Role of Bim in the survival pathway induced by Raf in epithelial cells

Article metrics

Abstract

Selective and sustained activation of the Raf/MAP kinase pathway in MCF-10A ΔRaf-ER cells, a spontaneously immortalized human mammary epithelial cell line, was previously shown to protect these cells from suspension-induced cell death, a critical feature of the Ras-transformed phenotype. Although autocrine signalling through the EGF receptor is crucial for the protection induced by Raf in these cells, we report here the existence of an additional, more direct survival mechanism, linking Raf activation to the inhibition of a proapoptotic member of the Bcl-2 family, Bim. While detachment from the matrix results in transcriptional induction of two variants of this BH3-only protein, BimEL and BimL, activation of the Raf/ERK signalling both prevents Bim upregulation specifically and leads to phosphorylation and degradation of the BimEL isoform. This represents an important route to protect epithelial cells from the proapoptotic effect of Bim.

Introduction

A critical component of the Ras-transformed phenotype is the protection of epithelial cells from detachment-induced apoptosis, or anoikis (Frisch and Francis, 1994). Unlike normal epithelial cells, in fact, transformed cells acquire the ability to survive in the absence of contact to extracellular matrix proteins and therefore they can proliferate away from their original correct location, contributing to the invasive phenotype. Thus, understanding how Ras signalling imparts adhesion independence may reveal important targets for therapeutic intervention. Although the Ras effector PI3K and its downstream target Akt were shown to be the principal mediators of Ras-mediated protection of epithelial cells from anoikis (Khwaja et al., 1997; Rytomaa et al., 2000), recent data highlight the essential contribution of the Raf/MEK/ERK pathway in mediating anoikis resistance (Le Gall et al., 2000; McFall et al., 2001; Schulze et al., 2001). In this signalling pathway, the survival signal is transduced from Ras through the serine/threonine kinase Raf, which in turn phosphorylates and activates mitogen-activated protein (MAP) kinase kinases MEK1 and 2 which are capable of phosphorylating and activating the p42/44 MAP kinases (p42 and p44 MAPKs) also known as extracellular signal-regulated kinases (ERK2 and 1, respectively) (reviewed in Downward, 2003).

Many central components of the apoptotic machinery are also controlled by Ras effector pathways, reflecting the fact that cell survival and cell proliferation are tightly linked to assure tissue homeostasis. This is highlighted by the finding that key regulators of proliferation signalling also function in the control of cell survival, for instance by regulating the activity of members of the Bcl-2 family. This includes both pro- as well as anti-apoptotic molecules that play a critical role in most cell-death decisions by controlling the release of mitochondrial apoptogenic factors. What distinguishes these proteins as part of the Bcl-2 family is the presence of at least one of four evolutionarily conserved motifs, known as the Bcl-2 homology (BH) domains, termed BH1, BH2, BH3 and BH4. While antiapoptotic members, such as Bcl-2, Bcl-XL and Mcl-1, contain all four domains, the proapoptotic counterparts are subdivided into two groups: multidomain proteins such as Bax and its close relative Bak, which possess BH1, BH2 and BH3 regions, and BH3 domain-only proteins such as Bim, Bid, Bad and Bmf. The BH3-only members are thought to act as sensors of a wide variety of cell death signals, which depend on mitochondrial dysfunction for their execution, and promote cell death at a point upstream of mitochondria and Bax subfamily activation (reviewed in Borner, 2003).

In the case of the Raf/MEK/ERK signalling pathway, activation of the MEK/ERK cascade has been shown to influence the apoptotic machinery by inhibiting the proapoptotic activity of the BH3-only protein Bad (Bonni et al., 1999; Scheid et al., 1999) as well as by upregulating the antiapoptotic proteins Bcl-2, Bcl-XL and Mcl-1(Liu et al., 1999; Leu et al., 2000; Jost et al., 2001).

In recent years, a great deal has been learnt about another member of the Bcl-2 family, the proapoptotic BH3-only protein Bim/BOD. This was originally identified by two independent groups as a Bcl-2- (O’Connor et al., 1998), or a Mcl-1-interacting protein (Hsu et al., 1998). Although several Bim/BOD isoforms have been characterized (Marani et al., 2002, U et al., 2001), only the proapoptotic activity of the longer variants, BimEL and BimL, is controlled by sequestration to the cytoskeleton-associated dynein motor complex bound to dynein light chain LC8 (Puthalakath et al., 1999). Apoptotic signals that impinge on the microtubular network can trigger the release of Bim from the microtubules and its association with other family members (Bouillet et al., 1999), a requisite for Bim-induced apoptosis. In this context, Bim represents a potential candidate for mediating cell death induced by loss of matrix attachment, an apoptotic signal that impinges on the cytoskeleton and motor complexes.

In this report, we provide evidence for a role for Bim in the mechanism of Ras-induced protection from anoikis. We employed the cell line MCF-10A ΔRaf-ER, a normal, spontaneously immortalized human luminal mammary epithelial cell line, a cell type which requires integrin-mediated adhesion to the extracellular matrix for survival (Streuli and Gilmore, 1999). Previous data showed that although activation of Raf alone in these cells is not sufficient to induce transformation, it provides protection from detachment-induced apoptosis (Schulze et al., 2001). This effect was shown to be mediated by an autocrine activation of the EGF receptor, through the expression of the EGF-like growth factors HB-EGF, TGFα and Amphiregulin, leading to stimulation of the PI3-kinase and PKB/Akt activities. We show here that Bim is a direct target of Raf in this system, where sustained stimulation of the Raf/MEK/ERK pathway specifically prevents BimEL and BimL upregulation, which is induced upon detachment of the cells from the matrix. Furthermore, activation of Raf leads to phosphorylation and downregulation of the longest isoform BimEL, presumably as a parallel route to ensure inhibition of its proapoptotic activity. Removal of Bim through RNA interference (RNAi) provides some protection against detachment-induced apoptosis in this cell line, indicating that the modification induced in BimEL upon Raf activation, both at the transcriptional and post-translational level, might contribute to the acquisition of resistance against detachment-induced apoptosis.

