pH 6.8), and 10 mM EDTA and sonicated twice for 10 s. Protein content was measured by the Lowry procedure, using bovine serum albumin as a standard. These cellular extracts or proteins from pull-down or immunoprecipitation experiments were electrophoresed in SDS-polyacrylamide gels essentially as described previously (34). After electrophoresis, the proteins were transferred to Immobilon-P strips for 2 h at 60 V. The sheets were preincubated in Tris-buffered saline (TBS) (20 mM Tris-HCl [pH 7.5], 150 mM NaCl)–0.05% Tween 20–5% defatted milk powder for 1 h at room temperature and then incubated in TBS–0.05% Tween 20–1% bovine serum albumin–0.5% defatted milk powder containing the appropriate antibodies for 1 h at room temperature. The antibodies used were monoclonal antibodies against pan-Ras (Oncogene Science OP40; 1:100 dilution), N-Ras (Santa Cruz sc-31; 1:100 dilution), K-Ras (Santa Cruz sc-30; 1:100 dilution), K-RasA (Santa Cruz sc-522; 1:100 dilution), K-RasB (Santa Cruz sc-521; 1:100 dilution), Sos1 (Transduction Laboratories S-15520; 1:1,000 dilution), Grb2 (a gift from J. Ureña; 1:1,000 dilution), NF1 (NF1-C [rabbit antibodies against the C-terminal 14 amino acids of human NF1]; 1:1,000 dilution), Raf-1 (Transduction Laboratories R-19120; 1:1,000 dilution), MEK (Transduction Laboratories M-17020; 1:1,000 dilution), ERK1/2 (Zymed Laboratories 03-6600; 1:500 dilution), and phospho-PKB (Cell Signaling Technology no. 9276; 1:1,000 dilution) and polyclonal antibodies against H-Ras (Santa Cruz sc-520; 1:100 dilution), p120GAP (Santa Cruz sc-425; 1:500 dilution), phospho-ERK1/2 (Cell Signaling Tech. no. 9101, 1:500 dilution), and PKB (Cell Signaling Technology no. 9272; 1:1,000 dilution). After being washed in TBS–0.05% Tween 20 (three times, 10 min each), the sheets were incubated with a peroxidase-coupled secondary antibody (1:2,000 dilution) (Bio-Rad) for 1 h at room temperature. After incubation, the sheets were washed twice in TBS–0.05% Tween 20 and once in TBS. The reaction was visualized with the ECL (Amersham) or Super-Signal (Pierce) system. Bacterially expressed H-Ras, N-Ras, and K-RasB proteins used as Western blot controls were from Oncogene.

Affinity chromatography with CaM-Sepharose.

Human recombinant CaM was coupled to BrCN-activated Sepharose 4B according to the manufacturer's procedures except that the buffer used for the coupling was 100 mM borate (pH 8.2)–400 mM NaCl–50 μM CaCl₂. For pull-down assays with cellular lysates, cells (5 × 10⁶) were washed twice in ice-cold phosphate-buffered saline, lysed with 0.5 to 1 ml of pull-down buffer (50 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1% [vol/vol] Triton X-100, 1 mM dithiothreitol [DTT]) plus protease and phosphatase inhibitors (0.1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride, 10 mM β-glycerophosphate, 2 μg of aprotinin per ml, and 10 μg of leupeptin per ml) for 30 min at 4°C, and clarified by centrifugation. Lysates (equalized for protein content) were incubated with 30 μl of CaM-Sepharose for 2 h at 4°C in the presence of 0.1 mM CaCl₂ or 1 mM EGTA. The unbound fraction was collected by centrifugation, and the remaining bound fraction was washed four times with pull-down buffer containing CaCl₂ or EGTA. An aliquot (25 to 50 μl) of the unbound fraction and all of the bound fraction were analyzed by electrophoresis and Western blotting. A lysate from NIH 3T3 or Swiss 3T3 cells was always loaded in the same gel as a control for the mobility of each protein. For in vitro binding experiments with purified proteins, these were incubated for 1 h at room temperature with 20 μl of CaM-Sepharose in pull-down buffer (50 μl) but with 300 mM NaCl in the presence of 1 mM CaCl₂ or 5 mM EGTA. The bound and unbound fractions were obtained as indicated above and analyzed by Western blotting or directly by Coomassie blue staining of the gel. For competition experiments, 5 to 10 nmol of CaMKII^290–309 peptide (Sigma) (in 50 μl of pull-down buffer with 1 mM CaCl₂) was preincubated for 20 min with CaM-Sepharose.

