Involvement of EGF receptor and c-Src in the survival signals induced by TGF-β1 in hepatocytes

Abstract

Transforming growth factor beta1 (TGF-β1) belongs to a family of polypeptide factors, whose cytostatic and apoptotic functions help restrain the growth of mammalian cells. Although solid data established the role of TGF-β's as suppressor factors in tumorigenic processes, in the context of an advanced stage of disease, TGF-β's could also play a pro-oncogenic role. We have previously shown that TGF-β1 induces both pro- and antiapoptotic signals in foetal rat hepatocytes. In this work, we have focused on its antiapoptotic mechanism. We show that TGF-β1 activates the epidermal growth factor receptor (EGFR) and phosphorylates c-Src. EGFR is required for Akt activation. Blocking EGFR signalling amplifies the apoptotic response to TGF-β1. TGF-β1 induced a rapid activation of the tumour necrosis factor-α-converting enzyme (TACE/ADAM (a disintegrin and metalloprotease) 17). Inhibitors of TACE considerably attenuated Akt activation, which suggests that TGF-β1 activates EGF signalling in hepatocytes by promoting shedding of EGF-like ligands. The activation of c-Src by TGF-β1 is EGFR dependent and is required for full Akt phosphorylation and cell survival. Inhibition of EGFR does not block the epithelial–mesenchymal transition (EMT) induced by TGF-β1 in hepatocytes, which indicates that activation of EGFR plays an essential role in impairing apoptosis, but it is dispensable for the EMT process.

Introduction

The transforming growth factor beta1 (TGF-β1) belongs to a family of structurally related polypeptide factors, whose cytostatic and apoptotic functions help restrain the growth of mammalian tissues (Siegel and Massagué, 2003). Although there is no doubt about the role of TGF-β's as suppressor factors in tumorigenic processes (Akhurst and Derynck, 2001; Roberts and Wakefield, 2003), it is now appreciated that metastasis of many different types of tumour cells requires TGF-β activity, and that, in the context of an advanced stage of disease, TGF-β's could play a pro-oncogenic role (Kang et al., 2003). During multistage tumorigenesis, TGF-β growth-inhibitory and apoptotic effects are lost, frequently by subversion of the normal signalling pathways due to the activation of other signalling molecules such as phosphatidylinositol-3-kinase (PI-3K) and/or Ras. Meanwhile, other TGF-β responses prevail, favouring tumour progression (Akhurst and Derynck, 2001; Akhurst 2002; Oft et al., 2002; Roberts and Wakefield, 2003).

We have previously described that TGF-β1 inhibits growth and induces apoptosis in foetal rat hepatocytes (Sánchez et al., 1996). However, a subpopulation of these cells survives concomitant with changes in morphology and phenotype, reminiscent of an epithelial–mesenchymal transition (EMT) (Sánchez et al., 1999; Valdés et al., 2002). Analyses of the intracellular signals that could impair the apoptotic effects of TGF-β1 indicated that the PI-3K/Akt pathway was activated in these resistant cells (Valdés et al., 2004). Experiments in foetal rat hepatocytes showed that TGF-β1 was able to activate transiently PI-3K/Akt, by a mechanism independent of protein synthesis, but dependent on a tyrosine kinase activity. Proapoptotic signals, such as oxidative stress and caspases, contributed to the loss of Akt at later times. Inhibiting PI-3K sensitized foetal hepatocytes to the apoptosis induced by TGF-β1 and caused spontaneous death in the resistant cells (Valdés et al., 2004). Taken together, these results indicated that TGF-β1 could be inducing both proapoptotic and survival signals in hepatocytes, the balance among them decides cell fate.

Previous studies have indicated that TGF-β1 can activate PI-3K in mammary epithelial cell lines, which would play a role in cell survival (Shin et al., 2001; Janda et al., 2002) and might be required for TGF-β1-mediated EMT and cell migration (Bakin et al., 2000; Gotzman et al., 2002). Interestingly, recent evidences indicate that the sensitivity to TGF-β1-induced apoptosis is regulated by crosstalk between the Akt/PKB and Smad through a mechanism that involves a direct interaction between Akt and Smad3 (Conery et al., 2004; Remy et al., 2004). Furthermore, Smad proteins activated by TGF-β1 form a complex with FoxO proteins to turn on the growth-inhibitory gene p21Cip1, this process being negatively controlled by the PI-3K pathway (Seoane et al., 2004). Thus, activation of PI-3K could confer resistance to TGF-β-mediated cytostasis and apoptosis.

In spite of the recent advances on the crosstalks between PI-3K/Akt and Smad signalling, very little is known about the mechanisms mediating Akt activation by TGF-β1. In this work, we have tried to go deep into this question, by using the foetal rat hepatocyte model. Since our previous results indicated that a tyrosine kinase activity was required for TGF-β1-dependent Akt activation (Valdés et al., 2004), we focused on searching for possible/s upstream kinase/s. Results have indicated that TGF-β1 activates epidermal growth factor receptor (EGFR) and c-Src family and both contribute to Akt phosphorylation and cell survival. The relevance of these results, in the context of TGF-β1 protumorigenic actions, is discussed.