Results

Bim expression is induced upon detachment of epithelial cells from the matrix

As previously reported, MCF-10A ΔRaf-ER cells undergo apoptosis upon culture in suspension and this effect is prevented by prestimulation with 4-OHT in order to activate the Raf signalling pathway (Schulze et al., 2001). To study whether Bim might have a role in the cell death signalling triggered by detachment of MCF10A cells from the culture dish, we followed the same experimental conditions by plating the cells on poly-HEMA-coated dishes in minimal medium and changes in Bim expression were assessed up to 30 h after detachment, when cell death is clearly evident (Schulze et al., 2001). Expression of BimEL, and to a lesser extent BimL, was strongly induced in a time-dependent manner following culture of the cells in suspension (Figure 1a), followed by cleavage of PARP, as a marker of apoptosis (Figure 1b). Importantly, prestimulation of the cells with 4-OHT for 48 h before detachment, prevented the induction of BimEL as well as PARP cleavage (Figure 1a and b). Treatment of the cells in suspension with the broad-spectrum caspase inhibitor Z-VAD-fmk did not prevent the increase in BimEL and BimL protein levels, indicating that the effect of Raf/ERK on BimEL is caspase independent (Figure 1a). Caspase inhibition did, however, protect the cells from detachment-induced apoptosis (Schulze et al., 2001) and inhibited the cleavage of PARP (Figure 1b).

Figure 1
figure1

Bim protein is induced upon culture of MCF 10A ΔRafER cells in suspension and this effect is specifically prevented upon Raf activation. (a) MCF 10A ΔRafER cells were grown on poly-HEMA-coated dishes with or without 4-OHT or the caspase inhibitor Z-VAD-fmk. Total cell lysates were analysed by Western blotting for Bim and Bmf. (b) Total cell lysates as in (a) analysed by Western blotting for PARP cleavage. (c) Analysis of Bcl-XL expression upon culture of the cells in suspension. (d) Total cell lysates as in (a) analysed by Western blotting for Bad and Bid. Equal loading was verified by blotting for Actin

MAPK activation in mammary epithelial cells has previously been shown to affect the balance of other members of the Bcl-2 family, by either an upregulation of antiapoptotic proteins, such as Bcl-XL (Schulze et al., 2001), or a post-translational inactivation of proapoptotic counterparts, such as Bad (Gilmore et al., 2002). To test whether Raf/MEK/ERK activation affects the expression of other members of the Bcl-2 family upon culture of MCF-10A ΔRaf-ER cells in suspension, expression of the BH3-only proteins Bid, Bad and Bmf and of the antiapoptotic protein Bcl-XL was analysed under the same experimental conditions. Bcl-XL protein expression was downregulated in a time-dependent manner and was almost absent 30 h after detachment-induced apoptosis (Figure 1c), in agreement with the previous data (Schulze et al., 2001). The expression of Bad was induced shortly after detachment, presumably as an effect of growth factor deprivation, and then gradually decreased throughout the course of the experiment, opposite to the observed upregulation of BimEL (Figure 1d). Similarly, Bid expression was weakly induced only at early time points and decreased thereafter. Importantly, the changes in the expression of these BH3-only proteins were not affected by activation of Raf (Figure 1d). Expression of another BH3-only member, Bmf, was also assessed, as this protein has been shown to be activated by anoikis (Puthalakath et al., 2001). Expression of Bmf, however, was only slightly induced at 30 h after detachment and was not influenced by pretreatment of the cells with 4-OHT (Figure 1a). Taken together, these data suggest that among these Bcl-2 family members, Bim is the BH3-only target downstream of the Raf/MEK/ERK signalling pathway whose downregulation and/or degradation coincides with the protection of MCF-10A ΔRaf-ER cells from the apoptotic signal initiated upon detachment of the cells from the matrix.

To study whether the increase in the amount of BimEL protein is due to a transcriptional effect induced by detachment, total RNA was isolated from cells maintained in suspension in minimal medium for up to 24 h and real-time PCR was performed using specific BimEL primers. In parallel experiments, cells were challenged with 4-OHT in the presence or absence of the specific MEK inhibitor U0126. As shown in Figure 2, the level of BimEL mRNA increased in a time-dependent manner upon detachment-induced apoptosis and this effect was clearly blocked by activation of Raf, suggesting that de novo transcription of Bim mRNA is required for the commitment of these cells to death upon detachment. Addition of the MEK inhibitor U0126 prevented the effect of Raf on BimEL transcription, which is therefore dependent on the MEK/ERK kinase activity. The increase in Bim mRNA upon detachment was unaffected by the addition of caspase inhibitors (Figure 2).

Figure 2
figure2

Raf prevents the transcriptional induction of Bim upon suspension-induced cell death. MCF 10A ΔRafER cells were grown on poly-HEMA-coated dishes with or without 4-OHT, 4-OHT and U0126 or the caspase inhibitor Z-VAD-fmk for the indicated time. Total RNA was isolated and subjected to real-time PCR using BimEL-specific primers. BimEL transcript levels were normalized for 18S RNA. Bars represent standard deviation

Activation of Raf in MCF-10a cells induces a post-translational modification in BimEL

The activation of Raf in MCF-10A ΔRaf-ER cells grown in suspension results in a mobility shift in the BimEL protein, a post-translational modification indicative of phosphorylation (Figure 1a), similar to recent data obtained in CC139 fibroblasts (Weston et al., 2003). This effect was also evident when cells adherent to the matrix were grown overnight in growth factor-poor medium (minimal medium) and then treated with 4-OHT for 24 h. Similar to the data obtained with cells grown in suspension, Raf activation induced a mobility shift specifically in the longest isoform BimEL (Figure 3).