Immunoprecipitation.

Immunoprecipitations were performed as described previously (24). Briefly, cells (3 × 10⁷) were lysed at 4°C with pull-down buffer (2 ml) plus 0.1 mM CaCl₂ and protease and phosphatase inhibitors. Lysates were sonicated twice for 10 s at 4°C and clarified by centrifugation at 10,000 × g for 10 min. Half of the lysate was incubated with 4 μg of anti-CaM monoclonal antibody (Upstate Biotechnology, Inc., product no. 05-173), and the other half was incubated with a control nonrelated anti-human mouse antibody for 2 h at 4°C. Protein immunocomplexes were then incubated with 15 μl of protein G-Sepharose (Sigma), collected by centrifugation, and washed four times in pull-down buffer. Immunoprecipitated Ras was then analyzed by SDS–12% polyacrylamide gel electrophoresis and Western blotting using pan-Ras monoclonal antibody. A lysate from NIH 3T3 cells was loaded in the same gel as a control for the mobility of Ras

Purification of Ras proteins and in vitro binding studies.

Ras proteins were purified as glutathione S-transferase (GST) fusion proteins expressed in insect cells (Sf9 cells) following infection with baculoviruses encoding H-RasV12 or K-RasBV12. Briefly, at 3 days postinfection Sf9 cells were collected by centrifugation, washed twice with ice-cold phosphate-buffered saline, and lysed with Sf9 cell lysis buffer (50 mM Tris-HCl [pH 7.5], 100 mM NaCl, 1 mM EDTA, 5 mM MgCl₂, 10% [vol/vol] glycerol, 0.1% [vol/vol] Triton X-100, and protease inhibitors) for 15 min at 4°C. Lysates were sonicated three times for 30 s, clarified by centrifugation at 10,000 × g for 10 min, and then incubated with glutathione-Sepharose beads for 1 h at 4°C. Beads were collected by centrifugation, washed five times with lysis buffer, and then resuspended in exchange buffer (20 mM Tris-HCl [pH 7.5], 50 mM NaCl, 5% [vol/vol] glycerol, 1 mM DTT, 10 mM EDTA, 1 mM GTP or GDP) for 30 min at 30°C to load Ras proteins with GTP or GDP. Loading was terminated by adding MgCl₂ to 15 mM, and then beads were washed in ice-cold lysis buffer. Finally, GST-Ras proteins were eluted with 20 mM glutathione in lysis buffer, dialyzed overnight against 2 liters of dialysis buffer (10 mM Tris-HCl [pH 7.5], 150 mM NaCl, 2 mM MgCl₂, 10% [vol/vol] glycerol, 0.1 mM DTT), and frozen at −80°C. The different isoforms of truncated Ras (from amino acid 1 to 166) were obtained by PCR and cloned into a pGexKG plasmid in order to obtained the different GST-Ras^1–166 fusion proteins. Fusion proteins were expressed in Escherichia coli BCl21 and then purified and loaded with the specific nucleotide as indicated above.

Measurement of Ras activation.

The capacity of Ras-GTP to bind to the Ras-binding domain of Raf-1 (RBD) was used to analyze the amount of active Ras (13). Cells (5 × 10⁶ to 10 × 10⁶) were lysed in the culture dish with Ras extraction buffer (20 mM Tris-HCl [pH 7.5], 2 mM EDTA, 100 mM NaCl, 5 mM MgCl₂, 1% [vol/vol] Triton X-100, 5 mM NaF, 10% [vol/vol] glycerol, 0.5% [vol/vol] 2-mercaptoethanol) plus protease and phosphatase inhibitors. Cleared (10,000 × g) lysate was assayed for protein concentration by the Bradford method, and protein-equalized supernatants were incubated for 2 h at 4°C with glutathione-Sepharose 4B beads precoupled with GST-RBD (1 h, 4°C). Beads were washed four times in the lysis buffer. Bound proteins were solubilized by the addition of 30 μl of Laemmli loading buffer and run on SDS–12.5% polyacrylamide gels. The amount of Ras in the bound fraction was analyzed by Western blotting.

RESULTS

CaM inhibition synergizes with low concentrations of FBS, PDGF, EGF, and Bombesin to induce ERK1/2 activation in Swiss 3T3 cells.

