Rac1 has a crucial role in epidermal growth factor (EGF)-induced membrane ruffling, lamellipodial protrusion, and cell migration. Several guanine nucleotide exchange factors (GEFs) including Sos1, Sos2, Tiam1 and Vav2 have been shown to transduce the growth signal from the EGF receptor to Rac1. To clarify the role of each GEF, we time-lapse imaged the EGF-induced activity change of Rac1 in A431 cells transfected with siRNA targeting each Rac1 GEF. Because knockdown of these GEFs suppressed EGF-induced Rac1 activation only partially, we looked for another Rac1 GEF downstream of the EGF receptor and found that Asef, a Rac1-Cdc42 GEF bound to the tumor suppressor APC, also contributed to EGF-induced Rac1 activation. Intriguingly, EGF stimulation induced phosphorylation of Tyr94 within the APC-binding region of Asef in a manner dependent on Src-family tyrosine kinases. The suppression of EGF-induced Rac1 activation in siRNA-treated cells was restored by wild-type Asef, but not by the Tyr94Phe mutant of Asef. This observation strongly argues for the positive role of Tyr94 phosphorylation in EGF-induced Asef activation following the activation of Rac1.
Introduction
The Rho-family GTPase Rac1 induces membrane ruffles and lamellipodial protrusion of the cell membrane in a growth-factor-receptor-dependent manner (Ridley et al., 1992). Many growth factors stimulate phosphatidylinositol 3-phosphate kinase (PI 3-kinase), resulting in an increase in phosphatidylinositol (3,4,5)-phosphate (PtdIns(3,4,5)P3) at the plasma membrane. Because not only classical guanine nucleotide exchange factors (GEFs) with Dbl homology (DH) and pleckstrin homology (PH) domains (Schmidt and Hall, 2002), but also atypical GEFs called DOCK180 superfamily proteins (Cote and Vuori, 2002) bind to PtdIns(3,4,5)P3 (reviewed in Schmidt and Hall, 2002; Cote et al., 2005), this local accumulation of PtdIns(3,4,5)P3 recruits and activates many GEFs for Rac1 at the plasma membrane. In particular, epidermal growth factor receptor (EGFR) has been shown to activate Rac1 through Tiam1 (Ray et al., 2007), Sos1 (Scita et al., 1999) and Vav2 (Marcoux and Vuori, 2003).
Rho guanine nucleotide exchange factor 4 (ARHG4, also known and hereafter referred to as Asef) was originally identified as a Rac1 GEF bound to the tumor suppressor APC and has been shown to promote reorganization of the actin cytoskeleton and to drive aberrant migration of colorectal tumor cells (Kawasaki et al., 2000; Kawasaki et al., 2003). Similar to other canonical GEFs for Rho-family GTPases, Asef contains a DH domain and a PH domain, the latter of which mediates PtdIns(3,4,5)P3-dependent membrane translocation of Asef (Muroya et al., 2007). More recently, Asef has been shown to activate Cdc42 in vitro more potently than Rac1 (Gotthardt and Ahmadian, 2007), and such substrate specificity could vary in a cell-type-specific manner (Kawasaki et al., 2007). These findings clearly indicate that Asef activates Rac1 and Cdc42 in the Wnt signaling cascade (Akiyama and Kawasaki, 2006); however, its role in epidermal growth factor (EGF) signaling remains to be studied.
The observations that multiple GEFs contribute to the EGF-induced Rac1 activation indicate simple redundancy or suggest specific roles of each GEF depending on the different cellular contexts. Here, we examined the contribution of each GEF to the EGF-induced activation of Rac1 in A431 epithelial cells by a small interfence RNA (siRNA)-mediated knockdown approach and by activity imaging with a biosensor based on Förster resonance energy transfer (FRET). In addition to the GEFs known to be involved in EGF signaling, we found that Asef was also involved in the EGF-induced activation of Rac1 and morphological changes. Furthermore, we found that tyrosine phosphorylation of Asef within the APC-binding region is essential for its activity.
Results
Effect of Sos-, Vav2- or Tiam1-deficiency on the EGF-dependent activation of Rac1
Among several GEFs that have been proposed to transduce growth signals to Rac1, we first examined the contribution of Sos proteins, Vav2 and Tiam1 by RNA interference (RNAi) (Fig. 1A). There are two Sos isoforms, Sos1 and Sos2, to which we refer collectively as Sos proteins. We confirmed by quantitative immunoblotting analysis that the knockdown efficiency was at least 70% (Fig. 1B). The EGF-induced activation of Rac1 was monitored by the FRET biosensor Raichu-Rac1, as reported previously (Kurokawa and Matsuda, 2005) (Fig. 1C). Knockdown of Vav2 or Tiam1 significantly suppressed EGF-induced Rac1 activation to ∼50% of the control, as measured by FRET imaging (Fig. 1C) and pull-down assay (Fig. 1D and supplementary material Fig. S1A). We also confirmed that EGF stimulation induced the phosphorylation of Vav2 in A431 cells (supplementary material Fig. S2). However, knockdown of Sos proteins (Fig. 1C) or expression of a dominant-negative mutant of Ras (data not shown) did not impair EGF-induced Rac1 activation to a detectable level. Of note, we confirmed that knockdown of Sos proteins abrogated EGF-induced activation of Ras (supplementary material Fig. S3). The modest effect of the knockdown of GEFs that have been regarded as the primary Rac1 activators motivated us to search for individual GEFs that contribute to the EGF-induced activation of Rac1 in A431 cells.
Effect of Sos, Vav2, and Tiam1 depletion on EGF-induced Rac1 activation. (A) Schematic, showing the current view of regulation of Rac1 by EGF. (B) A431 cells were transfected with either Sos1- and Sos2-specific siRNAs, Vav2-specific siRNA or Tiam1-specific siRNA. Forty-eight hours after transfection, cells were lysed and subjected to immunoblot analysis using antibodies against Sos1, Sos2, Vav2 and Tiam1. (C) A431 cells were transfected with siRNA as indicated. One day after siRNA transfection, cells were further transfected with pRaichu-Rac1. After serum starvation for 6 hours, cells were stimulated with 25 ng/ml EGF and examined for Rac1 activity by FRET microscopy as described in the text. Averages of the fold increase after EGF addition over the control cells are given + s.d. Numbers of cell examined under each condition were as following: control, n=85; Sos1,2 KD, n=13; Vav2 KD, n=9; Tiam1 KD, n=90. *P<0.05, significant difference as compared with the control (Student's t-test). (D) Control and knockdown cells were starved for 6-12 hours, stimulated with 25 ng/ml EGF for 2 minutes or left unstimulated, and examined for active Rac1 by pull-down assay. Experiments were repeated at least three times, and average values of fold increase over the control cells are given + s.d. *P<0.05, significant difference as compared with the control (Student's t-test).