Figure 3
figure3

Raf activation induces a mobility shift in BimEL indicative of phosphorylation. Western blot analysis of lysates from MCF-10A ΔRaf-ER cells grown (adherent) overnight in minimal medium and either left untreated (Con) or treated with 4-hydroxytamoxifen (+4-OHT) for 24 h

Analysis of the human and mouse Bim protein using the Scansite program (http://scansite.mit.edu/) predicted several conserved sites that conform to the consensus motifs for the substrate of ERK1 kinase (Figure 4). Serine 69, which is recognized with high stringency, together with serine 59 and serine 77, both recognized with low stringency, is located within the third translated exon of the human Bim gene, which is present only in the longest isoform BimEL. Two additional predicted ERK phosphorylation sites, serine 118, predicted at medium stringency, and threonine 116, predicted at low stringency, are found also in BimL. These regions are, however, absent in the truncated variants BimS and BimAD. Although BimL is also predicted to contain potential phosphorylation sites for ERK1 kinase, changes in the mobility of BimL upon Raf activation were not detected (Figure 1a and 3), in line with the finding that these sites are instead specific consensus motifs for JNK kinase (Lei and Davis, 2003).

Figure 4
figure4

Motifs in Bim that are potential substrates for ERK. The sequence of human Bim protein was analysed using Scansite predictive software (http://scansite.mit.edu/)

Recent reports have also shown that the mobility shift induced in Bim by either NGF-deprivation (Biswas and Greene, 2002) or activation of JNK kinase (Lei and Davis, 2003) is indeed due to phosphorylation.

The post-translational modification of BimEL is an early event upon ERK kinase activation and is dependent on MEK/ERK kinase activity

The activation of Raf in MCF-10A cells was shown to induce the production of autocrine EGF-like factors, such as HB-EGF, resulting in an autocrine loop through the EGF receptor and activation of further kinase activities (McCarthy et al., 1997; Schulze et al., 2001). To study whether activated Raf could affect BimEL protein mobility directly, this effect was assessed after a shorter 4-OHT treatment (as short as 10 min), when it is predicted that the autocrine loop has not yet been activated. It was in fact reported previously that the appearance of mature HB-EGF mRNA occurs within 1–2 h upon Raf activation (McCarthy et al., 1995). Phosphorylation of both p44ERK1 and p42ERK2 was clearly evident as early as 30 min after 4-OHT treatment in MCF-10A ΔRaf-ER cells (Figure 5a). ERK activation coincided with the appearance of slowly migrating bands in BimEL, hence suggesting that this isoform is a direct substrate of ERK kinase in vivo (Figure 5b). The immunoblot analysis also indicated that a fraction of endogenous BimEL remains nonphosphorylated even after a prolonged treatment with 4-OHT.

Figure 5
figure5

Phosphorylation of ERK kinase following Raf activation results in changes in BimEL migration which depend on MEK/ERK activity. (a) Western blot analysis of a time course experiment in MCF-10A ΔRaf-ER cells after 4-OHT treatment. EGF was used as a positive control for ERK activation. (b) Effect of Raf/MEK/ERK activation on Bim mobility. A specific inhibitor of ERK activation, U0126, completely prevents the mobility shift of BimEL in the presence of Raf activation. Equal loading was verified by blotting for Actin. (c) Another specific inhibitor of ERK activation, PD098059 (PD) completely prevents the mobility shift of BimEL in the presence of Raf activation (6.5 h of 4-OHT treatment)

In line with the finding that activation of ΔRaf-ER in CC139 fibroblasts results in the rapid and sustained activation of ERK1 but not JNK1 or p38α, during serum withdrawal (Weston et al., 2003), we did not observe any activation of JNK kinase upon specific Raf activation in our system, indicating that the early changes in BimEL migration are a specific and direct effect of ERK kinase activation (data not shown). Furthermore, treatment of MCF-10A ΔRaf-ER cells with 4-OHT together with either of two specific MEK1/2 inhibitors, PD098059 or U0126, used at concentrations previously shown to block MEK activity in these cells (Schulze et al., 2001), completely prevented the changes in the BimEL mobility (Figure 5b and c), showing that this activity is MEK/ERK dependent.

We next tested whether the PI3K/Akt pathway, which is activated as a result of the autocrine loop (Schulze et al., 2001), could contribute to the effect of Raf activation on Bim protein level and BimEL mobility on cells grown in suspension. For this purpose, MCF-10A ΔRaf-ER cells were grown on poly-HEMA-coated dishes in the presence of 4-OHT and either the PI3K inhibitor LY294002 or the MEK inhibitor U0126. While the latter one clearly prevented both the transcriptional effect on BimEL and BimL and BimEL mobility shift induced by Raf activation, similar to the data shown in Figure 5b, LY294002 was ineffective, indicating that Bim is not a target of PKB/Akt upon activation of the autocrine loop in this system (Figure 6). This is in contrast to the situation in haematopoietic cells where the activation of FoxO Forkhead transcription factors on PKB/Akt inhibition has been shown to induce Bim transcription (Dijkers et al., 2002).