We have previously shown that in NIH 3T3 cells, CaM inhibition is able to synergize with FBS to induce ERK1/2 phosphorylation. In order to analyze which signaling pathways were cooperating with CaM inactivation to lead to this effect on ERK1/2 phosphorylation, Swiss 3T3 cells were stimulated with diverse growth factors known to use different intracellular pathways to activate Ras. The synergism between FBS and CaM inactivation was first analyzed. The anti-CaM drug used was W13, while W12 was used as a control because it is chemically very similar to W13 but has much lower affinity for CaM. W13 has been used extensively to inhibit CaM in cell cultures, and it is known to be highly specific at the doses used in this work (10, 37, 56). Thus, quiescent Swiss 3T3 cells were incubated overnight with serum-free medium or medium containing 0.2, 0.5, or 1% FBS and then treated with W13 (15 μg/ml) in order to inhibit CaM or with the control drug W12 (15 μg/ml) for 30 min, and ERK1/2 phosphorylation was analyzed by Western blotting. As shown in Fig. 1A, while in W12-treated cells ERK1/2 phosphorylation was observed only in the presence of 1% FBS, in W13-treated cells activation was observed with only 0.2% FBS and was increased further in 0.5% FBS. No significant activation of ERK1/2 was observed in W13-treated cells without FBS. To analyze whether the enhancement of ERK1/2 phosphorylation induced by CaM inhibition was dependent on the factor used to activate ERK1/2, low concentrations of purified growth factors instead of FBS were added to the quiescent cells together with the anti-CaM drug. Synergism for ERK1/2 activation was observed with low concentrations of EGF, bombesin, or PDGF. In the presence of W13, 0.5 ng of EGF per ml, 0.5 nM bombesin, and 0.025 nM PDGF were able to induce ERK1/2 phosphorylation, which was very low or not detected in W12-treated cells (Fig. 1B). In order to ensure that the effects observed with W13 were due to CaM inhibition, other anti-CaM drugs were tested. As shown in Fig. 1C, both W7 and trifluoroperazine also induced activation of ERK1/2 at 0.025 nM PDGF.

FIG. 1.

CaM inhibition synergizes with different growth factors to induce ERK1/2 activation. (A) Quiescent Swiss 3T3 cells were incubated overnight with medium containing 0, 0.2, 0.5, or 1% FBS and then treated with the anti-CaM drug W13 (15 μg/ml) or the control drug W12 (15 μg/ml) for 30 min. (B) Quiescent Swiss 3T3 cells were incubated for 30 min with the indicated concentrations of EGF, Bombesin, or PDGF plus W13 (15 μg/ml), W12 (15 μg/ml), or nothing (−). (C) Quiescent Swiss3T3 cells were incubated with PDGF at the indicated concentration and, in the lanes indicated, W7 (25 μM) or trifluoroperazine (TFP) (12.5 and 25 μM) was added. For all panels phosphorylation of ERK1/2 was analyzed by Western blotting using specific anti-P-ERK1/2 antibodies as indicated in Materials and Methods.

CaM inhibition induces Ras activation in the absence of any other stimuli.

We found previously that W13 induces Ras activation in the presence of 0.5% FBS (7). Since synergism with CaM inhibition to induce ERK1/2 phosphorylation was observed independently of the growth factor used, we tested the possibility that anti-CaM drug addition was enough to induce Ras activation in the absence of any other stimuli. As previously shown, CaM inhibition in serum-starved cells in the presence of 0.5% FBS induced an increase of Ras-GTP detected by GST–Raf-1-RBD pull-down analysis. This effect was induced by both W13 and trifluoroperazine (Fig. 2A). Interestingly, when quiescent NIH 3T3 cells incubated in serum-free medium were treated for 5 min with the CaM inhibitor W13 in the absence of any other stimuli, an increase of Ras-GTP was also produced (Fig. 2B). Although the levels of Ras-GTP produced after W13 treatment in the absence of FBS were not as high as those in the presence of 0.5% FBS, a reproducible activation of Ras was observed compared with nontreated or W12-treated cells.

FIG. 2.

CaM inhibition induces Ras activation in the absence of any other stimuli. Subconfluent NIH 3T3 cells were incubated for 24 h with 0.5% FBS (A) or serum-free medium (B), and then PDGF (0.4 nM), W13 (15 μg/ml), W12 (15 μg/ml), trifluoroperazine (TFP) (25 μM), or nothing (−) was added to the medium and left for 5 min. The amount of Ras-GTP was analyzed by pull-down assay with RBD-Sepharose and Western blotting using a pan-Ras antibody. Total Ras was also analyzed by Western blotting directly from an aliquot of the corresponding cell lysate. Results from a representative experiment out of five for each condition are shown.

