Gallium compound GaQ₃-induced Ca²⁺ signalling triggers p53-dependent and -independent apoptosis in cancer cells

Abstract

BACKGROUND AND PURPOSE

A novel anti-neoplastic gallium complex GaQ₃ (KP46), earlier developed by us, is currently in phase I clinical trial. GaQ₃ induced S-phase arrest and apoptosis via caspase/PARP cleavage in a variety of cancers. However, the underlying mechanism of apoptosis is unknown. Here, we have explored the mechanism(s) of GaQ₃-induced apoptosis in cancer cells, focusing on p53 and intracellular Ca²⁺ signalling.

EXPERIMENTAL APPROACH

GaQ₃-induced cytotoxicity and apoptosis were determined in cancer cell lines, with different p53 status (p53^+/+, p53^−/− and p53 mutant). Time course analysis of intracellular Ca²⁺ calcium release, p53 promoter activation, p53-nuclear/cytoplasmic movements and reactive oxygen species (ROS) were conducted. Ca²⁺-dependent formation of the p53–p300 transcriptional complex was analysed by co-immunoprecipitation and chromatin immunoprecipitation. Ca²⁺ signalling, p53, p300 and ROS were serially knocked down to study Ca²⁺–p53–ROS ineractions in GaQ₃-induced apoptosis.

KEY RESULTS

GaQ₃ triggered intracellular Ca²⁺ release stabilizing p53–p300 complex and recruited p53 to p53 promoter, leading to p53 mRNA and protein synthesis. p53 induced higher intracellular Ca²⁺ release and ROS followed by activation of p53 downstream genes including those for the micro RNA mir34a. In p53^−/− and p53 mutant cells, GaQ₃-induced Ca²⁺-signalling generated ROS. ROS further increased membrane translocation of FAS and FAS-mediated extrinsic apoptosis.

CONCLUSIONS AND IMPLICATIONS

This study disclosed a novel mechanism of Ca²⁺-signalling-mediated p53 activation and ROS up-regulation. Understanding the mechanism of GaQ₃-induced apoptosis will help establish this gallium-based organic compound as a potent anti-cancer drug.

Introduction

A range of gallium compounds have been investigated in clinical trials against a number of malignancies including lymphomas and bladder cancer (Einhorn, 2003; Straus, 2003; Chitambar, 2004; Jakupec and Keppler, 2004). Gallium nitrate has shown significant anti-tumour activity against non-Hodgkin's lymphoma and bladder cancer (Chitambar et al., 2007), and induced apoptosis in mantle cell lymphoma cells through increased Bax and reactive oxygen species (ROS), along with down-regulation of cyclin D1 (Joseph et al., 2005). Gallium maltolate, a novel hydroxy-pyrone gallium compound, induced apoptosis and circumvented lymphoma cell resistance to gallium nitrate (Chitambar et al., 2007). Gallium (III) complexes that act as a potent proteasome inhibitors have great potential to be developed into novel anti-cancer drugs (Chen et al., 2007). The gallium (III) complexes of 2-pyridineformamide thio-semicarbazones have been shown to induce cytotoxicity in both malignant glioblastoma cells RT2 (p53^+/+) and T98 cells (p53 mutant) (Mendes et al., 2009).

Recently, we have successfully assessed the therapeutic potential of an orally bio-available gallium complex, tris(8-quinolinolato)gallium(III) (KP46; here referred to as GaQ₃) against cancer cells in phase I clinical trials (Valiahdi et al., 2009). The effects of GaQ₃ on cancer cells involve S-phase arrest of the cell cycle and apoptosis via activation of caspase-9, cleavage of PARP protein and formation of apoptotic bodies (Valiahdi et al., 2009). However, the mechanism of GaQ₃-induced cell death is unknown. Interestingly, the activity profile of GaQ₃ is distinct from that of inorganic gallium salts, as cells exposed to gallium nitrate displayed increased generation of ROS (Joseph et al., 2005). Inorganic gallium salts interfere with physiological and intracellular procurement of iron, thus causing its depletion (Chitambar et al., 2000). These compounds also interact with iron-dependent proteins like ribonucleotide reductase, prolyl-hydroxylase, transferrin and are effective by inhibiting DNA synthesis and transferrin-mediated endocytosis, ultimately resulting in apoptosis (Zanias et al., 2010).

The widely used anti-cancer drug tamoxifen enhances Ca²⁺ signalling in both breast cancer and several primary glioma cells; increasing both the radius of calcium wave propagation by local stimulation and the amplitude of agonist-induced calcium elevations, while retarding the normalization of cytosolic calcium (Zhang et al., 2000). Ca²⁺–S100B-mediated regulation of p53 transcriptional activity also provides a link between Ca²⁺ signalling and oncogenic processes in which the tumour suppressor protein p53 plays key roles (Ikura and Yap, 2000).; This was confirmed by the fact that Ca²⁺-activated RAF/MEK/ERK signalling pathway mediated p53-dependent apoptosis (Li et al., 2005a). Furthermore, the Ca²⁺ binding protein S100A2 binds to the p53 C-terminus in cancer cells, and this interaction was Ca²⁺-dependent and altered p53 transcriptional activity (Mueller et al., 2005). Patients with p53 deletion exhibit higher serum calcium levels (Chang et al., 2005).

By contrast, the chemotherapeutic drug bleomycin induced transient nuclear accumulation of wild-type p53 and (APO-1/FAS) receptor in HepG2 cells (Muller et al., 1997). A direct correlation between wild-type p53 activity and FAS up-regulation after treatment with ionizing radiation was demonstrated, thus suggesting that post-irradiation FAS up-regulation was dependent on activity of wild-type p53 (Sheard et al., 1997). Regulation of FAS (CD95)-induced apoptotic and necrotic cell death has also been linked to ROS in macrophages (Medan et al., 2005). On the other hand, p53 in mitochondria binds to and inhibits manganese-superoxide dismutase (Mn-SOD), thus playing a direct role in ROS generation and apoptosis (Zhao et al., 2005). Each cellular concentration and distribution of p53 has a distinct cellular function, and ROS acts as both an upstream signal that triggers p53 activation and as a downstream factor that mediates apoptosis (Liu et al., 2008). A microarray analysis of H₂O₂-treated cells identified one-third of the 48 highly H₂O₂-responsive genes as targets of p53 (Desaint et al., 2004).

In this study, we have established the mechanism of action of a novel anti-cancer compound GaQ₃. We have shown that GaQ₃ induced both p53-dependent intrinsic and p53-independent extrinsic pathways of apoptosis in cancer cell. GaQ₃ initiated intracellular release of Ca²⁺ ions in p53^+/+, p53 mutant and p53^−/− cells. In p53^+/+ cells, intracellular Ca²⁺ stabilized the p53–p300 transcriptional complex followed by p53 transactivation in order to trigger p53-dependent apoptosis, whereas in p53 null and p53 mutant cells, Ca²⁺ signalling increased ROS, which induced FAS membrane translocation and FAS-dependent extrinsic apoptosis.

Methods

Cell culture

Four cancer cell lines (MCF-7, HepG2, PC3, H1299), normal keratinocytes and peripheral blood mononuclear cells (PBMC) were obtained from the National Centre for Cell Science (Pune, India) and were maintained in DMEM. The cells were cultured as monolayers in DMEM supplemented with 10% (v/v) heat-inactivated fetal bovine serum and antibiotics, and incubated at 37°C in a humidified atmosphere of 95% air and 5% CO₂ All the transfections were carried out using effectene transfection reagent (Qiagen, Gaithersburg, MD, USA) according to manufacturer's instructions.

Plasmids and reagents

p53 siRNA, miR-34a and miR-34b/c promoter luciferase constructs were used as previously described byLu et al. (2006). p300 siRNA SC-29431-V (Santa Cruz Biotechnology, Santa Cruz, CA, USA) was used for p300 gene silencing. p53 2.5 KB promoter luciferase constructs in pGL2 vector were used to study p53 full promoter activity. The ROS scavenger N-acetyl cysteine (NAC) and the Ca²⁺ quencher TMB8 were purchased from Sigma Aldrich Co. (St Louis, MO, USA) FAS siRNA SR300253 (Origene, Rockville, MD, USA) was used for Fas gene silencing.

Time course analysis

For the various time course studies, five Petri dishes for each time point were used.

Cell proliferation and toxicity assays

Cell proliferation was assayed by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay performed according to the manufacturer's protocol. The MTT cell proliferation assay kit (catalogue no. 10009365) was obtained from Cayman Chemicals (Ann Arbor, MI, USA). The LDH assay was performed with the LDH cell proliferation assay kit (catalogue no. 10008882; Cayman Chemicals), according to the manufacturer's protocol. DNA synthesis was assayed by the immunofluorescence assay for the detection of 5-bromo-2′-deoxyuridine (BrdU) incorporated into cellular DNA, as described previously (Garcia-Berrocal et al., 2007), and using the In situ cell proliferation kit fluos (Roche Applied Science, Indianapolis, IN, USA).