Figure 6
figure6

Changes in Bim expression and migration do not depend upon PKB/Akt activation through the autocrine loop. MCF-10A ΔRaf-ER cells were grown on poly-HEMA-coated dishes in minimal medium for 14 h in the presence or absence of 4-OHT, U0126 and LY294002, as indicated. Whole-cell lysates were analysed by Western blotting for Bim, Phospho-p42ERK2/p44ERK1 MAPK and phospho-PKB/Akt. Equal loading was verified by blotting for Actin

Bim is degraded following Raf activation

The mobility shift in BimEL induced by activation of the Raf signalling pathway is accompanied by a clear decrease in the expression level of the protein (Figures 1a and 5b). This effect is seen early after the appearance of the slowing migrating bands, when ERK1 kinase is still fully active (Figure 5a). This raised the possibility that the post-translational modification induced by ERK could flag Bim as a target for degradation. The ubiquitin proteasome system is a likely candidate for Bim degradation as it has been shown to control the stability of several other members of the family (Breitschopf et al., 2000; Li and Dou, 2000; Marshansky et al., 2001). To test this hypothesis, MCF-10A ΔRaf-ER cells were challenged with 4-OHT in the presence of the specific proteasome inhibitor Lactacystin. As shown in Figure 7a, addition of Lactacystin prevented the decrease in Bim protein levels, hence suggesting a role of the ubiquitin proteolytic system in controlling Bim stability. Consistently, addition of proteasome inhibitors was able to stabilize the BimEL protein after addition of 4-OHT to cells grown in suspension (Figure 7b). The activation of the Raf/MEK/ERK pathway was recently shown to induce phosphorylation and proteasome degradation of BimEL in CC139 fibroblasts (Ley et al., 2003), and therefore may represent a conserved mechanism for Raf-induced survival.

Figure 7
figure7

The changes in BimEL induced by Raf activation, highlight it as a target of the ubiquitin proteolytic system. (a) Addition of the specific proteasome inhibitor Lactacystin (L) together with 4-OHT for 6.5 h, prevents the decrease in Bim expression, hence suggesting a role of the proteasome in controlling Bim stability. Equal loading of the proteins was verified by blotting for Actin. (b) Similar experiment for cells grown in suspension for the indicated time with or without 4-OHT or Lactacystin

Mapping of the phosphorylation sites

In view of the findings that the BimEL isoform is specifically modified upon Raf activation in our system and that serine 69, the only site recognized with high stringency as a consensus motif for the substrate of ERK1 kinase, is present only in the BimEL variant, we reasoned that this site might be the target of activated MEK/ERK. Site-directed mutagenesis was therefore performed by converting this potential phosphorylation site to an alanine residue. In parallel, mutations of threonine 116 and serine 118 were also performed. Several mutants were generated, in which either one, two or all three sites were mutated to alanine residues, as detailed in the Material and methods section. Each mutant was transiently transfected into 293 cells and the electrophoretic mobility shift was investigated upon treatment with EGF, a strong inducer of ERK kinase activity (Figure 8a). In cells stimulated with EGF, wild-type (wt) BimEL appeared as a triplet and no differences were found in the electrophoretic mobility of the mutants BimELT116A and BimELS118A. The slowest migrating band was instead lost in all those constructs in which serine 69 was mutated to alanine, suggesting that this site might also be important for the modification induced by Raf activation. These mutants, however, still migrated as doublets, indicating that some additional sites might be important in BimEL regulation upon EGF stimulation (as discussed later). To investigate whether the mutation of serine 69 in BimEL also influences its proapoptotic activity by, perhaps, protecting BimEL from degradation, BimEL and BimELS69A were transiently transfected into NIH3T3 cells and the β-galactosidase (β-Gal) assay was performed as previously described (Marani et al., 2002). Surprisingly, in cells transfected with the BimELS69A mutant, there was no consistent difference in cell death as compared to wt BimEL, at all the DNA doses tested (Figure 8b). It is possible that the lack of proapoptotic activity of this mutant is due to a compensatory effect of alternative ERK kinase sites present in exon 3.

Figure 8
figure8

Mutation of serine 69 in BimEL affects its migration upon epidermal growth factor (EGF) stimulation but not its proapoptotic activity. (a) Several mutants of BimEL were generated in which the potential phosphorylation sites for ERK1 kinase have been replaced by alanines. These constructs were transfected into 293 cells and the mobility shift was examined upon treatment with EGF. (b) NIH3T3 cells were cotransfected with different doses of BimEL or BimELS69A mutant along with a β-Gal expressing plasmid as a marker for transfected cells. After 14 h, cells were fixed and stained for β-Gal expression. Cell death was calculated as the percentage of round blue cells relative to the total population of blue cells. Bars indicate the standard deviation of the results of three independent experiments

Effect of Bim removal on the survival of MCF-10A ΔRaf-ER cells in suspension

As these findings provide evidence that Bim is a direct target of Raf in epithelial cells and a potential mediator of Raf-induced protection from anoikis, it is important to assess the contribution of Bim to the survival potential of this Ras effector. Since Raf activation leads to protection from anoikis in this system, a key question is whether removal of endogenous Bim influences the survival of MCF-10A ΔRaf-ER cells after detachment from the matrix. To answer this question, we took advantage of recent progress in RNA interference technology to knock down endogenous Bim RNA expression. This method of RNA silencing is based on the finding that short double-stranded RNA molecules can cause degradation of homologous mRNA in mammalian cells (Elbashir et al., 2001). Two double-stranded small interfering RNA (siRNA) oligonucleotides specific for Bim were designed as detailed in the Materials and methods section. MCF-10A ΔRaf-ER cells were then transfected with these synthetic oligonucleotides and Bim expression was assessed by immunoblot analysis 24 h after transfection (Figure 9a). In the case of one Bim siRNA, named Bim2, elimination of Bim expression was complete at this time, while the Bim3 oligonucleotide had only a partial effect. It is interesting to note that the targeting region of Bim2 lies within the first 50 nucleotides downstream of the start codon of the hBim gene, a region that most guidelines would have advised avoiding because of the regulatory protein binding sites that could interfere with binding of the siRNA. The experiment was repeated for the Bim2 oligonucleotide and extended up to 72 h after transfection, to show that the silencing effect was still present (Figure 9b). The effect of Bim2 and Bim3 siRNAs was further analysed by real-time PCR, using specific Bim primers. The decrease in the RNA level of Bim paralleled the changes in the protein levels, with approximately 80% reduction in the case of the Bim2 oligonucleotide, and only 13% reduction with the Bim3 oligonucleotide (Figure 9c).

