Introduction

Angiogenesis, the formation of new capillaries from pre-existing ones, is the major mechanism that underlies neovascularization in adults. It is a multistep process that includes degradation of the basement membrane by proteases, endothelial cell migration and proliferation, formation of a new basement membrane through the recruitment of supporting pericytes and assembly of the new vessels (Polverini, 2002). Angiogenesis is required in many physiological and pathological processes, including embryonic development, wound healing, tissue regeneration, rheumatoid arthritis, tumor growth and metastasis. Angiogenesis is regulated by a tight balance between antiangiogenic agents, such as angiostatin and proangiogenic agents like vascular endothelial growth factor (VEGF).

VEGF is a key regulator of both physiological and pathological angiogenesis. It belongs to a family of dimeric glycoproteins of the PDGF superfamily that includes VEGF-A to -E and PlGF (Meyer et al., 1999; Rousseau et al., 2000a). VEGF-A exists under five different isoforms. VEGF₁₆₅ is the most abundant and biologically active form of VEGF-A and is referred here as VEGF (Petrova et al., 1999). VEGF stimulates angiogenesis in vivo by positively regulating various steps of the angiogenic process. It promotes endothelial cell proliferation and migration and increases the production of plasminogen activators, and the permeability of the endothelial layer. VEGF is expressed and secreted by tumor cells in response to hypoxia, which contributes to tumor expansion associated with neovascularization (Barleon et al., 1997). Interestingly, the VEGF gene is upregulated in most types of cancers (Nakamura et al., 2002; Van Trappen et al., 2002).

On endothelial cells, VEGF binds to two tyrosine kinase receptors, vascular endothelial growth factor receptor (VEGFR)1/Flt-1 and VEGFR2/KDR, and also to neuropilin-1, a coreceptor that increases the affinity of VEGFR2 for VEGF (Petrova et al., 1999; Bernatchez et al., 2002). Knockout mice for VEGFR1 and VEGFR2 are embryonic lethal, which indicates that both receptors are essential for angiogenesis. VEGFR2 is essential for the earliest steps of angiogenesis like blood islands and large vessel formation, whereas VEGFR1/Flt-1 plays a later role, as denoted by the presence of already formed, albeit poorly organized, vessels in the VEGR1-/- mice (Fong et al., 1995; Shalaby et al., 1995; Carmeliet et al., 1996; Ferrara and Bunting, 1996). Upon ligand binding, VEGFRs undergo homodimerization and oligomerization that activate their intrinsic tyrosine kinase activity. The formation of heterodimers between VEGFR1 and VEGFR2 has recently been reported, but their role is unclear (Rousseau et al., 2000a; Sato et al., 2000). VEGFR1 is poorly autophosphorylated in response to VEGF in endothelial cells (Petrova et al., 1999). In contrast, ligand-induced homodimerization of VEGFR2 leads to a strong autophosphorylation of VEGFR2 on tyrosine residues, which drives activation of major signaling pathways that include ERK and stress-activated protein kinase (SAPK)2/p38 mitogen-activated protein kinases (MAP kinases) (Seetharam et al., 1995; Rousseau et al., 1997). Activation of ERK leads to increased proliferation of endothelial cells, whereas activation of SAPK2/p38 triggers the cytoskeleton remodeling that is required to drive actin-based motility (Huttenlocher et al., 1995; Rousseau et al., 2000a). Major autophosphorylation sites on VEGFR2 have been ascribed as Y1175 and Y1214 (Takahashi et al., 2001). Other putatively important phosphorylated sites include Y951 and Y996 in the kinase insert domain and Y1054 and Y1059 in the tyrosine kinase catalytic domain. These tyrosine residues, when phosphorylated, are involved as docking sites to recruit molecules containing SH2, SH3 or PTB domains and to convey signals to downstream pathways (Petrova et al., 1999). In particular, phosphorylation of Y1175 by VEGF is crucial to recruit and activate PLC as well as to convey ERK-mediated signal to DNA synthesis (Takahashi et al., 2001). Interestingly, the adapter protein Sck is recruited to Y1175 and is,thereby, plausibly involved in coupling VEGFR2 to ERK (Ratcliffe et al., 2002). The adapter protein VRAP is possibly involved in connecting PI-3 kinase and PLC to VEGFR2, presumably through interaction with Y951 (Wu et al., 2000). Shc is recruited to a yet unknown tyrosine on the phosphorylated VEGFR2, which triggers connection to the adapter protein Grb2 and then transmits the mitogenic signals to Sos, Ras and ERK (Kroll and Waltenberger, 2000). The tyrosine residues on VEGFR2 as well as the adapter molecules that couple VEGFR2 to SAPK2/p38 are unknown.

The Rho-related members of the Ras GTPase superfamily comprise a number of proteins that include Cdc42, Rac-1 and RhoA. The Rho proteins act as molecular switches that contribute to confer specificity to upstream signals by directing them to the appropriate MAP kinase module (Clerk et al., 2001). In the GDP-bound form, they are inactive, and they are activated by the exchange of GDP for GTP, a reaction that is catalysed by guanine nucleotide exchange factors. GTPase-activating proteins enhance the intrinsic GTPase activity of the Ras and Rho proteins, returning them to their inactive state. Ras is localized to the plasma membrane and it becomes activated upon binding of growth factors to their specific tyrosine receptor kinase. One of the effects of activated Ras-GTP is to bind to c-Raf, translocating it to the plasma membrane for activation. Activated c-Raf phosphorylates and activates the MAPK kinases MEK1 and MEK2, which phosphorylates and activates the MAPKs ERK1 and ERK2. Similarly, Rac1 and Cdc42 are both implicated in the activation of JNKs and SAPK2/p38-MAPKs by various stresses and physiological agonists (Fan et al., 1998; Clerk et al., 2001), an effect that may be mediated through p21-activated kinases (PAKs) (Bagrodia and Cerione, 1999). Intriguingly, activation of Cdc42 downstream of Flt-1 has been associated with the modulation of antiproliferative activity in response to VEGF (Zeng et al., 2002a). However, nothing is known with respect to the GTPases that couple the VEGF signal that emanates from VEGFR2 to SAPK2/p38. Yet, the involvement of the Rho GTPases in the VEGFR2-SAPK2/p38 pathway can be expected since these proteins, as well as SAPK2/p38, are involved in regulating actin dynamics. In the present study, we show that the activation of SAPK2/p38 by VEGF requires the activation of Cdc42 and the phosphorylation of Y1214 on VEGFR2.