PKB phosphorylation is not observed following CaM inhibition at low FBS concentration.

The effect of CaM inhibition on another Ras effector pathway, the PI3K/PKB pathway, was analyzed. Serum-starved NIH 3T3 cells were incubated with anti-CaM drugs in medium containing 0.5% FBS for the various time periods. Activation of the PI3K/PKB pathway was analyzed by Western blotting using a phospho-specific anti-PKB antibody. As shown in Fig. 3, whereas W13 treatment induced ERK1/2 phosphorylation, under the same conditions no phosphorylation of PKB was detected, indicating that PKB activation by W13 is at least lower than ERK1/2 activation. In contrast, a positive control showed both ERK1/2 and PKB phosphorylation following activation with 10% FBS for 10 min. Thus, the activation of Ras induced by CaM inhibition preferentially activated the Raf/MEK/ERK pathway.

FIG. 3.

CaM inhibition at low FBS concentration does not lead to PKB phosphorylation. Subconfluent NIH 3T3 cells were incubated for 24 h with medium containing 0.5% FBS, and then W13 (15 μg/ml) was added to the medium and left for 5, 10, 20, 30, and 45 min. A negative control (untreated cells) and a positive control (cells treated for 10 min with 10% FBS) were also loaded in the same gel. Total PKB, phosphorylated PKB (P-PKB), and phospho-ERK1/2 were analyzed by Western blotting using specific antibodies.

Analysis of the interaction between CaM and different proteins of the Ras/Raf/MEK/ERK1/2 pathway.

As the functions of CaM are mediated by its Ca²⁺-dependent association with specific target proteins, the presence of a CaMBP associated with any of the proteins involved in the regulation of the Ras/Raf/MEK/ERK1/2 pathway was analyzed. Cell lysates from NIH 3T3 cells were incubated with CaM-Sepharose in the presence of Ca²⁺ or EGTA. After pulling down the proteins bound to CaM-Sepharose, the presence of the different proteins of the Ras/Raf/MEK/ERK1/2 pathway in the bound and unbound fractions was analyzed by Western blotting. Among the proteins directly involved in the regulation of Ras-GTP levels, Grb2, SOS, p120GAP, and NF1 were analyzed, and none of them was found to bind to CaM in the presence of either Ca²⁺ or EGTA. MEK and ERK1/2 were also found in the unbound fractions. In contrast, Ras and Raf-1 were able to bind to CaM in the presence of Ca²⁺ and not when Ca²⁺ was chelated by EGTA (Fig. 4A). As shown in Fig. 4B, binding of Ras and Raf to CaM-Sepharose was specific, since no binding to control Sepharose was observed. Furthermore, upon loading the cellular extract on CaM-Sepharose, Ras could be eluted specifically with EGTA (5 mM), and almost no Ras remained bound to CaM-Sepharose (Fig. 4C).

FIG. 4.

Ca²⁺-dependent binding of Ras and Raf-1 from cellular lysates to CaM-Sepharose. (A) Cellular lysates (0.5 ml) from NIH 3T3 cells (5 × 10⁶) were incubated with CaM-Sepharose (Seph) in the presence of Ca²⁺ or EGTA as indicated in Materials and Methods. The presence of Ras, Raf-1, MEK, ERK, GRB-2, p120 GAP, Sos1, and NF1 in the bound and unbound fractions was analyzed by Western blotting using specific antibodies. All bound fraction and 50 μl of the unbound fraction were loaded. (B) As in panel A, but half of the cellular lysate was applied in the presence of Ca²⁺ to CaM-Sepharose and half was applied to control Sepharose. The presence of Ras and Raf-1 in the bound fractions was analyzed by Western blotting. (C) Cellular lysate (1 ml) was incubated with CaM-Sepharose as indicated in Materials and Methods in the presence of Ca²⁺. The unbound fraction was collected, and after washing with Ca²⁺-containing buffer, bound proteins were eluted sequentially (E1, E2, and E3) with 40 μl of the same buffer supplied with EGTA (5 mM). Finally, the remaining bound proteins were eluted with SDS-containing buffer (E_SDS). Twenty-five microliters of the unbound fraction and all eluted fraction were loaded onto a SDS-acrylamide gel, and the amount of Ras present in each fraction was analyzed by Western blotting. Pan-Ras antibody was used to detect Ras in all panels.