Effect of Sos, Vav2, and Tiam1 depletion on EGF-induced Rac1 activation. (A) Schematic, showing the current view of regulation of Rac1 by EGF. (B) A431 cells were transfected with either Sos1- and Sos2-specific siRNAs, Vav2-specific siRNA or Tiam1-specific siRNA. Forty-eight hours after transfection, cells were lysed and subjected to immunoblot analysis using antibodies against Sos1, Sos2, Vav2 and Tiam1. (C) A431 cells were transfected with siRNA as indicated. One day after siRNA transfection, cells were further transfected with pRaichu-Rac1. After serum starvation for 6 hours, cells were stimulated with 25 ng/ml EGF and examined for Rac1 activity by FRET microscopy as described in the text. Averages of the fold increase after EGF addition over the control cells are given + s.d. Numbers of cell examined under each condition were as following: control, n=85; Sos1,2 KD, n=13; Vav2 KD, n=9; Tiam1 KD, n=90. *P<0.05, significant difference as compared with the control (Student's t-test). (D) Control and knockdown cells were starved for 6-12 hours, stimulated with 25 ng/ml EGF for 2 minutes or left unstimulated, and examined for active Rac1 by pull-down assay. Experiments were repeated at least three times, and average values of fold increase over the control cells are given + s.d. *P<0.05, significant difference as compared with the control (Student's t-test).
Involvement of Asef in EGF-induced activation of Rac1 in A431 cells
Since Asef has been shown to have a crucial role in migration of colorectal tumor cells (Kawasaki et al., 2003), we examined its involvement in EGF-induced activation of Rac1 by using shRNA (Fig. 2A-C). In the control A431 cells stimulated with EGF, Rac1 was rapidly and diffusely activated in broad areas of the plasma membrane, followed by localized and intermittent activation at nascent lamellipodia and membrane ruffles. However, EGF-induced Rac1 activation was significantly attenuated in Asef-depleted A431 (Fig. 2D) and HeLa cells (supplementary material Fig. S4). The suppression of Rac1 activation was more prominent at the cell periphery than the perinuclear area. Accordingly, EGF-induced lamellipodial induction and membrane ruffles were also suppressed in Asef-knockdown cells. Quantitative analysis of FRET images and the results of pull-down assays showed that the level of suppression of Rac1 activation in Asef-knockdown cells was similar to that in Vav2- or Tiam1-knockdown cells (Fig. 2E-G). This observation clearly indicates that there are multiple GEFs that lead to Rac1 activation in EGF-stimulated A431 cells. We attempted to doubly or triply knockdown Asef, Tiam1 and Vav2, but we observed that a large proportion of the transfected cells detached from the culture dishes and could not find any synergistic effect using the cells remaining on the dishes (data not shown). Probably these GEFs are also involved in the cell adhesion and/or survival, and only the cells that had failed to uptake siRNAs remained on the culture dishes. Thus, it is currently unclear whether simultaneous knockdown of Tiam1, Vav2 and Asef could completely ablate the EGF-induced activation of Rac1.
Suppression of EGF-dependent activation of Cdc42 by depletion of Asef
Since Asef has also been shown to activate Cdc42 (Gotthardt and Ahmadian, 2007; Kawasaki et al., 2007), we examined the effect of Asef knockdown on the EGF-induced activation of Cdc42 using FRET imaging with the FRET biosensor Raichu-Cdc42 or pull-down assay. In A431 cells or HeLa cells deficient for Asef, EGF-induced Cdc42 activation was significantly reduced, similar to cells deficient for Vav2 (Fig. 3A,B, supplementary material Fig. S1B and Fig. S4). These results suggested that Vav2 and Asef are involved not only in the activation of Rac1 but also in that of Cdc42. To further understand the activation mechanisms, we examined the activation of Rac1 in the presence of a dominant-negative Cdc42 mutant and the activation of Cdc42 in the presence of a dominant-negative Rac1 mutant. As previously reported (Kurokawa et al., 2004), dominant-negative mutants of Cdc42 suppressed the activation of Rac1 and vice versa (supplementary material Fig. S5). Combining this with our previous observation we reasoned that there is a positive feedback cycle between Rac1 and Cdc42 in EGF-stimulated cells (Kurokawa et al., 2004). This observation prevented us from determining whether or not Asef activated Rac1 directly or indirectly through the activation of Cdc42 in A431 cells.
Identification of Tyr94 within the APC-binding region as the phosphorylation site of Asef
We next sought to determine the pathway leading to Asef activation in EGF-stimulated A431 cells. In colorectal tumor cells, the armadillo repeat domain of APC (APCARM) interacts with and activates Asef (Kawasaki et al., 2000). However, we did not observe any detectable increase in the amount of Asef in complex with APCARM upon EGF stimulation, negating the involvement of this pathway in the EGF-mediated Asef activation (data not shown). We therefore examined the phosphorylation status of Asef and found that Asef was phosphorylated on a tyrosine residue in an EGF-dependent manner (Fig. 4). Intriguingly, neither A2VEA9 (also known and hereafter referred to as Asef2), a close homolog of Asef, nor Tiam1 was tyrosine-phosphorylated to a detectable level. We took advantage of the observation that EGF induces tyrosine phosphorylation of Asef but not Asef2, despite their high sequence similarity, and attempted to determine the phosphotyrosine residue(s) in Asef by preparing chimeric proteins (Fig. 5A,B). Two chimeric mutants, Asef1-Asef2 and Asef2-Asef1, were constructed by swapping the NcoI-digested restriction fragments of the respective Asef and Asef2 cDNAs (shown in Fig. 5B). Tyrosine phosphorylation of these chimeric mutants was determined as described above (Fig. 5C). Asef1-Asef2, which consisted of the N-terminal fragment of Asef and the C-terminal fragment of Asef2, was tyrosine-phosphorylated as efficiently as Asef, whereas Asef2-Asef1, the reverse of Asef1-Asef2, was not. Asef contains three tyrosine residues before the NcoI site. These three tyrosine resides were replaced with glutamate to identify the phosphorylation site(s) (Fig. 5D). The substitution of Tyr94 caused the most remarkable reduction of phosphotyrosine on Asef, followed by the substitution of Tyr104. Substitution of both Tyr94 and Tyr104 almost completely abrogated EGF-induced tyrosine phosphorylation of Asef. Thus, the primary and secondary tyrosine phosphorylation sites in Asef were identified as Tyr94 and Tyr104, respectively. Intriguingly, Tyr94 is located within the APC-binding region (ABR) of Asef.