Figure 9
figure9

RNA interference for hBim. (a) Western blot analysis of levels of endogenous Bim after RNA interference with the indicated siRNA oligonucleotides (Bim2 and Bim3) in MCF-10A ΔRaf-ER cells. The control (Con) shows the expression of Bim in cells transfected with a control oligonucleotide. Cells were lysed 24 h after the transfection. (b) Immunoblot analysis from a parallel experiment where cells were lysed 72 h after transfection. Equal loading of the proteins was verified by blotting for Actin. (c) Real-time PCR for BimEL and BimL with total RNA extracted from cells treated as in (a). Bim transcript levels were normalized for 18S

To study whether the absence of Bim could affect the survival of MCF-10A ΔRaf-ER cells upon detachment from the matrix, we performed a similar experiment to the one reported by Schulze et al. (2001). MCF-10A ΔRaf-ER cells were either transfected with the Bim2 siRNA or a control oligonucleotide (negative control) and then either left untreated or treated with 4-OHT for 48 h. Cells were then detached from the culture dishes and plated on normal (adherent) or poly-HEMA-coated dishes (suspension) in minimal medium. Apoptosis was measured 24 h later by quantifying DNA fragmentation using a Cell Death Detection ELISA kit.

Parental cells underwent apoptosis upon culture in suspension, an effect prevented by prestimulation with 4-OHT, as previously reported (Figure 10a). Loss of Bim expression in the transfected cells also resulted in a significant, albeit less pronounced, protective effect. Autocrine signalling through the EGF receptor is clearly important in the protection of these cells from anoikis at this time point. Analysis of Bim expression indicated that the silencing effect is still present upon detachment from the matrix (Figure 10b). However, approximately 20% of the protein, as compared to the control cells, is still expressed (as calculated using a Phosphor-Imager by normalizing the intensity of the immunoreactive bands to the appropriate actin intensity), and it is therefore possible that a greater protective effect could be obtained in the complete absence of Bim. This experiment supports the proposal for a role of Bim in the response of MCF-10A cells to anoikis.

Figure 10
figure10

The absence of Bim partially protects MCF-10A ΔRaf-ER cells from detachment-induced apoptosis. (a) MCF-10A ΔRaf-ER cells were either left untreated (−) or pretreated with 4-OHT for 48 h (+), detached by trypsin treatment, and plated on poly-HEMA-coated dishes (suspension) in minimal medium. After 24 h, cell death was monitored by assaying DNA fragmentation. ‘Bim’ refers to cells transfected with the Bim2 siRNA oligonucleotide to silence the Bim gene. Bars indicate the standard deviation of the results of three independent experiments. (b) Western blot analysis of Bim expression from a parallel experiment to show that the silencing effect is still present upon detachment from the matrix. Equal loading was verified by blotting for Actin

Discussion

Apoptosis induced by loss of matrix attachment, a phenomenon termed anoikis (Frisch and Francis, 1994), is a common event in untransformed adherent cells, such as epithelial (Rytomaa et al., 2000; Schulze et al., 2001), endothelial (Meredith et al., 1993) and some fibroblastoid cell lines (Le Gall et al., 2000). The dependence on matrix interaction for cell survival is important in preventing inappropriate population expansion and metastatic spread of normal epithelial cells. Several signal transduction pathways have been shown to modulate anoikis sensitivity. Survival of fibroblastoid and epithelial cells in suspension is in part mediated by the MEK/MAPK signalling (Le Gall et al., 2000; Gilmore et al., 2002).

The aim of this study was to gain further insights into the molecular basis by which activation of Raf leads to protection of epithelial cells from detachment-induced apoptosis, a very important part of the Ras transformed phenotype. The use of an inducible system to activate Raf, allows the study of the early changes induced by this Ras effector that contribute to the establishment of the transformed state. The data presented here show that the proapoptotic BH3-only protein Bim represents a direct target of Raf in the human luminal mammary epithelial cell line MCF-10A. Selective activation of Raf was shown to protect these cells against detachment-induced apoptosis (Rytomaa et al., 2000; Schulze et al., 2001), and we provide the evidence that at least part of this effect is due to inhibition of Bim expression both at the transcriptional and post-translational level.

BimEL expression was specifically induced upon suspension-induced apoptosis in a time-dependent manner, paralleled by a downregulation of the antiapoptotic protein Bcl-XL, as previously reported (Rytomaa et al., 2000; Schulze et al., 2001). Addition of 4-OHT specifically prevented the upregulation of BimEL and BimL seen shortly after detachment of the cells from the matrix, and the parallel changes in the mRNA levels of BimEL indicate that Raf signalling downregulate Bim by a transcriptional mechanism. The BH3-only proteins Bad, Bmf and Bid seem not to have a clear role downstream of Raf in this system, as addition of 4-OHT did not change their pattern of expression upon culture in suspension. At the same time, Raf activation induced a mobility shift, indicative of phosphorylation, specifically in the BimEL isoform, a modification which presumably highlights BimEL for degradation. The loss of BimEL protein is indeed inhibited by the proteasome inhibitor Lactacystin, implicating the ubiquitin proteolytic system in controlling Bim stability. The effect of Raf on BimL seems to be only at the transcriptional level, as BimL does not appear to be a target of phosphorylation in this system (Figures 1a and 3), and addition of proteasome inhibitors does not alter the decrease in the protein level upon addition of 4-OHT (Figure 7b).

