- Email: [email protected]
APO-1) expression
Experimental Cell Research 300 (2004) 354 – 364 www.elsevier.com/locate/yexcr
GSK-3h inhibition by lithium confers resistance to chemotherapy-induced apoptosis through the repression of CD95 (Fas/APO-1) expression Ele´onore Beurela, Michel Kornprobstb, Marie-Jose´ Blivet-Van EggelpoJla, Carmen Ruiz-Ruizc, Axelle Cadoreta, Jacqueline Capeaua, Christe`le Desbois-Mouthona,* a
INSERM U.402, Faculte´ de Me´decine Saint-Antoine, Paris, France b IFR65, Faculte´ de Me´decine Saint-Antoine, Paris, France c Department of Biochemistry and Molecular Biology, University of Granada, Granada, Spain Received 1 June 2004, revised version received 2 August 2004 Available online 3 September 2004
Abstract Lithium exerts neuroprotective actions that involve the inhibition of glycogen synthase kinase-3h (GSK-3h). Otherwise, recent studies suggest that sustained GSK-3h inhibition is a hallmark of tumorigenesis. In this context, the present study was undertaken to examine whether lithium modulated cancer cell sensitivity to apoptosis induced by chemotherapy agents. We observed that, in different human cancer cell lines, lithium significantly reduced etoposide- and camptothecin-induced apoptosis. In HepG2 cells, lithium repressed drug induction of CD95 expression and clustering at the cell surface as well as caspase-8 activation. Lithium acted through deregulation of GSK-3h signaling since (1) it provoked a rapid and sustained phosphorylation of GSK-3h on the inhibitory serine 9 residue; (2) the GSK-3h inhibitor SB415286 mimicked lithium effects by repressing drug-induced apoptosis and CD95 membrane expression; and (3) lithium promoted the disruption of nuclear GSK-3h/p53 complexes. Moreover, the overexpression of an inactivated GSK-3h mutant counteracted the stimulatory effects of etoposide and camptothecin on a luciferase reporter plasmid driven by a p53-responsive sequence from the CD95 gene. In conclusion, we provide the first evidence that lithium confers resistance to apoptosis in cancer cells through GSK-3h inhibition and subsequent repression of CD95 gene expression. Our study also highlights the concerted action of GSK-3h and p53 on CD95 gene expression. D 2004 Elsevier Inc. All rights reserved. Keywords: Lithium; Apoptosis; GSK-3h; p53; Chemotherapy; CD95; Caspase-8
Introduction Lithium is the most widely used treatment for bipolar mood disorder. Recent evidence shows that lithium can protect neuronal cells from apoptosis induced by a wide array of neurotoxic insults [1,2]. The neuroprotective properties of lithium could be an important component of its therapeutic action. Analysis of the molecular mechanisms underlying lithium-induced neuroprotection has revealed that lithium inhibits the activity of glycogen synthase * Corresponding author. INSERM U.402, Faculte´ de Me´decine SaintAntoine, 27 rue Chaligny, 75571 Paris Cedex 12, France. Fax: +33 1 40 01 13 52. E-mail address: [email protected] (C. Desbois-Mouthon). 0014-4827/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2004.08.001
kinase-3h (GSK-3h), and based on this finding, lithium is now commonly used as a potent inhibitor of GSK-3h [1]. GSK-3h is a serine/threonine kinase active in resting cells and implicated in metabolism, proliferation, and gene transcription as a result of its ability to phosphorylate and inactivate key proteins governing these processes such as glycogen synthase, cyclin D1, and h-catenin [3,4]. GSK-3h undergoes rapid and transient inhibition through sitespecific phosphorylation on the serine 9 residue in response to diverse stimuli including insulin and insulin-like growth factors [3,4]. Recent studies suggest that sustained GSK3hSer9 phosphorylation could contribute to tumorigenesis. Thus, we and others have reported that GSK-3h is hyperphosphorylated on serine 9 in human hepatoma cell lines as well as in human and murine tumoral livers [5–7]. In
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
addition, chemically induced mouse skin carcinogenesis has been associated with GSK-3hSer9 hyperphosphorylation [8]. Interestingly, it has been reported that lithium increases GSK-3hSer9 phosphorylation [9–11], thus showing that lithium inhibits GSK-3h activity through a mechanism similar to that observed in cancer cells. Evidence obtained in neuronal cells has revealed that GSK-3h is a proapoptotic kinase. Thus, overexpression or hyperactivation of GSK-3h sensitizes neuronal cells to apoptosis [10,12–15]. Recently, studies by Jope et al. [16– 18] have shown that GSK-3h is located in mitochondria where it is highly active during apoptosis. The functional outcome of this mitochondrial interaction is unclear, but it might facilitate apoptosis since lithium-mediated GSK-3h inhibition attenuates cytochrome c release from the mitochondria [18,19]. Because GSK-3h is an important mediator of apoptotic signals, it is plausible that GSK-3h deregulation observed in cancer cells confers resistance to chemotherapy-induced apoptosis which is a major cause of treatment failure in human cancers [20]. In this context, we were interested to investigate whether lithium modulated chemotherapyinduced apoptosis in cancer cells. We observed that lithium repressed apoptosis induced by etoposide and camptothecin in different cancer cell lines. The molecular mechanisms responsible for chemoresistance were then examined, focusing the analyses on the CD95 (Fas/APO-1) death receptor signaling pathway.
Materials and methods Cell culture HepG2 (hepatoblastoma) cells were maintained in minimal essential medium (MEM) containing Earle salts, 1% nonessential amino acids, 1 mM sodium pyruvate, and 10% fetal calf serum. A549 (lung carcinoma) and MCF-7 (breast carcinoma) cell lines were cultured in Dulbecco MEM supplemented with 10% fetal calf serum. Cell lines were from the American Type Culture Collection. Before experiments, cells were incubated overnight in serum-free medium. In some assays, cells were treated with dimethylsulfoxide (DMSO), etoposide, camptothecin, lithium chloride (SigmaAldrich Co.), a general caspase inhibitor (Z-VAD-fmk), a caspase-8 inhibitor (Z-IETD-fmk, R and D Systems, Inc.), or with a GSK-3h small inhibitor (SB- 415286, Tocris). DNA fragmentation Both adherent and floating cells were collected and lysed in 1% Nonidet P-40, 20 mM EDTA, and 50 mM Tris-HCl (pH 7.5). After centrifugation, supernatants were incubated with 1% SDS and 5 mg/ml ribonuclease A for 3 h at 568C. Then, 2.5 mg/ml proteinase K was added overnight at 378C. DNAs were precipitated at 208C in the presence of 10 M
355
ammonium acetate and ethanol, resuspended in water, and analyzed by electrophoresis on 1% agarose gels. Flow cytometry analysis of apoptosis Both adherent and floating cells were collected, washed, fixed in ice-cold 70% ethanol at 208C, and stained with 20 Ag/ml propidium iodide in the presence of 100 Ag/ml ribonuclease A for 30 min at 378C in the dark. DNA content was analyzed by flow cytometry (FACS Calibur, Becton Dickinson). Apoptotic cells with hypodiploid DNA staining were found in the Tsub-G1r peak. Confocal laser microscopy HepG2 cells cultured on 12-mm glass coverslips were fixed in 4% paraformaldehyde (Sigma-Aldrich Co.) for 10 min and incubated for 2 h at room temperature with an antiCD95 antibody (ZB4, MBL International Co.) diluted 1:100 in PBS containing 1% bovine serum albumin. Samples were then incubated for 45 min with a secondary antibody (Alexa FluorR 488 conjugate, Molecular Probes Inc., 1:100 dilution). Confocal microscopy was performed using a Leica TCS SP microscope equipped with a 63 objective (Lasertechnik GmbH). Reverse transcription–polymerase chain reaction (RT–PCR) CD95 and CD95 ligand mRNAs were analyzed by RT–PCR using the following primers: CD95, 5VGCGAAAGCCCATTTTTCTTCC-3V and 5V-ATTTATTGCCACTGTTTCAGG-3V; CD95 ligand, 5V-CAG C T C T T C C A C C T A C A G A A G G A - 3 V a n d 5 VGAGAGCTCAGATACGTTGAC-3V. For semiquantitative analysis, concurrent h-actin PCR products (primers 5VAT C AT G T T T G A G A C C T T C A A - 3 V a n d 5 V- T T G CGCTCAGGAGGAGCAAT-3V) were generated in parallel to ensure that equivalent amounts of template were amplified. For CD95 quantitation, real-time PCR was performed with the LightCycler system (Roche Molecular Biochemicals) using the SYBR green fluorophore. Each sample was normalized on the basis of its endogenous content in 18S rRNA (primers: 5 V-GAGCGAAAGCATTTGCCAAG3Vand 5V-GGCATCGTT-TATGGTCGGAA-3V). Flow cytometry detection of CD95 Both adherent and floating cells were collected and fixed in 4% paraformaldehyde (Sigma-Aldrich Co.) for 10 min and incubated for 1.5 h at room temperature with an antiCD95 antibody (ZB4) diluted 1:300 in PBS containing 1% bovine serum albumin. Samples were then incubated for 45 min with a secondary antibody (Alexa FluorR 488 conjugate, 1:500 dilution). Unstained cells and cells stained with the secondary antibody alone were run in parallel as
356
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
negative controls. After staining, cells were resuspended in PBS and analyzed by flow cytometry. Transient transfections Cells were plated at a density of 3 105 cells in sixwell plates the day before transfection. Cells were transfected using Transfast (Promega Corp.) with 0.5 Ag of the luciferase reporter plasmid pI-CD95 1448Luc or pCD95 1448Luc (generous gifts from Dr. A. LopezRivas) [21] and 0.25 Ag of ph-gal-Control (BD Biosciences Clontech) together with 0.5 Ag of either an empty expression vector (pcDNA3, InVitrogen) or of an expression vector encoding an inactive K85R-GSK-3h (generous gift from Dr. C. Michiels). Luciferase and hgalactosidase activities were evaluated as previously reported [22], and luciferase activities were normalized to h-galactosidase activities. Immunoprecipitation and Western blotting Whole-cell and cytosolic extracts were prepared as previously reported [22]. Lysates for caspase expression were prepared as recommended by Cell Signaling Technology Inc. Nuclear extracts were prepared from about 1 106 cells by using NE-PERk nuclear and cytoplasmic kit (Pierce). For immunoprecipitation, nuclear extracts were incubated overnight at 48C with an antiphospho-Ser15 p53 (Cell Signaling Technology, Inc.) or an anti-GSK-3h (Transduction Laboratories) antibody at a 1:200 dilution together with 20 Al of protein A/G PLUSagarose (Santa Cruz Biotechnology, Inc.). For Western blot analysis, nitrocellulose membranes were probed with the following antibodies: anti-phospho-Ser9 GSK-3h, anticaspase-3, anti-caspase-8, anti-phospho-Ser15 p53 (Cell Signaling Technology, Inc.), anti-poly(ADP-ribose) polymerase (Santa Cruz Biotechnology, Inc.), anti-p53 (clone BP53-12, Upstate Biotechnology), and anti-GSK-3h. Membranes were reprobed with an anti-h-actin antibody (Sigma-Aldrich Co.) to ensure equivalent loading. Caspase-3 and caspase-8 assays Cells were pelleted, washed with PBS, lysed in 50 mM Tris-HCl (pH 7.5), 0.03% Nonidet P-40, and 1 mM dithiotreitol, and centrifuged. Assays were set up in 96well plates containing supernatants (100 Ag proteins), 0.2 mM of caspase-3 substrate (Ac-DEVD-pNA) or of caspase-8 substrate (Ac-IETD-pNA) (BIOMOL Research Laboratories, Inc.) in a caspase reaction buffer (100 mM HEPES pH 7.5, 10% sucrose, 0.1% CHAPS, 10 mM DTT). Assays were performed at 378C, and release of pNa was detected by periodic readings of absorbance at 405 nm from 0 to 5 h to mark the linearity of the enzymatic reaction in time. Enzyme activities were measured as initial velocities.