Effect of Asef deficiency on EGF-induced activation of Rac1. (A) A431 cells were transfected with pSuper-Luc or pSuper-Asef. After selection with puromycin, the expression level of Asef mRNA was examined by RT-PCR (left). (B) To confirm the the efficiency of knockdown, cells transfected with a control shRNA expression vector and an Asef expression vector were analyzed by immunoblotting (right). Signal intensities of each band were quantified and are shown below each band as a relative strength to the control (in %). (C) A431 cells were transfected with control siRNA or Asef siRNA. Two days later, knockdown efficiency was examined by immunoblotting. Signal intensities of each band were quantified and are shown below each band as a relative strength to the control (in %) in B and C. (D-F) A431 cells were transfected with an empty pSuper vector (control) or pSuper-Asef. After puromycin selection, cells were further transfected with pRaichu-Rac1. After 3-6 hours of serum starvation, cells were stimulated with 25 ng/ml EGF. (D) Representative FRET images at the indicated time points are shown in the intensity-modulated display mode. (E) Normalized FRET:CFP ratios as described in the text. (F) From the peak values of the normalized ratio, the effect of Asef knockdown on EGF-induced Rac1 activation was calculated as described in the legend to Fig. 1D. The error bars indicate + s.d. Numbers of cells under each condition were as follows: control, n=13; Asef KD, n=14. *P<0.05, significant difference as compared with control (Student's t-test). (G) Control and Asef knockdown cells were starved for 6-12 hours, stimulated with 25 ng/ml EGF for 2 minutes or left unstimulated, and were examined for active Rac1 by pull-down assay. Experiments were repeated at least three times, and average values of the fold increase over control cells are given + s.d. *P<0.05, significant difference as compared with the contro (Student's t-test).
Effect of Asef deficiency on EGF-induced activation of Rac1. (A) A431 cells were transfected with pSuper-Luc or pSuper-Asef. After selection with puromycin, the expression level of Asef mRNA was examined by RT-PCR (left). (B) To confirm the the efficiency of knockdown, cells transfected with a control shRNA expression vector and an Asef expression vector were analyzed by immunoblotting (right). Signal intensities of each band were quantified and are shown below each band as a relative strength to the control (in %). (C) A431 cells were transfected with control siRNA or Asef siRNA. Two days later, knockdown efficiency was examined by immunoblotting. Signal intensities of each band were quantified and are shown below each band as a relative strength to the control (in %) in B and C. (D-F) A431 cells were transfected with an empty pSuper vector (control) or pSuper-Asef. After puromycin selection, cells were further transfected with pRaichu-Rac1. After 3-6 hours of serum starvation, cells were stimulated with 25 ng/ml EGF. (D) Representative FRET images at the indicated time points are shown in the intensity-modulated display mode. (E) Normalized FRET:CFP ratios as described in the text. (F) From the peak values of the normalized ratio, the effect of Asef knockdown on EGF-induced Rac1 activation was calculated as described in the legend to Fig. 1D. The error bars indicate + s.d. Numbers of cells under each condition were as follows: control, n=13; Asef KD, n=14. *P<0.05, significant difference as compared with control (Student's t-test). (G) Control and Asef knockdown cells were starved for 6-12 hours, stimulated with 25 ng/ml EGF for 2 minutes or left unstimulated, and were examined for active Rac1 by pull-down assay. Experiments were repeated at least three times, and average values of the fold increase over control cells are given + s.d. *P<0.05, significant difference as compared with the contro (Student's t-test).
Production and characterization of anti-phospho-Asef antibody
To further characterize the role of Tyr94 phosphorylation, we tried to raise antiserum against an oligopeptide containing phospho-Tyr94 (Fig. 5E). However, judging from the reactivity to the phosphorylated EGFR, this antiserum seemed to non-specifically recognize phosphotyrosine and not just phospho-Tyr94. Therefore, we used this serum against Asef phosphorylated at Tyr94 in the presence of phenylphosphate, which remarkably reduced the reactivity of the serum to EGFR, but not to Asef. Under this condition, the serum against Asef phosphorylated at Tyr94 reacted with the wild-type Asef in an EGF-dependent manner but not with mutants in which Tyr94 was replaced with either phenylalanine (Y94F) or glutamate (Y94E) (Fig. 5F).
Src-mediated phosphorylation of Asef
We next examined whether Asef was phosphorylated directly by the EGFR or by another tyrosine kinase. The peptide sequence encompassing Tyr94 of Asef, but not that of the homologous peptide sequence of Asef2, was found to match the substrate sequence for Src family kinases (for more information see http://scansite.mit.edu/). Thus, we examined the effects of the Src kinase inhibitor PP2 on tyrosine phosphorylation of Asef in response to EGF. In the presence of 4 μM PP2, Asef phosphorylation by EGF was attenuated markedly (Fig. 6A). We confirmed that autophosphorylation of the EGFR was not inhibited in the presence of 4 μM PP2 (data not shown). In agreement with this result, we found that v-Src and active c-Src, but not Abl, phosphorylated Asef very efficiently (Fig. 6B). The Asef Y94F mutant was not phosphorylated by active c-Src, negating non-specific phosphorylation by Src kinase (Fig. 6C). To corroborate the direct phosphorylation by Src, we incubated either wild-type or the Y94F-mutant Asef with the purified c-Src protein. In vitro, Src phosphorylates Asef on mutiple tyrosine residues, but apparently Tyr94 seemed to be the major site (Fig. 6D). The time course of Asef phosphorylation as examined by using the serum against Asef phosphorylated at Tyr94 was very similar to that of the tyrosine phosphorylation of EGFR (Fig. 6E).
Effect of depletion of GEFs on EGF-induced Cdc42 activation. (A) Cells were prepared as described for Figs 1 and 2, except that pRaichu-Cdc42 was transfected instead of pRaichu-Rac1. After serum starvation, cells were stimulated with 25 ng/ml EGF and examined for Cdc42 activity by FRET microscopy as described in the text. The averages of the fold increase over the control cells are shown + s.d. Numbers of cells examined for each condition were as follows: control siRNA, n=23; Vav2 KD, n=6; Tiam1 KD, n=13; control shRNA, n=6; Asef KD, n=3. *P<0.05, significant difference as compared with the control by t-test analysis (Student's t-test). (B) Control and knockdown cells were starved for 6-12 h, stimulated with 25 ng/ml EGF for 2 minutes or left unstimulated, and were examined for active Cdc42 by pull-down assay. Experiments were repeated at least three times, and average values of the fold increase over the mock-treated cells are shown + s.d. *P<0.05, significant difference as compared with the control by t-test analysis (Student's t-test).
