Background
Soft tissue sarcomas are a rare and heterogeneous entity of tumors with an annual incidence of 2-4/100,000 [
1]. In contrast to carcinomas, soft tissue tumors are of mesenchymal origin. According to the tumor grading the 5-year survival ranges from 72-83% in well differentiated G1 sarcomas, 53-59% in G2 sarcomas to 26-42% in G3 sarcomas [
2]. Metastatic disease becomes evident within the first 2–3 years after initial diagnosis and is the main cause of mortality in these patients [
3]. Despite multi-disciplinary treatment (i.e. surgery, chemotherapy and radiation therapy), the rate of recurrence remains higher than 50%, and results in diffuse metastatic disease and the death of the patients [
4]. Chemotherapy for advanced tumors remains unsatisfactory due to low chemosensitivity despite combination chemotherapeutics. The most effective chemotherapeutic agents are the anthracyclines doxorubicin and epirubicin [
5‐
7]. Doxorubicin still remains the first line chemotherapeutic for soft tissue sarcomas. Unfortunately, its cytostatic effect in therapeutic doses is frequently insufficient (monotherapy response rate of only 18-29%); but the use of higher doxorubicin doses is limited by the development of systemic toxicity, especially cardiotoxicity. In addition to the poor response rate to doxorubicin, the development of drug resistance remain an unresolved problem [
8‐
12].
The anionic and weakly acidic antibiotic salinomycin acts in different biological membranes including cytoplasmic and mitochondrial membranes. In addition to its antimicrobial properties, salinomycin selectively depletes breast cancer stem cells from tumorspheres and impedes breast tumor growth in mice xenograft experiments [
13]. The activation of apoptotic pathways by salinomycin is independent of p53 and caspase activation [
14]. Further, it has been reported that salinomycin sensitized cancer cells by reducing p21 levels [
15,
16]. However, little is known about its impact on sarcomas.
In order to improve the oncological treatment of patients with soft tissue sarcoma mechanisms to overcome drug resistance and reduce drug toxicity must be identified. Therefore, the aim of this study was to evaluate the effect of salinomycin on the chemosensitivity to doxorubicin in three different soft tissue sarcoma cell lines.
Methods
Cell culture
The human soft tissue sarcoma cell lines SW872 (liposarcoma cell line), A204 (rhabdomyosarcoma cell line) and HT-1080 (fibrosarcoma cell linie) (ATCC-LGC Standards, Wesel, Germany) were cultured in DMEM (PAN-Biotech, Aidenbach, Germany) media containing 10% fetal bovine serum (Hyclone-Thermo Scientific, Bonn, Germany) and 1% penicillin/streptomycin (PAN-Biotech, Aidenbach, Germany) and were incubated at 37°C and 5% CO2.
Cell viability assay
Cells were seeded at 2×104 per well in 24 well plates and treated 16 h later at the indicated drug concentrations. Forty-eight hours after the application of salinomycin (Sigma, Taufkirchen, Germany), doxorubicin (Sigma, Taufkirchen, Germany) or both compounds together, respectively, 100 μl of MTT solution (5 mg/mL) was added per well. Cells were lysed with 250 μl triplex solution (1 mM HCl; 5% iso-butanol; 1% SDS), after 3 h of incubation. Optical density was measured at 562 nm with a background correction at 630 nm. All data points were normalized with respect to the DMSO control.
Cytotoxicity assay
Alternatively, determination of cytotoxicity was carried out using the MultiTox-Glo assay (Promega, Mannheim, Germany). For this, 3000 cells per well were seeded in 96 well plates (Corning, Amsterdam, Netherlands), and treated 16 h after cell seeding in the presence of a dose–response of doxorubicin in the presence and absence of 1 μM Salinomycin (i.e. 751 ng/mL). Toxicity was measured after 24 h according to the manufacturer’s instructions in four independent measurements. The cytotoxicity is expressed as the ratio of live cell fluorescence to dead cell luminescence, relative to the vehicle control.
Caspase assay
Caspase activities were measured using the Caspase Glo 3/7 and 9 assays from Promega (Mannheim, Germany), according to the manufacturer’s instructions. The resulting luminescence was measured with a Tecan M200 microplate reader (Tecan, Crailsheim, Germany) for 10 s measurement period. Values were corrected for differences in cell numbers by simultaneously conducting a MTT assay (Carl Roth, Karlsruhe, Germany).
Annexin V analysis
Cells were plated at 1.5×105 cells per well in a 6 well plate 16 h before treatment and treated as indicated. DMSO served as a vehicle control. Twenty-four hours after treatment, cells were harvested and apoptosis was determined by Annexin V-Alexa 488 (Life Technologies, Darmstadt, Germany) staining. Cells were counterstained with 0.2 μM TO-PRO-3 iodide (Life Technologies, Darmstadt, Germany) to discriminate between vital and dead cells. After staining the cells were measured on a Guava HT flow cytometer (Millipore, Schwalbach am Taunus, Germany) in triplicates. Data analysis was carried out using the Express Pro 2.0 software (Millipore, Schwalbach am Taunus, Germany).