Results

Activation of EGFR, c-Src family and PI-3K/Akt in foetal rat hepatocytes treated with TGF-β1

Previous results had indicated that TGF-β1-induced PI-3K/Akt was inhibited by genistein, a pan inhibitor of tyrosine kinases (Valdés et al., 2004). Thus, we started analysing possible proteins that were immunoprecipitated with a pan-tyrosine kinase antibody in protein extracts from TGF-β1-treated cells. Preliminary results revealed several bands in the 50–55 kDa range (results not shown), which lead us to study whether TGF-β1 could activate the c-Src family of tyrosine kinases. Furthermore, a band around 160–170 kDa was repeatedly present. Since EGFR is a 170 kDa protein and we had previously shown that EGF activates PI-3K/Akt in hepatocytes (Fabregat et al., 2000), we decided to study whether TGF-β1 could activate the EGF signalling in hepatocytes. Results shown in Figure 1a indicate that 30 min after the addition of TGF-β1 to the hepatocytes, EGFR and c-Src family were activated, as it was observed in Western blot experiments with antiphospho-EGFR (Tyr 1068) and antiphospho-Src (Tyr 416) antibodies. Phosphorylation of EGFR and c-Src was coincident with PI-3K activation (Figure 1b), which was followed by Akt phosphorylation (Figure 1c).

Figure 1
figure1

Activation of EGF receptor, Src family kinases and PI-3K/Akt in foetal rat hepatocytes treated with TGF-β1. Cells were incubated with TGF-β1 at different times, as indicated. (a and c) Western blot analysis. (b) PI-3K assay. Autoradiography of the radioactive PIP (phosphatidylinositol-3-phosphate). In all cases, a representative experiment of three is shown

Activation of EGFR by TGF-β1 mediates its antiapoptotic signals: involvement of tumour necrosis factor-α (TNF-α)-converting enzyme (TACE/ADAM (a disintegrin and metalloprotease) 17)

To analyse the role of EGFR in the antiapoptotic signals induced by TGF-β1, we decided to use the EGFR inhibitor AG1478. The mitogenic activity of EGF on hepatocytes was inhibited by AG1478, in a dose-dependent manner. AG1478 (10 μ M), dose at which maximal inhibitory effect was observed (Figure 2a), completely blocked the activation of Akt by TGF-β1 in hepatocytes (Figure 2b). Since recent evidences indicated that Akt/PKB interacts with Smad3, blocking its translocation to the nucleus and impairing its apoptotic activity (Conery et al., 2004; Remy et al., 2004), we analysed the effect of AG1478 on Smad3 nuclear levels (Figure 2c). In the absence of AG1478, Smad3 was initially translocated to the nucleus; however, nuclear levels clearly decreased when Akt was activated (3 h). When EGFR was inhibited, nuclear Smad3 levels significantly increased at all the times studied. The effect was specific for Smad3. Furthermore, AG1478 considerably increased the apoptosis induced by TGF-β1 in hepatocytes, analysed as the number of viable cells (Figure 3a), caspase-3 activity (Figure 3b) and percentage of hypodiploid (apoptotic) cells (Figure 3c). These results suggested that EGFR is required for Akt activation and survival induced by TGF-β1.

Figure 2
figure2

EGF receptor mediates Akt phosphorylation induced by TGF-β1 in hepatocytes. (a) Foetal hepatocytes were incubated for 48 h without (control: C) or with EGF, as indicated, and with different concentrations of AG1478. The number of viable cells was analysed and results, expressed as the percentage of untreated cells, are the mean±s.e.m. of three independent experiments, in duplicate. (b) Effect of AG1478 (10 μ M) on Akt activation by TGF-β1. Western blot analysis. A representative experiment of three is shown. (c) Effect of AG1478 (10 μ M) on nuclear translocation of Smad2 and Smad3. Western blot analysis of nuclear extracts. For Smad3, an antitotal-Smad3 antibody was used. For Smad2, due to the availability of a useful antiphospho-Smad, the phosphorylated form was analysed. Equal protein loading was visualized by Ponceau red staining of the membranes. A representative experiment of two is shown

Figure 3
figure3

EGF receptor is required for Akt activation by TGF-β1 in hepatocytes. (a) Cells were treated for 24 h with or without TGF-β1 and with or without different concentrations of AG1478. The number of viable cells was analysed and results, expressed as the percentage of untreated cells, are the mean±s.e.m. of three independent experiments, in duplicate. (b) Cells, incubated for 8 h without (c) or with TGF-β1 and with or without AG1478 (10 μ M), were lysed and caspase-3 activity analysed. Results are the mean±s.e.m. of three independent experiments, in duplicate. (c) Foetal hepatocytes were treated as described in (b). DNA content was analysed by flow cytometry. The percentage of hypodiploid (apoptotic) cells is shown. Data are the mean±s.e.m. of four independent experiments. In (b and c), values from TGF-β1 plus AG-treated cells were compared with those from TGF-β1-only treated cells by a Student's t-test: *P<0.01; **P<0.005

