The growth rate was measured by the xCELLigence RTCA (real-time cell analysis) System. These non-invasive cell-based assays provide physiologically relevant data in the case of treatment with siRNA-TNF. Knockdown of TNF-α gene was associated with a cell proliferation inhibition, as we can see from the electronic readout of cell sensor impedance that is translated as the reduction of the cell index (Figure 1) when compared to untreated cells (control) or to the transfection agent (siPort NeoFX).

After 72 h of monitoring cell proliferation with the xCELLigence System, we can observe a continuous decrease in cell growth and motility after treatment with siRNA-TNF. In the last years, TNF-α was associated with the stimulation and expression of the NF-κB gene, an important modulator that activates many genes implicated in cell survival, growth and expression of pro-inflammatory cytokines, including TNF-α, which acts as a positive feedback signal to stimulate and continuously support cell growth [14]. This is in correlation with the well known fact that tumors secret TNF-α [14,16].

The flow cytometry results obtained indicate that blocking the expression of TNF-α in the triple negative breast cancer cell line leads to apoptosis, as we can see from Figure 2. So, after 24 h of treatment with siRNA-TNF-α, we obtained a high percentage of apoptotic cells.

PCR array technology was used to examine the transcript levels of 84 genes involved in apoptosis. After data analysis with the ΔΔC_t method, we obtained a relevant p-value for 15 genes after comparing the treated with untreated groups. From these, three were down-regulated and 12 up-regulated, as we can see in Table 1. Most of the genes from Table 1 are members of the extrinsic apoptosis pathway, mediated by receptors.

RTCA is a powerful biotechnology tool suited for large scale screening [18], used in evaluating the antiproliferative effect and inhibition of migratory processed by knockout of TNF-α. The role of TNF-α is controversial; some investigations have proved apoptotic or necrotic effects of TNF-α, while others furnished evidence that endogenous TNF-α activates cellular growth and tumor progression [19].

In a similar study, TNF-α induced inhibition of proliferation and enhanced the expression of p21^cip/waf1 and p27^kip1 in human glioma cells. p21 might be regulated by NF-κB or p53 independently, as confirmed by the present experiment. The inhibition of cell proliferation does not have a direct role in rendering the cells resistant to TNF-α mediated cytotoxicity [20]. Dong et al. reveals that TNF-α can promote epithelial-mesenchymal transition (EMT) of MCF-7 cells and activates cell migration, being in agreement with our findings according to which down-regulation of TNF-α inhibits cell migration.

Our study supports the hypothesis that the TNF-α cytokine sustains the growth and spread of breast cancer. The constitutive synthesis of TNF-α [20] observed in many tumor types leads to the activation of cell survival signaling pathways and the synthesis of cytokines (including TNF-α), chemokines and angiogenic factors [17]. There is evidence that indicates that TNF-α is involved in the transformation, proliferation, angiogenesis, invasion and metastasis of many cancers [21]. Tumor necrosis factor-α has an important role in the tumor microenvironment, where its secretion is increased in both stromal and tumor cells [22]. This is an example of a feed-back mechanism, where the secretion of TNF-α stimulates its own formation, promoting tumor cell growth, survival, invasion, metastasis and neoangiogenesis [23,24].

After blocking the expression of TNF-α in the triple negative breast cancer cell line, our cell viability and flow citometry data suggest that cells undergo apoptosis. Down regulation of TNF-α blocks the expression of NF-κB pathway that plays a key role in tumor cell survival. Constitutive activation of the NF-κB in tumor cells [17,25] is triggered by TNF-α. Soluble TNF-α binds to TNFR1 that is highly expressed on tumor cells and determines its trimerization that internalizes the message with the recruitment of an adaptor protein, TNF-R1-associated death domain protein (TRADD). The signaling cascade continues with the phosphorylation of TRAF2 with the help of RIP (a death domain kinase) and leads to the activation of the IKK complex [26]. This leads to the degradation of IkB and the release of the active heterodimer NF-κB, which translocates to the nucleus. Here, the transcription factor induces the transcription of the target survival genes.

Our PCR array data show that removing the ligand leads to the removal of the adaptor protein as well. Down-regulation of TRADD completely blocks the cell survival and inflammation supporting pathways, leaving room for new interactions between proteins that participate in other signaling pathways to control the development of tumor cell growth or apoptosis.

The TNF-α receptor family is mediated through several regulatory factors that are pro/anti-apoptosis regulators. By inhibiting the expression of TNF-α, the cell survival pathway controlled by TNFR1 is blocked, but the apoptosis signals internalized by the TNFSF10 and TNFRSF1A are not blocked, as observed in our data, leading to the activation of JNK kinase. When NF-κB is suppressed, JNK is activated and the TNF receptor signaling pathway shifts the fate of the cell from survival toward apoptosis [27].