Effect of depletion of GEFs on EGF-induced Cdc42 activation. (A) Cells were prepared as described for Figs 1 and 2, except that pRaichu-Cdc42 was transfected instead of pRaichu-Rac1. After serum starvation, cells were stimulated with 25 ng/ml EGF and examined for Cdc42 activity by FRET microscopy as described in the text. The averages of the fold increase over the control cells are shown + s.d. Numbers of cells examined for each condition were as follows: control siRNA, n=23; Vav2 KD, n=6; Tiam1 KD, n=13; control shRNA, n=6; Asef KD, n=3. *P<0.05, significant difference as compared with the control by t-test analysis (Student's t-test). (B) Control and knockdown cells were starved for 6-12 h, stimulated with 25 ng/ml EGF for 2 minutes or left unstimulated, and were examined for active Cdc42 by pull-down assay. Experiments were repeated at least three times, and average values of the fold increase over the mock-treated cells are shown + s.d. *P<0.05, significant difference as compared with the control by t-test analysis (Student's t-test).
Localization of tyrosine-phosphorylated Asef at lamellipodia and membrane ruffles
It has previously been reported that Asef is localized at both cytoplasm and membrane ruffles (Kawasaki et al., 2000). We therefore examined by immunostaining whether EGF stimulation recruits Asef to the plasma membrane. Wild-type Asef and the Asef-Y94F mutant were recruited to the membrane ruffles upon EGF stimulation in A431 cells (Fig. 7), suggesting that the EGF-induced recruitment of Asef to the plasma membrane does not depend on tyrosine phosphorylation, but probably on the PH domain (Muroya et al., 2007). Notably, phosphorylated Asef accumulated at the lamellipodia and membrane ruffles of the EGF-stimulated A431 cells (Fig. 7). These observations suggest that, upon EGF stimulation, Asef is translocated to the plasma membrane in a PH-domain-dependent manner, and is then phosphorylated and activated in a Src-dependent manner.
Tyr94 phosphorylation is essential in Asef-mediated activation of Rac1 and Cdc42
To understand the role of Tyr94 phosphorylation on the activity of Asef, we tested whether the wild-type and Y94F mutant of Asef can restore the EGF-induced activation of Rac1 and Cdc42 in Asef-knockdown cells by using RNAi-resistant expression vectors (Fig. 8A). In contrast to wild-type Asef, the AsefY94F mutant could not rescue the EGF-induced activation of Rac1 (Fig. 8B) and Cdc42 (Fig. 8C). This observation strongly suggested that Asef is regulated positively by phosphorylation of Tyr94.
EGF-induced tyrosine phosphorylation of Asef. A431 cells expressing HA-Asef, HA-Tiam1-C1199 or HA-Asef2 were serum starved for 24 hours and stimulated with 100 ng/ml EGF for 2 minutes or left unstimulated. Immunoprecipitates and total cell lysates were immunoblotted using anti-phosphotyrosine (pTyr) or anti-HA antibody.
EGF-induced tyrosine phosphorylation of Asef. A431 cells expressing HA-Asef, HA-Tiam1-C1199 or HA-Asef2 were serum starved for 24 hours and stimulated with 100 ng/ml EGF for 2 minutes or left unstimulated. Immunoprecipitates and total cell lysates were immunoblotted using anti-phosphotyrosine (pTyr) or anti-HA antibody.
Discussion
Various growth factors activate Rac1 and/or Cdc42 through a number of GEFs. Owing to this redundancy, it is largely unknown to what extent each GEF contributes to the individual signaling cascades. At the beginning of this research, we postulated that the EGF signal leading to the activation of Rac1 and Cdc42 is transduced from the EGFR primarily by three GEFs: Sos proteins, Tiam1 and Vav2 (reviewed in Scita et al., 2000). By using the siRNA-mediated knockdown and FRET analysis, we found that the contribution of Asef to the EGF-induced Rac1 and Cdc42 activation was comparable to that of Tiam1 or Vav2 (Fig. 1B). To further examine the contribution of each GEF, we attempted the double or triple knockdown of GEFs. However, a substantial proportion of cells that had been transfected with siRNAs doubly or triply targeting Tiam1, Vav2 and Asef detached from culture dishes, suggesting that these GEFs cooperatively function to maintain the adherence of A431 cells. Notably, we could not observe any effect on EGF-induced Rac and Cdc42 activation when knocking down Sos. It has been reported that the Sos1-Grb2 complex and the Sos1-E3b1-Eps8 complex exert Ras- and Rac-specific GEF activities, respectively (Scita et al., 1999; Innocenti et al., 2002). In A431 cells, however, the double knockdown of Sos1 and Sos2 abrogated EGF-induced activation of Ras (supplementary material Fig. S3), but not Rac1 (Fig. 1C). Thus, the Sos-E3b1-Eps8 complex does not seem to have a major role in EGF-induced Rac1 activation or in the resulting induction of membrane ruffles in A431 cells.
Considering the large number of GEFs that activate Rac1 and Cdc42, it is not surprising to observe that none of the siRNAs targeting Rac1 or Cdc42 GEFs tested in this study completely inhibited the EGF-induced Rac1 and Cdc42 activation. Unexpectedly, however, we also did not observe any detectable difference in the spatio-temporal changes of Rac1 or Cdc42 activity in these A431 cells, i.e. the initial robust activation at the diffuse area and the following localized activation at membrane ruffles were similarly inhibited by siRNAs targeting Tiam1, Vav2 and Asef. Nevertheless, this observation does not exclude the possibility that each GEF has different roles in different cellular milieus. It is possible that we simply could not detect the difference at the resolution of the fluorescence microscope: the activities of Rac1 and Cdc42 might be differentially regulated, for example, between raft and non-raft membrane compartments. In this regard, it would be interesting to compare the activity change of the Rac1 and Cdc42 effectors in cells deficient for each GEF.
Identification of the principal phosphorylation site of Asef. (A) Domain structures of Asef and Asef2. ABR, APC-binding region; SH3, Src-homology 3; DH, Dbl-homology; PH, pleckstrin homology. (B) Alignment of amino-terminal sequences of Asef and Asef2. Tyr94 and Tyr104 are shown in bold, and the ABR and SH3 domains are underlined and boxed, respectively. The NcoI recognition site used to construct Asef2-Asef1 and Asef1-Asef2 mutants (Asef2/1 and Asef/2, respectively) is shown. (C,D) HA-tagged proteins were expressed transiently in A431 cells and immunoprecipitated with anti-HA antibody before or after 25 ng/mL EGF stimulation for 2 minutes. Immunoprecipitates and total-cell lysates were resolved by SDS-PAGE and probed with anti-phosphotyrosine (pTyr) or anti-HA antibody. (E) A431 cells expressing HA-tagged Asef were stimulated with or without 100 ng/mL EGF for 2 minutes. Cell lysates were separated by SDS-PAGE and analyzed by immunoblotting with the antibodies indicated above each lane. In the case of anti-phosphorylated-Y94Asef (α-pY94Asef), phenylphosphate was included in the reaction buffer to reduce the signals of non-specific anti-phosphotyrosine antibody. (F) HA-tagged constructs were expressed in A431 cells and immunoprecipitated with anti-HA antibody as in C. Immunoprecipitates and total cell lysates were analyzed using the antibodies as indicated.