Statistical analysis Results are given as the mean F SEM. Statistical comparison of mean values was performed using the Student t test.
Results Lithium confers resistance to etoposide- and camptothecin-induced apoptosis in human cancer cells We first examined the effect of lithium on the sensitivity of hepatoma HepG2 cells to inhibitors of topoisomerases I (camptothecin) and II (etoposide). To this aim, cells were treated for 48 h with solvent alone (dimethylsulfoxide), etoposide (85 AM), or camptothecin (0.5 AM) in the presence or absence of lithium (20 mM) before being analyzed for DNA fragmentation by gel electrophoresis and for the fraction of hypodiploid cells by flow cytometry. We used lithium at a 20-mM concentration since this concentration has been previously reported to potently inhibit GSK-3h in vitro [9–11]. Etoposide and camptothecin induced apoptosis in HepG2 cells since (i) oligonucleosomal DNA fragmentation typical of apoptosis was observed in cells treated with each of these drugs (Fig. 1A); (ii) etoposide and camptothecin increased the number of sub-G1 cells from 5.9% F 0.4% to 16.8% F 1.9% and to 44.8% F 3.2%, respectively (Fig. 1B); and (iii) the general caspase inhibitor Z-VAD-fmk totally inhibited the effects of etoposide and camptothecin (Fig. 1B, inset). We observed that lithium treatment strongly reduced oligonucleosomal DNA fragmentation (Fig. 1A) as well as the number of apoptotic cells induced by each drug (Fig. 1B, 7.7% F 1.4% and 19.7% F 2.6% in the presence of etoposide and camptothecin, respectively), indicating that lithium conferred resistance to chemotherapy-induced apoptosis in HepG2 cells. To examine whether lithium interfered with etoposideand camptothecin-induced apoptosis in other cancer cell types, flow cytometry analyses of sub-G1 cells were also conducted in A549 (lung carcinoma) and MCF-7 (breast carcinoma) cells. Again, we observed that lithium significantly reduced the number of apoptotic cells induced by etoposide and camptothecin in the two lines studied (Fig. 2). The maximal inhibitory effects of lithium were observed with the HepG2 cell line and, thereby, the molecular mechanisms whereby lithium conferred resistance to chemotherapy drugs were analyzed in these cells. Lithium inhibits GSK-3b during etoposide- and camptothecin-induced apoptosis in HepG2 cells We next examined the impact of lithium on GSK-3h activity in HepG2 cells by measuring the phosphorylation
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
357
Lithium prevents activation of the CD95 death pathway during etoposide- and camptothecin-induced apoptosis in HepG2 cells
Fig. 1. Lithium confers resistance to etoposide- and camptothecin-induced apoptosis in HepG2 cells. HepG2 cells were treated for 48 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium chloride (LiCl, 20 mM) and analyzed for DNA fragmentation by electrophoresis (A) and for hypodiploid cell (sub-G1) contents by flow cytometry (B). Results are means F SEM of six independent experiments. Inset, HepG2 cells were treated for 48 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without a general inhibitor of caspases (Z-VAD-fmk, 20 AM), and hypodiploid cells were analyzed by flow cytometry. Results are representative of two independent experiments. ***P b 0.01 compared with cells treated with drug alone.
level of GSK-3h at serine 9 as an indicator of GSK-3h inactivation. We observed that lithium rapidly increased GSK-3h phosphorylation on the serine 9 regulatory site. GSK-3hSer9 phosphorylation occurred within 5 min of lithium treatment and was maintained for at least 48 h, without any modification in the total GSK-3h level (Fig. 3A). In HepG2 cells, etoposide or camptothecin altered neither the basal level of phosphorylated GSK-3h nor that induced by lithium (Fig. 3B). Of note, lithium also increased GSK-3hSer9 phosphorylation in A549 and MCF-7 cells (data not shown). To confirm the contribution of GSK-3h inhibition to chemoresistance, HepG2 cells were incubated with or without SB-415286, a small inhibitor of GSK-3h which acts as an ATP competitive inhibitor [23]. We observed that 30 AM SB-415286 reduced by 48% and 38% apoptosis induced by a 12-h treatment with etoposide and camptothecin, respectively, thus supporting a role for GSK-3h inhibition in lithium-induced resistance to chemotherapy (Fig. 3C).
Etoposide and camptothecin induce apoptosis in cancer cells via the CD95 death pathway through the enhancement of CD95 ligand and/or CD95 expression [24, 25]. Activation of the CD95 signaling pathway results from CD95 receptor trimerization and clustering at the cell surface, which in turns leads to the recruitment of the adaptor protein FADD and caspase-8 and the subsequent activation of downstream caspases. Experiments were designed to examine whether lithium conferred resistance to etoposide- and camptothecin-induced apoptosis through the modulation of the CD95 signaling pathway. We first used confocal laser microscopy to analyze the effects of anticancer drugs and lithium on CD95 aggregation at the cell surface. While control cells (DMSO) had diffuse staining of CD95, cells treated with etoposide or camptothecin exhibited CD95 clustering as evidenced by a dense staining that was primarily membranelocalized (Fig. 4A). In the presence of lithium, drug effects were diminished. Induction of CD95 ligand mRNA was not observed in etoposide- and camptothecin-treated HepG2 cells (Fig. 4B). However, etoposide and camptothecin increased CD95 mRNA expression in HepG2 cells evaluated by semiquantitative (Fig. 4C, left) and quantitative real-time RT–PCR (Fig. 4C, right) as well as CD95 expression at the cell surface evaluated by flow cytometry (Fig. 4D, left). Importantly, we observed that lithium prevented etoposide- and camptothecininduced up-regulation of CD95 mRNA (Fig. 4C) and protein expression at the cell surface (Fig. 4D). Similarly, SB-415286 repressed the induction of CD95 expression at the cell surface by etoposide (Fig. 4E) and camptothecin (data not shown), thus supporting a role for GSK3h inhibition in the repressive effect of lithium on CD95 expression. Upon CD95 activation, caspase-8 is the main initiation caspase. In accordance with these data, we observed that etoposide and camptothecin induced caspase-8 activation in HepG2 cells measured by the appearance of cleaved active caspase-8 bands and by a colorimetric assay (Fig. 5A). Moreover, etoposide- and camptothecin-induced apoptosis was inhibited by the caspase-8 inhibitor ZIETD-fmk (data not shown). Cell treatment with lithium prevented etoposide and camptothecin induction of caspase-8 activity (Fig. 5A). Lithium also repressed drug activation of the executioner caspase-3 measured by the appearance of cleaved active caspase-3 bands, by a colorimetric assay, and by the proteolysis of the caspase3 substrate PARP (Fig. 5B). Altogether, these data show that lithium interferes with etoposide and camptothecin induction of the CD95 signaling pathway by repressing CD95 expression (mRNA and protein) and clustering as well as caspase-8 activation.
358
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
Fig. 2. Lithium confers resistance to etoposide- and camptothecin-induced apoptosis in human cancer cell lines. A549 and MCF-7 cells were treated for 48 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM for A549 cells and 0.125 AM for MCF-7 cells) with or without lithium chloride (LiCl, 20 mM), and hypodiploid cells were analyzed by flow cytometry. Results are means F SEM of three independent experiments. *P b 0.05, **P b 0.02 compared with cells treated with drug alone.
GSK-3b inhibition blocks p53-dependent transcriptional activation from the CD95 gene promoter in HepG2 cells
Fig. 3. Lithium inhibits GSK-3h during etoposide- and camptothecininduced apoptosis in HepG2 cells. Whole-cell extracts from HepG2 cells were analyzed by Western blotting for GSK-3hSer9 phosphorylation and total GSK-3h. (A) Cells were treated with or without lithium (LiCl, 20 mM) for various incubation times. (B) Cells were treated for 24 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM) or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM). Blots are representative of three independent experiments. (C) Cells were treated with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without SB-415286 (30 AM) for 12 h. Hypodiploid cells were quantified by flow cytometry. Results are means F SEM of two independent experiments. *P b 0.05, compared with cells treated with drug alone.
It has been clearly demonstrated that, in cancer cells, upregulation of CD95 gene expression after genotoxic treatment is mediated through p53 activation [21,26–28]. We then examined the impact of GSK-3h inhibition on the transcriptional regulation of the CD95 gene. To this end, transient transfection experiments were performed using the plasmid pICD95 1448Luc in which the expression of a luciferase reporter gene is driven by regulatory DNA elements from the human CD95 gene [21]. These regulatory sequences include 1448 bp of the 5V-promoter and 500 bp of the first intron containing a binding element for p53. It has been previously shown that this p53 consensus site is absolutely required for induction of pICD95 1448Luc expression upon chemotherapy treatment [21]. In HepG2 cells, we observed that etoposide and camptothecin stimulated luciferase expression from the pICD95 1448Luc construct, while both drugs did not increase luciferase expression in the absence of the intronic region (pCD95 1448Luc construct) (Fig. 6A). When cells were transfected with a plasmid encoding a dominant-negative GSK-3h (K85R-GSK-3h), we observed that the stimulatory effects of drugs on pI-CD95 1448Luc were significantly decreased as compared to cells transfected with an empty vector (pcDNA3) (Fig. 6B). These findings show that GSK-3h inhibition prevents the stimulation of p53-dependent pI-CD95 1448Luc expression by chemotherapy drugs in HepG2 cells. Lithium inhibits p53 accumulation and nuclear association of p53 with GSK-3b during etoposide- and camptothecin-induced apoptosis in HepG2 cells The above results prompted us to analyze the effect of lithium on GSK-3h and p53 expression and interaction
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
359
Fig. 4. Lithium prevents activation of the CD95 death pathway during etoposide- and camptothecin-induced apoptosis in HepG2 cells. HepG2 cells were treated with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM) or SB-415286 (SB, 30 AM). (A) Cells treated for 12 h were fixed and labeled for confocal laser microscopy using the antihuman CD95 antibody ZB4. (B) After a 12-h incubation, CD95 ligand expression was analyzed by semiquantitative RT–PCR. mRNA obtained from HepG2 cells treated with methotrexate (Metho, 100 Ag/ml) was used as a positive control. (C) After a 12-h treatment, CD95 expression was analyzed by semiquantitative (left) and quantitative real-time (right) RT–PCR. (D) Cells treated for 24 h were collected and labeled for flow cytometry analysis using the antihuman CD95 antibody ZB4. Grey histogram, nonspecific; black histogram, DMSO-treated; green histogram, ETP-treated; blue histogram, CPT-treated; histograms with dotted lines, cotreatment with lithium. (E) Cells treated for 24 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without SB-415286 (30 AM) were collected and labeled for flow cytometry analysis using the antihuman CD95 antibody ZB4. Grey histogram, nonspecific; black histogram, DMSO-treated; green histogram, ETP-treated; histograms with dotted lines, cotreatment with SB-415286. Results are representative of three independent experiments.
