SEARCH

SEARCH BY CITATION

Keywords:

  • HGF;
  • cisplatin;
  • paclitaxel;
  • p38 MAPK;
  • apoptosis

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We recently showed that Hepatocyte Growth Factor (HGF), known as a survival factor, unexpectedly enhances apoptosis in human ovarian cancer cells treated with the front-line chemotherapeutics cisplatin (CDDP) and paclitaxel (PTX). Here we demonstrate that this effect depends on the p38 mitogen-activated kinase (MAPK). In fact, p38 MAPK activity is stimulated by HGF and further increased by the combined treatment with HGF and either CDDP or PTX. The expression of a dominant negative form of p38 MAPK abrogates apoptosis elicited by drugs, alone or in combination with HGF. HGF and drugs also activate the ERK1/2 MAPKs, the PI3K/AKT and the AKT substrate mTOR. However, activation of these survival pathways does not hinder the ability of HGF to enhance drug-dependent apoptosis. Altogether data show that p38 MAPK is necessary for HGF sensitization of ovarian cancer cells to low-doses of CDDP and PTX and might be sufficient to overcome activation of survival pathways. Therefore, the p38 MAPK pathway might be a suitable target to improve response to conventional chemotherapy in human ovarian cancer. © 2006 Wiley-Liss, Inc.

The sensitivity of cells to chemotherapy depends on the balance between death and survival signals. We have recently shown that the apoptotic response of human ovarian carcinoma cell lines to the first-line chemotherapeutics cisplatin (CDDP) or paclitaxel (PTX) is enhanced by the Hepatocyte Growth Factor (HGF).1 This was surprising, as HGF was reported to protect a variety of normal and cancer cells from several apoptotic stimuli, including chemotherapeutics.2, 3, 4 Integration of HGF-driven cell survival with proliferation and motility elicits “invasive growth,” a biological program that confers cells the capability to invade secondary districts and promotes tumor expansion and invasion when deregulated.5 To foster invasive growth, HGF engages its cognate receptor MET that orchestrates the activation of several transduction pathways.6 In our study, we have explored the HGF/MET-elicited pathways that contribute to apoptosis enhancement in human ovarian carcinoma cells treated with either CDDP or PTX.

The primary intracellular targets of CDDP and PTX are distinct, but both these compounds prompt apoptosis and activate the MAPK pathways.7, 8, 9 The MAPK family of serine/threonine kinases includes the pro-apoptotic p38 MAP and JUN kinases and the primarily anti-apoptotic ERKs.10 HGF promotes survival by triggering ERK1/2 pathway as well,11, 12 but also by activating the PI3K/AKT signaling.3, 11, 13, 14

We here determine that signaling through the p38 MAPK pathway is a major effector of CDDP- or PTX-induced apoptosis and of death sensitization by HGF in ovarian cancer cells. We found that HGF and chemotherapeutics activate, as expected, survival signaling pathways. However, in ovarian cancer cells the activation of the p38 MAPK overwhelms these signals and commits cells to drug-induced apoptosis.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Chemicals and antibodies

Recombinant HGF (rHGF) was obtained from culture supernatant of Sf9 cells infected with the baculovirus vector containing the full-size human factor. Pure human recombinant HGF (pure rHGF) was from R&D Systems (Minneapolis, MN). Cisplatin (CDDP) and paclitaxel (PTX) were from Bristol–Myers Squibb (Rocky Hill, NJ). LY294002 and PD98059 were from Calbiochem (Darmstadt, Germany), rapamycin was from Sigma (St. Louis, MO). Anti-phospho-Thr180/Tyr182 p38 MAPK, anti p38 MAPK, anti-phospho-Ser473 Akt, anti Akt, anti-phospho-Thr202/Tyr204 Erk1/2 and anti-phospho-Thr389 p70S6K rabbit polyclonal antibodies were from Cell Signaling Technology (Beverly, MA). Anti Erk2 rabbit polyclonal antibody and anti p70S6K mouse monoclonal antibody were from Santa Cruz Biotechnology (Santa Cruz, CA); anti FLAG mouse monoclonal antibody was from Stratagene (La Jolla, CA).

