Transformation potency of ErbB heterodimer signaling is determined by B-Raf kinase

Abstract

Cellular transformation occurs only in cells that express both ErbB1 and ErbB4 receptors, but not in cells expressing only one or the other of these receptors. However, when both receptors are coexpressed and ligand-stimulated, they interact with virtually the same adaptor/effector proteins as when expressed singly. To reveal the underlying regulatory mechanism of the kinase/phosphatase network in ErbB homo- and heterodimer receptor signaling, extracellular signal-regulated kinase (ERK) and Akt activities were evaluated in the presence of several enzyme inhibitors in ligand-induced cells expressing ErbB1 (E1), ErbB4 (E4), and ErbB1/ErbB4 (E1/4) receptor. The PP2A inhibitor okadaic acid showed receptor-specific inhibitory profiles for ERK and Akt activities. Moreover, B-Raf isolated only from E1/4 cells could induce in vitro phosphorylation for MEK; this B-Raf kinase activity was abolished by pretreatment of the cells with okadaic acid. Our study further showed that the E1/4 cell-specific B-Raf activity was stimulated by PLCγ and subsequent Rap1 activation. The present study suggests that B-Raf kinase, which was specifically activated in the cells coexpressing ErbB1 and ErbB4 receptors, elevates total ERK activity within the cell and, therefore, can induce cellular transformation.

Introduction

ErbB receptors are a family of membrane receptor tyrosine kinases that activate and recruit intracellular signaling pathways after the binding of growth factors. These receptors play essential roles in cellular proliferation and differentiation, and their overexpression and mutation are implicated in a variety of mammalian cancers (Olayioye et al., 2000; Yarden and Sliwkowski, 2001). The ErbB receptor family comprises the ErbB1/EGF receptor (EGFR), ErbB2, ErbB3, and ErbB4. The ErbB proteins share conserved intrinsic tyrosine kinase domains and extracellular ligand-binding domains distinguish their binding properties and affinities with several types of epidermal growth factor (EGF)-like ligands (Riese et al., 1996; Tzahar et al., 1996; Jones et al., 1999). These distinct binding properties result in diverse biological outputs. Further, the binding affinity of the ligand with an ErbB receptor and the cellular transformation potency are also modulated by coexpression of another ErbB receptor. For example, the binding of EGF with EGFR is enhanced by the coexpression of ErbB2 or ErbB3, and the binding of heregulin (HRG) with ErbB3 or ErbB4 is enhanced by the coexpression of EGFR or ErbB2 (Riese et al., 1995; Cohen et al., 1996; Zhang et al., 1996; Wang et al., 1998).

The ErbB1 and ErbB4 receptors possess strong affinities for EGF and HRG, respectively, and induce the MAPK cascade (Raf-MEK-ERK pathway) and PI3K-Akt/protein kinase B pathway (Yarden and Sliwkowski, 2001) upon ligand binding. However ligand-induced transformation of 3T3-7d cells takes place only when both receptors are cotransfected in the same cells (Cohen et al., 1996). The mechanism responsible for this biological response induced by the coexpression of different ErbB receptors remains unclear. Even when ErbB1 and ErbB4 are coexpressed and form a heterodimer upon ligand binding, the two receptors interact with the same signaling molecules with which each receptor interacts when expressed singly. For example, the ErbB1–ErbB4 heterodimer interacts with the Grb2, Shc, the p85 subunit of phosphatidylinositol 3′-kinase (PI3K), Cbl, and PLCγ in response to EGF in the same way that the ErbB1 homodimer does. Similarly, the same heterodimer behaves identically to the ErbB4 homodimer in binding to Grb2, Shc, and p85 in response to HRG (Cohen et al., 1996; Olayioye et al., 1998). Thus, the increase in biological response elicited by the coexpression of ErbB receptors cannot be simply explained by the common signaling cassettes recruited to the same ligand–receptor complex. Rather, it seems that some quantitative (strength or duration of the initial or intermediate signals) or qualitative differences (e.g., different regulation systems on the same signaling cascades are activated in the cells) influence the signal amplitudes (Simon, 2000; Heinrich et al., 2002) and thereby produce dissimilar biological responses in vivo. To understand the mechanism of cellular transformation in ErbB signal transduction, it is required to identify the factors that influence the cellular transformation potency produced by coexpression of different ErbB receptors.

In a previous study, we found that HRG-activated ErbB4 receptor induced ERK and Akt activation, and that Akt negatively regulated Raf-1 and the subsequent ERK activity in CHO cells expressing ErbB4 receptor (Hatakeyama et al., 2003). Furthermore, a computer simulation of the ErbB4 signal transduction suggested that the dynamics of ERK and Akt activities were modulated by protein phosphatase 2A (PP2A). The major regulatory signaling cassettes in ligand-stimulated ErbB4 receptor signaling are Shc-MAPK and PI3K-Akt pathways, and PP2A targets the dephosphorylation of MEK and Akt for signal attenuation within these pathways. In contrast, PP2A is known to activate Raf-1 in the MAPK pathway by dephosphorylating the inhibitory phosphoserine residue of the molecule (Dhillon et al., 2002). Thus, PP2A modulates the activation and attenuation of kinase activities in ErbB4 receptor signaling pathways.