Both the transcriptional and post-translational effects induced by Raf on BimEL are dependent on MEK/ERK activity, as specific inhibitors of these kinases completely prevented both the effect on Bim mRNA and the modification and the degradation of BimEL (Figures 2,5b and c). The picture looks similar to recent studies in fibroblasts, where part of the effect of the ERK pathway is to repress Bim transcription (Weston et al., 2003), and at the same time phosphorylate BimEL to destabilize the protein via the proteasome (Ley et al., 2003).

Mutagenesis studies of the potential phosphorylation sites for ERK1 kinase within the Bim protein showed that mutation of serine 69 to alanine prevents the appearance of the slowest migrating band in BimEL upon EGF treatment, suggesting that this site is a potential target for Bim phosphorylation. A second phosphorylated band, however, was retained, indicating that other sites may influence the mobility shift of BimEL in this system. Furthermore, the cell-death activity of the BimELS69A mutant is similar to that of wt BimEL, presumably due to a compensatory effect of other MAPK-sites. Threonine 116 and serine 118 have been recently shown to be important targets of JNK (Lei and Davis, 2003) and we found that mutagenesis of these sites does not prevent the mobility shift on BimEL induced by Raf/ERK activation in our system. In the context of specific Raf activation, the presence of additional phosphorylation sites in the NH2-terminal insertion of BimEL (exon 3), which is absent in BimL, would explain why we could not detect any changes in the mobility of BimL upon Raf activation. Two additional serine–proline motifs are indeed found in this region and it would be informative to address whether mutation of these serine residues together with serine 69 is able to inhibit the appearance of all the multiple phosphorylated forms of BimEL upon EGF or 4-OHT treatment, and whether this would affect the apoptotic potential of BimEL.

The use of RNA interference technology allowed us to study the contribution of Bim to anoikis sensitivity upon Bim removal. These data provide a first indication of a role of Bim in the response of MCF-10A cells to anoikis, and support a novel function of Bim as a direct target of Raf in epithelial cells, a system where the K-ras proto-oncogene is most commonly mutated (Bos, 1989). Very recently it has been reported that integrin engagement and EGF receptor activation act together to suppress Bim expression in MCF-10A cells (Reginato et al., 2003). Although the lack of complete Bim removal in the experiments performed here might explain why these cells are only partly protected from detachment-induced apoptosis, we cannot exclude the possibility that parallel pathways to cell death via detachment might exist. In support of this, the stronger protection provided by Raf activation has been shown to be mediated by an autocrine activation of the EGF receptor, through the expression of the EGF-like growth factors HB-EGF, TGFα and Amphiregulin, leading to the stimulation of the PI3-kinase and PKB/Akt activities (Schulze et al., 2001). Although Bad, Bid and Bmf seem not to be alternative targets downstream of Raf in this mammary epithelial cell line, other BH3-only proteins might play a role in this system.

In view of the finding that BimEL and BimL proapoptotic activity is controlled by sequestration to motor complexes, a role of these proteins as mediators of those apoptotic signals which impinge on the cytoskeleton and motor complexes, such as anoikis, has already been speculated (Puthalakath et al., 1999). The finding that a proapoptotic gene is directly regulated by the Ras-Raf signalling cascade may prove important in order to understand the process of oncogenic transformation.

Materials and methods

Cell culture and drug treatment

MCF-10A ΔRaf-ER mammary epithelial cells were kindly donated by Almut Schulze (Schulze et al., 2001). This cell line was cultured in Ham's nutrient mixture F12/Dulbecco's Modified Eagles Medium (DMEM) (1 : 1) containing 5% horse serum (GIBCO-Life Technology) and 10 μg/ml insulin (Sigma), 20 ng/ml EGF (Sigma), 5 μg/ml hydrocortisone (Sigma) and 100 ng/ml cholera toxin (Sigma) (full medium). Minimal medium consisted of Ham's nutrient mixture F12/DMEM (1 : 1) containing 5% horse serum. NIH3T3 cells were grown in DMEM medium supplemented with 5% donor calf serum, 50 μg/ml streptomycin and 50 IU/ml penicillin. Human embryonal kidney epithelial 293 cells were cultured in DMEM supplemented with 10% foetal bovine serum (FBS), 50 μg/ml streptomycin and 50 IU/ml penicillin. Cultures were incubated at 37°C in a humidified atmosphere of 10% CO2.

The drugs used in this study were as follows: 4-hydroxytamoxifen (4-OHT) (Sigma) was dissolved in ethanol and used at 100 nM; PD98059 (Calbiochem) was dissolved in DMSO and used at 30 μ M; U0126 (Promega) and Lactacystin (Calbiochem) were dissolved in DMSO and used at 10 μ M; LY294002 (Calbiochem) was dissolved in DMSO and used at 20 μ M. The broad-spectrum caspase inhibitor Z-VAD-fmk (Calbiochem) was dissolved in DMSO and used at 50 μ M.

Site-directed mutagenesis of Bim

Site-directed mutagenesis of Bim was performed using the QuickChange™ Site-Directed Mutagenesis kit (Stratagene) according to the manufacturer's recommended procedure. The BimEL mutants were generated with the following paired primers: S69A 5′-IndexTermGCC CCA CCT GCC GCC CCT GGC CCT TTT GC-3′ and 5′-IndexTermGCA AAA GGG CCA GGG GCG GCA GGT GGG GC-3′; S116A 5′-IndexTermGAC AAA TCA ACA CAA GCC CCA AGT CCT CCT TGC-3′ and 5′-IndexTermGCA AGG AGG ACT TGG GGC TTG TGT TGA TTT GTC-3′; S118A 5′-IndexTermCAACACAAACCCCAGCTCCTCCTTGCCAGGCC-3′ and 5′-IndexTermGGC CTG GCA AGG AGG AGC TGG GGT TTG TGT TG-3′. Mutations were confirmed by DNA sequencing.