Cell lines and apoptosis induction

Experiments were performed on the following cell lines: TOV21G cells, propagated from a primary ovarian adenocarcinoma; the OV-90 cells and the SK-OV-3 cells from ascites of ovarian adenocarcinomas. Cell lines tested were all purchased from ATCC (American Type Culture Collection, Manassas, VA) and grown as suggested by the provider in media supplemented with 5% fetal bovine serum (Euroclone, Wetherby, UK) and 2 mM L-glutamine. Experiments were performed by culturing exponentially growing cells for 48 hr in presence of HGF at the concentration of either 100 ng/ml (rHGF) or 50 ng/ml (pure rHGF) unless otherwise stated, or control medium. Apoptosis was then induced by adding fresh medium, with or without HGF, supplemented with PTX (10 nM) or CDDP (20 μM) for the reported times. Control medium in each experiment contains equal amount of the relevant solvent. The PI3K inhibitor LY294002, the MEK1 inhibitor PD98059 and the mTOR inhibitor rapamycin were added to the medium 30 min before apoptosis induction. Each experiment was repeated 5–7 times.

Flow cytometry analysis of apoptosis induction

Flow cytometry recordings of several independent apoptotic changes were performed by a single-tube analysis, as described.15 Briefly, after induction of apoptosis, cells were resuspended in HEPES buffer (10 mM HEPES, 135 mM NaCl, 5 mM CaCl2) and incubated for 15 min at 37°C in tetramethylrhodamine methyl ester (TMRM, 200 nM), FITC-conjugated Annexin-V and propidium iodide (PI, 1 μg/ml), to detect changes of mitochondrial inner membrane electrochemical gradient, loss of plasma membrane asymmetry (i.e. phosphatidylserine exposure on the cell surface) and integrity (i.e. cell permeability to PI), respectively. Cell morphology changes were analyzed following variations of the forward (FSC) and side light scatter (SSC). Samples were analyzed on a FACSCalibur flow cytometer (Becton Dickinson, San Diego, CA). Data acquisition was performed using CellQuest software and data analysis with WinMDI software. FITC-Annexin-V vs TMRM fluorescent signals were displayed as density plot diagrams on a logarithmic scale. A quadrant was set on the diagram experiment-by-experiment and it was kept constant in all the conditions of each experiment so as to point out the different cell populations. PI-positive cells were detected on a PI vs TMRM plot that was not reported in the figures. The percentage of the apoptotic cells (n=4±S.D.) are also shown as diagrams in Supplementary Material.

Western immunoblot analysis

Cells were lysed at 4°C in a buffer containing 135 mM NaCl, 20 mM Tris/HCl pH 7.5, 1 mM CaCl2, 1% NP40, in the presence of phosphatase and protease inhibitors (1 mM vanadate, 1 mg/ml leupeptine, 1 mM pepstatine, 1 mM PMSF, 100 mg/ml soybean trypsin inhibitor). Proteins were separated on SDS-polyacrylamide gels and transferred onto Hybond-C Extra membranes (Amersham, Little Chalfont, U.K.) following standard methods. Primary antibodies were incubated 16 hr at 4°C, and horseradish peroxidase-conjugated secondary antibodies were added for 1 hr. Proteins were visualized by enhanced chemiluminescence (Amersham).