Results

Complementary role of RhoA and SAPK2/p38 in transducing the VEGF signal that leads to the formation of actin stress fibers

A major phenotype associated with the activation of VEGFR2 by VEGF in endothelial cells is a massive remodeling of the actin cytoskeleton that reorganizes, within 5 min, into transcytoplasmic stress fibers (Figure 1a, b). This phenotype is abolished and accompanied by a decrease of polymerized actin following the inhibition of SAPK2/p38 activity with the inhibitor SB203580, which indicates that SAPK2/p38 activity is essential for actin remodeling into stress fibers (Figure 1b, e). The VEGF-induced reorganization into stress fibers is similar to the effect that results from the activation of RhoA in fibroblasts and other types of cells (Ridley, 1999). In fact, as early as within 30 s, VEGF increases the activity of RhoA in the human umbilical vein endothelial cell (HUVEC) (Zeng et al., 2002b). Moreover, activating the RhoA pathway with the typical RhoA agonist lysophosphatidic acid also leads to actin remodeling into stress fibers (Figure 1c). However, in contrast to VEGF, inhibiting SAPK2/p38 with SB203580 did not impair the effect of LPA on the formation of stress fibers (Figure 1c, f). Along these lines, we found that the expression of RhoA N19, a dominant-negative form of RhoA, impairs the formation of stress fibers by VEGF in porcine endothelial cells that express VEGFR2 (Figure 2A, panels a–d). In the RhoA N19-expressing cells compared to the surrounding cells, F-actin remained diffuse in the cytoplasm, suggesting that it could not bundle into stress fibers, even when the cells were treated with VEGF. This finding is in marked contrast with the depletion of F-actin that characterized the inhibition of SAPK2/p38 (Figure 1e). These results suggest that RhoA and SAPK2/p38 activities are both involved in the formation of stress fibers induced by VEGF, but that they act through complementary mechanisms. To demonstrate that RhoA was not upstream of SAPK2/p38 in the formation of stress fibers by VEGF, we used a cellular and functional approach. We cotransfected cells with expression vectors carrying dominant-negative forms of both Rac (Rac N17) and Cdc42 (Cdc42 N17) along with a vector expressing an activated form of RhoA (RhoA V14). Then, the cells were treated or not with VEGF in the presence or absence of 1M SB203580, a concentration that completely inhibited the VEGF-induced increase of SAPK2/p38 activity as reflected by the inhibition of MAP kinase-activated protein kinase 2 (MAPKAP K2), a functional target of SAPK2/38 in various types of cells (Figure 2C). Thereafter, the formation of stress fibers was examined in fluorescence microscopy (Figure 2B). As expected, we found that the expression of RhoA V14 induced the formation of stress fibers, even in the presence of the dominant-negative forms of the other two GTPases (Figure 2B, panels a–d). Interestingly, inhibiting SAPK2/p38 with SB203580 did not inhibit the formation of stress fibers in the transfected cells treated with VEGF (Figure 2B panels e, f vs c, d). This further supports the fact that RhoA can mediate stress fiber formation, independently of SAPK2/p38. In accordance, LPA did not activate SAPK2/p38, even in concentrations (200 ng/ml) and times (15 min) that maximally induced the formation of stress fibers in HUVEC (data not shown). Overall, these results suggest that RhoA and SAPK2/p38 are both implicated in the formation of stress fibers, but that RhoA is not upstream of SAPK2/p38. Hence, in the case of VEGF, both SAPK2/p38 and RhoA are essential and act complementarily to form stress fibers. In the case of LPA-induced formation of stress fibers, RhoA is essential but SAPK2/p38 is dispensable.

Figure 1.

SAPK2/p38 is required for VEGF- but not LPA-induced actin remodeling into stress fibers in HUVEC. Quiescent HUVEC plated on gelatin-coated labtek chambers were pretreated for 15 min with 5 M SB203580 (d–f) or 0.02% DMSO (a–c). Cells were then exposed or not (a, d) for another 15 min to 5 ng/ml VEGF (b, e) or 200 ng/ml LPA (c, f). F-actin was detected using FITC-conjugated phalloidin, as described in Materials and methods. Cells were examined by fluorescent microscopy. Representative fields are shown. Similar results were obtained in two separate experiments

Full figure and legend (230K)

Figure 2.

RhoA and SAPK2/p38 are both essential to induce the formation of stress fibers by VEGF. (a) PAE cells stably expressing VEGFR2 (PAE/VEGFR2 cells) were plated on gelatin-coated labtek chambers for 24 h. Then, they were transfected by lipofection with 0.11 g of pCDNA3-myc-RhoA N19 (dominant negative). At 24 h after transfection, quiescent cells were left untreated (a, b) or exposed to 5 ng/ml VEGF (c, d) for 15 min. Cells were fixed and incubated with 9E10 myc-tag antibody that was detected with a biotinylated secondary mouse antibody. They were then stained with rhodamine–streptavidin to visualize RhoA-positive cells (b, d). F-actin was visualized using FITC-conjugated phalloidin (a, c). Cells were examined by fluorescent microscopy. Representative fields are shown. Similar results were obtained in two separate experiments. (b) PAE/VEGFR2 cells were plated on gelatin-coated labtek chambers for 24 h. Then, they were transfected by lipofection with 0.037 g (0.11 g total) of each of the following plasmids: pCDNA3-myc-Cdc42 N17 (dominant negative), pCDNA3-myc-Rac N17 (dominant negative) and pCDNA3-myc-RhoA V14 (constitutively active). After 24 h, quiescent cells were pretreated with 1 M SB203580 (e, f) or 0,02% DMSO (a–d) for 15 min. Cells were then treated or not for another 15 min with 5 ng/ml VEGF (c–f). Then F-actin (a, c, e) and transfected cells (b, d, f) were detected in fluorescence microscopy as in (b). Representative fields are shown. (c) PAE/VEGFR2 cells were pretreated for 15 min with the indicated increasing concentrations of SB203580 to inhibit SAPK2/p38 activity. Thereafter, they were exposed for 10 min to VEGF (5 ng/ml). After treatments, cells were extracted and the same amount of proteins were loaded on SDS–PAGE, transferred onto nitrocellulose and blotted for phospho-MAPKAP kinase-2

Full figure and legend (331K)

Small GTPase Cdc42 is essential to drive VEGFR2-mediated activation of SAPK2/p38 by VEGF

The finding that RhoA is not upstream of SAPK2/p38 activation in response to VEGF indicates that a GTPase distinct from RhoA is involved in connecting the SAPK2/p38 MAP kinase module to VEGFR2. Notably, Cdc42 and Rac have been shown to be involved in the activation of both SAPK1/JNK and SAPK2/p38 in response to various stimuli (Tibbles and Woodgett, 1999). To ascertain whether Cdc42 might be involved in activating SAPK2/p38 in response to VEGF, we first determined whether VEGF induced the activation of this GTPase. HUVEC were treated or not for various periods of time with 5 ng/ml VEGF. Then, extracts were prepared and the activation of Cdc42 was evaluated by measuring the amount of Cdc42-GTP that binds to glutathione S-transferase (GST)-PAK-Rac-binding domain fusion protein (GST-CRIB) in a pull-down assay (Li et al., 2002). Results showed that VEGF induced a time-dependent activation of Cdc42 that reached a maximum after 1 min (Figure 3a). Similarly, VEGF induced a twofold activation of Cdc42 in NIH-3T3 cells that transiently express VEGFR2. In these cells, the activity of Cdc42 was increased or blocked by the expression of Cdc42 V14 or Cdc42 N17, respectively (Figure 3b). We next ascertained whether Cdc42 was upstream of SAPK2/p38 in response to VEGF. We then developed NIH 3T3 cell lines that stably express VEGFR2, since HUVEC and porcine aortic endothelial (PAE) cells cannot be efficiently transfected to be used for biochemical assays. Several clones expressing VEGFR2 were obtained. 4F2 clone was selected for the experiments since it expressed the highest amount of VEGFR2 (Figure 3c). 4F2 cells were cotransfected with vectors expressing Ha-tagged p38 and LT-tagged MAPKAP K2, a direct target of SAPK2/p38, along with an empty vector or with vectors expressing Cdc42 V12 or Cdc42 N17. Then, the transfectants were treated or not with VEGF and the activity of LT-tagged MAPKAP K2 was assayed as a functional measure of SAPK2/p38 activity in the transfected cells. As shown in Figure 3d, VEGF induced the activation of LT-tagged MAPKAP K2 when added to control cells expressing the empty vectors or the vectors expressing activated Cdc42 (Cdc42 V12). In contrast, the expression of Cdc42 N17 in 4F2 cells abolished the VEGF-induced activation of LT-tagged MAPKAP K2. These results indicate that Cdc42 is essential for the activation of SAPK2/p38 by VEGF. As expected, the expression of Cdc42 N17 also impaired the VEGF-induced activation of HA-SAPK2/p38 (Figure 3e).