Senescence-associated β-galactosidase (SA-β-Gal) staining

MCF-7 and H1299 cells were stained for SA-β-Gal activity using the senescence detection kit (Cell Signaling Technology, Irvine, CA, USA). Briefly, cells were washed with PBS, fixed for 15 min at room temperature, washed again with PBS and treated overnight at 37°C in SA-β-Gal staining reagent (1 mg·mL⁻¹ of X-Gal). Cells were then washed with PBS and images were taken at 200× magnification, with phase-contrast microscopy.

Cell cycle analysis

Cells were collected by centrifugation at 500×g at 4°C for 5 min. Cell pellet was re-suspended in 1 mL of cold PBS. Cells were then fixed by adding 4 mL of cold absolute ethanol and stored at −20°C in this fixation buffer until ready for analysis. Fixed cells were then centrifuged (500×g, 5 min) and the pellet was re-suspended in 1 mL of PBS. 100 µL of 200 µg·mL⁻¹ DNase-free, RNaseA was added to cells and incubated at 37°C for 30 min. Then 100 µL of 1 mg·mL⁻¹ propidium iodide (PI) was added and incubated at room temperature for 5–10 min. Finally, the samples were transferred to 12 × 75 Falcon tubes and read on Becton Dickinson (Franklin Lakes, NJ, USA) (LSR) flow cytometer.

Annexin V staining

Becton Dickinson flow cytometer was used to detect the apoptotic cell surface shift of phosphatidylserine by the binding of fluorescein isothiocyanate (FITC)-conjugated Annexin V to the outer membrane of intact cells. Floating cells were collected by centrifugation (500×g, 5 min) and these, as well as the attached cells, were washed in PBS (Ca²⁺- or Mg²⁺-free)/0.1% EDTA. The attached cells were gently scraped off the dish and centrifuged (500×g, 5 min). The pelleted cells were treated with 500 µL binding buffer (Abcam, Cambridge, MA, USA) then 5 µL of Annexin V–FITC, and 5 µL PI (10 µg mL⁻¹) was added. Cells were incubated at room temperature for 5 min. These cells were then filtered (70 µm mesh), to eliminate cell aggregates, and were analysed by flow cytometry.

DNA fragmentation assay

Treated cells (10⁶) were lysed in 50 µL of NTE buffer (100 mM NaCl, 40 mM Tris–Cl, pH 7.4, 20 mM EDTA) containing 0.5% SDS. The lysate was heated at 65°C for 10 min to inactivate nucleases and digested by a 2 h treatment with 0.5 mg·mL⁻¹ proteinase K followed by a 2 h treatment with 0.2 mg·mL⁻¹ RNase A at 50°C. The DNA fragmentation was analysed on a 1.8% agarose gel in the presence of 0.5 µg·mL⁻¹ ethidium bromide.

Terminal deoxynucleotidyl transferase (TdT)-mediated dUTP nick-end labelling (TUNEL) assay

TUNEL assay was performed using apop-tag fluorescein in situ apoptosis detection kit (Invitrogen, Carlsbad, CA, USA). Briefly, the cell lines were grown on glass-bottomed dishes and treated with GaQ₃ as described above. Cells were first washed in equilibration buffer, treated with TdT enzyme in a humidified chamber at 37°C for 1 h; cells were then washed and treated (room temperature, 30 min) in the dark with fluorescein-conjugated anti-digoxigenin. The washed specimens were counterstained with4′-6-diamidino-2-phenylindole DAPI; 1 µg·mL⁻¹) and visualized with Zeiss Axio vision fluorescent microscope.

Assay for Ca²⁺mobilization

Ca²⁺ was measured using the cell permeable Ca²⁺-sensitive fluorescent dye Fluo-3 acetoxymethyl ester (Kowaltowski et al., 1998). Where indicated, BAPTA acetoxymethyl ester (10 µM) was added to the culture medium of cells in 10 cm plastic tissue culture plates for a 1 h exposure prior to the loading procedure with Fluo-3 acetoxymethyl ester. The medium was removed from the tissue culture plates and replaced with 4 µM Fluo-3 acetoxymethyl ester diluted in Krebs–Ringer buffer (KRB) [10 mM d-glucose, 120 mM NaCl, 4.5 mM KCl, 0.7 mM Na₂HPO₄, 1.5 mM NaH₂PO₄ and 0.5 mM MgCl₂ (pH 7.4 at 37°C)] (Sigma) for 20 min. The dishes were washed once with 5 mL KRB to remove the residual dye. The cells were harvested by trypsinization, washed in 5 mL of Ca²⁺-free PBS at 37°C, pelleted by centrifugation (500×g, 5 min), re-suspended in 1 mL of Ca²⁺-free PBS at 37°C and analysed immediately for Fluo-3 fluorescence intensity by flow cytometry.

Luciferase assay

Cells were plated in 35 mm Petri dishes the day before transfection so that they reached 60–80% confluence upon transfection. Reporter plasmids (1.0–1.5 mg·mL⁻¹) were transfected with effectene transfection reagent (Qiagen) according to the manufacturer's instructions. After desired treatment period, the cells were washed in cold PBS three times and lysed with 200 mL of the lysis buffer by a freeze–thaw cycle, and lysates were collected by centrifugation at 3700×g for 2 min in a bench top centrifuge. Twenty microlitres of supernatant was used for the assay of luciferase activity using a kit (Promega, Madison, WI, USA) according to the manufacturer's instruction. Luciferase activity of 2.5 kb p53 promoter was assayed using hp53-luc plasmid.

Real-time PCR

Real-time PCR was performed using the 7500 fast real-time system (Applied Biosystems, Alameda, CA, USA) using TaqMan probe (Applied Biosystems) (Lu et al., 2006). Glyceraldehyde-3-phosphate dehydrogenase served as an endogenous control to normalize expression. Each sample was analysed in quadruplicate. Relative expression and standard error were calculated by the supplied fast 7500 real-time system software.

Reverse transcriptase PCR

Cells were lysed in appropriate amount of Trizol (1 mL Trizol per well of a six-well plate for cultured cells). Cells were repeatedly and vigorously pipetted. Cells were then kept at room temperature for 5–10 min, and then 200 µL of chloroform per 1 mL of Trizol was added and mixed thoroughly. The cells were again left at room temperature for 10 min. Cells were then centrifuged at 3700×g at 4°C for 15 min, and the upper aqueous colourless layer was transferred to a fresh Eppendorf tube. To this Eppendorf tube, 75 µL LiCl (lithium chloride) followed by 1 mL chilled EtOH (ethanol) were added and kept at −20°C for 2–3 h. The Eppendorf tube was centrifuged at 3700×g for 15 min at 4°C. The supernatant was discarded, and 250 µL of 70% EtOH was added, and the tube was kept at room temperature for 2 min. The tube was again centrifuged at 2500×g for 5 min at 4°C. Finally, the supernatant was discarded, and the pellet was re-suspended in RNA grade water till it was completely dissolved. Single-strand c-DNA was synthesized for treatment with sense and anti-sense primers using revert aid TM h minus first strand cDNA synthesis (Fermentas, Amherst, NY, USA). The resulting cDNA was diluted 1:10 before proceeding with the PCR reaction. PCR was conducted in Mastercycler gradient (Brinkmann Instruments, Inc., Westbury, NY, USA). Each PCR reaction used 50 µL cDNA, 2.5 U Taq polymerase (Eppendorf Scientific, Inc., Ocala, FL, USA), 0.2 mM dNTPs and 0.5 µM primer. PCR products were resolved on 2% agarose gel containing 0.01% (v/v) ethidium bromide and visualized by u.v. illuminator. The size of the PCR amplicon was determined by comparison with 100 bp DNA ladder (Promega). The RT-PCR primers were as follows: β-actin, forward, 5′–ATGAAGTGTGACGTTGACATCCG-3′; reverse, 5′–GCTTGCTGATCCACATCTGCTG-3′; p53, forward,5 CTGGCCCCTGTCATCTTCTGTC-3′; reverse, 5′–CACGCAAATTTCCTTCCACTCG-3′: FAS, forward, 5′-TTGGTGGACCCGCTCAGTA-3′; reverse, 5′-AATCTAGCAACAGACGTAAGAACCAG-3′.

Micro RNA (miRNA) measurements using real-time PCR: miRNA expression analysis

TaqMan miRNA assays were used to quantify levels of mature miRNAs, as described previously (Lu et al., 2006). Primary miR-34 transcript levels were determined with the superscript III SYBR green one-step qRT-PCR system (Invitrogen).

Western blotting

Whole-cell lysates were prepared using RIPA buffer [10 mM Tris–HCl (pH 7.4), 150 mM NaCl, 1% NP-40, 1 mM EDTA, 0.1% SDS and 1 mM DTT]. Proteins were resolved by 10% or 12% SDS-PAGE and transferred onto PVDF membranes (Invitrogen). Incubations with primary antibodies were followed by incubations with the appropriate secondary antibodies. To quantify Western blotting signals, densitometry was performed using ImageJ software (Bethesda, MD, USA).