Lentiviral vector production

To generate Lentiviral vectors by transient transfection, the three-plasmid expression system was used, as previously described.16 The three plasmids were the packaging plasmid, pCMVΔR8.74, designed to provide the HIV proteins needed to produce the virus particle; the envelope-coding plasmid, pMD.G, for pseudotyping the virion with VSV-G, and the self-inactivating (SIN) transfer vector plasmid pRRL.sin.PPT.hCMV.Flag-p38α(AF).pre. The transfer vector plasmid pRRL.sin.PPT.hCMV.Flag-p38α(AF) was obtained substituting the luciferase reporter gene fragment in the pRRL.sin.PPT.hCMV.Luci.pre plasmid with the cDNA coding for a dominant negative Flag tagged p38α MAPK containing T180A and Y182F amino acid substitutions17 obtained from the mammalian expression vector pcDNA3. The transfer vector pRRL.sin.PPT. Ecadpr.Luciferase.pre containing wild-type E-cadherin promoter was used as control vector in experiments. Vector stocks were produced by calcium phosphate transient transfection, co-transfecting the three plasmids in 293T human embryonic kidney cells. Serial dilutions of freshly harvested 293T cell media were used to infect cells in the presence of Polybrene (8 μg/ml). The viral p24 antigen concentration was determined by HIV-1 p24 core profile ELISA Innotest HIV antigen monoclonal antibody (Innogenetics N.V., Ghent, Belgium).

Analysis of JNK and p38 MAPK activity

The enzymatic activity of the p38 MAPK was measured using a kinase assay kit (Cell Signaling Technology). Briefly, active p38 MAPK was extracted and immobilized with anti-phospho-p38 MAPK monoclonal antibody. Then, phosphorylation of its substrate ATF-2 added to the mixture was revealed with anti-phospho-Thr71 ATF-2 antibody in Western blotting. The enzymatic activity of the JNK was measured using a kinase assay kit (Cell Signaling Technology). Briefly, active JNK was extracted and immobilized using its substrate JUN. After the addition of ATP to the mixture, JUN phosphorylation was revealed with anti-phospho-Ser63 JUN antibody in Western blotting.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

HGF enhances chemotherapeutic-induced pro-apoptotic activation of the p38 MAPK pathway in human ovarian cancer cells

The p38 MAPK pathway is activated by cell stress and is involved in cancer cell apoptosis induced by several chemotherapeutics, including platinum drugs and taxanes.7, 18 HGF-driven activation of the p38 MAPK was reported to induce proliferation, migration and morphogenesis,19, 20 but no data link HGF/Met recruitment of this pathway and cell death/survival.

Thus, we first investigated whether HGF activates the p38 MAPK in ovarian cancer cells. We observed that either HGF, CDDP or PTX triggered the p38 MAPK activity, assessed as phosphorylation of the enzyme itself, and that HGF pretreatment enhanced activation caused by each drug alone (Fig. 1a). In addition, HGF strongly increased the p38 MAPK activity toward its substrates. The combined treatment with HGF and drugs amplified both the extent and the duration of the phosphorylation of ATF-2 (Fig. 1b), a transcription factor which, upon phosphorylation, mediates many of the p38 MAPK biological effects.21

thumbnail image

Figure 1. HGF synergizes with chemotherapeutics in activating p38 MAPK, but does not activate JNK. (a): Western blot analysis of p38 MAPK phosphorylation. SK-OV-3 cells were treated with either recombinant purified HGF (50 ng/ml) or CDDP (20 μM) or PTX (10 nM), alone or in combination, for the time indicated on top of panels. Upper panel: detection of phospho-p38 MAPK (P-p38) with an anti-phospho-Thr180/Tyr182. Lower panel: the same blot was re-probed with an anti-p38 MAPK antibody. (b): Detection of p38 MAPK enzymatic activity toward its substrate ATF-2. Immobilized active p38 MAPK was used to phosphorylate exogenous ATF-2. This was revealed with an anti-phospho ATF-2 (phospho-Thr71) specific antibody in Western blotting. In the upper and lower panel short and long term time-courses are shown, respectively. (c): HGF and PTX transiently activate the JUN kinase. Activation of the JNK was detected as enzymatic activity toward its substrate JUN. Immobilized endogenous JUN captures JNK and JUN phosphorylation is revealed with anti-phospho- JUN (phospho-Ser63) antibody in Western blotting.