In the present study, we observed distinctive regulation of PP2A on ERK and Akt activities in ligand-induced cell lines expressing ErbB1 (E1), ErbB4 (E4), and ErbB1/ErbB4 (E1/4) receptors. As for sensitivity toward a PP2A inhibitor okadaic acid, E4 and E1/4 cells showed virtually the same response and resulted in the inhibition of ERK activities. In addition, specific activation of B-Raf kinase, where PP2A acts as a positive regulator, was observed only in E1/4 cells and not in E1 or E4 cells. Further study showed that this E1/4 cell-specific B-Raf kinase activity was followed by PLCγ and Rap1 activation upon EGF stimulation. This phenomenon was consistent with data showing that the highest net ERK activity and cellular transformation was observed in E1/4 cells, and that B-Raf kinase seemed to contribute to the rise in overall ERK activity within the cells. Our results suggest that a difference in the distinct regulation of kinase/phosphatase in the signaling pathways results in different amplitudes in the conclusive kinase activities in ErbB1-, ErbB4 homo-, and heterodimer signaling, and determines cellular transformation potency.

Results

Only E1/4 cells induce cellular transformation

First, we compared the transformation abilities of CHO cells that express ErbB1 (E1), ErbB4 (E4), and ErbB1-ErbB4 (E1/4) in the presence of EGF or HRG. Foci formation was observed in ligand-stimulated E1/4 cells, but not in E1 or E4 cells (Figure 1). This result was consistent with a former study using ErbB1 and ErbB4 expressing NIH3T3 cells (Cohen et al., 1996), and it confirmed that cellular transformation occurs only when different ErbB receptors are coexpressed.

Figure 1
figure1

Focus formation assay for wild-type CHO cells, E1, E4, and E1/4 cells. Wild-type CHO cells (WT) and CHO cells expressing ErbB1 (E1), ErbB4 (E4) and both ErbB1 and ErbB4 receptors (E1/4) were incubated in the presence or absence of 1 nM EGF or 1 nM HRG in a serum-free medium. After incubation for 14 days, the cells were rinsed with PBS and fixed with 10 % formalin and stained with Giemsa reagent. Cont, absence of ligand; EGF, incubation with EGF; HRG, incubation with HRG

Receptor phosphorylation is a specific event associated with ligand-induced activation of ErbB protein tyrosine kinases. In our study, significant tyrosine phosphorylation was observed on the ErbB1 receptor in EGF-stimulated E1 and E1/4 cells and the ErbB4 receptor in HRG-stimulated E4 and E1/4 cells, respectively (Figure 2a). The data showed that the ligands specifically induced tyrosine phosphorylation on their high-affinity receptors.

Figure 2
figure2

Ligand-induced receptor phosphorylation in E1, E4, and E1/4 cells and formation of heterodimer in E1/4 cells. (a) E1, E4, and E1/4 cells were treated with 10 nM EGF or HRG for 1, 5, and 10 min. Cell lysates containing equal amounts of protein were subjected to immunoprecipitation (IP) using anti-ErbB1 or ErbB4 receptor antibody. After SDS–PAGE and membrane transfer, resolved protein bands were immunoblotted with anti-phosphotyrosine (PY20) antibody and later re-blotted with the ErbB1 or ErbB4 receptor antibody. (b) E1/4 cells were incubated in the presence or absence of 10 nM EGF or HRG for 30 s. Cell lysates containing equal amounts of protein were subjected to IP using anti-ErbB1, ErbB4 or PY20 antibodies, respectively. After Western blot analysis, resolved protein bands were immunoblotted with anti-ErbB1 or anti-ErbB4 antibody. Data show the representative figure of two independent experiments

To confirm the heterodimer formation in E1/4 cells, we isolated ErbB1, ErbB4, and tyrosine-phosphorylated proteins from ligand-stimulated E1/4 cells using the corresponding specific antibodies followed by Western blot analysis. Isolated proteins were detected with ErbB1 and ErbB4 receptor antibodies. Stronger band intensities of ErbB1 and ErbB4 proteins were observed after ligand stimulation in the ErbB4 and ErbB1 immunoprecipitates, respectively (Figure 2b). EGF treatment of E1/4 cells for 30 s enhanced ErbB1 phosphorylation by 1.68±0.54-fold and ErbB4 phosphorylation by 1.67±0.02-fold. Similarly, HRG treatment enhanced ErbB1 phosphorylation by 1.67±0.54-fold and ErbB4 phosphorylation by 2.43±0.37-fold (based on two independent experiments). This result is consistent with published data showing that coexpressed receptors form heterodimers in response to ligand binding (Tzahar et al., 1997; Wang et al., 1998). In this test, the wild-type CHO cells did not show detectable ErbB receptors (data not shown); accordingly, we estimated that the endogenous ErbB receptor level in the wild-type cells is extremely low.

Effect of wortmannin on ERK and Akt activities

Next, to observe the underlying regulatory mechanism of kinases and phosphatases in ErbB signaling, we tested the effect of several protein kinase and phosphatase inhibitors on ERK and Akt activation in wild-type CHO cells, E1, E4, and E1/4 cells. Activated ErbB1 and ErbB4 receptors have been shown to stimulate the MAPK cascade through the binding of Shc or Grb2 adaptor proteins. Subsequently, these receptors activate the PI3K-Akt cascade through the binding of the p85 subunit of PI3K to the receptor directly (in the case of ErbB4) (Cohen et al., 1996) or indirectly via the docking protein Gab1 (for ErbB1) (Rodrigues et al., 2000). As PI3K is an upstream regulator of Akt, the PI3K inhibitor wortmannin was added to observe the effect of PI3K on ERK and Akt activation in the cells. The addition of wortmannin suppressed the EGF-induced ERK activation in E1 cells. However, wortmannin treatment did not induce a remarkable change in ERK activation in the EGF-treated E4 and E1/4 cells and in the HRG-treated cells (Figure 3). Under these conditions, the Akt activity was suppressed in all of the cell lines tested. The suppression of ERK by wortmannin in EGF-induced E1 cells seemed to result from the PI3K-dependent MAPK activation (Oehrl et al., 2003); however, our data were insufficient to evaluate the direct PI3K-Ras interaction in E1 cells. On the other hand, our study showed Akt and ERK activation within the E4 cells where ErbB4 receptor phosphorylation was not detected by anti-phosphotyrosine antibody (PY20). Similar EGF-induced ERK activation in the ErbB4 receptor-expressing cell line was also observed in an earlier study (Shelly et al., 1998; Tzahar et al., 1998).