In order to analyze whether the binding of Ras to CaM was dependent on the nucleotide bound to Ras, the same experiment was performed using lysates of quiescent NIH 3T3 cells that were extracted with pull-down buffer containing 5 mM MgCl₂. Under these conditions nucleotide exchange is inhibited, and most Ras was expected to be loaded with GDP. In this case, Ras was found only in the protein fraction not bound to CaM-Sepharose, while Raf-1 still bound to CaM in a Ca²⁺-dependent manner (Fig. 5A). When no MgCl₂ was added to the cellular lysates, Ras was able to bind CaM. It should be mentioned that under these conditions of cell lysis, Ras-GTP was detected in the lysate by GST–Raf-1-RBD pull-down analysis (Fig. 5B). To explore the possibility that Ras-GTP binding to CaM was dependent on its association with Raf-1, cells were treated for 12 h with 5 or 10 μM geldanamycin. This drug inhibits HSP90 and induces the degradation of Raf-1 (54). Lysates from these cells were mixed with CaM-Sepharose, and the presence of Raf-1 and Ras in the bound and unbound fractions was analyzed. As shown in Fig. 5C, confirming its induced degradation, Raf-1 was almost undetectable either in the bound or in the unbound fraction in cells treated with geldanamycin. In contrast, Ras was still able to bind to CaM in a Ca²⁺-dependent manner even in the absence of Raf-1. In order to corroborate the interaction between Ras and CaM, a coimmunoprecipitation assay was performed. As shown in Fig. 6, Ras was detected by Western blotting in the immunoprecipitates of NIH 3T3 cellular lysates with anti-CaM monoclonal antibody but not with a nonrelated anti-mouse control antibody. From these results, it can be concluded (i) that Raf-1 is associated with a CaMBP or is itself a CaMBP and (ii) that Ras-GTP but not Ras-GDP is associated with a CaMBP distinct from Raf or is itself a CaMBP.

FIG. 5.

Binding of Ras-GTP from cellular lysates to CaM-Sepharose (Seph) independently of the presence of Raf. (A) Subconfluent NIH 3T3 cells (5 × 10⁶) were incubated for 24 h with medium containing 0.5% FBS and then lysed with 1 ml of lysis buffer. In the indicated lanes, cellular lysates were made with a buffer containing 5 mM MgCl₂. CaM pull-down assays were performed with Ca²⁺ or EGTA, and the presence of Ras and Raf-1 in the bound and unbound fractions was analyzed by Western blotting using specific antibodies. All bound fraction and 25 μl of the unbound fraction were loaded. (B) Subconfluent NIH 3T3 cells were incubated for 24 h with medium containing 0.5% FBS. Cells were lysed with CaM pull-down buffer in the presence or absence of 5 mM MgCl₂ and incubated for 1 h at 4°C, and then the amount of Ras-GTP was analyzed by the RBD pull-down method. (C) NIH 3T3 cells were treated with the indicated concentrations of geldanamycin (GA) for 12 h. CaM pull-down assays were performed in the presence of Ca²⁺ or EGTA. The amounts of Ras and Raf-1 present in the bound and unbound fractions were analyzed by Western blotting using specific antibodies.

FIG. 6.

Ras coimmunoprecipitates with CaM. NIH 3T3 cell extracts were incubated with anti-CaM antibodies (α-CaM) or a nonrelated monoclonal antibody (mAb), and the immunocomplex was pulled down using protein G-Sepharose. The presence of Ras in the immunoprecipitate was analyzed by Western blotting using pan-Ras antibodies. An aliquot of the cellular lysate was also loaded (lane L).

Binding of K-Ras, but not H- or N-Ras, from cellular lysates to CaM.