Download figure to PowerPoint

We then investigated whether HGF-elicited p38 MAPK activation is indeed required for sensitization to CDDP- and PTX-induced apoptotic death. We stably expressed a dominant negative form of p38 MAPK (DN-p38), using a lentiviral vector carrying the double T180A/Y182F Flag-tagged p38 MAPK mutant (see Methods). Lentiviral vectors drive transgene random integration in cell genomic DNA, thus allowing the study of the bulk unselected cell population.16 We transduced SK-OV-3, TOV21G and OV90 human ovarian carcinoma cell lines, which are sensitized by HGF to both CDDP- and PTX-induced cell death (Fig. 2a and 2b and data not shown). Expression of the DN-p38 was detected in 85–90% of the transduced ovarian cancer cell lines, as assessed by Western blot analysis with anti-Flag antibody and FACS analysis (data not shown). Remarkably, the inhibition of the p38 MAPK pathway abolished apoptosis elicited by either CDDP or PTX and enhanced by HGF in all the ovarian cancer cell lines (Fig. 2a and 2b, Fig. 2c and 2d of Supplementary Material and data not shown).

thumbnail image

Figure 2. HGF synergizes with chemotherapeutics in activating p38 MAPK-dependent apoptosis. Apoptosis of ovarian carcinoma cells, SK-OV-3 in (a) and TOV21G in (b), was measured with a multiparametric FACS analysis (see Methods), which measures hallmarks of early, intermediate and late phases of apoptosis, i.e. disruption of the mitochondrial membrane electrochemical gradient and loss of plasma membrane asymmetry and integrity. Cells displaying mitochondrial depolarization are in the lower parts of each plot (reduced TMRM staining), whereas cells exposing phosphatidylserine on their surface are in the right quarters of the diagrams (increased Annexin V-FITC staining). The percentages of healthy cells (H, delimited in the quadrant) and dead, propidium iodide-permeable cells (PI+, not shown in the diagrams), are indicated as numbers in the lower part of each panel. A representative experiment is shown. (a): CDDP- or PTX-mediated apoptosis of SK-OV-3 cells stably expressing either the DN-p38 or a control transgene. SK-OV-3 cells were exposed for 72 hr to PTX (10 nM) or CDDP (20 μM). Where indicated, cells were pretreated for 48 hr with HGF (50 ng/ml). (b): CDDP- mediated apoptosis of TOV21G cells transduced with either a control lentiviral vector or a lentiviral vector carrying the DN-p38. Cells were treated as in panel (a). The percentage of apoptotic cells (n=4±S.D.) is shown in Supplementary Material panel 2c and d.

Download figure to PowerPoint

These data indicate that the p38 MAPK activation is required and has a central role in the induction and modulation of chemotherapeutic-driven apoptosis in ovarian cancer cells.

Chemotherapeutics and HGF transiently activate the JNK pathway in ovarian cancer cells

Similar to p38 MAPK, signaling by the JNK family of MAPKs is fostered by stress stimuli and has been linked to apoptosis,22 also in response to chemotherapeutics.23, 24 In hepatocytes and liver myofibroblasts,25, 26 the pro-apoptotic activity of HGF is associated to JUN kinase (JNK) activation. However, in ovarian cancer cells JNK was not phosphorylated (i.e. it was not activated) in response to either HGF or CDDP and only transiently phosphorylated in response to PTX (data not shown). Then we tested the enzymatic activity toward the JNK substrate, the transcription factor JUN (Fig. 1c). JUN was only weakly and transiently phosphorylated in response to either HGF or PTX, while the combined treatment did not result in increased phosphorylation after short or prolonged treatments. Accordingly, the JNK peptide inhibitor D-JNKI did not affect HGF-mediated sensitization to drugs (data not shown).