Identification of the principal phosphorylation site of Asef. (A) Domain structures of Asef and Asef2. ABR, APC-binding region; SH3, Src-homology 3; DH, Dbl-homology; PH, pleckstrin homology. (B) Alignment of amino-terminal sequences of Asef and Asef2. Tyr94 and Tyr104 are shown in bold, and the ABR and SH3 domains are underlined and boxed, respectively. The NcoI recognition site used to construct Asef2-Asef1 and Asef1-Asef2 mutants (Asef2/1 and Asef/2, respectively) is shown. (C,D) HA-tagged proteins were expressed transiently in A431 cells and immunoprecipitated with anti-HA antibody before or after 25 ng/mL EGF stimulation for 2 minutes. Immunoprecipitates and total-cell lysates were resolved by SDS-PAGE and probed with anti-phosphotyrosine (pTyr) or anti-HA antibody. (E) A431 cells expressing HA-tagged Asef were stimulated with or without 100 ng/mL EGF for 2 minutes. Cell lysates were separated by SDS-PAGE and analyzed by immunoblotting with the antibodies indicated above each lane. In the case of anti-phosphorylated-Y94Asef (α-pY94Asef), phenylphosphate was included in the reaction buffer to reduce the signals of non-specific anti-phosphotyrosine antibody. (F) HA-tagged constructs were expressed in A431 cells and immunoprecipitated with anti-HA antibody as in C. Immunoprecipitates and total cell lysates were analyzed using the antibodies as indicated.
Role of Src-family tyrosine kinases on the Asef phosphorylation. (A) HA-Asef-expressing A431 cells were stimulated with 25 ng/ml EGF for 2 minutes. When indicated, cells were pre-treated with 4 μM PP2 for 30 minutes or with 10 μM STI-571 for 2 hours. HA-Asef was immunoprecipitated and analyzed using anti-phosphotyrosine (pTyr) or anti-HA antibodies. (B) HA-Asef expressed alone or co-expressed with v-Src, c-Src-Y537F mutant or c-Abl, was immunoprecipitated and analyzed by SDS-PAGE and immunoblotting using anti-phosphotyrosine (pTyr) or anti-HA antibodies. (C) A431 cells were transfected with HA-Asef (WT) or HA-AsefY94F (YF) in the presence or absence of the c-Src Y527F mutant. Asef was immunoprecipitated and analyzed using antibody against phospho-Tyr94 Asef (anti-pY94Asef) or anti-HA antibody. (D) HA-Asef and HA-AsefY94F expressed in 293F cells were immunoprecipitated using anti-HA antibody, phosphorylated using purified Src protein, and analyzed by immunoblotting as indicated. (E) HA-Asef-expressing A431 cells were stimulated with 50 ng/ml EGF for the indicated period and analyzed as in C.
Role of Src-family tyrosine kinases on the Asef phosphorylation. (A) HA-Asef-expressing A431 cells were stimulated with 25 ng/ml EGF for 2 minutes. When indicated, cells were pre-treated with 4 μM PP2 for 30 minutes or with 10 μM STI-571 for 2 hours. HA-Asef was immunoprecipitated and analyzed using anti-phosphotyrosine (pTyr) or anti-HA antibodies. (B) HA-Asef expressed alone or co-expressed with v-Src, c-Src-Y537F mutant or c-Abl, was immunoprecipitated and analyzed by SDS-PAGE and immunoblotting using anti-phosphotyrosine (pTyr) or anti-HA antibodies. (C) A431 cells were transfected with HA-Asef (WT) or HA-AsefY94F (YF) in the presence or absence of the c-Src Y527F mutant. Asef was immunoprecipitated and analyzed using antibody against phospho-Tyr94 Asef (anti-pY94Asef) or anti-HA antibody. (D) HA-Asef and HA-AsefY94F expressed in 293F cells were immunoprecipitated using anti-HA antibody, phosphorylated using purified Src protein, and analyzed by immunoblotting as indicated. (E) HA-Asef-expressing A431 cells were stimulated with 50 ng/ml EGF for the indicated period and analyzed as in C.
Earlier studies have shown that Asef is a GEF for Rac1 and a link between APC and G-protein signaling (Kawasaki et al., 2000; Kawasaki et al., 2003). Thus, Asef transduces a Wnt signal to Rac1, thereby regulating the actin cytoskeleton and cell migration (Akiyama and Kawasaki, 2006). More recently, it has been demonstrated that both Asef and Asef2 also show GEF activity for Cdc42 (Gotthardt and Ahmadian, 2007; Hamann et al., 2007; Kawasaki et al., 2007). Through the seminal work of Hall and colleagues, it has been demonstrated that Rho-family proteins function in a hierarchical cascade, wherein Cdc42 activates Rac1 and Rac1 – in turn – activates RhoA (Nobes and Hall, 1995). In this context, the suppression of EGF-induced Rac1 activity in Asef-knockdown A431 cells might reflect the suppression of Cdc42 activation. It should also be taken into consideration that both Rac1 and Cdc42 function in a synergistic manner in EGF-stimulated COS cells (Kurokawa and Matsuda, 2005). Therefore, the precise mechanism of the suppression of Rac1 and Cdc42 in Asef-knockdown cells remains to be clarified in a future study.
Detection of tyrosine-phosphorylated Asef at lamellipodia and membrane ruffles. (A,B) A431 cells expressing HA-Asef (A) or HA-AsefY94F (B) were serum starved for 24 hours and then stimulated with 25 ng/ml EGF for 2 minutes. Cells fixed and stained using antibody against phospho-Tyr94 Asef (anti-pY94Asef) or anti-HA antibody were analyzed using a confocal microscope. (C,D) Fluorescence intensities of HA-Asef-WT (C) or HA-Asef-Y94F (D) are plotted along the red lines in A and B.