Figure 3.

Cdc42 is activated by VEGF and is upstream of SAPK2/p38 in VEGF/VEGFR2 signaling. (a) Quiescent HUVEC were treated for various periods of time with 5 ng/ml VEGF. They were then extracted and incubated with GST-CRIB to adsorb activated Cdc42. The samples were then loaded on 12% SDS–PAGE, transferred onto a nitrocellulose membrane and blotted for Cdc42. (b) Using the CaCl₂ transfection method, NIH 3T3 cells were transfected with either pAlter-VEGFR2, pCDNA3-myc-Cdc42 V12 or pCDNA3-myc-Cdc42 N17. At 48 h after transfection, quiescent cells were left untreated or exposed to 10 ng/ml VEGF for 1 min. Cells were then processed for Cdc42 assay as in (a). (c) NIH 3T3 cells were transfected with pAlterMax/VEGFR2 and pCDNA3 and selected for stable expression using 400 g/ml G418. Different clones were obtained and cells from these clones were grown in selective media, plated and extracted in Laemli sample buffer. A total of 35 g of proteins were loaded on 8.5% SDS–PAGE, transferred onto nitrocellulose and analysed for VEGFR2 expression. The clone 4F2 expresses the highest amount of VEGFR2 and was selected for further experiments. In (d) 4F2 cells were plated for 24 h. Then, using the CaCl₂ method, they were cotransfected with pCDNA3-HA-p38, pSVT7-LT-MAPKAP K2 and pCINeo (empty vector), or pCDNA3-myc-Cdc42 V12 or pCDNA3-myc-Cdc42 N17. After 48 h, cells were left untreated or exposed to 10 ng/ml VEGF for 10 min. Cells were extracted and LT-MAPKAP kinase-2 was immunoprecipitated using a KT3 mouse antibody. Then LT-MAPKAP K2 was assayed in the presence of [-³²P]ATP using recombinant HSP27 as a substrate. Proteins were separated by SDS–PAGE and the kinase activity was visualized using a PhosphorImager system by measuring ³²P incorporation into the specific substrate. Results are expressed as the ratio of kinase activity of stimulated cells over unstimulated cells. Representative results from two distinct experiments are shown. In (e), 4F2 cells were cotransfected with pCDNA3-HA-p38 and pCDNA3 (empty vector) or pCDNA3-myc-Cdc42 N17. After 48 h, quiescent cells were left untreated or exposed to 10 ng/ml VEGF for 10 min. Cells were extracted and HA-p38 was immunoprecipitated using the HA-tag antibody. The immunocomplexes were loaded on 10% SDS–PAGE, transferred onto nitrocellulose and blotted with phospho-p38 antibody. The membrane was stripped and reprobed with HA-tag to ensure equal protein loading

Full figure and legend (157K)

As shown in Figure 1, the activation of SAPK2/p38 by VEGF is involved in triggering the formation of stress fibers (Figure 1b, d). To confirm our findings that Cdc42 was upstream of SAPK2/p38 activation by VEGF, we transfected two different constitutively active forms of this GTPase, Cdc42 V12 or Cdc42 L61 (Lamarche et al., 1996), in PAE/VEGFR2 cells and looked for the formation of stress fibers following exposures to VEGF in the presence or absence of SB203580. We found that the expression of both Cdc42 V12 (Figure 4a, d vs b, e) and Cdc42 L61 (Figure 4g, j vs h, k) increased the formation of stress fibers by VEGF and that the increase was blocked by inhibiting SAPK2/p38 activity with SB203580 (Figure 4b, e vs c, f and h, A vs i, l and Figure 2c). Notably, the number of cells showing transcytoplasmic stress fibers increased by 1.5-fold in the VEGF-treated cells expressing Cdc42 L61, and pretreatments with SB203580 completely inhibited this increase (data not shown). Moreover, the stress fibers observed in cells expressing Cdc42 V14 or Cdc42 L61 were much more thick (Figures 4b, h). In contrast, the expression of a constitutively active form of Rac (Rac V12) was associated with the formation of membrane ruffles, but did not lead to the formation of stress fibers, neither in the absence nor in the presence of VEGF (Figures 5a, d vs b, e). This indicates that Rac cannot be upstream of SAPK2/p38 in signaling to actin remodeling into stress fibers. Interestingly, Rac V12-mediated formation of membrane ruffles was insensitive to the inhibition of SAPK2/p38 with SB203580 (Figure 2C), which supports the fact that Rac is not upstream of SAPK2/p38 in regulating the formation of ruffles (Figure 5b, e vs c, f). Collectively, these results indicate that the small GTPase Cdc42 is involved in connecting the SAPK2/p38 MAP kinase module to VEGFR2.

Figure 4.

Cdc42 is required for VEGF-induced formation of stress fibers. PAE/VEGFR2 cells plated on gelatin-coated labtek chambers were lipofected with pCDNA3-myc-Cdc42 V12 (a–f) or pCDNA3-myc-Cdc42 L61 (g–l). At 24 h after transfection, cells were pretreated for 15 min with 1 M SB203580 (c, f, i, l) or 0.02% DMSO (a, b, d, e and g, h, j, k), and were then left untreated (a, d, g, j) or were exposed to 5 ng/ml VEGF (b, c, e, f and h, i, k, l) for another 15 min. Then, F-actin (a–c and g–i) and transfected cells (d–f and j–l) were visualized in fluorescent microscopy as in Figure 3. Representative fields are shown. Similar results were obtained in two separate experiments

Full figure and legend (236K)

Figure 5.

Rac induces the formation of ruffles in an SAPK2/p38-independent manner in PAE/VEGFR2 cells. PAE/VEGFR2 cells plated on gelatin-coated labtek were lipofected with pCDNA3-myc-Rac V12 (constitutively active). After 24 h, cells were pretreated for 15 min with 1 M SB203580 (c, f) or 0.02% DMSO (a, b, d, e), and were then left untreated (a, d) or were exposed to 5 ng/ml VEGF (b, c, e, f) for another 15 min. Then, F-actin (a–c) and transfected cells (d–f) were visualized as in Figure 3. Representative fields are shown. Similar results were obtained in two separate experiments

Full figure and legend (188K)

Tyrosine 1214 on VEGFR2 is required to drive activation of Cdc42 and SAPK2/p38 by VEGF