Chemotherapeutics and HGF activate the ERK1/2 MAPK survival pathway in ovarian cancer cells

The activation of the ERK1/2 MAPK pathway promotes cell proliferation, migration and differentiation and delivers anti-apoptotic signals following cell stimulation with a variety of growth factors.10 HGF acts as a survival factor via ERK1/2 in both normal, e.g. myocardial27 and endothelial28 cells, and cancer cells.29 In ovarian cancer cells, HGF markedly activated ERK1/2 (Fig. 3a). However, HGF did not further modulate CDDP- or PTX-triggered ERK1/2 activation (Fig. 3a). Pharmacological inhibition of ERK1/2 enhanced apoptosis in drug-treated cells (Fig. 3b and Fig. 3c of Supplementary Material). This indicates that in ovarian cancer cells the activation of ERK-1/2 mediates survival signals, albeit it is not sufficient to inhibit the apoptotic program induced by drugs and to interfere with HGF sensitization to drugs.

thumbnail image

Figure 3. Chemotherapeutics and HGF activate the ERK1/2 MAPK survival pathway. (a): Western blot analysis of ERK1/2 phosphorylation. SK-OV-3 cells were treated with either HGF or CDDP or PTX alone or in combination as described in the legend to Figure 1 for the time indicated on top of panels. Upper panels: detection of phospho-ERK1/2 (P-ERK1/2) with an anti-phospho-Thr202/Tyr204. The MEK1 inhibitor PD98059 (40 μM) was added where indicated (bottom line). Lower panel: the same blot was re-probed with with an anti-ERK2 protein antibody. (b): CDDP- or PTX-mediated apoptosis of SK-OV-3 cells exposed for 72 hr to drugs. Where indicated, cells were pretreated for 48 hr with HGF (50 ng/ml) and maintained in the presence of PD98059 for the experiment time length. A representative experiment demonstrating the output of multiparametric FACS analyses is shown. Diagrams and percentages of the different cell populations are as indicated in the legend to Figure 2. The percentage of apoptotic cells (n=4±S.D.) is shown in Supplementary Material panel 3c.

Download figure to PowerPoint

Chemotherapeutics and HGF activate the PI3K/AKT/mTOR/S6K survival pathway in ovarian cancer cells

Signaling provided by PI3K/AKT mediates HGF protection from apoptotic stimuli in cells propagated from several cancer types.6 In some instances, the activation of PI3K/AKT antagonizes CDDP- and PTX-mediated cytotoxicity.30, 31 It is known that both PI3K and AKT are constitutively active in most human ovarian cancers and tumor-derived cell lines, mainly because of over-expression associated to gene amplification32, 33, 34 and in a few cases because of activating mutations.35 Accordingly, we observed that AKT is constitutively active in ovarian cancer cell lines (Fig. 4a). However, HGF further stimulated AKT phosphorylation, which was also increased by either CDDP or PTX (Fig. 4a). In all experimental conditions, AKT activation relied on PI3K, as it was strongly impaired by the PI3K inhibitor LY294002. The PI3K inhibitor induced apoptosis by itself and added to the pro-apoptotic effect obtained with CDDP or PTX (Fig. 4b and Fig. 4c of Supplementary Material). Altogether, data show that HGF sensitization to drug-induced apoptosis is not mediated by HGF-dependent PI3K/AKT modulation. It is intriguing that even if HGF activates the PI3K/AKT survival pathway, this does not counteract HGF-driven cell sensitization to drugs.