Immunoprecipitation

Cells were washed with PBS, harvested, lysed in radioimmune precipitation buffer (50 mM Tris–HCl, pH 7.4, 150 mM NaCl, 1.0% Triton X-100, 0.1% SDS and 1.0% sodium deoxycholate). The whole cell lysate was pre-cleared, and 1.0 µg of polyclonal antibodies was used for immunoprecipitation. After a 2 h incubation with antibodies, 10 µL of protein A-agarose was added to the lysate and further incubated at 4°C for 4 h. Washing was done twice with Nonidet P-40 buffer (20 mM Tris–HCl, pH 7.4, 100 mM NaCl, 10% glycerol, 1.0% Nonidet P-40 and 1 mM EDTA) and once with radioimmune precipitation buffer. Immunocomplexes were released by the addition of SDS loading dye, boiled for 5 min, centrifuged and loaded to 10% SDS-PAGE. Transfer of proteins to nitrocellulose membrane was done and immunoblotted with target antibodies.

p53 elisa kit

Pathscan total p53 sandwich elisa kit was purchased from Cell Signaling Technology. The endogenous p53 level was measured in MCF-7 and HepG2 cells, at different times as indicated. Cells were collected and washed twice with PBS. The cells were lysed and processed for detection of p53 according to the manufacturer's protocol.

DNA damage assay

GaQ₃-induced DNA-damage was quantified with the ‘DNA Damage Quantification Kit’ (JM-K253-25, MBL International Corporation, Woburn, MA, USA) according to the instruction manual.

ROS measurement

ROS levels were determined by incubating the cells in PBS containing 10 µM 2′,7′ di-chloro-dihydro-fluorescein-diacetate (H-DCFDA; Molecular Probes, Eugene, OR, USA) for 30 min at 37°C. DCFDA was metabolized by non-specific esterases to the non-fluorescent product, 2′,7′-di-chloro-dihydro-fluorescein, which was oxidized to the fluorescent product, DCF, by ROS. The cells were washed twice in PBS, trypsinized, re-suspended in PBS and their ROS content measured, using a Becton Dickinson flow cytometer.

Preparation of nuclear and cytoplasmic extracts

Cells were collected with a rubber policeman and pelleted by centrifugation (500×g, 5 min). 200 µL of cytoplasmic extraction reagent (CER-I) according to manufacturer's instructions (#78833; NE-PER nuclear and cytoplasmic extraction kit, Pierce inc.) was added per 20 µL of packed cell volume, and the cell pellet was vortexed for 15 s. Cells were treated in presence of CER I for 10 min, followed by treatment with 11 µL of CER II and were further treated for 1 min. Lysed cells were centrifuged at 4100×g for 5 min to pellet the intact nuclei. The supernatant containing the cytoplasmic fraction was carefully separated. The pelleted nuclei were re-suspended in 100 µL of NER I, vortexed for 15 s and treated on ice for 45 min with periodic vortexing after every 10 min. At this point, the suspension was centrifuged (4100×g, 5 min) and the supernatant carrying the nuclear proteins was stored at −80°C.

Protein elisa

The wells of PVC microtitre plate were coated with polyclonal antibodies at a concentration of 1–10 µg·mL⁻¹ in carbonate/bicarbonate buffer (pH 7.4). The plate was covered with an adhesive plastic and incubated overnight at 4°C. The coating solution was removed, and the plate was washed twice by filling the wells with 200 µL PBS. The solutions were removed by flicking the plate over a sink. The remaining protein-binding sites in the coated wells were blocked by adding 200 µL blocking buffer, 5% non-fat dry milk/PBS, per well. Again, the plate was covered and treated overnight at 4°C. 100 µL of appropriately diluted cell suspension was added to each well and treated for 90 min at 37°C. Diluted monoclonal detection antibody (0.5 µg·in 100 µL) was added to each well. The plate was covered with an adhesive plastic and incubated for 2 h at room temperature. The plates were washed four times with PBS. 100 µL (containing 0.01 µg) of secondary antibody conjugated with horseradish peroxidase was added to the plate. The plate was covered with an adhesive plastic and treated for 1–2 h at room temperature. The plate was washed four times with PBS.

Chromatin immunoprecipitation

Formaldehyde was added at a final concentration of 1% directly to cell culture media. Fixation proceeded at 22°C for 10 min and was stopped by the addition of glycine to a final concentration of 0.125 M. The cells were collected by centrifugation (500×g, 5 min) and rinsed in cold PBS. The cell pellets were re-suspended in swelling buffer [10 mM potassium acetate, 15 mM magnesium acetate, 0.1 M Tris (pH 7.6), 0.5 mM PMSF and 100 ng of leupeptin and aprotinin/mL], incubated on ice for 20 min, and then homogenized (Dounce homogenizer). The nuclei were collected by micro-centrifugation (5000×g, 30 min) and then re-suspended in sonication buffer [1% SDS, 10 mM EDTA, 50 mM Tris–HCl (pH 8.1), 0.5 mM PMSF and 100 ng of leupeptin and aprotinin/mL] and incubated on ice for 10 min. Before sonication, 0.1 g of glass beads (212 to 300 µm diameter; Sigma) was added to each sample. The samples were sonicated on ice with an ultrasonic sonicator at setting 10 for six 20 s pulses to an average length of approximately 1000 bp and then micro-centrifuged (5000×g, 45 min). The chromatin solution was pre-cleared with the addition of Staphylococcus aureus protein A-positive cells for 15 min at 4°C. Prior to use, the Staph A cells were blocked with sheared herring sperm DNA (1 µg µL⁻¹) and BSA (1 µg µL⁻¹) for at least 4 h at 4°C. Pre-cleared chromatin from 10⁷ cells was incubated with 1 µg of affinity-purified rabbit polyclonal antibody or no antibody and rotated at 4°C for approximately 12 to 16 h. Before the first wash, 20% of the supernatant from the reaction with no primary antibody for each time point was saved as total input chromatin and was processed with the eluted immunoprecipitates beginning at the cross-link reversal step. Cross-links were reversed by the addition of NaCl to a final concentration of 200 mM, and RNA was removed by the addition of 10 µg of RNase A per sample followed by incubation at 65°C for 4 to 5 h. The samples were then precipitated at −20°C overnight by the addition of 2.5 volumes of ethanol and then pelleted by micro-centrifugation (5000×g, 15 min). The samples were re-suspended in 100 µL of Tris-EDTA (pH 7.5), 25 µL of 5 × proteinase K buffer [1.25% SDS, 50 mM Tris (pH 7.5) and 25 mM EDTA], and 1.5 µL of proteinase K (Boehringer Mannheim, Indianapolis, IN, USA) and incubated at 45°C for 2 h. Samples were extracted with phenol–chloroform–isoamyl alcohol (25:24:1) followed by extraction with chloroform–isoamyl alcohol and then precipitated with 1/10 volume of 3 M sodium acetate (pH 5.3), 5 µg of glycogen and 2.5 volumes of ethanol. The pellets were collected by micro-centrifugation (5000×g, 10 min), re-suspended in 30 µL of H₂O and analysed by PCR.

Data analysis

Data are shown as means ± SD and differences between group means were assessed with anova. Values of P < 0.05 were taken to show a significant difference.

Results

GaQ₃ causes cytotoxicity and cellular senescence in p53^+/+ and p53^−/− cancer cells

GaQ₃ has been previously shown to have anti-proliferative activity in a variety of human cancer cells (Valiahdi et al., 2009). In this study, the role of the p53 tumour suppressor protein in GaQ₃-mediated cytotoxicity and cellular senescence was analysed. The IC₅₀ values of GaQ₃ were determined in three types of caner cells; those with normal, wild-type, p53 (p53^+/+) i.e., MCF-7 and HepG2 cells), those with deletion of p53 (p53^−/−) i.e., H1299 cells and those with mutant forms of p53 (p53 mutant) i.e., PC3 cells (Figure 1A), using MTT and LDH assays (Figures S1–S6). The IC₅₀ values of GaQ₃ for MCF-7, HepG2, H1299 and PC3 cells were 4.1, 4.2, 5.5 and 5.9 µM respectively. A significant difference in the IC₅₀ values of GaQ₃₃was observed in p53^+/+, p53^−/− and p53 mutant cells (Figure 1A). In non-cancer, control cells, such as normal keratinocytes and normal PBMCs, the IC₅₀ was 18 and 16 µM GaQ₃ respectively (Figure 1A). These data show that GaQ₃ is selectively toxic towards cancer cells since the IC₅₀ values of GaQ₃ for normal cells were higher than those in the cancer cell lines. To confirm the role of p53 in GaQ₃-induced cytotoxicity, two closely related cell lines, HCT-p53^+/+ and HCT-p53^−/− cells, were used and IC₅₀ values of 4.4 µM and 7.78 µM, respectively, were obtained.

Figure 1.