360
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
Fig. 5. Lithium prevents activation of caspase-8 and caspase-3 during etoposide- and camptothecin-induced apoptosis in HepG2 cells. HepG2 cells were treated with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM). (A) After a 4-h incubation, cell lysates were prepared and analyzed for caspase-8 activation by Western blotting using an antibody that allows detection of full-length (p57/55) and cleaved (p43/41) caspase-8 (left) or by a colorimetric assay using a caspase-8-specific peptide conjugated to a chromogene (Ac-IETD-pNA) (right). (B) After a 24-h incubation, cell lysates were prepared and analyzed for caspase-3 activation by Western blotting using antibodies that allow detection of cleaved (p19/17) caspase-3 or of cleaved (p85) poly(ADP-ribose) polymerase (PARP) (left) or by a colorimetric assay using a caspase-3-specific peptide conjugated to a chromogene (Ac-DEVD-pNA) (right). Blots and data are representative of three independent experiments.
during etoposide and camptothecin induction of apoptosis. In HepG2 cells, we observed that etoposide and camptothecin increased total (Fig. 7A) and nuclear (Fig. 7B) levels of p53. In the presence of lithium, p53 accumulation induced by etoposide and camptothecin was markedly reduced (Figs. 7A,B). Phosphorylation of NH2-terminal residues of p53 mediates its stabilization and nuclear accumulation after anticancer drug treatment [29]. Consistent with this notion, we observed that etoposide and camptothecin stimulated p53Ser15 phosphorylation in nuclear extracts (Fig. 7C). Again, in the presence of lithium, drug induction of p53Ser15 phosphorylation was markedly reduced. Etoposide and camptothecin promoted p53 nuclear accumulation without modifying phosphorylated and total levels of GSK-3h, while cotreatment with lithium reduced p53 accumulation and increased GSK-3hSer9 phosphorylation but not GSK-3h expression (Fig. 7B). Examination of a direct association of p53 with GSK-3h was performed by coimmunoprecipitation experiments. Immunoblots revealed that etoposide and camptothecin increased the coimmunoprecipitation of phospho-p53Ser15 with GSK-3h in nuclear fractions from HepG2 cells (Fig. 7D). In the presence of lithium, drug induction of nuclear phospho-p53Ser15/GSK-3h complexes was clearly diminished. Taken together, these findings show that lithium prevents p53 accumulation induced by etoposide or camptothecin treatment which subsequently reduces the formation of nuclear GSK-3h/ p53 complexes.
Discussion Lithium provides effective pharmacotherapy for bipolar mood disorder. Although it is not carcinogenic by itself [30], it is not excluded that, in some cellular context, lithium could favor the propensity for developing certain cancers. For example, in a murine model genetically predisposed to intestinal adenomas, lithium treatment significantly increases tumor number in colon and overall tumor size [31]. In addition, lithium has been shown to stimulate proliferation in some cancer cell types [32]. Moreover, we and others have reported that lithium increases nuclear levels of h-catenin [6,11,23,33], a protein which associates with LEF/TCF transcription factors to stimulate expression of proliferative genes. In this report, we now provide evidence that, in human cancer cells, lithium confers resistance to apoptosis induced by two different chemotherapy drugs. These findings were obtained with 20 mM lithium which is a concentration usually used in in vitro studies to maximize the observable changes. In patients treated for bipolar disorders, the plasma concentration range of lithium is 0.5–1.0 mM. Long-term treatment of mice with such doses causes a large increase in the level of GSK-3hSer9 phosphorylation in brain [9] and in liver (E. Beurel, personal observation). Taken together, these findings could suggest that short-term in vitro experiments conducted with a high dose of lithium mimic cellular responses obtained in vivo after the chronic administration of a low dose of lithium.
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
361
first time that lithium also potently inhibits etoposide and camptothecin stimulation of the CD95 signaling pathway. We observed that lithium repressed drug induction of CD95 expression and clustering at the cell surface as well as caspase-8 activation. Of note, neither etoposide nor camptothecin induced CD95 ligand mRNA expression in HepG2 cells. This suggests that drug stimulation of CD95 clustering most probably resulted from up-regulation of CD95
Fig. 6. GSK-3h inhibition blocks p53-dependent transcriptional activation from the CD95 gene promoter in HepG2 cells. (A) HepG2 cells were transiently transfected with 0.25 Ag ph-gal-Control together with 0.5 Ag of pI-CD95 1448Luc containing a 1448-bp fragment of the CD95 promoter and a p53 binding element from the first intron or of pCD95 1448Luc deleted of the intronic sequence. Twenty-four hours posttransfection, cells were starved and treated with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) for additional 24 h before luciferase and h-galactosidase activities were determined. Luciferase activities were normalized to h-galactosidase activities, and results are means F SEM of three independent experiments. (B) HepG2 cells were transiently transfected with 0.5 Ag pI-CD95 1448Luc, 0.25 Ag ph-galControl together with 0.5 Ag of an empty vector (pcDNA3) or of a plasmid encoding a dominant-negative GSK-3h (K85R-GSK-3h). Cells were then treated as reported in (A). Results are means F SEM of four independent experiments. *P b 0.05, compared with cells transfected with the empty vector (pcDNA3).
Up to now, the antiapoptotic action of lithium has been ascribed to its ability to inhibit mitochondria dysfunction in neuronal cells [12,19,34]. Further evidence for a mitochondrial action of lithium was obtained recently by showing that GSK-3h is localized in mitochondria where it is highly active during apoptosis [16–18]. In agreement with these studies, we observed that lithium counteracted etoposide and camptothecin activation of the mitochondria death pathway in HepG2 cells by inhibiting Bax translocation to mitochondria and cytosolic release of cytochrome c (data not shown). Importantly, the present study reveals for the
Fig. 7. Lithium inhibits p53 accumulation and nuclear association of p53 with GSK-3h during etoposide- and camptothecin-induced apoptosis in HepG2 cells. (A) HepG2 cells were treated for 12 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM). Whole-cell extracts were analyzed for p53 expression by Western blot analysis. (B, C) HepG2 cells were treated for 4 h with DMSO, etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM). Nuclear extracts were prepared and analyzed for p53, p53Ser15 phosphorylation, GSK-3hSer9 phosphorylation, and total GSK-3h expression levels by Western blotting. (D) GSK-3h was immunoprecipitated from nuclear extracts, and immunoprecipitates were immunoblotted for phospho-p53Ser15 and GSK-3h expression. Representative immunoblots are shown. Quantitative values are means F SEM from three independent experiments. *P b 0.05; ***P b 0.01 compared with cells treated with etoposide or camptothecin alone.
362
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
expression than from CD95/CD95 ligand interaction, as previously reported in other cell systems [35,36]. Several lines of evidence have been obtained indicating that lithium acts through GSK-3h inhibition for inducing chemoresistance and down-regulation of CD95 expression. Thus, lithium promoted the rapid and sustained phosphorylation of GSK-3h on the serine 9 residue. More importantly, the GSK-3h inhibitor SB-415286 mimicked lithium effects by repressing drug induction of apoptosis and CD95 membrane expression. In addition, overexpression of an inactivated GSK-3h mutant counteracted the stimulatory effects of etoposide and camptothecin on a luciferase reporter plasmid driven by DNA sequences from the CD95 gene. Since activation of this construct is tightly dependent upon the presence of a p53 enhancer, our findings suggest that inactivated GSK-3h exerts a repressive action on p53dependent activation of CD95 gene expression. Recent data obtained in different cell contexts indicate that GSK-3h can act as a positive or negative physiological regulator of the p53 protein. Thus, it has been shown that, during endoplasmic reticulum stress, p53 is inhibited through a mechanism involving its phosphorylation at serine 315 and serine 376 by GSK-3h [37]. On the other hand, it has been reported that treatment of neuroblastoma cells with camptothecin induces GSK-3h binding to p53 in both nucleus and mitchondria promoting responses to p53 which include increases in p21 and Bax expressions as well as mitochondrial apoptotic signaling [18,38]. In the present study, we observed that etoposide and camptothecin induced a physical interaction between GSK-3h and stabilized phospho-p53Ser15 in nucleus and that lithium disrupted GSK-3h/p53 association by promoting p53 destabilization. Taken into consideration with results from transfection experiments, these findings support the notion that p53 and GSK-3h cooperate during etoposide- and camptothecininduced apoptosis in our cell system. This cooperation could promote the transcriptional action of p53 on the CD95 gene during apoptosis. Further studies are needed to elucidate the role of GSK3h inhibition in lithium-induced p53 destabilization. As p53 phosphorylation on multiple NH2-terminal residues is responsible for p53 stabilization during genotoxic treatments [29], it is possible that GSK-3h inhibition represses the signaling pathways involved in p53 phosphorylation. Such an effect could be ascribed to a direct phosphorylating action of GSK-3h on p53 [39] or to an indirect action on p53 kinases such as cyclin-dependent kinases [40,41] or the ATM/ATR family of kinases [42,42–44]. Although our results suggest that GSK-3h cooperates with p53 for inducing CD95 expression during etoposideand camptothecin-induced apoptosis, it is not excluded that GSK-3h interacts with other transcription factors to regulate CD95 expression. Indeed, it has been previously shown that c-Jun cooperates with STAT3 to suppress CD95 transcription [45] and that GSK-3h inactivates c-Jun by phosphorylation [3,4]. Therefore, it is possible that lith-
ium-mediated GSK-3h inhibition alleviates an inhibitory constraint on c-Jun, thus contributing to the repression of CD95 mRNA expression. Our study provides new insights into the contribution of GSK-3hSer9 phosphorylation to chemoresistance. Thus, increased GSK-3hSer9 phosphorylation has been recently reported in different cancers [5–8]. The fact that lithium confers resistance to etoposide- and camptothecin-induced apoptosis in different cancer cell lines by increasing GSK3hSer9 phosphorylation suggests that the acquisition and the maintenance of sustained GSK-3hSer9 phosphorylation by tumor cells confer resistance to anticancer therapy. Sustained GSK-3h phosphorylation could be considered as a new mechanism repressing p53 activation after genotoxic treatment and accounting for chemoresistance in tumors which have retained wild-type p53. In conclusion, our results demonstrate that lithium confers resistance to chemotherapy-induced apoptosis in cancer cells through GSK-3h inhibition, disruption of GSK3h/p53 cooperation, and repression of CD95 expression. Our findings highlight the potential ambivalence of strategies designed to inhibit GSK-3h. Indeed, the development of a new generation of GSK-3h inhibitors is currently under investigation since hyperactivation of GSK-3h seems to be implicated in the pathogenesis of diabetes and devastating neurodegenerative diseases such as Alzheimer disease [46]. While these inhibitors could be beneficial in these disorders, they could be detrimental in a cancerous context by conferring chemoresistance.
Acknowledgments We thank Dr. Azeddine Atfi (INSERM U.482, Paris, France) for providing us with MCF-7 cells, Dr. A. LopezRivas (Instituto de Parasitologı´a y Biomedicina, CSIC, Granada, Spain) and Dr. C. Michiels (Laboratory of Biochemistry and Cellular Biology, University of Namur, Namur, Belgium) for providing us with plasmids, and Philippe Fontange (IFR65, Hoˆ pital Tenon, Paris) for confocal microscopy analysis. E. Beurel is a recipient of the Ministe`re de la Recherche.
References [1] A.J. Harwood, G. Agam, Search for a common mechanism of mood stabilizers, Biochem. Pharmacol. 66 (2003) 179 – 189. [2] R.S. Jope, G.N. Bijur, Mood stabilizers, glycogen synthase kinase3beta and cell survival, Mol. Psychiatry 7 (Suppl. 1) (2002) S35 – S45. [3] C.A. Grimes, R.S. Jope, The multifaceted roles of glycogen synthase kinase 3beta in cellular signaling, Prog. Neurobiol. 65 (2001) 391 – 426. [4] S. Frame, P. Cohen, GSK3 takes centre stage more than 20 years after its discovery, Biochem. J. 359 (2001) 1 – 16. [5] K.C. Ban, H. Singh, R. Krishnan, H.F. Seow, GSK-3beta phosphorylation and alteration of beta-catenin in hepatocellular carcinoma, Cancer Lett. 199 (2003) 201 – 208. [6] C. Desbois-Mouthon, M.J. Blivet-Van Eggelpoel, E. Beurel, M.