Figure 3
figure3

Inhibitory profile of wild-type CHO cells, E1, E4, and E1/4 cells by a PI3K inhibitor. Wild-type CHO cells, E1, E4, and E1/4 cells were pretreated in the presence or absence of a PI3K inhibitor wortmannin (Wort; 100 μ M) 10 min prior to the addition of the growth hormone. After treatment with growth hormone (10 nM EGF or HRG) for 10 min, cell lysates were subjected to Western blot analysis, and blotted with antibodies to detect phospho-p44/42 ERK (pERK) or phospho-Ser473 Akt (pAkt), and later re-blotted with anti-ERK (ERK) or anti-Akt (Akt) antibody, respectively. Data show the representative figure of three independent experiments

ERK and Akt were distinctly regulated by PP2A in different sets of ErbB receptors

In all cell lines, the MEK inhibitor PD98059 suppressed ERK activity but showed no effect on Akt activity (Figure 4), indicating that the MAPK pathway did not activate the PI3K-Akt pathway. When the PP2A inhibitor okadaic acid was tested, ERK and Akt activation were upregulated in the wild-type and E1 cells irrespective of the type of ligands (Figure 4). However, okadaic acid inhibited ERK activity in the E4 and E1/4 cells. As for Akt activity, EGF and HRG treatment resulted in different responses in E4 and E1/4 cells; in both cell lines, okadaic acid promoted Akt activity in response to EGF but showed no effect in response to HRG. The PP2A regulation of ERK activity in E4 and E1/4 cells was distinct from that in wild-type and E1 cells; we therefore inferred that expression of the ErbB4 receptor stimulated a specific pathway for ERK activation. Furthermore, our data showed that treatment with the PP2A inhibitor promoted Akt phosphorylation in the absence of ligands in the all cell lines tested. These results suggested that the PP2A action sites in the regulation of Akt and ERK activation are distinct from each other. We used two different ErbB-expressing clones to test the reproducibility of sensitivities toward these enzyme inhibitors. As these clones showed similar inhibitory patterns, we concluded that the expression level of the receptor in the cells did not alter the inhibitory profiles. The Akt activity observed in Figure 3 was not observed in the EGF-stimulated E1 cells in Figure 4, but this difference was due to the shorter exposure time used in Figure 4.

Figure 4
figure4

Inhibitory profile of wild-type CHO cells, E1, E4, and E1/4 cells by MEK inhibitor and PP2A inhibitor. Wild-type CHO cells, E1, E4, and E1/4 cells were pretreated in the presence or absence of PD98059 (PD; 50 μ M) or okadaic acid (OA; 2 μ M) 10 min prior to the addition of the 10 nM growth hormone (EGF or HRG). After treatment for 10 min, the cell lysates were subjected to Western blot analysis, blotted with antibodies to detect phospho-p44/42 ERK (pERK) or phospho-Ser473 Akt (pAkt), and later re-blotted with anti-ERK (ERK) or anti-Akt antibody (Akt), respectively. Data show the representative figure of three independent experiments

In MAPK (Raf-MEK-ERK) and PI3K-Akt pathways, PP2A is known to attenuate the activities of MEK (Keyse, 2000; Zhou et al., 2002) and Akt (Andjelkovic et al., 1996; Ivaska et al., 2002). These previous findings are consistent with our results showing PP2A inhibition of ERK and Akt in wild-type and E1 cells. However, this regulatory mechanism is unlikely to explain the positive stimulatory effect of PP2A on ERK activity in E4 and E1/4 cells. Accordingly, we speculated that the Raf activation by PP2A (Dhillon et al., 2002; Kubicek et al., 2002; Strack, 2002) might be responsible for the positive regulatory effect of PP2A on ERK activity.

B-Raf kinase is specifically activated in E1/4 cells

Earlier studies showed that dephosphorylation of Raf-1 by PP2A enhances the membrane translocation of Raf-1, which, in turn, facilitates Raf-1 kinase activation (Dhillon et al., 2002; Kubicek et al., 2002). Furthermore, the PKA-catalysed phosphoserine/threonine residues on Raf-1 were understood to be PP2A action targets (Peraldi et al., 1995). The mechanism of action of B-Raf is not as clear as that of Raf-1; however, it has been shown in the earlier study that B-Raf is also positively regulated by PP2A (Strack, 2002). Since Raf-1 is a common mediator of the MAPK cascade in ErbB receptor signaling (Yarden and Sliwkowski, 2001), we tested B-Raf activity. Since, B-Raf is also activated, it is likely that okadaic acid causes the inhibition of ERK in ErbB4-expressing cells as a result of Raf-1 and B-Raf inhibition even though MEK is activated by the same reagent.