GaQ₃ induces cytotoxicity in p53^+/+ and p53^−/− cells. (A) IC₅₀ values of GaQ₃-treated normal keratinocytes and PBMCs are calculated based on MTT and LDH assays. IC₅₀ values of GaQ₃ in normal keratinocytes and PBMCs were higher than those in MCF-7, HepG2, PC3 and H1299 cells. In the middle graph, the IC₅₀ values are lower for HepG2 cells and p53^+/+ (MCF-7) cells than in p53^−/− (H1299) cells and p53 mutant (PC3) cells. In the last graph the IC₅₀ for GaQ₃ in HCT-p53^+/+ cells was lower than that in HCT-p53^−/− cells. Data shown are means ± SD (n= 10), *P < 0.05, significant difference between p53^+/+ and p53^−/− or p53 mutant cells; anova. (B) Effect of GaQ₃ on cellular DNA synthesis determined by BrdU assay in MCF-7, H1299 and PC3 cells. Immunofluorescence microscopy was used to visualize BrdU incorporation in control (second panel) and GaQ₃-treated cells (third panel). DAPI was used to stain the nucleus (n= 7). GaQ₃ was used for 24 h, at the IC₅₀ corresponding to each cell line. (C) p53^+/+ (MCF-7), p53^−/− (H1299) and p53 mutant (PC3) cells were treated with GaQ₃ at corresponding IC₅₀, for 24 h and assayed for cellular senescence using SA-β Gal assay (10× and 40×, phase contrast; n= 5).

The BrdU assay was used to analyse the effect of GaQ₃ treatment on cellular DNA synthesis (Figure 1B) and showed inhibition of DNA synthesis (low BrdU uptake) in GaQ₃-treated MCF-7, H1299 and PC3 cells, whereas the untreated cells did not show any decline in BrdU uptake. This suggested that GaQ₃ abolished DNA synthesis in both p53^+/+ and p53^−/− cells. The SA-β Gal assay (Figure 1C) showed high SA-β Gal staining in GaQ₃-treated cells in MCF-7, H1299 and PC3 cells. We thus concluded that cancer cells with wild-type p53 were more susceptible to GaQ₃; although GaQ₃ was cytotoxic to all three types of cancer cell (p53^+/+, p53^−/− and p53 mutant). Furthermore, GaQ₃ also induced inhibition of DNA synthesis and cellular senescence but this function was independent of p53 status in cancer cells.

GaQ₃ induces G₁/S-phase arrest and apoptosis in cancer cells

The results obtained from the BrdU assay suggested that GaQ₃ might also interfere with cell cycle progression in cancer cells. The cell cycle distribution curve of GaQ₃-treated MCF-7, H1299 and PC3 cells was analysed with PI staining and flow cytometry (Figures 2A and S7). This analysis showed G₁/S-phase arrest and a significant sub-G₁ peak. The G₁–S phase arrest indicated that GaQ₃ might interfere with initiation of DNA replication and synthesis of DNA during S phase. Annexin V staining showed a significant apoptotic fraction in MCF-7 (67%), H1299 (51%) and PC3 (58%) cells after treatment with GaQ₃ (Figures 2B and S8). These data indicated that GaQ₃ induced apoptosis via both p53-dependent and -independent pathways. The nature of cell death (either apoptotic or necrotic) induced by GaQ₃ in cancer cells was examined by inhibiting caspase-3 in the treated cells (Figure S9). Inhibition of caspase-3, using DEVD-FMK, significantly reduced GaQ₃-induced apoptosis, showing that GaQ₃ induced only apoptotic cell death and not necrosis. DNA fragmentation in GaQ₃-treated MCF-7, H1299 and PC3 cells also showed the typical apoptotic DNA ladder formation (Figure 2C, lanes 4, 6 and 8). TUNEL assay using fluorescence-based and elisa-based methods showed higher apoptosis in MCF-7, H1299 and PC3 cells (Figures 2D and S10). Fluorescence microscopy-based TUNEL-analysis showed a significant proportion of TUNEL-positive cells after GaQ₃ treatment, whereas 97% of untreated cells were TUNEL-negative. elisa-based TUNEL assays also showed a significant rise in TUNEL-positive cells upon GaQ₃ treatment in HCT-p53^+/+, HCT-p53^−/− cells (Figure 2D) and MCF-7, H1299 and PC3 cells (Figure S10). These data showed a higher level of apoptotic cells in HCT-p53^+/+ cultures, again implying an important role of p53 in GaQ₃-induced apoptosis.

Figure 2.

GaQ₃ mediated cell cycle arrest and apoptosis in cancer cells. (A) Effects of GaQ₃ on cell cycle progression in MCF-7 and H1299 cells using PI staining. GaQ₃-treated MCF-7 and H1299 cells show G₁–S phase arrest (n= 10). (B) GaQ₃-induced apoptosis in MCF-7 and H1299 cells using Annexin V-FITC staining, showing 67% and 51% apoptosis, respectively (n= 15). (C) DNA fragmentation assay in GaQ₃-treated MCF-7, H1299 and PC3 cells shows a characteristic DNA ladder indicating GaQ₃-induced apoptosis in all three cell lines (n= 3). (D) TUNEL assay using fluorescence microscopy in MCF-7 and H1299 cells (left panel) and using elisa-based assays in HCT-p53^+/+ and HCT-p53^−/− cells. In both sets of experiments, GaQ₃-treated cells were TUNEL positive (n= 8).

GaQ₃ induces intracellular Ca²⁺ release both in p53^+/+ and p53^−/− cells

Metal-based anti-tumour drugs are known to induce Ca²⁺ signalling via intracellular release of Ca²⁺ ions in the cytoplasmic pool (Takemura et al., 1989; Yamagami et al., 1998; Faurskov and Bjerregaard, 2002). We therefore measured intracellular Ca²⁺ release in GaQ₃-treated MCF-7 (p53^+/+) and H1299 (p53^−/−) cells using Fluo-3-AM staining and flow cytometry, at different times up to 10 h after GaQ₃ treatment). As shown in Figure 3A, GaQ₃ significantly increased release of intracellular Ca²⁺ within 30 min incubation and a more gradual increase in intracellular Ca²⁺ levels over the next 7–10 h (panels 3 and 4). Interestingly, the 7 and 10 h intracellular calcium levels were 29% higher in MCF-7 cells than in H1299 cells, although the GaQ₃-induced Ca²⁺ release was not totally dependent on p53 status in these cells. A detailed time course analysis showed that GaQ₃ initiated intracellular Ca²⁺ release by 30 min in MCF-7, H1299 and PC3 cells; however, the pattern of Ca²⁺ release between 0 and 24 h was different in p53^+/+ (MCF-7), p53^−/− (H1299) and p53 mutant (PC3) cells (Figure 3B,C). In H1299 (Figure 3B) and PC3 (Figure 3C) cells, intracellular Ca²⁺ increased slowly and gradually over 24 h of incubation. On the other hand, in MCF-7 cells (Figure 3B), there was a substantial burst of intracellular calcium release at 8 h of incubation,. After silencing the p53 gene in MCF-7 cells, GaQ₃-treatment did not induce the 8 h burst of intracellular Ca²⁺ release, although the gradual increase was maintained (Figure 3D). These data suggested that p53 might play a role in regulation of intracellular Ca²⁺ release in cancer cells. The levels of p53 protein, using Western blotting, were significantly increased in GaQ₃-treated MCF-7 cells (Figure 3E). MCF-7 cells treated with p53 siRNA or tamoxifen were used as negative and positive controls respectively.

Figure 3.

GaQ₃ induced intracellular release of Ca²⁺ in cancer cells. (A) GaQ₃-induced intracellular Ca²⁺ release in MCF-7 and H1299 cells using Fluo-3-AM dye and flow cytometry, determined after 30 min, 7 h and 10 h of GaQ₃ incubation. Results show increased Ca²⁺ release at 30 min and a slower, maintained, increase over 24 h. MCF-7 cells show higher Ca²⁺ release than H1299 cells (n= 15). (B) Time course of intracellular Ca²⁺ release in GaQ₃-treated MCF-7 and H1299 cells, using flow cytometry. There was a sharp increase in Ca²⁺ release at 8 h in MCF-7 (p53^+/+) cells, but not in H1299 (p53^−/−) cells. Data shown are means ± SD (n= 15). (C) Intracellular Ca²⁺ release in p53 mutant (PC3) cells using flow cytometry. Results show a constant increase in intracellular calcium release after GaQ₃ treatment (n= 15, means ± SD). (D) The role of p53 in intracellular Ca²⁺ release in MCF-7 cells analysed by silencing p53 gene using p53 siRNA. Time course of intracellular Ca²⁺ release in GaQ₃-treated MCF-7 cells, with or without p53 siRNA showed that the GaQ₃-induced sharp increase in the intracellular Ca²⁺ release at 8 h was abolished, after p53 gene silencing. Data shown are means ± SD (n= 15). (E) p53 protein level in GaQ₃-treated MCF-7 cells, by Western blot, was increased (lane 2), relative to untreated cells (control). Lanes 3 and 4 show cells treated with p53 siRNA and tamoxifen as negative and positive controls respectively, Tubulin is used as loading control.