Cell death assay

MCF-10A ΔRaf-ER cells were challenged by detachment-induced apoptosis by plating them at a density of 105 cells/ml on poly-HEMA-coated dishes in minimal medium. In parallel experiments, cells were pretreated with 4-OHT for 48 h or alternatively subjected to RNA interference (as described below) before culture in suspension. After 24 h, cells were harvested and DNA fragmentation was quantified using a Cell Death Detection ELISA kit (Roche) according to the manufacturer's recommended procedures.

NIH3T3 cells were transiently cotransfected with different doses of BimEL or BimELS69A mutant (0.5, 0.75 and 1 μg) along with 0.5 μg of pcDNA3.1/V5-His-TOPO/lacZ (Invitrogen) containing the gene for β-Gal using Superfect™. After 14 h, cells were fixed and stained for β-Gal expression, as previously described (Marani et al., 2002). The percentage of apoptotic cells was quantified by microscopically counting the number of apoptotic (round) blue cells over the total number of blue cells counted (500–800 cells) from randomly chosen fields.

Antibodies and immunoblotting

Antibodies used in this study were as follows: a rabbit polyclonal antibody against amino acids 22–40 of human Bim protein (Anti-Bim, N-Terminal; Calbiochem); a goat polyclonal antibody against Actin (Actin (C-11): sc-1615; Santa Cruz Biotechnology); a mouse monoclonal antibody against phospho-p42/p44 ERK1/2 (T202, Y204) (New England Biolabs); a goat polyclonal antibody against Bid (R&D systems); a mouse monoclonal antibody against Bad (B36420; Transduction Laboratories); a mouse monoclonal antibody against Bmf (Apotech); a mouse monoclonal antibody against PARP was kindly donated by Miguel Martins (University of Leicester). A rabbit polyclonal antibody against phospho-S473-PKB/Akt has been described previously (Khwaja et al., 1997).

Cells were lysed in RIPA buffer or in SDS sample buffer and proteins were separated by SDS–PAGE and transferred onto PVDF membrane. After incubation with the relevant antibodies, the antigen–antibody complex was visualized using the ECLTM system (Amersham), according to the manufacturer's instructions. To ensure equal loading and transfer, membranes were probed for Actin.

RNA interference

MCF-10A ΔRaf-ER cells were plated onto six-well plates at 105 cells/well in complete medium without antibiotics. After 24 h, cells were transfected with double-stranded siRNA at a final concentration of 60 nM using EffecteneTM (Qiagen), according to the manufacturer's instructions. Silencing was assayed by resolving the proteins on SDS–PAGE and immunoblotting with the polyclonal anti-Bim antibody.

The optimal target sequence was designed by using the Extractor-2 program and the selected oligonucleotides were obtained from Xeragon (Zurich) and had the following sequences:

  • Bim (2) sense, 5′-r(IndexTermGCAACCUUCUGAUGUAAGU)d(TT)-3′

  • Bim (2) antisense 5′-r(IndexTermACUUACAUCAGAAGGUUGC)d(TT)-3′

  • Bim (3) sense, 5′-r(IndexTermGACCACCCACGAAUGGUUA)d(TT)-3′

  • Bim (3) antisense 5′-r(IndexTermUAACCAUUCGUGGGUGGUC)d(TT)-3′

Annealed double-stranded siRNA oligonucleotides were obtained as lyophilized powder and resuspended in 1 ml of the provided sterile buffer (100 mM potassium acetate, 30 mM HEPES–KOH, 2 mM magnesium acetate, pH 7.4) to obtain a 20 μ M solution. The suspension was heated to 90°C for 1 min and then incubated at 37°C for 1 h, in order to disrupt higher aggregates that may have formed in the lyophilization process. The clear solution was stored at −20°C.

Real-time PCR

Primers specific for the hBimEL isoform or for both the hBimEL and the hBimL isoforms (exon 4/5 boundary) were kindly donated by Justin Cross (Cancer Research UK) and were as follows:

  • BimEL forward 5′-IndexTermGCCCCACCTGCCAGC-3′

  • BimEL reverse 5′-IndexTermACAGCAGGGAGGATCTTCTCATAA-3′

  • Bim EL+L forward 5′-IndexTermTCCTCCTTGCCAGGCCTT-3′

  • Bim EL+L reverse 5′-IndexTermCTGCAGGTTCAGCCTGCC-3′

Total RNA was extracted using Trizol reagent (Invitrogen), followed by further purification using Rneasy Mini Kit (Qiagen). The quality of the extracted RNA was determined by agarose gel electrophoresis, using standard techniques. Reverse transcription was performed using Taqman reverse transcription reagents (ABI). Real-time PCR experiments were performed using the SYBR green method (Molecular Bioproducts). All samples were read in triplicate and relative quantification of gene expression was obtained using 18S RNA as an internal standard. Specific primers for 18S were as follows: forward primer (5′-IndexTermCGCCGCTAGAGGTGAAATTC-3′); reverse primer (5′-IndexTermCATTCTTGGCAAATGCTTTCG-3′).