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
[22]
[23]
[24]
Boissan, R. Delelo, A. Cadoret, J. Capeau, Dysregulation of glycogen synthase kinase-3beta signaling in hepatocellular carcinoma cells, Hepatology 36 (2002) 1528 – 1536. J. Gotoh, M. Obata, M. Yoshie, S. Kasai, K. Ogawa, Cyclin D1 overexpression correlates with beta-catenin activation, but not with H-ras mutations, phosphorylation of Akt, GSK3 beta and ERK1/2 in mouse hepatic carcinogenesis, Carcinogenesis 24 (2003) 435 – 442. H. Leis, C. Segrelles, S. Ruiz, M. Santos, J.M. Paramio, Expression, localization, activity of glycogen synthase kinase 3beta during mouse skin tumorigenesis, Mol. Carcinog. 35 (2002) 180 – 185. P. De Sarno, X. Li, R.S. Jope, Regulation of Akt and glycogen synthase kinase-3beta phosphorylation by sodium valproate and lithium, Neuropharmacology 43 (2002) 1158 – 1164. L. Song, P. De Sarno, R.S. Jope, Central role of glycogen synthase kinase-3beta in endoplasmic reticulum stress-induced caspase-3 activation, J. Biol. Chem. 277 (2002) 44701 – 44708. F. Zhang, C.J. Phiel, L. Spece, N. Gurvich, P.S. Klein, Inhibitory phosphorylation of glycogen synthase kinase-3 (GSK-3) in response to lithium. Evidence for autoregulation of GSK-3, J. Biol. Chem. 278 (2003) 33067 – 33077. G.N. Bijur, P. De Sarno, R.S. Jope, Glycogen synthase kinase-3beta facilitates staurosporine- and heat shock-induced apoptosis. Protection by lithium, J. Biol. Chem. 275 (2000) 7583 – 7590. R.J. Crowder, R.S. Freeman, Glycogen synthase kinase-3 beta activity is critical for neuronal death caused by inhibiting phosphatidylinositol 3-kinase or Akt but not for death caused by nerve growth factor withdrawal, J. Biol. Chem. 275 (2000) 34266 – 34271. M. Hetman, J.E. Cavanaugh, D. Kimelman, Z. Xia, Role of glycogen synthase kinase-3beta in neuronal apoptosis induced by trophic withdrawal, J. Neurosci. 20 (2000) 2567 – 2574. T.D. King, G.N. Bijur, R.S. Jope, Caspase-3 activation induced by inhibition of mitochondrial complex I is facilitated by glycogen synthase kinase-3beta and attenuated by lithium, Brain Res. 919 (2001) 106 – 114. G.N. Bijur, R.S. Jope, Rapid accumulation of Akt in mitochondria following phosphatidylinositol 3-kinase activation, J. Neurochem. 87 (2003) 1427 – 1435. G.N. Bijur, R.S. Jope, Glycogen synthase kinase-3 beta is highly activated in nuclei and mitochondria, NeuroReport 14 (2003) 2415 – 2419. P. Watcharasit, G.N. Bijur, L. Song, J. Zhu, X. Chen, R.S. Jope, Glycogen synthase kinase-3beta (GSK3beta) binds to and promotes the actions of p53, J. Biol. Chem. 278 (2003) 48872 – 48879. R.W. Chen, D.M. Chuang, Long term lithium treatment suppresses p53 and Bax expression but increases Bcl-2 expression. A prominent role in neuroprotection against excitotoxicity, J. Biol. Chem. 274 (1999) 6039 – 6042. F.H. Igney, P.H. Krammer, Death and anti-death: tumour resistance to apoptosis, Nat. Rev., Cancer 2 (2002) 277 – 288. C. Ruiz-Ruiz, G. Robledo, E. Cano, J.M. Redondo, A. Lopez-Rivas, Characterization of p53-mediated up-regulation of CD95 gene expression upon genotoxic treatment in human breast tumor cells, J. Biol. Chem. 278 (2003) 31667 – 31675. C. Desbois-Mouthon, A. Cadoret, M.J. Blivet-Van Eggelpoel, F. Bertrand, G. Cherqui, C. Perret, J. Capeau, Insulin and IGF-1 stimulate the beta-catenin pathway through two signalling cascades involving GSK-3beta inhibition and Ras activation, Oncogene 20 (2001) 252 – 259. M.P. Coghlan, A.A. Culbert, D.A. Cross, S.L. Corcoran, J.W. Yates, N.J. Pearce, O.L. Rausch, G.J. Murphy, P.S. Carter, L. Roxbee Cox, D. Mills, M.J. Brown, D. Haigh, R.W. Ward, D.G. Smith, K.J. Murray, A.D. Reith, J.C. Holder, Selective small molecule inhibitors of glycogen synthase kinase-3 modulate glycogen metabolism and gene transcription, Chem. Biol. 7 (2000) 793 – 803. I. Petak, J.A. Houghton, Shared pathways: death receptors and cytotoxic drugs in cancer therapy, Pathol. Oncol. Res. 7 (2001) 95 – 106.
363
[25] V. Poulaki, C.S. Mitsiades, N. Mitsiades, The role of Fas and FasL as mediators of anticancer chemotherapy, Drug Resist. Updat. 4 (2001) 233 – 242. [26] M. Muller, S. Strand, H. Hug, E.M. Heinemann, H. Walczak, W.J. Hofmann, W. Stremmel, P.H. Krammer, P.R. Galle, Drug-induced apoptosis in hepatoma cells is mediated by the CD95 (APO-1/Fas) receptor/ligand system and involves activation of wild-type p53, J. Clin. Invest. 99 (1997) 403 – 413. [27] M. Muller, S. Wilder, D. Bannasch, D. Israeli, K. Lehlbach, M. LiWeber, S.L. Friedman, P.R. Galle, W. Stremmel, M. Oren, P.H. Krammer, p53 Activates the CD95 (APO-1/Fas) gene in response to DNA damage by anticancer drugs, J. Exp. Med. 188 (1998) 2033 – 2045. [28] E. Lorenzo, C. Ruiz-Ruiz, A.J. Quesada, G. Hernandez, A. Rodriguez, A. Lopez-Rivas, J.M. Redondo, Doxorubicin induces apoptosis and CD95 gene expression in human primary endothelial cells through a p53-dependent mechanism, J. Biol. Chem. 277 (2002) 10883 – 10892. [29] T. Caspari, How to activate p53, Curr. Biol. 10 (2000) R315 – R317. [30] Y. Cohen, A. Chetrit, P. Sirota, B. Modan, Cancer morbidity in psychiatric patients: influence of lithium carbonate treatment, Med. Oncol. 15 (1998) 32 – 36. [31] T.D. Gould, N.A. Gray, H.K. Manji, Effects of a glycogen synthase kinase-3 inhibitor, lithium, in adenomatous polyposis coli mutant mice, Pharmacol. Res. 48 (2003) 49 – 53. [32] J. Nordenberg, E. Fenig, M. Landau, R. Weizman, A. Weizman, Effects of psychotropic drugs on cell proliferation and differentiation, Biochem. Pharmacol. 58 (1999) 1229 – 1236. [33] R.H. Chen, W.V. Ding, F. McCormick, Wnt signaling to beta-catenin involves two interactive components. Glycogen synthase kinase-3beta inhibition and activation of protein kinase C, J. Biol. Chem. 275 (2000) 17894 – 17899. [34] T.D. King, G.N. Bijur, R.S. Jope, Caspase-3 activation induced by inhibition of mitochondrial complex I is facilitated by glycogen synthase kinase-3beta and attenuated by lithium, Brain Res. 919 (2001) 106 – 114. [35] Y. Aragane, D. Kulms, D. Metze, G. Wilkes, B. Poppelmann, T.A. Luger, T. Schwarz, Ultraviolet light induces apoptosis via direct activation of CD95 (Fas/APO-1) independently of its ligand CD95L, J. Cell Biol. 140 (1998) 171 – 182. [36] O. Micheau, E. Solary, A. Hammann, M.T. Dimanche-Boitrel, Fas ligand-independent, FADD-mediated activation of the Fas death pathway by anticancer drugs, J. Biol. Chem. 274 (1999) 7987 – 7992. [37] L. Qu, S. Huang, D. Baltzis, A.M. Rivas-Estilla, O. Pluquet, M. Hatzoglou, C. Koumenis, Y. Taya, A. Yoshimura, A.E. Koromilas, Endoplasmic reticulum stress induces p53 cytoplasmic localization and prevents p53-dependent apoptosis by a pathway involving glycogen synthase kinase-3beta, Genes Dev. 18 (2004) 261 – 277. [38] P. Watcharasit, G.N. Bijur, J.W. Zmijewski, L. Song, A. Zmijewska, X. Chen, G.V. Johnson, R.S. Jope, Direct, activating interaction between glycogen synthase kinase-3beta and p53 after DNA damage, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 7951 – 7955. [39] G.A. Turenne, B.D. Price, Glycogen synthase kinase3 beta phosphorylates serine 33 of p53 and activates p53’s transcriptional activity, BMC Cell Biol. 2 (2001) 12. [40] Y. Wang, C. Prives, Increased and altered DNA binding of human p53 by S and G2/M but not G1 cyclin-dependent kinases, Nature 376 (1995) 88 – 91. [41] L.J. Ko, S.Y. Shieh, X. Chen, L. Jayaraman, K. Tamai, Y. Taya, C. Prives, Z.Q. Pan, p53 is phosphorylated by CDK7-cyclin H in a p36MAT1-dependent manner, Mol. Cell. Biol. 17 (1997) 7220 – 7229. [42] S. Banin, L. Moyal, S. Shieh, Y. Taya, C.W. Anderson, L. Chessa, N.I. Smorodinsky, C. Prives, Y. Reiss, Y. Shiloh, Y. Ziv, Enhanced phosphorylation of p53 by ATM in response to DNA damage, Science 281 (1998) 1674 – 1677. [43] C.E. Canman, D.S. Lim, K.A. Cimprich, Y. Taya, K. Tamai, K. Sakaguchi, E. Appella, M.B. Kastan, J.D. Siliciano, Activation of the
364
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
ATM kinase by ionizing radiation and phosphorylation of p53, Science 281 (1998) 1677 – 1679. [44] S. Saito, A.A. Goodarzi, Y. Higashimoto, Y. Noda, S.P. Lees-Miller, E. Appella, C.W. Anderson, ATM mediates phosphorylation at multiple p53 sites, including Ser(46), in response to ionizing radiation, J. Biol. Chem. 277 (2002) 12491 – 12494. [45] V.N. Ivanov, M. Krasilnikov, Z. Ronai, Regulation of Fas
expression by STAT3 and c-Jun is mediated by phosphatidylinositol 3-kinase-AKT signaling, J. Biol. Chem. 277 (2002) 4932 – 4944. [46] A. Martinez, A. Castro, I. Dorronsoro, M. Alonso, Glycogen synthase kinase 3 (GSK-3) inhibitors as new promising drugs for diabetes, neurodegeneration, cancer, inflammation, Med. Res. Rev. 22 (2002) 373 – 384.