Briefly, we tested EGF-induced B-Raf phosphorylation using a phospho-(Ser/Thr) PKA substrate motif antibody capable of detecting RXXT/S (PKA recognition site) phosphorylation. Our results showed that PKA phosphorylation in B-Raf was decreased after EGF administration in E1/4 cells, but increased in E1 and E4 cells (Figure 5a). Pretreatment of the cells with okadaic acid inhibited this de-phosphorylation (data not shown). Furthermore, B-Raf immunoprecipitates isolated from the EGF-treated E1/4 cells induced in vitro phosphorylation of MEK, and this B-Raf activity was inhibited by pretreatment with okadaic acid, whereas those from the E1 or E4 cells induced no such effect (Figure 5b). Thus, it was inferred that B-Raf kinase activity in E1/4 cells is suppressed by PKA phosphorylation at the basal state, and promoted by PP2A dephosphorylation after ligand stimulation.

Figure 5
figure5

E1/4 cell-specific B-Raf dephosphorylation and activation. (a) E1, E4, and E1/4 cells were treated in the presence or absence of 10 nM EGF for 3, 5, and 10 min. Cell lysate was immunoprecipitated (IP) with anti-B-Raf antibody and blotted with phospho-(Ser/Thr) PKA substrate motif antibody, and later re-blotted with anti-B-Raf antibody. (b) E1, E4, and E1/4 cells were incubated in the presence or absence of 10 nM EGF for 10 min with or without pre-incubation with okadaic acid (OA; 2 μ M). After the EGF treatment, B-Raf was isolated from the cell lysates using B-Raf-specific antibody with the Catch-and-Release immunoprecipitation system. The obtained soluble B-Raf fraction was incubated with recombinant MEK-1 and [γ-32P] ATP in the kinase buffer for 30 min at room temperature. Finally, 32P incorporation of the MEK-1 substrate was detected by autoradiogram after SDS–PAGE. Data show the representative figure of two independent experiments

It has been reported that B-Raf is activated through PLCγ and subsequent small GTPase Rap1 activation (York et al., 1998; Zwartkruis et al., 1998). Rap1-induced B-Raf activation is known to stimulate sustained activation of the MAPK cascade (York et al., 1998; Kao et al., 2001). To elucidate the B-Raf activation mechanism in E1/4 cells, we analysed in vitro B-Raf kinase activity in the presence of the PLCγ inhibitor U73122. Figure 6a shows that the kinase activity of B-Raf isolated from E1/4 cells was suppressed by this inhibitor. Similarly, Rap1 activity was specifically elevated in E1/4 cells in response to EGF administration and diminished by the same inhibitor (Figure 6b). Under these conditions, EGF did not elevate Rap1 activity in E4 cells. Basal Rap1 activity, however, was inhibited by U73122. Overall, B-Raf activation seemed to be a specific event for E1/4 cells and to induce an additional route for MEK-ERK activation (Guo et al., 2001). These findings suggest that this E1/4 cell-specific B-Raf activation by PLCγ as well as PP2A contributes to the elevation of total ERK activity in E1/4 cells.

Figure 6
figure6

Inhibition of B-Raf kinase activity by PLCγ inhibitor and E1/4 cell-specific Rap1 activation. (a) E1/4 cells were incubated in the presence or absence of EGF (10 nM) and PLC γinhibitor U73122 (10 μ M). B-Raf kinase activity was assayed as described in the Materials and methods section. The band intensities were quantified using a densitometer. (b) E1, E4, and E1/4 cells were incubated in the presence or absence of EGF and U73122. Cell lysates were incubated with GST-Ral GDS-RBD agarose to isolate Rap1. Agarose beads were subjected to Western blot analysis, and protein bands were detected using anti-Rap1 antibody (upper panel). The band intensities were quantified using a densitometer and are shown in arbitrary units (AU) (lower panel). Data show the representative figure of two independent experiments

To confirm this, we compared the levels of EGF-induced ERK phosphorylation between E1 and E1/4 cells and between E4 and E1/4 cells. ERK and phospho-ERK proteins were isolated from the cells by immunoprecipitation after treating the cells with 10 nM EGF for 1, 5, 10, and 30 min, and placed side by side on the SDS–PAGE gel. After membrane transfer, both ERK and phospho-ERK proteins were detected with anti-ERK antibodies (Figure 7a). The signal intensity ratios showing phospho-ERK versus ERK were compared between the E1 and E1/4 cells (Figure 7b) and between the E4 and E/4 cells (Figure 7c). The results showed that ERK originating from the EGF-induced E1/4 cells was much more highly phosphorylated than that from the E1 and E4 cells (Figure 7). However, this tendency was specific to EGF and could not be observed in association with the HRG-induced event, which results in identical levels of ERK activation in E4 and E1/4 cells (data not shown).

Figure 7
figure7

Comparison of the amplitudes of EGF-induced ERK activity in E1, E4, and E1/4 cells. E1, E4, and E1/4 cells were treated in the presence or absence of 10 nM EGF for 1, 5, 10, and 30 min. Immunoprecipitates against anti-ERK (IP: E) and anti-phospho-ERK (IP: P) antibodies from E1 and E1/4 cells, or E4 and E1/4 cells, at each time point, were placed side by side on the SDS–PAGE gel. After Western blot analysis, the proteins were detected using anti-ERK antibody. Finally, protein band intensities were quantified using a densitometer. (a) Comparisons of phospho-ERK levels in E1 and E1/4 cells (upper panel), and in E4 and E1/4 cells (lower panel). (b) Graph representing the phospho-ERK/ERK ratio in E1 (solid line) and E1/4 (dotted line) cells. (c) Graph representing the phospho-ERK/ERK ratio in E4 (solid line) and E1/4 (dotted line) cells

Discussion

Cells that coexpress different ErbB receptors tend to undergo cellular transformation more frequently than cells that express a single receptor (Cohen et al., 1996; Zhang et al., 1996). In this study, we compared cells expressing ErbB1, ErbB4, and ErbB1–4 and found that distinct PP2A and PLCγ regulatory mechanisms exist among these cell lines. We also discovered that cells coexpressing both the ErbB1 and ErbB4 receptors could activate B-Raf kinase and induce cellular transformation.