Cell cycle analysis
Cells were plated at 1.5×105 cells per well in a 6 well plate 16 h before treatment. Cells were trypsinized, collected by centrifugation, and fixed in ice-cold 70% ethanol, 48 h post treatment. Fixed cells were collected by centrifugation and washed once with PBS. Subsequently, the cells were stained in PBS buffer containing 50 μg/mL propidium iodide (MP Biochemicals, Illkirch, France), 0.1% Triton X-100 (Sigma, Taufkirchen, Germany) and 10 U/mL RNase A (Sigma, Taufkirchen, Germany). After incubation for 30 min at 37°C, cell cycle profiles were measured on a Guava HT flow cytometer following data analysis using Express Pro 2.0.
Reporter NF-κB analysis
NF-κB activity was analyzed in triplicates by transfecting 5×104 HT-1080 cells with 300 ng pGL4.32[luc2P/NF-kB-RE/Hygro] vector (Promega, Mannheim, Germany), using X-tremeGENE (Roche, Mannheim, Germany); 10 ng pRL-TK (Promega, Mannheim, Germany) served as transfection control. Firefly and Renilla luciferase activities were measured 6 h and 10 h post treatment. The luciferase-signals were measured for 10s (Tecan M2000, Crailsheim, Germany). The Renilla signal was used for normalization. Mean values and SEM were calculated from triplicates.
Western blot analysis
HT-1080 cells were seeded with 1×106 cells per 10 cm dish. Sixteen hours post seeding, the cells were subjected for 6 h to the different treatments. The isolation of nuclear and cytoplasmic fractions were carried out after cells were allowed to swell on ice for 10 min in 500 μl of hypotonic buffer (20 mM Tris–HCl, pH 7.4, 5 mM MgCl2, 1.5 mM KCl, 0.1% NP-40, 50 mM NaF, 2 mM sodium orthovanadate, and protease inhibitors (Complete, Roche)). Cells were subsequently disrupted by passing them several times through a 26 ½ gauge syringe needle, followed by a centrifugation at 800×g (5 min; 4°C). The supernatants were further centrifuged at 10,000×g (15 min; 4°C) to remove insoluble pellets, and the resulting supernatants were collected as the cytoplasmic fractions. The pellets were resuspended in 100 μl of TKM buffer (20 mM Tris-acetate; pH 7.4, 50 mM KCl, 5 mM MgCl2, containing protease and phosphatase inhibitors). After centrifugation (800×g; 10 min; 4°C), the supernatants were collected like the cytoplasmic fractions. From each fraction, 30 μg total protein were subjected to 4-12% BisTris-PAGE and transferred onto PVDF membranes (Millipore, Schwalbach, Germany) with 2 mA/cm2 for 1 h. After protein transfer membranes were blocked in PBS-T containing 5% (w/v) skimmed milk, for 1 h and incubated with anti-pS15 p53 antibody (Cell Signaling, Frankfurt am Main, Germany) and anti-p53 (Clone DO-1, Sigma-Aldrich, Taufkirchen, Germany) overnight (1:1000 in PBS-T). As loading control for the cytoplasmic fraction, anti-α-tubulin antibody (Sigma-Aldrich, Taufkirchen, Germany) was used at 1:2500 dilution in PBS-T for 1 h at room temperature whereas anti-lamin (Cell Signaling, Frankfurt am Main, Germany) at 1:1000 served as loading control for the nuclear fraction. Membranes were incubated for detection with secondary antibodies raised against rabbit labeled with CyDye800 (Licor, Bad Homburg, Germany) and mouse labeled with CyDye700 (Licor, Bad Homburg, Germany) for 1 h at room temperature. Signals were detected by Odyssee Scanner (Licor, Bad Homburg, Germany).
RNA isolation and RT-PCR
RNA was isolated using the RNeasy mini kit (Qiagen, Hilden, Germany), according to the manufacturer’s instructions. To remove possible genomic contamination, DNA digestion was performed by using the Ambion TurboDNAse purification kit (Life Technologies, Darmstadt, Germany) as described in the kit’s manual. The RNA concentration was measured with a Tecan M200 (Tecan, Crailsheim, Germany). For quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR), first-strand cDNA was synthesized from 1 μg of total RNA using the Applied Biosystems High Capacity cDNA reverse transcription kit (Life Technologies, Darmstadt, Germany). cDNA was amplified on an Eppendorf Realplex4 thermal cycler (Hamburg, Germany) using Promega GoTaq qPCR Master Mix. The sequence for the PCR primers are: p21: 5′-GGCGGCAGACCAGCATGACAGATT-3′ and 5′-GCAGGGGGCGGCCAGGGTAT-3′; p53: 5′-CTAAGCGAGCACTGCCCAAC-3′ and 5′-GGCCTCATTCAGCTCTCGGA-3′; actin: 5′-CATGCCATCCTGCGTCTGGACC-3′ and 5′-ACATGGTGGTGCCGCCAGACAG-3′, whereas PUMA expression was analyzed using the QuantiTect Primer Assay (Qiagen, Hilden, Germany). After an initial activation at 94°C for 3 min, 40 cycles of 94°C for 15 s, 55°C for 30 seconds, and 72°C for 45 s. Experiments were done in triplicates and fold changes calculated based on the ∆∆Ct method.