We next decided to study whether the EGFR was activated by binding to any of its ligands, which could be expressed and secreted by the hepatocytes. Firstly, we collected conditioned media from cells either untreated or treated with TGF-β1 during 48 h. Then, we used these conditioned media in apoptosis experiments in hepatocytes. As shown in Figure 4a, conditioned medium from TGF-β1-treated hepatocytes (CM-FH+TGF-β1), but not from untreated hepatocytes (CM-FH), was able to block completely cell death induced by TGF-β1. These results indicated that TGF-β1-treated hepatocytes were secreting a survival factor.

Figure 4
figure4

Role of EGF family ligands in the antiapoptotic signal induced by TGF-β1 in hepatocytes. (a) Conditioned media collected from foetal hepatocytes incubated for 48 h either in the absence (CM-FH) or in the presence of TGF-β1 (CM-FH+TGF-β1) were used to incubate new untreated cells for 24 h with new TGF-β1. Results show the percentage of viable cells and are the mean±s.e.m. of three independent experiments, in duplicate. (b) Effect of 20 μ M TAPI-1 on Akt and Stat-3 phosphorylation by TGF-β1. Western blot analysis. A representative experiment of three is shown. (c) Cells, incubated with TGF-β1 at different times, were lysed and TACE activity was analysed. A representative experiment of four is shown

We have previously reported that Akt activation by TGF-β1 was not prohibited by cycloheximide (Valdés et al., 2004). Neither early EGFR activation nor c-Src phosphorylation was blocked by cycloheximide (results not shown), suggesting that early antiapoptotic response induced by TGF-β1 does not require de novo synthesis of proteins. Soluble EGF family growth factors are all derived from proteolytic cleavage of the ectodomains of integral membrane precursors (Jorissen et al., 2003). The identity of the processing enzymes has remained rather obscure until identification of a novel protease, TACE/ADAM 17, whose role in mediating the release of different EGF family members has been well established (Sunnarborg et al., 2002; Borrell-Pages et al., 2003). Taken together, these results lead us to hypothesize that TGF-β1 could be activating TACE. To test this possibility, we analysed the effect of two TACE inhibitors: TAPI-1 and TAPI-2. As shown in Figure 4b, 20 μ M TAPI-1 clearly attenuated Akt activation. Phosphorylation of Stat-3, a known target of EGFR, was also analysed. In a similar way, 20 μ M TAPI-1 attenuated Stat-3 phosphorylation induced by TGF-β1. Similar results were obtained when TAPI-2 was used (results not shown). Finally, in vitro TACE-protease assays revealed a rapid (15–30 min) and transient activation of the proteolytic activity of a TACE-specific substrate, after incubation with TGF-β1 (Figure 4c).

These results suggest that TGF-β1 could be activating EGF signalling in hepatocytes by activating the shedding of EGF-like ligands.

EGFR mediates Src activation by TGF-β1, which is required for Akt phosphorylation and cell survival

Since we observed phosphorylation of the c-Src family of protein tyrosine kinases in TGF-β1-treated hepatocytes (Figure 1a), we next wondered whether c-Src activation was a consequence of EGF signalling or it was an independent and parallel event. We first tested the effect of PP2, a well-known inhibitor of c-Src family, in the mitogenic effect of EGF in hepatocytes. Results are shown in Figure 5a. The mitogenic effect of EGF was completely blocked by PP2, in a dose-dependent manner. In fact, EGF activated c-Src in hepatocytes and Src activation cooperated in Akt phosphorylation, as demonstrated by a significant decrease in Akt activation under PP2 treatment (Figure 5b). These results suggest that c-Src is downstream EGFR in hepatocytes and is required for its mitogenic and antiapoptotic signalling.

Figure 5
figure5

EGF phosphorylates Src family kinases in hepatocytes, which is required for its mitogenic and antiapoptotic signalling. (a) Foetal hepatocytes were incubated for 48 h without (control: C) or with EGF, as indicated, and with different concentrations of PP2. The number of viable cells was measured and results, expressed as the percentage of untreated cells, are the mean±s.e.m. of three independent experiments, in duplicate. (b) Cells were incubated for 15 min in the absence or presence of EGF and/or PP2 (10 μ M). Western blot analysis. A representative experiment of three is shown

Taking into account these results, we decided to explore the possibility that EGFR could mediate c-Src activation by TGF-β1 in hepatocytes. In the presence of AG1478, c-Src phosphorylation by TGF-β1 was considerably attenuated (Figure 6a), indicating that EGFR activation is required for c-Src phosphorylation. Furthermore, PP2 attenuated Akt phosphorylation by TGF-β1 (Figure 6b), in a pattern similar to that found when EGF was the stimulatory signal (Figure 5b). Finally, we observed that inhibition of c-Src activity resulted in a significant increase in TGF-β1-induced toxicity (Figure 6c).