Detection of tyrosine-phosphorylated Asef at lamellipodia and membrane ruffles. (A,B) A431 cells expressing HA-Asef (A) or HA-AsefY94F (B) were serum starved for 24 hours and then stimulated with 25 ng/ml EGF for 2 minutes. Cells fixed and stained using antibody against phospho-Tyr94 Asef (anti-pY94Asef) or anti-HA antibody were analyzed using a confocal microscope. (C,D) Fluorescence intensities of HA-Asef-WT (C) or HA-Asef-Y94F (D) are plotted along the red lines in A and B.
In this study we have shown that EGF stimulation phosphorylates Asef at Tyr94 within the APC-binding region in an Src-dependent manner. This finding adds Asef to the list of Src-activated GEFs targeting Rac1 and/or Cdc42, which includes Vav proteins, ARHG7 (also known as Cool-1 or beta-PIX), RGRF1 (also known as CDC25 or GRF1) and FRG (also known as FARP2) (Crespo et al., 1997; Kiyono et al., 2000; Miyamoto et al., 2003; Feng et al., 2006). It has been shown that phosphorylation of a tyrosine residue in the conserved acidic region of Vav1, Vav2 or Vav3 releases the auto-inhibition of their GEF activity (Llorca et al., 2005). It is uncertain, however, that a similar mechanism operates in Asef because we failed to observe any increase in the GEF activity upon Asef tyrosine phosphorylation or upon the introduction of a phospho-mimetic amino acid into Tyr94 of Asef in the overexpression experiments (data not shown). Recently, the crystal structure of Asef has been solved and it was found that the SH3 domain in the N-terminal regulatory domain has a crucial role in the auto-inhibition of GEF activity (Murayama et al., 2007; Mitin et al., 2007). In these crystallography experiments, the structure of the APC-binding region including Tyr94 was not determined. Thus, it will be necessary to solve the whole structure of Asef to understand the contribution of Tyr94 in the regulation of GEF activity.
The finding that Asef is phosphorylated within the APC-binding domain led us to examine the effect of Asef phosphorylation on binding to APC. Against our expectation, we failed to detect any effect when substituting Tyr94 and Tyr104 for either Phe or Glu (supplementary material Fig. S6). Thus, it will also be necessary to examine whether Tyr94 and Tyr104 are exposed at the APC-binding interphase by solving the structure of the APC-Asef complex.
Although the structural role of phosphorylated Tyr94 on the regulation of GEF activity, and the phosphorylation stoichiometry of Tyr94 of endogenous Asef remain unknown owing to technical reasons, the complementation experiment using the Asef-Y94F mutant strongly supported the notion that Tyr94 phosphorylation has a crucial role in EGF-induced Rac1 activation. Thus, this study adds Asef to the list of EGF-regulated activators of Rac1 and Cdc42. To untangle the specific role of Asef from the roles of the other GEFs, we probably need to examine the activity changes of Rac1 and Cdc42 in tissues or living animals. We are now developing new FRET biosensors that could be applicable in vivo imaging.
Materials and Methods
FRET biosensors and plasmids
The plasmids that encode FRET biosensors to measure Rac1 and Cdc42 activities, Raichu-Rac1/1011× and Raichu-Cdc42/1054×, have been described previously (Itoh et al., 2002). The RNA-targeting constructs were generated using pSuper.retro.puro vector (OligoEngine, Seattle, WA). The shRNA sequences used for Asef and firefly luciferase were 5′-AAGCCGACTTCCAGATCTACTCGGAGTACTG-3′ and 5′-GATTATGTCCGGTTATGTA-3′, respectively. pcDNA-HA-Asef and pcDNA-HA-Asef2 contain the hemagglutinin (HA)-tagged Asef and Asef2 cDNA, respectively, in pcDNA3.1 (+) expression vectors. In Asef1-Asef2 and Asef2-Asef1 mutants, the N-terminal regions of Asef and Asef2 were exchanged at the single NcoI site. Tyr32, Tyr94, Tyr104 and Tyr94-Tyr104 were replaced with Glu (E) or Phe (F) in AsefY32E, AsefY94E, AsefY94F, AsefY104E, AsefY104F and AsefY94E-Y104E. All mutations were introduced into the cDNA by PCR-based mutagenesis and verified by DNA sequencing. The pERedNLS-3HA expression vector contains a triple-HA tag and an internal ribosome entry site (IRES) followed by the cDNAs of DsRed-Express (Clontech, Mountain View, CA) and a nuclear localization signal. Dominant-negative mutants of Rac1 and Cdc42 have been reported previously (Kurokawa and Matsuda, 2005). For RNAi rescue experiments, an RNAi-resistant construct was created by introducing seven silent point mutations in the target sequences. pcDNA-HA-Tiam1C1199 was obtained from John G. Collard (The Netherlands Cancer Institute, Amsterdam, The Netherlands). pBabe-neo-v-Src and pMEX-neo-c-SrcY527F were provided by Shinya Tanaka (Hokkaido University Graduate School of Medicine, Japan). pEBG-cAbl was provided by Bruce J. Mayer (University of Connecticut Health Center, Farmington, CT).
Asef phosphorylation at Tyr94 has an essential role in EGF-induced activation of Rac1 and Cdc42. (A) A431 cells were transfected with vectors expressing wild-type Asef (WT) and siRNA-resistant Asef (WTR) were harvested at 48 hours after transfection and analyzed by SDS-PAGE and immunoblotting using the indicated antibodies. pSuper-Asef is an shRNA expression vector targeting Asef; pERedNLS-3HA-Asef-WT is a wild-type Asef expression vector; and pERedNLS-3HA-Asef-WTR is a wild-type Asef expression vector carrying RNAi-resistant nuleotide substitutions. R, siRNA-resistant expression plasmids. (B) A431 cells were transfected with expression vectors as indicated below each bar (WTR, pERedNLS-3HA-Asef-WTR; Y94FR, pERedNLS-3HA-Asef-Y94FR). After selection with puromycin, cells were transfected with pRaichu-Rac1 for 24 hours. After 3-6 hours of serum starvation, cells were stimulated with 25 ng/ml EGF and analyzed as described for Figs 2 and 3. Numbers of cells examined for each condition were as follows: control, n=22; Asef KD, n=21; Asef KD and Asef WTR; n=11; Asef KD and Asef Y94FR, n=11. *P<0.05, significant difference over the control cells (Student's t-test). *P<0.05. (C) Experiments were performed as in B except that pRaichu-Cdc42 was used instead of pRaichu-Rac1. Numbers of cells examined for each condition were as follows: control, n=10; Asef KD, n=10; Asef KD and Asef WTR, n=14; Asef KD and Asef Y94FR, n=11.