Evidence is accumulating that the diverse Ras isoforms may have different functions and regulation. The possibility that one of the Ras isoforms specifically bound to CaM was explored. CaM-Sepharose pull-down experiments were performed with NIH 3T3 cellular lysates (Fig. 7) in the presence of either Ca²⁺ or EGTA. The presence of K-Ras, H-Ras, or N-Ras in the bound and unbound fractions was analyzed by Western blotting using specific antibodies for each of the Ras isoforms. As shown in Fig. 7A, K-Ras was the only isoform able to bind to CaM-Sepharose. There are two K-Ras proteins, A and B, which originate from alternative splicing of the K-Ras gene and differ principally in their COOH-terminal regions. We analyzed which of the two K-Ras isoforms was binding to CaM by using antibodies specific for each of these isoforms. Interestingly, K-RasB was able to bind to CaM-Sepharose, while K-RasA was not able to at all (Fig. 7B). The specificity of the antibodies used was verified (Fig. 7C).

FIG. 7.

Binding of K-RasB but not K-RasA, H-Ras, or N-Ras from cellular lysates to CaM-Sepharose. NIH 3T3 cellular lysates were incubated with CaM-Sepharose (Seph) in the presence of Ca²⁺ or EGTA. (A) The presence of the different Ras isoforms N-Ras, H-Ras, and K-Ras in the bound and unbound fractions was analyzed by Western blotting using specific antibodies. (B) Same as in panel A, but the Western blot was incubated with either K-RasA or K-RasB antibodies. (C) Fifty-nanogram quantities of H-Ras–GST and K-RasB–GST fusion proteins expressed in Sf9 cells and of bacterially expressed H-Ras, N-Ras, and K-Ras (Oncogene) were loaded onto five different gels, and Western blotting with the indicated antibodies was performed.

Preferential activation of K-Ras by CaM inhibition.

Because the binding of Ras to CaM was isoform specific, we tested whether the activation of Ras observed after CaM inhibition was also isoform specific. Quiescent NIH3T3 cells (in 0.5% FBS-containing medium) were treated with either W13, W12, or PDGF (0.4 nM) for 5 min. The amounts of active H-Ras, N-Ras, and K-Ras were analyzed by RBD pull-down assay followed by Western blotting with specific Ras antibodies against each of the isoforms. As shown in Fig. 8, PDGF (0.4 nM) was able to induce activation of all Ras isoforms. In contrast, W13 treatment induced a significant increase only in the levels of active K-Ras with respect to control nontreated or W12-treated serum-starved cells. Thus, CaM inhibition specifically induced K-Ras activation.

FIG. 8.

Preferential activation of K-Ras by CaM inhibition. Subconfluent NIH 3T3 cells were incubated for 24 h with 0.5% FBS, and then PDGF (0.4 nM), W13 (15 μg/ml), W12 (15 μg/ml), or nothing (−) was added to the medium and left for 5 min. The activation of the different Ras isoforms was analyzed by pull-down assay with RBD-Sepharose and Western blotting using antibodies specific for each of the isoforms. (A) Quantification of the scanned Western blots corresponding to three different experiments was performed, and the relationship between the intensities of the bands after W12 or W13 treatment with respect to nontreated cells (Q) is shown. Error bars indicate standard deviations. (B) Results from a representative experiment out of three performed are shown.

Direct binding of purified K-Ras to CaM.

We then analyzed whether the interaction of K-RasB with CaM, observed from cellular lysates, was due to direct binding of K-RasB to CaM or to the mediation of a CaMBP. For this purpose, K-RasB and H-Ras were expressed in Sf9 insect cells as GST-fused proteins. Purified proteins were loaded with either GTP or GDP and then incubated with CaM-Sepharose in the presence of Ca²⁺ or EGTA. As shown in Fig. 9A, GST–K-RasB, when loaded with GTP, was able to bind to CaM in a Ca²⁺-dependent way. No specific binding of H-Ras to CaM was observed. In order to further prove the specificity of the binding of K-RasB–GTP to CaM, competition with the CaM-binding domain of CaMKII was performed. CaM-Sepharose was incubated with an excess of CaMKII peptide prior to K-RasB–GTP addition. As shown in Fig. 9B, the CaM-binding domain of CaMKII was able to compete for the binding of K-RasB–GTP to CaM.

FIG. 9.