Figure 6
figure6

EGF receptor mediates Src family kinases activation by TGF-β1, which is required for full Akt phosphorylation and cell survival. (a) Effect of AG1478 (10 μ M) on phosphorylation of Src family kinases (p-Src) by TGF-β1. Western blot analysis. (b) Effect of PP2 (10 μ M) on Akt activation by TGF-β1. Western blot analysis. In both (a and b), a representative experiment of three is shown. (c) Hepatocytes were incubated for 24 h with or without TGF-β1 in the presence of different concentrations of PP2. The number of viable cells was analysed and results, expressed as the percentage of untreated cells, are the mean±s.e.m. of three independent experiments, in duplicate

These results suggest that TGF-β1 is activating c-Src through an EGFR-dependent mechanism. Src activation is required for full Akt phosphorylation and cell survival.

EGFR activation is not required for the EMT process induced by TGF-β1 in hepatocytes

The subpopulation of foetal hepatocytes that survives to TGF-β1-induced apoptosis show changes in morphology and phenotype, replacement of cytokeratin filaments by vimentin, upregulation of Snail and downregulation of E-cadherin, reminiscent of an EMT (Valdés et al., 2002 and Figure 7 in this work). After the clear evidence of the involvement of EGFR in the survival signals induced by TGF-β1 in hepatocytes, we wondered whether or not EGFR would also be required for the EMT process. Even though 70–80% of cells died within 24 h after TGF-β1 treatment when AG1478 was present (Figure 3a), the remaining cells had undergone an EMT process, as evidenced by the presence of vimentin, instead of cytokeratin 18, and appearance of smooth-muscle actin α. The expression level of these proteins was even higher when EGFR was inhibited (Figure 7a). This is probably due to cell selection, caused by amplification and acceleration of the cell death process. Snail, a transcription factor that controls EMT by repressing E-cadherin (Batlle et al., 2000; Cano et al., 2000) and is transiently (3–5 h) upregulated by TGF-β1 in hepatocytes (Valdés et al., 2002), was identically upregulated when EGFR activation was blocked (Figure 7b).

Figure 7
figure7

EGF receptor activation is not required for the EMT process induced by TGF-β1 in hepatocytes. (a) Foetal hepatocytes were incubated in the absence or presence of TGF-β1 and/or AG1478 (10 μ M) for 24 h. Cytoskeleton protein extracts were collected to analyse the levels of smooth-muscle actin (SMA), vimentin (Vim) and cytokeratin 18 (CK-18) by Western blot. (b) Levels of Snail transcripts were analysed by RT–PCR after 3 h of treatment without (C) or with TGF-β1 (T) in the absence or presence of AG1478 (10 μ M). GAPDH expression is a control of the cDNA template in each sample. The −RT lane shows the results of amplification in the absence of reverse transcriptase

These results suggest that, although activation of EGFR by TGF-β1 in foetal rat hepatocytes appears to play an essential role in impairing apoptosis, it is not involved in the EMT process.

Discussion

TGF-β1 inhibits growth and induces apoptosis in foetal rat hepatocytes (Sánchez et al., 1996; Herrera et al., 2001). However, a subpopulation of these cells (60–70%) survives to its cytotoxic effects (Sánchez et al., 1999) and responds to this factor inducing an EMT process, which increases cell migration and invasiveness capability (Valdés et al., 2002). Understanding the intracellular signals that allow cells to resist to the apoptotic effects of TGF-β1 might contribute to exploit effectively the TGF-β system in new therapeutic approaches to cancer.