Asef phosphorylation at Tyr94 has an essential role in EGF-induced activation of Rac1 and Cdc42. (A) A431 cells were transfected with vectors expressing wild-type Asef (WT) and siRNA-resistant Asef (WTR) were harvested at 48 hours after transfection and analyzed by SDS-PAGE and immunoblotting using the indicated antibodies. pSuper-Asef is an shRNA expression vector targeting Asef; pERedNLS-3HA-Asef-WT is a wild-type Asef expression vector; and pERedNLS-3HA-Asef-WTR is a wild-type Asef expression vector carrying RNAi-resistant nuleotide substitutions. R, siRNA-resistant expression plasmids. (B) A431 cells were transfected with expression vectors as indicated below each bar (WTR, pERedNLS-3HA-Asef-WTR; Y94FR, pERedNLS-3HA-Asef-Y94FR). After selection with puromycin, cells were transfected with pRaichu-Rac1 for 24 hours. After 3-6 hours of serum starvation, cells were stimulated with 25 ng/ml EGF and analyzed as described for Figs 2 and 3. Numbers of cells examined for each condition were as follows: control, n=22; Asef KD, n=21; Asef KD and Asef WTR; n=11; Asef KD and Asef Y94FR, n=11. *P<0.05, significant difference over the control cells (Student's t-test). *P<0.05. (C) Experiments were performed as in B except that pRaichu-Cdc42 was used instead of pRaichu-Rac1. Numbers of cells examined for each condition were as follows: control, n=10; Asef KD, n=10; Asef KD and Asef WTR, n=14; Asef KD and Asef Y94FR, n=11.
Cells and reagents
Cells of the human epithelial carcinoma cell line A431 and and HeLa cells were maintained in DMEM (Sigma-Aldrich, St Louis, MO) supplemented with 10% fetal bovine serum. The cells were plated on 35-mm glass-base dishes (Asahi Techno Glass, Tokyo, Japan) that had been coated with collagen type I (Nitta Gelatin Inc., Osaka, Japan). FreeStyle 293F cells (derived from primary embryonal human kidney cells) were obtained from Invitrogen and maintained in Free Style 293 Expression medium (Invitrogen, San Diego, CA). Phenylphosphate disodium salt and puromycin were obtained from Sigma (St Louis, MO). EGF and Src family protein kinase inhibitor PP2 were purchased from Invitrogen and Calbiochem (EMD Chemical Inc., San Diego, CA), respectively. Expression plasmids were transfected into A431 cells by Polyfect (Qiagen, Valencia, CA), 293fectin or LipofectAMINE 2000 (Invitrogen).
Antibodies
A polypeptide corresponding to amino acids 90-99 of Asef, EEDL(P)YDDLHS, including phospho-Tyr94, was synthesized chemically. The antibody against Asef phosphorylated at Tyr94 (pY94Asef) was purified by affinity chromatography with a column to which the phospho-Asef peptide used for immunization had been linked. To remove the non-specific anti-phosphotyrosine antibody from the serum, the anti-pY94Asef antibody was used in the presence of 20 μg/ml phenylphosphate, a phosphotyrosine mimetic. The following antibodies were also used in this study: anti-HA high-affinity rat monoclonal antibody (Roche, Basel, Switzerland); antibody against phosphotyrosine (PY20), horseradish-conjugated anti-phosphotyrosine (RC20) antibody and anti-Sos1 (BD Transduction Laboratories, San Jose, CA), anti-phosphotyrosine monoclonal antibody (4G10; Upstate Biotechnology, Lake Placid, NY), monoclonal antibody against EGFR (Medical Biological Laboratories, Tokyo, Japan). Monoclonal antibody against Myc (9E10), polyclonal antibodies against Vav2 phosphorylated at Tyr172 (Tyr172), Vav2 (H-200) and Sos2 (C-19) were all from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal antibody against tubulin (Ab-1) was from Calbiochem, and Alexa-Fluor-488-conjugated antibody against rabbit IgG and Alexa-Fluor-594-conjugated antibody against rat IgG were from Molecular Probes (Eugene, OR).
RNA interference experiments
siRNA oligomers for RNA interference (RNAi) experiments comprise 19 nucleotides and were synthesized in sense and antisense directions with dTdT overhangs at each 3′ terminus. A human Sos1-targeting siRNA (sense, 5′-GGAACGUGUUCAAAAAAGU-3′) was purchased from Ambion (Austin, TX). A human Sos2-targeting siRNA (sense 5′-GCCUUUGCUAGAAAAUGCAGAAACU-3′, antisense 5′-AGUUUCUGCAUUUUCUAGCAAAGGCAU-3′; iGENE, Tsukuba, Japan), and human Vav2-targeting siRNA (sense 5′-CUACCUAAUUCACCUUCAAGGAAAGAG-3′, antisense, 5′-CUUUCCUUGAAGGUGAAUUAGGUAGAU-3′; iGENE). Synthetic siRNAs targeting Tiam1 were prepared as described previously (Malliri et al., 2004). Negative control siRNA tagged with Alexa-Fluor-555 was obtained from QIAGEN. siRNAs were transfected using Oligofectamine (Invitrogen) in A431 cells according to the manufacturer's instructions. After transfection, cells were incubated for at least 48 hours before analysis. A431 cells were transfected with the desired pSuper constructs by using Lipofectamine 2000 (Invitrogen). As a control shRNA, we used pSuper.retro.puro vector or pSuper-Luc designed to silence luciferase, an exogenous gene not present in A431 cells. After recovery, the cells were selected by two-day-incubation with 3.5 μg/ml puromycin and then used for further analysis. For FRET imaging, pRaichu-Rac1 was transfected into siRNA-transfected or the short-hairpin RNA (shRNA)-expressing cells one-day after RNAi treatment. In the case of the shRNA experiments, after two-day-incubation with puromycin, the cells were starved and used for imaging. For the rescue experiments, the cells were co-transfected with shRNA and the RNAi-resistant gene, then selected with 3.5 μg/ml puromycin for 4 days before imaging. In some FRET-imaging experiments, we also transfected the cells with Alexa-555-conjugated control siRNA (Alexa555 control siRNA:target siRNA, 1:10) to selectively mark the siRNA-transfected cells. For Bos's pull down assay, scramble siRNA and stealth siRNA against human Vav2, human Tiam1 and human Asef were purchased from Invitrogen. Target sequence of each siRNA are as follows: Vav2 (5′-AACUCCAGGAGGCUGUCGAAUUUCU-3′), Tiam1 (5′-UGAAGACGAAUGCUGCCAACUCUGG-3′) and Asef (5′-UUCAGUGAUGGAGAAGCCUGUCUCC-3′).