Ligand-stimulated ErbB1 and ErbB4 receptors activate the MAPK cascade and PI3K-Akt pathway. Raf-1, located at the first step in the MAPK cascade, determines the amplitude of ERK activity. Three types of Raf have been identified in mammals: A-Raf, B-Raf, and Raf-1 (=C-Raf). All the three Raf isoforms share MEK as a common downstream substrate (Kolch, 2000; Chong et al., 2003), and PP2A has been shown to be a common positive regulator for Raf-1 (Abraham et al., 2000; Jaumot and Hancock, 2001) and B-Raf (Strack, 2002). The ubiquitous Raf-1 is the most studied Raf isoform, but the role of B-Raf in ErbB signaling has not been intensively studied, notwithstanding the importance of B-Raf in human cancer (Brose et al., 2002). Since B-Raf is an upstream regulator of MEK (Busca et al., 2000) and displays higher MEK kinase activity than Raf-1 (Papin et al., 1998), B-Raf activation should greatly facilitate final ERK activity.

In our study, cellular responses to several enzyme inhibitors, including okadaic acid, were virtually the same in both E4 and E1/4 cells; however, B-Raf kinase activity was specifically elevated in E1/4 cells. B-Raf is known to be activated (MacNicol and MacNicol, 1999) or inhibited (Peraldi et al., 1995) by PKA in a cell-specific manner. In our study, the basal inhibitory phosphorylation of B-Raf by PKA and PP2A-induced stimulatory dephosphorylation of B-Raf seemed to be observed only in E1/4 cells. On the other hand, in E1 and E4 cells, EGF-stimulated PKA seemed to phosphorylate B-Raf, thereby inhibiting its kinase activity. The difference in the roles of PKA and PP2A for B-Raf activation in these cell lines is not clear. B-Raf activation in E1/4 cells seems to be more relevant than Raf-1 activation (e.g. activation by small G-protein binding and PP2A dephosphorylation), while the suppression mechanism of B-Raf activity in E1 and E4 cells seems to be more complex.

Phosphorylation and dephosphorylation regulate and determine the function of the proteins in the signal transduction cascade in response to extracellular stimuli. If we focus on an MAPK (Raf-1-MEK-ERK) cascade, there are two PP2A action sites. One is MEK, where PP2A acts as a signal attenuator by dephosphorylating the protein. The other is Raf-1; however, at this site, PP2A acts as a positive regulator. Accordingly, whether PP2A stimulates or inhibits the entire MAPK cascade depends on the contribution made by the ratio of the Raf and MEK kinase in the cascade. It is clear from Figure 4 that PP2A acted as a negative regulator of the MAPK cascade in wild-type and E1 cells, and therefore it is inferred that MEK makes a major contribution to activation of the cascade in these cell lines. With respect to the cells expressing both ErbB1 and ErbB4, our data showed that there were additional positive regulatory sites for PP2A, B-Raf, in addition to Raf-1. Consequently, it was inferred from Figure 4 that PP2A acts as a positive regulator in the entire MAPK cascade in this cell line beyond its inhibitory role on MEK. However, since the ERK activity in E4 cells was inhibited by the PP2A inhibitor but did not show B-Raf activity, there remains a possibility that B-Raf is not a control point for ERK activation in ErbB4-expressing cells. More intensive study is needed to elucidate B-Raf activation mechanisms caused by different ErbB receptor-expressing cells.

In earlier investigations of EGF-induced ErbB1 receptor signaling, PP2A was shown to suppress Akt activity through direct dephosphorylation of the enzymes (Andjelkovic et al., 1996; Keyse et al. 2000; Ivaska et al., 2002; Zhou et al., 2002). When we used okadaic acid to suppress PP2A, Akt activation took place in all cell lines tested in concert with a direct phosphorylation signal after EGF stimulation when detected by anti-PKA substrate motif (RXXT/S) antibody (data not shown). Therefore, we inferred that constitutive PP2A suppressed this PKA-induced Akt activity in the basal state. PKA substrate motifs are located at Thr72, Ser124, and Ser246 in the Akt molecule. A PKA recognition site, Arg69-X-X-Thr72, was found to reside in the PH domain, a domain essential for the interaction of Akt with lipid membrane (Thomas et al., 2002). Therefore, we postulate that Thr72 phosphorylation by PKA may facilitate the localization of Akt, required for its activation (Aoki et al., 1998; Sable et al., 1998).

There is also the unsolved question of why PLCγ in E1/4 cells can activate B-Raf through Rap1 even though the ErbB1 receptor also induces PLCγ activation in response to EGF (Cohen et al., 1996; Olayioye et al., 1998). One clear difference between E1, E4, and E1/4 cells is that E1/4 cells form receptor heterodimers, which may facilitate the specific signaling that causes the PLCγ activation and PKA-PP2A regulation that is essential for B-Raf activation. However, further study is needed to find a direct link among this heterodimer formation, specific PLCγ activation, and PKA-PP2A regulation.

Overall, our results indicate that specific regulation of the kinases and phosphatases promotes cellular transformation of the cells that coexpress different ErbB receptors. In CHO cells, coexpression of ErbB1 and ErbB4 receptors specifically induced B-Raf activation through PLCγ, and this additional route for MAPK cascade, in addition to the Raf-1 route, resulted in a rise in total ERK activity. Even though the direct link between heterodimer formation and PLCγ activation has yet to be discovered, we surmised that ErbB1/4 heterodimer formation triggers a specific signaling event that activates PLCγ. Furthermore, the present study shows a specific example that the signaling amplitude of conclusive kinase activities leading to cellular transformation may be determined by the underlying differential regulatory mechanisms of PKA and PP2A on the effector enzymes in cellular signal transduction.