Activation of MAP kinases by tyrosine kinase receptors like VEGFR2 involved autophosphorylation and transphosphorylation of tyrosine residues. Then, adapter proteins are recruited to specific phosphorylated tyrosines. This initiates the activation of the small GTPases and then of the MAP kinase module. In this study, we aimed to identify which tyrosine residue on VEGFR2 leads to the activation of SAPK2/p38 via Cdc42. The cytoplasmic domain of VEGFR2 contains at least 11 Y residues that once phosphorylated can theoretically interact with SH2 and/or SH3 and/or PTB domain contained in signaling molecules. Major autophosphorylated sites on VEGFR2 have been ascribed as Y1175 and Y1214 (Takahashi et al., 2001). Whereas activation of ERK is linked to phosphorylation of Y1175, nothing is known concerning signaling from Y1214 (Takahashi et al., 2001). We then tested the hypothesis that Y1214 was required to activate SAPK2/p38. By site-directed mutagenesis, we proceeded to the mutation of site Y1214 of VEGFR2 replacing the tyrosine (Y) residue for a phenylalanine (F) residue and we transiently transfected this mutant and the wild-type form of VEGFR2 into wild-type PAE or NIH 3T3 cells. Then, we monitored the activation of SAPK2/p38, as well as ERK, in response to VEGF by means of Western blotting using antibodies that specifically recognized the phosphorylated/activated form of the kinases. We found that Y1214 was necessary to allow the transduction of the VEGF signal to SAPK2/p38, since the VEGF-induced activation of SAPK2/p38 was completely abolished in cells transiently expressing VEGFR2-Y1214 relative to cells that express the wild-type form of the receptor (Figure 6a). As expected, VEGF has no effect when added to control cells expressing an empty vector. Interestingly, the Y1214 mutated receptor still transduced the signal to ERK (Figure 6b), indicating that signaling from Y1214 is specific for SAPK2/p38 activation. Overall, these results indicate that tyrosine 1214 on VEGFR2 is essential for SAPK2/p38 activation in response to VEGF. As shown above, activation of Cdc42 by VEGF is required to drive the activation of SAPK2/p38. In this context, we next ascertained whether Y1214 was required for the activation of Cdc42 by VEGF. We transfected NIH 3T3 cells with a wild-type form of VEGFR2 or with the VEGFR2-Y1214F mutant in NIH 3T3 cells, and then treated them with VEGF. We found that wild-type VEGFR2 could drive the activation of Cdc42 with a 2.5–3-fold peak at 1 min. This is consistent with our findings that the VEGF-induced activation of Cdc42 follows the same kinetics in normal VEGFR2-bearing cells (Figure 3a). Conversely, VEGF did not activate Cdc42 in cells that expressed VEGFR2-Y1214F (Figure 7). We conclude that phosphorylation of Y1214 on VEGFR2 is essential to activate Cdc42 and SAPK2/p38 sequentially in response to VEGF.

Figure 6.

Phosphorylation on Y1214 of VEGFR2 is required for VEGF-induced activation of SAPK2/p38. (a) Wild-type PAE cells (null for VEGFR2) were lipofected or not with pAlterMax-VEGFR2 or pAlterMax-VEGFR2/Y1214F. After 24 h, the cells were treated or not with 10 ng/ml VEGF for 10 min. Cells were extracted in Laemli sample buffer and equal amounts of proteins were separated on two different SDS–PAGE. Proteins were transferred onto nitrocellulose and were processed for immunodetection of phospho-p38 (upper panel) or of p38 (lower panel). Results are expressed as the ratio of kinase activity of stimulated cells over unstimulated cells. Representative results from two distinct experiments are shown. (b) NIH 3T3 cells (null for VEGFR2) were transfected or not with either pAlterMax VEGFR2 or pAlterMax VEGFR2/Y1214F. After 48 h, cells were treated or not with 10 ng/ml VEGF for 10 min. Cells were extracted in Laemli sample buffer and equal amounts of proteins were separated on two separate SDS–PAGE. Proteins were transferred onto nitrocellulose and were processed for immunodetection of phospho-ERK (upper panel), ERK2 (middle panel) or VEGFR2 (lower panel). Representative results from two distinct experiments are shown

Full figure and legend (142K)

Figure 7.

Phosphorylation on Tyr1214 on VEGFR2 is required for VEGF-induced activation of Cdc42. NIH 3T3 cells (null for VEGFR2) were transfected, using the CaCl₂ method, with either pAlter-Max VEGFR2 or pAlterMax-VEGFR2/Y1214F. At 48 h after transfection, quiescent cells were exposed to 10 ng/ml VEGF for 1 or 5 min, as indicated. Cells were then extracted and were used for Cdc42 assays using GST-CRIB pull-down assay as described in Figure 3. Results are expressed as the mean ratio of Cdc42 activity of stimulated cells over unstimulated cells. Data are expressed as the means.e. of triplicate values

Full figure and legend (65K)

Discussion

In endothelial cells, VEGF binds to the tyrosine kinase receptor VEGFR2, which activates several pathways that convey the VEGF signal to targets involved in modulating its biological functions. In particular, binding of VEGF to VEGFR2 triggers the activation of the MAP kinase SAPK2/p38, which leads to activation in cascade of MAPKAP kinase 2/3, phosphorylation of heat-shock protein (HSP)27, actin remodeling and increased actin-based motility (Rousseau et al., 2000b; Masson-Gadais et al., 2003). Autophosphorylation of tyrosine kinase receptors like VEGFR2 is considered to be important to impart substrate specificity by recruiting limited substrates for the receptors (Fantl et al., 1992). The small GTPases of the Rho family further contribute to impart specificity to the signal by differently connecting the MAP kinase module to the receptors. In the present study, we found that activation of SAPK2/p38 by VEGFR2 requires its phosphorylation on Y1214 and that it further requires the activation of the small GTPase Cdc42 downstream of Y1214.

The major autophosphorylation sites on VEGFR2 are Y1175 and Y1214 (Takahashi et al., 2001). Here, we obtained evidence that phosphorylation of Y1214 is required to drive the activation of SAPK2/p38. This is supported by the finding that the activation of SAPK2/p38 in cells expressing a nonphosphorylatable mutant for Y1214 (Y1214F) is completely impaired following exposure to VEGF, in contrast to cells that express a wild form of VEGFR2. This identifies for the first time Y1214 as required for the activation of SAPK2/p38. SAPK2/p38 is a major pathway that drives the actin remodeling that underlies actin-based motility in various cell systems including endothelial cells. In accordance, this pathway is involved in modulating angiogenesis and its inhibition impairs angiogenesis (Mudgett et al., 2000; Javerzat et al., 2002). SAPK2/p38 is, however, a nonspecific pathway in the sense that it transduces signals elicited by other growth factors and stress stimuli (Kyriakis and Avruch, 2001). Hence, its direct inhibition cannot be specifically targeted against angiogenesis. Our finding that activation of SAPK2/p38 by VEGF required the tyrosine phosphorylation of Y1214 of VEGFR2 raises the possibility to develop PY1214-blocking compounds to be used to inhibit specifically the increased endothelial cell motility associated with activation by VEGF. In this context, identification of the earliest adapter molecules that are recruited to phosphorylate Y1214 will greatly help in designing compounds that block PY1214. The amino-acid environment that surrounds Y1214 revealed the GRB2-binding site (Songyang et al., 1993). Intriguingly, GRB2 is recruited to ligand-bound VEGFR2 in porcine endothelial cells, whereas it is not associated with VEGFR2 at detectable levels in primary cultures of sinusoidal endothelial cells activated by VEGF (Kroll and Waltenberger, 1997; Takahashi et al., 1999). Hence, its role in connecting VEGFR2 to SAPK2/p38 remains to be clarified. Data bank searching revealed that Y1214 is also in an environment that can bind to Src-related proteins. In accordance, we found that inhibiting Src activities could also inhibit SAPK2/p38 (our unpublished results), which suggests that Src might be recruited to activated VEGFR2 and be involved in connecting Y1214 to activation of SAPK2/p38. Interestingly, the expression of Y1214F does not inhibit the activation of ERK by VEGF, which indicates that phosphorylation of Y1214 is not required for ERK activation by VEGF. This is consistent with the previous findings showing that phosphorylation of Y1175 of VEGFR2 is required to recruit and activate PLC, a step that is upstream of ERK activation by VEGF (Bernatchez et al., 2001; Takahashi et al., 2001). In turn, this will initiate DNA synthesis, a well-known event downstream of ERK (Rousseau et al., 1997; Takahashi et al., 2001).