Time-lapse FRET imaging
Cells expressing FRET biosensors were starved for 6-12 hours with Phenol-Red-free MEM or DMEM/F12 medium, and then treated with 25 ng/ml of EGF essentially as described previously (Kurokawa and Matsuda, 2005). Cells were imaged with an IX81 inverted microscope (Olympus, Tokyo, Japan) equipped with a Cool SNAP-HQ cooled CCD camera (Roper Scientific, Trenton, NJ), an IX2-ZDC laser-based autofocusing system (Olympus), and an MD-XY30100T-Meta automatically programmable XY stage (SIGMA KOKI, Tokyo, Japan). The filters used for the dual-emission imaging were obtained from Omega Optical (Brattleboro, VT): an XF1071 (440AF21) excitation filter, an XF2034 (455DRLP) dichroic mirror, and two emission filters [XF3075 (480AF30) for CFP and XF3079 (535AF2) for FRET]. Cells were illuminated with a 75-W Xenon lamp through a 6% ND filter and viewed through a 60× oil immersion objective lens (PlanApo 60×/1.4). The exposure times for 4×4 binning were 500 ms for CFP and FRET images, and 100 ms for differential interference contrast images, respectively. After background subtraction, FRET:CFP ratio images were created with MetaMorph software (Universal Imaging, West Chester, PA) and these images were used to represent FRET efficiency. Normalized FRET:CFP ratio graphs were obtained as follows. First, in each sample, we determined the average ratio over the whole cell before EGF addition and used this ratio as the reference value. Then, the raw FRET:CFP ratio of each pixel was divided by the reference value, and this normalized ratio value was used to generate a normalized ratio graph. For the quantitative analysis of EGF-stimulated cells under various conditions, we obtained a peak value, which was taken as the highest FRET:CFP ratio value after EGF addition. This peak value of the control cells, i.e. without any inhibitors, expression vectors or siRNAs, was set as 100%. The peak values in test samples were similarly obtained and used to examine the effect of various reagents.
In vitro analysis of Rac1/Cdc42 activation
The Rac1 and Cdc42 activities in mock- or EGF-treated cells were measured by the Bos' pull-down method as described previously (Kurokawa et al., 2004). Briefly, cells were harvested in ice-cold lysis buffer (50 mM Tris pH 7.5, 100 mM NaCl, 2 mM MgCl2, 1% Nonidet P-40, 10% glycerol, 1 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride containing GST-PAK-CRIB. The cleared lysates were incubated with glutathione-Sepharose beads (Amersham Biosciences) for 30 minutes at 4°C. The washed beads were boiled in sample buffer, and both the bound proteins and cell lysates were analyzed by immunoblotting with anti-Rac1 or anti-Cdc42 antibody. Blots were quantified with the luminescent image analyzer LAS 1000 Plus (Fuji Film, Tokyo, Japan). The amount of GTP-Rac1 or GTP-Cdc42 was divided by that of total Rac1 or Cdc42 in cell lysates, respectively.
Reverse transcription (RT)-PCR
For the Asef expression analysis by using reverse transcription (RT)-PCR, shRNA-expressing single colonies were isolated after selection in the presence of 3.5 μg/ml puromycin. Total RNAs were extracted by using Sepasol-RNAI super (Nacalai-Tesque, Kyoto, Japan), and cDNA was synthesized from 2 μg total RNA by using the First-Strand cDNA Synthesis kit (Amersham Biosciences, Piscataway, NJ). The two pairs of human Asef primers were forward 5′-TTAGGAACTACACTGGCACC-3′ and reverse 5′-TTCTCCAGACTCTTTGGTCC-3′; and forward 5′-GGAGCATCAAGCCGACTTCC-3′ and reverse 5′-CAGTAGATGAGCTGGTGGTC-3′. For β-actin, the primer pair was forward 5′-GCGGGAAATCGTGCGTGACATT-3′ and reverse 5′-CGTGGATGCCACAGGACTCCATGC-3′. Using an Expand High Fidelity PLUS PCR System (Roche), human Asef cDNA and actin cDNA were amplified for 40 cycles and 20 cycles, respectively.
Immunoprecipitation
Transfected A431 cells were harvested in ice-cold lysis buffer (20 mM Tris-HCl pH 7.5, 100 mM NaCl, 1% Triton-X-100, 1 mM Na3VO4, 1 mM phenylmethylsulfunyl fluoride, 10 μg/ml aprotinin, and 3 μg/ml leupeptin). Anti-HA antibody was added to the cleared lysates. After 2 hours of incubation with protein G-Sepharose (Amersham Biosciences) at 4°C, the beads were washed and boiled in sample buffer. The bound proteins were analyzed by immunoblotting.
In vitro phosphorylation of Asef
293F cells were transfected with the expression plasmids for HA-Asef and AsefY94F. After 48 hours, cells were lysed in ice-cold lysis buffer (50 mM Tris pH 7.5, 100 mM NaCl, 2 mM MgCl2, 1% Nonidet P-40, 10% glycerol, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride), and immunoprecipitated with anti-HA antibody and protein G beads. The beads were washed twice with lysis buffer, and once with in vitro kinase (IVK) buffer (60 mM HEPES pH 7.5, 5 mM MgCl2, 5 mM MnCl2, 3 mM Na3VO4, 1.25 mM DTT). Beads were suspended in 25 μl of IVK buffer containing 200 mM ATP and 50 ng of GST-Src (Cell Signaling). After 30 minutes reaction at room temperature, the immunoprecipitates were washed with IVK buffer and subjected to SDS-PAGE, followed by immunoblotting using antibodies against HA, phosho-Tyr or phospho-Y94Asef.
Immunostaining
To stain exogenous HA-Asef protein, A431 cells were fixed with 3.7% formaldehyde, followed by permeabilization with 0.2% Triton X-100. After having been soaked for 1 hour in PBS containing 3% BSA and 0.02% Triton X-100, cells were incubated overnight at 4°C with pY94Asef antibody in the presence of 10 μg/ml phenylphosphate, washed with PBS, and then incubated for 30 minutes at room temperature with Alexa-Fluor-488 conjugated to anti-rabbit IgG. To stain HA, Alexa-Fluor-568 conjugated to anti-rat IgG was used as the secondary antibody. After washing, cells were imaged using an FV-500 confocal microscope equipped with an Arg laser and a HeNe laser microscope (Olympus).
Acknowledgements
We thank J. G. Collard, Y. Fukui, B. J. Mayer, and S. Tanaka for the provision of reagents and A. Nishiyama and Y. Kasakawa for their technical assistance. This work was supported by a Grant-in-Aid for Scientific Research on the Priority Area `Integrative Research toward the Conquest of Cancer' from the Ministry of Education, Culture, Sports, Science and Technology of Japan and by a grant from the New Energy and Industrial Technology Development Organization. R.E.I. was supported by research fellowships from the Japan Society for the Promotion of Science for Young Scientists.