thumbnail image

Figure 4. Chemotherapeutics and HGF activate the PI3K/AKT survival pathway. (a): Western blot analysis of AKT phosphorylation. SK-OV-3 cells were treated with either HGF or CDDP or PTX alone or in combination as described in the legend to Figure 1 for the time indicated on top of panels. Upper panels: detection of phospho-AKT (P-AKT) with an anti-phospho-Ser473. The PI3K inhibitor LY294002 (20 μM) was added where indicated (bottom line). Lower panel: the same blot was re-probed with anti-AKT antibody. (b): CDDP- or PTX-mediated apoptosis of SK-OV-3 cells exposed for 72 hr to drugs. Where indicated, cells were pretreated for 48 hr with HGF (50 ng/ml) and maintained in the presence of LY294002 for the experiment time length. A representative experiment demonstrating the output of multiparametric FACS analyses is shown. Diagrams and percentages of the different cell populations are as indicated in the legend to Figure 2. The percentage of apoptotic cells (n=4±S.D.) is shown in Supplementary Material panel 4c.

Download figure to PowerPoint

One of the downstream effectors of AKT signaling, the mammalian target of rapamycin (mTOR), has been implicated in apoptosis regulation with diverse effects. The mTOR inhibitor rapamycin induces apoptosis of several cancer cells,36 whereas it counteracts death of cells treated with microtubule-damaging agents.37 In addition, it has been shown that HGF activates mTOR.38, 39 To investigate the status of mTOR in ovarian cancer cells, we studied the phosphorylation of the mTOR substrate p70S6K. We observed p70S6K phosphorylation on Thr389 (Fig. 5a) and on Thr421/Ser424 (data not shown) residues in a rapamycin-sensitive manner after cell treatment with HGF and PTX. Phosphorylation at Thr389 was also modestly enhanced by the combined treatment with HGF and CDDP (Fig. 5a). However, rapamycin did not interfere with CDDP- or PTX-mediated apoptosis (Fig. 5b and Fig. 5c of Supplementary Material). It is noteworthy that Thr389 is the p70S6K activation site,40 suggesting that either HGF or each drug mainly regulates p70S6K-dependent processes other than survival in human ovarian cancer cells.

thumbnail image

Figure 5. Chemotherapeutics and HGF activate the mTOR/S6K pathway. (a): Western blot analysis of p70S6K phosphorylation. SK-OV-3 cells were treated with either HGF or CDDP or PTX alone or in combination as described in the legend to Figure 1 for the time indicated on top of panels. Upper panels: detection of phospho-p70S6K (P-p70S6K) with an anti-phospho-Thr389. The mTOR inhibitor rapamycin (100 ng/ml) was added where indicated (bottom line). Lower panel: the same blots were re-probed with anti-p70S6K antibody. (b): CDDP- or PTX-mediated apoptosis of SK-OV-3 cells exposed for 72 hr to drugs. Where indicated, cells were pretreated for 48 hr with HGF (50 ng/ml) and maintained in the presence of rapamycin for the experiment whole time length. A representative experiment demonstrating the output of multiparametric FACS analyses is shown. Diagrams and percentages of the different cell populations are as indicated in the legend to Figure 2. The percentage of apoptotic cells (n=4±S.D.) is shown in Supplementary Material panel 5c.

Download figure to PowerPoint

P38 MAPK pathway is a key death switch in chemotherapeutics and HGF-induced apoptosis

We then wondered if the HGF-activated survival pathways were effective in ovarian cancer cells expressing the DN-p38. As shown in Figure 6ac, although p38 MAPK activation by HGF and drugs was fully impaired, AKT and ERK1/2 were activated by HGF as in control cells (compare Fig. 6 with Figs. 3 and 4). However, cells treatment with PI3K/AKT and ERK1/2 pathways inhibitors almost completely failed to induce apoptosis (Fig. 6d and Fig. 6e of Supplementary Material). Together these data indicate that p38 MAPK is a main death switch in chemotherapeutic-treated ovarian cancer cells, as its constitutive inhibition makes HGF-elicited survival pathways dispensable.