Binding of purified K-Ras to CaM-Sepharose. (A) Purified GST–H-RasV12 and GST–K-RasBV12 proteins expressed in Sf9 cells were loaded with either GTP or GDP as indicated in Materials and Methods. Proteins were then incubated with CaM-Sepharose (Seph) (in the presence of Ca²⁺ or EGTA [E]) or with control Sepharose (in the presence of Ca²⁺). The amounts of proteins in the unbound and bound fractions were analyzed by Western blotting using pan-Ras antibodies. (B) GTP-loaded GST–K-RasBV12 was incubated with CaM-Sepharose in the presence of Ca²⁺ or EGTA, and in the indicated lane CaM-Sepharose was preincubated with the indicated amounts of CaMKII^290–309 peptide in the presence of Ca²⁺. The amount of Ras in the bound fractions was analyzed by Western blotting using pan-Ras antibodies. (C) CaM-Sepharose or Sepharose control pull-down assays were performed with bacterially expressed and purified GST–K-RasB^1–166 and GST–H-Ras^1–166 in the presence of Ca²⁺ or EGTA. The amounts of Ras in the unbound and bound fractions were analyzed by Western blotting using pan-Ras antibodies.

To test whether the region of K-Ras responsible for the binding to CaM was the N-terminal conserved domain (amino acids 1 to 166) or the C-terminal variable domain, the binding of C-terminally truncated K-RasB and H-Ras to CaM was analyzed. GST–K-RasB^1–166 and GST–H-Ras^1–166 were loaded with GTP and then mixed with CaM-Sepharose with either Ca²⁺ or EGTA. The C-terminally truncated forms of both K-RasB and H-Ras were able to bind in a Ca²⁺-dependent way to CaM, while no binding to control Sepharose was observed (Fig. 9C). The same results were obtained with the proteins without the GST (data not shown). Therefore, the conserved region of the two Ras isoforms had the capability to bind to CaM, and most probably the variable region was modulating this capability.

DISCUSSION

Activation of Ras induces a variety of cellular responses, depending on the effectors that become activated and the intensity and amplitude of this activation. A great deal of research in this field is focused on how specific effectors are activated and how the intensity, timing, and localization of the signals are regulated. We have previously shown that Ca²⁺ and CaM are able to down-regulate the Ras/Raf/MEK/ERK pathway, impairing its activation at low serum concentration and preventing a too-high and too-sustained response of this pathway to growth factors (7). We report here new data concerning the down-regulation of Ras by Ca²⁺ and CaM and thus a new point of convergence between Ca²⁺-mediated signaling and the Ras/Raf/MEK/ERK pathway. We show that K-Ras is a CaMBP, and we propose that binding of CaM to K-Ras inhibits in vivo its signaling to Raf and consequent ERK1/2 activation.

In order to gain insight into the mechanism of how CaM down-regulates the Ras/Raf/MEK/ERK pathway, we analyzed whether different extracellular stimuli were able to cooperate with CaM inhibition to induce ERK1/2 activation. As we had previously described, low doses of FBS were essential to induce ERK1/2 activation in cells treated with W13. Therefore, there must be some basal signals provided by those low concentrations of FBS cooperating with W13 to induce ERK1/2 activation. To further elucidate which signals could be involved in this process, we investigated whether low doses of different growth factors could contribute to W13-dependent activation of ERK1/2. A synergism to activate ERK1/2 was observed between CaM inhibition and low doses of PDGF, EGF, and bombesin in Swiss 3T3 cells. Both the EGF and PDGF receptors are tyrosine kinase receptors, and the bombesin receptor is a G-protein-coupled receptor. Although some of the G-protein-coupled receptor agonists have been shown to transactivate tyrosine kinase receptors (23), this seems not to be the case for bombesin in Swiss 3T3 cells, thus indicating that the activation of ERK1/2 by CaM inhibition does not require activation of a tyrosine kinase receptor and suggesting instead that CaM modulates the activity of a commonly used regulator of the Ras/ERK pathway. Interestingly, Ras and ERK activations by W13 treatment appear to require distinct basal conditions: ERK activation clearly requires an additional signal, but this seems not to be the case for Ras activation, as we have found activation with W13 under serum-free conditions (although the activation induced in cells incubated with 0.5% FBS is stronger). Perhaps the higher level of active Ras in cells incubated with 0.5% FBS and treated with W13 is able to activate ERK, in contrast to what happens in serum-free cells treated with W13, although the serum requirement for ERK activation is more likely to be explained by an extra signaling input provided by serum to achieve Raf-1 or MEK activity despite the level of Ras activation. Whatever the case, CaM is essential to lower the activation of the pathway, as blocking of CaM function by itself leads to an activation of Ras, suggesting that CaM is setting a threshold for Ras downstream signaling under basal conditions. This may not be the unique role of CaM-dependent down-regulation of Ras activity. As previously shown, down-regulation of the Ras/Raf/MEK/ERK1/2 pathway by CaM after proliferative stimulation of fibroblasts is important to prevent a too-high increase in the amount of the cell cycle inhibitor p21^cip1 (7). Most recently, Ras activation has been shown to induce mdm2 transcription through the ERK1/2 pathway. A strong Ras signal makes cells more resistant to p53-dependent apoptosis following exposure of the cells to DNA damage due to a destabilization of p53 by the high basal levels of mdm2 (47). Down-regulation of the Ras pathway by CaM could also be essential to modulate the basal levels of mdm2. Furthermore, there are other physiological circumstances, such as cell detachment, in which Ras/Raf/MEK/ERK1/2 pathway activation is inhibited even in the presence of growth factors (46). Although diverse mechanisms have been proposed to inhibit this pathway under these conditions, it would be interesting to analyze CaM participation.