GSK-3h inhibition by lithium confers resistance to chemotherapy-induced apoptosis through the repression of CD95 (Fas/APO-1) expression Ele´onore Beurela, Michel Kornprobstb, Marie-Jose´ Blivet-Van EggelpoJla, Carmen Ruiz-Ruizc, Axelle Cadoreta, Jacqueline Capeaua, Christe`le Desbois-Mouthona,* a
INSERM U.402, Faculte´ de Me´decine Saint-Antoine, Paris, France b IFR65, Faculte´ de Me´decine Saint-Antoine, Paris, France c Department of Biochemistry and Molecular Biology, University of Granada, Granada, Spain Received 1 June 2004, revised version received 2 August 2004 Available online 3 September 2004
Abstract Lithium exerts neuroprotective actions that involve the inhibition of glycogen synthase kinase-3h (GSK-3h). Otherwise, recent studies suggest that sustained GSK-3h inhibition is a hallmark of tumorigenesis. In this context, the present study was undertaken to examine whether lithium modulated cancer cell sensitivity to apoptosis induced by chemotherapy agents. We observed that, in different human cancer cell lines, lithium significantly reduced etoposide- and camptothecin-induced apoptosis. In HepG2 cells, lithium repressed drug induction of CD95 expression and clustering at the cell surface as well as caspase-8 activation. Lithium acted through deregulation of GSK-3h signaling since (1) it provoked a rapid and sustained phosphorylation of GSK-3h on the inhibitory serine 9 residue; (2) the GSK-3h inhibitor SB415286 mimicked lithium effects by repressing drug-induced apoptosis and CD95 membrane expression; and (3) lithium promoted the disruption of nuclear GSK-3h/p53 complexes. Moreover, the overexpression of an inactivated GSK-3h mutant counteracted the stimulatory effects of etoposide and camptothecin on a luciferase reporter plasmid driven by a p53-responsive sequence from the CD95 gene. In conclusion, we provide the first evidence that lithium confers resistance to apoptosis in cancer cells through GSK-3h inhibition and subsequent repression of CD95 gene expression. Our study also highlights the concerted action of GSK-3h and p53 on CD95 gene expression. D 2004 Elsevier Inc. All rights reserved. Keywords: Lithium; Apoptosis; GSK-3h; p53; Chemotherapy; CD95; Caspase-8
Introduction Lithium is the most widely used treatment for bipolar mood disorder. Recent evidence shows that lithium can protect neuronal cells from apoptosis induced by a wide array of neurotoxic insults [1,2]. The neuroprotective properties of lithium could be an important component of its therapeutic action. Analysis of the molecular mechanisms underlying lithium-induced neuroprotection has revealed that lithium inhibits the activity of glycogen synthase * Corresponding author. INSERM U.402, Faculte´ de Me´decine SaintAntoine, 27 rue Chaligny, 75571 Paris Cedex 12, France. Fax: +33 1 40 01 13 52. E-mail address: [email protected] (C. Desbois-Mouthon). 0014-4827/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2004.08.001
kinase-3h (GSK-3h), and based on this finding, lithium is now commonly used as a potent inhibitor of GSK-3h [1]. GSK-3h is a serine/threonine kinase active in resting cells and implicated in metabolism, proliferation, and gene transcription as a result of its ability to phosphorylate and inactivate key proteins governing these processes such as glycogen synthase, cyclin D1, and h-catenin [3,4]. GSK-3h undergoes rapid and transient inhibition through sitespecific phosphorylation on the serine 9 residue in response to diverse stimuli including insulin and insulin-like growth factors [3,4]. Recent studies suggest that sustained GSK3hSer9 phosphorylation could contribute to tumorigenesis. Thus, we and others have reported that GSK-3h is hyperphosphorylated on serine 9 in human hepatoma cell lines as well as in human and murine tumoral livers [5–7]. In
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
addition, chemically induced mouse skin carcinogenesis has been associated with GSK-3hSer9 hyperphosphorylation [8]. Interestingly, it has been reported that lithium increases GSK-3hSer9 phosphorylation [9–11], thus showing that lithium inhibits GSK-3h activity through a mechanism similar to that observed in cancer cells. Evidence obtained in neuronal cells has revealed that GSK-3h is a proapoptotic kinase. Thus, overexpression or hyperactivation of GSK-3h sensitizes neuronal cells to apoptosis [10,12–15]. Recently, studies by Jope et al. [16– 18] have shown that GSK-3h is located in mitochondria where it is highly active during apoptosis. The functional outcome of this mitochondrial interaction is unclear, but it might facilitate apoptosis since lithium-mediated GSK-3h inhibition attenuates cytochrome c release from the mitochondria [18,19]. Because GSK-3h is an important mediator of apoptotic signals, it is plausible that GSK-3h deregulation observed in cancer cells confers resistance to chemotherapy-induced apoptosis which is a major cause of treatment failure in human cancers [20]. In this context, we were interested to investigate whether lithium modulated chemotherapyinduced apoptosis in cancer cells. We observed that lithium repressed apoptosis induced by etoposide and camptothecin in different cancer cell lines. The molecular mechanisms responsible for chemoresistance were then examined, focusing the analyses on the CD95 (Fas/APO-1) death receptor signaling pathway.
Materials and methods Cell culture HepG2 (hepatoblastoma) cells were maintained in minimal essential medium (MEM) containing Earle salts, 1% nonessential amino acids, 1 mM sodium pyruvate, and 10% fetal calf serum. A549 (lung carcinoma) and MCF-7 (breast carcinoma) cell lines were cultured in Dulbecco MEM supplemented with 10% fetal calf serum. Cell lines were from the American Type Culture Collection. Before experiments, cells were incubated overnight in serum-free medium. In some assays, cells were treated with dimethylsulfoxide (DMSO), etoposide, camptothecin, lithium chloride (SigmaAldrich Co.), a general caspase inhibitor (Z-VAD-fmk), a caspase-8 inhibitor (Z-IETD-fmk, R and D Systems, Inc.), or with a GSK-3h small inhibitor (SB- 415286, Tocris). DNA fragmentation Both adherent and floating cells were collected and lysed in 1% Nonidet P-40, 20 mM EDTA, and 50 mM Tris-HCl (pH 7.5). After centrifugation, supernatants were incubated with 1% SDS and 5 mg/ml ribonuclease A for 3 h at 568C. Then, 2.5 mg/ml proteinase K was added overnight at 378C. DNAs were precipitated at 208C in the presence of 10 M
355
ammonium acetate and ethanol, resuspended in water, and analyzed by electrophoresis on 1% agarose gels. Flow cytometry analysis of apoptosis Both adherent and floating cells were collected, washed, fixed in ice-cold 70% ethanol at 208C, and stained with 20 Ag/ml propidium iodide in the presence of 100 Ag/ml ribonuclease A for 30 min at 378C in the dark. DNA content was analyzed by flow cytometry (FACS Calibur, Becton Dickinson). Apoptotic cells with hypodiploid DNA staining were found in the Tsub-G1r peak. Confocal laser microscopy HepG2 cells cultured on 12-mm glass coverslips were fixed in 4% paraformaldehyde (Sigma-Aldrich Co.) for 10 min and incubated for 2 h at room temperature with an antiCD95 antibody (ZB4, MBL International Co.) diluted 1:100 in PBS containing 1% bovine serum albumin. Samples were then incubated for 45 min with a secondary antibody (Alexa FluorR 488 conjugate, Molecular Probes Inc., 1:100 dilution). Confocal microscopy was performed using a Leica TCS SP microscope equipped with a 63 objective (Lasertechnik GmbH). Reverse transcription–polymerase chain reaction (RT–PCR) CD95 and CD95 ligand mRNAs were analyzed by RT–PCR using the following primers: CD95, 5VGCGAAAGCCCATTTTTCTTCC-3V and 5V-ATTTATTGCCACTGTTTCAGG-3V; CD95 ligand, 5V-CAG C T C T T C C A C C T A C A G A A G G A - 3 V a n d 5 VGAGAGCTCAGATACGTTGAC-3V. For semiquantitative analysis, concurrent h-actin PCR products (primers 5VAT C AT G T T T G A G A C C T T C A A - 3 V a n d 5 V- T T G CGCTCAGGAGGAGCAAT-3V) were generated in parallel to ensure that equivalent amounts of template were amplified. For CD95 quantitation, real-time PCR was performed with the LightCycler system (Roche Molecular Biochemicals) using the SYBR green fluorophore. Each sample was normalized on the basis of its endogenous content in 18S rRNA (primers: 5 V-GAGCGAAAGCATTTGCCAAG3Vand 5V-GGCATCGTT-TATGGTCGGAA-3V). Flow cytometry detection of CD95 Both adherent and floating cells were collected and fixed in 4% paraformaldehyde (Sigma-Aldrich Co.) for 10 min and incubated for 1.5 h at room temperature with an antiCD95 antibody (ZB4) diluted 1:300 in PBS containing 1% bovine serum albumin. Samples were then incubated for 45 min with a secondary antibody (Alexa FluorR 488 conjugate, 1:500 dilution). Unstained cells and cells stained with the secondary antibody alone were run in parallel as
356
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
negative controls. After staining, cells were resuspended in PBS and analyzed by flow cytometry. Transient transfections Cells were plated at a density of 3 105 cells in sixwell plates the day before transfection. Cells were transfected using Transfast (Promega Corp.) with 0.5 Ag of the luciferase reporter plasmid pI-CD95 1448Luc or pCD95 1448Luc (generous gifts from Dr. A. LopezRivas) [21] and 0.25 Ag of ph-gal-Control (BD Biosciences Clontech) together with 0.5 Ag of either an empty expression vector (pcDNA3, InVitrogen) or of an expression vector encoding an inactive K85R-GSK-3h (generous gift from Dr. C. Michiels). Luciferase and hgalactosidase activities were evaluated as previously reported [22], and luciferase activities were normalized to h-galactosidase activities. Immunoprecipitation and Western blotting Whole-cell and cytosolic extracts were prepared as previously reported [22]. Lysates for caspase expression were prepared as recommended by Cell Signaling Technology Inc. Nuclear extracts were prepared from about 1 106 cells by using NE-PERk nuclear and cytoplasmic kit (Pierce). For immunoprecipitation, nuclear extracts were incubated overnight at 48C with an antiphospho-Ser15 p53 (Cell Signaling Technology, Inc.) or an anti-GSK-3h (Transduction Laboratories) antibody at a 1:200 dilution together with 20 Al of protein A/G PLUSagarose (Santa Cruz Biotechnology, Inc.). For Western blot analysis, nitrocellulose membranes were probed with the following antibodies: anti-phospho-Ser9 GSK-3h, anticaspase-3, anti-caspase-8, anti-phospho-Ser15 p53 (Cell Signaling Technology, Inc.), anti-poly(ADP-ribose) polymerase (Santa Cruz Biotechnology, Inc.), anti-p53 (clone BP53-12, Upstate Biotechnology), and anti-GSK-3h. Membranes were reprobed with an anti-h-actin antibody (Sigma-Aldrich Co.) to ensure equivalent loading. Caspase-3 and caspase-8 assays Cells were pelleted, washed with PBS, lysed in 50 mM Tris-HCl (pH 7.5), 0.03% Nonidet P-40, and 1 mM dithiotreitol, and centrifuged. Assays were set up in 96well plates containing supernatants (100 Ag proteins), 0.2 mM of caspase-3 substrate (Ac-DEVD-pNA) or of caspase-8 substrate (Ac-IETD-pNA) (BIOMOL Research Laboratories, Inc.) in a caspase reaction buffer (100 mM HEPES pH 7.5, 10% sucrose, 0.1% CHAPS, 10 mM DTT). Assays were performed at 378C, and release of pNa was detected by periodic readings of absorbance at 405 nm from 0 to 5 h to mark the linearity of the enzymatic reaction in time. Enzyme activities were measured as initial velocities.