Results presented herein indicate that abrogation of TGF-β1 proapoptotic signalling could be mediated by TGF-β1-induced activation of the EGFR pathway. Several lines of evidence support this hypothesis. Firstly, TGF-β1, by itself, is able to activate transiently EGFR, which is followed by the activation of the PI-3K/p-Akt pathway in foetal rat hepatocytes (Figure 1). Secondly, inhibiting EGFR with AG1478 completely blocks Akt activation (Figure 2), increasing TGF-β-induced nuclear translocation of Smad3 and sensitizing hepatocytes to the apoptosis induced by TGF-β1, provoking the death of 70–80% of the cells in 24 h (Figure 3). Previous reports had indicated that TGF-β1 activates TGF-α expression in endothelial cells (Viñals and Pouysségur, 2001). However, the lack of effect of cycloheximide on EGFR activation and Akt phosphorylation by TGF-β1 in hepatocytes (Valdés et al., 2004 and results not shown in this work) excludes the possibility that a protein needs to be synthesized. In fact, our results would indicate that TGF-β1 induces a rapid secretion of a member of the EGF family of growth factors by activation of the metalloprotease TACE/ADAM 17 (Figure 4). Although TACE was initially identified as the metalloprotease responsible for shedding of the pro-TNF-α (Black et al., 1997; Moss et al., 1997), TACE is also necessary for shedding of the EGF family of growth factors (Merlos-Suárez et al., 2001; Sunnarborg et al., 2002; Borrell-Pages et al., 2003). Interestingly, it has been recently reported that the activity of TACE is required for the activation of the EGFR by TNF-α in hepatocytes, contributing to its mitogenic ability (Argast et al., 2004), which indicates that different cytokines could activate TACE in hepatocytes. Interestingly, development of tumours in nude mice requires TACE activity, indicating a crucial role of TACE in tumorigenesis (Borrell-Pages et al., 2003). In this same line of evidence, cannabinoids induce cancer cell proliferation via TACE-mediated transactivation of the EGFR (Hart et al., 2004).

Results presented here also suggest that TGF-β1 activates c-Src family of tyrosine kinases in foetal rat hepatocytes, with a similar timing to that one observed for EGFR phosphorylation (Figure 1). TGF-β1 can stimulate Src kinase activity in epithelial cells (Kim and Joo, 2002), and it has been recently suggested that activated c-Src might contribute to resistance against the apoptotic and/or antiproliferative properties of TGF-β1 in hepatoma cells (Park et al., 2004). We describe here that activation of c-Src might be mediated by activation of the EGFR (Figures 5 and 6), Src activation being required for full Akt activation and survival (Figure 6). Src family tyrosine kinases are activated by tyrosine kinase receptors and, in turn, can promote signalling from growth factor receptors in a number of ways including DNA synthesis, control of receptor turnover, actin cytoskeleton rearrangements, motility and survival (Bromann et al., 2004). There are many reports indicating that c-Src protein expression and/or activity is elevated in epithelial cancers (Biscardi et al., 1999) and activation of c-Src in hepatocellular carcinoma cells is highly correlated with the indices of early-stage phenotype (Ito et al., 2001).

On the basis of our results, we hypothesize that TGF-β1 is inducing both pro- and antiapoptotic signals in foetal rat hepatocytes (Figure 8). The activation of PI-3K/Akt precedes the apoptotic pathways and is mediated by the activation of TACE/ADAM 17, which increases the shedding of members of the EGF family of growth factors. Apoptotic machinery, which requires the novo synthesis of proteins (Sánchez et al., 1997) and is mediated by an oxidative stress process (Herrera et al., 2001), could interfere with the survival signals activated by TGF-β1. Thus, it is worth noting that Akt appears to be rapidly downregulated at later times in response to TGF-β1 in a caspases- and reactive oxygen species-dependent manner (Valdés et al., 2004) and c-Src is also caspase-mediated degraded (Park et al., 2004). The balance between both processes (survival and apoptotic signals) will decide cell fate. This scenario, which we observe in primary cells, could occur during the different stages of a carcinogenesis process, where different members of the TGF-β family are highly expressed. TGF-β1, or other members of the family (Lu et al., 2004), might induce both proapoptotic and survival signals. Any response in favour of the antiapoptotic signals would protect cells from death and would allow TGF-β to induce other responses such as increase in motility and migration. Preliminary results in this work indicate that EGFR is required for cell survival, but it is dispensable for the EMT process initiated by TGF-β1 in hepatocytes (Figure 7).

Figure 8
figure8

Proposed model for the pro- and antiapoptotic signals induced for TGF-β1 in hepatocytes

An important consequence of the results presented in this paper is that TGF-β1, by itself, is able to induce both pro- and antitumorigenic effects in a normal cell. The hypothesis that TGF-β1 becomes a protumorigenic factor when intracellular signals, such as Ras or PI-3K, are overexpressed or overactivated will probably change in the next future, if evidence accumulates indicating that TGF-β1 is a double agent even in nontumorigenic cells. Aberrant activation of the EGFR is frequently observed in neoplasia, notably in tumours of epithelial origin. Interestingly, it has been proposed that control of cell survival through EGFR activation is conditional in the sense that it is rate limiting to tumour cell survival, but not to survival of normal epithelial cells (Kari et al., 2003). Considering that TGF-β1 production is elevated in advanced stages of tumorigenesis (Roberts and Wakefield, 2003), it could be possible that tumour cells require EGFR to overcome TGF-β1-proapoptotic effects and respond to this factor inducing cellular events associated with tumour progression. In this line of evidence, recent results in mammary tumours have demonstrated that antimitogenic and prometastatic effects of TGF-β1 can exist simultaneously and HER2 (a human member of the EGFR family) synergies with TGF-β1 in accelerating metastatic tumour progression (Muraoka et al., 2003; Ueda et al., 2004).