Stabilization of p53–p300 transcriptional complex and activation of p53 gene in GaQ₃-treated cells

As the p53 protein level was increased in GaQ₃-treated MCF-7 cells and because p53 requires a transcriptional co-activator p300 to form a transcriptional complex and initiate transcription, the effect of GaQ₃ treatment on p53–p300 binding was analysed. As shown in Figure 4A, p53 did not bind to p300 in untreated MCF-7 cells, but upon addition of GaQ₃, p53–p300 binding was observed, suggesting that GaQ₃ treatment up-regulated p53 and stabilized its transcriptional complex with p300. Mdm2 is the known ubiquitin ligase of p53, which participates in p53 degradation and inhibits p53 activity and we looked for an effect of GaQ₃ treatment on p53–mdm2 interactions. In untreated cells (lane 1), p53 and mdm2 were bound together but this complex was broken in presence of GaQ₃ (lane 2). We also assessed the role of Ca²⁺signalling in p53–p300 binding following GaQ₃ (Figure 4B). As intracellular Ca²⁺ release increased within 30 min of GaQ₃ incubation, MCF-7 cells were treated with GaQ₃ for up to 24 h and the p53–p300 complex assayed at 30 min, 1 h and 3 h. We found p53 bound to p300 at 24 h (also shown in Figure 4A), but not at 30 min or 1 h of GaQ₃ incubation (lanes 4 and 5). Interestingly, at 3 h of GaQ₃ incubation, the p53–p300 complex was stabilized, and after quenching intracellular Ca²⁺ release using TMB-8 (100 µM), this p53–p300 complex was again broken. These data suggested that Ca²⁺ signalling induced binding of p53 to p300 and stabilized the p53 transcriptional complex. We used the non-immunoprecipitated total protein extract (referred to as Input), p53 siRNA and p300 siRNA as controls. As the p53–p300 complex was established, but p53 cellular level was low at the 3 h time point, we hypothesized that this GaQ₃-induced early p53–p300 binding was responsible for up-regulation of p53 mRNA and protein expression. In order to study p53 gene activation by Ca²⁺-induced p53–p300 complex, we used chromatin immunoprecipitation (Figure 4C) to observe p53 and p300 binding on the p53 minimal promoter (Tripathi et al., 2007) (−157 to −397; upstream of p53 transcription start site). This region of the p53 promoter carries p53 binding sites, and p53–p300-mediated activation of this promoter would increase p53 mRNA and protein level. Results showed that both p53 and p300 were bound to the same site as a complex on the p53 minimal promoter after 24 h of GaQ₃ incubation (Figure 4C, lane 3). Interestingly upon p53 silencing, both p53 and p300 did not show binding to the p53 minimal promoter (lane 5). Upon silencing of p300 using p300 siRNA, the binding of both p53 and p300 at the p53 minimal promoter was abolished, suggesting that both p53 and p300 are crucial for the GaQ₃-induced activation of p53 minimal promoter (Figure 4C, lane 6). p53 and p300 both did not bind at the p53 minimal promoter after 30 min and 1 h of incubation with GaQ₃. At 3 h of GaQ₃ incubation, both p53 and p300 showed binding at the p53 minimal promoter, which was abolished after Ca²⁺ quenching (lane 10). Upon silencing both p53 and p300 using their respective siRNAs (Figure 4C, lanes 11 and 12), the 3 h binding of both p53 and p300 to the minimal promoter was abolished. These data suggested that Ca²⁺ signalling induced binding or p53 to its co-activator p300 and then recruited this complex to the p53 minimal promoter to induce up-regulation of p53 gene and protein. Furthermore, the effect of p53–p300 binding to the p53 minimal promoter was assayed by analysing activation of the p53 full (2.5 kb) promoter (Figure 4D). The time course of p53 2.5 kb promoter activity was followed, using the luciferase assay on GaQ₃-treated MCF-7 cells. GaQ₃ activated the p53 promoter at 6 h post GaQ₃ treatment followed by a constant increase in the luciferase activity until 8 h. Ca²⁺ quenching reversed p53 promoter activity in GaQ₃-treated MCF-7 cells (Figure S11). As GaQ₃ induced the p53 minimal promoter by stabilizing the p53–p300 complex, the role of p300 in activation of the p53 promoter was assessed by p53 minimal promoter luciferase analysis in GaQ3-treated MCF-7 cells with p300 silencing (Figure S12). p53 mRNA levels in GaQ₃-treated MCF-7 cells were analysed using both real-time and RT-PCR assay (Figure 4E, RT-PCR in inset). p53 mRNA was increased at 6 h in GaQ₃-treated cells and Ca²⁺ quenching reversed p53 mRNA levels (Figure S13). Furthermore, both p53 mRNA and protein synthesis were reversed in GaQ3-treated cells after p300 gene silencing (Figure S14). These data confirmed the role of calcium-induced p53–p300 complex in p53 up-regulation via activation of the p53 minimal promoter. The time course of increased p53 levels and p53 nuclear translocation was followed in cytoplasmic and nuclear fractions of GaQ₃-treated MCF-7 cells, using immunoprecipitation and in vivoelisa with anti-p53 antibodies (Figure 4F, IPP in inset) and showed significantly increased p53 protein level and p53 nuclear translocation at 9 h of treatment. The purity of nuclear and cytoplasmic fractions was checked by the absence of tubulin in the nuclear fractions (Figure S15, lane 2), and the absence of PARP in the cytoplasmic fractions (lane 3), using immunoprecipitation (Figure S15). The non-immunoprecipitated total protein extract (Input) was used as the control (Figure S15, lane 1).

Figure 4.

GaQ₃ induced p53 expression through Ca²⁺-mediated stabilization of p5–-p300 transcriptional complex. (A) Co-immunoprecipitation (Co-IPP) of p53 with p300 and mdm2 shows p53–p300 binding in GaQ₃-treated cells (lane 2), whereas p53–mdm2 interaction is only observed in untreated cells (lane 1). Tamoxifen (positive control) also induces p53–p300 binding (lane 4) (n= 10). (B) The p53–p300 binding is observed in MCF-7 cells after 24 h, 30 min, 1 h and 3 h of GaQ₃ incubation. Results show that p53 binds to p300 at 3 h of GaQ₃ incubation (lane 6). p53–p300 interaction was not observed at 30 min and 1 h of GaQ₃ incubation; however, p53 and p300 were in a complex after treatment with GaQ₃ for 24 h (Figure 3C, as well). Quenching the release of intracellular Ca²⁺ using TMB-8 (100 µM) abolishes the p53–p300 binding induced upon 3 h of GaQ₃ incubation (lane 7). p53 siRNA and p300 siRNA are used as controls (lanes 8 and 9). In this experiment, immunoprecipitation used an anti-p53 N-terminus antibody (Ab421), and Western blots were developed using both anti-p53 and anti-p300 Abs (n= 8). (C) The binding of p53–p300 complex to the p53 minimal promoter at 30 min, 1 h and 3 h of GaQ₃ incubation (lanes 7, 8 and 9). Chromatin immunoprecipitation (ChIP) shows strong binding of both p53 and p300 on the −157 to −397 base pairs of p53 minimal promoter in MCF-7 cells treated with GaQ₃ for 24 h (lane 3). No binding of either p53 or p300 was observed at 30 min and 1 h of GaQ₃ incubation (lanes 7 and 8); however, the binding was present at 3 h (lane 9). Upon quenching of intracellular Ca²⁺ release using TMB-8, the GaQ₃-induced p53–p300 binding on p53 minimal promoter is abolished (lane 10) (n= 8). p53 and p300 did not show binding to the p53 minimal promoter in the untreated cells (lane 2). The non-immunoprecipitated total protein extract (Input), scrambled primers, p53 siRNA and p300 siRNA are used as controls. (D) The activity of p53 2.5 Kb promoter analysed in GaQ₃-treated MCF-7 cells over 24 h. Results show a significant increase in p53 2.5 Kb promoter activity at 6 h of GaQ₃ incubation. Furthermore, a rise in the promoter activity was observed until 18 h of incubation (*P < 0.05, for values from 6–18 h, n= 4, means ± SD, anova). (E) p53 mRNA level was determined by real-time PCR in GaQ₃-treated MCF-7 cells in a time-dependent manner (every hour). Results show a significant increase in p53 mRNA level from 6 h of GaQ₃ incubation. (n= 6, means ± SD). The mRNA expression is also analysed using RT-PCR at 0, 6, 12, 18 and 24 h of GaQ₃ incubation (inset) (n= 4). (F) Time-dependent analysis of the increase in p53 protein expression and p53 nuclear translocation in GaQ₃-treated MCF-7 cells. elisa using p53 whole cell, nuclear and cytoplasmic fractions show a significant increase in p53 protein level and p53 nuclear translocation at 6 h of GaQ₃ incubation (n= 8, SD, anova). Furthermore, the p53 nuclear translocation is confirmed by immunoprecipitation experiment using anti p53 C-terminus antibody (421) in GaQ₃-treated MCF-7 cells at 0 and 24 h time points (WC: whole cell, NF: nuclear fraction, CF: cytoplasmic fraction) (n= 6).