References

  1. Biswas SC and Greene LA . (2002). J. Biol. Chem., 277, 49511–49516.

  2. Bonni A, Brunet A, West AE, Datta SR, Takasu MA and Greenberg ME . (1999). Science, 286, 1358–1362.

  3. Borner C . (2003). Mol. Immunol., 39, 615–647.

  4. Bos JL . (1989). Cancer Res., 49, 4682–4689.

  5. Bouillet P, Metcalf D, Huang DC, Tarlinton DM, Kay TW, Kontgen F, Adams JM and Strasser A . (1999). Science, 286, 1735–1738.

  6. Breitschopf K, Zeiher AM and Dimmeler S . (2000). J. Biol. Chem., 275, 21648–21652.

  7. Dijkers PF, Birkenkamp KU, Lam EW, Thomas NS, Lammers JW, Koenderman L and Coffer PJ . (2002). J. Cell. Biol., 156, 531–542.

  8. Downward J . (2003). Nat. Rev. Cancer, 3, 11–22.

  9. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K and Tuschl T . (2001). Nature, 411, 494–498.

  10. Frisch SM and Francis H . (1994). J. Cell. Biol., 124, 619–626.

  11. Gilmore AP, Valentijn AJ, Wang P, Ranger AM, Bundred N, O'Hare MJ, Wakeling A, Korsmeyer SJ and Streuli CH . (2002). J. Biol. Chem., 277, 27643–27650.

  12. Hsu SY, Lin P and Hsueh AJ . (1998). Mol. Endocrinol., 12, 1432–1440.

  13. Jost M, Huggett TM, Kari C, Boise LH and Rodeck U . (2001). J. Biol. Chem., 276, 6320–6326.

  14. Khwaja A, Rodriguez-Viciana P, Wennstrom S, Warne PH and Downward J . (1997). EMBO J., 16, 2783–2793.

  15. Le Gall M, Chambard JC, Breittmayer JP, Grall D, Pouyssegur J and Van Obberghen-Schilling E . (2000). Mol. Biol. Cell, 11, 1103–1112.

  16. Lei K and Davis RJ . (2003). Proc. Natl. Acad. Sci. USA, 18, 18.

  17. Leu CM, Chang C and Hu C . (2000). Oncogene, 19, 1665–1675.

  18. Ley R, Balmanno K, Hadfield K, Weston C and Cook SJ . (2003). J. Biol. Chem., 278, 18811–18816.

  19. Li B and Dou QP . (2000). Proc. Natl. Acad. Sci. USA, 97, 3850–3855.

  20. Liu YZ, Boxer LM and Latchman DS . (1999). Nucleic Acids Res., 27, 2086–2090.

  21. Marani M, Tenev T, Hancock D, Downward J and Lemoine NR . (2002). Mol. Cell. Biol., 22, 3577–3589.

  22. Marshansky V, Wang X, Bertrand R, Luo H, Duguid W, Chinnadurai G, Kanaan N, Vu MD and Wu J . (2001). J. Immunol., 166, 3130–3142.

  23. McCarthy SA, Chen D, Yang BS, Garcia Ramirez JJ, Cherwinski H, Chen XR, Klagsbrun M, Hauser CA, Ostrowski MC and McMahon M . (1997). Mol. Cell. Biol., 17, 2401–2412.

  24. McCarthy SA, Samuels ML, Pritchard CA, Abraham JA and McMahon M . (1995). Genes Dev., 9, 1953–1964.

  25. McFall A, Ulku A, Lambert QT, Kusa A, Rogers-Graham K and Der CJ . (2001). Mol. Cell. Biol., 21, 5488–5499.

  26. Meredith Jr. JE, Fazeli B and Schwartz MA . (1993). Mol. Biol. Cell, 4, 953–961.

  27. O’Connor L, Strasser A, O'Reilly LA, Hausmann G, Adams JM, Cory S and Huang DC . (1998). EMBO J., 17, 384–395.

  28. Puthalakath H, Huang DC, O'Reilly LA, King SM and Strasser A . (1999). Mol. Cell, 3, 287–296.

  29. Puthalakath H, Villunger A, O'Reilly LA, Beaumont JG, Coultas L, Cheney RE, Huang DC and Strasser A . (2001). Science, 293, 1829–1832.

  30. Reginato MJ, Mills KR, Paulus JK, Lynch DK, Sgroi DC, Debnath J, Muthuswamy SK and Brugge JS . (2003). Nat. Cell Biol., 5, 733–740.

  31. Rytomaa M, Lehmann K and Downward J . (2000). Oncogene, 19, 4461–4468.

  32. Scheid MP, Schubert KM and Duronio V . (1999). J. Biol. Chem., 274, 31108–31113.

  33. Schulze A, Lehmann K, Jefferies HB, McMahon M and Downward J . (2001). Genes Dev., 15, 981–994.

  34. Streuli CH and Gilmore AP . (1999). J. Mammary Gland Biol. Neoplasia, 4, 183–191.

  35. U M, Miyashita T, Shikama Y, Tadokoro K and Yamada M. (2001). FEBS Lett., 509, 135–141.

  36. Weston CR, Balmanno K, Chalmers C, Hadfield K, Molton SA, Ley R, Wagner EF and Cook SJ . (2003). Oncogene, 22, 1281–1293.

Download references

Acknowledgements

This work was supported by programme grants from Cancer Research UK. Michela Marani and Nick Lemoine thank the Mike Stone Cancer Research Fund and the Special Trustees of Hammersmith Hospitals Trust for support; Rita Lopes is supported by the Fundaçâo para Ciência e Tecnologia of Portugal.

Author information

Correspondence to Nicholas R Lemoine.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Marani, M., Hancock, D., Lopes, R. et al. Role of Bim in the survival pathway induced by Raf in epithelial cells. Oncogene 23, 2431–2441 (2004) doi:10.1038/sj.onc.1207364

Download citation

Keywords

  • Bim
  • anoikis
  • phosphorylation
  • EGF receptor

Further reading