The BAK-1 gene expression is also down-regulated, leading us to believe that cell apoptosis is not regulated through the cell death pathway that is mediated by mitochondria. BAK-1 increases apoptosis by accelerating the opening of the mitochondrial voltage dependent anion channel [28] and counteracts the protection from apoptosis provided by Bcl-2 [29].

There are studies indicating that TNF-α is implicated in drug resistance [30] through the activation of the NF-κB cell survival pathway leading to the inhibition of apoptosis. Other studies suggest that blocking the TNF-α expression may compromise the docetaxel chemotherapy efficiency [31]. Our results conclude that TNF-α is proving to be an efficient target in triple negative breast cancer.

Therefore, the direct targeting of TNF-α in tumor cells could be an approach to novel treatment designs. Targeting the ability of tumor cells to proliferate by blocking the intracellular cell survival pathways activated by TNF-α definitely leads tumor cells to apoptosis. At the same time, a combination with other conventional or targeted therapies might be more efficient or at least stop the tumor development and metastasis.

In our experiment, we used Hs578T, a triple negative breast cancer cell line. Cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) with a high glucose concentration, supplemented with 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO, USA), glutamine 2 mM, penicillin 100 UI/mL and insulin. Cells were grown in a humidified 5% CO₂ incubator at 37 °C. Cells were treated using the reverse transfection method that involves transfecting and plating the cells simultaneously. Half a million cells were seeded in a six-well plate and treated with siRNA-TNF from Silencer^® siRNA Transfection II Kit (Ambion, Austin, TX, USA). For each well, we used 5 μL siPORT NeoFX transfection agent dissolved in 95 mL Opti-MEM I (Gibco-Invitrogen, Paisley, UK). After 10 min of incubation at RT for each well, 2.5 μL siRNA-TNF were diluted in 97.5 μL Opti-MEM I and mixed with the transfection agent in order to achieve 50 nM in the cell culture medium. We incubated the mixture for 10 min at RT and then distributed it on the plate. Cells were cultured in a total volume of 4 mL Opti-MEM I/well for 24 h at 37 °C, 5% CO₂ before analysis.

The breast cancer cell line Hs578T was seeded in 5000 cells/well, the optimal cell density for cell proliferation assay. The cell growth curves were recorded on the xCELLigence System in real time, every 30 min, using an E-plate 16. Cells adhere to the bottom of each well, covering the surface of the sensor that monitors cells by measuring their cell index (CI). Cell migration was dynamically recorded in real time without labeling cells. The RTCA DP instrument uses the CIM-Plate 16 for cell migration assay. A cell density of 20,000 cells/well was used in order to assess the capacity of migration from the upper chamber through the porous membrane, where the cell-sensor is embedded, into the bottom chamber of each CIM-Plate 16 well, in response to fetal serum as a chemoattractant. The impedance is correlated with a numerical increase of the cells that migrated on the underside of the membrane by measuring cell index. The CI value represents the ratio between the Rn-Rb/Rb, measured at each time point. Rn represents the cell-electrode impedance of the well with the cells, and Rb is the background impedance of the well with media alone.

In this study, we used a marked-antibody staining protocol from Annexin V: FITC assay kit (AbD seroTec, Oxford, UK). For each sample, we used triplicates and evaluated apoptosis after 24 h of treatment with siRNA-TNF-α. After removing the cells from the culture plates, they were resuspended in 300 mL Binding Buffer and 1 μL Annexine V-FITC and incubated for 15 min at room temperature, protected from light. Just before flow cytometry analysis, we added 1 μL of PI. The rate of apoptosis was evaluated using FACSCantoII flow cytometer, and the data obtained was analyzed with BD FACSDiva software.

After 24 h from treatment, total ARN was extracted using TriReagent (Sigma-Aldrich, St. Louis, MO, USA), according to the manufacturer’s protocol. The extracted ARN was purified using the RNeasy Mini Kit (Qiagen, Hilden, Germany). ARN purity and concentration was measured using Agilent 2100 Bioanalyzer, and the spectrophotometer Nano Drop 1100. 350 μg of total ARN was reversely transcribed using the C03 Kit from Quiagen. For the PCR array analysis, we used 7 μg of cDNA for each well, and all experiments were performed in triplicates. In the array analysis, we evaluated 84 genes (the design of the PCR-array plate and gene description is presented in supplementary Tables S1 and S2) involved in apoptosis with the Human Apoptosis RT²Profiler™ PCR Array plate, and the PCR-array reaction was done on the LightCycler 480 instrument (Roche, Rotkreuz, Switzerland) and the PCR cycles performed according to the manufacturer instructions.

The Ingenuity System Pathway Analysis was used to interpret the data and generate a canonical network with the available interactions among the statistically significant genes implicated in apoptosis from the 84 genes evaluated with Human Apoptosis RT²Profiler™ PCR Array plate.