CaM operates its Ca²⁺-signaling outputs through binding and modulation of several CaMBPs. In an attempt to find the CaMBP that could be involved in the regulation of Ras activation, we have analyzed by affinity chromatography with CaM-Sepharose both upstream and downstream Ras regulators, such as GEFs, GAPs, and members of the Ras/ERK signaling pathway. None of them but Raf-1 and Ras were able to bind to CaM in cell lysates, and in both cases the binding was Ca²⁺ dependent. Ras was also shown to immunoprecipitate with anti-CaM antibodies. Further analysis suggested that Ras binding to CaM was GTP dependent, because Ras from lysates of quiescent cells in the presence of 5 mM MgCl₂ was not able to bind to CaM. We have also proved that Ras binds to CaM irrespective of Raf-1, as Raf-1 depletion by geldanamycin treatment did not impair Ras binding. Moreover, binding of Raf to CaM was also Ras independent. Interestingly, K-RasB but not K-RasA, H-Ras, or N-Ras from cellular lysates was able to bind to CaM. Furthermore, direct in vitro studies using H-Ras and K-RasB expressed in insect cells showed that K-RasB itself, but not H-Ras, is a CaMBP, because it is able to bind directly to CaM in a Ca²⁺-dependent manner and this can be inhibited by competition with a well-known CaM-binding peptide, the CaM-binding domain of CaMKII. Finally, K-RasB binding to CaM is clearly favored when it is GTP bound, compared to that of GDP-bound K-Ras.

The high degree of homology between the different Ras isoforms suggested that they would be functionally identical, but there is evidence pointing to a preferential activation of specific effectors by the different Ras isoforms. K-Ras has been shown to activate Raf preferentially with respect to PI3K (61). This would agree with our finding that CaM inhibition activated K-Ras preferentially and with the preferential induction of ERK1/2 phosphorylation with respect to PKB phosphorylation. Furthermore, the fact that H-Ras but not K-Ras is located in cholesterol-enriched fractions of the plasma membrane reinforces the idea of distinct functions of the isoforms (50). Our data showing a distinct modulation of the activation of Ras isoforms by CaM, together with the finding of the specific binding to CaM of only one of the Ras isoforms in a conformation-dependent manner (GTP bound), clearly suggest a novel signaling difference between Ras family members affecting its negative control. The differential down-regulation of the Ras isoforms together with the specificity of the effectors may help to maintain a distinct timing of activation of the diverse Ras downstream pathways. Surprisingly, while the full-length K-RasB but not the H-Ras binds to CaM, we have found that the truncated forms of both K-RasB and H-Ras bind to CaM. This suggest that the conserved regions of all Ras isoforms have the capability to bind to CaM but that the variable carboxy-terminal regions of H-Ras, N-Ras, and K-RasA inhibit CaM interaction while the carboxy-terminal region of K-RasB does not. Analysis of the CaM-binding domain will allow us to design a K-RasB mutant unable to bind to CaM and thus determine the function of CaM–K-RasB interaction.

Although it is possible that CaM binding to K-Ras and activation of K-Ras by CaM inhibition are independent events, we favor the hypothesis that CaM binding to K-Ras leads to its inactivation. One mechanism for this could be that CaM binding to K-Ras-GTP increases its GTPase activity. A second mechanism we propose is that CaM binding to K-Ras–GTP inhibits the transmission of the signal to Raf. Of course, both mechanisms raise many intriguing questions regarding to the precise nature of the modulation of Ras-effector interaction by CaM binding. Experiments to further elucidate these interactions and their physiological consequences in the cell are under way in our laboratory.