Materials and methods

Materials

Recombinant human heregulin-β176–246 (HRG) was purchased from R&D Systems (Minneapolis, MN, USA). EGF was purchased from PeproTech House (London, England). Antibodies for detecting phospho-p44/42 ERK, phospho-Ser473 Akt, phospho-Ser/Thr PKA substrate motif, ERK, and Akt were purchased from Cell Signaling Technology, Inc. (Beverly, MA, USA). Anti-ErbB1 receptor, anti-ErbB4 receptor, anti-phosphotyrosine (PY20), anti-Rap1 antibody, and anti-B-Raf were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Human recombinant MEK1 was purchased from Upstate Biotechnology (Lake Placid, NY, USA). Okadaic acid (PP2A inhibitor), PD98059 (MEK inhibitor), U73122 (PLCγ inhibitor), and wortmannin (PI3K inhibitor) were obtained from Calbiochem (San Diego, CA, USA). GST-Ral GDS-Rap binding domain (RDB) agarose was purchased from Upstate Biotechnology (Lake Placid, NY, USA).

A method for constructing Chinese hamster ovary (CHO) cells expressing full-length human ErbB1 (EGFR) or ErbB4 receptor is described elsewhere (Kim et al., 2002). The cells coexpressing ErbB1 and ErbB4 receptors were constructed as follows: ErbB1/pcDNA3.1/Zeo was digested by NheI–XbaI, and inserted into the NheI–XbaI site of the mammalian expression plasmid pcDNA3.1/Zeo (+) (Invitrogen Corp., Carlsbad, CA, USA). The plasmid was then transfected into ErbB4/Zeo cells by FuGENE6 (Roche Diagnostics, Basel, Switzerland), and stable transfectants were selected by G418 (neomycin).

Cell culture

CHO cells expressing ErbB1, ErbB4 or ErbB1–4 receptors (E1, E4, and E1/4 cells, respectively) were routinely maintained in DMEM/F12 (Gibco BRL, Githersburg, MD, USA) medium supplemented with 10% bovine calf serum and antibiotics. For detection of the effect of growth hormones, the cells were starved in serum-free DMEM/F12 for 16–24 h prior to the experiment. To test the effect of kinase and phosphatase inhibitors, the cells were pretreated with the inhibitors 10 min prior to the addition of the growth hormone.

Focus assay

E1, E4, and E1/4 cells were transferred to six-well dishes and grown in phenol red-free DMEM supplemented with 5% calf serum that had been heat-inactivated and treated with 0.25%. (w/v) dextran-coated charcoal (5% CDCS-DMEM) to remove mitogenic compounds. When the cells reached confluence, the medium was replaced with serum-free DMEM and incubated for 14 days in the presence or absence of 1 nM EGF or 1 nM HRG. After the incubation, the cells were fixed with 10% formalin and stained with Giemsa reagent.

Western blot analysis

The growth hormone-stimulated cells were rinsed with ice-cold PBS and lysed with cell lysis buffer (pH 7.4) containing 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS, PBS and protease inhibitors. Cell lysate was cleared by centrifugation, and the protein concentration of the supernatant was determined using protein assay reagent (Bio-Rad laboratories, Hercules, CA, USA). For the detection of phosphorylated-ErbB receptors, cell lysate samples containing equal amounts of protein were subjected to immunoprecipitation using each anti-ErbB antibody and then analysed by Western blot analysis. The resolved proteins were blotted with anti-phosphotyrosine (PY20) antibody. Alternatively, the protein bands were later re-blotted with the corresponding anti-ErbB antibody. To test for receptor heterodimer formation, cell lysates containing equal amounts of protein were immunoprecipitated with antibodies against ErbB1, ErbB4 receptor, or phosphotyrosine (PY20) and detected using ErbB4 or ErbB1 receptor antibody, respectively.

We examined ERK phosphorylation as a downstream marker of the MAPK cascade and Akt phosphorylation at Ser473 as a downstream marker of the PI3K-Akt pathway (Resjö et al., 2002). Since the doubly phosphorylated forms of ERK (Payne et al., 1991) and Akt (Andjelkovic et al., 1999) function as active enzymes, we used the phosphorylation of these enzymes as a direct activation marker. To detect the active form of ERK, protein was subjected to Western blot analysis and detected using anti-phospho-p44/42 ERK antibody. To compare the level of ERK phosphorylation in EGF-induced cells expressing ErbB1, ErbB4, and ErbB1–4, we performed immunoprecipitation using anti-ERK and anti-phospho-p44/42 antibodies for the same protein samples, respectively, and placed the immunoprecipitates side by side on the gel. After membrane transfer, both proteins were detected using anti-ERK antibody and the band intensities were quantified using a densitometer (Fuji Film Corp, Tokyo, Japan). The ratio of phosphorylated-p44/42 ERK proteins to total ERK proteins was calculated.

To detect the phospho-PKA substrate motif in B-Raf, B-Raf was immunoprecipitated with the Catch-and-Release immunoprecipitation system (Upstate Biotechnology) according to the manufacturer’s protocol and subjected to Western blot analysis. Protein bands corresponding to B-Raf were detected using phospho-PKA substrate motif antibody.