Statistics
Significance testing between pairs of treatments was done by unpaired two tailed Student’s
t-test using Welch’s correction if the F-test indicated a significant difference between variances. Differences were considered significant if
p < 0.05. The IC
50 was estimated from the MTT absorbance data, using the four parameter logistic model in R 2.15.1. The synergy between the effects of salinomycin and Doxorubicin was determined by the combination index (CI) based on IC
50 isobologram [
17]. A CI < 0.9 was considered as synergism between the two compounds whereas 0.9 – 1.1 indicates an additive effect.
Discussion
Malignant soft tissue tumors are composed of a heterogeneous cell population which exhibits varying degrees of chemosensitivity. A high rate of recurrence would be expected even if only a minor percentage of the cancer cells with a high resistance to systemic therapy persist in the patient [
21]. This hypothesis is also reflected by the clinical characteristics of these tumors with a marked chemoresistance, high rate of relapse and metastasis. The antibiotic salinomycin has been demonstrated to overcome drug resistance in various apoptosis-resistant human cancer cells including CSCs [
13,
22,
23]. Furthermore, it has been shown that the cell death induced by salinomycin occurs independent of p53 and caspase activation [
14,
24], pathways that are frequently disturbed in tumors. Therefore, we studied whether salinomycin could have a therapeutic use in sarcomas.
The sarcoma cell lines HT-1080, A204, and SW872 displayed a slight reduction of the cell growth in MTT assays between 0.5 μM to 10 μM, without any significant changes at concentrations higher than 1 μM, but further analysis revealed no increases in caspase 3/7 activity or the sub-G1 fraction and only a minor increase in Annexin V staining. These results indicated that at the tested concentration of 1 μM salinomycin no apoptotic cellular response occurred. In contrast, we were able to show that even at low salinomycin doses, which did not directly provoke cell death, salinomycin was able to enhance the cellular response to doxorubicin. The concurrent administration of 1 μM salinomycin in combination with doxorubicin doses ranging from 30 ng/mL to 500 ng/mL showed a synergistic effect on apoptosis. This was supported by the lowered doxorubicin IC
50 in all cell lines in the presence of salinomycin. The salinomycin concentration in the present study was 10–20 fold below the concentrations used in a previously published study [
15], and below the direct toxic dose of salinomycin for the analyzed sarcoma cell lines. These findings further suggest that salinomycin is acting synergistically to doxorubicin therapy even if used at a sublethal concentration. This proved that the combination is more effective in the treatment of sarcoma cells than the doxorubicin monotherapy on its own.
The development of multiple mechanisms of drug and apoptosis resistance is a hallmark of soft tissue sarcomas. For sarcoma cell lines the abrogation of p53-induced apoptosis by blocking NF-κB is described as a mechanism of drug resistance [
25,
26], whereas HT-1080 cells acquire chemoresistance through the activation of NF-κB to mediate cell survival [
18]. This demonstrates the dual function of NF-κB in the regulation of pro- and anti-apoptotic cellular responses. In this study we observed that the activity of NF-κB is higher in cells which were simultaneously treated with doxorubicin and salinomycin than in the doxorubicin monotherapy. Therefore we propose that in the presence of 1 μM salinomycin, the NF-κB signaling induced by doxorubicin is shifted towards a pro-apoptotic response rather than cell survival in HT-1080 cells. NF-κB has been proposed to be a transcription factor of p53 [
27,
28]. This led us to analyze the canonical p53 targets PUMA and p21. PUMA plays a pivotal role in p53 dependent and independent apoptosis. Normally expressed at low concentrations, PUMA is markedly induced following DNA damage. This is further supported by the strong increase of PUMA at the transcription level 10 h after drug administration, whereas the p21 induction is only slightly different from doxorubicin monotreatment. In addition to the p53 dependent induction, Wang
et al. reported that PUMA expression can also be induced directly by NF-κB independent of p53 signaling [
29]. The significant increase of PUMA expression by the combined treatment is further supported by the stronger activation of caspase 9, a canonical downstream effector of PUMA. Therefore our data support a model in which a sub-lethal dose of salinomycin in combination with doxorubicin enables the pro-apoptotic function of NF-κB and enhances the activation of PUMA-mediated apoptosis. Furthermore, Liu
et al. reported that PUMA is also important in p53-dependent apoptosis for the depletion of adult stem cells [
30].
Competing interests
All authors declare that they have no competing interest in regard to this manuscript.
Authors’ contributions
STL, DJT, IS, IT and AM conceived and designed the experiments. STL, DJT, CGJ, BV and MV performed the experimental work. STL, DJT, IS, BV, IT and AM participate in data analysis and interpretation. SAB, HUS and AT revised the manuscript critically for important intellectual content. All authors read and approved the final manuscript.