thumbnail image

Figure 6. In cells expressing DN-p38, the PI3K/AKT and the ERK1/2 MAPK survival pathways are activated by HGF but do not influence cell survival. (a): Western blot analysis of p38 MAPK phosphorylation. SK-OV-3 cells were treated with either recombinant purified HGF (50 ng/ml) or CDDP (20 μM) or PTX (10 nM), alone or in combination, for the time indicated on top of panels. In the upper panel phospho-p38 MAPK (P-p38) and in the lower panel total p38 MAPK are detected as in the legend to Figure 1. (b): Western blot analysis of ERK1/2 phosphorylation. SK-OV-3 cells were treated with either HGF or CDDP or PTX alone or in combination as in panel (a) for the time indicated on top of panels. In the upper panel phospho-ERK1/2 (P-ERK1/2) and in the lower panel total ERK1/2 are detected as in the legend to Figure 3. (c): Western blot analysis of AKT phosphorylation. SK-OV-3 cells were treated with either HGF or CDDP or PTX alone or in combination as in panel (a) for the time indicated on top of panels. In the upper panel phospho-AKT (P-AKT) and in the lower panel total AKT are detected as in the legend to Figure 4. (d): CDDP-mediated apoptosis of SK-OV-3 cells exposed for 72 hr to the drug. Where indicated, cells were pretreated for 48 hr with HGF (50 ng/ml) and/or maintained in the presence of either PD98059 or LY294002 for the experiment time length. A representative experiment demonstrating the output of multiparametric FACS analyses is shown. Diagrams and percentages of the different cell populations are as indicated in the legend to Figure 2. The percentage of apoptotic cells (n=4±S.D.) is shown in Supplementary Material panel 6e.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

In our study, we show that the p38 MAPK pathway activation is the switch that makes HGF, a known survival factor, able to sensitize ovarian cancer cells to cisplatin and paclitaxel. HGF commits ovarian cancer cells to drug-induced cell death, even though it activates the PI3K/AKT and ERK1/2 survival pathways.

The mechanisms by which MET activation by its ligand HGF impinges upon the regulation of the cell death/survival balance are only partially understood. HGF/MET axis protects most cancer cells from apoptosis,6 but pro-apoptotic function of HGF/MET has been observed by us and others in human ovarian carcinoma1 and sarcoma cell lines.41, 42 Here we show that in ovarian cancer cells HGF triggers survival signaling pathways, but these are overwhelmed by the dominant role of p38 MAPK in apoptosis induction in cells treated with drugs. In fact, neither PI3K/AKT nor ERK1/2 MAPK activity interferes with HGF sensitization to chemotherapeutics, while DN-p38 fully protects cells from death. Cell death is a binary decision, and the apoptotic machinery is a rheostat-like system, whose equilibrium derives from the integration of diverse signals. For instance, in some cell lines, elevated levels of EGFR and ErbB-2 result in the induction of ligand-dependent cell death rather than proliferation, the former being in all cases mediated by p38 MAPK.43 The output signal of HGF seems to depend on the context, and this could explain the reported antinomic effects of HGF/MET on survival in different cancer cells models. However, in the panel of ovarian carcinoma cell lines that we studied, HGF invariably sensitizes cells to drug-induced apoptosis, and p38 MAPK activity is necessary. This effect seems to be characteristic of ovarian cancer cells, as shown in this and our previous work.1 We could form the hypothesis that the p38 MAPK pathway might be particularly effective in counteracting the anti-apoptotic action of the oncogenic pathways consistently activated in ovarian cancer cells, such as those of growth factor receptors and the downstream kinases that transduce signals from the membrane to the nucleus.33, 44, 45

Here we show that the p38 MAPK in the main effector of the HGF sensitization to drugs in human ovarian cancer cells. In fact, dominant negative p38 MAPK not only rescued drug-induced cell death but also abrogated HGF-dependent enhanced susceptibility to chemotherapeutics. We did not explore several other pathways that could be implicated in modulating ovarian cancer cell death. However, we show that the p38 MAPK activation might be sufficient to commit ovarian cancer cells to death as it overwhelms the concomitant activation of important and effective survival pathways.