Rap1 pull-down assay

Rap1 activation assay was performed exactly as described by Zwartkruis et al. (1998). Briefly, E1/4 cells were grown to 70% confluency and serum-starved as described above. Following EGF stimulation for 5 min, the cells were washed twice with PBS and lysed in Ral-buffer (10% glycerol, 1% Nonidet P-40, 50 mM Tris, pH 7.4, 200 mM NaCl, 2.5 mM MgCl2, protease inhibitors, 1 mM NaF, and 1 mM Na3VO4). Lysates were clarified by centrifugation and the supernatants were incubated with 15 μg GST-Ral GDS-RDB agarose for 1 h to isolate Rap1. Agarose beads were washed three times in Ral buffer and subjected to Western blot analysis. Protein bands were detected using anti-Rap1 antibody. To observe the effect of PLCγ on Rap1 activation, 10 μ M U73122 was added to the incubation medium 10 min prior to the addition of EGF. The band intensities were quantified by a densitometer.

B-Raf immunoprecipitation and in vitro kinase activity assay

Cells were cultured in a serum-free medium and treated with 10 nM EGF for 10 min with or without pre-incubation with the enzyme inhibitor for 10 min, then lysed in a lysis buffer containing 50 mM HEPES (pH 7.5), 10 nM EDTA, 150 mM NaCl, 10 mM sodium pyrophosphate, 2 mM sodium orthovanadate, 100 mM NaF, 1% Triton X-100 (v/v), 100 U/ml aprotinin, 20 mM leupeptin, and 0.18 mg/ml PMSF. B-Raf was immunoprecipitated using the B-Raf-specific antibody with the Catch-and-Release immunoprecipitation system according to the manufacturer’s protocol. The obtained soluble B-Raf fraction was incubated with 1 μg of recombinant MEK-1 and 160 μ M [γ32-P] ATP (6000 Ci/mmol) in a kinase buffer containing 20 mM MOPS (pH 7.2), 500 μ M ATP, 75 mM MgCl2, 25 mM glycerophosphate, 5 mM EGTA, 1 mM sodium orthovanadate, and 1 mM dithiothreitol. After incubation of the reaction mixture for 30 min at room temperature, the B-Raf kinase reaction was stopped by the addition of SDS–PAGE loading buffer and the proteins were subjected to SDS–PAGE. The band corresponding to the 32P-labeled MEK protein was analysed using the BAS2000 system (Fuji Film Corp., Tokyo, Japan).

Abbreviations

CHO:

Chinese hamster ovary

EGFR:

EGF receptor

EGF:

epidermal growth factor

HRG:

heregulin

MAPK:

mitogen-activated protein kinase

MEK:

extracellular signal-regulated kinase kinase

ERK:

extracellular signal-regulated kinase

SH2:

Src homology 2 domain

Grb2:

growth factor receptor-binding protein 2

Shc:

Src homology and collagen domain protein

PI3K:

phosphatidylinositol 3′-kinase

PLCγ:

phosphoinositide-specific phospholipase C-γ

PKA:

protein kinase A

PH domain:

pleckstrin homology domain

PP2A:

protein phosphatase 2A

References

  1. Abraham D, Podar K, Pacher M, Kubicek M, Welzel N, Hemmings BA, Dilworth SM, Mischak H, Kolch W and Baccarini M . (2000). J. Biol. Chem., 275, 22300–22304.

  2. Andjelkovic M, Jakubowicz T, Cron P, Ming XF, Han JW and Hemmings BA . (1996). Proc. Natl. Acad. Sci. USA, 93, 5699–5704.

  3. Andjelkovic M, Maira SM, Cron P, Parker PJ and Hemmings BA . (1999). Mol. Cell Biol., 19, 5061–5072.

  4. Aoki M, Batista O, Bellacosa A, Tsichlis P and Vogt PK . (1998). Proc. Natl. Acad. Sci. USA, 95, 14950–14955.

  5. Brose MS, Volpe P, Feldman M, Kumar M, Rishi I, Gerrero R, Einhorn E, Herlyn M, Minna J, Nicholson A, Roth JA, Albelda SM, Davies H, Cox C, Brignell G, Stephens P, Futreal PA, Wooster R, Stratton MR and Weber BL . (2002). Cancer Res., 62, 6997–7000.

  6. Busca R, Abbe P, Mantoux F, Aberdam E, Peyssonnaux C, Eychene A, Ortonne JP and Ballotti R . (2000). EMBO J., 19, 2900–2910.

  7. Chong H, Vikis HG. and Guan K-L . (2003). Cell. Signal., 15, 463–469.

  8. Cohen BD, Kiener PA, Green JM, Foy L, Fell HP and Zhang K . (1996). J. Biol. Chem., 271, 30897–30903.

  9. Dhillon AS, Pollock C, Steen H, Shaw PE, Mischak H and Kolch W . (2002). Mol. Cell Biol., 22, 3237–3246.

  10. Guo FF, Kumahara E and Saffen D . (2001). J. Biol. Chem., 276, 25568–25581.

  11. Hatakeyama M, Kimura S, Naka T, Kawasaki T, Yumoto N, Ichikawa M, Kim JH, Saito K, Saeki M, Shirouzu M, Yokoyama S and Konagaya A . (2003). Biochem. J., 373, 451–463.