The small GTPase Ras and members of the Rho family, RhoA, Rac and Cdc42 are essential intermediates that are involved in signaling to MAP kinases. Signaling by small GTPases is interconnected and the implication of a given GTPase in the activation of a given MAP kinase depends upon the stimuli and the cellular context. A major accomplishment of our study is to have brought evidence that Cdc42 is essential to connect phosphorylation of Y1214 of VEGFR2 to activation of SAPK2/p38 and SAPK2/p38-mediated formation of stress fibers in response to VEGF. This is supported by the findings that VEGF activates Cdc42 within 1 min in HUVEC exposed to VEGF and by the findings that the expression of a dominant-negative form of Cdc42 in NIH-3T3 cells that stably express VEGFR2 abrogated the activation of SAPK2/p38 by VEGF. Moreover, the expression of two different constitutively active forms of Cdc42 (Cdc42 V12 and Cdc42 L61) induces the formation of stress fibers in porcine endothelial cells and the effect is increased by VEGF and blocked by SB203580, making the link with SAPK2/p38. In addition, we found that the expression of the nonphosphorylatable mutant Y1214F of VEGFR2 inhibits both the activation of Cdc42 and SAPK2/p38. From these results, we conclude that SAPK2/p38-mediated actin reorganization into stress fibers in response to VEGF requires the phosphorylation of Y1214 on VEGFR2, which then triggers, through yet unknown adapter proteins, the activation of Cdc42 and then of the SAPK2/p38 module (MAP3K, MAP2K and SAPK2/p38). In turn, activation of SAPK2/p38 triggers the activation of MAPKAP kinase 2/3 and phosphorylation of HSP27 initiating actin remodeling and actin-based motility. The p21-activated protein kinase PAK-1 is an effector of the GTP-bound Cdc42 that possibly links activated Cdc42 to the SAPK2/p38 module. This is supported by the fact that several studies showed that activation of PAK-1 is associated with the subsequent activation of SAPK2/p38 and with the regulation of the cytoskeleton dynamics in smooth muscle cells and other types of cells (Zhang et al., 1995; Dechert et al., 2001). It remains to be determined whether PAK1 is activated in endothelial cells exposed to VEGF. Incidentally, the findings that Cdc42 is downstream of Y1214 is consistent with the possibility that Src might be recruited to Y1214 since Src is involved in the activation of Cdc42 as well as PAK and cell migration (Sells et al., 1997; Renkema et al., 2002). In addition, the fact that activation of Cdc42 by VEGF was observed in cells that express only VEGFR2 is in line with our previous observations that VEGFR2, but not VEGFR1, is sufficient to activate SAPK2/p38 and actin-based motility (Rousseau et al., 2000a, 2000b). Intriguingly, in HUVEC that express EGLT, a chimeric receptor containing the extracellular domain of the EGFR and the catalytic domain of VEGFR1, the addition of EGF triggers the activation of Cdc42, which suggests that VEGFR1 can be involved in activating Cdc42 (Zeng et al., 2002a).

In various types of cells, the formation of stress fibers is tightly associated with the activation of RhoA (Ridley, 1999, 2000). In accordance, we found that exposure of endothelial cells to LPA, a typical activator of RhoA, leads to the formation of stress fibers. Interestingly, we found that the expression of a dominant-negative form of RhoA inhibited the formation of stress fibers induced by VEGF, indicating that the road taken by the VEGF signal to induce stress fibers transits through RhoA. However, the finding that RhoA V14 induced the formation of stress fibers even in cells that express Cdc42 N17 indicates that the RhoA-mediated effect occurs independent of SAPK2/p38, since Cdc42 is required for activation of SAPK2/p38 by VEGF. In corollary, this finding suggests that RhoA is not upstream of SAPK2/p38 activation by VEGF, which is in accordance with the observation that the formation of stress fibers by VEGF/RhoA V14 was insensitive to inhibition of SAPK2/p38 activity by SB203580. This is also in line with the fact that LPA induced the formation of stress fibers independently of SAPK2/p38, since this actin phenotype was not inhibited by SB203580 and since LPA did not activate SAPK2/p38. Incidentally, activation of RhoA by VEGF is not downstream of SAPK2/p38 either since SB203580 did not inhibit the VEGF-induced activation of RhoA, both in HUVEC and in 4F2 cells (data not shown). From all these findings, one can conclude that the formation of stress fibers in response to VEGF requires signals that transit through two distinct and complementary pathways involving RhoA and Cdc42/SAPK2/p38. Activation of SAPK2/p38 by VEGF is importantly involved in triggering actin polymerization through phosphorylation of the actin-polymerizing factor HSP27 (Rousseau et al., 2000b). On its side, activation of RhoA is tightly associated with the formation of focal adhesions that result from the increased cellular tension mediated through the activation of Rock1, MLCK and phosphorylation of myosin light chain (Sastry and Burridge, 2000). It is thus possible that the RhoA-mediated formation of focal adhesions is required to anchor the actin filaments that are generated through the SAPK2/p38 pathway allowing bundling of actin into stress fibers. In this context, we previously reported that the proper assembly of focal adhesions, through a process that requires HSP90-dependent phosphorylation of focal adhesion kinase (FAK), was required to allow the SAPK2/p38 remodeling of F-actin into stress fibers in response to VEGF (Rousseau et al., 2000). In accordance, FAK is tyrosine phosphorylated downstream of RhoA in Swiss 3T3 cells activated with serum and HSP90 is required for the activation of RhoA and cytoskeleton remodeling in response to thrombin (Barry and Critchley, 1994; Pai et al., 2001; Pai and Cunningham, 2002). A complementarity between the RhoA and Cdc42/SAPK2/p38 pathways in regulating actin remodeling has also been reported in response to TGF in pancreatic carcinoma cells (Edlund et al., 2002). In that case, however, the formation of stress fibers is delayed and seems to require new protein synthesis that may involve the activation of different genes downstream of RhoA, SAPK2/p38 as well as Smad4. Incidentally, in response to TGF, SAPK2/p38 is also required for the formation of ruffles, but in a Rac-independent manner. In contrast, we found in endothelial cells that the formation of ruffles requires Rac activation, since they are inhibited by a dominant-negative form of Rac (data not shown). Moreover, they are independent of SAPK2/p38 since they are not blocked by SB203580. These findings clearly illustrate that similar actin phenotypes can be achieved using different signaling pathways and mechanisms.

In summary, we showed in the present study that the VEGF signal that leads to the activation of SAPK2/p38 and SAPK2/p38-mediated formation of stress fibers requires the phosphorylation of autophosphorylation site Y1214 on VEGFR2, and involves the activation of the small GTPase Cdc42. We also found that the activation of the small GTPase RhoA is required for the formation of stress fibers, but independent of SAPK2/p38. We propose that the formation of stress fibers by VEGF requires increased actin polymerization generated through the activation of the VEGFR2-Cdc42-SAPK2/p38 pathway and proper assembly of focal adhesions downstream of RhoA (see the model in Figure 8). The identification of phosphorylation of Y1214, as a necessary step for the activation of SAPK2/p38, may help to develop new drugs aimed at inhibiting the SAPK2/p38-mediated increase in cell migration and angiogenesis from its inception, imparting strong specificity to the intervention.