Statistical analysis Results are given as the mean F SEM. Statistical comparison of mean values was performed using the Student t test.
Results Lithium confers resistance to etoposide- and camptothecin-induced apoptosis in human cancer cells We first examined the effect of lithium on the sensitivity of hepatoma HepG2 cells to inhibitors of topoisomerases I (camptothecin) and II (etoposide). To this aim, cells were treated for 48 h with solvent alone (dimethylsulfoxide), etoposide (85 AM), or camptothecin (0.5 AM) in the presence or absence of lithium (20 mM) before being analyzed for DNA fragmentation by gel electrophoresis and for the fraction of hypodiploid cells by flow cytometry. We used lithium at a 20-mM concentration since this concentration has been previously reported to potently inhibit GSK-3h in vitro [9–11]. Etoposide and camptothecin induced apoptosis in HepG2 cells since (i) oligonucleosomal DNA fragmentation typical of apoptosis was observed in cells treated with each of these drugs (Fig. 1A); (ii) etoposide and camptothecin increased the number of sub-G1 cells from 5.9% F 0.4% to 16.8% F 1.9% and to 44.8% F 3.2%, respectively (Fig. 1B); and (iii) the general caspase inhibitor Z-VAD-fmk totally inhibited the effects of etoposide and camptothecin (Fig. 1B, inset). We observed that lithium treatment strongly reduced oligonucleosomal DNA fragmentation (Fig. 1A) as well as the number of apoptotic cells induced by each drug (Fig. 1B, 7.7% F 1.4% and 19.7% F 2.6% in the presence of etoposide and camptothecin, respectively), indicating that lithium conferred resistance to chemotherapy-induced apoptosis in HepG2 cells. To examine whether lithium interfered with etoposideand camptothecin-induced apoptosis in other cancer cell types, flow cytometry analyses of sub-G1 cells were also conducted in A549 (lung carcinoma) and MCF-7 (breast carcinoma) cells. Again, we observed that lithium significantly reduced the number of apoptotic cells induced by etoposide and camptothecin in the two lines studied (Fig. 2). The maximal inhibitory effects of lithium were observed with the HepG2 cell line and, thereby, the molecular mechanisms whereby lithium conferred resistance to chemotherapy drugs were analyzed in these cells. Lithium inhibits GSK-3b during etoposide- and camptothecin-induced apoptosis in HepG2 cells We next examined the impact of lithium on GSK-3h activity in HepG2 cells by measuring the phosphorylation
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
357
Lithium prevents activation of the CD95 death pathway during etoposide- and camptothecin-induced apoptosis in HepG2 cells
Fig. 1. Lithium confers resistance to etoposide- and camptothecin-induced apoptosis in HepG2 cells. HepG2 cells were treated for 48 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium chloride (LiCl, 20 mM) and analyzed for DNA fragmentation by electrophoresis (A) and for hypodiploid cell (sub-G1) contents by flow cytometry (B). Results are means F SEM of six independent experiments. Inset, HepG2 cells were treated for 48 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without a general inhibitor of caspases (Z-VAD-fmk, 20 AM), and hypodiploid cells were analyzed by flow cytometry. Results are representative of two independent experiments. ***P b 0.01 compared with cells treated with drug alone.
level of GSK-3h at serine 9 as an indicator of GSK-3h inactivation. We observed that lithium rapidly increased GSK-3h phosphorylation on the serine 9 regulatory site. GSK-3hSer9 phosphorylation occurred within 5 min of lithium treatment and was maintained for at least 48 h, without any modification in the total GSK-3h level (Fig. 3A). In HepG2 cells, etoposide or camptothecin altered neither the basal level of phosphorylated GSK-3h nor that induced by lithium (Fig. 3B). Of note, lithium also increased GSK-3hSer9 phosphorylation in A549 and MCF-7 cells (data not shown). To confirm the contribution of GSK-3h inhibition to chemoresistance, HepG2 cells were incubated with or without SB-415286, a small inhibitor of GSK-3h which acts as an ATP competitive inhibitor [23]. We observed that 30 AM SB-415286 reduced by 48% and 38% apoptosis induced by a 12-h treatment with etoposide and camptothecin, respectively, thus supporting a role for GSK-3h inhibition in lithium-induced resistance to chemotherapy (Fig. 3C).
Etoposide and camptothecin induce apoptosis in cancer cells via the CD95 death pathway through the enhancement of CD95 ligand and/or CD95 expression [24, 25]. Activation of the CD95 signaling pathway results from CD95 receptor trimerization and clustering at the cell surface, which in turns leads to the recruitment of the adaptor protein FADD and caspase-8 and the subsequent activation of downstream caspases. Experiments were designed to examine whether lithium conferred resistance to etoposide- and camptothecin-induced apoptosis through the modulation of the CD95 signaling pathway. We first used confocal laser microscopy to analyze the effects of anticancer drugs and lithium on CD95 aggregation at the cell surface. While control cells (DMSO) had diffuse staining of CD95, cells treated with etoposide or camptothecin exhibited CD95 clustering as evidenced by a dense staining that was primarily membranelocalized (Fig. 4A). In the presence of lithium, drug effects were diminished. Induction of CD95 ligand mRNA was not observed in etoposide- and camptothecin-treated HepG2 cells (Fig. 4B). However, etoposide and camptothecin increased CD95 mRNA expression in HepG2 cells evaluated by semiquantitative (Fig. 4C, left) and quantitative real-time RT–PCR (Fig. 4C, right) as well as CD95 expression at the cell surface evaluated by flow cytometry (Fig. 4D, left). Importantly, we observed that lithium prevented etoposide- and camptothecininduced up-regulation of CD95 mRNA (Fig. 4C) and protein expression at the cell surface (Fig. 4D). Similarly, SB-415286 repressed the induction of CD95 expression at the cell surface by etoposide (Fig. 4E) and camptothecin (data not shown), thus supporting a role for GSK3h inhibition in the repressive effect of lithium on CD95 expression. Upon CD95 activation, caspase-8 is the main initiation caspase. In accordance with these data, we observed that etoposide and camptothecin induced caspase-8 activation in HepG2 cells measured by the appearance of cleaved active caspase-8 bands and by a colorimetric assay (Fig. 5A). Moreover, etoposide- and camptothecin-induced apoptosis was inhibited by the caspase-8 inhibitor ZIETD-fmk (data not shown). Cell treatment with lithium prevented etoposide and camptothecin induction of caspase-8 activity (Fig. 5A). Lithium also repressed drug activation of the executioner caspase-3 measured by the appearance of cleaved active caspase-3 bands, by a colorimetric assay, and by the proteolysis of the caspase3 substrate PARP (Fig. 5B). Altogether, these data show that lithium interferes with etoposide and camptothecin induction of the CD95 signaling pathway by repressing CD95 expression (mRNA and protein) and clustering as well as caspase-8 activation.
358
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
Fig. 2. Lithium confers resistance to etoposide- and camptothecin-induced apoptosis in human cancer cell lines. A549 and MCF-7 cells were treated for 48 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM for A549 cells and 0.125 AM for MCF-7 cells) with or without lithium chloride (LiCl, 20 mM), and hypodiploid cells were analyzed by flow cytometry. Results are means F SEM of three independent experiments. *P b 0.05, **P b 0.02 compared with cells treated with drug alone.
GSK-3b inhibition blocks p53-dependent transcriptional activation from the CD95 gene promoter in HepG2 cells
Fig. 3. Lithium inhibits GSK-3h during etoposide- and camptothecininduced apoptosis in HepG2 cells. Whole-cell extracts from HepG2 cells were analyzed by Western blotting for GSK-3hSer9 phosphorylation and total GSK-3h. (A) Cells were treated with or without lithium (LiCl, 20 mM) for various incubation times. (B) Cells were treated for 24 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM) or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM). Blots are representative of three independent experiments. (C) Cells were treated with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without SB-415286 (30 AM) for 12 h. Hypodiploid cells were quantified by flow cytometry. Results are means F SEM of two independent experiments. *P b 0.05, compared with cells treated with drug alone.
It has been clearly demonstrated that, in cancer cells, upregulation of CD95 gene expression after genotoxic treatment is mediated through p53 activation [21,26–28]. We then examined the impact of GSK-3h inhibition on the transcriptional regulation of the CD95 gene. To this end, transient transfection experiments were performed using the plasmid pICD95 1448Luc in which the expression of a luciferase reporter gene is driven by regulatory DNA elements from the human CD95 gene [21]. These regulatory sequences include 1448 bp of the 5V-promoter and 500 bp of the first intron containing a binding element for p53. It has been previously shown that this p53 consensus site is absolutely required for induction of pICD95 1448Luc expression upon chemotherapy treatment [21]. In HepG2 cells, we observed that etoposide and camptothecin stimulated luciferase expression from the pICD95 1448Luc construct, while both drugs did not increase luciferase expression in the absence of the intronic region (pCD95 1448Luc construct) (Fig. 6A). When cells were transfected with a plasmid encoding a dominant-negative GSK-3h (K85R-GSK-3h), we observed that the stimulatory effects of drugs on pI-CD95 1448Luc were significantly decreased as compared to cells transfected with an empty vector (pcDNA3) (Fig. 6B). These findings show that GSK-3h inhibition prevents the stimulation of p53-dependent pI-CD95 1448Luc expression by chemotherapy drugs in HepG2 cells. Lithium inhibits p53 accumulation and nuclear association of p53 with GSK-3b during etoposide- and camptothecin-induced apoptosis in HepG2 cells The above results prompted us to analyze the effect of lithium on GSK-3h and p53 expression and interaction
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
359
Fig. 4. Lithium prevents activation of the CD95 death pathway during etoposide- and camptothecin-induced apoptosis in HepG2 cells. HepG2 cells were treated with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM) or SB-415286 (SB, 30 AM). (A) Cells treated for 12 h were fixed and labeled for confocal laser microscopy using the antihuman CD95 antibody ZB4. (B) After a 12-h incubation, CD95 ligand expression was analyzed by semiquantitative RT–PCR. mRNA obtained from HepG2 cells treated with methotrexate (Metho, 100 Ag/ml) was used as a positive control. (C) After a 12-h treatment, CD95 expression was analyzed by semiquantitative (left) and quantitative real-time (right) RT–PCR. (D) Cells treated for 24 h were collected and labeled for flow cytometry analysis using the antihuman CD95 antibody ZB4. Grey histogram, nonspecific; black histogram, DMSO-treated; green histogram, ETP-treated; blue histogram, CPT-treated; histograms with dotted lines, cotreatment with lithium. (E) Cells treated for 24 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without SB-415286 (30 AM) were collected and labeled for flow cytometry analysis using the antihuman CD95 antibody ZB4. Grey histogram, nonspecific; black histogram, DMSO-treated; green histogram, ETP-treated; histograms with dotted lines, cotreatment with SB-415286. Results are representative of three independent experiments.