GaQ₃ damaged DNA in treated cells and DNA damage and p53 stabilization are directly linked (Campaner et al., 2011). In order to establish the role of calcium in stabilization of p53 and to exclude the possibility of DNA-damage-induced p53 activation in GaQ₃-treated cells, the time course of cellular DNA damage was analysed in GaQ₃-treated MCF-7, H1299 and PC3 cells (Figure S16), using apurinic/apyrimidinic (AP) sites as the measure of cellular DNA damage. GaQ₃ treatment did not induce any DNA damage till 15 h of incubation, whereas Ca²⁺-induced stabilization of p53–p300 complex was present at 3 h and p53 stabilization and up-regulation occurred 6–8 h after GaQ₃ incubation. Thus, GaQ₃-induced DNA damage had no role in p53 up-regulation and p53 gene transcription was up-regulated in MCF-7 cells due solely to GaQ₃-induced intracellular Ca²⁺ release.

Intracellular Ca²⁺ release induces ROS signalling in GaQ₃-treated cells

Ca²⁺ signalling induced ROS in cancer cells (Lemarie et al., 2004; Son et al., 2010) and we analysed the effect of GaQ₃ on cellular ROS, using DCFH-DA staining and flow cytometry (Figures 5A and S17). GaQ₃ induced cellular ROS in MCF-7, PC3 and H1299 cells, although the ROS levels in MCF-7 cells (panel 1) were nearly twofold higher than in H1299 cells (panel 2) and PC3 cells (Figure S18). GaQ₃-induced ROS was independent of p53 status since ROS was up-regulated in both p53^+/+ (MCF-7) and p53^−/− (H1299) cells. However, in MCF-7 cells, p53 gene silencing using p53 siRNA significantly decreased cellular ROS. This suggested that GaQ₃ might initiate ROS synthesis in a p53-independent manner, but the presence of p53 in MCF-7 cells appeared to play a role in amplifying the cellular ROS response (Figure 5A). As p53 was not required for ROS generation in GaQ₃-treated MCF-7 and H1299 cells, we searched for the cellular factors that might be responsible for ROS, by conducting a flow-cytometry-based, time course analysis of ROS level in MCF-7, H1299 and PC3 cells (Figures 5B and S18). Interestingly, the ROS profile in MCF-7 and H1299 cells synchronized with the intracellular Ca²⁺ release profile. Quenching intracellular Ca²⁺ (using TMB-8; 100 µM) in MCF-7 cells led to inhibition of ROS, implying Ca²⁺-mediated activation of ROS in GaQ₃-treated cancer cells.

Figure 5.

Intracellular Ca²⁺ release induced ROS in GaQ₃-treated cells. (A) Flow cytometry was used to assess ROS generation in GaQ₃-treated MCF-7 and H1299 cells (panels 1 and 2).Results show a significant increase in ROS in both cell lines, and the role of p53 in GaQ₃-induced ROS is confirmed upon silencing p53 gene (blue, untreated cells; red, GaQ₃-treated cells; pink, p53 siRNA in GaQ₃-treated cells). Results show that GaQ₃ induced higher ROS synthesis in p53^+/+ (MCF-7) cells than in p53^−/− (H1299) cells (n= 7; Student's t-test). (B) Time-dependent analysis of ROS synthesis, using flow cytometry in GaQ₃-treated MCF-7, GaQ₃-treated H1299 and GaQ₃-treated + Ca²⁺ quenched MCF-7 cells. p53^+/+ (MCF-7) cells show higher generation of ROS around 7 h of GaQ₃ treatment, than in p53^−/− (H1299) cells, which shows a gradual and constant increase in ROS. Ca²⁺ quenching using TMB-8 abolishes GaQ₃-induced ROS, suggesting that ROS generation is dependent on the intracellular Ca²⁺ release (n= 8, means ± SD). (C) (Panel i) In vivoelisa shows that quenching of the intracellular Ca²⁺ release using TMB-8 abolishes the p53 protein expression and p53 nuclear translocation in GaQ₃-treated MCF-7 cells. (Panel ii) ROS quenching using N-acetyl cysteine (NAC; 50 µM) shows no change in p53 protein expression and p53 nuclear translocation in GaQ₃-treated MCF-7 cells (n= 7, means ± SD). (D) Immunoprecipitation using anti-p53 N terminus (1801Ab) to confirm the results of in vivoelisa. Results (Input, non-immunoprecipitated total protein extract; WC, whole cells; NF, nuclear fraction; CF, cytoplasmic fraction) show that Ca²⁺ quenching, but not ROS quenching using NAC (50 µM), inhibited up-regulation and nuclear translocation of p53 protein at 10 h of GaQ₃ incubation (n= 5).

Because Ca²⁺ signalling induced both ROS and p53, and ROS is known to activate p53, there was a possibility of cross-talk between Ca²⁺, p53 and ROS. We therefore measured cellular and nuclear levels of p53 in GaQ₃-treated MCF-7 cells in presence of Ca²⁺ quenching (left panel) and ROS quenching (right panel) over 10 h, using elisa (Figure 5C). The GaQ₃-induced increase in p53 level and p53 nuclear translocation was abolished by quenching intracellular Ca²⁺ but was unaffected by quenching ROS. The elisa results were confirmed by immunoprecipitating p53 from the whole cell, nuclear and cytoplasmic fraction of GaQ₃-treated, Ca²⁺ and ROS-quenched MCF-7 cells (Figure 5D). Up-regulation and nuclear translocation of p53 was inhibited after Ca²⁺ quenching, whereas ROS quenching had no effect. These data established that intracellular Ca²⁺ release regulated both p53 and ROS separately, and that intracellular Ca²⁺ release, and not ROS, induced p53 activation in GaQ₃-treated cells.

GaQ₃ induced both FAS-mediated extrinsic and p53-mediated intrinsic apoptosis in cancer cells

As GaQ₃ induced apoptosis in both p53^+/+ and p53^−/− cells, we asked if GaQ₃ might initiate p53-independent extrinsic pathways of apoptosis. Intracellular Ca²⁺ release (Lu and Tian, 2005; Lu et al., 2006; Timmins et al., 2009) increased cellular ROS (Gulbins et al., 1996; Sato et al., 2004; Medan et al., 2005) and p53 was also known to induce and regulate the expression of FAS in cancer cells (Tamura et al., 1995; Bennett et al., 1998; Fukazawa et al., 1999; Sun et al., 2000; Yin et al., 2003; Li et al., 2005b). GaQ₃ induced 67% apoptosis in MCF-7 cells (Figure 6A) and p53 gene silencing significantly reduced the GaQ₃-induced apoptotic fraction (panel 3). This apoptotic fraction was decreased by FAS siRNA (panel 4), suggesting the contribution of an extrinsic apoptotic pathway. Silencing of both p53 and FAS totally prevented GaQ₃-induced apoptosis. Thus both p53-mediated intrinsic, and FAS-mediated extrinsic, pathways of apoptosis were activated by GaQ₃. In p53^−/− (H1299) cells (Figure 6B), either Fas silencing (panel 3) or Ca²⁺ quenching (panel 4) abolished GaQ₃-induced apoptosis. These data suggested that, in p53^−/− cells, GaQ₃ induced apoptosis only via extrinsic FAS-dependent pathways and that Ca²⁺ was responsible for FAS activation.

Figure 6.

GaQ₃ induces FAS-mediated apoptosis in p53^−/− cells. (A) FAS silencing, p53 silencing, Ca²⁺ quenching and combined silencing of p53 and FAS significantly reduced the GaQ₃-induced apoptosis in MCF-7 cells (n= 10). (B) Apoptosis was assessed by flow cytometry in GaQ₃-treated H1299 cells with FAS silencing and Ca²⁺ quenching (n= 10). (C) (Panel i) RT-PCR to show increased mRNA for FAS receptor in MCF-7, H1299 and PC3 cells treated with GaQ₃ at 0, 6, 12, 18 and 24 h respectively. (Panel ii) Western blot analysis to study expression of FAS protein in MCF-7, H1299 and PC3 cells treated with GaQ₃ for 0, 6, 12, 18 and 24 h respectively. (Panel iii) Western blot to study the role of p53 protein in FAS up-regulation upon silencing p53 in GaQ₃-treated MCF-7 cells (n= 7). (D) elisa of the cell membrane and the nuclear/cytoplasmic fractions of ROS-quenched, GaQ₃-treated MCF-7 and H1299 cells to assess translocation of FAS to the cell membrane. Data shown are means ± SD (n= 8). P < 0.05, significantly greater than untreated MCF-7 cells or H1299 cells, anova).