Figure 8.

Summary and working model. VEGF-A₁₆₅ binds to VEGFR2 on endothelial cells, which stimulates autophosphorylation on cytoplasmic tyrosine residues of VEGFR2, notably Tyr1214. SAPK2/p38 is activated by VEGF via tyrosine phosphorylation of Y1214 and activation of Cdc42. Activation of SAPK2/p38 activates MAPKAP K2, which, as we previously reported, phosphorylates HSP27 and increases actin polymerization (Rousseau et al., 1997, 2000a). RhoA is also activated by VEGF but the crosstalk between RhoA and SAPK2/p38 remains to be elucidated. We previously reported that phosphorylation of FAK by VEGF was sensitive to HSP90 inhibition by geldanamycin (Masson-Gadais et al., 2003; Rousseau et al., 2000b). However, the role of HSP90 in regulating the phosphorylation of FAK also remains to be clarified. We propose that the activation of RhoA contributes with the activation of FAK to assemble focal adhesions allowing the anchorage of the stress fibers that are generated through the increased actin polymerization that results from the activation of SAPK2/p38

Full figure and legend (161K)

Materials and methods

Chemicals

[-³²P]ATP was purchased from Amersham-Pharmacia Biotech (Montréal, Québec, Canada). FITC–phalloidin, endothelial cell growth supplement (ECGS), VEGFR, lysophosphatidic acid and Geneticin (G418) were obtained from Sigma-Aldrich (Oakville, Ontario, Canada). SB203580 was purchased from Calbiochem (San Diego, CA, USA) and was diluted in DMSO to make a stock solution of 20 mM. Pfu Turbo polymerase was purchased from Stratagene (LaJolla, CA, USA).

Antibodies

Anti-Cdc42 rabbit antibody was purchased from Cytoskeleton (Denver, CO, USA). The anti-Myc tag (9E10) mouse antibody was a gift from Dr Jacques Landry (Québec). Anti-ERK2 is a rabbit polyclonal antibody raised against a synthetic peptide that corresponds to the 14 carboxy-terminal amino acids of rat ERK2 (Huot et al., 1995). Anti-p38 is a polyclonal antibody raised in rabbit against the carboxy-terminal sequence PPLQEEMES of murine p38 (Huot et al., 1997). The phospho-p38 MAP kinase rabbit polyclonal antibody and the phospho-p44/p42 mouse monoclonal antibody were purchased from Cell Signaling (Beverly, MA, USA). The anti-phospho-MAPKAP K2 (Thr222) is a rabbit polyclonal antibody from Upstate (Lake Placid, NY, USA). The KT3 antibody is a mouse monoclonal antibody purchased from Covance (Richmond, CA, USA). Anti-HA-tag (clone HA-7) and anti-VEGFR2 (clone KDR-2) are both mouse monoclonal antibodies purchased from Sigma.

Plasmids

pCDNA3-myc-Cdc42 V12, pCDNA-myc-Cdc42 N17, pCDNA-myc-Cdc42 L61, pCDNA-myc-Rac-N17 and pCDNA-myc-Rac-V12 were obtained from Dr Nathalie Lamarche (McGill University, Montreal, Canada). pCDNA3-myc-RhoA V14 and pCDNA3-myc-RhoA-N19 were obtained from Dr Sylvain Bourgoin (Université Laval, Québec, Canada). The expression vectors containing Ha-p38 and LT-MAPKAP K2 were previously described (Huot et al., 1997).

Site-directed mutagenesis

Human VEGFR2 cDNA was obtained from Dr Johannes Waltenberger (Ulm, Germany). VEGFR2 was subcloned in the pAlterMax expression vector (Promega). Single point mutations were introduced by PCR-mediated mutagenesis. The mutagenic oligonucleotides designed to replace tyrosine (Y) residues for phenylalanine residues (F: bold) in Y1214 were: 5'-GTGACCCCAAGTTCCATTTTGACAACACAG-3' (sense) and 5'-CTGATTCCTGCTGTGTTGTCAAAATGGAATTTG-3' (reverse). The methylated parental cDNA was then digested using DpnI and was discarded by separation on agarose. Constructs were ligated and the mutated nucleotide was verified by sequencing.

Cell cultures and transfection

HUVECs were isolated by collagenase digestion of umbilical veins from undamaged sections of fresh cords (Huot et al., 1997). HUVEC were grown on gelatin-coated 75 cm² culture dishes in 199 medium supplemented with 20% heat-inactivated fetal bovine serum (FBS), 60 mg/l ECGS, glutamine, heparin and antibiotics. Subcultures were obtained by trypsination and were used at passages <5. At 16 h before experiments, HUVEC were incubated in ECGS-free medium containing 5% FBS before addition of VEGF, or LPA. PAE/VEGFR2 is a well-characterized cell line that stably expresses VEGFR2 (Waltenberger et al., 1994). Parental PAE cells and PAE/VEGFR2 cells were grown on gelatin-coated culture dishes in F12 medium supplemented with 10% heat-inactivated FBS and, in the case of PAE/VEGFR2 cells, with 400 g/ml G418. For transient transfection, PAE or PAE/KDR cells were lipofected with plasmid DNAs using a ratio of 4 : 1 Tfx-50 (Promega) for 90 min in the absence of serum. Cells were then overlaid with complete medium and assays were carried out 24 h post transfection. At 16 h before experiments, cells were incubated in serum-free F12 medium. NIH 3T3 were transfected with pAlterMax/VEGFR2 and pCDNA3 using the calcium chloride precipitation technique for generation of a fibroblastic cell line stably expressing VEGFR2. The cells were incubated in medium containing 400 g/ml G418. A pool of resistant cells was diluted until a single cell per well was obtained. Clone 4F2 expressed the highest level of VEGFR2 as determined by Western blotting and was thus selected for further experiments. NIH 3T3 parental cells and clone 4F2 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% calf serum, and, in the case of 4F2 cells, with 400 g/ml G418. For transient expression, 1.75 10⁶ NIH 3T3 or 4F2 cells were plated in 100 mm culture dishes and transfected with 20–25 g of plasmid DNAs using the standard calcium chloride precipitation technique. At 16 h before the addition of VEGF, cells were incubated in serum-free DMEM. Assays were carried out 48 h post transfection. All cultures were incubated in a humidified atmosphere containing 5% CO₂.

Kinase assays

MAPKAP K2/3 activity, a physiological target of SAPK2/p38, was assayed as a measure of the in vivo kinase activity of SAPK2/p38 (Huot et al., 1998). MAPKAP kinase 2/3 activity was assayed using recombinant HSP27 as a substrate (Huot et al., 1995). The assays were carried out in 25 l of kinase buffer: 100 M ATP, 3 Ci of [-³²P]ATP (3000 Ci/mol), 40 mM p-nitrophenyl phosphate, 20 mM MOPS pH 7.0, 10% glycerol, 15 mM MgCl₂, 0.05% Triton X-100, 1 mM dithiothreitol, 1 M leupeptin, 0.1 mM phenylmethylsulfonyl fluoride. The kinase activity was assayed for 30 min at 30°C and was stopped by adding 10 l of SDS–PAGE loading buffer. The activity of the kinase was quantified by measuring the incorporation of radioactivity into the specific substrate after SDS–PAGE. In certain experiments, p38 and ERK activities have been assayed by Western blotting using antibodies against phospho-p38 and phospho-ERK, respectively (Rousseau et al., 2000b).