360
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
Fig. 5. Lithium prevents activation of caspase-8 and caspase-3 during etoposide- and camptothecin-induced apoptosis in HepG2 cells. HepG2 cells were treated with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM). (A) After a 4-h incubation, cell lysates were prepared and analyzed for caspase-8 activation by Western blotting using an antibody that allows detection of full-length (p57/55) and cleaved (p43/41) caspase-8 (left) or by a colorimetric assay using a caspase-8-specific peptide conjugated to a chromogene (Ac-IETD-pNA) (right). (B) After a 24-h incubation, cell lysates were prepared and analyzed for caspase-3 activation by Western blotting using antibodies that allow detection of cleaved (p19/17) caspase-3 or of cleaved (p85) poly(ADP-ribose) polymerase (PARP) (left) or by a colorimetric assay using a caspase-3-specific peptide conjugated to a chromogene (Ac-DEVD-pNA) (right). Blots and data are representative of three independent experiments.
during etoposide and camptothecin induction of apoptosis. In HepG2 cells, we observed that etoposide and camptothecin increased total (Fig. 7A) and nuclear (Fig. 7B) levels of p53. In the presence of lithium, p53 accumulation induced by etoposide and camptothecin was markedly reduced (Figs. 7A,B). Phosphorylation of NH2-terminal residues of p53 mediates its stabilization and nuclear accumulation after anticancer drug treatment [29]. Consistent with this notion, we observed that etoposide and camptothecin stimulated p53Ser15 phosphorylation in nuclear extracts (Fig. 7C). Again, in the presence of lithium, drug induction of p53Ser15 phosphorylation was markedly reduced. Etoposide and camptothecin promoted p53 nuclear accumulation without modifying phosphorylated and total levels of GSK-3h, while cotreatment with lithium reduced p53 accumulation and increased GSK-3hSer9 phosphorylation but not GSK-3h expression (Fig. 7B). Examination of a direct association of p53 with GSK-3h was performed by coimmunoprecipitation experiments. Immunoblots revealed that etoposide and camptothecin increased the coimmunoprecipitation of phospho-p53Ser15 with GSK-3h in nuclear fractions from HepG2 cells (Fig. 7D). In the presence of lithium, drug induction of nuclear phospho-p53Ser15/GSK-3h complexes was clearly diminished. Taken together, these findings show that lithium prevents p53 accumulation induced by etoposide or camptothecin treatment which subsequently reduces the formation of nuclear GSK-3h/ p53 complexes.
Discussion Lithium provides effective pharmacotherapy for bipolar mood disorder. Although it is not carcinogenic by itself [30], it is not excluded that, in some cellular context, lithium could favor the propensity for developing certain cancers. For example, in a murine model genetically predisposed to intestinal adenomas, lithium treatment significantly increases tumor number in colon and overall tumor size [31]. In addition, lithium has been shown to stimulate proliferation in some cancer cell types [32]. Moreover, we and others have reported that lithium increases nuclear levels of h-catenin [6,11,23,33], a protein which associates with LEF/TCF transcription factors to stimulate expression of proliferative genes. In this report, we now provide evidence that, in human cancer cells, lithium confers resistance to apoptosis induced by two different chemotherapy drugs. These findings were obtained with 20 mM lithium which is a concentration usually used in in vitro studies to maximize the observable changes. In patients treated for bipolar disorders, the plasma concentration range of lithium is 0.5–1.0 mM. Long-term treatment of mice with such doses causes a large increase in the level of GSK-3hSer9 phosphorylation in brain [9] and in liver (E. Beurel, personal observation). Taken together, these findings could suggest that short-term in vitro experiments conducted with a high dose of lithium mimic cellular responses obtained in vivo after the chronic administration of a low dose of lithium.
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
361
first time that lithium also potently inhibits etoposide and camptothecin stimulation of the CD95 signaling pathway. We observed that lithium repressed drug induction of CD95 expression and clustering at the cell surface as well as caspase-8 activation. Of note, neither etoposide nor camptothecin induced CD95 ligand mRNA expression in HepG2 cells. This suggests that drug stimulation of CD95 clustering most probably resulted from up-regulation of CD95
Fig. 6. GSK-3h inhibition blocks p53-dependent transcriptional activation from the CD95 gene promoter in HepG2 cells. (A) HepG2 cells were transiently transfected with 0.25 Ag ph-gal-Control together with 0.5 Ag of pI-CD95 1448Luc containing a 1448-bp fragment of the CD95 promoter and a p53 binding element from the first intron or of pCD95 1448Luc deleted of the intronic sequence. Twenty-four hours posttransfection, cells were starved and treated with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) for additional 24 h before luciferase and h-galactosidase activities were determined. Luciferase activities were normalized to h-galactosidase activities, and results are means F SEM of three independent experiments. (B) HepG2 cells were transiently transfected with 0.5 Ag pI-CD95 1448Luc, 0.25 Ag ph-galControl together with 0.5 Ag of an empty vector (pcDNA3) or of a plasmid encoding a dominant-negative GSK-3h (K85R-GSK-3h). Cells were then treated as reported in (A). Results are means F SEM of four independent experiments. *P b 0.05, compared with cells transfected with the empty vector (pcDNA3).
Up to now, the antiapoptotic action of lithium has been ascribed to its ability to inhibit mitochondria dysfunction in neuronal cells [12,19,34]. Further evidence for a mitochondrial action of lithium was obtained recently by showing that GSK-3h is localized in mitochondria where it is highly active during apoptosis [16–18]. In agreement with these studies, we observed that lithium counteracted etoposide and camptothecin activation of the mitochondria death pathway in HepG2 cells by inhibiting Bax translocation to mitochondria and cytosolic release of cytochrome c (data not shown). Importantly, the present study reveals for the
Fig. 7. Lithium inhibits p53 accumulation and nuclear association of p53 with GSK-3h during etoposide- and camptothecin-induced apoptosis in HepG2 cells. (A) HepG2 cells were treated for 12 h with dimethylsulfoxide (DMSO), etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM). Whole-cell extracts were analyzed for p53 expression by Western blot analysis. (B, C) HepG2 cells were treated for 4 h with DMSO, etoposide (ETP, 85 AM), or camptothecin (CPT, 0.5 AM) with or without lithium (LiCl, 20 mM). Nuclear extracts were prepared and analyzed for p53, p53Ser15 phosphorylation, GSK-3hSer9 phosphorylation, and total GSK-3h expression levels by Western blotting. (D) GSK-3h was immunoprecipitated from nuclear extracts, and immunoprecipitates were immunoblotted for phospho-p53Ser15 and GSK-3h expression. Representative immunoblots are shown. Quantitative values are means F SEM from three independent experiments. *P b 0.05; ***P b 0.01 compared with cells treated with etoposide or camptothecin alone.
362
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
expression than from CD95/CD95 ligand interaction, as previously reported in other cell systems [35,36]. Several lines of evidence have been obtained indicating that lithium acts through GSK-3h inhibition for inducing chemoresistance and down-regulation of CD95 expression. Thus, lithium promoted the rapid and sustained phosphorylation of GSK-3h on the serine 9 residue. More importantly, the GSK-3h inhibitor SB-415286 mimicked lithium effects by repressing drug induction of apoptosis and CD95 membrane expression. In addition, overexpression of an inactivated GSK-3h mutant counteracted the stimulatory effects of etoposide and camptothecin on a luciferase reporter plasmid driven by DNA sequences from the CD95 gene. Since activation of this construct is tightly dependent upon the presence of a p53 enhancer, our findings suggest that inactivated GSK-3h exerts a repressive action on p53dependent activation of CD95 gene expression. Recent data obtained in different cell contexts indicate that GSK-3h can act as a positive or negative physiological regulator of the p53 protein. Thus, it has been shown that, during endoplasmic reticulum stress, p53 is inhibited through a mechanism involving its phosphorylation at serine 315 and serine 376 by GSK-3h [37]. On the other hand, it has been reported that treatment of neuroblastoma cells with camptothecin induces GSK-3h binding to p53 in both nucleus and mitchondria promoting responses to p53 which include increases in p21 and Bax expressions as well as mitochondrial apoptotic signaling [18,38]. In the present study, we observed that etoposide and camptothecin induced a physical interaction between GSK-3h and stabilized phospho-p53Ser15 in nucleus and that lithium disrupted GSK-3h/p53 association by promoting p53 destabilization. Taken into consideration with results from transfection experiments, these findings support the notion that p53 and GSK-3h cooperate during etoposide- and camptothecininduced apoptosis in our cell system. This cooperation could promote the transcriptional action of p53 on the CD95 gene during apoptosis. Further studies are needed to elucidate the role of GSK3h inhibition in lithium-induced p53 destabilization. As p53 phosphorylation on multiple NH2-terminal residues is responsible for p53 stabilization during genotoxic treatments [29], it is possible that GSK-3h inhibition represses the signaling pathways involved in p53 phosphorylation. Such an effect could be ascribed to a direct phosphorylating action of GSK-3h on p53 [39] or to an indirect action on p53 kinases such as cyclin-dependent kinases [40,41] or the ATM/ATR family of kinases [42,42–44]. Although our results suggest that GSK-3h cooperates with p53 for inducing CD95 expression during etoposideand camptothecin-induced apoptosis, it is not excluded that GSK-3h interacts with other transcription factors to regulate CD95 expression. Indeed, it has been previously shown that c-Jun cooperates with STAT3 to suppress CD95 transcription [45] and that GSK-3h inactivates c-Jun by phosphorylation [3,4]. Therefore, it is possible that lith-
ium-mediated GSK-3h inhibition alleviates an inhibitory constraint on c-Jun, thus contributing to the repression of CD95 mRNA expression. Our study provides new insights into the contribution of GSK-3hSer9 phosphorylation to chemoresistance. Thus, increased GSK-3hSer9 phosphorylation has been recently reported in different cancers [5–8]. The fact that lithium confers resistance to etoposide- and camptothecin-induced apoptosis in different cancer cell lines by increasing GSK3hSer9 phosphorylation suggests that the acquisition and the maintenance of sustained GSK-3hSer9 phosphorylation by tumor cells confer resistance to anticancer therapy. Sustained GSK-3h phosphorylation could be considered as a new mechanism repressing p53 activation after genotoxic treatment and accounting for chemoresistance in tumors which have retained wild-type p53. In conclusion, our results demonstrate that lithium confers resistance to chemotherapy-induced apoptosis in cancer cells through GSK-3h inhibition, disruption of GSK3h/p53 cooperation, and repression of CD95 expression. Our findings highlight the potential ambivalence of strategies designed to inhibit GSK-3h. Indeed, the development of a new generation of GSK-3h inhibitors is currently under investigation since hyperactivation of GSK-3h seems to be implicated in the pathogenesis of diabetes and devastating neurodegenerative diseases such as Alzheimer disease [46]. While these inhibitors could be beneficial in these disorders, they could be detrimental in a cancerous context by conferring chemoresistance.
Acknowledgments We thank Dr. Azeddine Atfi (INSERM U.482, Paris, France) for providing us with MCF-7 cells, Dr. A. LopezRivas (Instituto de Parasitologı´a y Biomedicina, CSIC, Granada, Spain) and Dr. C. Michiels (Laboratory of Biochemistry and Cellular Biology, University of Namur, Namur, Belgium) for providing us with plasmids, and Philippe Fontange (IFR65, Hoˆ pital Tenon, Paris) for confocal microscopy analysis. E. Beurel is a recipient of the Ministe`re de la Recherche.
References [1] A.J. Harwood, G. Agam, Search for a common mechanism of mood stabilizers, Biochem. Pharmacol. 66 (2003) 179 – 189. [2] R.S. Jope, G.N. Bijur, Mood stabilizers, glycogen synthase kinase3beta and cell survival, Mol. Psychiatry 7 (Suppl. 1) (2002) S35 – S45. [3] C.A. Grimes, R.S. Jope, The multifaceted roles of glycogen synthase kinase 3beta in cellular signaling, Prog. Neurobiol. 65 (2001) 391 – 426. [4] S. Frame, P. Cohen, GSK3 takes centre stage more than 20 years after its discovery, Biochem. J. 359 (2001) 1 – 16. [5] K.C. Ban, H. Singh, R. Krishnan, H.F. Seow, GSK-3beta phosphorylation and alteration of beta-catenin in hepatocellular carcinoma, Cancer Lett. 199 (2003) 201 – 208. [6] C. Desbois-Mouthon, M.J. Blivet-Van Eggelpoel, E. Beurel, M.