In summary, results presented in this paper clearly reveal the potency of TGF-β1 to act as a dual regulator in tumorigenic processes. Survival signals precede the apoptotic machinery and its signal intensity might alter the cell death response. The requirement of EGFR and c-Src for the survival signalling, if corroborated in tumour cells, could open new perspectives in understanding the effectiveness of EGFR tyrosine kinase inhibitors as therapeutic agents combating cancer (Arteaga and Baselga, 2004).

Materials and methods

Materials

Human recombinant TGF-β1, TAPI-1, TAPI-2, AG1478, PP2, and TACE Substrate IV were from Calbiochem (La Jolla, CA, USA). Collagenase was from Roche (Barcelona, Spain). Culture medium was from Invitrogen-Gibco (Carlsbad, CA, USA). Foetal bovine serum was from Sera Laboratories International (West Sussex, UK). EGF was a gift of Serono Lab (Spain). Monoclonal anti-β-actin antibody (A 5441) was from Sigma (Madrid, Spain). Antiphospho-EGF receptor (Tyr1068) (CS-2236), anti-AKT (CS-9272), antiphospho-AKT (Ser472) (CS-9271), anti-EGF receptor (CS-2232), antiphospho-Smad2 (Ser465/467) (CS-3101) and antiphosho-Src (Tyr416) (CS-2101) antibodies were from Cell Signalling Technology (Beverly, MA, USA). Anti-Smad3 was from Upstate (Dundee, UK) (# 06-920).

Foetal rat hepatocytes isolation and culture

Hepatocytes, obtained from 20-day-old foetal Wistar rats by collagenase disruption, were cultured in noncoated plastic dishes with arginine-free M-199 medium, as described previously (Valdés et al., 2004). After cell attachment, serum is removed and after a minimum of 2, or a maximum of 24 h, the factor/s were added. TGF-β1 is used at 1 ng/ml. EGF is used at 20 ng/ml. Inhibitors are added 30 min before the other factors (EGF or TGF-β1).

Western blot analysis

Total protein extracts and nuclear extracts were obtained as described previously (Valdés et al, 2004), separated in 8% polyacrylamide gels (10% for cytoskeleton proteins; 12% for nuclear extracts) by SDS electrophoresis and transferred to PVDF membranes. After blocking with a 5% nonfat dried milk TTBS solution, overnight incubation at 4°C with the corresponding antibody in a 0.5% nonfat dried milk TTBS solution (1 : 5000 for β-actin, 1 : 500 for phospho-Src and phospho-EGF receptor, 1 : 200 for cytokeratin 18 and 1 : 1000 for others), washing and incubation at room temperature with an appropriate peroxidase-conjugated antibody (1 : 5000) for 1 h, antibody binding was visualized using ECL (Amersham). β-actin analysis is shown as loading control.

PI-3K activity assay

After lysis of the cells, proteins were immunoprecipitated with a monoclonal anti-Tyr(P)(Py72) antibody (Fabregat et al., 2000). PI-3 kinase activity was measured in these immunoprecipitates, by in vitro phosphorylation of phosphatidylinositol/L-α-phosphatidyl-L-serine, as we have described previously (Fabregat et al., 2000). The organic phase was extracted, dried in vacuo and resuspended in chloroform. Samples were applied to a silica gel TLC plate (Whatman, Clifton, NJ, USA). TLC plates were developed in propanol-1-acetic acid (2 N; 65 : 35 (vol/vol)), dried and visualized by autoradiography.

Protein analysis

Protein concentration was measured using the Bio-Rad protein reagent.

Analysis of the number of viable cells

In proliferation experiments (i.e. to observe effects on the mitogenic activity of EGF), cells were incubated during 48 h in the absence or presence of EGF and/or inhibitors. In apoptosis experiments, cells were incubated for 24 h in the absence or presence of TGF-β and/or inhibitors). After treatment, cells were washed twice with phosphate-buffered saline and the remaining viable adherent cells were stained with crystal violet (0.2% in 2% ethanol), as described previously (Sánchez et al., 1996). The absorbance of each plate was read photometrically at 560 nm.

Analysis of caspase-3 and TACE activities

Cells were lysed with a 5 mM Tris-HCl, pH 8.0, 20 mM EDTA and 0.5% Triton X-100 buffer. For caspase-3 activity, a reaction mixture containing 20–40 μg of protein extract and 20 μmol/l caspase-3 substrate (Ac-DEVD-AMC) in a 20 mmol/l HEPES, pH 7.5, 10% glycerol and 2 mmol/l dithiothreitol buffer was incubated for 2 h in the dark at 37°C (Valdés et al., 2004). For TACE analysis, a reaction mixture containing 60 μg of protein extract and 10 μ M TACE Substrate IV in a 50 mM Tris-HCl, pH 7.4, 25 mM NaCl and 4% glycerol buffer was incubated at 37°C in the dark during 20 min (Jin et al., 2002). In both cases, fluorescence was measured in a Microplate Fluorescence Reader FL600 (Bio-Tek) (excitation, 380 nm; emission, 440 nm for caspase-3; excitation, 320 nm; emission, 420 nm for TACE). A unit of caspase-3 or TACE activity is the amount of active enzyme necessary to produce an increase in 1 fluorescence unit in the luminescence spectrophotometer. Subsequently, the protein concentration of the cell lysates was determined and the results presented as units of caspase-3 or TACE activity/h/mg of protein.