Cdc42 activation assays

Bacteria expressing GST-Pak-Rac-binding domain (CRIB) fusion protein were kindly provided by Dr Louise Larose (McGill University, Montréal, Canada). Bacteria were pelleted and were resuspended in lysis buffer: 50 mM Tris-HCl pH 7.5, 1% Triton X-100, 150 mM NaCl, 5 mM MgCl₂, 1 mM dithiothreitol, 10 g/ml aprotinin, 10 g/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride. The lysate was sonicated on ice and then was pelleted again. The supernatant was incubated for 1 h with 250 l of glutathione-coupled Sepharose-4B beads (Amersham) in lysis buffer. The beads were washed six times with buffer (50 mM Tris-HCl pH 7.5, 0.5% Triton X-100, 150 mM NaCl, 5 mM MgCl₂, 1 mM dithiothreitol, 1 g/ml aprotinin, 1 g/ml leupeptin, 0.1 mM phenylmethylsulfonyl fluoride) and finally resuspended in wash buffer containing 10% glycerol. GST-CRIB bound to Sepharose beads was aliquoted and stored at -80°C. The concentration was approximately 0.5 g/l as estimated on Coomassie-stained SDS–PAGE. Starved cells were stimulated with VEGF (see figures for the exact concentrations) and stimulation was stopped by the addition of ice-cold phosphate-buffered saline (PBS). Cells were lysed with 600 l cell lysis buffer (50 mM Tris-HCl pH 7.2, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 500 mM NaCl, 10 mM MgCl₂, 10 g/ml aprotinin, 10 g/ml leupeptin, 0.1 mM phenylmethylsulfonyl fluoride) and lysates were centrifugated at 16 000 g for 2 min. The supernatant was incubated with 50 g of GST-CRIB beads for 90 min on ice. Protein bound to beads were washed four times in Tris buffer (50 mM Tris-HCl pH 7.2, 1% Triton X-100, 150 mM NaCl, 10 mM MgCl₂, 10 g/ml aprotinin, 10 g/ml leupeptin, 0.1 mM phenylmethylsulfonyl fluoride) and resuspended in 15 l of Laemli sample buffer. The whole extract was loaded on 12% SDS–PAGE, transferred onto nitrocellulose and blotted with Cdc42 antibody.

Immunofluorescence

Fluorescence microscopy was used for visualization of F-actin and Myc-tagged proteins. HUVECs or PAE/KDR cells were plated on gelatin-coated Labtek. After treatments, cells were fixed with 3.7% formaldehyde and permeabilized with 0.1% saponin in PBS, pH 7.5. F-actin was detected using FITC-conjugated phalloidin (33.3 g/ml) diluted 1 : 50 in phosphate buffer. Myc epitope was detected using 9E10 mouse antibody. Myc antigen–antibody complexes were detected with biotin-labeled anti-mouse IgG and were revealed with Texas Red-conjugated streptavidin. Cells were examined under fluorescent microscopy with a Nikon Eclipse E600 equipped with a 40 0.85 NA objective lens. Images were captured as 16 bit TIFF files with a Micromax 130 YHS cooled (-30°C) camera (Princeton Instruments, Trenton, NJ, USA) driven by Metamorph software (Universal Imaging Corp., Downington, PA, USA).

Immunoprecipitation

After treatments, cells were lysed in 80 l of denaturation buffer (10 mM Tris-HCl pH 7.5, 1% SDS, 1 mM Na₃VO₄). Samples were diluted 10 times in buffer B (150 mM NaCl, 50 mM Tris-HCl pH 7.5, 0.1% Triton X-100, 0.1% sodium deoxycholate, 2 mM EDTA, 2 mM EGTA, 1 mM Na₃VO₄, 1 M leupeptin, 1 mM benzamidine, 50 mM sodium fluoride and 1 mM PMSF) and precleared with 12 l of a 50% v/v Protein-G Sepharose suspension for 15 min with shaking on ice. Supernatants were incubated on for 16 h with 2 l of mouse monoclonal anti-HA-tag antibody and then 12 l of 50% v/v Protein-G Sepharose was added and incubation was extended for 30 min on ice with shaking. Antibody–antigen complexes were washed four times with buffer B and SDS–PAGE loading buffer was added. Proteins were separated through SDS–PAGE and transferred onto nitrocellulose for Western blotting.