The mechanism of p38 MAPK downstream signaling to apoptosis has long been studied, but poorly clarified. Targets of the p38 MAPK are several molecules participating in transcription regulation21, 46, 47, 48 and in cell cycle control.49 Among them, some have been already associated to p38 MAPK activation by chemotherapeutics: in cardiomyocites undergoing apoptosis in response to doxorubicin, the transcription factor p300 is phosphorylated and degraded upon p38 MAPK activation by the drug.48

The finding that HGF sensitizes ovarian carcinoma cells to conventional anti-cancer drugs might suggest therapeutic implications. In ovarian cancer, first-line treatment with platinum drugs and taxanes has become the standard therapy after cytoreductive surgery. However, many patients develop resistance during postoperative chemotherapy. To the aim of overcoming unresponsiveness, innovative therapeutic strategies are under current investigation.50 In principle, HGF could be used to improve response to CDDP and PTX in a set of human ovarian carcinomas molecularly classified on the basis of HGF receptor expression. However, clinical use of HGF as such is likely to be unsuitable, as HGF is known as a pro-invasive and pro-angiogenic factor. In particular, HGF stimulates invasiveness of ovarian carcinoma cells in vitro.51, 52 The identification of downstream signaling molecules implicated in HGF sensitization to drugs might indicate how to circumvent HGF undesired biological effects in vivo. Altogether, these data might point to p38 MAPK, its regulatory molecules and its substrates as possible therapeutic targets. For instance, the Wip1 phosphatase has been identified as a negative regulator of p38 MAPK and proposed as a suitable therapeutic target.53 Interestingly, we found that the pharmacological inhibition of the ERK1/2 pathway potentiated drug-induced apoptosis, in particular when HGF is added. This opens the possibility of combinatorial therapy, given that several agents targeting the ERK1/2 pathway have been developed and are in clinical trials.54 Conversely, we also show that ovarian cancer cells are insensitive to both conventional chemotherapeutics and inhibitors of survival pathways when p38 MAPK function is impaired. This raises the possibility that human ovarian carcinomas showing low level or inactive p38 MAPK55 might resist to both conventional and innovative treatments.

Our findings might be relevant to a high number of human ovarian cancers. Among serine/threonine kinases, those of the p38 MAPK pathway are activated in a model of ovarian cancer progression.56 In addition, a survey of human ovarian serous carcinomas showed an increased expression and activation of p38 MAPKs in ∼60% of cases and a positive correlation between p38 MAPK activation and better overall patient survival.55 Understanding which biochemical processes are affected in tumor development is mandatory to achieve highly selective anti-tumor devices. Therefore, the p38 MAPK pathway might be an elective target to obtain the desired sensitization of ovarian cancer to first-line chemotherapeutics at low doses.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We thank Mr. Vincenzo De Sio, Mrs. Raffaella Albano and Miss Solange Tienga for technical help and Miss Anna Adezati for reading the English. We thank Dr. J. Han and Dr. Simone Grethe for providing the FlagDNp38 MAPK plasmid. We thank Paola Circosta for helping with FACS analysis. This work was supported by Italian Ministry of Research and Education (MIUR), by the Italian National Research Council (C.N.R.-MIUR Progetto Oncologia) and Regione Piemonte funding to M.D.R. and A.G. and by the Italian Association for Cancer Research (AIRC) funding to M.D.R, P.M.C. and A.G. A.R. and M.F. are recipient of Regione Piemonte Ricerca Scientifica Applicata fellowships. N.F. is recipient of FIRB-MIUR young researcher appointment.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This article contains supplementary material available via the Internet at http://www.interscience.wiley.com/jpages/0020-7136/suppmat .

FilenameFormatSizeDescription
jws-ijc.21766.tif20135KSupplementary Figures

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.