  12. Heinrich R, Neel BG and Rapoport TA . (2002). Mol. Cell, 9, 957–970.

  13. Ivaska J, Nissinen L, Immonen N, Eriksson JE, Kahari VM and Heino J . (2002). Mol. Cell Biol., 22, 1352–1359.

  14. Jaumot M and Hancock JF . (2001). Oncogene, 20, 3949–3958.

  15. Jones JT, Akita RW and Sliwkowski MX . (1999). FEBS Lett., 447, 227–231.

  16. Kao S, Jaiswal RK, Kolch W and Landreth GE . (2001). J. Biol. Chem., 276, 18169–18177.

  17. Keyse SM . (2000). Curr. Opin. Cell Biol., 12, 186–192.

  18. Kim JH, Saito K and Yokoyama S . (2002). Eur. J. Biochem., 269, 2323–2329.

  19. Kolch W . (2000). Biochem J., 351, 289–305.

  20. Kubicek M, Pacher M, Abraham D, Podar K, Eulitz M and Baccarini M . (2002). J. Biol. Chem., 277, 7913–7919.

  21. MacNicol MC and MacNicol AM . (1999). J. Biol. Chem., 274, 13193–13197.

  22. Oehrl W, Rubio I and Wetzker R . (2003). J. Biol. Chem., 278, 17819–17826.

  23. Olayioye MA, Graus-Porta D, Beerli RR, Rohrer J, Gay B and Hynes NE . (1998). Mol. Cell Biol., 18, 5042–5051.

  24. Olayioye MA, Neve RM, Lane HA and Hynes NE . (2000). EMBO J., 19, 3159–3167.

  25. Papin C, Denouel-Galy A, Laugier D, Calothy G and Eychène A . (1998). J. Biol. Chem., 273, 24939–24947.

  26. Payne DM, Rossomando AJ, Martino P, Erickson AK, Her JH, Shabanowitz J, Hunt DF, Weber MJ and Sturgill TW . (1991). EMBO J., 10, 885–892.

  27. Peraldi P, Frodin M, Barnier JV, Calleja V, Scimeca JC, Filloux C, Calothy G and Van Obberghen E . (1995). FEBS Lett., 357, 290–296.

  28. Resjö S, Göransson O, Härndahl L, Zolnierowicz S, Manganiello V and Degerman E . (2002). Cell. Signal., 14, 231–238.

  29. Riese II DJ, van Raaij TM, Plowman GD, Andrews GC and Stern DF . (1995). Mol. Cell Biol., 15, 5770–5776.

  30. Riese DJ, Kim ED, Elenius K, Buckley S, Klagsbrun M, Plowman GD and Stern DF . (1996). J. Biol. Chem., 271, 20047–20052.

  31. Rodrigues GA, Falasca M, Zhang Z, Ong SH and Schlessinger J . (2000). Mol. Cell Biol., 20, 1448–1459.

  32. Sable CL, Filippa N, Filloux C, Hemmings BA and Van Obberghen E . (1998). J. Biol. Chem., 273, 29600–29606.

  33. Shelly M, Pinkas-Kramarski R, Guarino BC, Waterman H, Wang LM, Lyass L, Alimandi M, Kuo A, Bacus SS, Pierce JH, Andrews GC and Yarden Y . (1998). J. Biol. Chem., 273, 10496–10505.

  34. Simon MA . (2000). Cell, 103, 13–15.

  35. Strack S . (2002). J. Biol. Chem., 277, 41525–41532.

  36. Thomas CC, Deak M, Alessi DR and van Aalten DMF . (2002). Curr. Biol., 12, 1256–1262.

  37. Tzahar E, Moyer JD, Waterman H, Barbacci EG, Bao J, Levkowitz G, Shelly M, Strano S, Pinkas-Kramarski R, Pierce JH, Andrews GC and Yarden Y . (1998). EMBO J., 17, 5948–5963.

  38. Tzahar E, Waterman H, Chen X, Levkowitz G, Karunagaran D, Lavi S, Ratzkin BJ and Yarden Y . (1996). Mol. Cell Biol., 16, 5276–5287.

  39. Tzahar E, Pinkas-Kramarski R, Moyer JD, Klapper LN, Alroy I, Levkowitz G, Shelly M, Henis S, Eisenstein M, Ratzkin BJ, Sela M, Andrews GC and Yarden Y . (1997). EMBO J., 16, 4938–4950.

  40. Wang LM, Kuo A, Alimandi M, Veri MC, Lee CC, Kapoor V, Ellmore N, Chen XH and Pierce JH. . (1998). Proc. Natl. Acad. Sci. USA, 95, 6809–6814.

  41. Yarden Y and Sliwkowski MX . (2001). Nat. Rev. Mol. Cell Biol., 2, 127–137.

  42. York RD, Yao H, Dillon T, Ellig CL, Eckert SP, McCleskey EW and Stork PJ . (1998). Nature, 392, 622–626.

  43. Zhang K, Sun J, Liu N, Wen D, Chang D, Thomason A and Yoshinaga SK . (1996). J. Biol. Chem., 271, 3884–3890.

  44. Zhou B, Wang Z, Zhao Y, Brautigan DL and Zhang Z . (2002). J. Biol. Chem., 277, 31818–31825.

  45. Zwartkruis FJ, Wolthuis RM, Nabben NM, Franke B and Bos JL . (1998). EMBO J., 17, 5905–5912.

Download references

Acknowledgements

We thank Ms Mihoro Saeki for constructing CHO cells expressing ErbB1 and ErbB4 receptors. We also thank Dr Takashi Naka for critically reading the manuscript and Mr Akinobu Fukuzaki for preparing the manuscript.

Author information

Affiliations

Authors

Corresponding author

Correspondence to M Hatakeyama.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hatakeyama, M., Yumoto, N., Yu, X. et al. Transformation potency of ErbB heterodimer signaling is determined by B-Raf kinase. Oncogene 23, 5023–5031 (2004). https://doi.org/10.1038/sj.onc.1207664

Download citation

Keywords

  • ErbB
  • B-Raf
  • transformation
  • PP2A
  • PKA

Further reading

Search