To further establish the involvement of FAS in GaQ₃-induced apoptosis, the time course of expression of FAS receptor mRNA and protein was analysed, every 6 h, in GaQ₃-treated MCF-7, H1299 and PC3 cells (Figure 6C) and showed a time-dependent increase in FAS mRNA and protein expression on incubation with GaQ₃. In a parallel experiment, p53 silencing decreased (30%) FAS mRNA expression in GaQ₃-treated MCF-7 cells (Figure 6C, right panel), suggesting that up-regulation of p53 increased FAS gene transcription. Because FAS protein becomes active and induces apoptosis only when it migrates to the cell membrane; we analysed by elisa the separated cytoplasmic/nuclear and cell membrane fractions of MCF-7, H1299 and PC3 cells (Figures 6D and S19) for FAS. GaQ₃ treatment increased the translocation of FAS from the cytoplasm to the cell membrane in both MCF-7 and H1299 cells. It is also interesting to note that ROS quenching decreased FAS expression in both cell lines, suggesting that Ca²⁺-induced ROS were also involved in FAS activation.

p53 downstream genes and mir34-a execute apoptosis in GaQ₃-treated cells

GaQ₃ induced apoptosis via both FAS-dependent extrinsic and p53-dependent intrinsic pathways. We therefore analysed the expression of p53 target genes involved in cellular apoptosis. We also analysed the promoter activation and mRNA expression of the p53 regulated family of micro-RNA, mir-34, and its putative role in p53-mediated apoptosis (Raver-Shapira et al., 2007). GaQ₃-treated MCF-7 and H1299 cells were transfected with luciferase constructs carrying mir-34a and mir-34b/c promoter sequences with p53 binding sites (Figure 7A) (He et al., 2007). Only in p53^+/+ (MCF-7) cells, GaQ₃ induced activity of mir-34a promoter. 18-fold with a 12-fold increase in the mir-34b/c promoter activity. Assays of mir-34a mRNA using real-time PCR, in GaQ₃-treated MCF-7 cells (Figure 7B), showed increased expression, suggesting that mir-34a might play a role in GaQ₃-mediated apoptosis. Western blotting showed that GaQ₃ induced expression of several p53 downstream apoptotic genes, as protein levels of BAX, PUMA, NOXA1, BID, SUMO, p21, APAF1 and PIG3 were significantly higher in GaQ₃-treated MCF-7 cells (Figure 7C).

Figure 7.

miR-34a was induced and in turn induced the expression of p53 downstream genes involved in apoptosis. (A) Luciferase assay to monitor the mir34a and mir34b/c promoter activity in both untreated and GaQ₃-treated MCF-7 and H1299 cells. Results show a significant increase in both mir-34a and mir34b/c promoter activity only in GaQ₃-treated MCF-7 cells; note that mir-34a activity is higher than mir-34b/c activity (panels 1 and 2). Data shown are means ± SD (n= 9); *P < 0.05, significant effect of GaQ₃, anova. (B) miR-34a RNA expression quantified in GaQ₃-treated MCF-7 cells using real-time PCR, showed increased miR34a mRNA expression in MCF-7 cells after GaQ₃ (Data shown are means ± SD (n= 5); *P < 0.05, significant effect of GaQ₃, anova). (C) Western blot analysis shows the up-regulation of p53 downstream genes in GaQ₃-treated MCF-7 cells. The expression of BAX, PUMA, NOXA, BID, SUMO, p21, APAF-1 and PIG3 increased in GaQ₃-treated MCF-7 cells (lane 2), and decreased after p53 silencing (lane 3). Transfection with FAS cDNA shows no effect on the expression of these proteins. The cDNA of all the respective proteins are transfected in the MCF-7 cells and used as positive control. Data shown are means ± SD *P < 0.05, significantly different as indicated, anova.

Discussion

In this study, we have established the mechanisms underlying the anti-neoplastic activity of a novel gallium-derived coordination compound, GaQ₃, that induced both extrinsic and intrinsic pathways of apoptosis. GaQ₃ initiated the intracellular release of Ca²⁺ in p53^+/+, p53^−/− and p53 mutant cell lines, which led to apoptosis. Heavy metals bind to and interact with receptor and intracellular proteins that affect Ca²⁺ homeostasis (Faurskov and Bjerregaard, 2002), and their cytotoxicity is known to be related to changes in intracellular Ca²⁺ levels (Lemarie et al., 2004; Wang et al., 2008). Ca²⁺ ions are the cellular messengers that control every aspect of cell and tissue physiology and play a major role in triggering mitotic division in numerous cell types and regulation of cell death (Giorgi et al., 2008). Alterations in the cytoplasmic level of intracellular Ca²⁺ induced apoptosis (McConkey et al., 1992; Berridge, 1993; Orrenius et al., 2003) and Ca²⁺ions could be turned into death signals when delivered at the wrong time and place (Berridge, 1993; Hajnoczky et al., 2006). Also the transfer of Ca²⁺ from the endoplasmic reticulum (ER) to the mitochondria was required for the initiation of apoptosis. However, some of the Ca²⁺-induced signals like Ca²⁺-activated RAF/MEK/ERK pathways also mediate p53-independent apoptosis (Li et al., 2005a).

Our MTT data showed that GaQ₃-mediated cell death occurred in p53^+/+ (MCF-7 and HepG2), p53 mutant (PC3) and p53^−/− (H1299) cells, although in p53^+/+ cells, GaQ₃ induced apoptosis at lower doses. In p53^+/+ cells, intracellular Ca²⁺ release stabilized the binding of p53 to its transcriptional co-activator p300 and stabilized the p53 transcriptional complex by recruiting p53–p300 complex to the p53-DNA binding site on the p53 minimal promoter (Tripathi et al., 2007). Ca²⁺ quenching abolished p53–p300 complex formation and the binding of p53–p300 complex to the p53-minimal promoter, suggesting that p53–p300 interaction was Ca²⁺-dependent. Silencing of p53 or p300 disrupted the p53–p300 complex and its subsequent binding to the p53 minimal promoter. Interestingly, silencing of p53 gene in p53^+/+ cells switched off intracellular Ca²⁺ release. After 3 h of incubation with GaQ3, p53–p300 complex bound to the p53 minimal promoter. Silencing of p53 or p300 at 3 h abolished the p53–p300 binding to p53 promoter, suggesting that Ca²⁺-induced formation of the p53–p300 complex was a crucial event in GaQ3-induced p53 activation.

The mechanism of Ca²⁺–p53 crosstalk is complicated by its different pattern of response to a range of drugs. Patients with p53 deletions have higher serum calcium (Chang et al., 2005). Although both Ca²⁺ binding proteins S100A2 (Mueller et al., 2005) and S100B (Rustandi et al., 2000) bind to the C-terminus of p53, S100A2 increased p53 transcriptional activity, whereas S100B inhibited p53 acetylation in a Ca²⁺-dependent manner. Furthermore, our data show that in cancer cells, there is a synergistic relation between Ca²⁺ release pattern and status of p53 protein. This conclusion was drawn based upon comparing the intracellular Ca²⁺ release patterns between p53^+/+ (MCF-7), p53 mutant (PC3) and p53^−/− (H1299) cells. In GaQ₃-treated H1299 cells, the release of Ca²⁺ between 0 and 24 h was gradual, whereas in GaQ₃-treated MCF-7 cells, there was a sharp increase in the release of intracellular Ca²⁺ levels at 8 h. This sharp rise in Ca²⁺ was further synchronized with p53 transactivation, as p53 silencing prevented Ca²⁺ release. One then could conclude that Ca²⁺ signalling is responsible for p53 transactivation, leading to p53-dependent apoptosis in cancer cells.

Furthermore, p53 up-regulation was earlier shown to be associated with the generation of ROS (Johnson et al., 1996; Liu et al., 2008). Our data indicated that GaQ₃ induced the generation of ROS in MCF-7, PC3 and H1299 cells, with a slow, gradual and continuous rise in ROS level in p53^−/− (H1299) and p53 mutant (PC3) cells, whereas a sudden rise was observed at 9 h in p53^+/+ (MCF-7) cells. Interestingly, Ca²⁺ quenching reversed ROS generation at 2 h in GaQ₃-treated cells. Intracellular Ca²⁺ release disrupted mitochondrial Ca²⁺ equilibrium, which facilitated the formation of ROS (Lemarie et al., 2004; Wang et al., 2008). In p53^−/− cells, GaQ₃-induced Ca²⁺ release must be p53-independent. Thus it was Ca²⁺ and not p53 that was responsible for the generation of ROS. However, in p53^+/+ cells, Ca²⁺ release up-regulated p53, and this in turn further increased intracellular ROS content at 8 h after GaQ₃ treatment. Our data suggest that Ca²⁺ signalling plays an important role in GaQ₃-induced toxicity, which may also be linked to ROS generation. The cellular redox state could significantly modulate Ca²⁺ signalling (Camello-Almaraz et al., 2006; Davidson and Duchen, 2006; Vercesi et al., 2006; Yan et al., 2006; Zima and Blatter, 2006; Hool and Corry, 2007) and it seems that GaQ₃ initiated a cross-talk between Ca²⁺-modulated ROS production and ROS-modulated Ca²⁺-signalling. The anti-cancer drug resveratrol induced an early biphasic increase in the concentration of free intracellular Ca²⁺ in breast cancer cells, probably resulting from depletion of ER Ca²⁺ stores (Sareen et al., 2007). A similar finding was noticed with the anti-cancer agent CDDO which increased cytoplasmic free Ca²⁺ in several cancer cell lines. A Ca²⁺ chelator reduced CDDO-induced increase in cytoplasmic free Ca²⁺ inhibiting caspase activation and DNA fragmentation, implying that Ca²⁺ played a pivotal role in signalling the initiation of apoptosis (Hail et al., 2004). By contrast a calcium mobilizer, calcimycin, activated the RAF/MEK/ERK pathway through RAS, which was essential for the induction of apoptosis (Li et al., 2005a).