Analysis of DNA content by flow cytometry

The ploidy determination of hepatocytes was estimated by flow cytometry DNA analysis, as described previously (Valdés et al., 2002).

RT–PCR

Microfast Track isolation kit (Invitrogen, Barcelona, Spain) was used for polyadenylated+RNA isolation from foetal hepatocytes. RT was carried out with oligodeoxythymidylate primer, whereas PCR reactions were performed using mouse specific primers for Snail, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Valdés et al., 2002). PCR products were obtained after 30–35 cycles of amplification at annealing temperatures of 62–65°C.

Abbreviations

TGF:

transforming growth factor

EMT:

epithelial mesenchymal transition

PI-3K:

phosphatidylinositol-3-kinase

EGF:

epidermal growth factor

TACE:

tumour necrosis factor-α-converting enzyme

ADAM:

a disintegrin and metalloprotease

References

  1. Akhurst JR and Derynck R . (2001). Trends Cell Biol., 11, S44–S51.

  2. Akhurst RJ . (2002). J. Clin. Invest., 109, 1533–1536.

  3. Argast GM, Campbell J, Brooling JT and Fausto N . (2004). J. Biol. Chem., 279, 34530–34536.

  4. Arteaga CL and Baselga J . (2004). Cancer Cell, 5, 525–531.

  5. Bakin AV, Tomlinson AK, Bhowmick NA, Moses HL and Arteaga CL . (2000). J. Biol. Chem., 47, 36803–36810.

  6. Batlle E, Sancho E, Franci C, Dominguez D, Monfar M, Baulida J and Garcia De Herreros A . (2000). Nat. Cell Biol., 2, 84–89.

  7. Biscardi JS, Tice DA and Parsons SJ . (1999). Adv. Cancer Res., 76, 61–119.

  8. Black RA, Rauch CT, Kozlosky CJ, Peschon JJ, Slack JL, Wolfson MF, Castner BJ, Stocking KL, Reddy P, Srinivasan S, Nelson N, Boiani N, Schooley KA, Gerhart M, Davis R, Fitzner JN, Johnson RS, Paxton RJ, March CJ and Cerretti DP . (1997). Nature, 385, 729–733.

  9. Borrell-Pages M, Rojo F, Albanell J, Baselga J and Arribas J . (2003). EMBO J., 22, 1114–1124.

  10. Bromann PA, Korkaya H and Courtneidge SA . (2004). Oncogene, 23, 7957–7968.

  11. Cano A, Perez-Moreno MA, Rodrigo I, Locascio A, Blanco MJ, del Barrio MG, Portillo F and Nieto MA . (2000). Nat. Cell Biol., 2, 76–83.

  12. Conery AR, Cao Y, Thompson EA, Townsend Jr CM, Ko TC and Luo K . (2004). Nat. Cell Biol., 6, 366–372.

  13. Fabregat I, Herrera B, Fernández M, Álvarez AM, Sánchez A, Roncero C, Ventura JJ, Valverde AM and Benito M . (2000). Hepatology, 32, 528–535.

  14. Gotzman J, Huber H, Thallinger C, Wolschek M, Jansen B, Schulte-Hermann R, Beug H and Mikulits W . (2002). J. Cell Sci., 115, 1189–1202.

  15. Hart S, Fischer OM and Ullrich A . (2004). Cancer Res., 64, 1943–1950.

  16. Herrera B, Álvarez AM, Sánchez A, Fernández M, Roncero C, Benito M and Fabregat I . (2001). FASEB J., 15, 741–751.

  17. Ito Y, Kawakatsu H, Takeda T, Sakon M, Nagano H, Sakai T, Miyoshi E, Noda K, Tsujimoto M, Wakasa K, Monden M and Matsuura N . (2001). J. Hepatol., 35, 68–73.

  18. Janda E, Lehmann K, Killisch I, Jechlinger M, Herzig M, Downward J, Beug H and Grunert S . (2002). J. Cell Biol., 156, 299–313.

  19. Jin G, Huang X, Black R, Wolfson M, Rauch C, McGregor H, Ellestad G and Cowling R . (2002). Anal. Biochem., 302, 269–275.