References

1. Bagrodia S and Cerione RA. (1999). Trends Cell Biol., 9, 350–355. | Article | PubMed | ISI | ChemPort |
2. Barleon B, Siemeister G, Martiny-Baron G, Weindel K, Herzog C and Marme D. (1997). Cancer Res., 57, 5421–5425. | PubMed | ISI | ChemPort |
3. Barry ST and Critchley DR. (1994). J. Cell Sci., 107, 2033–2045. | PubMed | ISI | ChemPort |
4. Bernatchez PN, Allen BG, Gelinas DS, Guillemette G and Sirois MG. (2001). Br. J. Pharmacol., 134, 1253–1262. | Article | PubMed | ISI | ChemPort |
5. Bernatchez PN, Rollin S, Soker S and Sirois MG. (2002). J. Cell Biochem., 85, 629–639. | PubMed |
6. Carmeliet P, Ferreira V, Breier G, Pollefeyt S, Kieckens L, Gertsenstein M, Fahrig M, Vandenhoeck A, Harpal K, Eberhardt C, Declercq C, Pawling J, Moons L, Collen D, Risau W and Nagy A. (1996). Nature, 380, 435–439. | Article | PubMed | ISI | ChemPort |
7. Clerk A, Pham FH, Fuller SJ, Sahai E, Aktories K, Marais R, Marshall C and Sugden PH. (2001). Mol. Cell. Biol., 21, 1173–1184. | Article | PubMed | ISI | ChemPort |
8. Dechert MA, Holder JM and Gerthoffer WT. (2001). Am. J. Physiol. Cell Physiol., 281, C123–C132. | PubMed |
9. Edlund S, Landstrom M, Heldin CH and Aspenstrom P. (2002). Mol. Biol. Cell, 13, 902–914. | Article | PubMed | ISI | ChemPort |
10. Fan WT, Koch CA, De Hoog CL, Fam NP and Moran MF. (1998). Curr. Biol., 8, 935–938. | Article | PubMed | ISI | ChemPort |
11. Fantl WJ, Escobedo JA, Martin GA, Turck CW, del Rosario M, McCormick F and Williams LT. (1992). Cell, 69, 413–423. | Article | PubMed | ISI | ChemPort |
12. Ferrara N and Bunting S. (1996). Curr. Opin. Nephrol. Hypertens., 5, 35–44. | PubMed | ChemPort |
13. Fong GH, Rossant J, Gertsenstein M and Breitman ML. (1995). Nature, 376, 66–70. | Article | PubMed | ISI | ChemPort |
14. Huot J, Houle F, Marceau F and Landry J. (1997). Circ. Res., 80, 383–392. | PubMed | ISI | ChemPort |
15. Huot J, Houle F, Rousseau S, Deschesnes RG, Shah GM and Landry J. (1998). J. Cell Biol., 143, 1361–1373. | Article | PubMed | ISI | ChemPort |
16. Huot J, Lambert H, Lavoie JN, Guimond A, Houle F and Landry J. (1995). Eur. J. Biochem., 227, 416–427. | Article | PubMed | ISI | ChemPort |
17. Huttenlocher A, Sandborg RR and Horwitz AF. (1995). Curr. Opin. Cell Biol., 7, 697–706. | Article | PubMed | ISI | ChemPort |
18. Javerzat S, Auguste P and Bikfalvi A. (2002). Trends Mol. Med., 8, 483. | Article | PubMed | ISI | ChemPort |
19. Kroll J and Waltenberger J. (1997). J. Biol. Chem., 272, 32521–32527. | Article | PubMed | ISI | ChemPort |
20. Kroll J and Waltenberger J. (2000). Z. Kardiol., 89, 206–218. | PubMed |
21. Kyriakis JM and Avruch J. (2001). Physiol. Rev., 81, 807–869. | PubMed | ISI | ChemPort |
22. Lamarche N, Tapon N, Stowers L, Burbelo PD, Aspenstrom P, Bridges T, Chant J and Hall A. (1996). Cell, 87, 519–529. | Article | PubMed | ISI | ChemPort |
23. Li X, Saint-Cyr-Proulx E, Aktories K and Lamarche-Vane N. (2002). J. Biol. Chem., 277, 15207–15214. | Article | PubMed | ISI | ChemPort |
24. Masson-Gadais B, Houle F, Laferrière J and Huot J. (2003). Cell Stress Chaperon., 8, 37–52.
25. Meyer M, Clauss M, Lepple-Wienhues A, Waltenberger J, Augustin HG, Ziche M, Lanz C, Buttner M, Rziha HJ and Dehio C. (1999). EMBO J., 18, 363–374. | Article | PubMed | ISI | ChemPort |
26. Mudgett JS, Ding J, Guh-Siesel L, Chartrain NA, Yang L, Gopal S and Shen MM. (2000). Proc. Natl. Acad. Sci. USA, 97, 10454–10459. | Article | PubMed | ChemPort |
27. Nakamura M, Abe Y and Tokunaga T. (2002). Pathol. Int., 52, 331–339. | PubMed |
28. Pai KS and Cunningham DD. (2002). J. Neurochem., 80, 715–718. | PubMed |
29. Pai KS, Mahajan VB, Lau A and Cunningham DD. (2001). J. Biol. Chem., 276, 32642–32647. | PubMed |
30. Petrova TV, Makinen T and Alitalo K. (1999). Exp. Cell Res., 253, 117–130. | Article | PubMed | ISI | ChemPort |
31. Polverini PJ. (2002). J. Dent. Educ., 66, 962–975. | PubMed |
32. Ratcliffe KE, Tao Q, Yavuz B, Stoletov KV, Spring SC and Terman BI. (2002). Oncogene, 21, 6307–6316. | Article | PubMed |
33. Renkema GH, Pulkkinen K and Saksela K. (2002). Mol. Cell. Biol., 22, 6719–6725. | Article | PubMed | ISI | ChemPort |
34. Ridley A. (2000). J. Cell Biol., 150, F107–F109. | PubMed |
35. Ridley AJ. (1999). Nat. Cell Biol., 1, E64–E66. | Article | PubMed | ISI | ChemPort |
36. Rousseau S, Houle F and Huot J. (2000a). Trends Cardiovasc. Med., 10, 321–327. | Article | PubMed | ISI | ChemPort |
37. Rousseau S, Houle F, Kotanides H, Witte L, Waltenberger J, Landry J and Huot J. (2000b). J. Biol. Chem., 275, 10661–10672. | Article | PubMed | ISI | ChemPort |
38. Rousseau S, Houle F, Landry J and Huot J. (1997). Oncogene, 15, 2169–2177. | Article | PubMed | ISI | ChemPort |
39. Sastry SK and Burridge K. (2000). Exp. Cell Res., 261, 25–36. | Article | PubMed | ISI | ChemPort |
40. Sato Y, Kanno S, Oda N, Abe M, Ito M, Shitara K and Shibuya M. (2000). Ann. NY Acad. Sci., 902, 201–205 (discussion 205–207). | PubMed | ChemPort |
41. Seetharam L, Gotoh N, Maru Y, Neufeld G, Yamaguchi S and Shibuya M. (1995). Oncogene, 10, 135–147. | PubMed | ISI | ChemPort |
42. Sells MA, Knaus UG, Bagrodia S, Ambrose DM, Bokoch GM and Chernoff J. (1997). Curr. Biol., 7, 202–210. | Article | PubMed | ISI | ChemPort |
43. Shalaby F, Rossant J, Yamaguchi TP, Gertsenstein M, Wu XF, Breitman ML and Schuh AC. (1995). Nature, 376, 62–66. | Article | PubMed | ISI | ChemPort |
44. Songyang Z, Shoelson SE, Chaudhuri M, Gish G, Pawson T, Haser WG, King F, Roberts T, Ratnofsky S, Lechleider RJ, Neel BG, Birge RB, Fajardo JE, Chou MM, Hanafusa H, Schaffhausen B and Cantley LC. (1993). Cell, 72, 767–778. | Article | PubMed | ChemPort |
45. Takahashi T, Ueno H and Shibuya M. (1999). Oncogene, 18, 2221–2230. | Article | PubMed | ISI | ChemPort |
46. Takahashi T, Yamaguchi S, Chida K and Shibuya M. (2001). EMBO J., 20, 2768–2778. | Article | PubMed | ISI | ChemPort |
47. Tibbles LA and Woodgett JR. (1999). Cell. Mol. Life Sci., 55, 1230–1254. | Article | PubMed | ISI | ChemPort |
48. Van Trappen PO, Ryan A, Carroll M, Lecoeur C, Goff L, Gyselman VG, Young BD, Lowe DG, Pepper MS, Shepherd JH and Jacobs IJ. (2002). Br. J. Cancer, 87, 537–544. | Article | PubMed | ChemPort |
49. Waltenberger J, Claesson-Welsh L, Siegbahn A, Shibuya M and Heldin CH. (1994). J. Biol. Chem., 269, 26988–26995. | PubMed | ISI | ChemPort |
50. Wu LW, Mayo LD, Dunbar JD, Kessler KM, Ozes ON, Warren RS and Donner DB. (2000). J. Biol. Chem., 275, 6059–6062. | Article | PubMed | ChemPort |
51. Zeng H, Zhao D and Mukhopadhyay D. (2002a). J. Biol. Chem., 277, 4003–4009. | Article | PubMed | ISI | ChemPort |
52. Zeng H, Zhao D and Mukhopadhyay D. (2002b). J. Biol. Chem., 19, 19.
53. Zhang S, Han J, Sells MA, Chernoff J, Knaus UG, Ulevitch RJ and Bokoch GM. (1995). J. Biol. Chem., 270, 23934–23936. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We thank Drs Nathalie Lamarche and Louise Larose from McGill University and Sylvain Bourgoin (Université Laval, Québec) for providing the various GTPase constructs and the GST-CRIB protein. We also thank Dr Jacques Landry (Québec) for providing the 9E10 antibody and NIH 3T3 cells and Drs Bruce Terman (New York) and Johannes Waltenberger (Ulm, Germany) for the VEGFR2 cDNA and PAE/VEGFR2 cells. We thank Dr François Marceau and the Department of Obstetrics of l'Hôpital St-François d'Assise for providing umbilical cords. We also thank André Lévesque and Éric Pèlerin for their help with microscopy and Dr Simon Rousseau and Éric Shink for the VEGFR2-NIH 3T3 cells. This work was supported by a grant from The Canadian Institutes of Health Research.