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20] [21]
[22]
[23]
[24]
Boissan, R. Delelo, A. Cadoret, J. Capeau, Dysregulation of glycogen synthase kinase-3beta signaling in hepatocellular carcinoma cells, Hepatology 36 (2002) 1528 – 1536. J. Gotoh, M. Obata, M. Yoshie, S. Kasai, K. Ogawa, Cyclin D1 overexpression correlates with beta-catenin activation, but not with H-ras mutations, phosphorylation of Akt, GSK3 beta and ERK1/2 in mouse hepatic carcinogenesis, Carcinogenesis 24 (2003) 435 – 442. H. Leis, C. Segrelles, S. Ruiz, M. Santos, J.M. Paramio, Expression, localization, activity of glycogen synthase kinase 3beta during mouse skin tumorigenesis, Mol. Carcinog. 35 (2002) 180 – 185. P. De Sarno, X. Li, R.S. Jope, Regulation of Akt and glycogen synthase kinase-3beta phosphorylation by sodium valproate and lithium, Neuropharmacology 43 (2002) 1158 – 1164. L. Song, P. De Sarno, R.S. Jope, Central role of glycogen synthase kinase-3beta in endoplasmic reticulum stress-induced caspase-3 activation, J. Biol. Chem. 277 (2002) 44701 – 44708. F. Zhang, C.J. Phiel, L. Spece, N. Gurvich, P.S. Klein, Inhibitory phosphorylation of glycogen synthase kinase-3 (GSK-3) in response to lithium. Evidence for autoregulation of GSK-3, J. Biol. Chem. 278 (2003) 33067 – 33077. G.N. Bijur, P. De Sarno, R.S. Jope, Glycogen synthase kinase-3beta facilitates staurosporine- and heat shock-induced apoptosis. Protection by lithium, J. Biol. Chem. 275 (2000) 7583 – 7590. R.J. Crowder, R.S. Freeman, Glycogen synthase kinase-3 beta activity is critical for neuronal death caused by inhibiting phosphatidylinositol 3-kinase or Akt but not for death caused by nerve growth factor withdrawal, J. Biol. Chem. 275 (2000) 34266 – 34271. M. Hetman, J.E. Cavanaugh, D. Kimelman, Z. Xia, Role of glycogen synthase kinase-3beta in neuronal apoptosis induced by trophic withdrawal, J. Neurosci. 20 (2000) 2567 – 2574. T.D. King, G.N. Bijur, R.S. Jope, Caspase-3 activation induced by inhibition of mitochondrial complex I is facilitated by glycogen synthase kinase-3beta and attenuated by lithium, Brain Res. 919 (2001) 106 – 114. G.N. Bijur, R.S. Jope, Rapid accumulation of Akt in mitochondria following phosphatidylinositol 3-kinase activation, J. Neurochem. 87 (2003) 1427 – 1435. G.N. Bijur, R.S. Jope, Glycogen synthase kinase-3 beta is highly activated in nuclei and mitochondria, NeuroReport 14 (2003) 2415 – 2419. P. Watcharasit, G.N. Bijur, L. Song, J. Zhu, X. Chen, R.S. Jope, Glycogen synthase kinase-3beta (GSK3beta) binds to and promotes the actions of p53, J. Biol. Chem. 278 (2003) 48872 – 48879. R.W. Chen, D.M. Chuang, Long term lithium treatment suppresses p53 and Bax expression but increases Bcl-2 expression. A prominent role in neuroprotection against excitotoxicity, J. Biol. Chem. 274 (1999) 6039 – 6042. F.H. Igney, P.H. Krammer, Death and anti-death: tumour resistance to apoptosis, Nat. Rev., Cancer 2 (2002) 277 – 288. C. Ruiz-Ruiz, G. Robledo, E. Cano, J.M. Redondo, A. Lopez-Rivas, Characterization of p53-mediated up-regulation of CD95 gene expression upon genotoxic treatment in human breast tumor cells, J. Biol. Chem. 278 (2003) 31667 – 31675. C. Desbois-Mouthon, A. Cadoret, M.J. Blivet-Van Eggelpoel, F. Bertrand, G. Cherqui, C. Perret, J. Capeau, Insulin and IGF-1 stimulate the beta-catenin pathway through two signalling cascades involving GSK-3beta inhibition and Ras activation, Oncogene 20 (2001) 252 – 259. M.P. Coghlan, A.A. Culbert, D.A. Cross, S.L. Corcoran, J.W. Yates, N.J. Pearce, O.L. Rausch, G.J. Murphy, P.S. Carter, L. Roxbee Cox, D. Mills, M.J. Brown, D. Haigh, R.W. Ward, D.G. Smith, K.J. Murray, A.D. Reith, J.C. Holder, Selective small molecule inhibitors of glycogen synthase kinase-3 modulate glycogen metabolism and gene transcription, Chem. Biol. 7 (2000) 793 – 803. I. Petak, J.A. Houghton, Shared pathways: death receptors and cytotoxic drugs in cancer therapy, Pathol. Oncol. Res. 7 (2001) 95 – 106.
363
[25] V. Poulaki, C.S. Mitsiades, N. Mitsiades, The role of Fas and FasL as mediators of anticancer chemotherapy, Drug Resist. Updat. 4 (2001) 233 – 242. [26] M. Muller, S. Strand, H. Hug, E.M. Heinemann, H. Walczak, W.J. Hofmann, W. Stremmel, P.H. Krammer, P.R. Galle, Drug-induced apoptosis in hepatoma cells is mediated by the CD95 (APO-1/Fas) receptor/ligand system and involves activation of wild-type p53, J. Clin. Invest. 99 (1997) 403 – 413. [27] M. Muller, S. Wilder, D. Bannasch, D. Israeli, K. Lehlbach, M. LiWeber, S.L. Friedman, P.R. Galle, W. Stremmel, M. Oren, P.H. Krammer, p53 Activates the CD95 (APO-1/Fas) gene in response to DNA damage by anticancer drugs, J. Exp. Med. 188 (1998) 2033 – 2045. [28] E. Lorenzo, C. Ruiz-Ruiz, A.J. Quesada, G. Hernandez, A. Rodriguez, A. Lopez-Rivas, J.M. Redondo, Doxorubicin induces apoptosis and CD95 gene expression in human primary endothelial cells through a p53-dependent mechanism, J. Biol. Chem. 277 (2002) 10883 – 10892. [29] T. Caspari, How to activate p53, Curr. Biol. 10 (2000) R315 – R317. [30] Y. Cohen, A. Chetrit, P. Sirota, B. Modan, Cancer morbidity in psychiatric patients: influence of lithium carbonate treatment, Med. Oncol. 15 (1998) 32 – 36. [31] T.D. Gould, N.A. Gray, H.K. Manji, Effects of a glycogen synthase kinase-3 inhibitor, lithium, in adenomatous polyposis coli mutant mice, Pharmacol. Res. 48 (2003) 49 – 53. [32] J. Nordenberg, E. Fenig, M. Landau, R. Weizman, A. Weizman, Effects of psychotropic drugs on cell proliferation and differentiation, Biochem. Pharmacol. 58 (1999) 1229 – 1236. [33] R.H. Chen, W.V. Ding, F. McCormick, Wnt signaling to beta-catenin involves two interactive components. Glycogen synthase kinase-3beta inhibition and activation of protein kinase C, J. Biol. Chem. 275 (2000) 17894 – 17899. [34] T.D. King, G.N. Bijur, R.S. Jope, Caspase-3 activation induced by inhibition of mitochondrial complex I is facilitated by glycogen synthase kinase-3beta and attenuated by lithium, Brain Res. 919 (2001) 106 – 114. [35] Y. Aragane, D. Kulms, D. Metze, G. Wilkes, B. Poppelmann, T.A. Luger, T. Schwarz, Ultraviolet light induces apoptosis via direct activation of CD95 (Fas/APO-1) independently of its ligand CD95L, J. Cell Biol. 140 (1998) 171 – 182. [36] O. Micheau, E. Solary, A. Hammann, M.T. Dimanche-Boitrel, Fas ligand-independent, FADD-mediated activation of the Fas death pathway by anticancer drugs, J. Biol. Chem. 274 (1999) 7987 – 7992. [37] L. Qu, S. Huang, D. Baltzis, A.M. Rivas-Estilla, O. Pluquet, M. Hatzoglou, C. Koumenis, Y. Taya, A. Yoshimura, A.E. Koromilas, Endoplasmic reticulum stress induces p53 cytoplasmic localization and prevents p53-dependent apoptosis by a pathway involving glycogen synthase kinase-3beta, Genes Dev. 18 (2004) 261 – 277. [38] P. Watcharasit, G.N. Bijur, J.W. Zmijewski, L. Song, A. Zmijewska, X. Chen, G.V. Johnson, R.S. Jope, Direct, activating interaction between glycogen synthase kinase-3beta and p53 after DNA damage, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 7951 – 7955. [39] G.A. Turenne, B.D. Price, Glycogen synthase kinase3 beta phosphorylates serine 33 of p53 and activates p53’s transcriptional activity, BMC Cell Biol. 2 (2001) 12. [40] Y. Wang, C. Prives, Increased and altered DNA binding of human p53 by S and G2/M but not G1 cyclin-dependent kinases, Nature 376 (1995) 88 – 91. [41] L.J. Ko, S.Y. Shieh, X. Chen, L. Jayaraman, K. Tamai, Y. Taya, C. Prives, Z.Q. Pan, p53 is phosphorylated by CDK7-cyclin H in a p36MAT1-dependent manner, Mol. Cell. Biol. 17 (1997) 7220 – 7229. [42] S. Banin, L. Moyal, S. Shieh, Y. Taya, C.W. Anderson, L. Chessa, N.I. Smorodinsky, C. Prives, Y. Reiss, Y. Shiloh, Y. Ziv, Enhanced phosphorylation of p53 by ATM in response to DNA damage, Science 281 (1998) 1674 – 1677. [43] C.E. Canman, D.S. Lim, K.A. Cimprich, Y. Taya, K. Tamai, K. Sakaguchi, E. Appella, M.B. Kastan, J.D. Siliciano, Activation of the
364
E. Beurel et al. / Experimental Cell Research 300 (2004) 354–364
ATM kinase by ionizing radiation and phosphorylation of p53, Science 281 (1998) 1677 – 1679. [44] S. Saito, A.A. Goodarzi, Y. Higashimoto, Y. Noda, S.P. Lees-Miller, E. Appella, C.W. Anderson, ATM mediates phosphorylation at multiple p53 sites, including Ser(46), in response to ionizing radiation, J. Biol. Chem. 277 (2002) 12491 – 12494. [45] V.N. Ivanov, M. Krasilnikov, Z. Ronai, Regulation of Fas
expression by STAT3 and c-Jun is mediated by phosphatidylinositol 3-kinase-AKT signaling, J. Biol. Chem. 277 (2002) 4932 – 4944. [46] A. Martinez, A. Castro, I. Dorronsoro, M. Alonso, Glycogen synthase kinase 3 (GSK-3) inhibitors as new promising drugs for diabetes, neurodegeneration, cancer, inflammation, Med. Res. Rev. 22 (2002) 373 – 384.