  20. Jorissen RN, Walker F, Pouliot N, Garrett TPJ, Ward CW and Burgess AW . (2003). Exp. Cell Res., 284, 31–53.

  21. Kang Y, Siegel PM, Shu W, Drobnjak M, Kakonen SM, Cordón-Cardo C, Guise TA and Massague J . (2003). Cancer Cell, 3, 537–549.

  22. Kari C, Chan TO, Rocha de Quadros M and Rodeck U . (2003). Cancer Cell, 5, 525–531.

  23. Kim JT and Joo CK . (2002). J. Biol. Chem., 277, 31938–31948.

  24. Lu T, Burdelya LG, Swiatkowski SM, Boiko AD, Howe PH, Stark GR and Gudkov AV . (2004). Proc. Natl. Acad. Sci. USA, 101, 7112–7117.

  25. Merlos-Suárez A, Ruiz-Paz S, Baselga J and Arribas J . (2001). J. Biol. Chem., 276, 48510–48517.

  26. Moss ML, Jin SL, Milla ME, Bickett DM, Burkhart W, Carter HL, Chen WJ, Clay WC, Didsbury JR, Hassler D, Hoffman CR, Kost TA, Lambert MH, Leesnitzer MA, McCauley P, McGeehan G, Mitchell J, Moyer M, Pahel G, Rocque W, Overton LK, Schoenen F, Seaton T, Su JL, Warner J, Willard D and Becherer JD . (1997). Nature, 385, 733–736.

  27. Muraoka RS, Koh Y, Roebuck LR, Sanders ME, Brantley-Sieders D, Gorska AE, Moses HL and Arteaga CL . (2003). Mol. Cell. Biol., 23, 8691–8703.

  28. Oft M, Akhurst RJ and Balmain A . (2002). Nat. Cell Biol., 4, 487–494.

  29. Park SS, Eom Y-W, Kim EH, Lee JH, Min DS, Kim S, Kim S-J and Choi KS . (2004). Oncogene, 23, 6272–6281.

  30. Remy I, Montmarquette A and Michnick SW . (2004). Nat. Cell Biol., 6, 358–365.

  31. Roberts AB and Wakefield LM . (2003). Proc. Natl. Acad. Sci. USA, 100, 8621–8623.

  32. Sánchez A, Álvarez AM, Benito M and Fabregat I . (1996). J. Biol. Chem., 271, 7416–7422.

  33. Sánchez A, Alvarez AM, Benito M and Fabregat I . (1997). Hepatology, 26, 935–943.

  34. Sánchez A, Álvarez AM, López-Pedrosa JM, Roncero C, Benito M and Fabregat I . (1999). Exp. Cell Res., 252, 281–291.

  35. Seoane J, Hong-Van L, Shen L, Anderson SA and Massagué J . (2004). Cell, 117, 211–223.

  36. Shin I, Bakin AV, Rodeck U, Brunet A and Arteaga CL . (2001). Mol. Cell. Biol., 12, 3328–3339.

  37. Siegel PM and Massagué J . (2003). Nat. Rev., 3, 807–821.

  38. Sunnarborg SW, Hinkle CL, Stevenson M, Russell WE, Raska CS, Peschon JJ, Castner BJ, Gerhart MJ, Paxton RJ, Black RA and Lee DC . (2002). J. Biol. Chem., 277, 12838–12845.

  39. Ueda Y, Wang S, Dumont N, Yi JY, Koh Y and Arteaga CL . (2004). J. Biol. Chem., 279, 24505–24513.

  40. Valdés F, Álvarez AM, Locascio A, Vega S, Herrera B, Fernández M, Benito M, Nieto A and Fabregat I . (2002). Mol. Cancer Res., 1, 68–78.

  41. Valdés F, Murillo MM, Valverde AM, Herrera B, Sánchez A, Benito M, Fernández M and Fabregat I . (2004). Exp. Cell Res., 292, 209–218.

  42. Viñals F and Pouysségur J . (2001). Mol. Cell. Biol., 21, 7218–7230.

Download references

Acknowledgements

We acknowledge Drs J Gil and MA Nieto for helpful discussions and A Vázquez for assistance in the flow cytometer. This work was supported by Grants from the Ministerio de Educación y Ciencia, Spain (BMC03-524) and the Comunidad Autónoma de Madrid (CAM 08.1/0003.1/2003). M Murillo and G del Castillo are the recipients of fellowships from the Ministerio de Educación y Ciencia, Spain.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Isabel Fabregat.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Murillo, M., Castillo, G., Sánchez, A. et al. Involvement of EGF receptor and c-Src in the survival signals induced by TGF-β1 in hepatocytes. Oncogene 24, 4580–4587 (2005). https://doi.org/10.1038/sj.onc.1208664

Download citation

Keywords

  • apoptosis
  • TGF-β
  • Akt
  • EGFR
  • c-Src
  • hepatocytes

Further reading

Search

Quick links