GaQ₃ activated mir34 a/b/c promoters, exclusively, in p53^+/+ MCF-7 cells, shown by the more than eight-fold rise in mRNA levels. A member of the miRNA34 family, miR-34a, which is commonly deleted in human cancers, is directly trans-activated by p53. Expression of miR-34a causes dramatic reprogramming of gene expression and promotes apoptosis. Therefore, it is likely that an important function of miR-34a is the modulation and fine-tuning of the gene expression program initiated by p53 (Chang et al., 2007; Hermeking, 2007). As p53 induces over-expression of miR-34a, it is a direct pro-apoptotic transcriptional target of p53 that can mediate some of the biological effects of p53 (Raver-Shapira et al., 2007).

Ca²⁺ signalling also activates the FAS-mediated pathway of cell death (Zhang et al., 2010)and FAS up-regulation as well as its translocation to the cell membrane causes apoptosis (Nagata, 1996). The fact that GaQ₃ up-regulated FAS in MCF-7, PC3 and H1299 cells further confirmed that the GaQ₃-mediated apoptotic cell death may be due to p53-dependent (intrinsic) and p53-independent (extrinsic) pathways. Strikingly, cisplatin-induced apoptosis, which was associated with p53-dependent and -independent responses, was also reported in human bladder cancer cells (Konstantakou et al., 2009), and both intrinsic and extrinsic pathways were observed in TNF-α induced apoptosis, which was further enhanced in presence of cisplatin (Barton et al., 2005). However, cisplatin exclusively induced intrinsic apoptosis in inner ear cells (Garcia-Berrocal et al., 2007). By contrast, bleomycin induced only the extrinsic pathway in pulmonary endothelial cells (Mungunsukh et al., 2010).

In conclusion, we have established the mechanism of action of a novel anti-cancer gallium complex GaQ₃ in a variety of cancer cells with different p53 status. A model has been proposed to explain the mechanism of GaQ₃-induced apoptosis in cancer cells (Figure 8). GaQ₃-mediated intracellular Ca²⁺ release is the primary event that triggers both p53-dependent intrinsic and p53-independent extrinsic apoptosis. A future strategy would be to explore the complex nature of p53–Ca⁺² cross-talk, with reference to GaQ₃-induced apoptosis.

Figure 8.

Scheme outlining the mechanisms involved in the action of GaQ₃ in cancer cells.

Glossary

BrdU

5-bromo-2′-deoxyuridine

DCFDA

2′,7′ di-chloro-dihydro-fluorescein-diacetate

GaQ₃

tris(8-quinolinolato)gallium(III)(KP-46)

Mn-SOD

manganese superoxide dismutase

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

NAC

N-acetyl cysteine

PBMC

peripheral blood mononuclear cell

ROS

reactive oxygen species

SA-β-Gal

senescence-associated β-galactosidase

TUNEL

terminal deoxynucleotidyl transferase-mediated dUTP nick end labelling

Conflict of interest

None of the authors have any conflict of interest in this study.

Supporting information

Additional Supporting Information may be found in the online version of this article:

Figure S1 MTT assay in MCF-7 cells (p53^+/+) treated with GaQ3 for 0, 24, 48 and 72 h. All the cells are treated with dose of GaQ3 varying from 1 to 10 μM(n = 10, mean; SD, statistical analysis; anova).

Figure S2 MTT assay in HepG2 cells (p53^+/+) treated with GaQ3 for 0, 24, 48 and 72 h. All the cells are treated with dose of GaQ3 varying from 1 to 10 μM (n = 10, mean; SD, statistical analysis; anova).

Figure S3 MTT assay in PC3 cells (p53 mutant) treated with GaQ3 for 0, 24, 48 and 72 h. All the cells are treated with dose of GaQ3 varying from 1 to 10 μM (n = 10, mean; SD, statistical analysis; anova).

Figure S4 MTT assay in H1299 cells (p53^−/−) treated with GaQ3 for 0, 24, 48 and 72 h. All the cells are treated with dose of GaQ3 varying from 1 to 10 μM (n = 10, mean; SD, statistical analysis; anova).

Figure S5 MTT assay in GaQ3-treated p53^+/+ (MCF-7 and HepG2), p53^−/− (H1299) and p53 mutant (PC3) cells. All the cells are treated for 24 h with doses of GaQ3 varying from 1 to 10 μM (n = 10, mean; SD, statistical analysis; anova).

Figure S6 LDH assay in GaQ3-treated p53^+/+ (MCF-7 and HepG2), p53^−/− (H1299) and p53 mutant (PC3) cells. All the cells are treated for 24 h with dose of GaQ3 varying from 1 to 10 μM (n = 10, mean; SD, statistical analysis; anova).

Figure S7 Cell cycle distribution curve of GaQ3-treated PC3 cells (p53 mutant) was analysed using flow cytometry. Results show that GaQ3 induces G₀–G₁ phase arrest in PC3 cells (n = 7). Lower panel: the cell cycle distribution curves of GaQ3-treated MCF-7, H1299 and PC3 cell are graphically represented (n = 5, mean; SD, statistical analysis; anova).

Figure S8 Percentage cellular apoptosis in GaQ3-treated MCF-7, H1299 and PC3 cells was measured using Annexin V staining and flow cytometry (n = 8, error bars; SD, statistical analysis; anova).

Figure S9 Percentage cellular apoptosis was measured in GaQ3 + DEVD-FMK (caspase inhibitor) treated MCF-7 cellsusing Annexin V staining and flow cytometry. Results show asignificant decrease in GaQ3-induced apoptosis upon incubation with DEVD-FMK (n = 10, mean; SD, statistical analysis; anova).

Figure S10 TUNEL staining shows cellular apoptosis in PC3 cells upon GaQ3 treatment (n = 4). Lower panel: percentage TUNEL-positive cells are shown in GaQ3-treated MCF-7, H1299 and PC3 cells (n = 10, mean; SD, statistical analysis; anova).

Figure S11 p53 2.5 Kb promoter activity was measured using luciferase assay, in GaQ3-treated MCF-7 cells in which intracellular calcium release is quenched using TMB-8. Results show that p53 promoter is inactive upon calcium quenching (n = 7, mean; SD, statistical analysis; anova).

Figure S12 p53 minimal promoter activity was measured using luciferase assay, in GaQ3-treated MCF-7 cells in which p300 gene is silenced using p300 siRNA. Results show that p53 minimal promoter is inactive in absence of p300 (n = 7, mean; SD, statistical analysis; anova).

Figure S13 p53 mRNA analysis (time course) in GaQ3-treated MCF-7 cells and TMB-8 + GaQ3-treated MCF-7 cells (calcium quenching), using real-time PCR. Results show that the GaQ3-induced 6 h rise in p53 mRNA level is abolished upon calcium quenching (n = 6, mean; SD, statistical analysis; anova).

Figure S14 p53 mRNA analysis using real-time PCR shows that p300 siRNA abolished the GaQ3-induced rise in p53 mRNA level (lane 5). This data suggest the crucial role of p53–p300 complex in p53 mRNA synthesis (n = 5, mean; SD, statistical analysis; anova). Lower panel: the role of p300 in p53 protein synthesis is observed. GaQ3-induced rise in p53 protein level (lane 2) is reversed upon p300 gene silencing (lane 3) in MCF-7 cells (n = 4).

Figure S15 Immunoprecipitation experiment shows the separation of the nuclear and cytoplasmic fractions of the MCF-7 cells. Immunoprecipitation (IPP) with PARP antibody only stains the nuclear fraction and not the cytoplasmic fractions, and IPP with tubulin antibody stains only the cytoplasmic fractions. The non-immunoprecipitated total protein extracts (Input) are usedas controls (n = 4).

Figure S16 Time-dependent analysis of GaQ3-induced DNA damage is conducted in MCF-7 cells. Results show that GaQ3 does not induce DNA damage until 15 h incubation with MCF-7 cells (n = 11, mean; SD, statistical analysis; anova). Lower panel: GaQ3-induced cellular DNA damage is analysed in PC3 and H1299 cells (n = 9, mean; SD, statistical analysis; anova).

Figure S17 The rise in cellular ROS is observed in GaQ3-treated H1299 cells using DCHF-DA staining in flow cytometry (n = 7).

Figure S18 Time-dependent analysis of GaQ3-induced rise in cellular ROS was measured in PC3 cells using flow cytometry. Results show that GaQ3 induces constant increase in the cellular ROS (red line). Upon quenching intracellular calcium release using TMB-8, the GaQ3-induced rise in cellular ROS is abolished (black line) (n = 11, mean; SD, statistical analysis; anova).

Figure S19 ELISA of the cell membrane (green) and the nuclear/cytoplasmic (black) fractions of ROS quenched,GaQ3-treated PC3 cells was performed to examine the translocation of FAS on the cell membrane (n = 8, mean; SD, statistical analysis; anova).

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